MATERIALS SCIENCE AND ENGINEERING

Chen • Liu
Advances in Materials Science and Engineering

Semiconductor Nanocrystals
and Metal Nanoparticles Semiconductor
Physical Properties and Device Applications Nanocrystals and Metal
Nanoparticles

Semiconductor Nanocrystals
and Metal Nanoparticles
Semiconductor nanocrystals and metal nanoparticles are the
building blocks of the next generation of electronic, optoelectronic,
and photonic devices. Covering this rapidly developing and
interdisciplinary field, this book examines in detail the physical
Physical Properties
properties and device applications of semiconductor nanocrystals
and metal nanoparticles. It begins with a review of the synthesis and
and Device Applications
characterization of various semiconductor nanocrystals and metal
nanoparticles and goes on to discuss in detail their optical, light
emission, and electrical properties. It then illustrates some exciting
applications of nanoelectronic devices (memristors and single-
electron devices) and optoelectronic devices (UV detectors, quantum
dot lasers, and solar cells), as well as other applications (gas sensors
and metallic nanopastes for power electronics packaging).

• Focuses on a new class of materials that exhibit fascinating

physical properties and have many exciting device applications.
• Presents an overview of synthesis strategies and
characterization techniques for various semiconductor
nanocrystal and metal nanoparticles.
• Examines in detail the optical/optoelectronic properties, light
emission properties, and electrical properties of semiconductor
nanocrystals and metal nanoparticles.
• Reviews applications in nanoelectronic devices, optoelectronic
devices, and photonic devices.

Edited by

Tupei Chen • Yang Liu

K13633
ISBN: 978-1-4398-7830-9
90000

9 781439 878309
w w w. c rc p r e s s . c o m

9781439878309_cover.indd 1 6/28/16 2:48 PM

 Semiconductor
Nanocrystals and
Metal Nanoparticles
Physical Properties
and Device Applications
 Advances in Materials Science and Engineering

Series Editor
Sam Zhang

Aerospace Materials Handbook, edited by Sam Zhang and Dongliang Zhao

Biological and Biomedical Coatings Handbook: Applications, edited by Sam Zhang
Biological and Biomedical Coatings Handbook: Processing and Characterization,
edited by Sam Zhang
Hierarchical Micro/Nanostructured Materials: Fabrication, Properties, and
Applications, Weiping Cai, Guotao Duan, Yue Li
Hydroxyapatite Coatings for Biomedical Applications, edited by Sam Zhang
Nanobiomaterials: Development and Applications, edited by Dong Kee Yi and
Georgia C. Papaefthymiou
Nanostructured and Advanced Materials for Fuel Cells, edited by San Ping Jiang
and Pei Kang Shen
Micro- and Macromechanical Properties of Materials, Yichun Zhou, Li Yang,
and Yongli Huang
Semiconductor Nanocrystals and Metal Nanoparticles: Physical Properties and
Device Applications, edited by Tupei Chen and Yang Liu
Thin Films and Coatings: Toughening and Toughness Characterization,
edited by Sam Zhang
 Semiconductor
Nanocrystals and
Metal Nanoparticles
Physical Properties
and Device Applications

Edited by
Tupei Chen • Yang Liu
MATLAB® is a trademark of The MathWorks, Inc. and is used with permission. The MathWorks does not warrant the accuracy
of the text or exercises in this book. This book’s use or discussion of MATLAB® software or related products does not consti-
tute endorsement or sponsorship by The MathWorks of a particular pedagogical approach or particular use of the MATLAB®
software.

CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742

© 2017 by Taylor & Francis Group, LLC

CRC Press is an imprint of Taylor & Francis Group, an Informa business

No claim to original U.S. Government works

Printed on acid-free paper

Version Date: 20160628

International Standard Book Number-13: 978-1-4398-7830-9 (Hardback)

This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to
publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials
or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material repro-
duced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any
copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint.

Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any
form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming,
and recording, or in any information storage or retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copy-
right.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400.
CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been
granted a photocopy license by the CCC, a separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identifica-
tion and explanation without intent to infringe.

Library of Congress Cataloging‑in‑Publication Data

Names: Chen, Tupei, editor. | Liu, Yang, 1975 September 12- editor.
Title: Semiconductor nanocrystals and metal nanoparticles : physical
properties and device applications / editors, Tupei Chen and Yang Liu.
Description: Boca Raton : Taylor & Francis, a CRC title, part of the Taylor &
Francis imprint, a member of the Taylor & Francis Group, the academic
division of T&F Informa, plc, [ 2017] | Series: Advances in materials
science and engineering | Includes bibliographical references and index.
Identifiers: LCCN 2016009756 | ISBN 9781439878309 (alk. paper)
Subjects: LCSH: Nanoelectronics--Materials. | Semiconductor nanocrystals. |
Semiconductor nanoparticles.
Classification: LCC TK7874.84 .S46 2017 | DDC 621.3815/2--dc23
LC record available at https://lccn.loc.gov/2016009756

Visit the Taylor & Francis Web site at

http://www.taylorandfrancis.com

and the CRC Press Web site at

http://www.crcpress.com
Contents
Preface..............................................................................................................................................vii
Editors................................................................................................................................................ix
Contributors.......................................................................................................................................xi

Chapter 1 Synthesis and Characterization of Nanocrystals and Nanoparticles............................1

Yang Liu

Chapter 2 Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices......... 39

S.K. Ray, N. Gogurla, and T. Rakshit

Chapter 3 From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art

Lasers for 1.55 µm Optical Communications: An Overview...................................... 95
Jacky Even, Cheng Wang, and Frédéric Grillot

Chapter 4 Optical, Photoluminescence, and Vibrational Spectroscopy of Metal

Nanoparticles............................................................................................................. 127
P. Gangopadhyay

Chapter 5 Silicon Nanocrystals: Properties and Potential Applications................................... 191

Spiros Gardelis

Chapter 6 Electronic and Optical Properties of Si and Ge Nanocrystals.................................. 215

Tupei Chen

Chapter 7 Light Emission Properties of Si Nanocrystals Embedded in a Dielectric Matrix.... 255

Tupei Chen

Chapter 8 Optical Properties of Semiconductor Nanoparticles in

Photoelectrochemical Cells....................................................................................... 283
M.H. Buraidah, S.N.F. Yusuf, I.M. Noor, and A.K. Arof

Chapter 9 Second-Order Nonlinear Susceptibility in Quantum Dot Structures.......................307

M. Abdullah, Farah T. Mohammed Noori, and Amin H. Al-Khursan

Chapter 10 Metallic Nanopastes for Power Electronic Packaging.............................................. 343

Denzel Bridges, Ruozhou Li, Zhiming Gao, Zhiqiang Wang, Zhiyong Wang,
Anming Hu, Zhili Feng, Leon M. Tolbert, and Ning-Cheng Lee

v
vi Contents

Chapter 11 Applications of Metal Nanoparticles and Nanostructures Fabricated Using

Ultrafast Laser Ablation in Liquids........................................................................... 367
S. Venugopal Rao, S. Hamad, and G. Krishna Podagatlapalli

Chapter 12 Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas: From

Fundamental to Technical Applications.................................................................... 423
Ngai Yui Chan, Fan Zhang, Kit Au, Wing Chong Lo, Helen La Wa Chan,
and Jiyan Dai

Chapter 13 Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles...... 461

Z. Liu

Index............................................................................................................................................... 501
Preface
A material particle or cluster having a size in the range of one to several hundreds of nanometers is
often referred to as “nanocrystal” or “nanoparticle.” Nanocrystal is considered as crystalline clus-
ters in either a single crystalline or a polycrystalline arrangement, while nanoparticle can represent
both crystalline and noncrystalline clusters. Semiconductor nanocrystals with small dimensions
are also described as quantum dots. Semiconductor nanocrystals and metal nanoparticles exhibit
fascinating behaviors and unique physical properties as a result of quantum size effect. For example,
for a semiconductor nanocrystal whose size is smaller than twice the size of its exciton Bohr radius,
the excitons are squeezed, leading to quantum confinement. Quantum confinement will modify the
electronic and optical properties of the semiconductor nanocrystal, for example, the bandgap of the
nanocrystal increases with the decrease of the nanocrystal size. A direct experimental observation
of the quantum confinement effect is that different sized quantum dots emit different color light.
A good example of the fascinating behaviors of metal nanoparticles is the localized surface plasmon
resonance (LSPR) in noble metal (for example, silver and gold) nanoparticles where the conduction
electrons oscillate coherently under irradiation by light in the visible and near-infrared regions of
the electromagnetic spectrum. The wavelength of excitation of the LSPR shows strong dependence
on the size, shape, and dielectric environment of the metal nanoparticles.
Semiconductor nanocrystals and metal nanoparticles are the building blocks of the next gen-
eration of electronic, optoelectronic, and photonic devices. For instance, semiconductor quantum
dots are particularly significant for optoelectronic/photonic device applications such as photovoltaic
devices, light-emitting devices, and photodetector devices due to their tunable absorption spectrum
and high extinction coefficient. In electronic applications, semiconductor nanocrystals or metal
nanoparticles with sizes smaller than ~5 nm can be used to realize single-electron or few-electron
devices (e.g., transistors and memory devices) as such tiny nanostructures exhibit the Coulomb
blockade effect at room temperature.
This book examines in detail the physical properties and device applications of semiconductor
nanocrystals and metal nanoparticles. It begins by giving a review on the synthesis and characteriza-
tion of various semiconductor nanocrystals and metal nanoparticles and goes on to discuss in detail
their electronic, optical, and electrical properties. Based on the knowledge of the physical proper-
ties, the book illustrates some exciting applications in nanoelectronic devices (e.g., memristors,
single-electron devices), optoelectronic devices (e.g., UV detectors, quantum dot lasers, solar cells),
and other applications (e.g., gas sensors, metallic nanopastes for power electronics packaging).
This book contains 13 chapters. It covers a rapidly developing, interdisciplinary field, and it is
also a compilation of some research efforts in the field. It is intended as a reference book for senior
undergraduate and graduate students and researchers in microelectronics/nanoelectronics, opto-
electronics, photonics, nanoscience and nanotechnology, physics, and materials science.

Tupei Chen
Nanyang Technological University, Singapore
Yang Liu
University of Electronic Science and Technology of China, China

vii
viii Preface

MATLAB® is a registered trademark of The MathWorks, Inc. For product information, please con-
tact the following:

The MathWorks, Inc.

3 Apple Hill Drive
Natick, MA 01760-2098 USA
Tel: 508-647-7000
Fax: 508-647-7001
E-mail: [email protected]
Web: www.mathworks.com
Editors
Tupei Chen received his PhD from the University of Hong Kong,
Hong Kong, in 1994. He is currently an associate professor in the
School of Electrical and Electronic Engineering, Nanyang
Technological University, Singapore. His research areas include
electronic and optoelectronic/photonic applications of nanoscale
materials, novel memory devices, memristors and applications in
artificial neural networks, Si photonics, and metal oxide thin films
and applications in flexible/transparent devices. He is the author or
coauthor of more than 250 peer-reviewed journal papers, more
than 120 conference presentations, and 6 book chapters. He has
filed several U.S. patents. He has supervised more than 20 PhD
students and over 10 postdoctoral research staff.

Yang Liu received his BSc in microelectronics from Jilin

University, China, in 1998 and his PhD from Nanyang
Technological University, Singapore, in 2005. From May 2005 to
July 2006, he was a research fellow at Nanyang Technological
University. In 2006, he was awarded the prestigious Singapore
Millennium Foundation Fellowship. In 2008, he joined the School
of Microelectronics, University of Electronic Science and
Technology, China, as a full professor. He is the author or coauthor
of over 120 peer-reviewed journal papers and more than 80 confer-
ence papers. He has filed one U.S. patent and more than 30 Chinese
patents. His current research includes advanced memory devices
and their applications in artificial neural networks, radio frequency
integrated circuits (RFICs), photonic/optoelectronic devices, and
integrated circuits (ICs).

ix
Contributors
M. Abdullah Tupei Chen
Nassiriya Nanotechnology Research School of Electrical and Electronic
Laboratory Engineering
Thi-Qar University Nanyang Technological University
Nassiriyah, Iraq Singapore, Singapore

Jiyan Dai
Amin H. Al-Khursan Department of Applied Physics
Nassiriya Nanotechnology Research The Hong Kong Polytechnic University
Laboratory Kowloon, Hong Kong, People’s Republic of
Thi-Qar University China
Nassiriyah, Iraq
Jacky Even
FOTON Laboratory
A.K. Arof National Institute for Applied Sciences
Centre for Ionics Rennes, France
Department of Physics
Faculty of Science Zhili Feng
University of Malaya Materials Science and Technology Division
Kuala Lumpur, Malaysia Oak Ridge National Laboratory
Oak Ridge, Tennessee
Kit Au P. Gangopadhyay
Department of Applied Physics Materials Science Group
The Hong Kong Polytechnic University Indira Gandhi Centre for Atomic Research
Kowloon, Hong Kong, People’s Republic of Kalpakkam, Tamil Nadu, India
China
Zhiming Gao
Denzel Bridges National Transportation Research Center
Department of Mechanical, Aerospace, Oak Ridge National Laboratory
and Biomedical Engineering Knoxville, Tennessee
The University of Tennessee, Knoxville
Spiros Gardelis
Knoxville, Tennessee
Department of Solid State Physics
Faculty of Physics
M.H. Buraidah National and Kapodistrian University of Athens
Centre for Ionics Athens, Greece
Department of Physics
Faculty of Science N. Gogurla
University of Malaya Department of Physics
Kuala Lumpur, Malaysia Indian Institute of Technology
Kharagpur, West Bengal, India

Ngai Yui Chan Frédéric Grillot

Department of Applied Physics Laboratory for Communication and Processing
The Hong Kong Polytechnic University of Information
Kowloon, Hong Kong, People’s Republic of Telecom Paris Tech—Université Paris-Saclay
China Paris, France

xi
xii Contributors

S. Hamad Wing Chong Lo

School of Physics Department of Applied Physics
University of Hyderabad The Hong Kong Polytechnic University
Hyderabad, Telangana, India Kowloon, Hong Kong, People’s Republic of
China
Anming Hu
Department of Mechanical, Aerospace, and Farah T. Mohammed Noori
Biomedical Engineering College of Science
The University of Tennessee, Knoxville Baghdad University
Knoxville, Tennessee Baghdad, Iraq
Ning-Cheng Lee
Indium Corporation I.M. Noor
Clinton, New York Centre for Ionics
Department of Physics
Helen La Wa Chan Faculty of Science
Department of Applied Physics University of Malaya
The Hong Kong Polytechnic University Kuala Lumpur, Malaysia
Kowloon, Hong Kong, People’s Republic of
China G. Krishna Podagatlapalli
Advanced Centre of Research in High Energy
Ruozhou Li Materials
School of Electronic Science and University of Hyderabad
Engineering Hyderabad, Telangana, India
Southeast University
Nanjing, Jiangsu, People’s Republic of China T. Rakshit
and Advanced Technology Development Centre
Indian Institute of Technology
Suzhou Key Laboratory of Metal Nano- Kharagpur, West Bengal, India
Optoelectronic Technology
Suzhou Research Institute of Southeast
University S.K. Ray
Suzhou, Jiangsu, People’s Republic of China Department of Physics
Indian Institute of Technology
Yang Liu Kharagpur, West Bengal, India
State Key Laboratory of Electronic Thin Films
and Integrated Devices Leon M. Tolbert
University of Electronic Science and Department of Electrical Engineering and
Technology of China Computer Science
Chengdu, Sichuan, People’s Republic of The University of Tennessee, Knoxville
China Knoxville, Tennessee

Z. Liu S. Venugopal Rao

School of Materials and Energy Advanced Centre of Research in High Energy
Guangdong University of Technology Materials
Guangzhou, Guangdong, People’s Republic of University of Hyderabad
China Hyderabad, Telangana, India
Contributors xiii

Cheng Wang Zhiyong Wang

School of Information Science and Technology Institute of Laser Engineering
ShanghaiTech University Beijing Institute of Technology
Pudong, Shanghai, People’s Republic of China Chaoyang, Beijing, People’s Republic of China
and S.N.F. Yusuf
FOTON Laboratory Centre for Ionics
National Institute for Applied Sciences Department of Physics
Rennes, France Faculty of Science
University of Malaya
Kuala Lumpur, Malaysia

Zhiqiang Wang Fan Zhang

Department of Electrical Engineering and Department of Applied Physics
Computer Science The Hong Kong Polytechnic University
The University of Tennessee, Knoxville Kowloon, Hong Kong, People’s Republic of
Knoxville, Tennessee China
 1 Synthesis and Characterization
of Nanocrystals and
Nanoparticles
Yang Liu

CONTENTS
1.1 Introduction...............................................................................................................................1
1.2 Semiconductor Nanocrystals..................................................................................................... 2
1.2.1 Compound Semiconductor Nanocrystals......................................................................2
1.2.2 Elemental Semiconductor Nanocrystals...................................................................... 14
1.3 Metal Nanoparticles................................................................................................................. 18
1.4 Nanocrystals and Nanoparticles Embedded in a Dielectric Film........................................... 23
References......................................................................................................................................... 31

1.1 INTRODUCTION
Nanocrystals and nanoparticles are described as clusters having diameters in the range of one to
several hundreds of nanometers. A “nanocrystal” is considered as a cluster of crystals, whereas a
“nanoparticle” can consist of both crystalline and noncrystalline clusters. The synthesis of nano-
crystals/nanoparticles has attracted extensive studies in the past several decades. These materials
show size-dependent physical properties that are not shown by their bulk counterparts. For example,
a Si nanocrystal shows photoluminescence (PL) and electroluminescence (EL), which cannot be
achieved in bulk silicon due to its indirect bandgap. Therefore, precise control during the synthesis
of nanocrystals/nanoparticles is of key importance for their applications.
Many techniques are used to synthesize nanocrystals/nanoparticles. Chemical vapor deposition
(CVD), physical vapor deposition, ion implantation, sputtering, laser ablation, and spray pyroly-
sis are all gas-based techniques and have been used successfully for the synthesis of nanocrys-
tals/nanoparticles. They provide the possibility of controlling the nanocrystal/nanoparticle size
by adjusting the synthesis parameters such as temperature, pressure, power density, gas flow rate,
and so on. Solution-based techniques are of intensive research interest, as they are highly effec-
tive in synthesizing nanocrystals/nanoparticles with good size control. Besides, they also have the
advantages of low synthesis temperature, flexibility, and low cost. A large variety of methods have
been developed to synthesize nanocrystals/nanoparticles, such as reduction, thermal decomposi-
tion, and processes with hydrolysis and alcoholysis. At the same time, various techniques have been
utilized to characterize the structural and physical properties of nanocrystals/nanoparticles, such
as transmission electron microscopy (TEM), scanning electron microscopy (SEM), atomic force
microscopy (AFM), X-ray photoemission spectroscopy (XPS), X-ray diffraction (XRD), Raman
scattering, and so on. In this chapter, we review the methods for synthesizing nanocrystals/nanopar-
ticles. We also discuss the characterization of nanocrystals/nanoparticles using various techniques.

1
2 Semiconductor Nanocrystals and Metal Nanoparticles

1.2 SEMICONDUCTOR NANOCRYSTALS

1.2.1 Compound Semiconductor Nanocrystals
Compound semiconductor nanocrystals such as CdSe, CdS, GaAs, InAs, and CuCl have been
widely investigated. In this section, we discuss the synthesis of compound semiconductor nanocrys-
tals as well as their characterization using various techniques.
Synthesis of II–VI (A2B6) nanocrystals/nanoparticles such as CdS and CdSe is attractive because
these materials have potential applications in photosensitive, photovoltaic, optical, and electronic
devices. The shape, concentration, and size of the nanocrystals, as well as their size distribution,
strongly depends on the synthesis process. CdSe nanoparticles showing quantum size effects was
reported in 1993 [1]. Research on nanocrystalline cadmium selenide (CdSe) has been going on for
years due to its excellent properties such as light emission, low bandgap (Eg = 1.75 eV), low toxicity,
facile surface modification, and thermostability, with applications in light-emitting diodes, solar
cells, hydrogen-producing catalysts, and biological imaging [2–15].
Chemical methods are simple, flexible, effective, and low cost to synthesize isolated CdSe nano-
crystals, and include solvothermal methods and hydrothermal methods. The solvothermal methods
always utilize TOP (tri-n-octylphosphine)/TOPO (tri-n-octylphosphine oxide), TOP/TOPO/HPA
(hexylphosphonic), TOP/TOPO/HDA (hexadecylam), or other reagents to synthesize CdSe nano-
crystals [16–23].
The TOP/TOPO method was reported in [16,20] and is illustrated in Figure 1.1. In [16], the
synthesis of ZnS-capped CdSe nanocrystals was carried out by using TOP/TOPO with a size of
2.7–3.0 nm. In [20], CdSe nanocrystals were synthesized with a diameter of ~3 nm with TOP/
TOPO. The nanocrystals were dried under N2 flow and dissolved in toluene. Finally, the nanocrys-
tal solution was filtered using a 0.45 mm syringe filter to remove other contaminations from the
synthesis process.
The TOP/TOPO/HPA method was used in [17], where Cd(CH3)2 in heptane and Se/TBP (tri-n-
butyphosphine) were mixed with TBP and the solution was injected into HPA (hexylphosphonic
acid) in TOPO. The size of SeCd nanocrystals could be controlled in the range 5–20 nm by vary-
ing the injection volume. The TOP/TOPO/HDA method was reported in [19,21]. In [19], Cd(CH3)2
and Se/TOP were injected into TOPO and HDA. The size of prepared CdSe nanocrystals was in
the range 4.5–5.0 nm. In [21], CdO/SA (stock solution of stearic acid with CdO) was injected into

Me2Cd + Heptane

TOP Se powder

TOPO

Filter

CdSe NC

FIGURE 1.1 Process of the TOP/TOPO method.

Synthesis and Characterization of Nanocrystals and Nanoparticles 3

TOPO and HDA, and the mixed solution was rapidly injected into Se/TOP. Finally, the mixture was
purified by centrifugation. The size of CdSe nanocrystals was distributed in the range 2.6–5.9 nm.
In [18], the different solvent reactions for synthesizing CdSe nanocrystals were discussed. The
Cd precursor was selected from Cd(Ac)2, CdCO3, or CdO, and the solvent from TOPO, SA/TOPO,
SA, Tech TOPO, HPA/TOPO, or DA (dodecylamine)/TOPO. A Se solution (Se powder dissolved in
TOP or TBP) was injected into the reaction vessel with the mixture of the above chemicals. CdSe
nanocrystals of 4–25 nm were finally obtained. In [22], Se powder and cadmium myristate were
added to a flask containing ODE (octadecene). Under a flow of Ar, the solution was stirred and
heated to 240°C at the rate of 25 K min−1. After the particle diameter reached ~3.0 nm, a solution of
oleic acid in ODE was added to stabilize the growth of the nanocrystals.
There are several reports on hydrothermal methods for the synthesis of CdSe nanocrystals [9–13].
In [24], a solution of Cd(ClO4)2 · 6H2O and the stabilizer in deionized (DI) water was adjusted to a
pH of 11.2 using NaOH. The solution was mixed with the oxygen-free NaHSe solution. The size
of CdSe nanocrystals obtained by this method ranged from 1.5 to 2.8 nm. In [27], a solution of
Cd(NO3)2 was dissolved in distilled water containing trisodium citrate as the stabilizer. H2Se gas
was slowly passed through the above solution together with N2 as buffer gas. The mean size of the
nanocrystals was ~45 nm in the CdSe sol. In [26], an aqueous solutions was used for the synthesis
of CdSe nanocrystals by reacting a solution CdCl2 and Na2SeSO3 in gelatine solution in distilled
water based on the following reaction:

Cd 2+ + SeSO32+ + 2OH ® CdSe + SO 4 2+ + H 2O

The size of the crystals could be controlled in the range 2.5–3.0 nm.
The synthesis of the MPA-capped CdSe nanocrystals was reported in [28]. NaHSeO3 and NaBH4
were injected into mixed solution of CdCl2 · 2.5H2O, 3-MPA, and NaOH. The MPA-capped CdSe
nanocrystals with an average size of 4–5 nm were finally obtained.
TEM, AFM, XRD, and XPS are often used to characterize nanocrystals/nanoparticles.
TEM is widely used to characterize the size and nanostructure of CdSe nanocrystals/nanoparticles
[16–19,21,22,24,25,28]. In [17], the TEM images of rod-, arrow-, teardrop-, and tetrapod-shaped
CdSe Nanocrystals were presented. In [19], the TEM images of CdSe nanocrystals and CdSe/ZnS
core–shell nanoparticles covered with monolayers of ZnS were shown. In [24], a TEM image of
thioglycerol-stabilized CdSe nanocrystals was presented together with a single CdSe nanocrystal,
and [25] provided TEM images of CdSe particles in different shapes. In [28], the TEM image
showed the existence of MPA-CdSe nanocrystals obtained after 7 h of refluxing. In [18], by select-
ing an appropriate solvent system, CdSe nanocrystals in a very broad size range could be obtained
from ~2 to >25 nm, as shown in Figure 1.2.
XRD is a very useful technique to obtain the material composition, internal information such as
atomic or molecular structure, or form by the analysis of the X-ray diffraction pattern. The diameter
can be determined by calculating the FWHM (full-width at half-maximum) of the XRD peaks with
the Debye–Scherrer formula:

kl
Dhkl = (1.1)
b cos q

where
Dhkl is the crystallite diameter along the crystal direction perpendicular to the (hkl) plane
k is the Scherrer constant, which is usually 0.89
λ is the wavelength of X-rays
β is the Bragg diffraction angle
θ is the FWHM of the prominent XRD peak
4 Semiconductor Nanocrystals and Metal Nanoparticles

50 nm 50 nm

50 nm 50 nm

50 nm 50 nm

FIGURE 1.2 TEM images of wurtzite CdSe nanocrystals of different sizes synthesized in different solvent
systems. (From Qu, L.H. et al., Nano Lett., 1, 333, 2001.)

XRD has been widely used to characterize CdSe nanocrystals/nanoparticles size and crystal ori-
entation [18,22,24,25,27,28]. In [18], CdSe nanocrystal of ~6 nm was demonstrated with high
crystalline characteristics. In [22], the XRD patterns of zinc blende (ZB) CdSe nanocrystals with
a diameter of 4.0 nm were provided. In [24], the authors presented the diffraction patterns of CdSe
crystals of wurzite hexagonal and zinc blende cubic structures capped with different stabilizers.
In [25], the relationship between CdSe nanocrystal structure and different reacting agents was
demonstrated at different reaction temperatures through their XRD patterns. In [27], the XRD pat-
tern of CdSe nanocrystals (zinc blende) grown by a reaction in an aqueous medium was shown. In
[28], XRD patterns of CdSe nanocrystals after different refluxing times were presented. In [22], a
typical XRD pattern of ZB CdSe nanocrystals with a diameter of 4.0 nm was illustrated, as shown
in Figure 1.3, where the inset shows TEM image of identical size.
Many methods for preparing CdS nanocrystals have been reported, mainly including aqueous-phase
synthesis [29–37], colloid technique [38–40], microemulsion by means of ultrasonic irradiation
[41,42], ion implantation method [43], and precipitation [44–46]. Aqueous-phase synthesis was
described in [29–37]. In [29], CdS nanocrystals were fabricated by in situ sulfuration of Cd2+ by
using some organic and inorganic molecules containing sulfur with average size from 4 nm to
several tens of nanometers. In [31], CdS nanocrystals of 3.2–6.0 nm were synthesized based on
a mixture of CdM2, oleic acid, toluene, and an aqueous solution of thiourea for different Cd/S
molar ratios by the seeding-growth technique at temperatures of 180°C, 150°C, or 120°C. In [32],
CdS nanocrystals of 20 nm were synthesized with a solution of thioacetamide (CH3CSNH2) and
CdCl2 mixed with sodium citrate in different concentrations. In [33], a CdCl2 solution and a GSH
Synthesis and Characterization of Nanocrystals and Nanoparticles 5

111

220
311

422
400
331

511

440

620
531
20 40 60 80 100 120
2θ (°)

FIGURE 1.3 X-ray diffraction pattern of zinc blende CdSe nanocrystals with a diameter of 4.0 nm.
The diffraction peak positions and relative intensities of bulk ZB CdSe are indicated. The inset
shows a high-resolution TEM image of a CdSe nanocrystal with a size of ~4 nm. (From Yang, Y.A.
et al., Angew. Chem., 117, 6870, 2005.)

(l-glutathione) solution with an appropriate pH value were mixed with thiourea. CdS nanocrystals
with an average diameter of 3.0 ± 0.3 nm were obtained by adding acetone followed by centrifug-
ing and drying the solution. In [34], CdS nanoparticles with diameters below 30 nm were prepared
by gradually adding Cd(NO3)2 · 4H2O solution into the copolymer DADMAC/NVP (N,N-dimethyl
N,N-diallyl ammonium chloride)/(N-vinyl pyrrolidinone) solution containing Na2S · H2O. In [35],
CdS nanocrystals with an average size of 3.1 nm were synthesized by using thioglycolic acid (TGA)
as stabilizer in an aqueous solution containing CdAc2 · 2H2O, Na2S, and NaOH. Finally, ethanol was
added and the solution was dried in vacuum. In [36], CdS nanocrystals of ~90 nm were synthesized
from a mixture of Cd(Ac)2 and Na2S solutions.
A colloid synthesis technique was reported in [38–40]. In [38], CdS nanocrystals of 1–2 nm
were prepared by a mixture of cadmium thioglicerate and sodium sulfide in dimethylformamide
(DMF). In [39], CdS nanocrystals of 3–5 nm were prepared with the block copolymer PEG-b-poly
(2-N,N-dimethylaminoerthyl methacrylate) (PAMA), Na2S, and CdCl2 solution. In [40], cadmium
acetate hydrate, myristic acid, and octadecene (ODE) solutions with sulfur and the nucleation initia-
tors (tetraethylthiuram disulfides [I1] and 2,2′-dithiobisbenzothiazole [I2]) were added into ODE. CdS
nanocrystals of ~3.8 nm were obtained by heating the resulting solution at 240°C in an argon ambient.
In [42], a mixture of ethylendiamine and CS2 solutions in oil phase was irradiated with ultra-
sound and was mixed with a solution of cadmium chloride under sonication. The resulting solution
was centrifuged, washed with distilled water and with ethanol, and finally dried in air at room
temperature. The CdS nanocrystal synthesized with this method was ~4 nm in size. In [45], CdS
nanocrystals were prepared via a co-precipitation method with Cd(NO3)2 · 4H2O, cetyltrimethyl
ammonium bromide (CTAB), and Na2S solution. The size of nanocrystal could be varied, by con-
trolling the heating temperature, from several nanometers to several hundred nanometers. In [46],
CdS nanocrystals of 3–4 nm were synthesized by a wet chemical precipitation method by using
cadmium sulfate (CdSO4), sodium sulfide (Na2S), and thio-glycerol.
TEM has been was extensively used to characterize CdS nanocrystals/nanoparticles [30–34,37,​
40–42,45,46]. In [30], the TEM images of CdS:Mn nanocrystals without a ZnS shell and with
the shell were presented. In [31], a TEM image of CdS nanocrystals synthesized by the seeding-
growth technique was shown, while [32] showed images of CdS nanocrystals synthesized with
different concentrations of sodium citrate. In [33], the TEM image of the as-prepared CdS mono-
disperse nanocrystals with an average diameter of 3.0 ± 0.3 nm was presented. In [34], the TEM
images showed that CdS nanocrystals were in the range 50–70 nm obtained with different stabi-
lizing copolymers. In [41], TEM images of the CdS nanocrystals synthesized with the ultrasound
technique were presented. Figure 1.4 shows the TEM images of CdS nanocrystals of <10 nm.
6 Semiconductor Nanocrystals and Metal Nanoparticles

10 nm

(a)

(002)

(002)

(112)
(110)

(b) (c)

FIGURE 1.4 (a) TEM image of CdS nanocrystals, (b) HRTEM image, and (c) SAED of the CdS nano-
crystals. (From Ghows, N. and Entezari, M.H., Ultrasonics Sonochem., 18, 269, 2011.)

Figure 1.4b presents the TEM image of a single CdS nanocrystal in the direction (002) with a
­lattice spacing of 0.34 nm. The crystal structure and crystallinity of the nanocrystal are also shown
in the selected area diffraction (SAED) pattern in Figure 1.4c.
XRD patterns of CdS nanocrystals were studied in [30–35,38,40–42,45,46]. In [30], the XRD
patterns of CdS:Mn/ZnS core/shell nanocrystals synthesized by varying the water-to-surfactant
molar ratios were presented, while in [32] the XRD patterns of CdS nanocrystals synthesized with
2.0 mM sodium citrate were shown. In [33], the XRD patterns exhibited a cubic zinc blende crystal
phase of crystalline CdS. In [34], the XRD pattern of CdS nanocrystals stabilized with DADMAC-
NVP (1/1) copolymer was shown. In [35], the XRD patterns of CdS nanocrystals showed the planes
of cubic zinc blende structure with an average size of 3.1 nm. In [40], the authors presented the XRD
pattern of CdS nanocrystals of the zinc blende phase. In [42], the XRD results of CdS nanocrystals
Synthesis and Characterization of Nanocrystals and Nanoparticles 7

synthesized from different precursors, different mole ratios of CdS, and different sonication tem-
peratures and durations were shown. In [45], the authors provided the XRD patterns of samples
for different synthesis durations and heating stages during thermal treatment. Figure 1.5 shows the
wide-angle X-ray diffraction (WAXD) patterns of the CdS nanocrystals prepared at different tem-
peratures and having the cubic zinc blende structure [31].
III–V compound semiconductor nanocrystals have attracted much attention because they exhibit
strong, size-dependent optical and electrical properties, which provide new possible applications includ-
ing large-gain low-threshold quantum lasers, light-emitting diodes, single-electron transistors, and so
on. GaAs is one of the most important semiconductor materials with applications in integrated circuits,
infrared detectors [47], quantum-well lasers [48], solar cells [49], and some other devices [50–53].
Electrochemical techniques were extensively used to synthesize GaAs nanocrystals with different
sizes and densities [54–59]. Metallic Ga and As2O3 of high purity were separately dissolved in con-
centrated HCl and diluted with DI water. The solution of Ga was mixed with the As2O3 solution. The
pH value was adjusted simultaneously by adding distilled water or acid [54–59]. A platinum anode
and an indium–tin–oxide (ITO)-coated glass cathode were dipped in the solution and the whole sys-
tem was kept in an ice bath at a constant current until it reached equilibrium. The GaAs nanocrystals
were prepared at different current densities [54,56,58] or at different temperatures [55,59]. The sizes
of GaAs nanocrystals were reported from 1.5 nm to several tens of nanometers [54–59].
Nanocrystalline GaAs was prepared by laser ablation [60–62], which is a promising way to
synthesize nanocrystals with small size and high purity. In [60], GaAs nanocrystals were syn-
thesized by laser-induced etching using a Nd:YAG (λ = 1.06 μm) laser. Figure 1.6 shows a sche-
matic illustration of laser-induced etching. The sample is irradiated with a laser power density
of 20 W cm−2 for a duration of 60 min to etch the GaAs wafer. After etching, GaAs nanocrys-
tals of 3.5–5.4 nm were obtained after rinsing with ethanol and drying in nitrogen ambient. In
[62], the deposition of GaAs onto Si was carried out using pulsed-laser deposition (PLD) with
a Q-switched Nd:YAG laser. The size of GaAs nanocrystals could be controlled by the number
of pulses of the laser from several to tens of nanometers. In [63,64], molecular beam epitaxy
(111)

(220)

(311)
(200)
Intensity (a.u.)

20 30 40 50 60
2θ (°)

FIGURE 1.5 Wide-angle X-ray diffraction patterns of the CdS nanocrystals synthesized at (a) 180°C,
(b) 120°C, and (c) 150°C, and (d) by the seeding-growth technique at 180°C. (From Wang, Q. et al., Chem.
Eur. J., 11, 3843, 2005.)
8 Semiconductor Nanocrystals and Metal Nanoparticles

Nd:Yag laser
λ = 1.06 μm

Mirror

Lens

Container
GaAs wafer

40% HF
solution Teflon plates

FIGURE 1.6 Schematic illustration of the setup for laser etching. (From Mavi, H.S. et al., Mater. Sci. Eng. B,
107, 148, 2004.)

(MBE) was used to prepare GaAs nanocrystals. GaAs nanocrystals were deposited on the Si
substrate at the pressure of ~9 × 10 −3 Pa for a size of ~20 nm [64].
A wet process was used to synthesize GaAs nanocrystals in [65]. Synthesis of GaAs nanocrystals
was carried out in a vacuum chamber under nitrogen atmosphere. Triglyme containing 0.5 mmol of
Ga(acac)3 and 0.5 mmol of As(SiMe3)3 was heated at 216°C for 70 h, and then was filtered through a
0.2 mm PTFE filter. GaAs nanocrystals of size 1.5–9 nm were obtained finally.
TEM was used to study the microstructure of GaAs nanocrystal size, surface morphology, and
lattice structure [55–59,61,65]. In [55–59], the researchers provided a series of TEM images and
discussed the relationship between the nanocrystals and the different synthesis conditions by the
electrochemical technique. In [61], the TEM characterization conducted after the sedimentation of
larger particles showed the existence of GaAs nanocrystals, prepared by laser ablation, with size
distribution ranging from 3 to 10 nm. In [65], TEM images of GaAs nanocrystals ranging from
1.5 to 9.0 nm were shown. A TEM image of 10 nm GaAs nanocrystals is shown in Figure 1.7 taken
from [55]. The d-spacing measured from the lattice image was found to be 0.1673 nm close to the
d(211) = 0.1686 nm of the orthorhombic GaAs phase.
In [54], the XRD patterns of GaAs nanocrystals prepared at different electrolysis current densi-
ties were presented. In [59], the authors discussed nanocrystalline GaAs deposited at different tem-
peratures. In [55], the XRD patterns of orthorhombic nanocrystalline GaAs were shown. Figure 1.8
(taken from [59]) shows the XRD pattern recorded from nanocrystalline GaAs deposited on an ITO
substrate. Three peaks were observed at 42.3191°, 44.307°, and 48.5141°, respectively. The d-values
were calculated to be 2.1414 Å (d1), 2.0499 Å (d2), and 1.8815 Å (d3), respectively. d1 and d2 were
close to the interplanar spacing of the (011) and (220) planes, respectively. d3 was close to the inter-
planar spacing of the (211) plane of cubic GaAs [55].
AFM was used to characterize GaAs nanocrystals [57,58,62]. In [57], the AFM image of PVA-
capped GaAs nanocrystal film deposited on an ITO substrate having crystal size of 80–100 nm was
presented. In [62], the authors showed the AFM images of PLD-synthesized GaAs for different num-
ber of pulses. Figure 1.9 (taken from [58]) shows the AFM images of GaAs nanocrystals synthesized
at current densities of 6.0, 3.5, and 1.5 mA cm−2, respectively, at 275 K. As shown in Figure 1.9a
through c, with increase in the current density, the size of GaAs nanocrystals increases. At a higher
temperature of 300 K, as shown in Figure 1.9d, GaAs nanocrystals have sizes of >250 nm.
Synthesis and Characterization of Nanocrystals and Nanoparticles 9

1 nm

0.1673 nm

FIGURE 1.7 HRTEM image of a nanocrystalline GaAs. (From Nayak, J. and Sahu, S.N., Appl. Surf. Sci.,
229, 97, 2004.)

800

600
(200)
Intensity (a.u.)

(220)

400
Bulk-GaAs

200
ITO substrate

0
30 35 40 45
2θ

FIGURE 1.8 X-ray diffraction pattern recorded from nanocrystalline GaAs on an ITO substrate. (From
Nayak, J. and Sahu, S.N., Physica E, 41, 92, 2008.)

Chemical synthesis, deposition, and epitaxy were used to synthesize InAs nanocrystals.
Chemical synthesis for InAs nanocrystals were described in [66–69]. In [66], InAs nanocrystals
of 2–6 nm were produced by the reaction between InCl3 and As[Si(CH3)3]3 with trioctylphosphine
(TOP) as the solvent and capping agent. In [67], InAs nanocrystals of 2–8 nm were fabricated by
solution-phase pyrolytic reaction of organometallic precursors. Surface passivation was carried
out with the organic ligand TOP. In [68], the authors reported the synthesis of InAs nanocrystals
with a size distribution of 7–60 nm by the solvothermal route at low temperature. In [70], InAs
nanocrystals were grown by depositing InAs in a two-dimensional supperlattice formed by the
GaAs matrix and AlAs barriers. In [71], the fabrication of InAs quantum dots of 60–65 nm on
10 Semiconductor Nanocrystals and Metal Nanoparticles

(a) (b)
100 nm 100 nm

(c) (d)
100 nm 100 nm

FIGURE 1.9 AFM image of GaAs thin films deposited on indium–tin–oxide (ITO) substrates at 275 K
and current densities (a) 6.0 mA cm−2, (b) 3.5 mA cm−2, (c) 1.5 mA cm−2. (d) Image with current density
2.5 mA cm−2 and temperature 300 K. (From Nayak, J. et al., Appl. Surf. Sci., 252, 2867, 2006.)

SiO2 patterned GaAs by metal organic chemical vapor deposition (MOCVD), followed by anneal-
ing in an arsine ambient, was described.
An epitaxial method was used to prepare InAs nanocrystals as reported in [75–79], such as
organometallic/metalorganic vapor-phase epitaxy (OMVPE/MOVPE) and MBE. In [76], an
InAs nanocrystal array of 75–280 nm was produced by MOVPE. A (100)-oriented InP wafer was
exposed to phosphine (PH3), with the introduction of arsine into the chamber. In [77], InAs nano-
crystals smaller than 25 nm were grown using a solid-source MBE and finally with rapid cooling
in arsenic ambient under certain pressure. In [78], InAs nanocrystals grown under a Volmer–
Weber-like mode on Si nanomembranes that were synthesized by selective etching silicon-on-
insulator substrate were described.
SEM images of InAs nanocrystals were presented in [71,78]. In [71], the authors showed
SEM micrographs of InAs nanocrystals with a center-to-center spacing of 80 nm, as shown in
Figure 1.10. The density of InAs nanocrystal could approach 1.8 × 1010 cm−2. In [78], an SEM
image of InAs nanocrystals deposited on a Si nanomembrane was presented.
TEM was extensively used to characterize InAs nanocrystals in [66,68,72–74]. In [73], TEM
characterization was carried out to study Si/InAs nanocrystals before and after thermal anneal-
ing. Besides, [76] provided a cross-sectional view of InAs nanocrystals grown on InP crystal along
Synthesis and Characterization of Nanocrystals and Nanoparticles 11

10.0 kV 10.8 mm ×200 k 200 nm

FIGURE 1.10 Array of InAs nanocrystals with a center-to-center spacing of 80 nm corresponding to a dot
density of 1.8 × 1010 cm−2. (From Dias, N.L. et al., Appl. Phys. Lett., 98, 141112, 2011.)

different directions. In [77], the authors presented images of InAs nanocrystals on SiO2 before
capping and after capping with 20 nm thick SiO2. Figure 1.11a shows InAs nanocrystals with an
average diameter of 5 nm. A higher resolution image of nanocrystals with diameters ranging from
5 to 10 nm is shown in Figure 1.11b [66].
AFM studies on InAs nanocrystals were presented in [71,74,76–78]. In [71], the authors showed
how single quantum dots and quantum dot clusters could be fabricated in pore diameters of 70 and
90 nm, respectively, as shown in Figure 1.12a and b. In [74], the topography of InAs nanocrystals/Si
was characterized by AFM, as well as the elemental redistribution of Si and As. In [77], the AFM
images of InAs nanocrystals grown at different temperatures by MBE were presented.
XRD patterns of InAs nanocrystals were shown in [66,68,73,78]. In [78], XRD patterns indicated a
nonuniform strain state of InAs nanocrystals. In [66], the authors presented the XRD patterns of InAs
nanocrystals of different sizes, as shown in Figure 1.13. The peak positions of the InAs nanocrystals
showed good consistency with those of bulk InAs, indicating the same cubic zinc blende lattice struc-
ture. The width of peaks varied considerably, indicating a substantial change in the crystalline size.
CuCl was used in applications such as electrooptic modulators, optical filters [80], solid-state
batteries [81], catalysts [82], adsorbent and air-purifying agents [83], and blue/UV light-emitting
devices [84]. It may be possible to apply CuCl nanocrystals in high-density optical storage, laser

5 nm

50 nm

(a) (b)

FIGURE 1.11 (a) TEM image of InAs nanocrystals of ~5 nm in diameter and (b) higher resolution TEM
image of InAs nanocrystals of 5–15 nm in diameter. (From Guzelian, A.A. et al., Appl. Phys. Lett., 69,
1432, 1996.)
12 Semiconductor Nanocrystals and Metal Nanoparticles

12.0 nm

0.0 nm
0 nm
1000 nm 1000 nm

750 750

500 500

250 250

(a) 250 500 750 1000 nm

12.0 nm

0.0 nm
0 nm
1000 nm 1000 nm

750 750

500 500

250 250

(b) 250 500 750 1000 nm

FIGURE 1.12 AFM profiles of quantum dots grown on 150 nm pitch pattern with (a) pore diameter of
70 nm and (b) pore diameter of 90 nm showing the transition from single QDs to QD clusters. (From Dias, N.L.
et al., Appl. Phys. Lett., 98, 141112, 2011.)

printing, projection displays, spectrophotometers, photocatalytic reactions, counterfeit detection,

chemical detection, traffic signals, and medicine.
Chemical methods were used to obtain CuCl nanocrystals, including hydrothermal methods
[85], ion implantation methods [86], double heat-treatment methods [87,88], and ionization in an
alkali halide matrix [89–97].
The hydrothermal methods were discussed in [85]. A green hydrothermal route was used to
prepare nanocrystalline CuCl powders via the reaction between CuCl2 and α-d-glucose in distilled
water. Double heat treatment methods were presented in [87,88]. The two-step glass melting was
conducted with a 10 Hz, Q-switched Nd:YAG laser on silica glass with NaCl and CuO additives.
CuCl microcrystallites were formed in the first heating step in silica glass. The second step of heat-
ing the glass containing CuCl microcrystallites induced nanometer-sized CuCl. The crystal size
could be controlled by varying the temperature due to laser heating.
In [89–98], CuCl nanocrystals synthesized in an alkali halide matrix were discussed; there were
two main approaches mentioned to fabricate single alkali halide crystals doped with CuCl, namely,
the Czochralski method and the Bridgman method. There appeared to be no significant differences
in the optical properties of crystals prepared by the two methods.
As can be seen in Figure 1.14, SEM characterization shows that the CuCl nanocrystals syn-
thesized through hydrothermal methods exhibit lamellar crystallites [85]. XRD is widely used to
Synthesis and Characterization of Nanocrystals and Nanoparticles 13

28 Å

39 Å
Counts (a.u.)

51 Å

62 Å

100 Å

20 30 40 50 60 70 80
2θ

FIGURE 1.13 XRD patterns of InAs nanocrystals with different sizes. (From Guzelian, A.A. et al., Appl.
Phys. Lett., 69, 1432, 1996.)

FIGURE 1.14 The SEM image of as-synthesized CuCl powders. (From Zhang, Y.C. and Tang, J.Y., Mater.
Lett., 61, 3708, 2007.)
14 Semiconductor Nanocrystals and Metal Nanoparticles

characterize CuCl nanocrystals and crystal orientations [94,95]. A typical XRD pattern is pre-
sented in Figure 1.15. The XRD pattern of the CuCl nanocrystals synthesized using hydrothermal
methods exhibits three diffraction peaks, namely, from the (111), (220), and (311) crystal planes of
cubic phase. The lattice constant calculated from XRD data was 5.414 Å. The XRD characteristics
of CuCl nanocrystals in NaCl matrix prepared with Czochralski method was discussed in [94,95].
Diffraction from the (111), (200), (220), (311), (400), and (331) planes of CuCl of zinc blende struc-
ture can be observed in the patterns.

1.2.2 Elemental Semiconductor Nanocrystals

Chemical etching, electrochemical etching, and dry etching have always been used to synthesize
freestanding Si nanocrystals [100–103]. In [100], Si nanocrystals of several tens of nanometers were
obtained from the pyrolysis of SiH4 and etched subsequently in a two-phase solution of HF. The
as-fabricated Si nanocrystals had a large size distribution from several nanometers to several tens
nanometers. In [101], the synthesis of Si nanocrystals was carried out by the pyrolysis of SiH4 and
subsequent size reduction by chemical etching in hydrofluoric acid in a two-phase cyclohexane/pro-
panol-2 solution. In [103], freestanding Si nanocrystals of 1.5–2 nm were obtained by etching off the
oxide from the SiO2/Si nanocomposite. The etching time was controlled for Si nanocrystal size reduc-
tion. Subsequently, organic ligands were attached to the Si nanocrystal surface via hydrosilylation.
Electrochemical etching was used to synthesize Si nanocrystals [104–106]. In [104], the
authors reported that Si nanocrystals with an average diameter of 3–5 nm were prepared by
­sonication of porous Si in an ultrasonic bath under Ar ambient. In [105], Si nanocrystals of
1.5–5 nm were synthesized by the electrochemical etching of boron and gallium-doped Si wafers
in a mixture of hydrofluoric acid with pure ethanol. The Si nanocrystals were scratched off
but aggregated in micrometer-sized grains, which were then subjected to a laser fragmenta-
tion ­process to achieve nanosized crystals. In [106], oxidized Si nanocrystals with the size of
~2.5–3 nm were prepared by electrochemical etching of monocrystalline Si wafers in a solution
of HF, ethanol, and hydrogen peroxide and subsequent mechanical pulverization. In [107], the
authors reported dry etching for obtaining Si nanocrystals smaller than 6 nm. Si nanocrystals
of larger sizes were synthesized in a SiH4-based plasma, which were then passed through a
111
Intensity (a.u.)

220

311

20 30 40 50 60
2θ (°)

FIGURE 1.15 XRD pattern of as-synthesized CuCl particles. (From Zhang, Y.C. and Tang, J.Y., Mater. Lett.,
61, 3708, 2007.)
Synthesis and Characterization of Nanocrystals and Nanoparticles 15

CF4-based plasma for etching into a desirable size. In [108] a method to produce Si nanocrystals
of about 2–3 nm capped with SiOx by dehydration of SiH4 was described.
Figure 1.16 shows SEM images of the as-prepared Si nanocrystals grains and after laser frag-
mentation for 55 min in ethanol [105]. Si nanocrystals were separated from the grains, and the
average size was decreased. After undergoing the laser fragmentation process for 55 min, the mean
size of the micrograins decreased from 25 μm to 5 µm [105]. In [103], the TEM image of large Si
nanocrystals revealed the crystalline structure of the nanocrystals, as shown in Figure 1.17.
In [101], AFM was utilized to show the array growth of Si nanospikes, as presented in Figure 1.18.
A monolayer of polystyrene balls having a diameter ~1.5 μm was prepared, as shown in Figure
1.18a and b. It was used as a mask to allow Si nanosize spike growth between three adjacent indi-
vidual balls. After the deposition, the polystyrene balls were removed by some solvent, as shown in
Figure 1.18c. An array of Si nanospikes was formed, as shown in Figure 1.18d.

50 μm 50 μm
SE WD15.3 mm 20.0 kV ×900 50 µm SE WD18.8 mm 20.0 kV ×1.0 k 50 µm
(a) (b)

FIGURE 1.16 SEM images of Si micrograins (a) as prepared and (b) fragmented by nanosecond laser
processing in ethanol for 55 min. (From Švrček, V. et al., Acta Mater., 59, 764, 2011.)

FIGURE 1.17 HRTEM image of Si nanocrystals. (From Maier-Flaig, F. et al., Chem. Phys., 405, 175, 2012.)
16 Semiconductor Nanocrystals and Metal Nanoparticles

1.200 μm
0 1200 1200
1000 1000
800 800
600 600
400 1.200 μm
400
10 200 200
0 0
μm

nm nm
0

10
20 20

μm
20 10
μm
0 10 20 0
(a) μm (b)

250.0 nm
0 250 60
200 50
150 40
100 60.00 nm 30
20
50 10
10
0 0
μm

nm 0 nm

10
20 20
μm

20 10 μm
0 10 20 0
(c) μm (d)

FIGURE 1.18 (a, b) A monolayer of polystyrene beads. (c) After removal of polystyrene. (d) Formation of
array of Si nanospikes. (From Fojtik, A. et al., Chin. Opt. Lett., 5, 250, 2007.)

Figure 1.19 shows the XRD pattern of Si nanocrystals synthesized at different pyrolysis tempera-
tures [99]. As can be observed in the figure, the cubic crystalline structure was confirmed and the
size of Si crystals was in the range 3–5 nm, which slightly increased with the pyrolysis temperature.
Methods of synthesizing freestanding Ge nanocrystals were reported, mainly using MBE [109]
and chemical synthesis [110–112].
In [109], Ge nanocrystals were grown on Si (001) substrates pre-covered with carbon. The sam-
ples were prepared by MBE in the temperature range 350°C–750°C. The nonuniform distribution of
carbon, together with the enhanced surface roughness, leads to the early onset of Ge nanocrystals
having a size of 10–50 nm.
In [110], Ge nanocrystals with a broad size distribution of 2–70 nm were synthesized by the
supercritical fluid method. Ge nanocrystals were synthesized by precipitation in hexane and octanol
at 400°C–550°C under 20.7 MPa in a continuous-flow reactor. Diphenylgermane (DPG) and tetra-
ethylgermane (TEG) were used as Ge precursors, which underwent thermolysis to crystalline Ge.
Octanol was added to control the particle size and to serve as capping ligand.
In [111], Ge nanocrystals of 2.5–4.7 nm were obtained by the reaction of GeCl4 and Mg in tet-
rahydrofuran (THF). The capping reaction of the remaining GeCl group was conducted by alkyl
Grignard reagents in THF. The synthesis route to Ge nanocrystals is illustrated in Figure 1.20.
In [112], Ge nanocrystals were synthesized in a plasma reactor chamber via the decomposition of
GeCl4. Nanocrystal formation was realized in the plasma by the clustering of Ge atoms. Ge nano-
crystals were collected from stainless steel meshes or substrates placed on meshes. The obtained
crystals were characterized by SEM and TEM, as shown in Figure 1.21. The SEM images showed
Synthesis and Characterization of Nanocrystals and Nanoparticles 17

Intensity (a.u.) T = 1200°C

T = 1175°C

T = 1100°C

T = 1050°C

T = 1000°C

0 10 20 30 40 50 60 70 80 90 100
2θ

FIGURE 1.19 XRD spectra of the Si nanocrystals at different pyrolysis temperatures. (From Sorarù, G.D.
et al., Appl. Phys. Lett., 83, 749, 2003.)

Cl Mg in THF RBr RBr at r.t.

Cl Ge Cl
ultrasonication 1 h at 10°C 0.5 h at 10°C overnight
Cl 1.5 h at 10°C 2 h at 50°C

FIGURE 1.20 Synthesis route of Ge nanocrystals with GeCl4 [p3]. R represents n-propyl or tert-butyl group.
(From Watanabe, A. et al., Appl. Organometal. Chem., 19, 530, 2005.)

100 nm
10 nm

(b)

Ge-NCs
Au
Si

5.0 kV X25,000 1 μm WD 10.1 mm

(a) (c)

FIGURE 1.21 (a) Image for TEM characterization for Ge nanocrystals on Au/Si substrate. (b) Image of
electron diffraction on a selected area. (c) Image of a cross-section of as-deposited sample by SEM. (From
Holman, Z.C. and Kortshagen, U.R., Langmuir, 25, 11883, 2009.)
18 Semiconductor Nanocrystals and Metal Nanoparticles

XRD size:

800°C ~25 nm

500°C 5.0 nm
Counts (a.u.)

300°C 4.9 nm

250°C 4.9 nm

200°C 4.9 nm

28°C 4.9 nm

20 40 60 80
2θ (°)

FIGURE 1.22 X-ray diffraction patterns of Ge nanocrystals of different sizes. (From Holman, Z.C. and
Kortshagen, U.R., Langmuir, 25, 11883, 2009.)

that the Ge nanocrystals were formed on the Au/Si substrate, and TEM images showed that the
nanocrystals had an average size of 4–5 nm. The XRD patterns in Figure 1.22 show Ge nanocrys-
tals having different sizes synthesized by plasma decomposition of GeCl4. The results indicated that
the nanocrystal size could be controlled by varying the annealing temperature.

1.3 METAL NANOPARTICLES

Synthesis of metal nanoparticles for applications such as catalysis, electronics, optics, environmen-
tal, and biotechnology is of substantial interest. Gold, silver, and copper have been used mostly
for the synthesis of stable nanoparticles, which are useful in areas such as photography, catalysis,
biological labeling, photonics, optoelectronics, and surface-enhanced Raman scattering detection.
Additionally, metal nanoparticles show surface-plasmon resonance absorption in the UV–visible
region. Generally, metal nanoparticles can be prepared and stabilized by physical and chemical
methods. Chemical approaches, such as chemical reduction, electrochemical techniques, and pho-
tochemical reduction, are widely used.
Many methods for preparing Au nanoparticles have been reported, mainly employing chemical,
physical, and biological methods. Chemical methods have been more widely used to synthesize Au
nanoparticles in aqueous media such as citrate or sodium borohydride [113–128].
In [113], the authors described Au nanoparticles with diameters of 1.5–20 nm synthesized by
the reaction between HAuCl4 · 3H2O and N(C8H17)4Br(s). Compounds C12H25NH2(s) and NaBH4 were
added to the above solution. Then black/brown precipitate in the organic solution was filtered and
washed by ethanol. The gold nanoparticles that were filtered out were then dried in a vacuum cham-
ber. Nanoparticles of 1–80 nm in diameter were prepared by mixing didodecyldimethylammonium
bromide (DDAB) dissolved in toluene and gold chloride dissolved in the micelle solution by sonica-
tion [115,116]. Then NaBH4 solution was added dropwise at room temperature. The color of the solu-
tion turned dark red after ~20 s. Au nanoparticles were obtained after stirring for another 15 min
for completing the reaction. In [117], the synthesis of Au nanoparticles between 1.5 and 7 nm was
reported. AuCl3 was dissolved in toluene with ammonium surfactants, and then tetrabutylammonium
Synthesis and Characterization of Nanocrystals and Nanoparticles 19

borohydride (TBAB) was mixed with hydrazine in toluene to achieve Au nanoparticles. In [118], the
authors reported Au nanoparticles of ~30 nm by the reaction of bis-(p-sulfonatophenyl)phenylphos-
phine dihydrate dipotassium (BSPP) and H2O2. Finally, sodium citrate solution and HAuCl4 were
added to obtain Au nanoparticles. In [119], dodecanethiol-capped Au nanoparticles of 2.1–8.3 nm
were prepared by mixing AuPPh3Cl and dodecanethiol in benzene with the introduction of tert-
butylamine-borane. Au nanoparticles were fabricated by HAuCl4 aqueous solution in toluene with
the aid of a phase-transfer agent (tetraoctylammonium bromide, TOAB) [121]. Phenylethylthiol was
then added with NaBH4 solution to complete the reaction. Au nanoparticles were dried by rotary
evaporation and then collected. In [122], the authors described Au nanoparticles with diameters
of 13 and 45 nm from a solution of gold chloride trihydrate with the addition of sodium citrate
tribasic dehydrate (C6H5Na3O7 · 2H2O). In [123], it was reported that a piece of glassy carbon plate
was placed into the seed solution of 4 nm Au colloid that was prepared by mixing NaBH4 solution,
HAuCl4 solution, and trisodium citrate. Finally, Au nanoparticles on the glassy carbon surface was
washed into a container with pure water, and then dried in nitrogen ambient. In [124], 15 ± 3 nm
Au nanoparticles were prepared by adding sodium citrate to heated hydrogen tetrachloroaurate
(HAuCl4) solution, which was followed by a washing process. In [125], the borohydride reduction
method was reported to synthesize Au nanoparticles of 8.1 ± 1.1 nm by injecting HAuCl4 solution
into sodium borohydride. In [126], gold (III) TOAB solution was synthesized by mixing a solution
of HAuCl4 · 3H2O and another of tetraoctylammonium bromide (TOAB) in toluene and thiol deriva-
tive of calixarene [4]; and then NaBH4 and MeOH solutions were added. Au nanoparticles of 2 ±
0.4 nm were obtained by dispersing and filtering the resulting solution. In [127], Au nanoparticles
of 12–41 nm were achieved by the reaction of propanoic acid 2-(3-acetoxy-4,4,14-trimethylandrost-
8-en-17-yl) (PAT) with an aqueous HAuCl4 solution. In [128], Au nanoparticles were fabricated by
mixing an ethanolic solution of tryptamine (TRA) and another ethanolic solution of KAuCl4,
followed with filtering, washing with pure cold water, and drying under vacuum.
Physical methods were also widely used to synthesize gold nanoparticles, such as RF sputtering
[129,130], laser ablation techniques [131,132], proton beam irradiation [133], UV irradiation [134–136],
atom beam co-sputtering [137], PLD [138], microwave (MW) irradiation [139], and multipulse meth-
ods (electrochemical synthesis) [140]. In [132], the authors reported the synthesis of Au nanoparticles
by focusing a Nd:YAG laser beam on a gold target in pure DI water or in an aqueous solutions of
PAMAM G5. In [133], proton beam irradiation was utilized to prepare 10 nm Au nanoparticles from a
mixed solution of aqueous HAuCl4, aqueous cetyltrimethylammonium bromide (CTAB), and aqueous
NaOH in a tube. In [135], Au nanoparticles were synthesized by UV irradiation of HAuCl4 dissolved
in distilled water in the presence of poly(vinyl pyrrolidone) (PVP) under a nitrogen atmosphere. In
[136], the authors reported the synthesis of Au nanoparticles of 10–40 nm by UV irradiation of an
aqueous solution containing AuCl4−. In [139] , Au nanoparticles of <5 nm were fabricated using auric
solutions placed inside a high-pressure microwave chamber. In [140], the synthesis of Au nanoparticles
of ~27 nm by five times oxidation/reduction sequence in a HAuCl4 solution was reported.
In [141–144], biological methods to synthesize Au nanoparticles were reported. In [141], a bio-
logical approach to fabricate Au nanoparticles of 6 nm was presented by adding trimethylphosphin-
chlorogold salt (ClAuPMe3) mixed with a reducing agent. In [142], the authors reported the synthesis
of Au nanocrystals with an average diameter of 5–13 nm by using HAuCl4 · 3H2O aqueous solution,
biochemical β-d-glucose solution, and NaOH. Au nanoparticles of 5–50 nm were prepared by mix-
ing the centrifuged supernatant and chloroauric acid (HAuCl4) aqueous solution [143]. In [144],
the synthesis of Au nanoparticles of 14–60 nm was reported using HAuCl4 solution and aqueous
Terminalia arjuna fruit extract, as shown in Figure 1.23.
TEM and HRTEM have been extensively used to characterize Au nanoparticles [114–122,124,​
126–133,135,136,138,139,141–144]. In [114], TEM images of amine-capped Au nanoparticles were
presented, while [116] presented the TEM images of ripening, segregation, and Au nanoparticle
formation in colloids. In [117], the authors presented the TEM images of Au nanoparticles obtained
using the single-phase approach. In [119], TEM images showed Au nanoparticles with different sizes
20 Semiconductor Nanocrystals and Metal Nanoparticles

HAuCl4
+
Water soluble polyphenols from aqueous
extract of Terminalia arjuna fruit
OH HO HO HO
Low HO O Moderate
+ HO + HO
availability OH availability
O OH HO HO
O
Small size anisotropic Au NPs
Large size anisotropic Au NPs
High availability

Small size uniform spherical Au NPs

FIGURE 1.23 Illustration of the synthesis of Au nanoparticles with Terminalia arjuna fruit extract.
(From Kumar, K.M. et al., Spectrochim. Acta A: Mol. Biomol. Spectrosc., 116, 539, 2013.)

produced by varying the reaction solvent at different temperatures. In [122], TEM images of core-
satellite structure containing Au nanoparticles were shown, whereas [126] showed the TEM images
of Au nanoparticles before and after the addition of Co(II) in various concentrations. In [128], TEM
studies showed a relatively narrow size distribution of Au nanoparticles in the range 36.6 ± 5.30 nm.
In [132], TEM images of Au nanoparticles in water and in water plus PAMAM G5 as a stabilizing
agent were shown. In [133], the authors presented TEM images of Au nanoparticles produced under
irradiation of different fluences, durations, and solutions. In [144], TEM images of Au nanoparticles
were shown with size and morphology variations depending on the quantities of extracts.
In [127], TEM images revealed the shape and size of Au nanoparticles synthesized using PAT
having a size of 12–41 nm. In [135], the HRTEM images of Au nanoparticles fabricated at low or
high concentration of HAuCl4 as a precursor were shown. In [139], TEM images of Au nanoparticles
obtained by a microwave-assisted method were presented. In [142], TEM images of β-d-glucose-
stabilized Au nanoparticles (5–11 nm) were shown. In [143], a TEM study characterized polygonal
Au nanoparticles of 5–50 nm. Figure 1.24 shows monodisperse Au nanoparticles of diameter of
2.2 nm (Figure 1.24a), and Au nanoparticles with lattice structure (Figure 1.24b), as given in [121].
XRD patterns of gold nanoparticles were shown in [113,114,121,127,128,134,140,143,144]. In [113],
the authors reported the patterns of Au nanoparticles of four different sizes. In [114], the XRD patterns

20 nm 2 nm

(a) (b)

FIGURE 1.24 (a) TEM and (b) HRTEM images of gold nanoparticles. (From Qian, H.F. et al., PNAS, 109,
696, 2012.)
Synthesis and Characterization of Nanocrystals and Nanoparticles 21

111

200

220 311

10 20 30 40 50 60 70 80 90
2θ

FIGURE 1.25 XRD patterns of gold nanoparticles exhibiting fcc structure. (From Qian, H.F. et al., PNAS,
109, 696, 2012.)

of amine-capped Au nanoparticles were presented. In [127], XRD patterns displayed PAT-stabilized

Au nanoparticles of the face-centered cubic (fcc) structure. The XRD technique was used in [128]
to investigate the fcc crystallinity of the Au-TRA nanoparticles. In [134], the XRD patterns of pure
PMMA and Au–PMMA nanocomposites were shown before and after UV irradiation. In [142], the
XRD patterns of β-d-glucose-stabilized Au nanoparticles were shown. In [143], XRD patterns of Au
nanoparticles with main crystallographic planes were presented, and [144] showed discrete diffraction
peaks in the XRD pattern that could be assigned to the planes of cubic Au nanoparticles. In [121], the
XRD pattern showed the fcc structure of Au nanoparticles in Figure 1.25.
In [114,130,134,136,137], analyses of Au nanoparticles using the XPS technique were presented.
In [114], the XPS spectra of amine-capped Au nanoparticles synthesized by different schemes were
shown, based on which the stability of Au nanoparticles obtained from different synthesis schemes
was discussed. In [130], the authors presented the XPS spectra of chemical state Au 4f from Au
nanoparticles formed in the SiO2 matrix. The binding energy difference between Au 4f 7/2 and
Au 4f5/2 was 3.62 eV, which was close to the standard value of 3.67 eV, indicating that metallic Au
nanoparticles were formed in the SiO2 matrix [130]. In [134], XPS core levels of Au–PMMA nano-
composites after UV irradiation were discussed. XPS spectra in [137] showed Au 4f and Si 3p core
levels of the as-deposited and annealed Au–SiOx nanocomposite films deposited on Si substrates.
Ag nanoparticles exhibit favorable properties for applications in optoelectronics, magnetic data
storage, killing microbes, printable electronics, ultrasensitive chemicals, biological sensors, cata-
lysts, and many others [145].
Many techniques were used to prepare Ag nanoparticles [146–158]. In [146], elemental Ag was
evaporated at temperatures ranging from 1200 to 1500 K in a helium ambient. The flow stream
was cooled over a short distance to 400 K to form Ag nanoparticles of 4–6 nm. Alkylthiol molecules
were used for the etching and passivation of Ag nanoparticles. In [147], silver nanoparticles were pro-
duced and passivated by dodecanethiol monolayers. In [148], Ag nanoparticles were fabricated using a
technique similar to that in [146,147] and were passivated by SC12H25 chain molecules.
The ion exchange/irradiation method was reported in [150,151]. The Na+ → Ag+ ion-exchange
technique was employed to form Ag ions in the surface layers of sodalime silicate glass by ion
22 Semiconductor Nanocrystals and Metal Nanoparticles

exchange in a AgNO3/NaNO3 solution at 300°C–380°C. Ions of 400–500 keV He, 1 MeV Ne, or
2 MeV Xe were used to induce the formation of Ag nanoparticles in sodalime silicate glass.
A chemical reduction method was reported for the synthesis of Ag nanoparticles [156,157], In
[156], Ag nanoparticles were formed by adding aqueous AgNO3 dropwise to a reducing solution
of NaBH4. Variations in particle size and distribution were obtained by adjusting the ratio between
AgNO3 and NaBH4. In [157], a solution of AgNO3 and sodium dodecyl sulfonate (SDS) was placed
in an ice bath, to which was then added NaBH4 with stirring for 3 h at room temperature. The pH
value of the reaction solution was adjusted by an acetate buffer solution. Ag nanoparticles with an
average size of 2 nm were then obtained.
TEM has been extensively used to characterize Ag nanoparticles [146–148,150,151,154–157].
In [147], the authors showed the TEM images of Ag nanoparticles. The atomic lattices and 3D
morphology of the silver nanoparticles were observed. In [148], bright-field and dark-field TEM
images of a monolayer of tetrahedral Ag nanoparticles were shown. In [151], the TEM images of Ag
ion–exchanged BK7 glass after 1 MeV Xe irradiation were presented. In [154], the authors showed
the images of Ag nanoparticle/CNT hybrid structure, and also revealed the morphologies of the
connection region between Ag nanoparticles and CNTs. In [157], the TEM image indicated that
the average size of Ag nanoparticles was about 2 nm and that the nanoparticles were spherical in
shape. Figure 1.26 shows the Ag nanoparticles with typical sizes ranging from 50 to >100 nm [158].

50 nm 50 nm
(a) (b)
–
C O

O–
–

O
O
O

C
C

O O
C

– OO O –
O C C OO O
O
C
C

O O
O COO OC O
Ag O –
O C
– O O C O
CO OC
O O
OO OO
O O
C

C
C
C

O
C

–
– O O 5 nm 5 nm
O
C O
O C
O

(c) (d) (e)

–
–

FIGURE 1.26 (a) Silver nanoparticles produced by traditional polyol process using polyvinylpyrrolidone
(PVP) as the surfactant. The coagulation of growing particles results in polycrystalline products as indicated
by the arrows. (b) A representative TEM image showing the uniform silver nanoparticles produced through
the modified polyol process using poly(acrylic acid) (PAA) as the surfactant. (c) Schematic illustration of
PAA-coated silver nanoparticles. HRTEM images of silver nanoparticles (d) without and (e) with defects.
(From Hu, Y.X. et al., J. Solid State Chem., 181, 1524, 2008.)
Synthesis and Characterization of Nanocrystals and Nanoparticles 23

1800

1500 111

1200

Intensity (a.u.)
900

200
600
220
311
300 222

0
10 20 30 40 50 60 70 80
2θ

FIGURE 1.27 XRD pattern of Ag nanoparticles. (From Zheng, C.Z. et al., J. Anal. Methods Chem., 2013,
Article ID 261648, 2013.)

Figure 1.26b shows the images of uniform silver nanoparticles produced using the polyol process.
Figure 1.26c illustrates the surface morphology of the silver nanoparticles synthesized with a sur-
factant. Figure 1.26d and e shows the lattice structures of Ag nanoparticles and the defects inside
them (Figure 1.26e) [158].
XRD was used to characterize Ag nanoparticles [150,154,157]. In [150], the authors showed the
XRD spectrum of an ion-exchanged sample irradiated with 500 keV He at a fluence of 1.1 × 1017
ions cm−2. In [154], the XRD patterns of Ag nanoparticles/CNTs hybrid structures were shown. The
angles 2θ ≈ 38.2° and 2θ ≈ 44.3° corresponded to Ag (111) and (200) planes, respectively. In [157],
a typical XRD pattern of Ag nanoparticles was presented, as shown in Figure 1.27. The 2θ values
of 38.12°, 44.28°, 64.43°, 77.48°, and 81.54° correspond to (111), (200), (220), (311), and (222) planes,
respectively.
STM images of Ag nanoparticles were presented in [149,153]. In [149], the STM images of Ag
nanoparticles and their lattice structure with different electronic transport properties were shown.
In [153], the authors reported an STM study of Ag nanoparticles grown on a single crystal of
SrTiO3(001). The STM images in Figure 1.28 show that the top surface of Ag nanoparticles have
two, three, or fivefold symmetry.

1.4 NANOCRYSTALS AND NANOPARTICLES EMBEDDED

IN A DIELECTRIC FILM
The observation of visible photoluminescence from indirect-bandgap material systems, that is,
nanocrystals in a dielectric matrix, presents potential applications in optoelectronic/optical devices
[159]. The nanocrystals embedded in a dielectric film can be synthesized using techniques such as
ion implantation, CVD, sputtering, and so on [159–199]. On the other hand, dielectric layers embed-
ded with nanocrystals/nanoparticles have also been widely studied due to their possible applications
in nonvolatile memory devices [160].
Si nanocrystals embedded in a dielectric matrix can be synthesized using techniques such as ion
implantation, CVD, and sputtering. One promising technique to realize Si nanocrystals embedded
in a dielectric matrix is by the implantation of Si ions into dielectric films [161–173]. First, dielectric
films such as SiO2, Si3N4, or Al2O3 were grown on a substrate. Then Si ions were accelerated in an
arc chamber and implanted into the film at selected energies of 1–200 keV with dose ranging from
24 Semiconductor Nanocrystals and Metal Nanoparticles

[010]

[100]

FIGURE 1.28 STM image of Ag nanoparticles deposited on a SrTiO3 substrate. (From Silly, F. and Castell,
M.R., Appl. Phys. Lett., 87, 213107, 2005.)

1015 to 1017 cm−2. A subsequent annealing at a high temperature (e.g., 1000°C) in N2 or Ar ambient
for a few minutes to a few hours was usually conducted to induce nanocrystal formation and remove
defects. The size of Si nanocrystal synthesized with this technique was usually 1–10 nm.
Si nanocrystals can be grown by CVD in dielectrics such as SiO2 and Si3N4. A Si-rich SiO2 (SiOx)
film was formed using very-high frequency (VHF) PECVD from a N2O:SiH4 mixture. A subsequent
high-temperature annealing (~1000°C–1300°C) of the SiOx films leads to the formation of nc-Si
embedded in the SiO2 matrix [184–187]. Low-pressure chemical vapor deposition (LPCVD) was
used to grow Si dots on Si substrate. Then a thin oxide/nitride film was grown on the substrate,
which also partially oxidized/nitridized the Si dots, resulting in Si nanocrystals embedded in the
SiO2/Si3N4 matrix [188,189]. Electron cyclotron resonance chemical vapor deposition (ECRCVD)
was employed to grow Si-rich silicon nitride films by controlling the SiH4 and PH3 flow rates.
A subsequent annealing was conducted to induce Si nanocrystal formation in Si3N4 [190].
Sputtering was used to synthesize Si nanocrystals embedded in a dielectric layer. In [176,179],
SiO2 films containing Si nanocrystals were prepared by co-sputtering Si and SiO2 targets in an
Ar gas atmosphere, which were annealed after deposition. Si nanocrystals embedded in SiC film
prepared by magnetron co-sputtering were reported in [181–183]. The Si-rich amorphous silicon
carbide films were deposited at room temperature by co-sputtering in an argon plasma. Subsequent
high-temperature annealing was carried out in nitrogen ambient. Si nanocrystals embedded in
Al2O3 film was synthesized by co-sputtering Si and Al2O3 targets [180].
PLD can be used to control Si nanocrystal size from 2 to 10 nm depending on the background
gas species and pressure [192]. In [193–195], mechanical ball milling followed by spin-coating was
used to prepare Si nanocrystals embedded in SiO2 films. Si nanocrystals embedded in SiO2 layer can
also be fabricated by atomic layer deposition [191]. Such films can also be synthesized by thermal
evaporation of SiO2 powder onto rotating substrates, followed by thermal annealing [196].
Synthesis and Characterization of Nanocrystals and Nanoparticles 25

TEM was extensively used to characterize Si nanocrystals embedded in a dielectric matrix

[164,166,169,170,175,177,181–183,188–190]. In [164], the authors presented a cross-sectional view
of Si nanocrystals prepared by Si ion implantation, as shown in Figure 1.29. Figure 1.29a shows the
Si nanocrystals embedded SiO2 film sandwiched between the capping layer and the Si substrate.
Figure 1.29b shows the lattice fringes of Si nanocrystals. Figure 1.29c shows that the as-prepared
Si nanocrystals had an average size of ~2 nm. In [169], the effect of Si implantation fluence on the
microstructure of the Si nanocrystals and the SiO2 matrix was examined using TEM; defects and
faceting were observed in Si nanocrystals. In [170], a TEM image of Si nanocrystals of ~3 nm
embedded in SiO2 matrix was shown. In [175], the TEM micrograph showed a high density of Si
nanocrystals with a mean radius of ~3 nm. In [177], TEM characterization showed Si nanocrystals
of 3–5 nm embedded in a SiC matrix. TEM images in [181] illustrated the formation of Si nanocrys-
tals/SiC films at different annealing temperatures. The size of Si nanocrystals in [182] character-
ized by TEM was in the range 3–5 nm. In [183], the TEM images of the Si-rich SiC films annealed
at high temperature containing Si nanocrystals of 2–5 nm were shown, and in [190] those of a
phosphorus-doped Si nanocrystals/SiNx film.
XRD studies of Si nanocrystals embedded in dielectric films were reported in [164,180,182,188,190].
In [180], the XRD patterns of Si nanocrystals in Al2O3 prepared with different magnetron powers
and annealing temperatures were shown. The formation of Si nanocrystals embedded in SiC after
annealing was confirmed by XRD measurements in [182]. And, [188] presented the XRD patterns
of single and multiple Si nanocrystals layers. The XRD patterns in [190] showed the average Si
nanocrystals/SiNx film fabricated by ECRCVD.
In [164], the small-angle XRD patterns for different ion implantation doses and annealing dura-
tions were given, which are shown in Figure 1.30. The Si atomic concentration of 15% and annealed
for 1 h exhibited a size distribution of 2–10 nm, while the sample containing Si at 10% and annealed
for 0.5 h showed a Gaussian distribution of nanocrystal sizes peaking at 6 nm.
XPS studied were carried out in [166,169,172,180,181,183,190]. In [166,197,198], the Si-implanted
profile was obtained by the XPS analysis and secondary ion mass spectrometry (SIMS). In [169], the
effect of Si implantation fluence on the microstructure of the Si nanocrystals and the SiO2 matrix
was examined by XPS. In [172], XPS was used to identify Si oxide and its electronic structure
at different annealing temperatures for various durations. In [180], XPS depth profiling revealed
the phase separation and formation of Si nanocrystals and SiO2. In [181], the authors provided the
chemical composition of the as-deposited layer containing C, Si, O, and N characterized by XPS.

I
Surface
I 80 nm

II
II
Si: SiO2

2 nm
Si substrate
3 nm
(a) (b) (c)

FIGURE 1.29 Cross-sectional TEM images of (a) the film structure. (b) Lattice structure of Si nanocrystals.
(c) Average size of Si nanocrystals for about 2 nm. (From Guha, S. et al., J. Appl. Phys., 88, 3954, 2000.)
26 Semiconductor Nanocrystals and Metal Nanoparticles

15 at.% Si 10 at.%
x-ray intensity (a.u.)

x-ray intensity (a.u.)

×10

0 5 10 0 5 10

(a) Diameter (nm) (b) Diameter (nm)

FIGURE 1.30 Small-angle X-ray diffraction intensity as a function of particle size from (a) 15 at.%
(1 h annealing) and (b) 10 at.% (1/2 h annealing). (From Guha, S. et al., J. Appl. Phys., 88, 3954, 2000.)

In [183], the as-deposited and annealed samples were characterized by XPS to study the chemical
structure of Si-rich SiC co-sputtered films, and [190] presented the XPS analysis of the core levels
of p-doped Si nanocrystals/SiNx films.
Figure 1.31 shows the deconvolution of the Si 2p peak for the as-implanted and annealed films.
In [199], Si nanocrystal formation in Si-implanted SiO2 films as a function of thermal annealing
was investigated. There were five subpeaks corresponding to Si suboxidation states Sin+ (n = 0, 1,
2, 3, and 4) in the SiO2 films, as shown in Figure 1.31. The XPS results showed the evolution of
the chemical structures and the formation of Si nanocrystals as functions of the annealing tem-
perature and annealing time. In [197], depth profiles of the charging effect in Si nanocrystals were
determined. The charging effect decreased with the increase of nc-Si concentration and vanished
when a densely stacked nanocrystal layer was formed as a result of charge diffusion among the
Si nanocrystals. In [198], the relative concentration of each oxidation state Sin+ (n = 0, 1, 2, 3, and 4)
at various depths was determined quantitatively by XPS analysis. The effects of annealing on both
the oxidation states and their depth distributions were presented.
Ge nanocrystals embedded in a dielectric matrix show promising applications in optoelectronic,
photovoltaic, and nonvolatile memory devices. Germanium has a larger dielectric constant and
Bohr radius compared than Si [200,201] and hence the quantum size effects should be prominent in
Ge nanocrystals even for a relatively large size.
Various techniques have been employed to synthesize Ge nanocrystals in the SiO2 matrix.
Ion implantation was found to be a promising technique to synthesize Ge nanocrystals in SiO2
[202–209]. Ge ions were implanted into thermally grown SiO2 films and followed by high-tem-
perature annealing in inert ambient. A broad distribution of Ge nanocrystals embedded in rela-
tively thick SiO2 films had been achieved by the ion implantation at high energies from 75 to 450
keV [202,205–207,210]. On the other hand, an orderly two-dimensional array of Ge nanocrystals
embedded in a thin SiO2 film was realized with a low-energy (<10 keV) Ge ion implantation [211].
A sol–gel method for the synthesis of Ge nanocrystals in a SiO2 matrix was reported in [212].
Silica glass containing 7 wt% Ge was prepared by the sol–gel method by using Si(OC2H5)4 and
GeCl4 as starting materials. Ge nanocrystals of ~5 nm embedded in SiO2 glass could be obtained
after a heating treatment. A sol–gel method using TEOS and 3-trichlorogermanium propanoic acid
(Cl3–Ge–C2H4 –COOH) as starting materials was purposed in [213,214]. The Ge nanocrystals were
Synthesis and Characterization of Nanocrystals and Nanoparticles 27

×101
160
As-implanted sample
Si 2p core level

120 Si4+
Intensity (CPS)

Si0

80 Si3+

Si2+ Si1+
40

0
108 106 104 102 100 98
(a) Binding energy (eV)

×102

100 Annealed at 1000°C for 20 min

Si 2p core level

80 Si4+
Intensity (CPS)

60

40
Si3+
Si2+ Si0
20 Si1+

0
106 104 102 100
(b) Binding energy (eV)

FIGURE 1.31 Deconvolution of Si 2p spectra for (a) the As-implanted sample and (b) the sample annealed
at 1100°C for 20 min (b). (From Liu, Y. et al., J. Phys. D: Appl. Phys., 36, L97, 2003.)

obtained in the range 1–13 nm by control of the reaction time and speed. Ge nanocrystals can be
deposited on SiO2 film surface by LPCVD and an additional capping SiO2 was grown to make nano-
crystals embedded in SiO2 [215,216,237,256].
The synthesis of Ge nanocrystals embedded in SiO2 has been demonstrated by the co-sputtering
small pieces of Ge wafers attached on the pure SiO2 target, followed by furnace-annealing at 800°C
for 30 min [217,218]. Spherical Ge nanocrystals dispersed in the SiO2 matrix with a mean size of
~6 nm could be achieved. To avoid the long process duration of the conventional furnace annealing,
rapid thermal annealing (RTA) was employed for the formation of Ge nanocrystals [219,220–223].
Selective oxidation of a thin polycrystalline Si1−xGex layer is another technique to synthesize Ge
nanocrystals embedded in SiO2 [224–233]. Deposition of a layer of Si1−xGex film followed by dry or
wet oxidation can realize Ge nanocrystals embedded in SiO2. The deposition of the Si1−xGex film can
28 Semiconductor Nanocrystals and Metal Nanoparticles

be done by many methods, including the co-sputtering of Si and Ge [224–226], LPCVD [227–230],
ultrahigh vacuum CVD (UHVCVD) [231], ion implantation [232], and MBE [233]. The annealing
ambient, that is, dry or wet oxidation, also affected the formation of Ge nanocrystals [227]. A proper
control of the annealing conditions, for example, ambient, temperature, and duration, was crucial
for the synthesis of Ge nanocrystals embedded in SiO2 using the selective oxidation of Si1−xGex.
Besides the above-mentioned techniques, other techniques such as e-beam evaporation
[234–236], PLD [238–240], and nano-patterning using a focused ion beam (FIB) [241,242]
were demonstrated for the synthesis of Ge nanocrystals embedded in a SiO2 matrix or on the
SiO2 surface.
TEM is an important technique to determine the crystallinity, external shape, size, and distribu-
tion of the nanocrystals [243–248]. Figure 1.32 shows typical cross-sectional TEM images of Ge
nanocrystals embedded in SiO2 synthesized by the co-sputtering technique. Figure 1.32a shows the
electron diffraction pattern of the crystal structure of the Ge nanocrystals, and Figure 1.32b shows
the structure of the Ge nanocrystal.
The coexistence of elemental Ge and Ge oxides in SiO2 embedded with Ge has been studied by
XPS [221,249,250]. Figure 1.33 shows a typical XPS analysis of the Ge nanocrystals embedded in a
SiO2 thin film deposited by CVD. The reduction of GeOx and the formation of elemental Ge could
be observed after thermal annealing. Moreover, the XPS analysis also provided useful information
about the changes in the electronic structure of nanocrystals as a shift in the core-level binding ener-
gies. Shifts to the high values of the binding energy with the reduction of particles size have been
reported in [251–255].
Al nanoparticles embedded in aluminum nitride (AlN) films or Al2O3 have attracted significant
interest due to their applications in memory devices. Charge trapping/distrapping in the Al nanopar-
ticles leads to a shift in the flat-band voltage (VFB) of the metal–insulator–semiconductor (MIS)
structure, which can be used in applications of memory devices [257–264].
The DC arc–discharge evaporation method was presented in [257]. The samples were prepared
by evaporation using a DC arc discharge of metallic aluminum in an ambient of N2 + NH3. The vol-
ume ratio between NH3 and (N2 + NH3) was adjusted to grow nanoparticles. AlN–Al nanocompos-
ites were formed at a slightly lower ratio of <5%. RF magnetron sputtering synthesis was described

(a) (b) (c)

10.0 nm 5.0 nm

FIGURE 1.32 (a) Electron diffraction pattern and cross-sectional HRTEM image for a typical SiO2 embedded
with Ge nanocrystals synthesized by the co-sputtering technique. (b) Image of Ge nanocrystals in low resolution.
(c) High resolution image of Ge nanocrystals. (From Fujii, M. et al., Jpn. J. Appl. Phys., 30, 687, 2002.)
Synthesis and Characterization of Nanocrystals and Nanoparticles 29

Annealed
Ge

GeOx

Intensity (a.u.)

As-deposited

45 20
Binding energy (eV)

FIGURE 1.33 XPS spectra for the as-deposited and annealed Ge nanocrystals in SiO2. (From Dutta, A.K.,
Appl. Phys. Lett., 68, 1189, 1996.)

in [258–264]. The deposition was carried out by RF (13.56 MHz) magnetron sputtering of a pure Al
target in a gas mixture of argon and nitrogen. The RF power was varied to achieve AlN films with
different Al:N ratios.
TEM was extensively used to characterize Al nanocrystals [257–264]. In [257], the TEM micro-
graph and SAD pattern of AlN–A1 nanocomposite were shown. In [258–264], the TEM images of
Al nanoparticles/nanoclusters embedded in the AlN matrix were presented. A typical micrograph
of Al nanoparticles is shown in Figure 1.34, where TEM image of Al nanocrystals (nc–Al) embed-
ded in the AlN matrix can be observed with size of 4–8 nm [260].

Al nanocrystal

AIN matrix

5 nm

FIGURE 1.34 TEM image of Al nanoparticles embedded in AlN matrix. (From Liu, Y. et al., Appl. Phys. A,
93, 483, 2008.)
30 Semiconductor Nanocrystals and Metal Nanoparticles

In [276], by varying the RF power between 100 W (sample 1), 150 W (sample 2), and
250 W (sample 3), three types of samples were obtained, which were characterized by XPS.
The chemical composition of the deposited films could be changed by varying the RF power.
The as-deposited films consisted of metallic Al embedded in the AlN and Al 2O3 matrix, as
shown in Figures 1.35 and 1.36, respectively. It was observed that at a higher RF power, the
Al concentration was lower.
Al2O3 is considered a promising candidate for the gate dielectric of field-effect transistors due
to its excellent electrical properties such as low leakage current, high dielectric constant (~8), and
high breakdown voltage (~9 MV cm−1). Recently, it was shown that Al nanoparticles embedded in
Al2O3 thin film can be used to realize resistive memory devices [266,269,271,275] and floating-gate
nonvolatile memory devices [273].
Two techniques were used for the synthesis of Al nanocrystals/nanoparticles embedded in
Al2O3 matrix. The RF sputtering of Al target was carried out in O2 ambient at a power of 250–310
W [265,267–270,272–275]. The thickness of the film was dependent on the sputtering duration.
The Ar/O2 ratio was set to different values to achieve different film properties. The films were
annealed with a rapid thermal process for a duration varying from 30 s to 5 min. Aluminum
anodization is another promising technique to fabricate Al nanoparticles [266,271]. A thin alumi-
num film was deposited on a silicon substrate by RF sputtering or other film growth techniques.
Several hundreds of nanometers may be anodized. A platinum foil was used as the cathode.
Anodization of the film was carried out in an oxalic acid solution at room temperature under a
DC voltage of ~40 V for a duration of ~2 h. The thickness of the anodized aluminum oxide was
nearly 30–50 nm [266,271].
XPS was used to characterize Al-rich Al2O3 thin films [267,270,272,273]. In [273], elemental Al
was observed in Al 2p core levels, as shown in Figure 1.36a. In [267], XPS analysis indicated that
thermal annealing could cause reactions at the interface between the Al-rich Al2O3 thin film and the
Si substrate, leading to an increase in the Al concentration at the interface of Al2O3/Al. TEM char-
acterizations were conducted in [270,273,274], where Al nanocrystals/nanoparticles were observed,
as shown in Figure 1.36b. The size of Al nanocrystal/nanoparticle prepared by sputtering was in the
range of several nanometers to tens of nanometers [270,273,274].
Photoemission intensity (a.u.)

AlN

Al

Al2O3

76 75 74 73 72
Binding energy (eV)

FIGURE 1.35 Peak decomposition of Al 2p core levels of the Al-rich AlN thin film. (From Liu, Y. et al.,
Appl. Phys. Lett., 87, 033112, 2005.)
Synthesis and Characterization of Nanocrystals and Nanoparticles 31

Experimental data
Background
Peak sum fitting

Photoemission internsity (a.u.)

Elemental Al
Al2O3

78 76 74 72 70
(a) Binding energy (eV)

nc-Al

5 nm Al2O3 matrix

(b)

FIGURE 1.36 TEM image of AlN thin film embedded with Al. (a) XPS result of Al rich Al2O3. (b) TEM image
of Al2O3 thin film embedded with Al nanocrystals. (From Liu, Z. et al., Appl. Phys. Lett., 96, 173110, 2010.)

REFERENCES
1. C.B. Murray, D.J. Norris, and M.G. Bawendi, J. Am. Chem. Soc. 115 (1993) 8706.
2. A.P. Alivisatos, Science 271 (1996) 933.
3. W.W. Yu, L. Qu, W. Guo, and X. Peng, Chem. Mater. 15 (2003) 2854.
4. M. Achermann, M.A. Petruska, S. Kos, D.L. Smith, D.D. Koleske, and V.I. Klimov, Nature 429 (2004) 642.
5. A.H. Mueller, M.A. Petruska, M. Achermann, D.J. Werder, E.A. Akhadov, D.D. Koleske, M.A.
Hoffbauer, and V.I. Klimov, Nano Lett. 5 (2005) 1039.
32 Semiconductor Nanocrystals and Metal Nanoparticles

6. N. Pradhan, D. Goorskey, J. Thessing, and X. Peng, J. Am. Chem. Soc. 127 (2005) 17586.
7. M.J. Bruchez, M. Moronne, P. Gin, S. Weiss, and A.P. Alivisatos, Science 281 (1998) 2013.
8. X.G. Peng, L. Manna, W.D. Yang, J. Wickham, E. Scher, A. Kadavanish, and A.P. Alivisatos, Nature
404 (2000) 59.
9. E.D. Sone, and S.I. Stupp, J. Am. Chem. Soc. 126 (2004) 12756.
10. R. Baron, C.H. Huang, D.M. Bassani, A. Onopriyenko, M. Zayats, and I. Willner, Angew. Chem., Int.
Ed. 44 (2005) 4010.
11. H. Tsuji and A.K. Kato, Angew. Chem., Int. Ed. 44 (2005) 3565.
12. C.B. Murray, D.J. Norris, and M.G. Bawendi, J. Am. Chem. Soc. 115 (1993) 8706.
13. R. Xie, X. Zhong, and T. Basche, Adv. Mater. 17 (2005) 2741.
14. M.J. Bowers II, J.R. McBride, and S.J. Rosenthal, J. Am. Chem. Soc. 127 (2005) 15378.
15. B.O. Dabbousi, J. Rodriguez-Viejo, F.V. Mikulec, J.R. Heine, H. Mattoussi, R. Ober, K.F. Jensen, and
M.G. Bawendi, J. Phys. Chem. B 101 (1997) 9463.
16. M.A. Hines and P. Guyot-Sionnest, J. Phys. Chem. 100 (1996) 468–471.
17. L. Manna, E.C. Scher, and A. Paul Alivisatos, J. Am. Chem. Soc. 122 (2000) 12700–12706.
18. L.H. Qu, Z.A. Peng, and X.G. Peng, Nano Lett. 1 (2001) 333–337.
19. D.V. Talapin, A.L. Rogach, A. Kornowski, M. Haase, and H. Weller, Nano Lett. 1 (2001) 207–211.
20. K. Walzer, U.J. Quaade, D.S. Ginger, N.C. Greenham, and K. Stokbro, J. Appl. Phys. 92 (2002) 1434.
21. D.G. Wu, M.E. Kordesch, and P. Gregory Van Patten, Chem. Mater. 17 (2005) 6436–6441.
22. Y.A. Yang, H.M. Wu, K.R. Williams, and Y.C. Cao, Angew. Chem. 117 (2005) 6870–6873.
23. S.J. Wang, H. Kim, H.H. Park, Y.S. Lee, and H. Jeon, J. Vac. Sci. Technol. A 28 (2010) 559.
24. A.L. Rogach, A. Kornowski, M.Y. Gao, A. Eychmüller, and H. Weller, J. Phys. Chem. B 103 (1999)
3065–3069.
25. Q. Peng, Y.J. Dong, Z.X. Deng, and Y.D. Li, Inorganic Chem. 41 (2002) 5249.
26. A.E. Raevskaya, A.L. Stroyuk, S.Y. Kuchmii, Y.M. Azhniuk, V.M. Dzhagan, V.O. Yukhymchuk, and
M.Y. Valakh, Proceedings of 25th International Conference on Microelectronics (MIEL), Beirut,
Lebanon, Vol. 1 (2006), p. 4244-0117.
27. J. Mu and X.B. Gao, J. Dispers. Sci. Technol. 26 (2005) 763–767.
28. X.M. Xu, Y.L. Wang, L.Y. Zhou, L.L. Wu, J.Z. Guo, Q. Niu, L.F. Zhang, and Q.F. Liu, Micro Nano Lett.
7 (2012) 589–591.
29. H. Mathieu, T. Richard, J. Allègre, P. Lefebvre, G. Arnaud, W. Granier, L. Boudes, J.L. Marc, A. Pradel,
and M. Ribes, J. Appl. Phys. 77 (1995) 287.
30. H. Yang, P.H. Holloway, G. Cunningham, and K.S. Schanze, J. Chem. Phys. 121 (2004) 10233.
31. Q. Wang, D.C. Pan, S.C. Jiang, X.L. Ji, L.J. An, and B.Z. Jiang, Chem. Eur. J. 11 (2005) 3843–3848.
32. T. Serranoa, I. Gómeza, R. Colásb, and J. Cavazosb, Colloids Surf. A: Physicochem. Eng. Aspects 338
(2009) 20–24.
33. L. Zou, Z. Fang, Z.Y. Gu, and X.H. Zhong, J. Lumin. 129 (2009) 536–540.
34. B. Girginer, G. Galli, E. Chiellini, and N. Bicak, Int. J. Hydrogen Energy 34 (2009) 1176–1184.
35. Q. Xiao and C. Xiao, Appl. Surf. Sci. 255 (2009) 7111–7114.
36. J. Chen, Y.J. Ma, G.C. Fan, Y.F. Li, J.Y. Jiang, and Z.Y. Huang, Mater. Lett. 65 (2011) 1768–1771.
37. P.Q. Zhao, S.J. Xiong, X.L. Wu, and P.K. Chu, Appl. Phys. Lett. 100 (2012) 171911.
38. M.V. Artemyev, V. Sperling, and U. Woggon, J. Appl. Phys. 81 (1997) 6975.
39. I. Umezua, R. Koizumia, K. Mandaia, T. Aoki-Matsumotoa, K. Mizunoa, M. Inada, A. Sugimura,
Y. Sunaga, T. Ishii, and Y. Nagasaki, Microelectron. Eng. 66 (2003) 53–58.
40. Y.C. Cao and J.H. Wang, J. Am. Chem. Soc. 126 (2004) 14336–14337.
41. N. Ghows and M.H. Entezari, Ultrasonics Sonochem. 18 (2011) 269–275.
42. N. Ghows and M.H. Entezari, Ultrasonics Sonochem. 18 (2011) 629–634.
43. D. Matsuura, Y. Kanemitsu, T. Kushida, C.W. White, J.D. Budai, and A. Meldrum, Appl. Phys. Lett. 77
(2000) 2289.
44. M.A. Zhukovskiy, A.L. Stroyuk, V.V. Shvalagin, N.P. Smirnova, O.S. Lytvyn, and A.M. Eremenko,
J. Photochem. Photobiol. A: Chem. 203 (2009) 137–144.
45. M.F. Kotkata, A.E. Masoud, M.B. Mohamed, and E.A. Mahmoud, Physica E 41 (2009) 1457–1465.
46. K.S. Rathore, P.D. Deepika, N.S. Saxena, and K. Sharma, AIP Conference Proceedings, Malaysia, Vol.
1249 (2010), p. 145.
47. A.L. Korotkov, J. Appl. Phys. 93 (2003) 786.
48. W. Schrenk, N. Finger, S. Gianordoli, E. Gornik, and G. Strasser, Appl. Phys. Lett. 77 (2000) 3328.
49. N. Usami, Y. Azuma, T. Ujihara, G. Sazaki, and K. Nakajima, Phys. Lett. 77 (2000) 3565.
50. Y. Kanemitsu, H. Tanaka, S. Mimura, and T. Kushida, J. Lumin. 83–84 (1999) 301–304.
Synthesis and Characterization of Nanocrystals and Nanoparticles 33

51. Y. Kanemitsu, H. Tanaka, T. Kushida, K.S. Min, and H.A. Atwater, Physica E 7 (2000) 322–325.
52. Y. Kanemitsu, H. Tanaka, T. Kushida, K.S. Min, and H.A. Atwater, J. Lumin. 87–89 (2000) 432–434.
53. K. Taniguchi, Y. Morishige, and Y. Kanemitsu, Physica E 17 (2003) 79–81.
54. J. Nayak and S.N. Sahu, Appl. Surf. Sci. 182 (2001) 407–412.
55. J. Nayak and S.N. Sahu, Appl. Surf. Sci. 229 (2004) 97–104.
56. J. Nayaka, R. Mythilib, M. Vijayalakshmib, and S.N. Sahu, Physica E 24 (2004) 227–233.
57. J. Nayak and S.N. Sahu, Physica E 30 (2005) 107–113.
58. J. Nayak, S.N. Sahu, and S. Nozaki, Appl. Surf. Sci. 252 (2006) 2867–2874.
59. J. Nayak and S.N. Sahu, Physica E 41 (2008) 92–95.
60. H.S. Mavi, A.K. Shukla, B.S. Chauhan, and S.S. Islam, Mater. Sci. Eng. B 107 (2004) 148–154.
61. R.A. Ganeev, A.I. Ryasnyanskiy, and T. Usmanov, Opt. Commun. 272 (2007) 242–246.
62. P. Dubček, B. Pivac, S. Milošević, N. Krstulović, Z. Kregar, and S. Bernstorff, Appl. Surf. Sci. 257
(2011) 5358–5361.
63. O.A. Neucheva, A.A. Evstrapov, Y.B. Samsonenko, and G.E. Cirlin, Tech. Phys. Lett. 33 (2007)
923–925.
64. H. Usui, S. Mukai, H. Asuda, and H. Mori, J. Cryst. Growth 311 (2009) 2269–2274.
65. H. Matsumoto, H. Uchida, H. Yoneyama, T. Sakata, and H. Mori, Res. Chem. Intermed. 20 (1994)
723–733.
66. A.A. Guzelian, U. Banin, A.V. Kadavanich, X. Peng, and A.P. Alivisatos, Appl. Phys. Lett. 69 (1996)
1432–1434.
67. O. Millo, D. Katz, Y. Levi, Y.W. Cao, and U. Banin, J. Low Temp. Phys. 118 (2000) 365–373.
68. X.X. Xu, K.H. Yu, W. Wei, B. Peng, S.H. Huang, Z.H. Chen, and X.S. Shen, Appl. Phys. Lett. 89 (2006)
253117.
69. S.A. Jewett and A. Ivanisevic, Acc. Chem. Res. 45 (2012) 1451–1459.
70. L. Rebohle, F.F. Schrey, S. Hofer, G. Strasser, and K. Unterrainer, Appl. Phys. Lett. 81 (2002) 2079.
71. N.L. Dias, A. Garg, U. Reddy, J.D. Young, and V.B. Verma, Appl. Phys. Lett. 98 (2011) 141112.
72. Y.L. Zi, K. Jung, D. Zakharov, and C. Yang, Nano Lett. 13 (2013) 2786–2791.
73. A. Tchebotareva, J.L. Brebner, S. Roorda, P. Desjardins, and C.W. White, J. Appl. Phys. 92 (2002) 4664.
74. S. Prucnal, S. Facsko, C. Baumgart, H. Schmidt, M.O. Liedke, L. Rebohle, A. Shalimov et al., Nano
Lett. 11 (2011) 2814–2818.
75. S.A. Dayeh, E.T. Yu, and D. Wang, J. Phys. Chem. C 111 (2007) 13331–13336.
76. H.D. Fonseca-Filho, C.M. Almeida, R. Prioli, M.P. Pires, and P.L. Souza, J. Appl. Phys. 107 (2010)
054313.
77. M. Hocevar, P. Regreny, A. Descamps, D. Albertini, and G. Saint-Girons, Appl. Phys. Lett. 91 (2007)
133114.
78. C. Deneke, A. Malachias, A. Rastelli, L. Merces, M. Huang, F. Cavallo, O.G. Schmidt, and M.G.
Lagally, Am. Chem. Soc. 6 (2012) 10287–10295.
79. M. Bennour, F. Saidi, L. Bouzaïene, L. Sfaxi, and H. Maaref, J. Appl. Phys. 111 (2012) 024310.
80. O. Ambacher, J. Phys. D Appl. Phys. 31 (1998) 2653.
81. A. Brune and S. Jiang, Mater. Res. Bull. 30 (1995) 573.
82. Z. Li, K.C. Xie, and R.C.T. Slade, Appl. Catal. A: Gen. 209 (2001) 107–115.
83. V. Recupero, L. Pino, M. Cordarö, A. Vita, F. Cipitĺ, and M. Laganà, Fuel Process. Technol. 85 (2004)
1445–1452.
84. F.O. Lucas, L. ÓReilly, G. Natarajan, P.J. McNally, S. Daniels, D.M. Taylor, S. William, D.C. Cameron,
A.L. Bradley, and A. Miltra, J. Cryst. Growth 287 (2006) 112–117.
85. Y.C. Zhang and J.Y. Tang, Mater. Lett. 61 (2007) 3708–3710.
86. O.A. Podsvirov, A.I. Sidorov, V.A. Tsekhomskii, and A.V. Vostokov, Phys. Solid State 52 (2010)
1906–1909.
87. A. Nakamura and H. Ohmura, J. Lumin. 83–84 (1999) 97–103.
88. M.Y. Lee, Bull. Korean Chem. Soc. 16 (1995) 126–129.
89. M. Haselho, K. Reimann, and H.-J. Weber, J. Cryst. Growth 196 (1999) 135–140.
90. M. Haselhoff and H.-J. Weber, Mater. Res. Bull. 30 (1995) 607–612.
91. D. FröGhlich, H.F. Haselhoff, and K. Reimann, Solid State Commun. 94 (1995) 189–191.
92. M. Haselho, K. Reimann, and H.-J. Weber, Eur. Phys. J. B 12 (1999) 147–155.
93. M. Haselhoff and H.-J. Weber, Phys. Rev. B 58 (1998) 5052.
94. S. Mahtout, M.A. Belkhir, and M. Samah, Acta Phys. Polonica A 105 (2004) 279–286.
95. S. Mahtout, M.A. Belkhir, and M. Samah, Semicond. Phys. Quant. Electron. Optoelectron. 7 (2004)
185–189.
34 Semiconductor Nanocrystals and Metal Nanoparticles

96. K. Yamanak, K. Edamatsu, and T. Itoh, J. Lumin. 76&77 (1998) 256–259.

97. P.M. Valov and V.I. Leiman, Phys. Solid State 51 (2009) 1703–1708.
98. T. Itoh, S. Yano, S. Iwai, K. Edamatsu, T. Goto, and A. Ekimov, Mater. Sci. Eng. A 217&218 (1996)
167–170.
99. G.D. Sorarù, S. Modena, P. Bettotti, G. Das, G. Mariotto, and L. Pavesi, Appl. Phys. Lett. 83 (2003) 749.
100. A. Fojtik, J. Valenta, The Ha Stuchlíková, J. Stuchlík, I. Pelant, and J. Kočka, Thin Solid Films 515
(2006) 775–777.
101. A. Fojtik, J. Valenta, I. Pelant, M. Kalal, and P. Fial, Chin. Opt. Lett. 5 (2007) 250–253.
102. T.L. Sudesh, L. Wijesinghe, E.J. Teo, and D.J. Blackwood, Electrochim. Acta 53 (2008) 4381–4386.
103. F. Maier-Flaig, E.J. Henderson, S. Valouch, S. Klinkhammer, C. Kubel, G.A. Ozin, and U. Lemmer,
Chem. Phys. 405 (2012) 175–180.
104. E. Froner, E. D’Amato, R. Adamo, N. Prtljaga, S. Larcheri, L. Pavesi, A. Rigo, C. Potrich, and M. Scarpa,
J. Colloid Interface Sci. 358 (2011) 86–92.
105. V. Švrček, T. Yamanari, Y. Shibata, and M. Kondo, Acta Mater. 59 (2011) 764–773.
106. K. Kůsová, J. Non-Cryst. Solids 358 (2012) 2130–2133.
107. X.D. Pi, R.W. Liptak, S.A. Campbell, and U. Kortshagen, Appl. Phys. Lett. 91 (2007) 083112.
108. J. Martin, F. Cichos, and C. von Borczyskowski, J. Lumin. 132 (2012) 2161–2165.
109. O. Leifeld, R. Hartmann, E. Müller, E. Kaxiras, K. Kern, and D. Gürtzmacher, Nanotechnology
10 (1999) 122–126.
110. X.M. Lu, K.J. Ziegler, A. Ghezelbash, K.P. Johnston, and B.A. Korgel, Nano Lett. 4 (2004) 969–974.
111. A. Watanabe, F. Hojo, and T. Miwa, Appl. Organometal. Chem. 19 (2005) 530–537.
112. Z.C. Holman and U.R. Kortshagen, Langmuir 25 (2009) 11883–11889.
113. D.V. Leff, P.C. Ohara, J.R. Heath, and W.M. Gelbart, J. Phys. Chem. 99 (1995) 7036–7041.
114. D.V. Leff, L. Brandt, and J.R. Heath, Langmuir 12 (1996) 4723–4730.
115. X.M. Lin and C.M. Sorensen, Chem. Mater. 11 (1999) 198–202.
116. X.M. Lin, C.M. Sorensen, and K.J. Klabunde, J. Nanopart. Res. 2 (2000) 157–164.
117. N.R. Jana and X.G. Peng, J. Am. Chem. Soc. 125 (2003) 14280–14281.
118. E. Hao, R.C. Bailey, G.C. Schatz, J.T. Hupp, and S.Y. Li, Nano Lett. 4 (2004) 327–330.
119. N.F. Zheng, J. Fan, and G.D. Stucky, J. Am. Chem. Soc. 128 (2006) 6550–6551.
120. C.-L. Lu, K.S. Prasad, H.-L. Wu, J. Annie Ho, and M.H. Huang, J. Am. Chem. Soc. 132 (2010)
14546–14553.
121. H.F. Qian, Y. Zhu, and R.C. Jin, PNAS 109 (2012) 696–700.
122. Z.Q. Weng, H.B. Wang, J. Vongsvivut, R.Q. Li, A.M. Glushenkov, J. He, Y. Chen, C.J. Barrow, and
W.R. Yang, Anal. Chim. Acta 803 (2013) 128–134.
123. G. Chang, H.H. Shu, K. Ji, M. Oyama, X. Liu, and Y.B. He, Appl. Surf. Sci. 288 (2014) 524–529.
124. U. Bubniene, M. Oćwieja, B. Bugelyte, Z. Adamczyk, M. Nattich-Rak, J. Voronovic, A. Ramanaviciene,
and A. Ramanavicius, Colloids Surf. A: Physicochem. Eng. Aspects 441 (2014) 204–210.
125. H.H. Deng, G.W. Li, L. Hong, A.L. Liu, W. Chen, X.H. Lin, and X.H. Xia, Food Chem. 147 (2014)
257–261.
126. D. Maity, R. Gupta, R. Gunupuru, D.N. Srivastava, and P. Paul, Sensors Actuat. B 191 (2014) 757–764.
127. M. Venkatachalam, K. Govindaraju, A. Mohamed Sadiq, S. Tamilselvan, V. Ganesh Kumar, and
G. Singaravelu, Spectrochim. Acta A Mol. Biomol. Spectrosc. 116 (2013) 331–338.
128. M.A. de Carvalho, P.F. Andrade, F.C.A. Corbi, M. do C. Goncalves, A.L.B. Formiga, I.O. Mazali,
J.A. Bonacin, and P.P. Corbi, Synthet. Metals 185–186 (2013) 61–65.
129. D. Barreca, A. Gasparotto, E. Tondello, G. Bruno, and M. Losurdo, J. Appl. Phys. 96 (2004) 1655.
130. C.C. Wang, J.Y. Tseng, T.B. Wu, L.J. Wu, C.S. Liang, and J.M. Wu, J. Appl. Phys. 99 (2006) 026102.
131. J.Y. Yang, J.H. Kim, W.J. Choi, Y.H. Do, C.O. Kim, and J.P. Hong, J. Appl. Phys. 100 (2006) 066102.
132. E. Giorgetti, A. Giusti, S.C. Laza, P. Marsili, and F. Giammanco, Phys. Status Solidi (a) 204 (2007)
1693–1698.
133. Y.-J. Kim, G. Cho, and J. Hee Song, Nucl. Instrum. Methods Phys. Res. B 246 (2006) 351–354.
134. M.K. Abyaneh, D. Paramanik, S. Varma, S.W. Gosavi, and S.K. Kulkarni, J. Phys. D Appl. Phys. 40
(2007) 3771–3779.
135. M.M. Chili, V.S.R. Rajasekhar Pullabhotla, and N. Revaprasadu, Mater. Lett. 65 (2011) 2844–2847.
136. S. Boufi, M.R. Vilar, A.M. Ferraria, and A.M. Botelho do Rego, Colloids Surf. A: Physicochem. Eng.
Aspects 439 (2013) 151–158.
137. S. Mohapatra, Y.K. Mishra, D.K. Avasthi, D. Kabiraj, J. Ghatak, and S. Varma, Appl. Phys. Lett. 92
(2008) 103105.
138. K.C. Chan, P.F. Lee, and J.Y. Dai, Appl. Phys. Lett. 95 (2009) 113109.
Synthesis and Characterization of Nanocrystals and Nanoparticles 35

139. C. Vargas-Hernandez, M.M. Mariscal, R. Esparza, and M.J. Yacaman, Appl. Phys. Lett. 96 (2010)
213115.
140. C. Fernández-Blanco, A. Colina, A. Heras, V. Ruiz, and J. López-Palacios, Electrochem. Commun. 18
(2012) 8–11.
141. R. Djalali, Y.F. Chen, and H. Matsui, J. Am. Chem. Soc. 125 (2003) 5873–5879.
142. J.C. Liu, G.W. Qin, P. Raveendran, and Y. Ikushima, Chem. Eur. J. 12 (2006) 2131–2138.
143. L. Karthik, G. Kumar, T. Keswani, A. Bhattacharyya, B. Palakshi Reddy, K.V. Bhaskara Rao, Nanomed.
Nanotechnol. Biol. Med. 9 (2013) 951–960.
144. K.M. Kumar, B.K. Mandal, H.A. Kiran Kumar, and S.B. Maddinedi, Spectrochim. Acta A Mol. Biomol.
Spectrosc. 116 (2013) 539–545.
145. V.І. Pokhmurs’kyi, V.М. Dovhunyk, М.М. Student, M.D. Klapkiv, V.М. Posuvailo, and А.R. Kytsya, Mat.
Sci. 48 (2013) 636–641.
146. S.A. Harfenist, Z.L. Wang, M.M. Alvarez, I. Vezmar, and R.L. Whetten, J. Phys. Chem. 100 (1996)
13904–13910.
147. S.A. Harfenist, Z.L. Wang, R.L. Whetten, I. Vezmar, and M.M. Alvarez, Adv. Mater. 9 (1997) 817–822.
148. Z.L. Wang, S.A. Harfenist, I. Vezmar, R.L. Whetten, J. Bentley, N.D. Evans, and K.B. Alexander, Adv.
Mater. 10 (1998) 808–812.
149. S.-H. Kim, G. Medeiros-Ribeiro, D.A.A. Ohlberg, R. Stanley Williams, and J.R. Heath, J. Phys. Chem.
B 103 (1999) 10341–10347.
150. D.P. Peters, C. Strohhöfer, M.L. Brongersma, J. van der Elsken, and A. Polman, Nucl. Instrum. Methods
Phys. Res. B 168 (2000) 237–244.
151. J.J. Penninkhof, A. Polman, L.A. Sweatlock, S.A. Maier, H.A. Atwater, A.M. Vredenberg, and B.J. Kooi,
Appl. Phys. Lett. 83 (2003) 4137–4139.
152. J.M. Warrender and M.J. Aziz, Appl. Phys. A 79 (2004) 713–716.
153. F. Silly and M.R. Castell, Appl. Phys. Lett. 87 (2005) 213107.
154. Z.X. Wang, X.N. Li, C.L. Ren, Z.Z. Yong, J.K. Zhu, W.Y. Luo, and X.M. Fang, Sci. China Ser. E Tech.
Sci. 52 (2009) 3215–3218.
155. P. Benzo, L. Cattaneo, C. Farcau, A. Andreozzi, M. Perego, G. Benassayag, B. Pécassou, R. Carles, and
C. Bonafos, J. Appl. Phys. 109 (2011) 103524.
156. A. Thøgersen, J. Bonsak, C.H. Fosli, and G. Muntingh, J. Appl. Phys. 110 (2011) 044306.
157. C.Z. Zheng, H.P. Wang, L.Z. Liu, M.J. Zhang, J.G. Liang, and H.Y. Han, J. Anal. Methods Chem. 2013
(2013) Article ID 261648.
158. Y.X. Hu, J.P. Ge, D. Lim, T. Zhang, and Y.D. Yin, J. Solid State Chem. 181 (2008) 1524–1529.
159. T. Shimizu-lwayama, K. Fujita, S. Nakao, K. Saitoh, T. Fujita, and N. Itoh, J. Appl. Phys. 75 (1994)
7779–7783.
160. S. Tiwari, F. Rana, H. Hanafi, A. Hartstein, and E.F. Crabbé, Appl. Phys. Lett. 68 (1996) 1377–1379.
161. C.E. Chryssou, A.J. Kenyon, T.S. Iwayama, C.W. Pitt, and D.E. Hole, Appl. Phys. Lett. 75 (1999) 2011.
162. S.-H. Choi and R.G. Elliman, Appl. Phys. Lett. 74 (1999) 3987.
163. P.G. Kik, M.L. Brongersma, and A. Polman, Appl. Phys. Lett. 76 (2000) 2325.
164. S. Guha, S.B. Qadri, R.G. Musket, M.A. Wall, and T. Shimizu-Iwayama, J. Appl. Phys. 88 (2000) 3954.
165. A.R. Wilkinson and R.G. Elliman, Appl. Phys. Lett. 83 (2003) 5512.
166. P. Pellegrino, B. Garrido, C. Garcia, J. Arbiol, J.R. Morante, M. Melchiorri, N. Daldosso, L. Pavesi,
E. Scheid, and G. Sarrabayrouse, J. Appl. Phys. 97 (2005) 074312.
167. I. Sychugov, A. Galeckas, N. Elfström, A.R. Wilkinson, R.G. Elliman, and J. Linnros, Appl. Phys. Lett.
89 (2006) 111124.
168. Y. Liu, T.P. Chen, L. Ding, S. Zhang, Y.Q. Fu, and S. Fung, J. Appl. Phys. 100 (2006) 096111.
169. V. Levitcharsky, R.G. Saint-Jacques, Y.Q. Wang, L. Nikolova, R. Smirani, and G.G. Ross, Surf. Coat.
Technol. 201 (2007) 8547–8551.
170. R.-J. Zhang, Y.-M. Chen, W.-J. Lu, Q.-Y. Cai, Y.-X. Zheng, and L.-Y. Chen, Appl. Phys. Lett. 95 (2009)
161109.
171. T.S. Iwayama, Vacuum 86 (2012) 1634–1637.
172. O. Korotchenkov, A. Podolian, V. Kuryliuk, B. Romanyuk, V. Melnik, and I. Khatsevich, J. Appl. Phys.
111 (2012) 063501.
173. R. Khelifi, D. Mathiot, R. Gupta, D. Muller, M. Roussel, and S. Duguay, Appl. Phys. Lett. 102 (2013)
013116.
174. G.S. Chang, J.H. Son, K.H. Chae, C.N. Whang, E.Z. Kurmaev, S.N. Shamin, V.R. Galakhov, A. Moewes,
and D.L. Ederer, Appl. Phys. A 72 (2001) 303–306.
175. S. Kim, Y.M. Park, S.-H. Choi, and K.J. Kim, J. Appl. Phys. 101 (2007) 034306.
36 Semiconductor Nanocrystals and Metal Nanoparticles

176. K. Watanabe, H. Tamaoka, M. Fujii, and S. Hayashi, J. Appl. Phys. 92 (2002) 4001.
177. D. Song, E.-C. Cho, G. Conibeer, Y. Huang, and M.A. Green, Appl. Phys. Lett. 91 (2007) 123510.
178. R. Limpens and T. Gregorkiewicz, J. Appl. Phys. 114 (2013) 074304.
179. J. Skov Jensen, D.A. Buttenschön, T.P. Leervad Pedersen, J. Chevallier, B. Bech Nielsen, and
A. Nylandsted Larsen, J. Appl. Phys. 101 (2007) 056108.
180. I. Dogan, I. Yildiz, and R. Turan, Physica E 41 (2009) 976–981.
181. D. Song, E.-C. Cho, G. Conibeer, Y.-H. Cho, Y. Huang, S. Huang, C. Flynn, and M.A. Green, J. Vac. Sci.
Technol. B 25 (2007) 1327.
182. D. Song, E.-C. Cho, G. Conibeer, Y. Huang, C. Flynn, and M.A. Green, J. Appl. Phys. 103 (2008)
083544.
183. A. Gencer Imer, I. Yildiz, and R. Turan, Physica E 42 (2010) 2358–2363.
184. T. Arguirov, T. Mchedlidze, M. Kittler, R. Rölver, B. Berghoff, M. Först, and B. Spangenberg, Appl.
Phys. Lett. 89 (2006) 053111.
185. S. Hernández, P. Pellegrino, A. Martínez, Y. Lebour, B. Garrido, R. Spano, M. Cazzanelli et al., J. Appl.
Phys. 103 (2008) 064309.
186. Y. Gong, J. Lu, S.-L. Cheng, Y. Nishi, and J. Vukovi, Appl. Phys. Lett. 94 (2009) 013106.
187. S. Prezioso, S.M. Hossain, A. Anopchenko, L. Pavesi, M. Wang, G. Pucker, and P. Bellutti, Appl. Phys.
Lett. 94 (2009) 062108.
188. S. Gardelis, A.G. Nassiopoulou, P. Manousiadis, S. Milita, A. Gkanatsiou, N. Frangis, and C.B. Lioutas,
J. Appl. Phys. 111 (2012) 083536.
189. P. Basa, Zs.J. Horváth, T. Jászi, A.E. Pap, L. Dobos, B. Pécz, L. Tóth, and P. Szöllösi, Physica E 38
(2007) 71–75.
190. P.-J. Wu, Y.-C. Wang and I.-C. Chen, Nanoscale Res. Lett. 8 (2013) 457.
191. T. Baron, A. Fernandes, J.F. Damlencourt, B. De Salvo, F. Martin, F. Mazen, and S. Haukka, Appl. Phys.
Lett. 82 (2003) 4151.
192. L. Patrone, D. Nelson, V.I. Safarov, M. Sentis, W. Marine, and S. Giorgio, J. Appl. Phys. 87 (2000) 3829.
193. H.W. Lau, O.K. Tan, Y. Liu, C.Y. Ng, T.P. Chen, K. Pita, and D. Lu, J. Appl. Phys. 97 (2005) 104307.
194. H.W. Lau, O.K. Tan, Y. Liu, D.A. Trigg, and T.P. Chen, Nanotechnology 17 (2006) 4078.
195. H.W. Lau, O.K. Tan, B.C. Ooi, Y. Liu, T.P. Chen, and D. Lub, J. Cryst. Growth 288 (2006) 92.
196. U. Kahler and H. Hofmeister, Appl. Phys. Lett. 75 (1999) 5.
197. Y. Liu, T.P. Chen, C.Y. Ng, L. Ding, S. Zhang, Y.Q. Fu, and S. Fung, J. Phys. Chem. B 110 (2006)
16499–16502.
198. Y. Liu, Y.Q. Fu, T.P. Chen, M.S. Tse, S. Fung, J.-H. Hsieh, and X.H. Yang, Jpn. J. Appl. Phys. 42 (2003)
L1394–L1396.
199. Y. Liu, T.P. Chen, Y.Q. Fu, M.S. Tse, J.H. Hsieh, and P.F. Ho, J. Phys. D Appl. Phys. 36 (2003)
L97–L100.
200. J.D. Choi, J.H. Lee, W.H. Lee, K.S. Shin, Y.S. Yim, J.D. Lee, Y.C. Shin et al., IEEE IEDM Tech. (2000)
767–770.
201. J.D. Lee, J.H. Choi, D. Park, and K. Kim, IEEE Electron Dev. Lett. 24 (2003) 748.
202. K. Heinig, B. Schmidt, A. Markwitz, R. Grotzschel, M. Strobel, and S. Oswald, Nucl. Instrum. Methods
Phys. Res. B 148 (1999) 969–974.
203. N. Arai, H. Tsuji, H. Nakatsuka, K. Kojima, K. Adachi, H. Kotaki, T. Ishibashi, Y. Gotoh, and
J. Ishikawa, Mater. Sci. Eng. B 147 (2008) 230.
204. H. Fukuda, S. Sakuma, T. Yamada, S. Nomura, and M. Nishino, J. Appl. Phys. 90 (2001) 3524.
205. Y. Chen, G.Z. Ran, Y.K. Sun, Y.Q. Wang, J.S. Fu, and W. Chen, Nucl. Instrum. Methods Phys. Res. B
183 (2001) 305.
206. J. Zhang, X. Wu, and X. Bao, Appl. Phys. Lett. 71 (1997) 2505.
207. A. Markwitz, L. Rebohle, H. Hofmeister, and W. Skorupa, Nucl. Instrum. Methods Phys. Res. B 147
(1999) 361.
208. S. Duguay, S. Burignat, P. Kern, J.J. Grob, A. Souifi, and A. Slaoui, Semicond. Sci. Technol. 22 (2007)
837.
209. M. Klimenkov, J. Von Borany, W. Matz, R. Grotzschel, and F. Herrmann, J. Appl. Phys. 91 (2002)
10062.
210. L. Rebohle, J. Von Borany, H. Fröb, and W. Skorupa, Appl. Phys. B 71 (2000) 131.
211. J. Zhao, D. Huang, Z. Chen, W. Chu, B. Makarenkov, A. Jacobson, B. Bahrim, and J. Rabalais, J. Appl.
Phys. 103 (2008) 124304.
212. M. Nogami and Y. Abe, Appl. Phys. Lett. 65 (1994) 2545.
213. H. Yang, X.S. Wang, H. Shi, F. Wang, X. Gu, and X. Yao, J. Cryst. Growth 236 (2002) 371.
Synthesis and Characterization of Nanocrystals and Nanoparticles 37

214. H. Yang, R. Yang, X. Wan, and W. Wan, J. Cryst. Growth 261 (2004) 549.
215. Y. Shi, S.L. Gu, X.L. Yuan, Y.D. Zheng, K. Saito, H. Ishikuro, and T. Hiramoto, Fifth International
Conference on Solid-State and Integrated Circuit Technology, 1998, Beijing, China, pp. 838–841.
216. T. Baron, P. Gentile, N. Magnea, and P. Mur, Appl. Phys. Lett. 79 (2001) 1175.
217. M. Fujii, S. Hayashi, and K. Yamamoto, Jpn. J. Appl. Phys. 30 (2002) 687.
218. Y. Maeda, N. Tsukamoto, Y. Yazawa, and Y. Kanemitsu, Appl. Phys. Lett. 59 (1991) 3168.
219. W. Choi, W. Chim, C. Heng, L. Teo, V. Ho, V. Ng, D. Antoniadis, and E. Fitzgerald, Appl. Phys. Lett.
80 (2002) 2014.
220. W. Choi, V. Ho, V. Ng, Y. Ho, S. Ng, and W. Chim, Appl. Phys. Lett. 86 (2005) 143114.
221. W. Choi, Y. Ho, S. Ng, and V. Ng, J. Appl. Phys. 89 (2001) 2168.
222. W. Choi, H. Thio, S. Ng, V. Ng, and B. Cheong, Philos. Mag. B 80 (2000) 729.
223. W.K. Choi, V. Ng, S.P. Ng, H.H. Thio, Z.X. Shen, and W.S. Li, J. Appl. Phys. 86 (1999) 1398.
224. E. Kan, W. Chim, C. Lee, W. Choi, and T. Ng, Appl. Phys. Lett. 85 (2004) 2349.
225. E.W.H. Kan, W. Choi, C.C. Leoy, W.K. Chim, D.A. Antoniadis, and E.A. Fitzgerald, Appl. Phys. Lett.
83 (2003) 2058.
226. E.W.H. Kan, W.K. Choi, W.K. Chim, E.A. Fitzgerald, and D.A. Antoniadis, J. Appl. Phys. 95 (2004)
3148.
227. A. Rodríguez, M. Ortiz, J. Sangrador, T. Rodríguez, M. Avella, A. Prieto, Á. Torres, J. Jiménez,
A. Kling, and C. Ballesteros, Nanotechnology 18 (2007) 065702.
228. J. Wu and P. Li, Semicond. Sci. Technol. 22 (2006) S89.
229. T.C. Chang, S.T. Yan, P.T. Liu, C.W. Chen, S.H. Lin, and S.M. Sze, Electrochem. Solid-State Lett.
7 (2004) G17.
230. M. Avella, A. Prieto, J. Jimenez, A. Rodriguez, J. Sangrador, and T. Rodriguez, Solid State Commun.
136 (2005) 224.
231. P. Li, W. Liao, S. Lin, P. Chen, S. Lu, and M. Tsai, Appl. Phys. Lett. 83 (2003) 4628.
232. Y.C. King, T.J. King, and C. Hu, IEEE Trans. Electron Dev. 48 (2001) 696.
233. J. Zhang, Q. Fang, A. Kenyon, and I.W. Boyd, Appl. Surf. Sci. 208–209 (2003) 364.
234. J.G. Couillard and H.G. Craighead, J. Mater. Sci. 33 (1998) 5665.
235. C. Heng and T. Finstad, Phys. E Low-Dimensional Syst. Nanostruct. 26 (2005) 386.
236. T. Kobayashi, T. Endoh, H. Fukuda, S. Nomura, A. Sakai, and Y. Ueda, Appl. Phys. Lett. 71 (1997) 1195.
237. B. De Salvo, G. Ghibaudo, and G. Pananakakis, IEEE Trans. Electron Dev. 48 (2001) 1789.
238. J.K. Kim, H.J. Cheong, Y. Kim, J.Y. Yi, H.J. Bark, S.H. Bang, and J.H. Cho, Appl. Phys. Lett. 82 (2003)
2527.
239.Y. Kim, H.J. Cheong, K.H. Park, T.H. Chung, H.J. Bark, S.H. Bang, and J. Yi, Semicond. Sci. Technol.
17 (2002) 1039.
240. X. Ma, Z. Yan, B. Yuan, and B. Li, Nanotechnology 16 (2005) 832.
241. I. Berbezier, A. Karmous, A. Ronda, T. Stoica, L. Vescan, R. Geurt, A. Olzierski, E. Tsoi, and
A. Nassiopoulou, J. Phys. Conf. Ser. 10 (2005) 73.
242. A. Karmous, I. Berbezier, and A. Ronda, Phys. Rev. B 73 (2006) 5.
243. M.L. Ostraat, J.W. De Blauwe, M.L. Green, L.D. Bell, M.L. Brongersma, J. Casperson, R.C. Flagan, and
H.A. Atwater, Appl. Phys. Lett. 79 (2001) 433.
244. Y.Q. Wang, R. Smirani, and G.G. Ross, Nano Lett. 4 (2004) 2041.
245. P. Dimitrakis, E. Kapetanakis, D. Tsoukalas, D. Skarlatos, C. Bonafos, G.B. Asssayag, A. Claverie et al.,
Solid-State Electron. 48 (2004) 1511.
246. G.B. Assayag, C. Bonafos, M. Carrada, A. Claverie, P. Normand, and D. Tsoukalas, Appl. Phys. Lett.
82 (2003) 200.
247. Y.Q. Wang, R. Smirani, G.G. Ross, and F. Schiettekatte, Phys. Rev. B 71 (2005) 161310.
248. Y.H. Kwon, C.J. Park, W.C. Lee, D.J. Fu, Y. Shon, T.W. Kang, C.Y. Hong, H.Y. Cho, and K.L. Wang,
Appl. Phys. Lett. 80 (2002) 2502.
249. A.K. Dutta, Appl. Phys. Lett. 68 (1996) 1189.
250. J.Y. Zhang, X.M. Bao, and Y.H. Ye, Thin Solid Films 323 (1998) 68.
251. V. Mulloni, P. Bellutti, and L. Vanzetti, Surf. Sci. 585 (2005) 137.
252. D. Schmeiber, O. Bohme, A. Yfantis, T. Heller, D.R. Batchelor, I. Lundstrom, and A.L. Spetz, Phys. Rev.
Lett. 83 (1999) 380.
253. K. Borgohain, J.B. Singh, M.V. Rama Rao, T. Shripathi, and S. Mahamuni, Phys. Rev. B 61 (2000)
11093.
254. T.P. Chen, Y. Liu, C.Q. Sun, M.S. Tse, J.H. Hsieh, Y.Q. Fu, Y.C. Liu, and S. Fung, J. Phys. Chem. B 108
(2004) 16609.
38 Semiconductor Nanocrystals and Metal Nanoparticles

255. C.Q. Sun, L.K. Pan, Y.Q. Fu, B.K. Tay, and S. Li, J. Phys. Chem. B 107 (2003) 5113.
256. T. Baron, F. Martin, P. Mur, C. Wyon, and M. Dupuy, J. Cryst. Growth 209 (2000) 1004.
257. S. Yu, D. Li, H. Sun, H. Li, H. Yang, and G. Zou, J. Cryst. Growth 183 (1998) 284–288.
258. Y. Liu, T.P. Chen, P. Zhao, S. Zhang, S. Fung, and Y.Q. Fu, Appl. Phys. Lett. 87 (2005) 033112.
259. Y. Liu, T.P. Chen, H.W. Lau, J.I. Wong, L. Ding, S. Zhang, and S. Fung, Appl. Phys. Lett. 89 (2006)
123101.
260. Y. Liu, T.P. Chen, H.W. Lau, L. Ding, M. Yang, J.I. Wong, S. Zhang, and Y.B. Li, Appl. Phys. A 93 (2008)
483–487.
261. M. Yang, T.P. Chen, Y. Liu, L. Ding, J.I. Wong, Z. Liu, S. Zhang, W. Zhang, and F. Zhu, IEEE Trans.
Electron Dev. 55 (2008) 3605–3609.
262. Z. Liu, T.P. Chen, Y. Liu, L. Ding, M. Yang, J.I. Wong, Z.H. Cen, Y.B. Li, S. Zhang, and S. Fung, Appl.
Phys. Lett. 92 (2008) 013102.
263. Y. Liu, T.P. Chen, M. Yang, Z.H. Cen, X.B. Chen, Y.B. Li, and S. Fung, Appl. Phys. A 95 (2009)
753–756.
264. Y. Liu, T.P. Chen, L. Ding, Y.B. Li, S. Zhang, and S. Fung, J. Nanosci. Nanotechnol. 10 (2010)
5796–5799.
265. W. Zhu, T.P. Chen, M. Yang, Y. Liu, and S. Fung, IEEE Trans. Electron Dev. 56 (2009) 2060–2064.
266. W. Zhu, T.P. Chen, Z. Liu, M. Yang, Y. Liu, and S. Fung, J. Appl. Phys. 106 (2009) 093706.
267. Z. Liu, T.P. Chen, Y. Liu, Z.H. Cen, S. Zhu, M. Yang, J.I. Wong, Y.B. Li, and S. Zhang, IEEE Trans.
Electron Dev. 58 (2011) 33–38.
268. Z. Liu, T.P. Chen, Y. Liu, M. Yang, J.I. Wong, Z.H. Cen, S. Zhang, and Y.B. Li, Appl. Phys. Lett. 94
(2009) 243106.
269. W. Zhu, T.P. Chen, Y. Liu, and S. Fung, J. Appl. Phys. 112 (2012) 063706.
270. Y. Liu, T.P. Chen, W. Zhu, M. Yang, Z.H. Cen, J.I. Wong, Y.B. Li, S. Zhang, X.B. Chen, and S. Fung,
Appl. Phys. Lett. 93 (2008) 142106.
271. W. Zhu, T.P. Chen, M. Yang, Y. Liu, and S. Fung, IEEE Trans. Electron Dev. 59 (2012) 2363–2367.
272. Z. Liu, T.P. Chen, Y. Liu, M. Yang, J.I. Wong, Z.H. Cen, and S. Zhang, ECS Solid State Lett. 1 (2012)
Q4–Q7.
273. Z. Liu, T.P. Chen, Y. Liu, M. Yang, J.I. Wong, and Z.H. Cen, Appl. Phys. Lett. 96 (2010) 173110.
274. Y. Liu, T.P. Chen, L. Ding, M. Yang, Z. Liu, J.I. Wong, and S. Fung, J. Appl. Phys. 110 (2011) 096108.
275. W. Zhu, T.P. Chen, Y. Liu, M. Yang, and S. Fung, IEEE Trans. Electron Dev. 58 (2011) 960–965.
276. W. Zhu, Electrical properties and memory applications of Al-rich Al-based dielectric thin films,
PhD thesis, Nanyang Technological University, Singapore, 2012.
 2 Size- and Shape-Controlled
ZnO Nanostructures for
Multifunctional Devices
S.K. Ray, N. Gogurla, and T. Rakshit

CONTENTS
2.1 Introduction............................................................................................................................. 39
2.2 Growth of ZnO Nanostructures............................................................................................... 41
2.3 Optical Properties of ZnO Tripods and Tetrapods.................................................................. 41
2.3.1 Undoped ZnO Tripods................................................................................................. 42
2.3.2 Undoped ZnO Tetrapods.............................................................................................44
2.3.3 Sn- and P-Doped ZnO Tetrapods................................................................................. 45
2.3.4 Mg- and Mn-Doped ZnO Tetrapods............................................................................ 49
2.4 Electrical Properties of ZnO Tetrapods................................................................................... 49
2.4.1 Undoped ZnO Tetrapods............................................................................................. 49
2.4.2 Sn-Doped ZnO Tetrapods............................................................................................ 50
2.5 Field-Emission Properties of ZnO Tetrapods.......................................................................... 53
2.5.1 Sn-Doped ZnO Tetrapods............................................................................................ 54
2.5.2 Mg-Doped ZnO Tetrapods........................................................................................... 55
2.6 Shape- and Size-Dependent Properties of ZnO Nanostructures............................................. 56
2.7 ZnO Nanostructures for Device Applications......................................................................... 56
2.7.1 Ultraviolet (UV) Emitters............................................................................................ 56
2.7.2 UV Photodetectors....................................................................................................... 58
2.7.2.1 Plasmonic ZnO Photodetector...................................................................... 58
2.7.3 ZnO Nanostructures for Photovoltaic Devices............................................................ 63
2.7.3.1 CdS-Sensitized ZnO Nanorods..................................................................... 63
2.7.3.2 PbS-Sensitized ZnO Nanorods.....................................................................66
2.7.3.3 Cosensitized (CdS and CdSe) ZnO Nanowires............................................. 68
2.7.4 Gas Sensors Using ZnO Nanostructures..................................................................... 69
2.7.4.1 Gas Sensing Parameters................................................................................ 70
2.7.4.2 Surface-Modified ZnO Sensors.................................................................... 71
2.7.5 Biosensors Using ZnO Nanostructures........................................................................ 86
2.8 Summary................................................................................................................................. 87
Acknowledgment.............................................................................................................................. 88
References......................................................................................................................................... 88

2.1 INTRODUCTION
Zinc oxide (ZnO), a II–VI compound semiconductor, has been in the focus of current research due
to its size- and morphology-dependent electrical, optical, and chemical properties at the nanoscale.
The thermodynamically stable phase of ZnO is the wurtzite hexagonal one with the lattice param-
eters a = 0.3249 and c = 0.5207 nm. This structure belongs to the point group 6mm and space group
P63mc. Besides wurtzite ZnO, it can also exist in other crystalline phases, such as zinc blende (ZB)

39
40 Semiconductor Nanocrystals and Metal Nanoparticles

and rock salt. A few experimental and theoretical studies have addressed the growth and funda-
mental properties of metastable ZB ZnO [1]. However, ZnO is most stable in the wurtzite structure
rather than the ZB form under ambient conditions due to its ionicity, which lies at the border line
between those of covalent and ionic materials. Some important physical properties of the wurtzite
and ZB ZnO structures are summarized in Table 2.1. Due to the noncentrosymmetric structure in
the wurtzite form and large electromechanical coupling of ZnO, it also exhibits excellent piezo-
electric and pyroelectric properties. The high breakdown strength and high saturation velocity of
ZnO are attractive for electronic applications. With a wide direct bandgap of 3.34 eV at room tem-
perature, ZnO is useful for high-temperature applications, transparent electrodes, and light emit-
ting devices. It also has a large exciton binding energy of 60 meV, which is higher than the thermal
energy at room temperature (26 meV). Therefore, it is expected to have an excitonic gain at room
temperature for ZnO-based blue/UV light emitting diodes.
ZnO is intrinsically an n-type semiconductor due to the native defects, and it is difficult to make
p–n homojunctions due to low solubility of p-type dopant. Much effort has been made to prepare
p-type ZnO by doping with phosphorous, nitrogen, and arsenic [3–5]. However, achieving repro-
ducible, reliable, high-quality p-type conductivity is still a challenge. Group III elements (Al, Ga,
and In) have also been doped to enhance the n-type conductivity in ZnO [6,7]. Al- and In-doped
ZnO thin films and nanostructures are widely used as transparent conductors for photonic devices.
Various kinds of defects in ZnO such as zinc vacancy (VZn), oxygen vacancy (VO), zinc intersti-
tials (Zni), and oxygen antisites (OZn) play a vital role in controlling the optical, electrical, and mag-
netic properties. An excellent control over the surface defects is mandatory for realizing efficient
optoelectronic devices. Under ambient condition, ZnO exhibits low conductivity due to the captur-
ing of free electrons by oxygen molecules on its surface. This oxygen adsorption and desorption play
significant roles on the photo- and gas-sensing properties of ZnO.
ZnO can be prepared in different morphologies using low-temperature processing techniques.
The performance of ZnO devices depends on the structural quality, morphology, surface effects,

TABLE 2.1
Some Important Properties of Wurtzite and Zinc Blende ZnO Structures
Property Wurtzite Zinc Blende
Lattice Hexagonal Cubic
Point group 6mm 43m
Space group P63mc F 43m
Parameters a = 0.3249 nm and c = 0.5207 nm a = 0.447 nm
Density 5.606 g/cm3
Melting point 1975°C
Lattice expansion coefficients A0: 6.531026/°C
c0: 3.031026/°C
Thermal conductivity 69║, 60┴ W/m/K
Refractive index 2.008, 2029
Static dielectric constant 8.656
Energy gap 3.34 eV 3.27 eV
Intrinsic carrier concentration <106 cm–3
Electron Hall mobility at 300 K 200 cm2/V/s
Hole Hall mobility at 300 K 5–50 cm2/V/s
Electron and hole effective masses 0.24 and 0.59 m0

Sources: Pearton, S.J. et al., Superlattices Microstruct., 34, 3, 2003; Ashrafi, A. and Jagadish, C., J. Appl.
Phys., 102, 071101(1–12), 2007.
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 41

and the rate of adsorption/desorption of oxygen molecules. Therefore, the structural properties
and surface engineering can enhance the performance of devices. In this chapter, we mainly
focus on the tunable properties of shape- and size-controlled ZnO nanostructures for multifunc-
tional applications. The optical and electrical properties of these nanostructures have been stud-
ied. Enormous effort has been made to implement efficient electrical and optical devices using
different morphologies of ZnO. However, there exist several challenges for the practical use of
the grown nanostructures for commercial device applications. Recently, several sensitizers have
been used to improve the performance of ZnO-based devices. Here, we review the progress of
semiconductor- and metal-sensitized ZnO nanostructures for the realization of efficient optoelec-
tronic and gas-sensing devices.

2.2 GROWTH OF ZnO NANOSTRUCTURES

Several techniques have been reported for growing ZnO nanostructures. The electrochemical method
[8], template-based growth [9], and sol–gel processing [10,11] are some of the widely used methods
for growing different types of ZnO nanostructures. Another effective solution-based method is the
hydrothermal process [12], by which the nanostructures can be grown at relatively low tempera-
tures. ZnO nanostructures have also been grown by other methods such as electrospinning [13], two-
step mechanochemical–thermal process [14], chemical vapor deposition (CVD) [15], sonochemical
synthesis [16], microwave-assisted combustion [17], anodization [18], precipitation [19], and so on.
Another widely used method for growing ZnO nanostructures is the vapor transport process, in
which Zn and oxygen or oxygen mixture vapors react with each other to form ZnO. Zn and oxygen
vapors can be produced in several ways. The simplest one uses the decomposition of ZnO [20]; how-
ever, it requires a very high temperature (~1400°C). A relatively low temperature (500°C–800°C)
can be used for growing ZnO nanostructures by heating Zn powder in an oxygen atmosphere [21,22].
The growth temperature, pressure, duration of evaporation, and gas flow rate need to be precisely
controlled for achieving the desired morphology of ZnO nanostructures. The carbothermal method,
which uses graphite and ZnO powder as the source materials [23], can be used for the growth at about
800°C–1100°C. Metalorganic vapor phase epitaxy using an organometallic zinc compound under
appropriate flow of oxygen or nitrous oxide can also be used for the growth of ZnO nanostructures
[24]. The widely used vapor transport method is usually classified into catalyst-assisted vapor–
liquid–solid (VLS) and catalyst-free vapor–solid (VS) processes. Both the processes are capable of
producing a wide variety of ZnO morphologies [25,26]. Among all the materials, ZnO nanostruc-
tures probably possess the widest range of morphologies, such as nanorods, nanowires, nanobelts,
nanorings, nanosprings, nanopropellers, tripods, tetrapods, and so on [26,27]. Some of the reported
morphologies of ZnO nanostructures are shown in Figure 2.1.
The morphology of ZnO nanostructures can be tuned by doping with appropriate elements.
Doping ZnO tetrapods with a high concentration of Sn changes the morphology to flower-like
multipods and nanowires [28]. ZnO nanocrystals exhibit different morphologies upon doping
with different elements such as Na, Li, Pr, Cu, and Mg [29]. The morphology of S-doped ZnO
nanostructures changes from nanonails to nanowires by varying the sulfur concentration [30].
The optical properties of ZnO nanostructures are largely influenced by this change in morphol-
ogy. The near band–edge and defect state emissions depend significantly on the doping con-
centration of S and Sn [28,30]. The electrical conductivity and field emission properties of ZnO
nanostructures can also be modulated upon doping with Sn and Mg [28,31].

2.3 OPTICAL PROPERTIES OF ZnO TRIPODS AND TETRAPODS

Tripods and tetrapods, consisting of three and four rod-shaped arms emerging from a central core,
respectively, are the commonly exhibited morphologies of ZnO nanostructures [26,28,31–37]. They
have a higher specific surface area than one-dimensional ZnO nanostructures such as nanorods,
42 Semiconductor Nanocrystals and Metal Nanoparticles

FIGURE 2.1 Micrographs of reported ZnO nanostructures. (Reprinted from Mater. Today, 7, Wang, Z.L.,
Nanostructures of zinc oxide, 26–33, © 2004, with permission from Elsevier.)

nanowires, and so on. Another advantage of tetrapods over other morphologies is their ability to
spontaneously orient over the substrate with an arm directing perpendicular to the substrate plane,
which can be useful for photovoltaic applications [38,39]. The fundamental study of the intrinsic
optical properties of ZnO is important for their applications in photonic devices. For this, different
types of experimental techniques such as optical transmission, absorption, reflection, photoreflec-
tion, cathodoluminescence, photoluminescence (PL), spectroscopic ellipsometry, and so on, have
been used. In the following sections, we will discuss the optical properties of ZnO tripods and
tetrapods studied using photoluminescence spectroscopy.

2.3.1 Undoped ZnO Tripods

Figure 2.2 shows the room-temperature PL characteristics of ZnO tripods and tetrapods, syn-
thesized by the catalyst-free vapor–solid method [26]. The spectrum of bulk ZnO is also shown
for comparison. A typical spectrum consists of a peak in the ultraviolet (UV) region, which is
related to the free excitonic (FX) emission, and a peak in the visible region, which is due to
defect-state transitions [40,41]. The defect-state emission intensity of ZnO tripods and tetrapods
are much lower than that of bulk ZnO, with tripods having the lowest. The PL spectra indicate
the improved structural ordering in ZnO tripods. Figure 2.3a shows the temperature-dependent
PL spectra of ZnO tripods. The peak at 3.407 eV (bound exciton) is the most dominant up to 50 K,
whereas the one at 3.425 eV (free exciton) becomes stronger with further increase of temperature
and dominates up to room temperature. The increase of the FX peak intensity is due to the strong
coupling between the excitons and phonons. The high intensity of the bound excitonic (BX) peak
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 43

(a) ZnO tripods

(c)
(b) ZnO tetrapods
(c) bulk ZnO

Intensity (a.u.)
(b)

(a)

3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0

Energy (eV)

FIGURE 2.2 Room-temperature photoluminescence spectra of ZnO: (a) tripods, (b) tetrapods, and (c) bulk
powder. (From Mandal, S. et al., J. Appl. Phys., 105, 033513, 2009.)

at a temperature below 50 K suppresses the evolution of the FX peak. The quenching of the FX
emission with temperature can be described by the equation [42]

I0
I= (2.1)
1 + A1 exp(- Ea /kBT )

where
I0 is the emission intensity at T = 0 K
A1 is the proportionality constant
kB is the Boltzmann’s constant
Ea is the activation energy

The activation energy has been found to be 66 meV, which is near the FX binding energy of bulk ZnO
(60 meV). Figure 2.3b shows the PL emission characteristics of ZnO tripods at 10 K. Besides FX and
BX emissions, the longitudinal optical (LO) phonon replicas of FX emission, which are designated as
FX-1LO, FX-2LO, FX-3LO, FX-4LO, and FX-5LO, could be detected at the energies of 3.355, 3.281,
3.209, 3.135, and 3.067 eV, respectively. Similarly, the LO phonon replicas of BX transition, desig-
nated as BX-1LO, BX-2LO, and BX-3LO, are observed in Figure 2.3b. The presence of eight phonon
replicas at 10 K indicates the growth of ZnO tripods with very low defect density. The variation of the
FX peak energy with temperature can be described by a Bose–Einstein-type expression [43]:
K
Eg (T ) = Eg (0) - (2.2)
exp(qE /T ) - 1

where
Eg(0) is the bandgap energy at T = 0 K
K is the electron–phonon coupling strength
θE is the Einstein temperature

The Einstein temperature obtained from the fitted plot is 175 K, compared to 240 K reported for
bulk ZnO crystals [44].
44 Semiconductor Nanocrystals and Metal Nanoparticles

(A)10 K
(B)25 K
(C)50 K
(D)75 K
(E)100 K

Intensity (a.u.)
(F)150 K
(G)200 K
(H)250 K (A)
(B)
(C)
(D)
(E)
(F)
(G)
(H)

3.045 3.150 3.255 3.360 3.465 3.570

(a) Energy (eV)

BX
FX
10 K BX-1LO
BX-2LO
BX-3LO
Intensity (a.u.)

FX-1LO

FX-2LO

FX-3LO

FX-4LO

FX-5LO

3.045 3.150 3.255 3.360 3.465 3.570

(b) Energy (eV)

FIGURE 2.3 (a) Temperature-dependent PL spectra and (b) PL characteristics at 10 K of ZnO tripods.
(From Mandal, S. et al., J. Appl. Phys., 105, 033513, 2009.)

2.3.2 Undoped ZnO Tetrapods

Figure 2.4a shows the temperature-dependent PL spectra and Figure 2.4b displays the spectrum
recorded at 20 K for ZnO tetrapods grown by the catalyst-free vapor–solid method [32]. The stron-
gest peak at 20 K is found to be at the energy of 3.352 eV, which is related to the bound excitonic
emission. The peak at 3.362 eV arises from the FX emission. The binding energy of the bound exci-
tonic peak is found to be 10 meV, suggesting its origin as due to excitons bound to neutral donors
(D0X). Another strong peak (3.296 eV) at 20 K is related to the first-order transverse optical (TO)
phonon replica of D0X, and assigned as D0X-TO. The high peak intensity indicates a strong cou-
pling between D0X and TO phonons. The peaks observed at 3.226 and 3.153 eV are the LO phonon
replicas of D0X-TO emission, which are designated as D0X-TO-1LO and D0X-TO-2LO transitions,
respectively. The deformation potential mainly gives rise to TO phonon scattering, whereas the LO
scattering occurs because of Fröhlich and deformation potentials [45]. The short-range interaction is
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 45

20 K
50 K
100 K
150 K
200 K

Intensity (a.u.)

(a)

D0X

D0X-TO
20 K

D0X-TO-1LO FX
e A0
Intensity (a.u.)

D0X-TO-2LO
eA0-1LO

(b)

2.94 3.00 3.06 3.12 3.18 3.24 3.30 3.36 3.42 3.48
Energy (eV)

FIGURE 2.4 (a) Temperature-dependent PL spectra and (b) PL spectrum at 20 K of ZnO tetrapods.
(Reprinted from Ceram. Int., 41, Roy, N. and Roy, A., Growth and temperature dependent photoluminescence
characteristics of ZnO tetrapods, 4154–4160. © 2015, with permission from Elsevier.)

usually influenced by the deformation potential in ZnO nanostructures [46]. Thus, the TO phonons
are accompanied by a number of LO phonon replicas. The phonon peaks merge with other neigh-
boring ones or become very weak at a high temperature, and therefore they can be distinguished
only at temperatures below 150 K, as shown in Figure 2.4a. The temperature-dependent PL spectra
in the range 20–150 K indicate the thermal ionization of donors at high temperature. The free elec-
tron–acceptor (eA0) transition [40,47] results in the evolution of a new electron–acceptor recombi-
nation peak at 3.195 eV. With increase in temperature, the population of acceptor levels gradually
increases, resulting in the decreased intensity of the eA0 peak. Also the FX peak becomes stronger
with increasing temperature and dominates over the eA0 transition. The LO phonon replica of
eA0 peak is also observed at 20 K in Figure 2.4b.

2.3.3 Sn- and P-Doped ZnO Tetrapods

The effect of doping with Sn and P on the optical properties of ZnO tetrapods has been reported
[28,33]. The introduction of Sn (up to 5 at.%) and P reduces the intensity of defect-state emis-
sion in ZnO tetrapods [28,33]. This may be due to the reduction in intrinsic defect density in ZnO
46 Semiconductor Nanocrystals and Metal Nanoparticles

0
SX D 1X D02X
FX
SX-1LO

SX-2LO
SX-3LO

Intensity (a.u.)
10 K
25 K
40 K
55 K
70 K
85 K
100 K
125 K
150 K
175 K
200 K
225 K
250 K
275 K
300 K

3.08 3.15 3.22 3.29 3.36 3.43

(a) Energy (eV)

D 03 X 0
D 1X
SX
D02X
SX-1LO FX

SX-2LO
SX-3LO
Intensity (a.u.)

10 K
25 K
40 K
55 K
70 K
85 K
100 K
125 K
150 K
175 K
200 K
225 K
250 K
275 K
300 K

3.08 3.15 3.22 3.29 3.36 3.43

(b) Energy (eV)

D03X D01X
SX D02X
SX-1LO 0
SX-2LO D 1X-1LO FX
SX-3LO D01X-2LO
D01X-3LO
Intensity (a.u.)

10 K
25 K
40 K
55 K
70 K
85 K
100 K
125 K
150 K
175 K
200 K
225 K

250 K

275 K
300 K

3.08 3.15 3.22 3.29 3.36 3.43

(c) Energy (eV)

FIGURE 2.5 Temperature-dependent PL spectra of (a) pure ZnO, (b) 3 at.% Sn-doped, and (c) 5 at.% Sn-doped
ZnO tetrapods. The spectrum at 300 K (solid line) has been fitted with Lorentzian curves (dash lines). The open
circles denote the sum of the three Lorentzian curves. (From Rakshit, T. et al., AIP Adv., 3, 112112, 2013.)
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 47

due to doping. The temperature-dependent excitonic PL emission bands of undoped ZnO, 3 at.%
Sn-doped ZnO, and 5 at.% Sn-doped ZnO tetrapods are shown in Figure 2.5a through c, respec-
tively. At 10 K, both undoped and Sn-doped ZnO samples have some common features [28]. The
peak at 3.375 eV is related to the FX emission, and those at 3.360 and 3.364 eV are assigned to exci-
tons bound to neutral donors, designated by D01X and D02X, respectively [41,48]. There is a transi-
tion at 3.311 eV in both pure and Sn-doped ZnO tetrapods. The binding energy of the peak (64 meV)
indicates its excitonic nature. Therefore, excitons bound to defect states (SX) could be considered
to be the cause of this emission [49,50]. The PL features observed at 3.238, 3.166, and 3.094 eV
are the LO phonon replicas of SX, denoted by SX-1LO, SX-2LO, and SX-3LO, respectively. An
additional peak is observed at 3.355 eV in Sn-doped ZnO tetrapods, which can be assigned to
excitons bound to neutral donors, designated by D03X. This emission may have originated from the
deep-level energy states formed in ZnO due to Sn doping. Some additional peaks could be observed
for 5 at.% Sn-doped ZnO samples. These peaks at 3.288, 3.216, and 3.144 eV are the LO phonon
replicas of D01X, and denoted by D01X-1LO, D01X-2LO, and D01X-3LO, respectively. In all these
samples, the D0X peak disappears at a temperature above 125 K. As the temperature rises, the
D0X peak thermally dissociates into free excitons, resulting in a decrease in the intensity of D0X
emission. The spectrum at 300 K of all the samples, fitted with a Lorentzian function, exhibited
deconvoluted peaks corresponding to FX, SX, and SX-1LO emissions. The variation of the FX peak
energy with temperature of the pure and Sn-doped ZnO tetrapods could be fitted with Equation 2.2.
The Einstein temperature from the plot is found to be 259 ± 3, 240 ± 2, and 179 ± 5 K, for pure
ZnO, 3 at.% Sn-doped ZnO, and 5 at.% Sn-doped ZnO tetrapods, respectively. In comparison, the
Einstein temperature is 175 K for pure ZnO tripods and 240 K for ZnO crystals.
The PL spectra of ZnO nanotetrapods grown by chemical vapor deposition revealed the presence
of only neutral-donor-bound and FX emissions [33]. The PL spectra recorded with various excita-
tion intensities at 10 K for P-doped ZnO nanotetrapods are shown in Figure 2.6a. The peaks at 3.313
and 3.355 eV are related to the free electrons to neutral-acceptor (FA) and neutral-acceptor-bound
excitonic (A0X) transitions, respectively. Both the peaks are invariant to the change in excitation
intensity. With the increase of excitation intensity from 0.3 to 30 mW/cm2, the peak at 3.287 eV
exhibits a blue shift of ~10 meV. The close donor–acceptor pair (DAP) transitions undergo faster
recombination, and therefore the Coulomb interaction between them is much stronger than that of
the distant pairs. This results in a blue shift in DAP transitions with increase in excitation intensity.
Hence, the peak at 3.287 eV is attributed to the DAP emission. The weak peak at 3.212 eV is related
to the LO phonon replica of DAP transition, and denoted by DAP-1LO. The radiative recombination
for DAP transitions results in a photon energy given by [51]

e2
hn DAP = Eg - E D - E A + (2.3)
4per

where
Eg is the bandgap energy
ED and EA are the donor and acceptor binding energy, respectively
e is the electronic charge
ε is the dielectric permittivity
r is the DAP distance

Using Equation 2.3, the acceptor binding energy was calculated to be 90–120 meV. Figure 2.6b
shows the temperature-dependent PL spectra of P-doped ZnO nanotetrapods. The A0X and DAP
emission undergoes red and blue shifts, respectively, with the increase of temperature. The intensity
of DAP emission shows a rapid reduction with temperature. There is no shift in the FA peak at tem-
peratures below 30 K, but undergoes a red shift with further increase of temperature. Such a shift in
48 Semiconductor Nanocrystals and Metal Nanoparticles

10 K A0X
0
10 K AX

PL intensity (a.u.)
FA
PL intensity (a.u.) DAP
DAP-LO

3.10 3.15 3.20 3.25 3.30 3.35 FA

Energy (eV)
DAP

DAP-LO

3.15 3.24 3.33

(a) Energy (eV)

A0X T (K)

10 K
20 K
30 K
40 K
PL intensity (a.u.)

50 K
60 K
80 K
FX 100 K
FA 120 K
DAP 150 K
180 K
DAP-LO 200 K

3.15 3.20 3.25 3.30 3.35

(b) Energy (eV)

FIGURE 2.6 (a) Excitation intensity–dependent PL spectra recorded at 10 K and (b) temperature-dependent
PL spectra of P-doped ZnO nanotetrapods. The inset in (a) shows the near-band edge emission of P-doped
ZnO nanotetrapods. (Reprinted with permission from Yu, D. et al., Photoluminescence study of novel phos-
phorus-doped ZnO nanotetrapods synthesized by chemical vapour deposition, J. Phys. D: Appl. Phys., 42,
055110, 2009. © 2009 by the IOP Publishing.)

the FA, DAP, and A0X peaks implies that a fraction of the doped phosphorus atoms are acting like
acceptors in ZnO nanotetrapods. The acceptors gradually ionize with the increase of temperature,
and thus the peaks disappear at high temperatures. Using Equation 2.1, the binding energy between
free excitons and acceptor was calculated to be 11.5 meV.
From the above discussion, it is apparent that doping plays a significant role in governing the
optical properties of ZnO tetrapods. Doping with Sn in ZnO tetrapods resulted in the formation of
a new peak due to excitons bound to neutral donors. Also, the LO phonon replicas of the emission
related to excitons bound to neutral donors were observed only in 5 at.% Sn-doped ZnO tetrapods.
When ZnO tetrapods were doped with P, the neutral-donor-bound excitonic peaks disappeared, and
several new PL peaks evolved such as free electrons to neutral-acceptor, neutral-acceptor-bound
excitonic transitions, as well as donor–acceptor pair emission and its LO phonon replica.
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 49

2.3.4 Mg- and Mn-Doped ZnO Tetrapods

The introduction of Mg and Mn dopants also significantly changes the optical properties of ZnO
tetrapods [34,35]. The near-band edge (NBE) emission underwent a blue shift of about 5 nm upon
adding Mg dopant in ZnO tetrapods [34]. However, the blue shift in the NBE peak was not mono-
tonic. Since MgO has a higher bandgap (6.7 eV) than ZnO, the addition of Mg to ZnO is expected
to increase the bandgap. But, at the same time, increasing the Mg fraction in ZnO changed the
dimension of the tetrapods, resulting in a lower quantum confinement effect and decrease of the
bandgap. Thus, no monotonic blue shift could be observed in the NBE transition with Mg addition.
The emission related to the native shallow donors decreased with the increase in Mg concentra-
tion. Such an emission peak is due to the transition between the zinc interstitial energy level and
the valence band [52]. With the increase of Mg mole fraction, the concentration of Zn interstitial
decreased, resulting in the reduction of the emission due to the native shallow donors. However,
Mn doping in ZnO tetrapods resulted in the reduction of the ratio of UV to defect-state emission,
without altering the peak position [35].

2.4 ELECTRICAL PROPERTIES OF ZnO TETRAPODS

The electrical properties of ZnO have been widely investigated for applications in nanoelectron-
ics. ZnO exhibits n-type semiconducting behavior due to the presence of native defects such
as zinc interstitials and oxygen vacancies. In this section, we discuss the study of the electri-
cal properties of pure and doped ZnO tetrapods using complex-plane impedance spectroscopy.
Impedance spectroscopy provides a correlation between the microstructural and electrical trans-
port properties. It is possible to separate the resistive and capacitive properties associated with
different parts of a material by relating the individual components to various relaxation times
(τ). The relaxation time, defined as τ = RC, is a unique and important property of the material
and does not change with the geometry of the sample. It is possible to obtain the time con-
stant associated with the relaxation processes taking place within the material by analyzing the
impedance spectrum.

2.4.1 Undoped ZnO Tetrapods

In order to find out the individual contribution of the arm and junction of a tetrapod to the overall
electrical conduction, Huh et al. [36] carried out impedance measurements on a single ZnO tetra-
pod. Figure 2.7a and b shows the scanning electron microscopy (SEM) images of the fabricated
devices (designated as devices A and B) used for electrical measurements. Terminals I, II, and
III are the Ti/Au electrodes, which were deposited at different positions in devices A and B, such
that the DC resistance measured between any pair of terminals in device A (RdcA ) consists of the
resistance due to the arm (R A) and the junction (R J) (i.e., RdcA @ RA + RJ ). On the other hand, the
measurement between terminals I and II in device B (RdcB ) provides the resistance due to the arm
(i.e., RdcB @ RA) only. Figure 2.7c and d shows the complex-plane plots measured at 380°C in the
frequency range 10 −1–107 Hz in devices A and B, respectively, using the terminals I and II in both
cases. The corresponding equivalent circuits are shown in the inset of the figures. The impedance
spectrum obtained from device A (Figure 2.7c) consists of two partially overlapping semicircles
appearing in the lower and higher frequency regions. These are due to the arm and junction of a
single tetrapod, which are structurally and electrically two different regions. On the other hand,
the impedance spectrum obtained from device B, shown in Figure 2.7d, comprises only a single
semicircle, corresponding to the arm of a tetrapod. The activation energy of R A (extracted from
the impedance spectra of Figure 2.7d) was found to be 0.95 eV. On the other hand, the activa-
tion energy of the resistance corresponding to the semicircles in the higher and lower frequency
regions (Figure 2.7c) was 0.9 and 0.73 eV, respectively. The activation energy values indicate that
50 Semiconductor Nanocrystals and Metal Nanoparticles

I
I

II
III
II
III
(a) (b)

4 × 106 3 × 101
380°C 380°C
I–II I–II 2 × 101
–Z'' (Ω)

2 × 106 Air Air

1 × 101
H2
0 0
0 2 × 106 4 × 106 6 × 106 0 2 × 101 4 × 101
(c) Z'(Ω) (d) Z'(Ω)

FIGURE 2.7 SEM micrographs of the fabricated single tetrapod devices for the measurement of dc resistance
at (a) arm and junction (device A) and (b) arm (device B), and complex-plane impedance spectra m ­ easured at
380°C under air (solid squares) in (c) device A and (d) device B. The impedance spectrum measured at 380°C
under hydrogen (open squares) is also shown in part (c). The solid curves in parts (c) and (d) are the corre-
sponding fit. The corresponding equivalent circuits are shown in the insets (c) and (d). The scale bar shown in
parts (a) and (b) corresponds to 1.0 μm. (Reprinted with permission from Huh, J., Kim, G.-T., Lee, J.S., and
Kim, S., Appl. Phys. Lett., 93, 042111, 2008. © 2008, American Institute of Physics.)

the semicircles in the higher and lower frequency regions of Figure 2.7c correspond to those of the
arm and the junction of the tetrapod, respectively. Thus, the individual contribution of the arm and
junction of a tetrapod to the overall electrical conductivity could be extracted, which may be use-
ful for ­several electronic devices. Moreover, the resistance of the junction is higher than the arm,
indicating the key role played by the junction in controlling the electrical property of a tetrapod, in
spite of its smaller volume fraction compared to the arm.

2.4.2 Sn-Doped ZnO Tetrapods

Impedance measurements have also been carried on Sn-doped ZnO tetrapods [28]. Unlike the
work reported on a single ZnO tetrapod [36], the study on Sn-doped ZnO [28] considered a high
density of tetrapods grown on SiO2/Si substrates. Al deposited over the tetrapods was used as
the electrodes. The insulating SiO2 layer between the tetrapods and the Si substrate eliminated
the contribution of the substrate in the electrical measurement. In high-density tetrapods, there
exists a connection between the arm of the tetrapod and the arm or junction of the neighboring
one. Conduction occurs between the two Al electrodes through these interconnected tetrapods.
Figure 2.8a through c shows the impedance spectra recorded for pure ZnO, 3 at.% Sn-doped
ZnO, and 5 at.% Sn-doped ZnO tetrapods, respectively. The measurements were carried out at
room temperature with the DC bias voltage varying from 0 to 1.5 V. In a typical plot, two over-
lapping semicircles are observed within the measured range: one in the low-frequency region
(denoted by semicircle 1) and another relatively small semicircle in the high-frequency region
(denoted by semicircle-2). The semicircle 2 is shown separately in the inset of the figures.
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 51

The semicircles could be attributed to the arm and junction of the tetrapods. The equivalent
circuit, shown in Figure 2.8d, results in an equivalent impedance, given by

-1 -1
æ 1 ö æ 1 ö
Z = RS + ç + i wC A ÷ + ç + i wC J ÷ (2.4)
è RA ø è RJ ø

where
RS is the resistor accounting for the shift occurring along the real Z-axis in the high-frequency
region
R A, CA are the resistance and capacitance due to the arms, respectively
RJ, CJ are the resistance and capacitance due to the junctions of the tetrapods, respectively

(A) 0V
5.0 × 103
(B) 0.5 V
(A) (C) 1V
4.0 × 103 (D) 1.5 V
(B)
(C)
Semicircle-1
3.0 × 103
–Im Z (Ω)

(D)

2.0 × 103

1.0 × 103

0 3.0 × 103 6.0 × 103 9.0 × 103 1.2 × 104

(a) Re Z (Ω)

(A) 0V
2.5 × 104 0.5 V
(B)
(C) 1V
(D) 1.5 V
2.0 × 104
(A)

1.5 × 104 (B)

–Im Z (Ω)

(C)
1.0 × 104 Semicircle-1 (D)

5.0 × 103

0 2.0 × 104 4.0 × 104 6.0 × 104 8.0 × 104

(b) Re Z (Ω)

FIGURE 2.8 Complex-plane impedance spectra for (a) pure ZnO, (b) 3 at.% Sn-doped. (From Rakshit, T.
et al., AIP Adv., 3, 112112, 2013.) (Continued)
52 Semiconductor Nanocrystals and Metal Nanoparticles

5.0 × 104
(A) 0V
(B) 0.5 V
4.0 × 104 (C) 1V
(D) 1.5 V
(A)
–Im Z (Ω) 3.0 × 104

2.0 × 104
(B)

1.0 × 104 (C)

Semicircle-1 (D)
0

0 3.0 × 103 6.0 × 103 9.0 × 103 1.2 × 104

(c) Re Z (Ω)

CA CJ

RS

RA RJ
(d)

FIGURE 2.8 (Continued) Complex-plane impedance spectra for (c) 5 at.% Sn-doped ZnO tetrapods. The
semicircle in the higher frequency region is shown in the inset of the figures. The experimental data are
denoted by open symbols and corresponding fit by the solid curves. (d) Equivalent circuit of the device. (From
Rakshit, T. et al., AIP Adv., 3, 112112, 2013.)

The impedance spectra were fitted with Equation 2.4, and the best fit is shown by the solid curves
in Figure 2.8a through c. The resistance extracted from semicircle 1 and semicircle 2, shown in
Figure 2.9a and b, respectively, is higher in Sn-doped ZnO tetrapods as compared to undoped ZnO.
As the DC bias increases, the resistance decreases for both semicircle 1 and 2. However, the change
of resistance with DC bias of semicircle 2 is much less than for semicircle 1. For semicircle 1, the
resistance of 5 at.% Sn-doped ZnO tetrapods is lower than that of 3 at.% Sn-doped ZnO samples
from 0.5 V onward. However, the resistance value increases uniformly with increase in Sn doping
for semicircle 2. When a bias is applied, carriers are injected into the tetrapods, which find a con-
ducting path due to the defects (such as oxygen vacancies) present in the arms and junctions, thereby
reducing the resistance. The defect density is generally more in junctions than in the solid arms. So,
the change of resistance of the arms is expected to be less than that of the junction. Therefore, semi-
circle 1 and semicircle 2 correspond to the junction and arms of tetrapods, respectively. In 3 at.%
Sn-doped ZnO tetrapods, the dominant carrier concentration is from defect states. When the doping
concentration is increased to 5 at.%, and a high bias is applied, the injected carriers dominate over
the defect-state contribution. This results in a decrease of the resistance of the junction of 5 at.%
Sn-doped ZnO tetrapods compared to 3 at.% Sn-doped ZnO tetrapods. On the other hand, tetrapod
arms have much lower defect density, so the conductivity is not much affected by the injected car-
riers. Therefore, an increase in the resistance of the arms with Sn doping is observed at all bias
voltages. The relaxation time of Sn-doped ZnO tetrapods is higher than that of pure ZnO. Thus, by
varying the concentration of Sn doping in ZnO tetrapods, both the conductivity and relaxation time
could be tuned.
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 53

1.2 × 105

Pure ZnO tetrapods

1.0 × 105 3 at.% Sn-doped ZnO
5 at.% Sn-doped ZnO
8.0 × 104
Resistance (Ω)

6.0 × 104

4.0 × 104

2.0 × 104

0.0 0.5 1.0 1.5

(a) Voltage (V)

Pure ZnO tetrapods

3 at.% Sn-doped ZnO
1.0 × 103
5 at.% Sn-doped ZnO

8.0 × 102
Resistance (Ω)

6.0 × 102

4.0 × 102

2.0 × 102

0.0 0.5 1.0 1.5

(b) Voltage (V)

FIGURE 2.9 Variation of resistance extracted from semicircles (a) in the low frequency region (semicircle-1)
and (b) in the high frequency region (semicircle-2) with dc bias voltage. (From Rakshit, T. et al., AIP Adv.,
3, 112112, 2013.)

2.5 FIELD-EMISSION PROPERTIES OF ZnO TETRAPODS

Field emission (FE) is strongly dependent on a number of factors of the emitter such as geom-
etry, conductivity, work function, and so on. ZnO has a low work function, is highly stable
both thermally and mechanically, and is extremely resistant to high-energy radiation. These
properties make ZnO nanostructures a potential candidate for FE devices. High FE current
density can be obtained at a relatively low electric field in ZnO nanotetrapods with high aspect
54 Semiconductor Nanocrystals and Metal Nanoparticles

ratio [53,54]. When ZnO tetrapods are used as field emitters, highly efficient electron emission
can be obtained from the four arms, with an arm always directed perpendicular to the anode,
resulting in a lower turn-on field [55]. Doping with selective elements such as Sn, Mg, and so
on, serves as an effective way in tuning the electrical properties and hence to improve the FE
performance [31,37].

2.5.1 Sn-Doped ZnO Tetrapods

The FE properties of Sn-doped ZnO tetrapods were evaluated by preparing cathodes with the
tetrapods [37]. Such tetrapods were synthesized by thermal evaporation, with different concentra-
tions of Sn dopant in the Zn and Sn mixture. The fraction of Sn was 0, 1, 3, 5, and 10 at.%, which
are denoted as Sn0, Sn1, Sn3, Sn5, and Sn10 samples, respectively, hereafter. Tetrapods with low
concentration of Sn doping (Sn1–Sn5) exhibited high aspect ratio with longer and thinner arms.
The variation of current density as a function of applied electric field is shown in Figure 2.10a.
The turn-on field is 3.91, 1.95, 1.96, 2.27, and 3.84 V/µm, for undoped ZnO, Sn1, Sn3, Sn5, and
Sn10 sample, respectively. The threshold field is found to be ~4.58 and 5.35 V/µm for Sn1 and
Sn3 samples, respectively. The highest current density of ~1.13 mA/cm 2 was obtained for 1 at.%
Sn:ZnO at an applied electric field of 4.71 V/µm. Thus, low concentration of Sn doping in ZnO
tetrapods shows significant improvement in the FE performance. The FE characteristics can be
analyzed using the Fowler–Nordheim (F–N) equation, which is given by [53]

æ b2 E 2 ö æ - Bj3 / 2 ö
J = Aç ÷ exp ç ÷ (2.5)
è j ø è bE ø

So,

æ J ö æ Ab2 ö æ - Bj3 / 2 ö 1
ln ç 2 ÷ = ln ç ÷+ç ÷ (2.6)
èE ø è j ø è b øE

where
J and E are the current density (A/m2) and the applied field, respectively
coefficients A = 1.54 × 10 −10 (A/V2 eV) and B = 6.83 × 109 (V/m/eV3/2)
φ is the work function of the emitter material (~5.4 eV for ZnO)
β is the field-enhancement factor [37], which is a measure of the ability of the emitter to increase
the local electric field

The field-enhancement factor strongly depends on the geometry, crystal structure, density of
nanostructures, work function, and conductivity [31]. The F–N plots for Sn-doped ZnO devices
are shown in Figure 2.10b, demonstrating that the characteristics are linearly dependent at high
applied fields. The field-enhancement factor calculated from the F–N plots for Sn0, Sn1, Sn3, Sn5,
and Sn10 was 1919, 9556, 9256, 8938, and 3614, respectively. Thus, this factor is much higher
in lightly doped ZnO sample (Sn1) compared to undoped and heavily doped ones. Sample Sn1
also showed a stable emission current density over a period of 200 min under an applied field of
~4 V/µm (Figure 2.10c). The superior FE performance of the lightly Sn-doped ZnO sample can be
attributed to the high electron density in the conduction band caused by the doping and high aspect
ratio of the tetrapods.
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 55

–4 Sn0
1.0
Sn1
Sn3
Current density (mA/cm2)

0.8 Sn5
–6 Sn10
0.6

Ln (J/E2)
Sn0
Sn1 –8
0.4 Sn3
Sn5
0.2 Sn10
–10

0.0
–12
0 1 2 3 4 5 6 0.0 0.2 0.4 0.6 0.8 1.0 1.2
(a) Electric field (V/μm) (b) 1/E (μm/V)

1.0

Sn1
0.8
Current density (mA/cm2)

0.6

0.4

0.2

0.0
0 50 100 150 200
(c) Time (min)

FIGURE 2.10 (a) Variation of the field emission current density as a function of applied electric field.
The field emission micrograph of 1 at.% Sn:ZnO (Sn1) is shown in the inset. (b) Fowler–Nordheim (F–N) plot
of the corresponding J–E curves. (c) Variation of field emission current density as a function of time for Sn1
sample under an applied field of ~4 V/µm. (Reprinted with permission from Zhou, X., Lin, T., Liu, Y., Wu, C.,
Zeng, X., Jiang, D., Zhang, Y.-A., and Guo, T., ACS Appl. Mater. Interfaces, 5, 10067–10073, 2013. © 2013,
American Chemical Society.)

2.5.2 Mg-Doped ZnO Tetrapods

The FE performance of Mg-doped ZnO tetrapods, grown by thermal evaporation, has also been
reported [31]. Two different types of samples were analyzed: one with higher Mg concentration
but with thicker and shorter tetrapod arms compared to the other. The turn-on field for the higher
and lower Mg atomic fraction was 2.6 and 2.8 V/µm, respectively. The corresponding current
density for the samples reached 0.76 and 0.58 mA/cm2, respectively, at an applied electric field
of 6.0 V/µm. Using Equation 2.6, the field-enhancement factor was found to be 2327 and 1690
for higher and lower Mg fraction, respectively. The results indicate that the FE performance of
the doped ZnO tetrapods with higher Mg content is better with higher current density. This could
be attributed to the higher conductivity for higher concentration of dopants. The Mg-doped ZnO
samples exhibited superior FE performance compared to the undoped ones due to the enhanced
conductivity induced by Mg doping.
56 Semiconductor Nanocrystals and Metal Nanoparticles

2.6 SHAPE- AND SIZE-DEPENDENT PROPERTIES OF ZnO NANOSTRUCTURES

The morphology of ZnO nanostructures largely controls their optical properties. Different exci-
tonic emissions have been observed from PL measurements in ZnO nanostructures. The NBE
emission of ZnO nanoneedles and nanonails at low temperatures consisted of several well-
resolved peaks, whereas well-resolved phonon peaks were absent in nanorods having rounded
tips [56]. The excitonic emissions together with their phonon replicas contributed differently to
the room temperature PL of ZnO nanorods and nanopencils, resulting in a red shift of 52 meV in
the NBE emission of nanopencils compared to the nanorods [57]. This was related to the differ-
ent exciton–phonon interaction, which depends on the surface defects. ZnO nanopencils, with a
large surface to volume ratio, exhibited strong exciton–phonon interactions leading to different
excitonic emissions from ZnO nanorods. Similar behavior was observed in the excitonic emis-
sion from ZnO nanosheets and nanowires [58]. The enhanced exciton–phonon interactions in
nanosheets resulted in a ~40 meV red shift of the NBE emission at room temperature compared
to the nanowires. The optical properties are also influenced by the size of ZnO nanostructures.
Room-temperature PL measurements on ultrathin ZnO nanobelts of 6 nm width exhibited a 120
meV blue shift in the NBE emission compared to 200 nm wide nanobelts [59]. A blue shift in the
NBE emission was also observed with decrease in the size of ZnO quantum dots due to quantum
confinement effect [60].
The shape and size of ZnO nanostructures also affect their electrical properties. ZnO nanow-
ires and nanorods were utilized for fabricating field-effect transistors (FETs) [61,62]; the electri-
cal transport properties such as carrier concentration and mobility of ZnO nanorods were almost
independent of the size of the nanorods [62]. The electric FE performance of ZnO nanowires,
nanoneedles, tetrapods, and so on. has been extensively studied [55,63,64]. At a current density of
0.1 µA/cm2, the turn-on field was 6 V/µm for ZnO nanowires [63]. An improved FE performance
was observed with ZnO tetrapods, in which a low turn-on field of 1.6 V/µm was found at a current
density of 1 µA/cm2 [55]. The higher aspect ratio of the tetrapods compared to nanowires could be
the reason for the enhanced performance.

2.7 ZnO NANOSTRUCTURES FOR DEVICE APPLICATIONS

Over the years, an enormous amount of research has been carried out on the synthesis and applica-
tion of ZnO nanostructures, motivated by their outstanding electrical and optical properties. ZnO
nanostructures have great potential for a variety of functional applications including photovoltaics
[65], photodetectors [66], light-emitting diodes (LEDs) [67], transparent conducting oxides [68],
optical waveguides [69], field emitters [70], piezoelectric transducers [71], spintronic devices, gas
sensors [72], and so on. Some of the device characteristics are discussed here.

2.7.1 Ultraviolet (UV) Emitters

The wide and direct bandgap of ZnO makes it a promising candidate for UV light emitters. Due
to the large exciton-binding energy of ZnO (60 meV), it has emerged as a potential candidate to
replace GaN (25 meV). However, because of the lack of high-quality and reliable p-type ZnO, the
homojunction ZnO UV emitters have rarely been investigated. Yang et al. have reported the elec-
troluminescence (EL) from p–n junction ZnO nanowires fabricated by As+ ion implantation [73].
The schematic of an As-doped ZnO nanowire diode p–n junction is shown in Figure 2.11a. The
EL spectra of the ZnO device with different forward injection currents are shown in Figure 2.11b.
An intense, narrow emission in the UV range is clearly observed. The inset of Figure 2.11b shows
the injection current–dependent EL intensity. The ZnO homojunction device exhibited enhanced
luminescence and lasing performance in the UV region at room temperature. A Fabry–Perot-type
ZnO UV laser using n-type ZnO film and p-type Sb-doped ZnO nanowires has been reported [74].
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 57

~20 nm Au layer

p-Zno:As

PMMA
Aligned
nanowires
ZnO seed
layer

~300 nm Au
layer Si (111) n substrate
(a)

AS+: 1014 cm–2

105
50 keV
Intensity (a.u.)

1.0 × 105
EL intensity (a.u.)

104

5.0 × 104
8 V, 9.2 mA UV peak
7 V, 7.1 mA
6 V, 5.2 mA 3 4 5 6 7 8 9 10
5 V, 3.9 mA Current (mA)

0.0
300 400 500 600 700 800 900

(b) Wavelength (nm)

FIGURE 2.11 (a) Schematic illustration of the p–n junction ZnO nanowire LED structure. (b) EL spectra
of the p–n ZnO device with different forward injecting currents. Inset shows the EL intensity as a function of
injecting current. (Reprinted with permission from Yang, Y., Sun, X.W., Tay, B.K., You, G.F., Tan, S.T., and
Teo, K.L., Appl. Phys. Lett., 93, 253107(1–3). © 2008, American Institute of Physics.)

The study demonstrated the potential use of ZnO for diode lasers with lower cost and higher power
as compared to GaN. However, efforts are still on to fabricate homojunction ZnO devices with
improved performance. The formation of ZnO heterojunctions with several p-type semiconductors
such as Si, SiC, GaN, AlGaN and NiO, and so on, has been investigated. Due to the well-matched
lattice constant and thermal expansion coefficient of ZnO and GaN, ZnO-based heterojunction pho-
tonic devices have been fabricated on p-type GaN or AlGaN. Coaxial n-GaN/ZnO nanorods on
p-GaN by metal-organic vapor phase epitaxy have been fabricated to achieve intense EL at near-UV
wavelengths [75]. LED devices using ordered and aligned ZnO nanorods on p-GaN have shown
strong UV emission at 390 nm [76]. The ZnO–MgO alloy resulted in UV emitters with broadened
emission wavelength. Bandgap tuning of ZnO by doping with several transition metals such as
Co, Mn, and Ni could be done to fabricate UV emitters over a broad spectral range.
58 Semiconductor Nanocrystals and Metal Nanoparticles

2.7.2 UV Photodetectors
UV detectors based on ZnO are attractive owing to their simple fabrication process, high on/off cur-
rent ratio, visible-light blindness, and potential for flexible electronics [77,78]. The first individual
ZnO nanostructure-based detector was reported in 2002 [79]. The device showed a high sensitiv-
ity with a decreased resistance of four to six times in magnitude upon illumination with 365 nm
[79]. A single ZnO nanowire UV photodetector with high internal photoconductive gain of ~108
has been reported [80]. Two main factors contribute to the high photoresponse of ZnO detectors:
large surface to volume ratio, and reduced charge carrier transit time in low-dimensional structures.
ZnO-based photodetectors can be implemented in different device configurations including metal–­
semiconductor–metal (MSM), Schottky barrier, and p–n or p–i–n junctions [81–83]. In particular,
MSM photodetectors or photoconductors have been widely used because of their simple structure
and ease of device fabrication. However, such photodetectors have slow response and poor gain due
to charge recombination or charge trapping/detrapping at the defect levels of ZnO [84]. In order
to improve the photoresponse of ZnO, the charge carrier density in ZnO needs to be increased.
Recently, there have been considerable efforts in enhancing the UV photoresponse of ZnO using
functionalized metal nanoparticles (NPs). Due to the localized surface plasmon resonances (LSPRs)
of metal NPs, the absorption of light is enhanced, resulting in higher charge density in ZnO. The
enhancement in photoresponse by metal nanoparticles can be explained by LSPR-induced absorp-
tion and scattering phenomena. The absorption (Cabs) and scattering cross-sections (Cscat) of metal
NPs of average radius a are given by [85]

é e -e ù
Cabs = 4pka 3 Im ê s d ú (2.7)
ë e s + 2e d û

2
8p 4 6 e s - e d
Cscat = k a (2.8)
3 e s + 2e d

where εs and εd are the dielectric constants of the metal sphere and surrounding dielectric medium,
respectively. Therefore, the absorption and scattering processes vary with the radius of the metal
spheres. Consequently, for smaller particles the absorption dominates, while for larger sized par-
ticles the scattering process dominates. Therefore, the plasmonic effect could be useful to tailor the
photoresponse over a broad wavelength range.

2.7.2.1 Plasmonic ZnO Photodetector

Metal-functionalized ZnO nanostructures can be prepared by various methods. A simple photo-
reduction method is one of the new approaches to prepare metal-functionalized ZnO nanocom-
posites. In a typical process, the photogenerated electrons from ZnO under UV illumination react
with metal salts in an aqueous solution to form metal NPs on the surface of ZnO [86]. Typical
low- and high-resolution transmission electron microscopy (TEM) images and selected area elec-
tron diffraction pattern (SAED) of synthesized Au–ZnO plasmonic nanocomposites are shown in
Figure 2.12. As shown in Figure 2.12a, the spherical Au NPs are well attached to ZnO nanosheets.
The lattice fringes of Au–ZnO nanocomposites shown in Figure 2.12b reveal an interplanar spac-
ing of 0.281 and 0.235 nm for ZnO(1100) and Au(111) planes, respectively. The combined SAED
pattern of Au–ZnO shows the hexagonal spot patterns with circular rings, as shown in Figure 2.12c.
The hexagonal SAED pattern can be indexed to the (1100) , (0110), (1010), and (1010) planes of
ZnO nanosheets. The circular rings corresponding to (111), (200), and (220) crystallographic
planes of Au NPs reveal their polycrystalline nature. The absorption spectra of ZnO and Au–ZnO
plasmonic nanocomposites are presented in Figure 2.13. The Au–ZnO sample clearly exhibits two
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 59

1 nm
0.28

Au Au
ZnO

5 nm
0.23
ZnO nanosheet
2 nm
10 nm

(a) (b)

(101

0)
1 10
0)

(220) (
Au (200) (0110)
(111)
(10
10)

[0001]
ZnO
21/nm

(c)

FIGURE 2.12 Typical TEM images of plasmonic Au–ZnO nanocomposites at (a) low and (b) high resolution.
(c) Combined SAED pattern of plasmonic Au–ZnO nanocomposites. (From Gogurla, N. et al., Sci. Rep.,
4, 6483(1–9), 2014.)

ZnO
Au–ZnO
Absorbance (a.u.)

Plasmon band

400 500 600 700

Wavelength (nm)

FIGURE 2.13 UV–visible absorption spectra of ZnO and plasmonic Au–ZnO nanocomposites. (From
Gogurla, N. et al., Sci. Rep., 4, 6483(1–9), 2014.)
60 Semiconductor Nanocrystals and Metal Nanoparticles

peaks compared to one for pure ZnO. The peak in the UV region is due to the band-edge absorp-
tion of ZnO and the one in the visible region is attributed to the surface plasmon absorption of
Au NPs. The surface plasmon peak of Au–ZnO nanocomposite is broad in the visible region due
to the wide particle size distribution of Au NPs.
Photoconductive gain and spectral response are the key parameters to test the performance
of photodetectors. ZnO and plasmonic Au–ZnO nanocomposite devices were tested in the lat-
eral photoconductor configuration under UV illumination, the schematic of which is shown in
Figure 2.14a. Typical I–V characteristics for ZnO and Au–ZnO devices in the dark and under
UV illumination are shown in Figure 2.14b. For both the devices, a pronounced increase in the
current under UV illumination is observed. The photocurrent of the Au–ZnO device is found to
be enhanced compared to that of the control ZnO sample. The spectral photoresponse of ZnO

10–5

10–6
Current (A)

10–7

10–8

10–9

hν 10–10
Au–ZnO
Pt
–10 –8 –6 –4 –2 0 2 4 6 8 10
Voltage (V)
ZnO dark
ZnO under UV
Au–ZnO dark
(a) (b) Au–ZnO under UV

60 ZnO
Au–ZnO hν
50
Responsivity (mA/W)

40 at 5 V
e
30 Interband e
e
transition Ee
20 Au
e e
10 ×20
Ev
0 d band ZnO

300 320 340 360 380 400

(c) Wavelength (nm) (d)

FIGURE 2.14 (a) Schematic illustration of a plasmonic Au–ZnO device. (b) Typical I–V characteristics
of ZnO and plasmonic Au–ZnO devices under dark and UV illumination. (c) Spectral response of ZnO
and Au–ZnO devices in the UV region at 5 V bias. (d) Schematic diagram representing the charge transfer
mechanism involved in the Au–ZnO device. (From Gogurla, N. et al., Sci. Rep., 4, 6483(1–9), 2014.)
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 61

and Au–ZnO devices at 5 V bias are shown in Figure 2.14c. The responsivity (R λ) of the devices
can be calculated from [87]
DI
Rl = (2.9)
PAd

where
P is the optical power density falling on the sample
∆I is the difference between the dark current and photocurrent
Ad is the area of the lateral device.
The Au–ZnO nanocomposite device exhibited an enhanced photoresponse over a broad wavelength
range compared to ZnO. The responsivity of Au–ZnO is notably higher for the wavelength range
300–350 nm and slowly decreases when the device is illuminated with a longer wavelength. As seen
in Figure 2.14c, the peak responsivity of the plasmonic Au–ZnO device is enhanced by 80 times at
325 nm over the control ZnO sample.
The origin of the enhanced photoresponse in metal-functionalized semiconductor UV photodetec-
tors can be explained by the LSPR effect of metal NPs [88,89]. However, in this case, Au nanoparticles
show LSPR only in the visible region, as can be seen in Figure 2.13. Therefore, the enhanced UV
photoresponse for the Au–ZnO device might be attributed to the interband transition in Au under UV
illumination or to photogenerated hole trapping in Au nanoparticles [90,91]. Under UV illumination,
the filled d-band electrons of Au NPs are excited to the conduction band and subsequently get trans-
ferred to ZnO due to the electric field at the junction formed between Au and ZnO, resulting in a higher
photoconductivity. In case of ZnO, the responsivity in the UV region is attributed to the band-edge
absorption and is much lower than that of the plasmonic nanocomposite. The absorption and transfer
processes are schematically shown in Figure 2.14d. It is also possible that the photogenerated holes
in ZnO can be trapped by Au nanoparticles, which may suppress the electron–hole recombination in
the plasmonic device. This would result in an enhanced photoresponse in the UV region. Therefore,
metal-sensitized ZnO nanostructures can be used as a novel platform for enhanced UV photodetectors.
Due to the negative real part and relatively low imaginary dielectric function of Al in the UV
region, it is also useful to enhance the UV photoresponse of ZnO. Enhanced UV photodetection char-
acteristics of ZnO with Al NPs have been reported [92]. ZnO nanorods prepared on quartz substrates
by vapor-phase transport were decorated by Al nanoparticles using RF sputtering. A schematic MSM
device diagram of the Al NP-decorated ZnO nanorods is shown in Figure 2.15a. Typical I–V charac-
teristics of ZnO MSM photodetectors without and with Al NPs under dark and with UV illumination
are shown in Figure 2.15b and c, respectively. The linear increment in the current with the voltage
indicates the ohmic contact between the ZnO nanorods and Ag electrodes at the interfaces. Due to the
lower work function of Al compared to ZnO [93], the electrons can move from Al to ZnO. These elec-
trons contribute to the conduction under 5 V bias, resulting in a higher dark current for the Al–ZnO
device. Under UV light illumination, the ZnO device with Al NPs has shown increased photocurrent
as compared to that without Al NPs. This can be attributed to the LSPR coupling between the Al NPs
and ZnO nanorods, leading to the enhanced electron–hole generation in ZnO.
Figure 2.16 shows the responsivity of ZnO UV photodetectors with and without Al NPs under
UV–vis light illumination at a bias of 5 V. The responsivity spectra show that the devices are sensi-
tive to UV light. It was also observed that the responsivity of the ZnO detector with Al NPs enhanced
greatly in the UV region. The responsivity peak was observed at 375 nm for both the devices. UV
radiation can be absorbed by Al NPs due to LSPR effect at the metal–dielectric interface. The LSPR
of Al NPs can couple with the excitons of ZnO due to the matching of surface plasmon energy with
ZnO interband transition. This significant energy coupling will lead to the pronounced excitation in
ZnO, resulting in a faster response under UV illumination.
Several other metal nanoparticles have also been used to enhance the photoresponse of ZnO
photodetectors. Enhanced photoresponse could be achieved in ZnO thin film MSM devices using Pt
62 Semiconductor Nanocrystals and Metal Nanoparticles

1E–6
W/O Al NPs
325 nm laser 1E–7

Current (A)
1E–8
1E–9 Dark
Ag Ag UV
1E–10
ZnO nanorods 1E–11

SiO2 –5 –4 –3 –2 –1 0 1 2 3 4 5
(a) (b) Voltage (V)

1E–5
with Al NPs
1E–6
Current (A)

1E–7

1E–8 Dark
UV
1E–9

1E–10
–5 –4 –3 –2 –1 0 1 2 3 4 5
(c) Voltage (V)

FIGURE 2.15 (a) Schematic illustration showing the device diagram of plasmonic Al–ZnO nanorods. I–V
characteristics of (b) bare ZnO and (c) Al-decorated ZnO devices in the dark and 325 nm illumination. (From
Lu, J., Xu, C., Dai, J. et al., Improved UV photoresponse of ZnO nanorod arrays by resonant coupling with
surface plasmons of Al nanoparticles, Nanoscale, 7, 3396–3403, 2015. Reproduced by permission of The
Royal Society of Chemistry.)

0.14
Responsivity (A/W)

0.12 W/O Al NPs

0.10
0.08
0.06
0.04
0.02
0.00
(a)
1.8
1.5
Responsivity (A/W)

With Al NPs
1.2
0.9
0.6
0.3
0.0

250 300 350 400 450 500 550 600 650 700
(b) Wavelength (nm)

FIGURE 2.16 Photoresponsivity spectra (a) without and (b) with Al-decorated ZnO devices at a bias of 5 V.
(From Lu, J., Xu, C., Dai, J. et al., Improved UV photoresponse of ZnO nanorod arrays by resonant coupling
with surface plasmons of Al nanoparticles, Nanoscale, 7, 3396–3403, 2015. Reproduced by permission of
The Royal Society of Chemistry.)
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 63

nanoparticles [94]. The photoresponse of ZnO devices with Pt NPs prepared by different sputtering
times were recorded at 3 V bias. The responsivity of ZnO increased from 0.836 to 1.306 A/W using
Pt nanoparticles [94]. The enhanced photoresponse in the Pt–ZnO devices could be well understood
from the absorption and scattering phenomena of Pt nanoparticles.

2.7.3 ZnO Nanostructures for Photovoltaic Devices

Due to their high transparency in the visible region and their high electron mobility, ZnO nanostruc-
tures are attractive for dye- and quantum-dot-sensitized solar cells. Several studies have been car-
ried out on ZnO nanostructure–based, dye-sensitized solar cells (DSSCs), which are summarized in
different reviews [95,96]. Although the obtained maximum power conversion efficiency of 5.6% for
ZnO-based DSSCs [96] is much lower than that of TiO2 (efficiency of 11%) [97], it is still potentially
attractive to replace TiO2 due to the easy crystallization and anisotropic growth of ZnO. However,
the degradation of dyes results in the lack of stability and long-term performance, which limits the
applications of DSSCs. An effective approach to overcome the problems associated with DSSCs is to
replace the dyes with polymers. Such hybrid solar cells comprise a polymer as the electron donor and
metal oxide as the electron acceptor. Hybrid solar cells are expected to exhibit superior performance
since they combine the merits of both polymers and metal oxides, for example, flexibility and low
processing cost of polymers as well as the thermal and chemical stability, high electron mobility, high
dielectric constant, and size-tunable optical properties of metal oxides. The ordering of donors and
acceptors in a polymer–metal oxide hybrid solar cell is commonly done in two different ways. In one
case, the polymer (donor) can be blended with nanostructured metal oxides (acceptor) to form poly-
mer–metal oxide hybrid bulk heterojunction (BHJ) solar cells. Such solar cells have the advantage of
low fabrication cost, since the blending of the polymer with ZnO nanostructures can be done by the
solution-growth technique. The blending also leads to large donor–acceptor interfacial areas, which
results in the formation of a large number of free charge carriers. However, the main challenge lies
in controlling the morphology of the heterojunction so that definite transport pathways for the charge
carriers can be obtained. The other approach is to grow ordered metal oxide nanostructures on the
substrate and infiltrate the polymer in between them. Such solar cells offer a definite transport pathway
for the charge carriers. But the main challenges to obtaining high performance in these solar cells are
(1) the complete infiltration of polymer into the nanostructures, (2) growth of metal oxide nanostruc-
tures of smaller dimensions to obtain higher interfacial area, and (3) controlling the distance between
the nanostructures to generate larger number of excitons within the excitonic diffusion length.
In particular, vertically aligned ZnO nanorods have been employed as electron acceptors along with
the conductive organic polymer in hybrid solar cells due to their low reflectivity as well as efficient
charge separation and collection [98]. In this regard, hybrid solar cells based on ZnO ­nanostructures
and the regioregular poly(3-hexylthiophene) (P3HT) polymer have been studied by several research
groups. Baeten et al. have demonstrated ZnO nanorods/P3HT solar cell with 0.76% efficiency [99]. The
power conversion efficiency of ZnO/P3HT hybrid devices has been observed to be very low because
of the incompatibility of the interface between ZnO and organic polymers. Some polymer processing
techniques such as annealing at their melting point and subsequent ­cooling, spin coating from different
organic solvents, and UV treatment on ZnO before polymer deposition can be employed to improve the
performance [100,101]. An effective way to improve the photovoltaic performance of hybrid solar cells
is to modify the surface of ZnO nanostructures with different inorganic semiconductors. Such modifi-
cation improves the interfaces between the ZnO nanostructures and polymers to form an optimum mor-
phology, which would ensure efficient charge transport with enhanced performance of the solar cells.

2.7.3.1 CdS-Sensitized ZnO Nanorods

Hybrid photovoltaic devices have been studied with an active layer consisting of a blend of CdS-
nanoparticle-decorated ZnO nanorods and P3HT polymer [102]. CdS was deposited by pulsed-laser
deposition on the surface of ZnO nanorods at varying deposition times of 10, 20, and 30 min, with the
64 Semiconductor Nanocrystals and Metal Nanoparticles

samples denoted as CSZO-1, CSZO-2, and CSZO-3, respectively. In CSZO-1, ZnO nanorods are only
partially covered by CdS nanoparticles, whereas in the CSZO-2 sample a larger portion becomes cov-
ered. A continuous CdS layer of about 80 nm thickness was formed on the ZnO nanorod surface in the
sample CSZO-3, as shown in Figure 2.17a. A photovoltaic device structure having Al/CdS-decorated
ZnO nanorods:P3HT/PEDOT:PSS/ITO was fabricated and tested. The conductive PEDOT:PSS layer
plays several roles: it assists in hole transport, blocks the excitons, and prevents the oxygen diffu-
sion from ITO by keeping the anode material away from the active layer to prevent the formation of
undesirable trap sites [103]. PL measurements showed that the intensity of P3HT quenched with the
increase of ZnO concentration. The quenching of P3HT intensity further enhanced on modification of
the surface of ZnO nanorods with CdS, indicating the higher rate of exciton dissociation.
The current density versus voltage characteristics of CdS-modified ZnO:P3HT devices (of
weight ratio 5:1) with varying CdS surface coverage are shown in Figure 2.17b. The measurement

(a)

1 μm

0.5

0.0
Current density (mA/cm2)

–0.5

–1.0

ZnO:P3HT blend
–1.5 CSZO-1:P3HT blend
CSZO-2:P3HT blend
CSZO-3:P3HT blend

–0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6

(b) Voltage (V)

FIGURE 2.17 (a) FESEM image of CdS-decorated ZnO nanorods (CSZO-3 sample). Inset shows the magni-
fied view of the micrograph. (b) Current density versus voltage characteristics of CdS-modified ZnO:P3HT
and control ZnO:P3HT hybrid solar cells (blended in weight ratio of 5:1) with varying CdS surface coverage.
(From Rakshit, T. et al., ACS Appl. Mater. Interfaces, 4, 6085, 2012.)
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 65

was performed under 100 mW/cm2 AM 1.5 simulated solar irradiation. The surface modification
of ZnO nanorods with CdS nanoparticles led to an increase in the open-circuit voltage (VOC) and
short-circuit current density (JSC). These parameters were found to increase with the increase in
CdS surface coverage. The highest power conversion efficiency (PCE) of 0.38% was obtained
in the CSZO-3:P3HT devices, with JSC of 1.67 mA/cm2, VOC of 557 mV, and fill factor (FF) of
40%. The efficiency of CdS-decorated ZnO (CSZO-3):P3HT device is higher by more than 300%
­compared to the control ZnO:P3HT device.
Figure 2.18a shows the external quantum efficiency (EQE) of ZnO:P3HT devices with different
ZnO:P3HT weight ratios. The EQE initially increases with ZnO addition in the blend up to a weight

7.5

6.0

4.5
EQE (%)

3.0

ZnO:P3HT weight ratio:

0.5:1
1.5 1:1
2.5:1
5:1
12:1
0.0
300 400 500 600 700

(a) Wavelength (nm)

16

12
EQE (%)

4 ZnO:P3HT blend
CSZO-1:P3HT blend
CSZO-2:P3HT blend
CSZO-3:P3HT blend
0
300 400 500 600 700

(b) Wavelength (nm)

FIGURE 2.18 External quantum efficiency of (a) ZnO:P3HT hybrid solar cells, blended in different weight
ratios, and (b) CdS-modified ZnO:P3HT and control ZnO:P3HT hybrid solar cells (blended in weight ratio of 5:1),
with varying CdS surface coverage. (From Rakshit, T. et al., ACS Appl. Mater. Interfaces, 4, 6085, 2012.)
66 Semiconductor Nanocrystals and Metal Nanoparticles

ratio of 5:1, but then decreases with the further increase of ZnO concentration. The highest EQE
obtained for ZnO:P3HT devices is ~7% at around 510 and 550 nm. The EQE of CdS-decorated
ZnO:P3HT devices (of weight ratio of 5:1) is shown in Figure 2.18b. The EQE of CdS-modified
ZnO devices is higher than that of the unmodified one in the wavelength range 300–620 nm,
which is enhanced with the increase of CdS surface coverage. The highest EQE for CdS-decorated
ZnO:P3HT devices was found to be ~16% at around 510 nm, which is a factor of 2 higher than that
of the control ZnO device.
The improved photovoltaic performance of CdS-decorated ZnO:P3HT devices can be explained
from the energy band diagram shown schematically in Figure 2.19. The electrons flow efficiently
from donors to acceptors if the energy difference between the lowest unoccupied molecular orbital
(LUMO) of donors and the conduction band edge of acceptors is ~0.3–0.5 eV [104]. However, this
difference is much higher (1.2 eV) in ZnO:P3HT devices. When the surface of ZnO nanorods is
modified with CdS nanoparticles, a cascaded band structure is formed, since the conduction band of
CdS has a higher energy than that of ZnO but lower than the LUMO of P3HT. This ensures efficient
flow of electrons from P3HT to ZnO. Moreover, the hole transfer from P3HT to ZnO is blocked by
CdS because of the alignment of the highest occupied molecular orbital (HOMO) of P3HT relative
to the valence band energy of CdS. This reduces the recombination in ZnO, thereby increasing the
lifetime of the carriers. Therefore a higher efficiency is observed in the CdS-modified ZnO devices.
This efficiency can be further improved by enhancing the CdS coverage on the surface of ZnO
nanorods. When the CdS nanoparticles partially cover the ZnO nanorod surface, a cascaded band
structure is absent in the unmodified portion. The formation of the cascaded band structure occurs
when the ZnO nanorods are fully covered with CdS nanoparticles, resulting in the smooth transfer
of electrons, which enhances the photovoltaic performance of CdS-decorated ZnO devices.

2.7.3.2 PbS-Sensitized ZnO Nanorods

Hybrid photovoltaic devices have been studied with PbS quantum dot (QD)-sensitized ZnO nanorod
arrays and poly[2-methoxy-5-(2-ethylhexyloxy-p-phenylenevinylene)] (MEH-PPV) polymer as the
active layer [105]. The PbS QDs were deposited on the surface of ZnO nanorods by chemical bath
deposition (CBD), with two, four, and six CBD cycles. For lower CBD cycles, PbS QDs partially

–2
e–

–3 e–
–3
e–
h+
–4 –3.8 h+
–4.2
Energy (eV)

–4.3
–5 –4.5
h+ –4.9
–5.2
–6
h+
–6.2

–7

–7.5
–8 PEDOT:
Al ZnO Cds P3HT ITO
PSS

FIGURE 2.19 Schematic energy band diagram of CdS-decorated ZnO:P3HT hybrid solar cell. (From Rakshit, T.
et al., ACS Appl. Mater. Interfaces, 4, 6085, 2012.)
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 67

covered the surface of ZnO nanorods, whereas ZnO nanorods were almost fully covered with six
CBD cycles. The device structure used for the photovoltaic measurement was Au/PEDOT:PSS/
MEH-PPV/PbS/ZnO-nanorod/ZnO-seed layer/ITO. The current density versus voltage charac-
teristics of PbS QD-sensitized ZnO devices, along with the control MEH-PPV/ZnO device, mea-
sured under 100 mW/cm2 AM 1.5 simulated solar irradiation condition are shown in Figure 2.20a.
The MEH-PPV/ZnO device exhibited JSC of 1.06 mA/cm2, VOC of 0.25 V, FF of 30%, and PCE of
0.09%. The efficiency was found to be enhanced with the introduction of PbS QDs on the surface
of ZnO nanorods. The highest efficiency of 0.42% was obtained for four-cycle CBD grown, PbS
QD-sensitized device. Thus, the efficiency obtained in PbS QD-sensitized ZnO devices is higher by
approximately five times that of the unmodified device.

1
Current density (mA/cm2)

–1
dark ZnO + 4Pbs + MEHPPV
ZnO + MEHPPV
–2 ZnO + 2PbS + MEHPPV
ZnO + 4PbS + MEHPPV
ZnO + 6PbS + MEHPPV
–3
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

(a) Voltage (V)

–2

3.0
–3
PEDOT:PSS

3.9
MEH-PPV

–4
Energy level (eV)

4.4
ITO PbS
–5 4.8 Au
5.1 5.0
5.2
5.4
–6 ZnO

–7

7.6
–8
(b)

FIGURE 2.20 (a) Current density versus voltage characteristics of MEH-PPV/PbS QD-sensitized ZnO and
control MEH-PPV/ZnO devices. ZnO + 2PbS, ZnO + 4PbS, and ZnO + 6PbS are PbS QD-sensitized ZnO
nanorods, grown with two, four, and six CBD cycles, respectively. (b) Schematic energy band diagram of
MEH-PPV/PbS QD-sensitized ZnO solar cell. (Reprinted with permission from Nanoscale Res. Lett., Hybrid
polymer/ZnO solar cells sensitized by PbS quantum dots, 7, 2012, 106-1, Wang, L., Zhao, D., Su, Z., and Shen, D.
© 2012 by the SpringerOpen.)
68 Semiconductor Nanocrystals and Metal Nanoparticles

Figure 2.20b shows the schematic energy band diagram of MEH-PPV/PbS QD-sensitized ZnO
devices. The introduction of PbS QDs on the surface of ZnO nanorods results in the formation of
a cascaded band structure, since the conduction band of PbS has a higher energy than that of ZnO.
Such a cascaded structure aids in efficient transport of electrons and reduces the recombination
of charge carriers. Also, the PbS coating passivates the surface states of ZnO nanorods, result-
ing in reduced recombination and charge trapping. This enhances the short-circuit current density
and open-circuit voltage of the MEH-PPV/PbS QD-sensitized ZnO devices. The ZnO seed layer
between the active layer and the ITO electrode prevents the leakage of holes into the ITO electrode.
These factors result in an improved efficiency in the four-cycle CBD grown PbS QD-sensitized
device. The decrease of efficiency at higher PbS coating (six CBD cycles) is attributed to the reduc-
tion in the exposed amount of ZnO. The higher PbS coating also reduces the spacing between the
nanorods, resulting in poor infiltration of MEH-PPV into the ZnO nanorods. Thus, the efficiency
of MEH-PPV/PbS QD-sensitized ZnO hybrid solar cells largely depends on the amount of surface
coverage of ZnO nanorods by PbS QDs.

2.7.3.3 Cosensitized (CdS and CdSe) ZnO Nanowires

Hybrid solar cells have also been fabricated with ZnO nanowires cosensitized with two inorganic
semiconductors, namely, CdS and CdSe [106]. A CdS layer of ~9 nm was uniformly deposited
on the surface of ZnO nanowires by successive ionic layer adsorption and reaction (SILAR), over
which CdSe was deposited by CBD. CdSe nanoparticles were deposited randomly on the surface
of the CdS/ZnO nanowires. The structure of the fabricated devices used for photovoltaic measure-
ments was Au/P3HT/CdSe/CdS/ZnO-nanowire/ZnO-film/ITO. The P3HT/CdSe/CdS/ZnO nanow-
ires showed a high absorbance in the visible region of 400–700 nm. The spectral response was much
broader than of pure ZnO, CdS/ZnO and CdSe/CdS/ZnO nanowires. Figure 2.21a shows the cur-
rent density versus voltage characteristics of hybrid solar cells fabricated with P3HT/CdSe/CdS/
ZnO, P3HT/CdS/ZnO, and P3HT/ZnO nanowires, measured under 100 mW/cm2 AM 1.5G solar
illumination. The introduction of CdS and CdSe quantum dots on the surface of ZnO nano­wires
greatly enhances the photovoltaic performance, yielding an efficiency of 1.5%, which is about 75
and 6 times higher than those of P3HT/ZnO and P3HT/CdS/ZnO devices, respectively. The corre-
sponding short-circuit current density, open-circuit voltage, and fill factor of P3HT/CdSe/CdS/ZnO
device are 4.2 mA/cm2, 675.2 mV, and 51.8%, respectively. The performance was found to be highly
stable even after 40 days of exposure in air.
The photovoltaic performance of P3HT/CdSe/CdS/ZnO nanowires was also studied as a func-
tion of the length of the ZnO nanowires varying from ~630 nm to 1.45 µm. The short-circuit current
density increased with the increase of the length of the ZnO nanowires up to ~1 µm. This could be
due to the higher loading of CdS and CdSe QDs in longer nanowires. However, further increase
in the length of ZnO nanowires lowered the infiltration of P3HT through the nanowires, which
reduced the short-circuit current density. On the other hand, the open-circuit voltage decreased with
the increase of the length of the ZnO nanowires. This may be due to the enhanced density of recom-
bination sites and inefficient transport of charges in longer ZnO nanowires. The highest efficiency of
~1.4% was obtained in devices with ~1 µm long ZnO nanowires. The cosensitized device exhibited
a higher incident photon-to-current conversion efficiency in the visible region of 350–700 nm, with
the maximum value of ~25% at 530 nm.
The energy band diagram and device structure of hybrid solar cells based on CdS- and CdSe-
sensitized ZnO nanowires are shown in Figure 2.21b. When the surface of ZnO nanowires is sen-
sitized with CdS and CdSe, a cascaded band structure is formed. The cascaded band structure is
useful in the dissociation and efficient transport of charge carriers. This reduces the recombination
probability and hence increases the carrier lifetime [107,108]. CdS and CdSe QDs also passivate the
surface states of ZnO nanowires, thus reducing the density of charge trapping and recombination
sites [107,108]. The shunt resistance of hybrid solar cells, which prevents the flow of photogen-
erated charge carriers through the internal defects, was also analyzed for cosensitized solar cells.
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 69

6
P3HT/CdSe/CdS/ZnO
5 P3HT/CdS/ZnO
P3HT/ZnO

Current density (mA/cm2)

4

0
0 100 200 300 400 500 600 700 800

(a) Voltage (mV)

e– e–
e–
e– Electrode
e– (Au)
–4.9 eV
CdSe
CdS P3HT e–
h+
ZnO NW h+
Transparent conducting substrate (ITO)
(b)

FIGURE 2.21 (a) Current density versus voltage characteristics of hybrid solar cells fabricated with P3HT/
CdSe/CdS/ZnO, P3HT/CdS/ZnO and P3HT/ZnO nanowires. (b) Schematic energy band diagram and device
structure of P3HT/CdSe/CdS/ZnO nanowire hybrid solar cell. (Reprinted with permission from Kim, H., Jeong,
H., An, T.K., Park, C.E., and Yong, K., Hybrid-type quantum-dot cosensitized ZnO nanowire solar cell with
enhanced visible-light harvesting, ACS Appl. Mater. Interfaces, 5, 268. © 2013, American Chemical Society.)

The shunt resistance of P3HT/ZnO solar cell was increased by about 17 times by modifying with
CdS and by ~26 times after cosensitizing with CdS and CdSe QDs. All these factors resulted in an
improved photovoltaic performance in P3HT/CdSe/CdS/ZnO nanowire solar cells.
Some of the reported results on hybrid solar cells based on semiconductor-sensitized ZnO
nanostructures are summarized in Table 2.2. It is apparent that the modification of the surface of
ZnO nanostructures with different inorganic semiconductors plays a significant role in control-
ling the photovoltaic performance of both ZnO/polymer BHJ and nanostructured ZnO/polymer
hybrid solar cells.

2.7.4 Gas Sensors Using ZnO Nanostructures

The change in electrical conductivity of ZnO thin films in the presence of reactive gases was discov-
ered in 1962 [115]. This property of ZnO has led many researchers to synthesize different kinds of
ZnO nanostructures for novel gas sensing applications. ZnO is one of the most widely investigated
sensor materials used for the detection of different gases and volatile organic compounds (VOCs)
[116–120]. ZnO nanostructures exhibit pronounced surface effects owing to their high surface to
volume ratio, leading to superior sensing performance. Gas sensors using a variety of ZnO nano-
structures have been widely reported. Jing et al. have fabricated a porous ZnO nanoplate-based
70 Semiconductor Nanocrystals and Metal Nanoparticles

TABLE 2.2
Photovoltaic Parameters for Hybrid Solar Cells Based on Semiconductor-Sensitized
ZnO Nanostructures
Short-Circuit Open-
Current Circuit Fill
Density, JSC Voltage, Factor, Efficiency, Irradiance
Device Structure (mA/cm2) VOC (V) FF (%) η (%) (mW/cm2) References
Al/CdS-decorated ZnO 1.674 0.557 40 0.38 100 [102]
Nanorods:P3HT/PEDOT:​
PSS/ITO
Au/P3HT/CdS/ZnO Nanofiber/ 2.25 0.619 46.5 0.65 100 [109]
ZnO thin film/ITO
Au/PEDOT:PSS/MEH-PPV/ 2.87 0.78 29 0.65 100 [110]
CdS/ZnO nanorod/ZnO seed
layer/ITO
Au/PEDOT:PSS/MEH-PPV/ 4.65 0.77 34.4 1.23 100 [111]
CdS/ZnO nanorod/ITO
Al/MoO3/P3HT/CdSe/ZnO 1.44 0.36 43 0.23 100 [112]
Nanorod/ZnO seed layer/ITO
Au/P3HT/CdSe/CdS/ZnO 4.2 0.6752 51.8 1.50 100 [106]
Nanowire/ZnO thin film/ITO
Au/PEDOT:PSS/MEH-PPV/ 2.68 0.55 29 0.42 100 [105]
PbS/ZnO nanorod/ZnO seed
layer/ITO
Pt/P3HT/CuS/ZnO Nanorod/ 4.9 0.391 53 1.02 100 [113]
ZnO seed layer/ITO
Au/P3HT/TiO2/ZnO nanorod/ 1.14 0.50 50 0.29 100 [114]
ZnO seed layer/ITO

sensor that exhibited high response to chlorobenzene at low operating temperatures [121]. A strong
response to ethanol was also observed for the sensor operated at high temperatures [121]. Park
et al. have reported the gas sensing properties of nanograined ZnO nanowires, which showed an
enhanced electrical response to NO2 at 300°C [122]. Different morphologies of ZnO nanostructures
have been utilized to sense ethanol and the results indicated that the nanofiber sensor shows
desirable response at 270°C [123]. Some theoretical simulations have also been performed on
ZnO gas sensors. The interaction of ZnO nanostructure (10 10) surface with NO2, NO, O, and N
was examined using density functional theory and ab initio molecular dynamics simulations [124].
A weak adsorption of nitrogen oxides and stable adsorption of O and N even at 700 K on the sur-
face of ZnO were observed. The calculations also indicated that all the adsorbates act as electron
acceptors on the surface. This potential sensing properties of ZnO nanostructures resulted in highly
responsive gas sensors. However, the applications of these ZnO nanostructures are often limited by
the high operating temperature and lack of stability, accuracy, and selectivity toward various gases
and VOCs. VOCs have quite similar molecular structure and composition, which increases the dif-
ficulty of distinguishing one VOC from another.

2.7.4.1 Gas Sensing Parameters

The adsorption of oxygen on the surface of ZnO nanostructures and the rate of reaction between the
target gas and oxygen depend strongly on the operating temperature. Other important parameters
determining the performance of gas sensors are sensitivity, selectivity, response time, recovery
time, and the limit of detection. These parameters are largely influenced by the surface area, donor
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 71

density, porosity of the sensing material, and the use of catalysts [125]. The crystallite size of metal
oxide also plays a significant role in determining the sensitivity of a gas sensor. When the size of the
crystallite (D) is less than twice the thickness of the depletion layer (L), that is, D < 2L, the entire
crystallite would become depleted of charge carriers. The energy bands become almost flat, leading
to easy charge transport within the crystallites. This inter-crystallite conductivity largely controls
the conductivity of the oxide. Then a small change in the thickness of the depletion layer due to
the reaction of the oxygen ion with the target gas causes a large variation in the conductivity of the
metal oxide. Thus, metal oxides having a small crystallite size are highly sensitive to the target gas.
One of the approaches of improving the selectivity and other sensing performances is to modify
the surface of ZnO nanostructures with different metals or metal oxides. When the surface of ZnO
nanostructures is sensitized by another metal oxide, the Fermi levels on both sides of the interface
reach equilibrium through charge transfer, and a depletion layer is formed. This depletion layer
plays a significant role in enhancing the performance of the gas sensor. To reduce the operating
temperature and improve the sensing performance, several approaches such as doping or function-
alizing with noble metals [126,127] and UV light stimulation [128] have been employed. UV light
stimulation effectively improves the sensing performances at low operating temperatures.
The sensitivity of a gas sensor is defined in various ways. The most widely used one is defined as
Ra/Rg for an n-type metal oxide in the presence of a reducing gas, where Ra and Rg are the resistance
of the sensor in air/carrier gas and target gas, respectively. Similarly, in the presence of an oxidizing
gas, the sensitivity is given by Rg/Ra for an n-type metal oxide. These become reversed in the case of a
p-type material. The response and recovery time are usually defined as the time needed for the sensor
to reach 90% of the variation in resistance upon exposure and removal of the target gas, respectively.

2.7.4.2 Surface-Modified ZnO Sensors

As discussed previously, several factors such as operating temperature, surface morphology, and
growth technique control the performance of ZnO nanostructure–based gas sensors. The size,
shape, and surface states of ZnO nanostructures largely depend on the growth techniques. Variation
of any of these parameters can change the performance of sensors. To improve the gas sensing per-
formance, the surface of ZnO nanostructures is often modified with different semiconductor metal
oxides or metal nanoparticles. In this section, we discuss the sensing performance of gas sensors
based on different semiconductor- and metal-sensitized ZnO nanostructures.

2.7.4.2.1 Semiconductor-Sensitized Gas Sensors

2.7.4.2.1.1 SnO2 Nanowire/ZnO Nanorod Heterostructures The sensing performance of brush-
like SnO2 nanowire/ZnO nanorod heterostructures, synthesized by a combination of pulsed laser
deposition and hydrothermal methods, was tested with different VOCs such as acetone, ethanol, meth-
anol, acetic acid, toluene, and triethylamine [129]. SnO2 nanowire/ZnO nanorod heterostructures with
varying length of the nanowires of ~25, 40, and 55 nm were utilized, which were designated as ZSO1,
ZSO2, and ZSO3 samples, respectively. The density of SnO2 nanowires was not uniform over the ZnO
nanorod, so the surface of the heterostructures was porous and rough. The optimum operating tem-
perature for sensing acetone, ethanol, acetic acid, toluene, and triethylamine was found to be 300°C,
whereas that for methanol sensing was 250°C. Figure 2.22a through c shows the sensing transients of
pure ZnO nanorods and SnO2/ZnO heterostructures toward toluene, ethanol, and acetone, respectively,
operated at a temperature of 300°C. The response (Ra/Rg) of ZSO1 sensor for toluene (14–115 ppm) is
~1.3 times higher than that of ZnO. However, the response of ZSO2 and ZSO3 sensors is lower than
that of ZnO. The sensing characteristics for ethanol (29–230 ppm), shown in Figure 2.22b, reveals that
the highest response is shown by ZSO3 sample, which is ~1.8–3 times higher than that of the control
sample. On the other hand, the ZSO2-based sensor exhibits the highest response of 21.4–48.4 toward
acetone (115–693 ppm), which is ~5.1–3 times higher than that of ZnO nanorods (Figure 2.22c).
Sensing transients of SnO2/ZnO heterostructures indicate that the performance depends strongly on
the length of the SnO2 nanowire brushes covering the ZnO surface.
72 Semiconductor Nanocrystals and Metal Nanoparticles

7
ZnO

115 ppm
ZSO1
6

57 ppm
ZSO2
ZSO3

29 ppm
5

Response (Ra/Rg)
4

14 ppm
3

1
0 1000 2000 3000
(a) Time (s)

16 ZnO

230 ppm
115 ppm
ZSO1
ZSO2
ZSO3
12
57 ppm
Response (Ra/Rg)

29 ppm

0
0 500 1000 1500 2000 2500
(b) Time (s)
693 ppm
462 ppm

50 ZnO
ZSO1
231 ppm

ZSO2
40 ZSO3
Response (Ra/Rg)

115 ppm

30

20

10

0
0 1000 2000 3000 4000 5000
(c) Time (s)

FIGURE 2.22 Sensing transients of ZnO nanorods and SnO2/ZnO heterostructures toward (a) toluene,
(b) ethanol, and (c) acetone, operated at 300°C. (From Rakshit, T. et al., RSC Adv., 4, 36749, 2014.)
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 73

21.4
Triethylamine
20 Toluene
Ethanol
Acetic acid
Response (Ra/Rg) 15 Acetone

12.7
Methanol

9.3
10

8.2

7.5
6.3

5.3
5.1
4.8

4.8

4.7
4.4

4.4
4.3
4.2

4.2
4.2
5

3.4
3.1

2.4
2.2
2.1
1.9

1.3
0
ZnO ZSO1 ZSO2 ZSO3

FIGURE 2.23 Sensing responses of ZnO nanorods and SnO2/ZnO heterostructures to 115 ppm acetone,
ethanol, methanol, acetic acid, toluene, and triethylamine. (From Rakshit, T. et al., RSC Adv., 4, 36749, 2014.)

The selectivity of ZnO nanorods and SnO2/ZnO heterostructures toward 115 ppm acetone,
ethanol, methanol, acetic acid, toluene, and triethylamine has been investigated, and the results
are shown in Figure 2.23. In pure ZnO nanorods and 25 nm long SnO2 nanowire heterostructure,
the response toward different VOCs is nearly same, and thus lacks any selectivity. The sensing
response of 40 nm long SnO2 nanowire heterostructure (ZSO2) is highest for acetone, which is
more than three times higher than that for the other VOCs. On the other hand, the response toward
ethanol and acetic acid is quite close to each other for ZSO3-based sensor, and thus lacks selectivity.
Thus, SnO2/ZnO heterostructures are capable of selectively detecting acetone by varying the length
of SnO2 nanowires on the surface of ZnO nanorods.
The response of a semiconducting oxide gas sensing element is usually represented as [130]

S = Ag Pgg (2.10)

where
Pg is the partial pressure of the test gas
Ag is the prefactor
γ is the exponent

The variation of response of ZSO1, ZSO2, and ZSO3 with toluene, acetone, and ethanol vapors,
respectively, when fitted with Equation 2.10, yields the γ value of 0.43 ± 0.06, 0.44 ± 0.07, and
0.48 ± 0.06. These values are slightly lower than that exhibited by an ideal microstructure (0.5).
This may be attributed to the less sensitive zones or agglomeration of the microstructure [130].
The limit of detection for toluene, ethanol, and acetone sensing has been estimated to be ~2, 1,
and <0.2 ppm, respectively, indicating the capability of SnO2/ZnO nanorod heterostructures of
detecting very low VOC concentrations.

2.7.4.2.1.2   Sensing Mechanism for SnO2/ZnO Heterostructures The gas sensing mechanism
of ZnO nanorods and SnO2/ZnO heterostructures can be described by the adsorption and desorption
processes occurring on the surface of nanostructures [131–133]. When exposed to air, the surface
of nanostructures adsorbs oxygen molecules, which turn into oxygen ions (O2−, O2−, and O −) by
accepting electrons from the conduction band, resulting in the formation of a depletion layer on
74 Semiconductor Nanocrystals and Metal Nanoparticles

the surface. This leads to narrowing of the conduction channel, which increases the resistance of
ZnO. Figure 2.24a through d shows the schematic representation of the sensing mechanism of ZnO
nanorods and SnO2 nanowire/ZnO nanorod heterostructures. When the nanostructures are exposed
to a VOC such as acetone, it reacts with the chemisorbed oxygen ions following the reaction [134]

C3H 6O + 8O - ® 3CO2 + 3H 2O + 8e - (2.11)

The electrons are thus released back to the conduction band of the nanostructures, resulting in the
reduction of the depletion layer thickness and widening of the conduction channel. The resistance
of the oxide thus decreases.
The electron transport properties of nanostructures depend on the formation of junctions at the
interfaces. In SnO2/ZnO heterostructures, a homojunction potential barrier develops at the interface
of SnO2 nanowires, whereas a heterojunction potential barrier forms at the SnO2 nanowire/ZnO
nanorod interface. Since ZnO has a higher work function than SnO2 [135], electron transfer will
occur from SnO2 to ZnO in the heterojunction. The process will continue till the Fermi levels of
ZnO and SnO2 equilibrate. Thus, a depletion layer is developed at their interface, resulting in bend-
ing of the energy band (shown in Figure 2.24e). The difference of the work function of ZnO and
SnO2 is ~0.3 eV, which is the magnitude of the potential formed at their interface. If the electron
transport is modulated by an effective potential barrier (Φeff), which includes the contribution of
both the homo- and heterojunction potential barriers, the conductivity can be expressed by [136]

æ -F eff ö
s = sc exp ç ÷ (2.12)
è kBT ø

O– O– O–
O– O–
O– O–
O– O– O–e–
– O–
O– –
O O– O–
O–e– O O– O– e– O–
e– O– O–
O– O– –
O– e– O
O– O– O–
O– e– O–
O– O– O– O– O–
O–
e–O– O– e– O– O–
O– O–
O–
(a) Exposed to air (b) Exposed to VOC (c) O– O– (d) O– O
–

ZnO nanorods
Oxygen ion
Depletion region formed due to
adsorption of oxygen
Potential Depletion region formed at the
barrier homojunction of SnO2 nanowires
Ec SnO2 nanowires
EF
SnO2 SnO2 Target gas
Ec Depletion region formed at the
Potential barrier heterojunction of ZnO nanorods
and SnO2 nanowires
EF
ZnO SnO2
(e)

FIGURE 2.24 Schematic sensing mechanism of (a) ZnO nanorods exposed to air, (b) ZnO nanorods exposed
to VOC, (c) SnO2/ZnO heterostructures exposed to air, and (d) SnO2/ZnO heterostructures exposed to VOC.
(e) Band structure of SnO2/ZnO heterostructures. (From Rakshit, T. et al., RSC Adv., 4, 36749, 2014.)
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 75

where σc is a constant pre-exponential factor. When nanostructured samples are exposed to air, the
adsorbed oxygen molecules extract electrons from the conduction band, increasing the height and
width of the effective potential barrier. These trapped electrons are released back to the conduc-
tion band upon exposing the nanostructures to VOCs, which reduces the effective potential barrier
[137,138]. Therefore, the electron transport in SnO2/ZnO heterostructures is largely modulated by
the homo- and heterojunction potential barriers formed at the interfaces, which results in a superior
gas sensing performance as compared to control ZnO nanorods.
The strong dependence of the response toward different VOCs on the morphology of the SnO2/
ZnO heterostructures can be explained by different inter-nanorod spacing caused by the variation in
the length of SnO2 nanowires and difference in the molecular size of VOCs. The inter-nanorod spac-
ing is quite large in short-length SnO2 nanowire (ZSO1) sensor, allowing only a small fraction of the
VOCs to be adsorbed on the surface of the heterostructures. This results in a low sensitivity toward
VOCs of the ZSO1 sample. The SnO2 nanowires in ZSO2 are longer than that of ZSO1, making
the spacing between the nanorods smaller. As a result, the surface of the heterostructures adsorbs a
relatively high fraction of different VOCs, leading to a higher response in ZSO2 compared to ZSO1.
The size of an acetone molecule (having a kinetic diameter of 0.469 nm [139]) is larger than that of
ethanol, methanol, and acetic acid molecule (having kinetic diameter of 0.45, 0.36, and 0.436 nm
[139,140], respectively). So a higher fraction of acetone molecules get absorbed, leading to a higher
acetone response in ZSO2. But at the same time, the narrow spacing between the nanorods makes
the diffusion of toluene and triethylamine molecules more difficult, due to the relatively larger
kinetic diameter of 0.585 and 0.78 nm [141,142], respectively. This results in a lower response toward
toluene and triethylamine in the ZSO2 sensor than in ZSO1. In the longest SnO2 nanowire (ZSO3)
sensor, the spacing between the nanorods further decreases, leading to a higher response toward
ethanol, methanol, and acetic acid than with ZSO2. But the diffusion of the larger acetone, toluene,
and triethylamine molecules becomes more difficult, leading to a reduced sensitivity. Therefore,
the selectivity observed in the SnO2/ZnO heterostructures can be explained by the difference in the
inter-nanorod spacing and molecular size of VOCs.

2.7.4.2.1.3   NiO-Sensitized ZnO Nanosheets To improve the sensing performance toward tri-
ethylamine, the surface of ZnO nanosheets was modified with a layer of NiO nanoparticles [143].
The interconnected nanosheets form a network-like structure, exhibiting a large surface area, which
could enhance the density of adsorption sites. This leads to the adsorption of a large fraction of
target gas molecules. The optimum operating temperature for sensing triethylamine with both ZnO
nanosheets and NiO/ZnO nanosheet sensors was found to be 320°C. The NiO/ZnO nanosheet sen-
sor exhibited a response (Ra /Rg) of 185.1–100 ppm for triethylamine at 320°C, whereas that for ZnO
nanosheets was 78.4. Thus, the response of the NiO/ZnO nanosheets was ~2.4 times higher than
that of the control ZnO. The fabricated sensors exhibited good repeatability, high stability, and low
detection limit of 2 ppm. The response and recovery characteristics of the sensors to 10 ppm tri-
ethylamine at 320°C were also studied. The sensors exhibited fast response with a response time of
6–7 s, and the recovery time of ZnO nanosheet sensor was 22 s. However, the NiO/ZnO nanosheets
required a slightly longer time of 33 s for recovery.
To examine the selectivity, the sensors were exposed to 100 ppm of different VOCs such
as triethylamine, acetone, ethanol, n-hexane, 2-propanol, p-xylene, C6H12, C6H6, and CH3OH at
320°C. The response of NiO/ZnO nanosheets for triethylamine was found to be much higher com-
pared to that for the other interfering gases, indicating its excellent selectivity toward triethylamine.
The effect of the variation of the thickness of NiO nanoparticle layer in NiO/ZnO nanosheets was
also examined. The response initially increased with the increase of the thickness of the NiO
nanoparticle layer, but decreased at very high thicknesses. The optimized NiO/ZnO nanosheet
sensor exhibited a response of 42 toward 2 ppm triethylamine.
The basic sensing mechanism of ZnO nanosheets, as shown schematically in Figure 2.25a
through d, is same as that of ZnO nanorods discussed earlier. However, the sensing mechanism of
76 Semiconductor Nanocrystals and Metal Nanoparticles

Conduction region
O2– O2–O– O2– O2– O– O2– O2–O–
Potential barrier e–

O2– O2–O–
O2– O2–O–
Ec e– e– e– e– e–
EF
Space-charge layer O2– O2– O2– O2–O– O2– O2–O– O2– O2–O–
O–
Depletion layer region
O2
(b) In air
Ev
Conduction region
(a) ZnO NS surface Air O2– O2–O–

O2– O2–O–
O2– O2–O–
e– e– e– e– e– e– e– e– e– e–

O2– O2–O–
Potential barrier –
Ec e Triethylamine Depletion layer region
EF O2– O2–O– (d) In triethylamine
Space-charge layer

CO2 + N2 + H2O O2– O2– O–

O2– O2–O– p-type NiO O2– O2– O–
Ev O2– O2– O–

O2– O2– O–
n-type ZnO NS
(c) ZnO NS surface Triethylamine
O2– O2– O– O2– O2– O– O2– O2– O–

Depletion layer region

p-type NiO Interface
(f ) In air
Ec n-type ZnO
CO2 + N2 + H2O
Ec Triethylamine
Eg = 4.2 eV
EF e– EF Electron–hole
Ev h+ O – O2– O–
Recombination h+ e– 2 p-type NiO
Eg = 3.37 eV n-type ZnO NS e–
Ev
O2– O2– O–
Depletion layer region
Electron Depletion layer CO2 + N2 + H2O
(e) Hole Triethylamine
(g) In triethylamine

FIGURE 2.25 (a) Energy band diagram and (b) schematic sensing mechanism of ZnO nanosheets in air.
(c) Energy band diagram and (d) schematic sensing mechanism of ZnO nanosheets in triethylamine. (e)
Energy band diagram of p-type NiO/n-type ZnO nanosheets. Schematic sensing mechanism of NiO/ZnO
nanosheets in (f) air and (g) triethylamine. (Reprinted from Sens. Actuators B Chem., 200, Ju, D., Xu, H.,
Qiu, Z., Guo, J., Zhang, J., and Cao, B., Highly sensitive and selective triethylamine-sensing properties of
nanosheets directly grown on ceramic tube by forming NiO/ZnO PN heterojunction, 288–296. © 2014, with
permission from Elsevier.)

NiO/ZnO nanosheets is different, due to the formation of a p–n heterojunction at the interface of
p-type NiO and n-type ZnO. When the surface of ZnO nanosheets is modified with a layer of NiO
nanoparticles, the diffusion of electrons of ZnO and holes of NiO occurs in opposite directions due
to the concentration gradient. This results in the formation of an internal electric field at the inter-
face of NiO and ZnO, and the carrier diffusion becomes balanced. The depletion layer developed at
the interface leads to the bending of energy bands till the Fermi levels of NiO and ZnO equilibrate,
as shown in Figure 2.25e. Thus, the combined effect of the depletion layer formed as a result of the
adsorbed oxygen molecules on the ZnO nanosheet surface and that formed at the interface of NiO
and ZnO leads to a higher resistance in NiO/ZnO nanosheets than the control ZnO sensor, upon
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 77

exposure to air (Figure 2.25f). When the NiO/ZnO nanosheets are exposed to triethylamine, the
adsorbed oxygen ions react with the target gas following the reaction [129]

N(C2H 5 )3 + O2 - ® H 2O + CO2 + N 2 + e - (2.13)

The electrons are thus released back to the conduction band of the nanosheets, thereby decreasing
the resistance. Also, the electrons released by triethylamine into NiO recombine with the holes,
leading to the reduction of the hole concentration of NiO. Following the law of mass action, the
electron concentration of NiO increases, which decreases the concentration gradient at the inter-
face. Thus, the depletion layer at the NiO/ZnO interfaces becomes narrow, resulting in the further
reduction of the resistance, as shown in Figure 2.25g. In addition, the bond energy of triethylamine
(C─N) is lower than that of acetone (C=O), hexane (C─C), ethanol/methanol (O─H), and benzene
(C=C), making the reaction activity of triethylamine high. This also enhances the response of NiO/
ZnO nanosheets to triethylamine. The low sensitivity of ZnO nanosheets covered by a very thick
NiO nanoparticle layer could be attributed to the reduction of the adsorption sites of ZnO.

2.7.4.2.1.4  CuO/ZnO Nanorod Heterostructures Gas sensors based on CuO/ZnO nanorod

heterostructures for the detection of H2S have been reported [144]. CuO nanoparticles were depos-
ited by a photochemical process on the surface of hydrothermally grown ZnO nanorods to form
CuO/ZnO nanorod heterostructures. The fabricated sensors exhibited a stable response to H2S with
rapid recovery in the temperature range 300°C–500°C. The sensitivity of CuO/ZnO nanorod
heterostructures increased with the increase in operating temperature from 300°C to 500°C, and
the response was found to be higher than that of control ZnO nanorods. The response and recovery
characteristics of the sensors were examined as a function of the operating temperature. With the
increase of the operating temperature, the recovery time of both CuO/ZnO heterostructures and
ZnO nanorods decreased. However, the response time of CuO/ZnO heterostructures showed a dif-
ferent trend; it increased with the rise of operating temperature. The gas sensing performance was
also studied by varying the surface coverage of ZnO nanorods by CuO nanoparticles. The response
toward H2S initially increased with the increase of the CuO surface coverage, but decreased at a
very high surface coverage.
The energy band diagram of CuO/ZnO heterostructures upon exposure to air and H2S gas is
shown in Figure 2.26a and b, respectively. When the surface of n-type ZnO nanorods is sensitized

CuO Metallic Cu2S

ZnO ZnO
ZnO ZnO
CB
CB EF
EF

VB VB

ZnO ZnO
CuO Cu2S
ZnO ZnO

(a) (b)

FIGURE 2.26 Schematic energy band diagram of CuO/ZnO heterostructures when exposed to (a) air and
(b) H2S gas. (Reprinted with permission from Kim, J., Kim, W., and Yong, K., J. Phys. Chem. C, 116, 15682–
15691, 2012. © 2012, American Chemical Society.)
78 Semiconductor Nanocrystals and Metal Nanoparticles

with p-type CuO nanoparticles, a p–n heterojunction is formed at the interface (Figure 2.26a).
The surface of the heterostructures adsorbs oxygen molecules upon exposure to air, leading to the
formation of a depletion layer and increase of resistance. When the CuO/ZnO heterostructures are
exposed to H2S gas, the gas reacts with CuO to form metallic Cu2S following the reaction [144]

6CuO(s) + 4H 2S- (ad) ® 3Cu2S(s) + 4H 2O(g) + SO2 (g) (2.14)

The formation of metallic Cu2S changes the energy band diagram, as shown in Figure 2.26b. This
leads to the reduction of the resistance of the heterostructures. When these heterostructures are
exposed to air for recovery, Cu2S reacts with the oxygen in air and converts back to CuO, following
the reaction [144]

Cu2S(s) + 2O2 ® 2CuO(s) + SO2 (g) (2.15)

As the operating temperature increases, the rate of conversion of CuO to Cu2S becomes slower,
resulting in an increase in the response time. When the surface coverage of ZnO nanorods by CuO
nanoparticles is very high, the active sensing area decreases because of the high Cu2S coverage.
Also, only a small fraction of CuO is converted to metallic Cu2S, and a large portion remains uncon-
verted. Therefore, a low sensitivity at very high CuO coverage is observed. Thus, there exists an
optimal surface coverage at which a portion of the ZnO nanorods remains exposed for H2S sensing,
and a considerable portion remains coated with CuO, for conversion to Cu2S.
Some of the reported results on gas sensors based on semiconductor sensitized ZnO nano-
structures are summarized in Table 2.3. It is clear that the gas sensing performance of ZnO nano-
structures is strongly dependent on the surface modification of ZnO nanostructures with different
inorganic semiconductors.

2.7.4.2.2 Metal-Sensitized ZnO Gas Sensors

2.7.4.2.2.1   Au-Functionalized ZnO Sensor To improve the gas sensing performance, ZnO nano-
structures have also been decorated with metal nanoparticles (Au, Pt, Pd, etc.). Metal nanoparticles
on the surface of ZnO can enhance the interaction of the gas molecules with the adsorbed oxygen
species. Gold-decorated ZnO nanowire CO gas sensors were fabricated by Chang et al. [153]. High-
density ZnO nanowires were prepared on patterned ZnO:Ga finger electrodes on SiO2/Si substrates.
Gold nanoparticles were decorated by immersing ZnO nanowire samples in the ethanol/HAuCl4
solution, and were kept under UV illumination for 4 min. The sensor measurements were performed
by injecting different concentrations of CO gas into the sealed chamber at different operating tem-
peratures. The measured sensitivities of Au/ZnO sensors operated at 350°C were around 30%,
37%, 46.5%, and 53% for 5, 20, 50, and 100 ppm concentrations of the CO gas, respectively [153].
It was also found that the Au/ZnO sensor response increased with temperature up to 250°C and then
decreased with further increase in temperature. The higher response for Au-decorated ZnO nanow-
ire device over ZnO might be attributed to the catalytic activity of Au nanoparticles at temperatures
below 250°C [153]. The lower response above 250°C might be due to the difficulty in exothermic CO
adsorption [153]. It is to be noted that the adsorbed oxygen species on the surface of ZnO play a sig-
nificant role in understanding the gas sensing mechanism for different gases and VOCs. The reducing
CO molecules react with oxygen and generate free electrons, which can be described as [153]

R + O - ® RO + e - (2.16)

where R is the reducing gas. The free electrons contribute to the electrical conduction in ZnO, lead-
ing to a decrease in resistance. For the Au/ZnO sensor, this reaction rate may be increased because
of the catalytic activity of Au nanoparticles. Thus, the Au/ZnO sensor showed higher response.
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 79

TABLE 2.3
Comparison of Sensing Performance of Semiconductor-Sensitized ZnO Nanostructures
for Detection of Various Gases and VOCs
Optimum Response Response
Operating Time (s)/
Sensitized Temperature Sensitivity, Conc. Temp. Recovery
Gas/VOC Heterostructures (°C) S (ppm) (°C) Time (s) References
Ethanol ZnO/α-Fe2O3 370 53.6172a 100 370 6/7 [145]
hierarchical
nanostructures
Ethanol Flower-like CuO/ 300 98.8a 100 300 7/9 [146]
ZnO nanorods
Ethanol PdO-decorated 320 35.4a 100 320 1/7 [147]
flower-like ZnO
structures
Ethanol ZnO–SnO2 400 280a 200 400 — [148]
core–shell
nanowires
Triethylamine NiO/ZnO 320 185.1a 100 320 7/33 [143]
nanosheets
Trimethylamine Cr2O3-decorated 400 17.79a 5 400 ~1–10/ [149]
ZnO nanowires 358–398
Acetone SnO2 nanowire/ 300 21.4a 115 300 — [129]
ZnO nanorod
heterostructures
Acetone NiO/ZnO 330 ~13a,* 100 330 — [150]
heterostructures
Nitrogen ZnO–SnO2 200 66.3b 10 200 ~50/60* [148]
dioxide core–shell
nanowires
Hydrogen ZnO–In2O3 — 15.5%c 100 Room — [151]
core–shell temp.
nanorods
Hydrogen CuO/ZnO 200 ~30d,* 5 200 360/1800 [152]
sulfide nanowires

a S = Ra/Rg.
b S = Rg/Ra.
c S = |(R − R )/R | × 100%.
g a a
d S = I /I , where I and I are current values of the sensor in air and test gas, respectively.
g a a g
* Value estimated from the reported graphical plot.

A chemical sensor based on Au-anchored ZnO hybrid material was studied by Liu et al. [154].
The dynamic response to different concentrations of ethanol at 310°C for ZnO and Au–ZnO sensors
is presented in Figure 2.27a. As compared to ZnO, the Au–ZnO sensor exhibits significant enhance-
ment in response toward each ethanol concentration, due to the presence of Au nanoparticles. The
sensitivity of the Au–ZnO sensor shows a maximum value of 52.5 at 310°C, and then decreases
with further increase in temperature [154]. It is known that the dynamic sensing reaction of a sensor
is dramatically influenced by the operating temperature. The reaction between ethanol vapor and
adsorbed oxygen species (O−, O2− and O2−) on ZnO slowly increases with temperature and reaches
80 Semiconductor Nanocrystals and Metal Nanoparticles

ZnO 600 ppm 1000 ppm

5 200 ppm
Au/ZnO

Voltage (V) 4 10 ppm

3
1 ppm
Gas out
2

1
Gas in

0
0 100 200 300 400 500
(a) Time (s)

10
Ethanol/100 ppm Gas out

7
T (°C)
350
6
340
Voltage (V)

330
5
320
310
4
300
290
3
280
270 τres
2
260
250
1
240
Gas in
0
0 50 100 150 200
(b) Time (s)

FIGURE 2.27 (a) Response of ZnO and Au/ZnO sensors to different ethanol concentrations operated at
310°C. (b) Sensing response and recovery curves of Au decorated ZnO device to 100 ppm ethanol at different
operating temperatures. (From Liu, X., Zhang, J., Guo, X., Wu, S., and Wang, S., Amino acid-assisted one-
pot assembly of Au, Pt nanoparticles onto one-dimensional ZnO microrods, Nanoscale, 2, 1178–1184, 2010.
Reproduced by permission of The Royal Society of Chemistry.)
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 81

a maximum at 310°C. Thereafter, it gradually decreases at higher temperatures. Figure 2.27b shows
the dynamic response and recovery of the Au–ZnO sensor for 100 ppm ethanol at different operating
temperatures. As observed in the figure, the response time decreases with increasing temperature
and there is a negligible effect on the recovery time. The response time is observed to be 45 and 7 s
for operating temperatures of 240°C and 280°C, respectively. Therefore, a relatively fast response
time of 3 s and recovery time of 2 min at an operating temperature exceeding 290°C suggests the
potential use of the Au NP-sensitized ZnO surface for gas sensors.
The enhanced performance of Au–ZnO sensors over ZnO might be attributed to the spillover
effect of Au nanoparticles or to the formation of an electron depletion layer at the interface of Au
and ZnO. Guo et al. have also fabricated a high-performance gas sensor based on ZnO nanowires
functionalized with Au nanoparticles and explained the sensing mechanism by the spillover effect
[155]. Figure 2.28 shows the sensing mechanism involved in the improved performance of the
Au–ZnO sensor. When the sensor is exposed to air, the oxygen molecules adsorb on the surface
of ZnO by capturing the electrons and become negatively charged oxygen species, as shown in
Figure 2.28a. These oxygen species react with the target gas molecules and provide the sensing
response (Figure 2.28b). The spillover effect of Au nanoparticles in the Au–ZnO sensor may help
adsorb more oxygen species on the surface of ZnO, which eventually enhances the interaction
with ethanol (Figure 2.28c and d). Thus, the Au–ZnO sensors showed an enhanced performance
compared to ZnO.

e– e– e– e–

e– e– e–
e–

e– e– e–
e–

e– e– e–
e–

(a) (b)

e––e– e– –
e
e–e– e––e– e–ee–
e ZnO
–
e–e– e– ee– e––e–
–
e–e– Oxygen
e– e e
– Electron depletion layer
e––e– e–ee– e––e– e–ee–
–
e e
Au
–
e––e– e–ee– e––e– e–ee–
–
Reducing gas
e e

(c) (d)

FIGURE 2.28 Schematic diagram showing the gas sensing mechanism for ZnO and Au/ZnO devices.
(a) Formation of depletion region on the surface of ZnO due to adsorption of oxygen molecules by trapping
free electrons. (b) Reaction of oxygen molecules with ethanol leads to thinner depletion layer. (c) Formation of
thicker depletion layer due to Au nanoparticles. (d) Thinner depletion layer formed as a result of the interac-
tion of oxygen molecules with ethanol. (Reprinted from Sens. Actuators B, 199, Guo, J., Zhang, J., Zhu, M.,
Ju, D., Xu, H., and Cao, B., High-performance gas sensor based on ZnO nanowires functionalized by Au
nanoparticles, 339–345. © 2014, with permission from Elsevier.)
82 Semiconductor Nanocrystals and Metal Nanoparticles

2.7.4.2.2.2  Photoinduced Au–ZnO Plasmonic Gas Sensors Photoinduced gas sensors

have attracted much interest to achieve lower operating temperatures and higher responsivity.
Room-temperature sensors have the benefits for low-power applications including good com-
patibility to integrate with other devices on flexible platform, reduced energy consumption, and
avoidance of major losses in passive electronic components. Although the UV-induced ZnO
gas sensing property has been widely investigated, its responsivity toward gas still needs to
be improved. Metal nanoparticles play a significant role in improving the sensing response for
ZnO gas sensors due to their localized surface plasmon resonance (LSPR) characteristics. The
LSPR effect mainly enhances the charge carrier density in ZnO surface via light scattering or
charge transfer.
Wavelength-tunable gas sensors were fabricated using Au-functionalized ZnO nanosheets at
room temperature [86]. The visible photoresponse of a Au–ZnO device is shown in Figure 2.29a.
The Au–ZnO device is known to display a broad photoresponse, while the ZnO device does not
show any photoresponse in the visible region. The visible photoresponse in Au–ZnO originates
from the charge excitation in Au NPs due to the decay of LSPR followed by charge transfer to the
conduction band of ZnO [156]. Therefore, the light illumination of Au–ZnO enhances the charge
density for ZnO in both UV and visible ranges, which can tune the gas sensing over a broad wave-
length range. Light-induced gas sensing study has been carried out in the presence of different gases
using photoresistance measurements. The sensor response (S) can be calculated for a particular gas
(NO) using the relation

RNO - Rair
S (%) = ´ 100 (2.17)
Rair

where
RNO is the resistance of the Au–ZnO film in the presence of NO gas
Rair is the film resistance in presence of dry air

0.4

ZnO
Au–ZnO
0.3
Responsivity (mA/W)

0.2

0.1

0.0

400 500 600 700 800

(a) Wavelength (nm)

FIGURE 2.29 (a) Photoresponsivity of plasmonic Au–ZnO device in the visible region at 5 V. (Continued)
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 83

800 350 nm Au–ZnO

Off Off On
On On Off
400

0
Response (%)
400 335 nm Au–ZnO

200

0
335 nm ZnO
300
200 6 ppm
4 ppm
100 2 ppm

0
0 1000 2000 3000 4000 5000 6000
(b) Time (s)

500 2 ppm
550 nm Au–ZnO
400
300
200
100
200 655 nm 2 ppm
Response (%)

160
120
80
40
0
120 725 nm 2 ppm

80
Off
40 On
0
0 500 1000 1500 2000 2500
(c) Time (s)

FIGURE 2.29 (Continued) (b) Sensing response to different NO molecule concentration of Au–ZnO and
ZnO devices on illumination with UV wavelengths at room temperature. (From Gogurla, N. et al., Sci. Rep.,
4, 6483(1–9), 2014.) (c) Response of Au–ZnO device to 2 ppm NO concentration at room temperature on
illumination with visible wavelengths. (From Gogurla, N. et al., Sci. Rep., 4, 6483(1–9), 2014.)

Figure 2.29b shows the room-temperature gas sensing response to different concentration of
NO gas in dry air for both ZnO and Au–ZnO devices under UV illumination of wavelengths
335 and 350 nm. The time pulse sequence was maintained as “ON” for 15 min and “OFF” for
15 min for both the devices. As seen in the figure, both the devices show a prominent change in
the response even at very low concentration levels. The response for Au–ZnO device upon expo-
sure to 4 ppm NO is found to be enhanced three times higher than the ZnO device at 325 nm.
84 Semiconductor Nanocrystals and Metal Nanoparticles

The sensing response of Au–ZnO device at different incident visible wavelengths at room tem-
perature is also presented in Figure 2.29c. The response at a wavelength of 550 nm is found to
be the maximum for the device.
The sensitivity as a function of wavelength of the Au–ZnO device at 2 ppm concentration of
NO is presented in Figure 2.30a. The device exhibits a higher response at 550 nm illumination
compared to other wavelengths. The sensitivity of the Au–ZnO device at 2 ppm is observed to be
194 and 50%/ppm for 550 and 335 nm illumination, respectively. The sensitivity at 550 nm is also
higher as compared to the reported value (0.326%/ppm at 532 nm) of Au–ZnO sensor for 25 ppm
C2H2 [157]. A histogram summarizing the selectivity of the Au–ZnO sensor to NO gas over CO and
some selected VOCs is shown in Figure 2.30b. The Au–ZnO sensor exhibits much higher selectivity
for NO compared to other gases measured at 335 nm. The limit of detection for the Au–ZnO sen-
sor for NO is found to be 0.1 ppb under 335 nm illumination. The ability of very low concentration
detection with high selectivity makes the photoconductive Au–ZnO nanocomposite attractive for
breath sensing applications, where NO is a signature for asthma disease. Thus, the sensitivity and
low detection limit of plasmonic devices operated at room temperature are comparable or even
better in magnitude than the reported values for conventional sensors using metal oxides operating
at higher temperatures [158,159].

2.7.4.2.2.3   Sensing Mechanism of Photoinduced ZnO Sensors The sensing mechanism of the
light-induced Au–ZnO sensor can be explained by the photoconductive property of ZnO in the UV
region and LSPR effect of Au nanoparticles in the visible region. Due to the large surface-to-volume
ratio of ZnO nanostructures, oxygen molecules create a depletion layer with lower conductivity by
capturing electrons: O2(gas) + e− → O2−(adsorption). Upon illumination, the photogenerated holes

200
Au–ZnO
180 550 nm
at 2 ppm
160
Sensitivity (%/ppm)

140

120

100

80

60

40

300 400 500 600 700 800

(a) Wavelength (nm)

FIGURE 2.30 (a) Sensitivity as a function of wavelength of Au–ZnO device at 2 ppm NO concentration.
(From Gogurla, N. et al., Sci. Rep., 4, 6483(1–9), 2014.) (Continued)
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 85

100
0.025

0.020
80

Sensitivity (%/ppm)
0.015
Sensitivity (%/ppm)

60 0.010

0.005
40
0.000
CO Ethanol Chloroform Chlorobenzene Toluene

20 VOCs

0
(b) NO CO Ethanol Chloroform Chlorobenzene Toluene

N O N O

hν

N O OO OO
N O
e e e
e e
e e
LSPR
Ec

Au

Ev
ZnO

(c)

FIGURE 2.30 (Continued) (b) Histogram plot of sensitivity of Au–ZnO device to different gases. Inset
shows the sensitivity of the sensor upon exposure to CO and VOCs. (c) Schematic diagram representing
the sensing mechanism of Au–ZnO device in the presence of light. (From Gogurla, N. et al., Sci. Rep., 4,
6483(1–9), 2014.)

in ZnO migrate to the surface along the potential gradient and desorb oxygen from the surface
(O2− (adsorption) + h+ → O2(gas)). This results in an increase in the free carrier concentration
in ZnO and a decrease in the width of the depletion layer. When the gas flow is switched “ON,”
because of the electron affinity of NO molecules, they can easily react with the photoelectrons
of ZnO. The reaction can be expressed by [86]

2NO + e - = N 2 (g) + O2 - (2.18)

86 Semiconductor Nanocrystals and Metal Nanoparticles

Here, the NO molecules interact with photoelectrons and generate O2 molecules, which are again
adsorbed on the surface of ZnO by capturing the photogenerated electrons, as shown in F­ igure 2.30c.
As a result, there is an increase in the photoresistance of the ZnO sensor due to the broadening of
surface depletion layer again. But for the Au–ZnO sensor, under the UV illumination, the excited
electrons of Au NPs due to interband transition also participate in the reaction. These excited elec-
trons in Au can be transferred to ZnO, or they can directly interact with the NO molecules. The
enhanced interaction with NO molecules occurring as a result of the increased charge density in
ZnO may lead to a higher photoresistance for Au–ZnO NO sensor.
The mechanism involved in the sensing in the visible region can be explained as due only to
the LSPR effect of Au nanoparticles. When light interacts with Au nanoparticles, more electrons
oscillate on the surface of Au due to plasmon resonance. The electron oscillation is more sensitive
to the dielectric environment and charge density of Au [160]. The strong electron affinity of the
NO molecule causes a strong interaction with the oscillating electrons of Au (Figure 2.30c). As
a result, there is decrease in the charge density of Au, which finally increases the photoresistance
of the film. It may be noted that the sensing response of the Au–ZnO sensor in the visible region
due to surface plasmons is even higher than in the UV region. These results imply the potential of
metal-functionalized ZnO for plasmonic gas sensors operated at room temperature and tunable over
different wavelengths.
UV illumination–assisted gas sensors based on ZnO nanofibers decorated with Au nanopar-
ticles were also fabricated for detecting ethanol at room temperature [161]. Under UV illumination
(365 and 254 nm), the Au–ZnO sensor showed a higher response upon exposure to ethanol than
that of pure ZnO nanofiber sensor operated at room temperature. The sensing mechanism involved
in enhancing the performance of the Au–ZnO sensor could be explained by the photocatalytic
­reaction. Due to the Schottky junction between Au and ZnO, the photogenerated electron–hole pairs
are separated and provide more charge carriers on the surface of ZnO for reaction. The photoxida-
tion rate of organic vapor on the ZnO surface can also be improved by the catalytic action of Au
nanoparticles.

2.7.5 Biosensors Using ZnO Nanostructures

ZnO nanostructures are also attractive for biosensing applications owing to their high surface-to-
volume ratio, high surface reactivity and catalytic efficiency, fast electron transport, and the ability
to adsorb and retain the activity of enzymes. The activity of the immobilized enzymes strongly
depends on the temperature, toxic chemicals, pH, and humidity [162]. ZnO, having a high isoelectric
point (IEP) of 9.5, presents a suitable electrode surface for immobilization of low-IEP enzymes such
as glucose oxidase (GOx; IEP ~4.2). The positively charged ZnO nanostructures are capable of eas-
ily immobilizing negatively charged GOx at a physiological pH of 7.4, due to the electrostatic attrac-
tive force between them. This force promotes the efficient immobilization of GOx in mild chemical
conditions while still retaining the enzymic bioactivity. Moreover, the high specific surface area of
ZnO nanostructures enables high loading of enzyme, and also assists in efficient transfer of elec-
trons between the electrode and GOx.
ZnO nanostructures of different morphologies have been used for fabricating glucose biosensors.
Glucose biosensors based on inverse opals showed a broad, linear detection range of 0.01–18 mM,
with a sensitivity of 22.5 μA/cm2/mM [163]. ZnO nanotube–based glucose biosensor exhibited a
sensitivity of 30.4 μA/cm2/mM, with a response time of less than 10 s [12]. The glucose biosensor
fabricated using a single ZnO nanofiber exhibited a high sensitivity of 70.2 μA/cm2/mM, with a fast
response time of <4 s and a low detection limit of 1 μM [164]. Glucose sensors were also fabricated
using ZnO tripods and nanorods [165]. ZnO tripods, synthesized by the vapor–solid method, were
distributed uniformly over the Si substrate, as shown in Figure 2.31a. ZnO tripods and nanorods
became interconnected with their neighboring ones after immobilization of GOx on their surfaces.
Figure 2.31b shows the amperometric response of ZnO nanorod- and tripod-based glucose sensors.
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 87

(a)

1 μm

20 μm

1.8 × 10–4
(B)
1.6 × 10–4 At V = 2V
26.4 mM
1.4 × 10–4
(A)
1.2 × 10–4 19.86 mM
Current (A)

1.0 × 10–4 26.4 mM

8.0 × 10–5 13.29 mM 19.86 mM

–5
6.0 × 10
13.29 mM
4.0 × 10–5 6.66 mM

2.0 × 10–5 6.66 mM

0.0
GOD attached ZnO

0 200 400 600 800 1000

(b) Time (s)

FIGURE 2.31 (a) Typical FESEM micrograph of ZnO tripods. Inset shows the magnified view of the micro-
graph. (b) Amperometric response of biosensors based on ZnO (A) nanorods and (B) tripods with successive
addition of glucose. (From Mandal, S. et al., Sens. Lett., 7, 635, 2009.)

Both sensors show a fast response on the addition of glucose, reaching 90% of the steady-state cur-
rent in <10 s. This indicates that a rapid exchange of electrons occurs between the ZnO nanostruc-
tures and GOx. The sensitivity of the ZnO tripod- and nanorod-based glucose sensor is found to be
36.8 and 27.7 μA/cm2/mM, respectively. The high specific surface area of ZnO tripods allows higher
loading of GOx, resulting in an enhanced sensitivity compared to those of ZnO nanorods. Thus,
the morphology of ZnO nanostructures plays an important role in controlling the performance of
biosensors.

2.8 SUMMARY
In this chapter, we presented a review on ZnO-based nanostructures sensitized by semiconduc-
tors and metal nanoparticles for optoelectronic and gas sensor applications. The optical properties
of ZnO tripods and tetrapods were discussed using low-temperature photoluminescence measure-
ments. The optical and electrical properties of ZnO tetrapods can be tuned by doping with selective
88 Semiconductor Nanocrystals and Metal Nanoparticles

elements such as Sn, P, Mg, Mn, and so on, which may be useful for different optoelectronic applica-
tions. The photovoltaic and gas sensing performance can be improved by modifying the surface of
ZnO nanostructures with selective inorganic semiconductors. The influence of metal nanoparticles
on the performance of ZnO-based UV photodetectors was also discussed. The responsivity of ZnO
devices could be enhanced 80 times with sensitized Au nanoparticles on the ZnO nanosheet sur-
face. The interband transition and hole trapping in Au nanoparticles may be the possible reasons
for enhanced photoresponse. Other metal nanoparticles (Al and Pt) are also used to enhance the
photoresponse of ZnO UV photodetectors. The observed enhancement values for these detectors
are found to be lower than those obtained for Au–ZnO nanocomposite devices. Metal nanoparticles
can also be used to improve the performance of ZnO gas sensors. These sensors exhibit enhanced
performance due to the LSPR effect of metal nanoparticles.

ACKNOWLEDGMENT
The academic contributions of Prof. I. Manna, Dr. S. Santra, and Dr. S. Mandal are gratefully
acknowledged for several research results reported in this chapter.

REFERENCES
1. Pearton, S. J., Norton, D. P., Ip, K., Heo, Y. W., and Steiner, T. 2003. Recent progress in processing and
properties of ZnO. Superlattices Microstruct. 34:3–32.
2. Ashrafi, A. and Jagadish, C. 2007. Review of zincblende ZnO: Stability of metastable ZnO phases.
J. Appl. Phys. 102:071101(1–12).
3. Kim, K.-K., Kim, H.-S., Hwang, D.-K., Lim, J.-H., and Park, S.-J. 2003. Realization of p-type ZnO thin
films via phosphorus doping and thermal activation of the dopant. Appl. Phys. Lett. 83:63–65.
4. Chavillon, B., Cario, L., Renaud, A. et al. 2012. P-type nitrogen-doped ZnO nanoparticles stable under
ambient conditions. J. Am. Chem. Soc. 134:464–470.
5. Fan, J. C., Zhu, C. Y., Fung, S. et al. 2009. Arsenic doped p-type zinc oxide films grown by radio frequency
magnetron sputtering. J. Appl. Phys. 106:073709(1–6).
6. Jun, M.-C., Park, S.-U., and Koh, J.-H. 2012. Comparative studies of Al-doped ZnO and Ga-doped ZnO
transparent conducting oxide thin films. Nanoscale Res. Lett. 7:639(6pp).
7. Kim, D. H., Cho, N. G., Kim, H. G., and Choi, W.-Y. 2007. Structural and electrical properties of indium
doped ZnO thin films fabricated by RF magnetron sputtering. J. Electrochem. Soc. 154:H939–H943.
8. Lee, J. and Tak, Y. 2001. Electrodeposition of ZnO on ITO electrode by potential modulation method.
Electrochem. Solid State Lett. 4:C63–C65.
9. Li, Y., Meng, G. W., Zhang, L. D., and Phillipp, F. 2000. Ordered semiconductor ZnO nanowire arrays
and their photoluminescence properties. Appl. Phys. Lett. 76:2011–2013.
10. Lakshmi, B. B., Dorhout, P. K., and Martin, C. R. 1997. Sol–gel template synthesis of semiconductor
nanostructures. Chem. Mater. 9:857–862.
11. Chen, Y. W. and Liu, Y. C. 2005. Optical properties of ZnO and ZnO:In nanorods assembled by sol–gel
method. J. Chem. Phys. 123:134701-1–134701-5.
12. Rakshit, T., Mandal, S., Mishra, P., Dhar, A., Manna, I., and Ray, S. K. 2012. Optical and bio-sensing
characteristics of ZnO nanotubes grown by hydrothermal method. J. Nanosci. Nanotechnol. 12:308–315.
13. Park, J.-A., Moon, J., Lee, S.-J., Lim, S.-C., and Zyung, T. 2009. Fabrication and characterization of
ZnO nanofibers by electrospinning. Curr. Appl. Phys. 9:S210–S212.
14. Rajesh, D., Lakshmi, B. V., and Sunandana, C. S. 2012. Two-step synthesis and characterization of
ZnO nanoparticles. Physica B 407:4537–4539.
15. Wu, J.-J. and Liu, S.-C. 2002. Low-temperature growth of well-aligned ZnO nanorods by chemical
vapor deposition. Adv. Mater. 14:215–218.
16. Zak, A. K., Majid, W. H. A., Wang, H. Z., Yousefi, R., Golsheikh, A. M., and Ren, Z. F. 2013.
Sonochemical synthesis of hierarchical ZnO nanostructures. Ultrason. Sonochem. 20:395–400.
17. Kooti, M. and Sedeh, A. N. 2013. Microwave-assisted combustion synthesis of ZnO nanoparticles.
J. Chem. 2013:1–4, Article ID: 562028.
18. Shetty, A. and Kar Nanda, K. 2012. Synthesis of zinc oxide porous structures by anodization with water
as an electrolyte. Appl. Phys. A 109:151–157.
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 89

19. Kumar, S. S., Venkateswarlu, P., Rao, V. R., and Rao, G. N. 2013. Synthesis, characterization and optical
properties of zinc oxide nanoparticles. Int. Nano Lett. 3:30-1–30-6.
20. Kong, X. Y. and Wang, Z. L. 2003. Spontaneous polarization-induced nanohelixes, nanosprings, and
nanorings of piezoelectric nanobelts. Nano Lett. 3:1625–1631.
21. Chang, P.-C., Fan, Z., Wang, D. et al. 2004. ZnO nanowires synthesized by vapor trapping CVD method.
Chem. Mater. 16:5133–5137.
22. Dang, H. Y., Wang, J., and Fan, S. S. 2003. The synthesis of metal oxide nanowires by directly heating
metal samples in appropriate oxygen atmospheres. Nanotechnology 14:738–741.
23. Huang, M. H., Wu, Y., Feick, H., Tran, N., Weber, E., and Yang, P. 2001. Catalytic growth of zinc oxide
nanowires by vapor transport. Adv. Mater. 13:113–116.
24. Park, W. I., Kim, D. H., Jung, S.-W., and Yi, G.-C. 2002. Metalorganic vapor-phase epitaxial growth of
vertically well-aligned ZnO nanorods. Appl. Phys. Lett. 80:4232–4234.
25. Wang, Z. L. 2004. Zinc oxide nanostructures: Growth, properties and applications. J. Phys.: Condens.
Matter 16:R829–R858.
26. Mandal, S., Dhar, A., and Ray, S. K. 2009. Growth and photoluminescence characteristics of ZnO
tripods. J. Appl. Phys. 105:033513-1–033513-6.
27. Wang, Z. L. 2004. Nanostructures of zinc oxide. Mater. Today 7:26–33.
28. Rakshit, T., Manna, I., and Ray, S. K. 2013. Shape controlled Sn doped ZnO nanostructures for tunable
optical emission and transport properties. AIP Adv. 3:112112-1–112112-12.
29. Jayanthi, K., Chawla, S., Sood, K. N., Chhibara, M., and Singh, S. 2009. Dopant induced morphology
changes in ZnO nanocrystals. Appl. Surf. Sci. 255:5869–5875.
30. Shen, G., Cho, J. H., Yoo, J. K., Yi, G.-C., and Lee, C. J. 2005. Synthesis and optical properties of
S-doped ZnO nanostructures: Nanonails and nanowires. J. Phys. Chem. B 109:5491–5496.
31. Pan, H., Zhu, Y., Sun, H., Feng, Y., Sow, C.-H., and Lin, J. 2006. Electroluminescence and field emission
of Mg-doped ZnO tetrapods. Nanotechnology 17:5096–5100.
32. Roy, N. and Roy, A. 2015. Growth and temperature dependent photoluminescence characteristics of
ZnO tetrapods. Ceram. Int. 41:4154–4160.
33. Yu, D., Hu, L., Qiao, S. et al. 2009. Photoluminescence study of novel phosphorus-doped ZnO nano­
tetrapods synthesized by chemical vapour deposition. J. Phys. D: Appl. Phys. 42:055110-1–055110-6.
34. Rackauskas, S., Mustonen, K., Järvinen, T. et al. 2012. Synthesis of ZnO tetrapods for flexible and
transparent UV sensors. Nanotechnology 23:095502-1–095502-7.
35. Roy, V. A. L., Djurišić, A. B., Liu, H. et al. 2004. Magnetic properties of Mn doped ZnO tetrapod
structures. Appl. Phys. Lett. 84:756–758.
36. Huh, J., Kim, G.-T., Lee, J. S., and Kim, S. 2008. A direct measurement of the local resistances in a ZnO
tetrapod by means of impedance spectroscopy: The role of the junction in the overall resistance. Appl.
Phys. Lett. 93:042111-1–042111-3.
37. Zhou, X., Lin, T., Liu, Y. et al. 2013. Structural, optical, and improved field-emission properties of
tetrapod-shaped Sn-doped ZnO nanostructures synthesized via thermal evaporation. ACS Appl. Mater.
Interfaces 5:10067–10073.
38. Zhao, L. and Hu, L. 2012. Synthesis and applications of CdSe nano-tetrapods in hybrid photovoltaic
devices. Pure Appl. Chem. 84:2549–2558.
39. Sun, B., Marx, E., and Greenham, N. C. 2003. Photovoltaic devices using blends of branched CdSe
nanoparticles and conjugated polymers. Nano Lett. 3:961–963.
40. Reynolds, D. C., Look, D. C., Jogai, B. et al. 1998. Neutral-donor–bound-exciton complexes in ZnO
crystals. Phys. Rev. B 57:12151–12155.
41. Teke, A., Özgür, Ü., Doğan, S. et al. 2004. Excitonic fine structure and recombination dynamics in
single-crystalline ZnO. Phys. Rev. B 70:195207-1–195207-10.
42. Jiang, D. S., Jung, H., and Ploog, K. 1988. Temperature dependence of photoluminescence from GaAs
single and multiple quantum-well heterostructures grown by molecular-beam epitaxy. J. Appl. Phys.
64:1371–1377.
43. Lautenschlager, P., Garriga, M., Logothetidis, S., and Cardona, M. 1987. Interband critical points of
GaAs and their temperature dependence. Phys. Rev. B 35:9174–9189.
44. Wang, L. and Giles, N. C. 2003. Temperature dependence of the free-exciton transition energy in zinc
oxide by photoluminescence excitation spectroscopy. J. Appl. Phys. 94:973–978.
45. Matsumoto, T., Kato, H., Miyamoto, K., Sano, M., Zhukov, E. A., and Yao, T. 2002. Correlation between
grain size and optical properties in zinc oxide thin films. Appl. Phys. Lett. 81:1231–1233.
46. Liang, W. Y. and Yoffe, A. D. 1968. Transmission spectra of ZnO single crystals. Phys. Rev. Lett.
20:59–62.
90 Semiconductor Nanocrystals and Metal Nanoparticles

47. Zhang, Y., Lin, B., Sun, X., and Fu, Z. 2005. Temperature-dependent photoluminescence of nano-
crystalline ZnO thin films grown on Si(100) substrates by the sol–gel process. Appl. Phys. Lett.
86:131910-1–131910-3.
48. Meyer, B. K., Alves, H., Hofmann, D. M. et al. 2004. Bound exciton and donor–acceptor pair recombi-
nations in ZnO. Phys. Stat. Sol. (b) 241:231–260.
49. Kurbanov, S. S. and Kang, T. W. 2010. Spectral behavior of the emission around 3.31 eV (A-line) from
ZnO nanocrystals. J. Lumin. 130:767–770.
50. Kurbanov, S. S., Panin, G. N., and Kang, T. W. 2009. Spatially resolved investigations of the emission
around 3.31 eV (A-line) from ZnO nanocrystals. Appl. Phys. Lett. 95:211902-1–211902-3.
51. Hwang, D.-K., Kim, H.-S., Lim, J.-H. et al. 2005. Study of the photoluminescence of phosphorus-
doped p-type ZnO thin films grown by radio-frequency magnetron sputtering. Appl. Phys. Lett.
86:151917-1–151917-3.
52. Janotti, A. and Van de Walle, C. G. 2007. Native point defects in ZnO. Phys. Rev. B 76:165202-1–165202-22.
53. Ma, L. A. and Guo, T. L. 2013. Morphology control and improved field emission properties of ZnO
tetrapod films deposited by electrophoretic deposition. Ceram. Int. 39:6923–6929.
54. Lee, G.-H. 2011. Synthesis and cathodoluminescence of ZnO tetrapods prepared by a simple oxidation
of Zn powder in air atmosphere. Ceram. Int. 37:189–193.
55. Wan, Q., Yu, K., Wang, T. H., and Lin, C. L. 2003. Low-field electron emission from tetrapod-like ZnO
nanostructures synthesized by rapid evaporation. Appl. Phys. Lett. 83:2253–2255.
56. Das, S. N., Kar, J. P., Choi, J.-H., Byeon, S., Jho, Y. D., and Myoung, J.-M. 2009. Influence of sur-
face morphology on the optical property of vertically aligned ZnO nanorods. Appl. Phys. Lett.
95:111909-1–111909-3.
57. Ahn, C. H., Mohanta, S. K., Lee, N. E., and Cho, H. K. 2009. Enhanced exciton–phonon interactions in
photoluminescence of ZnO nanopencils. Appl. Phys. Lett. 94:261904-1–261904-3.
58. Gu, X., Huo, K., Qian, G., Fu, J., and Chu, P. K. 2008. Temperature dependent photoluminescence from
ZnO nanowires and nanosheets on brass substrate. Appl. Phys. Lett. 93:203117-1–203117-3.
59. Wang, X., Ding, Y., Summers, C. J., and Wang, Z. L. 2004. Large-scale synthesis of six-nanometer-wide
ZnO nanobelts. J. Phys. Chem. B 108:8773–8777.
60. Lu, J. G., Ye, Z. Z., Zhang, Y. Z., Liang, Q. L., Fujita, S., and Wang, Z. L. 2006. Self-assembled ZnO
quantum dots with tunable optical properties. Appl. Phys. Lett. 89:023122-1–023122-3.
61. Fan, Z., Wang, D., Chang, P.-C., Tseng, W.-Y., and Lu, J. G. 2004. ZnO nanowire field-effect transistor
and oxygen sensing property. Appl. Phys. Lett. 85:5923–5925.
62. Yun, Y. S., Park, J. Y., Oh, H., Kim, J.-J., and Kim, S. S. 2006. Electrical transport properties of
size-tuned ZnO nanorods. J. Mater. Res. 21:132–136.
63. Lee, C. J., Lee, T. J., Lyu, S. C., Zhang, Y., Ruh, H., and Lee, H. J. 2002. Field emission from well-aligned
zinc oxide nanowires grown at low temperature. Appl. Phys. Lett. 81:3648–3650.
64. Zhu, Y. W., Zhang, H. Z., Sun, X. C. et al. 2003. Efficient field emission from ZnO nanoneedle arrays.
Appl. Phys. Lett. 83:144–146.
65. Park, H., Chang, S., Jean, J. et al. 2012. Graphene cathode-based ZnO nanowire hybrid solar cells. Nano
Lett. 13:233–239.
66. Tian, W., Zhang, C., Zhai, T. et al. 2014. Flexible ultraviolet photodetectors with broad photoresponse
based on branched ZnS–ZnO heterostructure nanofilms. Adv. Mater. 26:3088–3093.
67. Zhu, G. Y., Xu, C. X., Lin, Y. et al. 2012. Ultraviolet electroluminescence from horizontal ZnO microrods/
GaN heterojunction light-emitting diode array. Appl. Phys. Lett. 101:041110(1–4).
68. Lee, S.-H., Han, S.-H., Jung, H. S. et al. 2010. Al-Doped ZnO Thin film: A new transparent conducting
layer for ZnO nanowire-based dye-sensitized solar cells. J. Phys. Chem. C 114:7185–7189.
69. Chen, B., Meng, C., Yang, Z. et al. 2014. Graphene coated ZnO nanowire optical waveguides. Opt.
Express 22:24276–24285.
70. Li, C., Li, C., Di, Y., Lei, W., Chen, J., and Cui, Y. 2013. ZnO electron field emitters on three-dimensional
patterned carbon nanotube framework. ACS Appl. Mater. Interfaces 5:9194–9198.
71. Wang, X., Zhou, J., Song, J., Liu, J., Xu, N., and Wang, Z. L. 2006. Piezoelectric field effect transistor
and nanoforce sensor based on a single ZnO nanowire. Nano Lett. 6:2768–2772.
72. Hoffmann, M. W. G., Mayrhofer, L., Casals, O. et al. 2014. A highly selective and self-powered gas
sensor via organic surface functionalization of p-Si/n-ZnO diodes. Adv. Mater. 26:8017–8022.
73. Yang, Y., Sun, X. W., Tay, B. K., You, G. F., Tan, S. T., and Teo, K. L., 2008. A p-n homojunction ZnO
nanorod light-emitting diode formed by As ion implantation. Appl. Phys. Lett. 93:253107(1–3).
74. Chu, S., Wang, G., Zhou, W. et al. 2011. Electrically pumped waveguide lasing from ZnO nanowires.
Nat. Nanotechnol. 6:506–510.
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 91

75. An, S. J. and Yi, G.-C. 2007. Near ultraviolet light emitting diode composed of n-GaN/ZnO coaxial
nanorod heterostructures on a p-GaN layer. Appl. Phys. Lett. 91:123109(1–3).
76. Dong, J. J., Zhang, X. W., and Yin, Z. G. 2012. Ultraviolet electroluminescence from ordered ZnO
nanorod array/p-GaN light emitting diodes. Appl. Phys. Lett. 100:171109(1–4).
77. Bai, S., Wu, W., Qin, Y., Cui, N., Bayerl, D. J., and Wang, X. 2011. High-performance integrated ZnO
nanowire UV sensors on rigid and flexible substrates. Adv. Funct. Mater. 21:4464–4469.
78. Liu, N., Fang, G., Zeng, W. et al. 2010. Direct growth of lateral ZnO nanorod UV photodetectors with
schottky contact by a single-step hydrothermal reaction. ACS Appl. Mater. Interfaces 7:1973–1979.
79. Kind, H., Yan, H., Messer, B., Law, M., and Yang, P. 2002. Nanowire ultraviolet photodetectors and
optical switches. Adv. Mater. 14:158–160.
80. Soci, C., Zhang, A., Xiang, B. et al. 2007. ZnO nanowire UV photodetectors with high internal gain.
Nano Lett. 7:1003–1009.
81. Dai, J., Xu, C., Xu, X. et al. 2013. Single ZnO microrod ultraviolet photodetector with high photocurrent
gain. ACS Appl. Mater. Interfaces 5:9344–9348.
82. Cheng, G., Wu, X., Liu, B., Li, B., Zhang, X., and Du, Z. 2011. ZnO nanowire Schottky barrier ultravio-
let photodetector with high sensitivity and fast recovery speed. Appl. Phys. Lett. 99:203105(1–3).
83. Wang, Z., Yu, R., and Wen, X., 2014. Optimizing performance of silicon-based p-n junction photodetec-
tors by the piezo-phototronic effect. ACS Nano 8:12866–12873.
84. Afal, A., Coskun, S., and Unalan, H. E. 2013. All solution processed, nanowire enhanced ultraviolet
photodetectors. Appl. Phys. Lett. 102:043503(1–5).
85. van de Hulst, H. C. 1981. Light Scattering by Small Particles. New York: Dover Publications.
86. Gogurla, N., Sinha, A. K., Santra, S., Manna, S., and Ray, S.K. 2014. Multifunctional Au–ZnO
plasmonic nanostructures for enhanced UV photodetector and room temperature NO sensing devices.
Sci. Rep. 4:6483(1–9).
87. Li, L., Wu, P., Fang, X. et al. 2010. Single-crystalline CdS nanobelts for excellent field-emitters and
ultrahigh quantum-efficiency photodetectors. Adv. Mater. 22:3161–3165.
88. Bao, G., Li, D., Sun, X. et al. 2014. Enhanced spectral response of an AlGaN-based solar-blind ultravio-
let photodetector with Al nanoparticles. Opt. Express 22:24286–24293.
89. Liu, Y., Zhang, X., Su, J., Li, H., Zhang, Q., and Gao, Y. 2014. Ag nanoparticles@ZnO nanowire com-
posite arrays: An absorption enhanced UV photodetector. Opt. Express 22:30148–30155.
90. Xie, X. N., Zhong, Y. L., Dhoni, M. S. et al. 2010. UV-visible-near infrared photoabsorption and photo-
detection using close-packed metallic gold nanoparticle network. J. Appl. Phys. 107:053510(1–6).
91. Jin, Z., Gao, L., Zhou, Q., and Wang, J. 2014. High-performance flexible ultraviolet photoconductors
based on solution-processed ultrathin ZnO/Au nanoparticle composite films. Sci. Rep. 4:4268(1–8).
92. Lu, J., Xu, C., Dai, J. et al. 2015. Improved UV photoresponse of ZnO nanorod arrays by resonant cou-
pling with surface plasmons of Al nanoparticles. Nanoscale 7:3396–3403.
93. Aguilar, C. A., Haight, R., Mavrokefalos, A., Korgel, B. A., and Chen, S. 2009. Probing electronic
properties of molecular engineered zinc oxide nanowires with photoelectron spectroscopy. ACS Nano
3:3057–3062.
94. Tian, C., Jiang, D., Li, B. et al. 2014. Performance enhancement of ZnO UV photodetectors by surface
plasmons. ACS Appl. Mater. Interfaces 6:2162–2166.
95. Zhang, Q., Dandeneau, C. S., Zhou, X., and Cao, G. 2009. ZnO nanostructures for dye-sensitized solar
cells. Adv. Mater. 21:4087–4108.
96. Xu, F. and Sun, L. 2011. Solution-derived ZnO nanostructures for photoanodes of dye-sensitized solar
cells. Energy Environ. Sci. 4:818–841.
97. Chiba, Y., Islam, A., Watanabe, Y., Komiya, R., Koide, N., and Han, L. 2006. Dye-sensitized solar cells
with conversion efficiency of 11.1%. Jpn. J. Appl. Phys. 45:L638–L640.
98. Wu, F., Shen, W., Cui, Q. et al. 2010. Dynamic characterization of hybrid solar cells based on poly-
mer and aligned ZnO nanorods by intensity modulated photocurrent spectroscopy. J. Phys. Chem. C
114:20225–20235.
99. Baeten, L., Conings, B., Boyen, H.-G. et al. 2011. Towards efficient hybrid solar cells based on fully
polymer infiltrated ZnO nanorod arrays. Adv. Mater. 23:2802–2805.
100. Olson, D. C., Lee, Y.-J., White, M. S. et al. 2007. Effect of polymer processing on the performance of
poly(3-hexylthiophene)/ZnO nanorod photovoltaic devices. J. Phys. Chem. C 111:16640–16645.
101. Olson, D. C., Lee, Y.-J., White, M. S. et al. 2008. Effect of ZnO processing on the photovoltage of ZnO/
poly(3-hexylthiophene) solar cells. J. Phys. Chem. C 112:9544–9547.
102. Rakshit, T., Mondal, S. P., Manna, I., and Ray, S. K. 2012. CdS-decorated ZnO nanorod heterostructures
for improved hybrid photovoltaic devices. ACS Appl. Mater. Interfaces 4:6085–6095.
92 Semiconductor Nanocrystals and Metal Nanoparticles

103. Saunders, B. R. 2012. Hybrid polymer/nanoparticle solar cells: Preparation, principles and challenges.
J. Colloid Interface Sci. 369:1–15.
104. Sandberg, H. G. O., Frey, G. L., Shkunov, M. N. et al. 2002. Ultrathin regioregular poly(3-hexyl thio-
phene) field-effect transistors. Langmuir 18:10176–10182.
105. Wang, L., Zhao, D., Su, Z., and Shen, D. 2012. Hybrid polymer/ZnO solar cells sensitized by PbS quan-
tum dots. Nanoscale Res. Lett. 7:106-1–106-6.
106. Kim, H., Jeong, H., An, T. K., Park, C. E., and Yong, K. 2013. Hybrid-type quantum-dot cosensi-
tized ZnO nanowire solar cell with enhanced visible-light harvesting. ACS Appl. Mater. Interfaces
5:268–275.
107. Spoerke, E. D., Lloyd, M. T., McCready, E. M., Olson, D. C., Lee, Y.-J., and Hsu, J. W. P. 2009. Improved
performance of poly(3-hexylthiophene)/zinc oxide hybrid photovoltaics modified with interfacial nano-
crystalline cadmium sulfide. Appl. Phys. Lett. 95:213506-1–213506-3.
108. O’Regan, B. C., Scully, S., Mayer, A. C., Palomares, E., and Durrant, J. 2005. The effect of Al2O3 bar-
rier layers in TiO2/dye/CuSCN photovoltaic cells explored by recombination and DOS characterization
using transient photovoltage measurements. J. Phys. Chem. B 109:4616–4623.
109. Wu, S., Li, J., Lo, S.-C., Tai, Q., and Yan, F. 2012. Enhanced performance of hybrid solar cells based on
ordered electrospun ZnO nanofibers modified with CdS on the surface. Org. Electron. 13:1569–1575.
110. Wang, L., Zhao, D., Su, Z., Li, B., Zhang, Z., and Shen, D. 2011. Enhanced efficiency of polymer/ZnO
nanorods hybrid solar cell sensitized by CdS quantum dots. J. Electrochem. Soc. 158:H804–H807.
111. Cui, Q., Liu, C., Wu, F. et al. 2013. Performance improvement in polymer/ZnO nanoarray hybrid solar
cells by formation of ZnO/CdS-core/shell heterostructures. J. Phys. Chem. C 117:5626–5637.
112. Zhu, C., Pan, X., Ye, C., Wang, L., Ye, Z., and Huang, J. 2013. Effect of CdSe quantum dots on the
performance of hybrid solar cells based on ZnO nanorod arrays. Ceram. Int. 39:2975–2980.
113. Liu, Z., Han, J., Han, L. et al. 2013. Fabrication of ZnO/CuS core/shell nanoarrays for inorganic-organic
heterojunction solar cells. Mater. Chem. Phys. 141:804–809.
114. Greene, L. E., Law, M., Yuhas, B. D., and Yang, P. 2007. ZnO–TiO2 core–shell nanorod/P3HT solar
cells. J. Phys. Chem. C 111:18451–18456.
115. Seiyama, T., Kato, A., Fujiishi, K., and Nagatani, M. 1962. A new detector for gaseous components
using semiconductive thin films. Anal. Chem. 34:1502–1503.
116. Müller, J. and Weiβenrieder, S. 1994. ZnO-thin film chemical sensors. Fresenius J. Anal. Chem.
349:380–384.
117. Paraguay, D. F., Miki-Yoshida, M., Morales, J., Solis, J., and Estrada, L. W. 2000. Influence of Al, In,
Cu, Fe and Sn dopants on the response of thin film ZnO gas sensor to ethanol vapour. Thin Solid Films
373:137–140.
118. Sarala Devi, G., Bala Subrahmanyam, V., Gadkari, S. C., and Gupta, S. K. 2006. NH3 gas sensing prop-
erties of nanocrystalline ZnO based thick films. Anal. Chim. Acta 568:41–46.
119. Gaikwad, R. S., Patil, G. R., Pawar, B. N., Mane, R. S., and Han, S.-H. 2013. Liquefied petroleum gas
sensing properties of sprayed nanocrystalline zinc oxide thin films. Sens. Actuators B Phys. 189:339–343.
120. Kalyamwar, V. S., Raghuwanshi, F. C., Jadhao, N. L., and Gadewar, A. J. 2013. Zinc oxide nanostructure
thick films as H2S gas sensors at room temperature. J. Sens. Technol. 3:31–35.
121. Jing, Z. and Zhan, J. 2008. Fabrication and gas-sensing properties of porous ZnO nanoplates. Adv.
Mater. 20:4547–4551.
122. Park, S., An, S., Ko, H., Jin, C., and Lee, C. 2012. Synthesis of nanograined ZnO nanowires and their
enhanced gas sensing properties. ACS Appl. Mater. Interfaces 4:3650–3656.
123. Wei, S., Wang, S., Zhang, Y., and Zhou, M. 2014. Different morphologies of ZnO and their ethanol sens-
ing property. Sens. Actuators B 192:480–487.
124. Spencer, M. J. S. and Yarovsky, I. 2010. ZnO nanostructures for gas sensing: Interaction of NO2, NO, O,
and N with the ZnO (10 10) surface. J. Phys. Chem. C 114:10881–10893.
125. Choi, J.-K., Hwang, I.-S., Kim, S.-J. et al. 2010. Design of selective gas sensors using electrospun
Pd-doped SnO2 hollow nanofibers. Sens. Actuators B Chem. 150:191–199.
126. Hongsith, N., Viriyaworasakul, C., Mangkorntong, P., Mangkorntong, N., and Choopun, S. 2008.
Ethanol sensor based on ZnO and Au-doped ZnO nanowires. Ceram. Int. 34:823–826.
127. Chow, L., Lupan, O., Chai, G. et al. 2013. Synthesis and characterization of Cu-doped ZnO one-
dimensional structures for miniaturized sensor applications with faster response. Sens. Actuators A
189:399–408.
128. Lu, G., Xu, J., Sun, J., Yu, Y., Zhang, Y., and Liu, F. 2012. UV-enhanced room temperature NO2 sensor
using ZnO nanorods modified with SnO2 nanoparticles. Sens. Actuators B Chem. 162:82–88.
Size- and Shape-Controlled ZnO Nanostructures for Multifunctional Devices 93

129. Rakshit, T., Santra, S., Manna, I., and Ray, S. K. 2014. Enhanced sensitivity and selectivity of brush-like
SnO2 nanowire/ZnO nanorod heterostructure based sensors for volatile organic compounds. RSC Adv.
4:36749–36756.
130. Scott, R. W. J., Yang, S. M., Chabanis, G., Coombs, N., Williams, D. E., and Ozin, G. A. 2001. Tin
dioxide opals and inverted opals: Near-ideal microstructures for gas sensors. Adv. Mater. 13:1468–1472.
131. Tiemann, M. 2007. Porous metal oxides as gas sensors. Chem. Eur. J. 13:8376–8388.
132. Chiu, H.-C. and Yeh, C.-S. 2007. Hydrothermal synthesis of SnO2 nanoparticles and their gas-sensing
of alcohol. J. Phys. Chem. C 111:7256–7259.
133. Zhang, Y., Li, J., An, G., and He, X. 2010. Highly porous SnO2 fibers by electrospinning and oxygen
plasma etching and its ethanol-sensing properties. Sens. Actuators B Chem. 144:43–48.
134. Al-Hardan, N. H., Abdullah, M. J., and Abdul Aziz, A. 2013. Performance of Cr-doped ZnO for acetone
sensing. Appl. Surf. Sci. 270:480–485.
135. Park, J. Y., Choi, S.-W., and Kim, S. S. 2011. A model for the enhancement of gas sensing properties in
SnO2–ZnO core–shell nanofibres. J. Phys. D: Appl. Phys. 44:205403-1–205403-4.
136. Weis, T., Lipperheide, R., Wille, U., and Brehme, S. 2002. Barrier-controlled carrier transport in micro-
crystalline semiconducting materials: Description within a unified model. J. Appl. Phys. 92:1411–1418.
137. Sysoev, V. V., Goschnick, J., Schneider, T., Strelcov, E., and Kolmakov, A. 2007. A gradient microarray
electronic nose based on percolating SnO2 nanowire sensing elements. Nano Lett. 7:3182–3188.
138. Kolmakov, A., Klenov, D. O., Lilach, Y., Stemmer, S., and Moskovits, M. 2005. Enhanced gas sens-
ing by individual SnO2 nanowires and nanobelts functionalized with Pd catalyst particles. Nano Lett.
5:667–673.
139. Bowen, T. C., Noble, R. D., and Falconer, J. L. 2004. Fundamentals and applications of pervaporation
through zeolite membranes. J. Membr. Sci. 245:1–33.
140. Borjigin, T., Sun, F., Zhang, J., Cai, K., Ren, H., and Zhu, G. 2012. A microporous metal-organic frame-
work with high stability for GC separation of alcohols from water. Chem. Commun. 48:7613–7615.
141. Szostak, R. 1992. Handbook of Molecular Sieves. New York: Van Nostrand Reinhold.
142. Baertsch, C. D., Funke, H. H., Falconer, J. L., and Noble, R. D. 1996. Permeation of aromatic hydrocar-
bon vapors through silicalite–zeolite membranes. J. Phys. Chem. 100:7676–7679.
143. Ju, D., Xu, H., Qiu, Z., Guo, J., Zhang, J., and Cao, B. 2014. Highly sensitive and selective triethylamine-
sensing properties of nanosheets directly grown on ceramic tube by forming NiO/ZnO PN heterojunc-
tion. Sens. Actuators B Chem. 200:288–296.
144. Kim, J., Kim, W., and Yong, K. 2012. CuO/ZnO heterostructured nanorods: Photochemical synthesis
and the mechanism of H2S gas sensing. J. Phys. Chem. C 116:15682–15691.
145. Huang, L. and Fan, H. 2012. Room-temperature solid state synthesis of ZnO/α-Fe2O3 hierarchical nano-
structures and their enhanced gas-sensing properties. Sens. Actuators B Chem. 171–172:1257–1263.
146. Zhang, Y.-B., Yin, J., Li, L., Zhang, L.-X., and Bie, L.-J. 2014. Enhanced ethanol gas-sensing properties
of flower-like p-CuO/n-ZnO heterojunction nanorods. Sens. Actuators B Chem. 202:500–507.
147. Lou, Z., Deng, J., Wang, L., Wang, L., Fei, T., and Zhang, T. 2013. Toluene and ethanol sensing
performances of pristine and PdO-decorated flower-like ZnO structures. Sens. Actuators B Chem.
176:323–329.
148. Hwang, I.-S., Kim, S.-J., Choi, J.-K. et al. 2010 Synthesis and gas sensing characteristics of highly crys-
talline ZnO–SnO2 core–shell nanowires. Sens. Actuators B Chem. 148:595–600.
149. Woo, H.-S., Na, C. W., Kim, I.-D., and Lee, J.-H. 2012. Highly sensitive and selective trimethylamine sen-
sor using one-dimensional ZnO–Cr2O3 hetero-nanostructures. Nanotechnology 23:245501-1–245501-10.
150. Liu, Y., Li, G., Mi, R., Deng, C., and Gao, P. 2014. An environment-benign method for the synthesis
of p-NiO/n-ZnO heterostructure with excellent performance for gas sensing and photocatalysis. Sens.
Actuators B Chem. 191:537–544.
151. Huang, B.-R. and Lin, J.-C. 2012. Core–shell structure of zinc oxide/indium oxide nanorod based hydro-
gen sensors. Sens. Actuators B Chem. 174:389–393.
152. Datta, N., Ramgir, N., Kaur, M. et al. 2012. Selective H2S sensing characteristics of hydrothermally grown
ZnO-nanowires network tailored by ultrathin CuO layers. Sens. Actuators B Chem. 166–167:394–401.
153. Chang, S.-J., Hsueh, T.-J., Chen, I.-C., and Huang, B.-R. 2008. Highly sensitive ZnO nanowire CO
sensors with the adsorption of Au nanoparticles. Nanotechnology 19:175502(5pp.).
154. Liu, X., Zhang, J., Guo, X., Wu, S., and Wang, S. 2010. Amino acid-assisted one-pot assembly of Au,
Pt nanoparticles onto one-dimensional ZnO microrods. Nanoscale 2:1178–1184.
155. Guo, J., Zhang, J., Zhu, M., Ju, D., Xu, H., and Cao, B. 2014. High-performance gas sensor based on ZnO
nanowires functionalized by Au nanoparticles. Sens. Actuators B 199:339–345.
94 Semiconductor Nanocrystals and Metal Nanoparticles

156. Mubeen, S., Hernandez-Sosa, G., Moses, D., Lee, J., and Moskovits, M. 2011. Plasmonic photosen-
sitization of a wide band gap semiconductor: Converting plasmons to charge carriers. Nano Lett.
11:5548–5552.
157. Zheng, Z. Q., Wang, B., Yao, J. D., and Yang, G. W. 2015. Light-controlled C2H2 gas sensing based
on Au–ZnO nanowires with plasmon-enhanced sensitivity at room temperature. J. Mater. Chem.
C 3:7067–7074.
158. Koshizaki, N. and Oyama, T. 2000. Sensing characteristics of ZnO-based NO sensor. Sens. Actuators B
66:119–122.
159. Shishiyanu, S. T., Shishiyanu, T. S., and Lupan, O. I. 2005. Sensing characteristics of tin doped ZnO thin
films as NO2 gas sensor. Sens. Actuators B 107:379–386.
160. Joy, N. A., Rogers, P. H., Nandasiri, M. I., Thevuthasan, S., and Carpenter, M. A. 2012. Plasmonic-based
sensing using an array of Au-metal oxide thin films. Anal. Chem. 84:10437–10444.
161. Li, Y., Gong, J., He, G., and Deng, Y. 2012. Enhancement of photoresponse and UV-assisted gas sensing
with Au decorated ZnO nanofibers. Mater. Chem. Phys. 134:1172–1178.
162. Hahn, Y.-B., Ahmad, R., and Tripathy, N 2012. Chemical and biological sensors based on metal oxide
nanostructures. Chem. Commun. 48:10369–10385.
163. You, X., Pikul, J. H., King, W. P., and Pak, J. J. 2013. Zinc oxide inverse opal enzymatic biosensor. Appl.
Phys. Lett. 102:253103-1–253103-5.
164. Ahmad, M., Pan, C., Luo, Z., and Zhu, J. 2010. A single ZnO nanofiber-based highly sensitive ampero-
metric glucose biosensor. J. Phys. Chem. C 114:9308–9313.
165. Mandal, S., Sambasivarao, K., Mullick, H., Dhar, A., Maiti, T. K., and Ray, S. K. 2009. Amperometric
detection of glucose biomolecules using ZnO tripods and nanorods: A comparative study. Sens. Lett.
7:635–639.
 3 From Basic Physical Properties
of InAs/InP Quantum Dots
to State-of-the-Art Lasers
for 1.55 µm Optical
Communications
An Overview
Jacky Even, Cheng Wang, and Frédéric Grillot

CONTENTS
3.1 Introduction............................................................................................................................. 95
3.1.1 Toward InP-Based Quantum Dot Lasers for Optical Telecommunications................ 95
3.1.2 Lasers on (311)B InP Substrate....................................................................................97
3.1.3 Lasers on (100)InP Substrate....................................................................................... 98
3.1.4 Current Status of the Dynamic Performance of InP-Based Qdot Laser.....................99
3.2 Basic Physical Properties of Buried InAs/InP Quantum Dots.............................................. 101
3.2.1 Quantum Dot Electronic Structure............................................................................ 101
3.2.2 Carrier Scattering Processes in Quantum Dot.......................................................... 103
3.2.3 Gain, Refractive Index, and Line-Width Enhancement Factor................................. 104
3.3 Free-Running Quantum Dot Lasers...................................................................................... 107
3.3.1 Amplitude Modulation Response.............................................................................. 107
3.3.2 Line-Width Enhancement Factor (α-Factor).............................................................. 111
3.3.3 Impacts of Carrier Capture and Relaxation Processes.............................................. 116
3.4 Conclusion............................................................................................................................. 120
Acknowledgments........................................................................................................................... 120
References....................................................................................................................................... 120

3.1 INTRODUCTION
3.1.1 Toward InP-Based Quantum Dot Lasers for Optical Telecommunications
Semiconductor lasers play a crucial role in optical data communication and telecommunication
applications. Nowadays, high-definition television, video on demand, broadband internet, and
mobile phones are available all around the globe. The exponential rise in cost-effective information
transmission would not have been possible without the introduction of optical transmission systems,
which in turn are enabled by semiconductor lasers. A hundred million new semiconductor lasers
are deployed in communication systems every year, generating several billion dollars of annual
revenue at the component level [1]. Higher performance semiconductor laser sources are in strong

95
96 Semiconductor Nanocrystals and Metal Nanoparticles

demand because of the increasing data traffic in the WAN (wide area network), MAN (metropolitan
area network), and LAN (local area network), which drives the development of novel semiconductor
laser technologies.
Since the first demonstration of a semiconductor laser in 1962, the field has been witnessing the
development from bulk-structure lasers and quantum-well (Qwell) to the advanced nanostructure
quantum wire (Qwire), quantum dash (Qdash), and quantum dot (Qdot) lasers. In the 1960s, the
bulk laser was developed on the basis of semiconductor heterostructures, which provided efficient
confinement of charged carriers in the active region [2]. In particular, the double heterostructure
(DH), which also yields optical confinement, has transformed semiconductor lasers from the labo-
ratory to industry [3]. Quantum confinement occurs when one or more spatial dimensions of a
nanocrystal approach the de Broglie wavelength of the carrier (on the order of 10 nm). The confine-
ment of carriers leads to the quantization of the density of states, and splits the energy band of bulk
semiconductors into energetic subbands [4]. In the 1970s, the first Qwell laser, in which carriers are
confined in one dimension, was demonstrated [5]. Its recognized advantages over DH lasers were
the reduced threshold current by decreasing the thickness of the active layer and the tunability of
the wavelength via changing the Qwell thickness.
Increase of the confinement dimension leads to Qwire (2D confinement), Qdash (quasi-3D con-
finement), or Qdot structures (3D confinement). The 3D spatial confinement of Qdots results in
an atomic-like density of states. The concept of Qdot semiconductor was proposed by Arakawa
and Sakaki in 1982 [6], predicting temperature independence of the threshold current. Thereafter,
reduction in threshold current density, high spectral purity, enhancement of differential gain, and
chirp-free properties were theoretically discussed in the 1980s [7]. The most straightforward
technique to produce an array of Qdots is to fabricate suitably sized mesa-etched quantum wells
grown by metal-organic chemical vapor deposition (MOCVD) or molecular beam epitaxy (MBE).
However, nonradiative defects produced during the etching procedure leads to a degradation of the
material quality, which results in unsuitable structures for lasers. In the 1990s, both selective growth
and self-assembled growth technique, which can avoid nonradiative defects, were well developed.
Particularly, the Stranski–Krastanov growth mode turned out to be very successful for InGaAs/
GaAs systems [8,9]. The strain-induced self-organization of InGaAs/GaAs quantum dots [10,11]
yields threshold current densities as low as ~60 A cm–2 at room temperature [12]. Extensive work on
the GaAs-based Qdot system has been carried out, which resulted in tremendous improvement in
laser performance [13]. An ultralow threshold current density of 17 A cm–2 and a high output power
of 7 W were achieved in InAs/GaAs lasers [14,15]. Nowadays, the InAs/GaAs Qdot products have
already become commercially available in the market [8,16]. Meanwhile, several self-assembled
growth techniques, such as solid-state MBE [17], gas-source MBE [18], MOCVD [19], and chem-
ical-beam epitaxy (CBE) [20], have been improved and successfully used to grow Qdot materials.
Nevertheless, the GaAs-based Qdot laser devices emit usually in the O band (1260–1360 nm)
of the telecommunication windows and are hardly able to reach the desirable long-haul commu-
nication window of ~1.55 µm. Instead, the InAs active region grown on InP substrates allows the
realization of laser devices working in the C band window (1530–1565 nm) because of the smaller
lattice mismatch. In the current market, InP-based 1.55 µm Qwell laser devices have shown sub-
stantial improvement in the optical characteristics in comparison with their DH counterparts.
In order to improve the Qwell laser performance, InAs dots grown on InP substrates have attracted
much attention in the research field. However, the formation of nanostructures on InP is much
more challenging than on the GaAs substrate [21]. Although the InAs/InP and the InAs/GaAs
systems have the same material in the dots, they differ in three aspects: (a) lattice mismatch in
InAs/InP (~3%) is smaller than that in InAs/GaAs (~7%); (b) InAs/InP dots exhibit less confine-
ment potential for electrons, but a stronger confinement for holes; (c) InAs/InP material shares the
same cation (In), while the InAs/GaAs shares the same anion (As) at the interface [22]. The small
lattice mismatch and the complex strain distribution can result in the formation of a new class of
self-assembled Qdash nanostructures instead of Qdots. These are elongated, dot-like structures
From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art Lasers 97

exhibiting interesting mixed characteristics in between those of the Qwell and the Qdot [23].
Strongly anisotropic QDash nanostructures are even closer to Qwire in their electronic properties
[24]. Therefore, realization of real InAs/InP Qdots requires specific epitaxial growth procedures.
This is done by employing conventional or miscut (100) InP substrates as well as (311) InP misori-
ented substrates along with various innovations in the growth process [25]. Nevertheless, reduction
of the dot size and suppression of the size dispersion due to the self-assembly growth procedure are
still challenges in achieving high-quality epitaxial material and hence better device performance.
Currently, the size dispersions, characterized by the photoluminescence (PL) line-width in terms
of the full-width at half-maximum (FWHM), are ~20 meV at 10 K for Qdots and ~50 meV at
room temperature for Qdashes [26–29] and Qdots [30]. Further improvements in material quality
are still required for them to be competitive with the mature InAs/GaAs Qdot systems. InAs dots
formed on the (100)InP substrate usually have a low dot density (on the order of 109 –1010 cm−2)
[31]. These laser structures usually require multiple stacked layers for sufficient material gain. In
contrast, dots grown on high-index (311)B substrates can lead to a large increase of the Qdots
density, commonly in the 5 × 1010 –1011 cm−2 range [32,33]. Figure 3.1 shows the atomic force
microscopy (AFM) images of (a) InAs/InP Qdash, (b) Qdots on a (311)B substrate, and (c) Qdots
on a (100) substrate. The first InAs/InP Qdot laser was reported by Ustinov et al. in 1998 [34,35],
which emitted at ~1.9 µm at 77 K. Room-temperature operation was realized soon [36]. In the fol-
lowing, we review the development of InAs/InP Qdot lasers on both (311)B and (100)InP substrates
employing the InAs/InGaAsP or InAs/InGaAlAs active material regions.

3.1.2 Lasers on (311)B InP Substrate

High-indexed (311)B InP substrates can offer a high density of nucleation points for the Qdot islands,
which strongly reduces surface migration effects and leads to the formation of more symmetric
Qdots in the planar direction. As a result, high density and uniform distribution of Qdots can be
obtained. Many studies have been carried out on the InP (311)B substrate, and a relatively high den-
sity of 5–13 × 1010 cm−2 has been realized [36,38]. The inhomogeneous line-width broadening due to
Qdot size dispersion can be constrained within 50 meV by using the double-capping-layer technique
[30,39]. Figure 3.2a presents the AFM image of a typical InAs/InP dot on a (311)B substrate. The
Qdot base resembles a circle rather than a square with a typical size of 30–50 nm in diameter, and
the height for this kind of dot with cylindrical symmetry is usually several nanometers [33]. The
cross-sectional scanning tunneling microscopy (X-STM) image of the Qdot structure in Figure 3.2b
shows a truncated, faceted profile of the Qdot.

(100) Qdash (311)B Qdot (100) Qdot

200 nm 200 nm
400 nm

(a) (b) (c)

FIGURE 3.1 AFM image of InAs/InP (a) Qdashes on (100) orientation substrate, (b) Qdots on (311)B
­substrate, and (c) Qdots on (100) substrate. (a: From Zhou, D. et al., Appl. Phys. Lett., 93, 161104, 2008;
b: From Zhou, D. et al., Appl. Phys. Lett., 93, 161104, 2008; c: From Bertru, N. et al., Two-dimensional order-
ing of self-assembled InAs quantum dots grown on (311)B InP substrate, Proceedings of SPIE, QD laser on
InP substrate for 1.55 um emission and beyond, San Francisco, CA, Vol. 7608, p. 76081B, 2010.)
98 Semiconductor Nanocrystals and Metal Nanoparticles

AFM X-STM

8 nm

~ 30 nm

35 nm
(a) (b)

FIGURE 3.2 Structural investigations on InAs/InP Qdots on (311)B-oriented substrate. (a) AFM image: the
bright areas represent the top of the Qdot; (b) X-STM image: the bright areas represent the rich InAs areas.
(From Cornet, C. et al., Phys. Rev. B, 74, 035312, 2006.)

Using the InAs/InGaAsP active region, Nishi et al. demonstrated a Qdot laser with seven stack-
ing layers grown by MBE [36]. The laser device had a dot density of 2 × 1010 cm−2 and a threshold
current density of 4.8 kA cm−2, and emitted at 1.4 µm at room temperature. By using a double-
capping technique, Caroff et al. achieved a high dot density of 1.1 × 1011 cm−2, emitting close to
1.55 µm at room temperature [40]. The laser had a high modal gain of 7 cm−1 per dot layer. The
threshold current density and transparency current density were 190 A cm–2 (63 A cm–2 per layer) and
68 A cm–2 (23 A cm–2 per layer), respectively. These are the best threshold performances reported in
this material system. Homeyer et al. reported a broad-area laser emitting at 1.54 µm at room tem-
perature with a record internal quantum efficiency of 62% [41]. The laser consisted of only two stack
layers with a modal gain of 8 cm−1 per layer. The dot density was also as high as ~1.0 × 1011 cm−2.
A single-active-layer laser device with a gain of 13 cm−1 was demonstrated in 2007, but it lased only
up to 295 K [32]. Recently, a low-internal-loss laser of 6 cm−1 was reported, which included nine
stack layers with a total modal gain of 25 cm−1 [42]. However, the InAs/InGaAsP material system
usually has poor temperature stability with a characteristic temperature of only 25–50 K at room
temperature due to the low conduction band offset [25].
By employing the InAs/InGaAlAs active region, Saito et al. first demonstrated a Qdot laser at the
wavelength of 1.63 µm; the threshold current density was 660 A cm–2 (132 A cm–2 per layer) [43].
The extracted internal loss was 3.6 cm−1, which is the lowest reported value for any InAs/InP ­system.
Six years later, by exploiting the Al atoms in the spacer layers and employing the strain compensa-
tion technique, Akahane et al. demonstrated a 30-stack-layer laser with a threshold current density
of 2.7 kA cm–2 (90 A cm–2 per layer) [44]. Subsequently, with improvement in the material growth
quality, the threshold current density of the 30-stack laser was reduced to 1.72 kA cm–2 (57.4 A cm–2
per layer) with extremely high temperature stability. The characteristic temperature (T0) was 114 K
(20°C–75°C), which further improved to a record 148 K (25°C–80°C) in a 20-stack laser [45].
However, the laser had a high internal loss of ~26 cm−1 due to the imperfect coupling of the optical
mode with the multistack gain medium.

3.1.3 Lasers on (100)InP Substrate

The formation of self-assembled Qdot on the (100)InP substrate is more complicated. The forma-
tion of the dot or dash strongly depends on the growth conditions and the thickness of InAs layers.
The major problem in obtaining good performance of the laser is the low dot density on this kind
of substrate.
In the InAs/InGaAsP material system, the growth of dots is mostly based on CBE or MOCVD
techniques. Allen et al. reported a Qdot laser using the CBE technique, which was operated in
pulsed mode at room temperature [46]. The pulsed threshold current density was 3.56 kA cm–2
(713 A cm–2 per layer), where a dot height trimming procedure with growth interruptions was
From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art Lasers 99

employed. The dot density was increased by using the higher energy barrier to 3–6 × 1010 cm−2.
On the other hand, Lelarge et al. used a hybrid growth technique with an MBE-grown active region
in conjunction with MOCVD grown p-doped cladding and contact layers [47]. The continuous wave
(CW) threshold current density was 1.4 kA cm–2 (240 A cm–2 per layer) and T0 was 56 K at room
temperature. Particularly, the laser showed a very high modal gain of 64 cm−1 (10.7 cm−1 per layer).
The laser grown by MOCVD showed a reduced threshold current density of 615 A cm–2 (123 A cm–2
per layer) [48]. The transparency current density was as low as 30 A cm–2 and the internal loss was
only 4.2 cm−1.
In the InAs/InGaAlAs material system, Kim et al. first demonstrated a Qdot laser using an
assisted growth technique of a thin gas underlying layer before the growth of InAs dots in the
InGaAlAs matrix [49,50]. The achieved dot density was 6.0 × 1010 cm−2, and the laser exhibited a
threshold current density of 2.8 kA cm–2 (400 A cm–2 per layer). The measured T0 was 377 K for
temperatures up to 200 K, and 138 K above 200 K. More recently, Gilfert et al. reported a high-gain
Qdot laser using the MBE method [51]. A low internal loss of 4 cm−1 and a high gain of 15 cm−1 per
layer were reported. The lasing threshold current density was 1.95 kA cm–2 (325 A cm–2 per layer).
Very recently, Mollet et al. reported a very high modal gain of 97 cm−1 for an InAs/InP(100)
Qdash laser. However, the internal loss was also as high as 23 cm−1 [52]. Generally, the performance
of InAs/InP lasers has improved significantly since its first demonstration. However, it still needs
improvement to be comparable with InAs/GaAs laser devices.

3.1.4 Current Status of the Dynamic Performance of InP-Based Qdot Laser

In fiber-optic links, the laser transmitter may be either directly modulated, known as directly
modulated (DM) laser, or externally modulated using a modulator. In the direct modulation
scheme, the driving current carries the transmitted data and is directly applied to the laser. In
the external modulation scheme, the laser, which is subjected to a constant bias current, emits a
continuous wave, while an external modulator switches the optical power on or off according to
the data stream. In the external modulation scheme, electrooptic (EO) or electroabsorption (EA)
modulators are commonly used [53]. EO modulators, such as the Mach–Zehnder modulator, uti-
lizes a signal-controlled crystalline material exhibiting the EO effect (Pockels effect) to modulate
the CW laser light. The EO effect is the change in the refractive index of the material resulting
from the application of a DC or low-frequency electric field. The EA modulator controls the
intensity of a laser beam via an electric voltage. Its operation can be based on the Franz–Keldysh
effect [54], that is, a change in the absorption spectrum caused by an applied electric field, which
changes the bandgap energy but usually does not involve the excitation of carriers by the electric
field. However, most EA modulators are made in the form of a waveguide with electrodes for
applying an electric field in a direction perpendicular to the modulated light beam. For achieving
a high extinction ratio, one usually exploits the quantum-confined Stark effect, which describes
the effect of an external electric field upon the light absorption spectrum or emission spectrum of
a quantum-well structure. Both EO and EA modulators are operated at a few volts (below 10 V).
In comparison with EO modulators, EA modulators have the convenient feature that they can
be integrated with the laser on a single chip to create a data transmitter in the form of a photonic
integrated circuit [55,56].
In contrast, DM lasers are the most common, particularly for short-reach systems. They have
the lowest cost and least energy consumption, but they usually suffer the chirp characteristics [68].
DM lasers generally produce more chirps for higher extinction ratios, leading to an optimum setting
for trading off the signal-to-noise ratio and chirp penalty. Aiming to realize chirp-free DM laser
devices, many attempts have been made to develop nanostructured semiconductor lasers. Table 3.1
summarizes the reported dynamic performance of InP-based Qdot and Qdash lasers in the litera-
ture, including the modulation bandwidth and the α-factor, which is linked to the frequency chirp
under direct modulation.
100 Semiconductor Nanocrystals and Metal Nanoparticles

TABLE 3.1
Dynamic Characteristics of Qdot and Qdash Lasers Grown on InP Substrate
Differential
Gain, Gain
References Material Bandwidth α-Factor, ≤Ith α-Factor, >Ith Compression
[57] Qdot, (311)B 4.8 GHz ~1.8 ~6 7.3 × 10−15 cm2
6.4 × 10−16 cm3
[58] Qdot, (100) 5 GHz
15 Gbps
[59] Qdot, (100) 9 GHz
22 Gbps
[60] Qdot, (100) 14.4 GHz ~0 0.8 × 10−15 cm2
p-doped and
tunnel injection
5.4 × 10−17 cm3
[61] Qdash, (100) 7.6 GHz
[62] Qdash, (100) 6 GHz @ undoped ~1 @ p-doped
8 GHz @ p-doped ~0 @ p-doped and
12 GHz @ p-doped tunnel injection
and tunnel injection
[63] Qdash, (100) 8 GHz
p-doping
[64] Qdash, (100) 9.6 GHz 5–7
10 Gbps
[65] Qdash, (100) 10 GHz 1.1 × 10–15 cm2
p-doped
[66] Qdash, (100) 10 GHz
20 Gbps
[67] Qdash, (100) 10 Gbps 2.2
p-doped
[52] Qdash, (100) ~10 GHz ~5 @ undoped 1–2 × 10−15 cm2
~2.7 @ p-doped

Martinez et al. reported a Qdot laser grown on the (311)B InP substrate with a modulation band-
width of 4.8 GHz. The above-threshold α-factor was high (~6) and did not significantly depend on
the bias current [57]. However, most reported dynamics of 1.55 µm lasers are for nanostructures
grown on (100)InP substrates. Gready et al. demonstrated an InAs/InGaAlAs/InP Qdot laser with
a 3 dB bandwidth of 5 GHz [58]. Interestingly, the laser showed a much larger signal modulation
capability: 15 Gbps at 4 dB extinction ratio. The discrepancy between the small- and large-signal
performances was attributed to the large nonlinear gain compression effect [69]. By optimizing the
barrier width and the number of stack layers, the performance of the structure was improved by the
same group. The small-signal modulation bandwidth was increased to 9 GHz, and the large signal
modulation was operated up to 22 Gbps with an extinction ratio of 3 dB [59]. The best performance
of InAs/InP Qdot laser was achieved by Bhowmick et al. recently [60]. The laser material system
was InAs/InGaAlAs, and the active zone was grown on the (100)InP substrate. Tunnel injection and
p-doping techniques enhanced the modulation bandwidth up to 14.4 GHz, and a near-zero α-factor
was realized in this structure. For Qdash lasers on the (100)InP substrate, most works employed the
p-doping technique to improve the modulation bandwidth and to reduce the α-factor [52,62,67,70].
Mi et al. showed that the modulation bandwidth could be increased from 6 GHz for the undoped
From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art Lasers 101

laser to 8 GHz with the p-doped laser [62]. Mollet et al. demonstrated that the α-factor could be
reduced from 5 down to 2.7 by p-doping [52]. On the other hand, the tunnel injection technique
further increased the modulation bandwidth to more than 10 GHz while reducing the α-factor to
near zero [60,62]. Consequently, a low chirp of 0.06 nm at a modulation frequency of 8 GHz was
achieved in the Qdash laser and the same chirp level at 10 GHz in the Qdot laser.

3.2 BASIC PHYSICAL PROPERTIES OF BURIED InAs/InP QUANTUM DOTS

3.2.1 Quantum Dot Electronic Structure
Schematically, the active region of a Qdot laser often consists of a 3D separate confinement hetero-
structure (SCH, also known as barrier), a 2D carrier reservoir (RS, roughly corresponding to the
wetting layer), and dots spatially confined in three dimensions. Figure 3.3 shows an illustration of
the electronic structure in a Qdot laser.
Carriers in the barrier and the wetting layer can be treated as quasi-free particles. Thus, quasi-
continuum electronic states are formed in the SCH and the RS. The total densities of states for the
SCH (per volume) and that for the RS (per area) are, respectively, given by [71]
3/2
æ m* ö
rSCH = 2 ç 2 SCH pk BT ÷ (3.1)
ç 2 ÷
è ø

m*RS
r RS = kBT (3.2)
p 2

with m* the effective mass of either electrons or holes. The quasi-continuum carrier reservoir cou-
pling with the localized energy states of the dots results in smaller energy separations and thus
overlapping states at higher energies [72]. The discrete states lying at lower energies are separated
by a few tens of millielectronvolts in the conduction band (CB), while it is smaller in the valence
band (VB) due to higher effective hole mass.

SCH
RS

ES2
ES1 CB

GS
Photons

VB

FIGURE 3.3 Schematic of a QD laser electronic band structure of electrons and holes.
102 Semiconductor Nanocrystals and Metal Nanoparticles

Accurate simulation of the electronic structure of an InAs/InP quantum dot can be performed using
multiband k · p theory including strain and piezoelectric effects [73]. Based on the eight-band k · p
theory, Figure 3.4a through c shows the confinement potential of the InAs/InP Qdot in comparison
with the InAs/GaAs Qdot system [33]. In the absence of strain (Figure 3.4b), the band edge of InP is
different from that of GaAs relative to the active material InAs. InP confines the holes more strongly,
whereas GaAs does the same for the electrons. In the presence of strain, the band-edge energies are
altered as shown in Figure 3.4a for InAs/InP and in Figure 3.4c for InAs/GaAs, by mainly hydrostatic
strain in the conduction band and biaxial strain in the valence band. Consequently, the heavy/light hole
degeneracy is lifted at the Γ point. The change of band edge for InAs embedded in GaAs is stronger
because of the larger lattice mismatch (6.6% compared to only 3.1% for InP). The most striking fea-
ture is the smaller (strained) bandgap of InAs in InP than in GaAs, which enables the former to reach
the 1.55 µm emission, which is hard to achieve for InAs/GaAs Qdots. Figure 3.4a and c shows that the
depth of the electron confinement potential is similar in both systems, and therefore one can expect
a comparable spectrum for confined electron states provided the Qdots share the same morphologi-
cal properties. However, this does not hold for hole states since the confinement potential for InAs/
InP Qdots is much deeper and the heavy–light hole splitting is smaller. Figure 3.4d and e illustrates
the first three electron and hole wave functions (70% isosurface) of a single particle for both (100)
(Figure 3.4d) and (311)B InP (Figure 3.4e) substrates, which is obtained by solving the Schrödinger
equation. The symbols e0 and h0 stand for 1Se and 1Sh states, and e1, e2, and h1, h2 stand for 1Pe and
1Ph states. The single-particle states provide a basis for the configuration interaction model, which
can be applied to calculate the excitonic properties, including correlation and exchange. Finally, the
excitonic optical absorption spectra can be computed; the details can be found in [33].
Because of the Coulomb interaction, electrons and holes in the semiconductor can be bound into
electron–hole pairs, known as excitons. The distance between the electron and the hole within an
exciton is called the Bohr radius of the exciton. Typical exciton Bohr radius of semiconductors is a
few nanometers [74]. The exciton’s nature can be modified by the confinement structure, and thus
it can exhibit different optical properties. The assumption holds as long as the electron and hole
populations do not show significant deviations. Such a simplified picture was capable of describing
the basic optical properties.

1200

1000 CB

800 e0 e1 e2
600

400
GaAs
InAs
InP

200
h0 h1 h2
(d)
GaAs

GaAs
InAs

0
InAs
InP

InP

–200

–400
LH
–600 e0 e1 e2
HH
–800
SO h0
–1000 h1 h2
(a) (b) (c) (e)

FIGURE 3.4 (a) Vertical scan through the confinement potential for an InAs Qdot embedded in InP, (b) ener-
getic positions of the unstrained band edges of InP and GaAs relative to InAs, (c) vertical scan through the
confinement potential for an InAs Qdot embedded in GaAs. Wave function representations for electrons and
holes for the first InAs/InP Qdot states (d) on (100) substrate and (e) on (311)B substrate, with a Qdot height of
2.93 nm. e0, e1, and e2 stand for electronic states, and h0, h1, and h2 stand for hole states. The (311)B substrate
induces an anisotropy of the wave function. (From Cornet, C. et al., Phys. Rev. B, 74, 035312, 2006.)
From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art Lasers 103

3.2.2 Carrier Scattering Processes in Quantum Dot

In a Qdot device, once the current injection generates charge carriers in the 3D barrier, the carriers
will be transported into the 2D RS, which acts as a carrier reservoir for the localized discrete Qdot
states. The carrier capture process refers to the subsequent carrier capture from the RS to the ESs
of the dots. In the dots, the carriers relax from high energetic ESs down to the GS level. Finally,
radiative recombination of electrons and holes takes place and lasing occurs often on the GS. These
processes are well reflected by the time-resolved photoluminescence (TRPL) of an InAs/InP Qdot
device, as shown in Figure 3.5a. The spectra are fitted with three Gaussian peaks, which, respectively,
correspond to the Qdot excitonic GS at 0.94 eV, the first Qdot ES at 0.99 eV, and the RS ~1.05 eV.
The peak of the spectra shifts from RS at 10 ps via ES at 600 ps toward GS at 1500 ps with time
evolution after the optical excitation [75]. In addition, carrier capture directly from the RS into the
GS is also possible. This direct channel accelerates the carrier indirect process (via ES) to the lasing
GS, and plays an important role on the dual (GS and ES) lasing process, as pointed out in InAs/InP
Qdot lasers [76]. Moreover, those electronic states also exhibit inter-dot electronic coupling [72]. As
in Qwell lasers, the carrier transport process plays an important role in determining the Qdot laser’s
dynamics as well, which induces a parasitic-like roll-off that is indistinguishable from an RC roll-off
in the modulation response, and thus limits the modulation bandwidth [77,78]. TPRL shows that the
carrier transport time across barrier to the RS is several picoseconds (1–5 ps) depending on the thick-
ness of the SCH layer [79,80]. The carrier capture and relaxation transition processes are supported
mainly by two physical mechanisms: Coulomb-interaction-induced c­ arrier–carrier scattering (Auger
process), and carrier–LO phonon scattering. The scattering behavior is different at low and high
excitation carrier densities. At low excitation density, the carrier interaction with LO phonons can
provide efficient scattering channels provided that energy conservation is fulfilled. While the energy
separation of Qdot states typically does not match the LO phonon energy, this scattering mechanism
is often possible for holes due to their dense states [81]. The carrier–LO phonon scattering process is
found to be temperature-dependent: high temperature accelerates this scattering rate [82,83]. When
a high-density carrier plasma is created in the carrier reservoir, carrier–carrier scattering accounts
for the efficient capture from the RS into the localized Qdot states as well as the relaxation between

80
70 W cm–2

10 K 10 ps
60
Intensity (a.u.)

Rise time (ps)

600 ps 40
Data
fit

20
1500 ps T = 10 K

0.9 1.0 1.1 10 100 1000

(a) Energy (eV) (b) Optical density (W cm–2)

FIGURE 3.5 (a) TRPL spectra recorded at 10 K for 10, 600, and 1500 ps after the optical excitation at 790 nm
with an optical excitation density of 70 W cm–2. Spectra are fitted with three Gaussian curves. (b) TRPL analy-
sis of the rise time of the Qdot as a function of the excitation intensity at 10 K. (a: From Miska, P. et al., Appl.
Phys. Lett., 92, 191103, 2008; b: From Miska, P. et al., Appl. Phys. Lett., 94, 061916, 2009.)
104 Semiconductor Nanocrystals and Metal Nanoparticles

the discrete Qdot states [84]. Auger scattering can be categorized into three types according to the
initial electronic states of the carriers involved. One type involves two carriers in the RS states [85];
another type is with one carrier in an ES state while the other in the RS [86]; and the third type has
both carriers occupied by the ES [87]. In contrast to the carrier–phonon interaction process, the
Auger process is carrier-density-dependent, as shown in Figure 3.5b [88]. At 10 K, the carrier relax-
ation time into the GS of the Qdots reduces from ~75 ps under an optical excitation of 1 W cm–2 down
to about 10 ps for an excitation density of 2000 W cm–2.
From the analysis of TRPL rise time of Qdot devices, both the capture and relaxation times are
found to vary over a wide range from 1 up to 100 ps depending on the excitation intensity [82,89].
However, for a moderate RS carrier density of 1011–1012 cm−2, the typical carrier scattering times
are on the order of 1–10 ps [79,90]. For the processes related to Coulomb many-body interactions,
relaxation within the Qdot is typically on a faster timescale than the carrier capture from the RS into
the Qdot. Processes involving holes are typically faster than the corresponding processes involving
electrons, and capture to the excited states is faster than capture to the ground states. Hence, in a
dynamical scenario, first the holes are captured to the excited Qdot states and immediately scattered
via relaxation to the Qdot ground states. Capture of electrons is somewhat slower; the subsequent
relaxation for electrons is only slightly slower than for holes [81].
From the aspect of rate equation modeling of Qdot lasers, carrier distributions in all states are
usually assumed to be under the quasi-equilibrium condition with Fermi–Dirac distribution, which
is suitable for sufficiently rapid intraband relaxation processes [91]. Carrier occupations, at least in
the GS and in the RS, must be modeled in order to distinguish the Qdot laser from the Qwell laser.
However, for achieving moderate accuracy and correlating with experimental data in the InAs/InP
Qdot system, it is necessary to consider the population in the first ES, which can have significant
influence on the laser’s static and dynamic characteristics. Inclusion of more states would be more
accurate but at a price of losing simplicity as well as the intuitive physical image.
Rigorous calculation of carrier scattering rates is a stiff task, which requires sophisticated many-
body quantum theory that treats intraband collision processes. In [91–93], a phenomenological for-
mula was proposed to take into account the carrier-dependent capture and relaxation time in a
semiempirical model:

1
ti = (3.3)
Ai + Ci N RS

where
i denotes the capture or relaxation process
Ai is the phonon-assisted scattering rate
Ci is the coefficient determining the Auger-assisted scattering by carriers in the RS (NRS)

Although this expression leads to good agreement with the TRPL experiments in [75], parameters
Ai and Ci can be quite different from device to device [76], thereby limiting the applicability of this
expression. Because of the fact that once the laser is operated above threshold there is a large density
of carriers in the RS, which does not vary much with the bias current, it is a reasonable approxima-
tion to assume the carrier scattering time as constant, which simplifies the rate equation model for
the study of Qdot laser dynamics.

3.2.3 Gain, Refractive Index, and Line-Width Enhancement Factor

The laser field and the semiconductor gain medium are coupled by the gain and the carrier-induced
refractive index, or equivalently, by the complex optical susceptibility. To determine these quanti-
ties, it is necessary to solve the quantum mechanical gain medium equations of motion for the
microscopic polarization. In principle, these dynamical equations should be derived using the
From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art Lasers 105

full system Hamiltonian, which includes contributions from the kinetic energies, the many-body
Coulomb interactions, the electric–dipole interaction between the carriers and the laser field, as
well as the interactions between the carriers and phonons. The effects of injection current pumping
should also be included [92].
The connection between the classical electrodynamics and quantum mechanics is effected
through the macroscopic polarization P and the microscopic polarization pα:

åm p
1
P= a a (3.4)
V a

In Qdot lasers, the processes include the GS, ES, RS, and barrier transitions. V is the active region
volume, and the polarization summation is performed over all interband optical transitions. The
complex optical susceptibility is connected with the polarization via

1 P
c= (3.5)
e0 nb2 E

where ε0 and E are the vacuum permittivity of light and the amplitude of the electric field, respec-
tively. The gain g and the carrier-induced refractive index δn in the model are defined by

d cg wdn
E (t ) = G P E (t ) + j E (t ) (3.6)
dt 2nb nb

where
ω and c are the laser frequency and the velocity of light, respectively
nb is the refractive index
ΓP is the optical confinement factor

We can obtain the following relation between the gain, refractive index, susceptibility, and the
polarization [94]:

wnb w Im{P}
g=- Im{c} = - (3.7)
c e0 nbc E

nb 1 Re{P}
dn = Re{c} = (3.8)
2 2e0 nb E

Introducing the differential gain a, the phenomenological gain can be expressed as

g = a( N - N tr ) (3.9)

with N and Ntr being the injected carrier density and transparency carrier density of zero gain,
respectively.
In semiconductor lasers, the nonlinear gain phenomenon plays an important role in both static
and dynamic characteristics such as spectral properties, modulation bandwidth, and frequency
chirping [95]. The main physical mechanisms behind nonlinear gain are attributed to the spectral
hole-burning, spatial hole-burning, and carrier heating [96,97]. Furthermore, the gain nonlinearity
was found to enhance the quantum confinement of carriers and carrier relaxation processes [98].
106 Semiconductor Nanocrystals and Metal Nanoparticles

Spectral hole-burning is the formation of a dip in the gain spectrum due to stimulated emission.
The dip occurs by the recombination of electrons and holes at a specific energy and the subse-
quent redistribution of carrier energies due to carrier–carrier scattering. The scattering process
takes place on the timescale of the order of 50–100 fs and leads to a dip width of about 20–40 meV.
It also ensures that temperature equilibrium is established among the carriers within the same
timescale. Carrier heating is related to the fact that the carrier temperatures can be different from
the lattice temperature due to the stimulated emission and free-carrier absorption [96]. The carrier
temperatures relax toward the lattice temperature by electron–phonon scattering processes within
a timescale of 0.5–1 ps. The nonlinear gain effect is usually characterized by a phenomenological
gain compression factor, as

g
gnl = » g(1 - xS ) (3.10)
1 + xS

with S being the photon density. This expression shows that the linear gain g is reduced at high
power density.
The refractive index change related to the optical interband transition as expressed in Equation 3.8
is known as anomalous dispersion [99]. Another important contribution to the index change is the
free-carrier plasma originating from intraband transitions [100]. In Qwell lasers, this contribution
to the differential index is well described by the Drude model [101]:

G pe2 N
dn fc = - (3.11)
2nbe0 m*w2

with e the electronic charge. Analogous transitions in Qdot lasers can be envisaged between bound
Qdot states and the continuum levels of the RS and the barrier. It has been shown that the Drude for-
mula can also be applied to the case of Qdot lasers when the Qdot carriers are not tightly confined
and when working at photon energies in the 0.8–1.0 eV region [100].
In semiconductor lasers, it is well known that any change in the imaginary part of the suscep-
tibility (gain) will be accompanied by a corresponding change in its real part (refractive index)
via the Kramers–Kronig relations. The line-width enhancement factor (or α-factor) describes the
coupling between the carrier-induced variation of real and imaginary parts of susceptibility, and
is defined as [102]

¶ Re{c}/¶N
aH = (3.12)
¶ Im{c}/¶N

Employing Equations 3.7 and 3.8, the above definition is equivalent to the following often-used
expression:

w ¶n /¶N
a H = -2 (3.13)
c ¶g /¶N

In a practical case, the variation of the carrier concentration is usually small, which justifies tak-
ing the derivatives at the operating point and assuming a linear dependence of g(N) and n(N).
Equation 3.13 can therefore be written as

w Dn
a H = -2 (3.14)
c Dg
From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art Lasers 107

Through the relation between the refractive index and frequency variation, we can obtain the fol-
lowing equivalent formula:

nb Dw
aH = 2 (3.15)
c Dg

The α-factor plays a crucial role in driving fundamental features of semiconductor lasers such as
the spectral line-width broadening [103], frequency chirp [104], mode stability [105], and nonlinear
dynamics subject to optical injection [106] or optical feedback [107–109]. Typical Qwell lasers often
exhibit α-factor values in the order of 2–5 [78]. For Qdot lasers, earlier analyses have suggested a
zero or near-zero α-factor due to the delta-function-like discrete density of states. A symmetrical
gain curve indeed leads to a dispersive curve of the refractive index with a zero value at the gain
peak. However, experimental α-factor values in Qdot lasers vary over a wide range from zero up to
more than 10, particularly a giant value (as high as 60) was also reported [110–112]. The nonzero
α-factor is attributed to the large inhomogeneous broadening, the off-resonant bound and contin-
uum states, as well as the free-carrier plasma effect.

3.3 FREE-RUNNING QUANTUM DOT LASERS

3.3.1 Amplitude Modulation Response
High-speed, energy-effective, and low-cost optical communication networks primarily require
semiconductor laser sources of broad modulation bandwidth. In order to theoretically discuss the
amplitude modulation (AM) performance of Qdot lasers, we employ a semiclassical rate equation
model [76]. The model will be analyzed in a semianalytical approach, which has the merit of giving
an intuitive physical image. This numerical model of the Qdot laser holds under the assumption that
the active region consists of only one Qdot ensemble; that is, the inhomogeneous broadening due to
the dot size fluctuation is not considered. The electrons and holes are treated as electron–hole (e–h)
pairs, meaning that the system is in excitonic energy states. Two discrete states in Qdots are taken
into account: a twofold degenerate GS and a fourfold degenerate first ES. The Qdots are intercon-
nected by the 2D RS. This simplified picture corresponds to the TRPL experimental observations
in Figure 3.5 [75]. Carriers are supposed to be injected directly from the contacts into the RS, so the
carrier dynamics in the 3D barrier are not taken into account in the model.
Figure 3.6 shows the schematic of the carrier dynamics in the exiton framework. First, the exter-
nally injected carrier fills directly the RS reservoir; some of the carriers are then either captured

Current
injection

RS Excitons

τGS
RS
τES
RS
τRS
ES

ES

τGS
ES τGS τES
GS
RS

GS

FIGURE 3.6 Sketch of carrier dynamics model including a direct relaxation channel.
108 Semiconductor Nanocrystals and Metal Nanoparticles

into the ES within time tES RS

or directly into the GS within time tGS RS
, and some of them recombine
spontaneously with a spontaneous emission time t RS . Once in the ES, carriers can relax into the
spon

GS within time tGS ES

or recombine spontaneously. On the other hand, carriers can also be thermally
re-emitted from the ES to the RS with an escape time tES RS , which is governed by the Fermi distribu-
tion for the quasi-thermal equilibrium without external excitation [113]. A similar dynamic behavior
is followed for the carrier population on the GS level with regard to the ES. Direct injection from
the RS to the GS was introduced to reproduce the experimental results [76]. Stimulated emission
occurs from the GS when the threshold is reached, and that from the ES is not taken into account
in the model. Following Figure 3.6, the four coupled rate equations on carrier and photon densities
are described as follows:

dN RS I N N RS N RS N RS N GS
= + ES - RS (1 - rES ) - RS (1 - rGS ) - spon + GS (3.16)
dt qV t ES
RS t ES t GS t RS t RS

dN ES N RS N GS N ES N ES N ES
= RS (1 - rES ) + GS (1 - rES ) - ES - ES (1-
- rGS ) - spon (3.17)
dt tES tES t RS tGS tES

dN GS N RS N ES N GS N GS N GS
= RS (1 - rGS ) + ES (1 - rGS ) - GS (1 - rES ) - spon - GS - gGS vg SGS (3.18)
dt tGS tGS tES tGS tWL

dSGS S N GS
= G P gGS vg SGS - GS + bSP spon (3.19)
dt tp tGS

where
NRS, NES, and NGS are the carrier densities in the RS, ES, GS, respectively
SGS is the photon density in the cavity with GS resonance energy
βSP is the spontaneous emission factor
ΓP is the optical confinement factor
τp is the photon lifetime
vg is the group velocity
V is the volume of the laser’s active region

The GS gain is given by

N B æ 2 N GS ö
gGS = aGS - 1÷ (3.20)
H B çè 2 N B /H B ø

where
aGS is the differential gain
NB is the total Qdot surface density
HB is the height of the dots.

In what follows, it is important to stress that the effects of gain compression are not taken into
account. In Equations 3.16 through 3.18, ρGS,ES are the carrier occupation probabilities in the GS and
the ES, respectively:

N GS N ES
rGS = ; rES = (3.21)
2 N B /H B 4 N B /H B
From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art Lasers 109

Since the carrier escape from the GS to the RS has little effect on lasing properties [76], the N GS /tGS
RS
term in Equations 3.16 and 3.18 can be neglected.
The rate equations can be linearized through a small-signal analysis. Assuming a sinusoidal
current modulation dI = I1ejωt with modulation frequency ω, the corresponding carrier and photon
variations are of the form

dN RS, ES, GS = N RS1, ES1, GS1e jwt

(3.22)
dSGS = SGS1e jwt

Substituting the above formulas into the rate equations (Equations 3.16 through 3.18), we obtain the
linearized differential rate equation in matrix form:

é1 ù
é g11 + jw - g12 0 0 ù é N RS1 ù ê ú
ê úê ú
ê - g 21 g 22 + jw - g 23 0 ú ê N ES1 ú I1 ê0 ú
= ê ú (3.23)
ê - g 31 - g 32 g 33 + jw - g 34 ú ê N GS1 ú qV ê0 ú
ê úê ú ê ú
ë 0 0 - g 43 g 44 + jwû ë SGS1 û
ë0 û

with the following elements:

1 - rES 1 - rGS 1 1 N RS 1
g11 = + RS + spon ; g12 = ES + RS
tES
RS
tGS t RS t RS t ES 4 N B /H B
1 - rES 1- r 1 N RS 1 N 1 1
g 21 = ; g 22 = ESGS + ES + RS + GS + spon
tES
RS
tGS t RS tES 4 N B /H B tGS
ES 4 N B / H B t ES

1 - rES N ES 1 1- r 1- r N GS 1
g 23 = + ES ; g 31 = RSGS ; g 32 = ESGS + GS (3.24)
t ES
GS
tGS 2 N B /H B tGS tGS tES 4 N B /H B
1 - rES N ES 1 1
g 33 = + ES + vg aGS SGS + spon ; g 34 = -vg gGS
tGS
ES t GS 2 N B / H B t GS

G pbSP 1
g 43 = + G p vg aGS SGS ; g 44 = - G p vg gGS
tGS
spon
tp

Finally, the AM or intensity modulation response of the Qdot laser is calculated by

SGS1
HQdot (w) = (3.25)
I1 / (qV )

This equation is also known as the modulation transfer function. Through proper approximation,
the AM response is given by [114]

HQdot (w) » H1 (w) H 0 (w)

æ w2R öæ w2R 0 ö
=ç 2 ÷ç 2 ÷ (3.26)
è w R - w + jwG ø è w R 0 - w + jwG 0 ø
2 2
110 Semiconductor Nanocrystals and Metal Nanoparticles

where the first part, H1(ω), is dominated by the carrier–photon interaction processes. The resonance
frequency ωR and the damping factor Γ of the Qdot laser are, respectively, expressed as

vg aGS SGS G pbSP N GS éæ H B N ES 1 - rES ö 1 - bSP ù bSP

w2R = + spon êç + GS ÷ + tspon ú + tspon t (3.27)
tp tGS SGS ëè 2 N B tGS
ES
tES ø GS û GS p

æ H N ES 1 - rES ö 1 G bSP N GS
G = vg aGS SGS + ç B ES + GS ÷ + spon + pspon (3.28)
è 2 N B tGS tES ø tGS tGS SGS

where the steady-state relationship 1/t p - G p vg gGS = G pbSP N GS / (tGS

spon
SGS ) has been used. In compari-
son with Qwell lasers [78], both expressions have an additional term ( H B N ES / (2 N B tGS
ES
) + (1 - rES ) /tGS
ES ),
which describes the effective carrier scattering rate into and out of the GS. The second part, H0(ω),
is dominated by the carrier capture and relaxation processes, and the introduced two parameters
ωR0 and Γ0 are given by

æ 1- r 1 öæ 1- r 1 H N RS H N GS ö
w2R 0 = ç RSES + spon ÷ ç ESGS + spon + B RS + B GS ÷
è tES t RS ø è tGS tES 4 N B tES 4 N B tES ø

1 1 H N RS 1 - rES
+ - B RS (3.29)
t RS t RS
ES spon
4 N B tES tES
RS

æ 1- r 1- r 1 öæ 1- r 1 N RS 1 N GS 1 1 ö
G 0 = ç RSES + RSGS + spon ÷ ç ESGS + ES + RS + GS + spon ÷ (3.30)
è t ES tGS t RS ø è tGS t RS tES 4 N B /H B tES 4 N B /H B tES ø

Figure 3.7 presents an example of the calculated AM response of a Qdot laser. The solid curve
HQdot(ω) shows that the response is strongly damped as usually observed in experiments [115–117],
which can be attributed to the carrier relaxation and escape process of the GS as described in

0
AM response (dB)

–6

–12

HQdot(ω)
–18 H1(ω)
H0(ω)

–24
0 1 2 3 4 5 6 7 8 9 10
Frequency (GHz)

FIGURE 3.7 AM response HQdot(ω) (solid curve) of the Qdot laser. H1(ω) (dash curve) describes mainly
the contribution of carrier–photon interactions, and H0(ω) (dash-dot curve) gives the contribution of carrier
capture and relaxation processes. (From Wang, C. et al., IEEE J. Quantum Electron., 48, 1144, 2012.)
From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art Lasers 111

Equation 3.28. On the other hand, H0(ω) (dash-dot) exhibits a much smaller bandwidth than H1(ω)
(dash), which demonstrates that the finite carrier capture and relaxation times limit the modulation
bandwidth of Qdot lasers.

3.3.2 Line-Width Enhancement Factor (α-Factor)

The line-width enhancement factor (α-factor) plays an important role in determining fundamental
features of semiconductor lasers. Although the Qdot laser is predicted to have a delta-function-like
discrete density of states, the measured α-factor values in experiments vary over a wide range from
zero up to >10. The nonzero α-factor in Qdot lasers is partly attributed to the asymmetric gain
spectrum because of the inhomogeneous broadening [118] and the carrier population in off-resonant
states [119]. On the other hand, the free carrier plasma effect in the barrier and in the RS is reported
to contribute almost half of the total refractive index change [100,101]. In this section, we describe
an improved electric field model for Qdot lasers taking into account the contribution of off-resonant
states (ES and RS) on the refractive index change, which allows a semianalytical study of the
α-factor features in the Qdots. The model is capable of exploring the crucial physical mechanisms
driving the Qdot laser’s α-factor.
The conventional model describing the complex electric field of semiconductor lasers is given by

dE (t ) 1 æ 1 ö
= ç G P vg g - ÷ E (t ) + jDwN E (t ) (3.31)
dt 2è tP ø

where the first term on the right-hand side gives the gain and the photon loss of the laser cavity. The
second term, ΔωN, describes the carrier-induced frequency shift of the laser field with respect to the
frequency at the lasing threshold. ΔωN is usually expressed by the α-factor αH as

1æ 1 ö
DwN = G P vg g - ÷ aH (3.32)
2 çè tP ø

This model is able to study the impacts of a nonzero α-factor on the modulation dynamics and
nonlinear dynamics of Qdot lasers [120] but does not allow the study of αH itself under different
operating conditions. In order to investigate the α-factor, we need to obtain the expression of the
gain and the refractive index separately. In the semiclassical theory, the semiconductor laser system
can be fully described by the optical Bloch equations together with Maxell’s equations [121]. The
RS is treated as a discrete energy state of degeneracy DRS = kBTm* A RS /(πℏ2), with m* the reduced
carrier mass and A RS the RS surface area (see Section 3.1) [122]. The active region consists of only
one Qdot ensemble. In addition, the electrons and holes are treated as neutral pairs (excitons). Two
discrete states—the ground state (GS) and the first excited state (ES)—are considered in the dots.
With these assumptions, the slowly varying electric field E(t) is given by [123]

å (m* P ) - 2t
d jw0 G 1 1
E (t ) = LS P X X E (t ) (3.33)
dt 2ebg e0 ARS H B X = GS, ES, RS P

where
w0LS is the lasing frequency in the cold cavity
εbg and ε0 are the background and vacuum permittivity, respectively
ΓP is the optical confinement factor
HB is the height (equal to the dot’s height)
112 Semiconductor Nanocrystals and Metal Nanoparticles

The sum over X (X = GS, ES, RS) includes all possible optical transitions, with μX being the cor-
responding dipole transition matrix element and PX the microscopic polarization. τp is the photon
lifetime in the laser cavity. Assuming a sufficiently short dephasing time TD and adiabatically elimi-
nating the interband polarization yields the quasi-static relation

m X TD 1 + j (wX - w0LS )TD

PX (t ) = - j (2r X - 1) E (t ) (3.34)
2 1 + (wX - w0LS )2 TD2

where
ρX denotes the carrier occupation probability
ℏωX gives the transition energy of each state

Inserting Equation 3.34 into Equation 3.33, we obtain the complex gain

 (w0LS , t ) = æç 2m*GS jwLS G P 2 N B PGS (t ) ö÷ + æç 4m*ES jwLS G P 2 N B PES (t ) ö÷

0 0
G
è 2e0ebg H B E (t ) ø è 2e0ebg H B E (t ) ø

æ jw0 G 2 P (t ) ö
+ ç LS P DRS m*RS RS ÷ (3.35)
è 2e0ebg ARS H B E (t ) ø

The optical susceptibility can be derived from this complex gain expression through the relation-
 (w0LS, t ) / ( jwG P ). The real part of Equation 3.35 is related to the laser gain,
ship c(w0LS, t ) = 2ebgG
while the imaginary part gives the instantaneous frequency shift of the electric field. The three
terms on the right-hand side give contributions of the GS, ES, and RS, respectively. Introducing the
differential gain (aX), we have

2mGS m*GS wGSTD

aGS =
vg e0ebg

4m ES m*ES wESTD
aES = (3.36)
vg e0ebg

m RS m*RS wRSTD
aRS =
vg e0ebg

with vg being the group velocity of the light. The material gain of each state is then given by

aGS N B æ 2 N GS ö
gGS = - 1÷
1 + xSGS H B çè 2 N B /H B ø

N B æ 2 N ES ö
gES = aES - 1÷ (3.37)
H B çè 4 N B /H B ø

DRS æ 2 N RS ö
gRS = aRS ç - 1÷
ARS H B è DRS /( ARS H B ) ø

where
N X is the carrier density in each state
SGS is the photon density in the GS
ξ denotes the gain compression factor
From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art Lasers 113

Because the real part of the complex gain approaches zero very quickly when off resonance, the
field gain originates mainly from the resonant state. Considering the lasing emission in resonance
with the GS transition w0LS = wGS , we obtain

 (wGS )] » G P vg gGS
Re[G (3.38)

In contrast, the imaginary part of the complex gain decays slowly for off-resonant frequencies.
Thus, the off-resonant states can significantly influence the refractive index change, even though
their gain contribution to the GS lasing is almost nil. Carrier populations in the off-resonant ES and
RS induced frequency shifts of the laser field respectively are

1
DwES
N = G P vg gES FES
GS
(3.39)
2

1
DwRS
N = G P vg gRS FRS
GS
(3.40)
2

with coefficients

wGS (wES , RS - wGS )TD

, RS =
GS
FES (3.41)
wES , RS 1 + (wES , RS - wGS )2 TD2

From Equation 3.35, it is seen that the resonant GS has no contribution to the refractive index
change, which is the case when the laser is operated at the gain peak with a symmetric gain distri-
bution. Nevertheless, as mentioned in the introduction, due to the asymmetric Qdot size dispersion,
the resonant state induces a finite α-factor aGS
H , and the corresponding frequency shift with respect
to the cold cavity can be expressed by

1
DwGS
N = G P vg gGS aGS
H (3.42)
2

Employing Equations 3.37 through 3.42, the electric field (Equation 3.33) is re-expressed as

dE (t ) 1 æ ö
dt
= ç G P vg gGS -
2è
1
tP ø
( ES RS
)
÷ E (t ) + j DwN + DwN + DwN E (t )
GS
(3.43)

With carrier injection, the lasing frequency becomes wLS = wthLS + DwLS N , where DwN = DwN +
LS GS

DwES
N + DwN gives the total frequency shift of the electric field from its threshold value (wth ).
RS LS

jf( t )
Through the E (t ) = S (t )V / G P e relationship, the photon density S(t) and the phase ϕ(t) can be
separately described. Combining with the equations describing the carrier dynamics in Qdot lasers,
the laser system is finally given by

dN RS I N ES N RS N RS
= + ES - RS (1 - rES ) - spon (3.44)
dt qV t RS tES t RS

dN ES æ N RS N GS ö N ES N ES N ES
= ç RS + GS ÷ (1 - rES ) - ES (1 - rGS ) - ES - (3.45)
dt è t ES tES ø tGS t RS tspon
ES
114 Semiconductor Nanocrystals and Metal Nanoparticles

dN GS N ES N GS N GS
= ES (1 - rGS ) - GS (1 - rES ) - vg gGS SGS - spon (3.46)
dt tGS tES tGS

dSGS æ 1 ö N GS
= ç G p vg gGS - ÷ SGS + bSP tspon (3.47)
dt è t P ø GS

df
= DwGS
N + DwN + DwN
ES RS
(3.48)
dt

where
tGS
spon
is the spontaneous emission time
βSP is the spontaneous emission factor

Carriers in the RS are scattered into the dots through phonon-assisted and Auger-assisted processes
[82,83]. The latter makes the scattering rates nonlinearly depend on the carrier density in the RS.
However, for the sake of simplicity, the carrier capture time tES
RS
and the relaxation time tGS ES
are both
ES GS
treated as constants in this work. On the other hand, the carrier escape times (t RS ,tES ) are governed
by the Fermi distribution for a quasi-thermal equilibrium system [124]. For semiconductor lasers
operating under small-signal modulation with frequency ω, the bias current change δI induces vari-
ations of the carriers δN X, the photon δSGS, and the phase δϕ. In order to perform the analyses, the
differential rate equations are derived as follows:

é g11 + jw - g12 0 0 0 ù é dN RS ù é1 ù
ê úê ú ê ú
ê - g 21 g 22 + jw - g 23 0 0 ú ê dN ES ú 0
dI ê ú
ê 0 - g 32 g 33 + jw - g 34 ú ê
0 dN GS ú = ê 0ú (3.49)
ê úê ú qV ê ú
ê 0 0 - g 43 g 44 + jw 0 ú ê dSGS ú ê0 ú
êë - g 51 - g 52 - g 53 - g 54 ú ê
jwû ë df û ú êë0 úû

where

1 - rES 1 1 1 N RS 1- r
g11 = + spon ; g12 = ES + ; g 21 = RSES ;
tES
RS
t RS t RS 4 N B tES
RS
t ES
1 - rGS 1 1 1 N RS 1 N GS
g 22 = + ES + spon + +
tGS
ES
t RS tES 4 N B tES
RS
4 N B tGS
ES

1 - rES 1 N ES 1- r 1 N GS 1- r 1 1 N ES
g 23 = + ; g 32 = ESGS + ; g 33 = GSES + spon + + vg aSGS
tGS
ES 2 N B t ES
GS t GS 4 N B t GS
ES t ES t GS 2 N B tGS
ES

G pbSP
g 34 = -vg gGS + vg aP SGS ; g 43 = G p vg aSGS + ;
tGS
spon

1
g 44 = -G p vg gGS + + G p vg aP SGS ; g 51 = G p vg aRS FRS
tp

1 1 1
g 52 = G p vg aES FES ; g 53 = G p vg aaGS
H ; g 54 = - G p vg aP aGS
H (3.50)
4 2 2
From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art Lasers 115

with

¶gGS aGS ¶g x
a= = ; aP = - GS = gGS (3.51)
¶N GS 1 + xSGS ¶SGS 1 + xSGS

Based on the above differential rate equations, the α-factor of the Qdot laser is derived as

2 d éë DwN ( N ) ùû
LS
1 a dN a dN
aGS
H ,QD (w) = º aGS
H + FES ES ES + 2 FRS RS RS (3.52)
G P vg dgGS ( N ) 2 adN GS adN GS

Following the definition in Equation 3.13, it is noted that only the carrier contribution (δN) is included
in the above equation, while the photon contribution (δS) is excluded. In the following, it will be
shown that the α-factor of Qdot lasers presents peculiar characteristics under direct modulation.
Over the last decades, various techniques have been proposed for the measurement of the
α-factor. In this work, we employ the well-known “FM/AM” technique for the above-threshold
analysis and the widely used “Hakki-Paoli” method for the below-threshold analysis [102]. The FM/
AM technique relies on the direct current modulation of the laser, which generates both the optical
frequency (FM) and amplitude (AM) modulations [125]. With respect to the linearized rate equa-
tions, the ratio of the FM/AM index is derived as

b(w) dwLS /w
2 =2
m(w) dSGS /SGS

jw + (1/tP - G P vg gGS + G p vg aP SGS ) é GS æ aP dSGS ö 1 aES dN ES a dN ù

º ´ êa H ç 1 - ÷ + 2 FES adN + 2 FRS RS RS ú
jw ë è adN GS ø GS adN GS û
(3.53)

where the relation δωLS = jωδϕ is used in the above derivation. In this approach, the laser’s α-factor
is usually extracted through the formula a FM
H ,QD = min{2b(w) /m(w)}. This is indeed true for Qwell
/AM

or bulk lasers; however, we show that the α-factor of Qdot lasers is dependent on the modulation
frequency but we still take the minimum value to characterize the Qdot laser as reported in [119].
For semiconductor lasers operating below threshold, the Hakki-Paoli method relies on the direct
measurement of the optical spectra of amplified spontaneous emission (ASE) in the laser cavity.
Tuning the pump current slightly step by step (ΔI), the gain change can be extracted by the Hakki-
Paoli method, and the wavelength variation can be directly recorded using an optical spectrum
analyzer. Correspondingly, the below-threshold α-factor is calculated as

1 GS aES DN ES GS a RS DN RS
a HASE
,QD = a H +
GS
FES + 2 FRS (3.54)
2 aGS DN GS aGS DN GS

The laser parameters used in the simulation are listed in Table 3.2 [57,75,126]. It is noted that the
carrier occupation in the GS has a small contribution to the α-factor (<1) [127]; hence we assume the
value aGS
H = 0.5 in the simulation. Figure 3.8a depicts the carrier density variations in the three states
under small-signal modulation. For low frequencies (smaller than 0.1 GHz), all the carrier density
variations remain almost constant, but the variations of the ES (δNES) and RS (δNRS) populations
are 15 dB larger than that of the GS (δNGS) one. The small variation of the GS carrier population is
associated with the gain-clamping above the threshold. Both δNGS and δNES exhibit resonances at
~7 GHz. Beyond the resonance frequency, δNGS decays faster than δNES and δNRS. These features
significantly impact the behavior of the α-factor, as described in Equation 3.52.
116 Semiconductor Nanocrystals and Metal Nanoparticles

TABLE 3.2
Qdot Material and Laser Parameters for the Study of the α-Factor
Symbol Description Value Symbol Description Value
L Active region length 5 × 10 cm
−2 t RS
ES Capture time from RS to ES 6.3 ps
W Active region width 4 × 10−4 cm tGS
ES Relaxation time from ES to GS 2.9 ps
R1 = R2 Mirror reflectivity 0.32 aGS GS differential gain 5 × 10−15 cm2
nr Refractive index 3.5 aES ES differential gain 10 × 10−15 cm2
αi Internal modal loss 6 cm−1 aRS RS differential gain 2.5 × 10−15 cm2
NB Dot density 10 × 1010 cm−2 ξ Gain compression factor 2 × 10−16 cm3
HB Dot height 5 × 10−7 cm aGS
H GS induced α-factor 0.5
ERS RS transition energy 0.97 eV TD Dephasing time 0.1 ps
EES ES transition energy 0.87 eV ΓP Optical confinement factor 0.06
EGS GS transition energy 0.82 eV βSP Spontaneous emission factor 1 × 10−4

Figure 3.8b compares the difference between the α-factor aGS H ,QD (w) and the ratio ∣2β(ω)/m(ω)∣ as

a function of the modulation frequency. At low frequencies (<0.1 GHz), there is a large discrepancy
between the two parameters. As expected, ∣2β(ω)/m(ω)∣ exhibits large values due to the gain com-
pression and the large carrier variations in the ES and in the RS. Nevertheless, aGS H ,QD (w) remains
constant. On increasing the modulation frequency beyond several gigahertz, the two values of both
parameters decrease down to a plateau, which gives the conventional α-factor, indicated by the
horizontal line. As can be seen, a FMH ,QD is almost the same as a H ,QD , which indicates that the FM/
/ AM GS

AM method is a reliable technique for the measurement of Qdot laser’s α-factor. Further increase
of the modulation frequency raises again both the values, as observed experimentally in a Qdot
laser (inset of Figure 3.8b) [129]. It is emphasized that such a situation is not encountered in Qwell
lasers [125]. This behavior is attributed to the different decay rates (versus modulation frequency)
of carrier variations in each state, as shown in Figure 3.8a. In addition, Figure 3.8b shows that the
ES (dash-dot curves) contributes more to the α-factor aGS H ,QD than the RS due to its smaller energy
separation with the resonant GS.
Based on the ASE and the FM/AM methods, Figure 3.9 illustrates the α-factor as a function of
the normalized pump current I/Ith. Below threshold, carrier populations in both the resonant and off-
resonant states increase with the pump current. In consequence, the α-factor increases nonlinearly.
Above threshold, the carrier population in the GS is clamped, while the off-resonant state popula-
tions keep increasing. Thus, the α-factor varies almost linearly above threshold as usually measured
in experiments [57,71]. At threshold, the α-factor extracted from the ASE method is similar to that
using the FM/AM technique. In addition, the α-factor is larger than the sole GS-induced value of
aGS
H = 0.5 both below and above threshold, which means the off-resonant ES and RS contribute to
the increase of the α-factor in the Qdot laser. This is explained by the fact that the coefficients FES GS
GS
and FRS are both positive since the ES and RS have higher energies than the GS (see Equation 3.41).

3.3.3 Impacts of Carrier Capture and Relaxation Processes

As discussed previously, the Qdot laser involves a carrier capture process from the 2D RS to the
localized ES and a carrier relaxation process from the ES to the GS inside the dots. This section
discusses the influences of these processes on the laser’s modulation dynamics.
In order to study the impacts of the carrier capture process, the carrier relaxation time tGS
ES
is
fixed at 2.9 ps, while the carrier capture time tES is varied from 0.1 up to 50 ps. Figure 3.10a
RS

shows the variation of the AM response for different capture times. Slow capture reduces the 3 dB
From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art Lasers 117

25
δNES
20

Carrier variation (dB) 15

δNRS

10

δNGS
5

–5

–10
0.01 0.1 1 10 100

(a) Modulation frequency (GHz)

100
100 100
|2β(ω)/m(ω)|

10 10 10

|αH,QD(ω)|
GS
0.1 1 10

1 1

α-factor
ES
0.1 0.1
RS

0.01 0.1 1 10 100

(b) Modulation frequency (GHz)

FIGURE 3.8 (a) Small-signal carrier density variations in the GS (solid line), ES (dashed line), and RS
(dash-dot line) versus the modulation frequency. The bias current is I = 1.2 × Ith, with the threshold current
Ith = 49 mA. The carrier variation is normalized to the value δNGS of 0.01 GHz. (b) Modulation-frequency
dependence of the FM/AM ratio (dash) and of the α-factor (thick solid). The minimum level indicated by the
horizontal line gives the laser’s conventional α-factor. The thin dash-dotted curve represents the sole contribu-
tion of the ES or the RS to the α-factor, respectively. The inset shows an experimental curve of the FM/AM
ratio for a Qdot laser. (From Wang, C. et al., Appl. Phys. Lett., 105, 221114, 2014.)

modulation bandwidth from 11 GHz down to 7 GHz. Besides, the resonance peak is also slightly
reduced. Interestingly, for capture times larger than 30 ps, a parasitic-like roll-off (dip) appears in
the response, which is similar to the effect of the slow carrier transport process from the 3D barrier
to the 2D RS [78]. In the same approach, by fixing the capture time at 6.3 ps, Figure 3.10b shows the
impact of carrier relaxation time on the AM response. The modulation bandwidth is significantly
reduced by the slow relaxation process from 10.8 GHz for tGS ES
= 0.1 ps to 1.6 GHz for tGS
ES
= 50 ps.
In addition, the response is strongly damped for large relaxation times. It is noted that the evolu-
tion of the AM response shape is quite different from that for the capture process in Figure 3.10a.
118 Semiconductor Nanocrystals and Metal Nanoparticles

1.0 ASE
ASE method, αH,QD
FM/AM
FM/AM method, αH,QD
0.9

α-factor 0.8

0.7

0.6

0.5
0.0 0.5 1.0 1.5 2.0
I/Ith (mA)

FIGURE 3.9 α-Factor as a function of the normalized bias current I/Ith. (From Wang, C. et al., Appl. Phys.
Lett., 105, 221114, 2014.)

6 12

0 6
AM response (dB)
AM response (dB)

dip
0
–6
–6
–12 Increase
–12 Increase
capture time
capture time
–18 –18
0 5 10 15 20 0 5 10 15 20
(a) Modulation frequency (GHz) (b) Modulation frequency (GHz)

7.5 10 80
33
Resonance frequency (GHz)
Resonance frequency (GHz)

7.0 70
30
Damping factor (GHz)
Damping factor (GHz)

6.5 9 60
27
6.0
24 50
5.5 8
21 40
5.0
18 7 30
4.5
4.0 15 20

3.5 12 6 10
0 10 20 30 40 50 0 10 20 30 40 50
(c) Capture time (ps) (d) Relaxation time (ps)

FIGURE 3.10 Influence of carrier scattering times on the AM response, resonance frequency, and damping
factor. (a) and (c) are for the variation of capture time with fixed relaxation time at 2.9 ps. (b) and (d) are for
the variation of relaxation time with fixed capture time at 6.3 ps.
From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art Lasers 119

For relaxation times tGS ES

> 25 ps, the response shows a clear resonance at ~8 GHz. From the eigen-
value analysis of the Qdot laser system [130], the resonance frequency f R and the damping factor
Γ are extracted for various carrier scattering times. It is shown that both f R and Γ decrease linearly
with increase in carrier capture time (Figure 3.10c). In contrast, for the carrier relaxation process
(Figure 3.10d) the behavior is much more complex. The resonance frequency first decreases for
tGS
ES
< 10 ps, while the damping factor increases with the relaxation time. At tGS ES
= 10 ps, the AM
response is rather flat. However, for tGS > 10 ps the resonance again increases until tGS
ES ES
= 25 ps,
while the damping factor reaches the maximum at tGS = 15 ps. Beyond the peak values, both the
ES

resonance and damping decrease as a function of the relaxation time. Lastly, we note that the damp-
ing factors of the Qdot laser in both Figure 3.10c and d are much larger than those of Qwell lasers,
which is attributed to the carrier occupation in the off-resonant states as well as the carrier scatter-
ing processes [131].
Figure 3.11a illustrates that the FM/AM index ratio 2β/m exhibits a significant re-increase beyond
10 GHz for a slow carrier capture process (large capture time). With respect to Equation 3.52, this
can be attributed to the larger carrier variation in the RS, δNRS, under high-frequency modulation

0.88
0.86
α-factor

10
0.84
0.82
0.80
2β/m

0 10 20 30 40 50
Capture time (ps)
Increase
capture time

0.1 1 10 100
(a) Modulation frequency (GHz)

100
1.1

1.0
α-factor

0.9

0.8
10 0 10 20 30 40 50
2β/m

Relaxation time (ps)

Increase
relaxation time

0.1 1 10 100
(b) Modulation frequency (GHz)

FIGURE 3.11 Influence of (a) the carrier capture time and (b) relaxation time on the ratio of FM/AM index.
Insets show the α-factor variation extracted from the minimum of the FM/AM index ratio.
120 Semiconductor Nanocrystals and Metal Nanoparticles

since the number of available carriers in the RS is larger. The inset of Figure 3.11a indicates that
increasing the carrier capture time enhances the α-factor by 11% from 0.80 for tES RS
= 0.10 ps to
0.88 for tES = 50 ps. Figure 3.11b depicts that a slow carrier relaxation process induces a steep
RS

­re-increase of 2β/m for modulation frequencies larger than 10 GHz. This is due to the increased
carrier populations and variations in the ES and RS. The inset of Figure 3.11b shows the α-factor
extracted from the minimum value of 2β/m, which increases by about 36% from 0.78 for tGS ES
= 0.1 ps
to 1.06 for tGS = 50 ps.
ES

3.4 CONCLUSION
In this chapter, we discussed the electronic and optical features of InP-based nanostructure semi-
conductor lasers. In contrast to conventional Qwell lasers, in Qdot and Qdash lasers the existence
of a carrier reservoir, discrete excited states, and the consequent carrier scattering processes bring
unique characteristics to the dynamic modulation response and the line-width enhancement factor.
Regarding the dynamical performance, it can be well improved by the excited-state lasing instead of
ground-state lasing [132]. On the other hand, nonlinear photonic techniques such as optical injection
[133,134], optical feedback [135], and optoelectronic feedback [136] can be employed for further
enhancement of the dynamical performance of nanostructure lasers.

ACKNOWLEDGMENTS
This work was supported in part by the Partenariat Hubert Curien under Grant No. 30794RC
(Campus France/DAAD) and by the European Office for Aerospace Research (EOARD) under
grant FA9550-15-1-0104.

REFERENCES
1. E. Murphy, Enabling optical communication, Nat. Photon. 4, 287 (2010).
2. H. Kroemer, Theory of a wide-gap emitter for transistors, Proc. IRE 45, 1535 (1957).
3. Z. I. Alferov and R. F. Kazarinov, Semiconductor laser with electric pumping, Inventor’s Certificate
181737 in Russian, Application 950840, priority as of March 30, 1963.
4. R. Dingle and C. H. Henry, Quantum effects in heterostructure lasers, U.S. Patent No. 3,982,207, filed
on March 7, 1975, issued September 21, 1976.
5. J. P. van der Ziel, R. Dingle, R. C. Miller, W. Wiegmann, and W. A. Nordland, Laser oscillations
from quantum states in very thin GaAs-Al0.2Ga0.8As multilayer structures, Appl. Phys. Lett. 26, 463
(1975).
6. Y. Arakawa and H. Sakaki, Multidimensional quantum well laser and temperature dependence of its
threshold current, Appl. Phys. Lett. 40, 939 (1982).
7. M. Asada, Y. Miyamoto, and Y. Suematsu, Gain and the threshold of three-dimensional quantum-box
lasers, IEEE J. Quantum Electron. 22, 1915 (1986).
8. D. Bimberg, M. Grundmann, and N. N. Ledentsov, Quantum Dot Heterostructures, New York: Wiley,
1998.
9. Y. Arakawa, Progress in growth and physics of nitride-based quantum dots, Phys. Status Solidi (a) 188,
37 (2001).
10. N. Kirstaedter, N. N. Ledentsov, M. Grundmann, D. Bimberg, V. M. Ustinov, S. S. Ruvimov,
M. V. Maximov, P. S. Kop’ev, and Zh. I. Alferov, Low threshold, large T0 injection laser emission from
(InGa)As quantum dots, Electron. Lett. 30, 1416 (1994).
11. D. Bimberg et al., InAs-GaAs quantum pyramid lasers: In situ growth, radiative lifetimes and polariza-
tion properties, Jpn. J. Appl. Phys. 35, 1311 (1996).
12. N. N. Ledentsov et al., Direct formation of vertically coupled quantum dots in Stranski–Krastanow
growth, Phys. Rev. B 54, 8743 (1996).
13. M. T. Crowley, N. A. Naderi, H. Su, F. Grillot, and L. F. Lester, GaAs based quantum dot lasers,
in Semiconductors and Semimetals: Advances in Semiconductor Lasers, J. J. Coleman and A. C. Bryce
(Eds.), 86 (2012).
From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art Lasers 121

14. H. Y. Liu, K. M. Groom, D. T. D. Childs, D. J. Robbins, T. J. Badcock, M. Hopkinson, D. J. Mowbray,

and M. S. Skolnick, 1.3 μm InAs/GaAs multilayer quantum-dot laser with extremely low room tempera-
ture threshold current density, Electron. Lett. 40, 1412 (2004).
15. M. V. Maksimov et al., High-power 1.5 μm InAs-InGaAs quantum dot lasers on GaAs substrates,
Semiconductors 38, 732 (2004).
16. A. Zhukov, M. Maksimov, and A. Kovsh, Device characteristics of long-wavelength lasers based on
self-organized quantum dots, Semiconductors 46, 1225 (2012).
17. P. B. Joyce, T. J. Krzyzewski, P. H. Steans, G. R. Bell, J. H. Neave, and T. S. Jones, Variations in critical
coverage for InAs/GaAs quantum dot formation in bilayer structures, J. Cryst. Growth 244, 39 (2002).
18. F. Y. Chang, C. C. Wu, and H. H. Lin, Effect of InGaAs capping layer on the properties of InAs/InGaAs
quantum dots and lasers, Appl. Phys. Lett. 82, 4477 (2003).
19. J. Oshinowo, M. Nishioka, S. Ishida, and Y. Arakawa, Highly uniform InGaAs/GaAs quantum dots
(15 nm) by metal organic chemical vapor deposition, Appl. Phys. Lett. 65, 1421 (1994).
20. P. J. Poole, K. Kaminska, P. Barrios, Z. Lu, and J. Liu, Growth of InAs/InP-based quantum dots for
1.55 μm laser applications, J. Cryst. Growth 311, 1482 (2009).
21. N. Bertru et al., QD laser on InP substrate for 1.55 µm emission and beyond, Proc. SPIE 7608, 76081B
(2010).
22. M. Gong, K. Duan, C. F. Li, R. Magri, A. Narvaez, and L. He, Electronic structure of self-assembled
InAs/InP quantum dots: Comparison with self-assembled InAs/GaAs quantum dots, Phys. Rev. B 77,
045326 (2008).
23. R. H. Wang, A. Stintz, P. M. Varangis, T. C. Newell, H. Li, K. J. Malloy, and L. F. Lester, Room-temperature
operation of InAs quantum-dash lasers on InP(001), IEEE Photon. Technol. Lett. 13, 767 (2001).
24. P. Miska, J. Even, C. Platz, B. Salem, and T. Benyattou, Experimental and theoretical investigation of
carrier confinement in InAs quantum dashes grown on InP(001), J. Appl. Phys. 95, 1074 (2004).
25. M. Z. M. Khan, T. K. Ng, and B. S. Ooi, Self-assembled InAs/InP quantum dots and quantum dashes:
Material structures and devices, Prog. Quantum Electron. 38, 6, 237 (2014).
26. J. P. Reithmaier, G. Eisenstein, and A. Forchel, InAs/InP quantum-dash lasers and amplifiers, Proc.
IEEE 95, 1779 (2007).
27. D. Zhou, R. Piron, F. Grillot, O. Dehaese, E. Homeyer, M. Dontabactouny, T. Batte, K. Tavernier,
J. Even, and S. Loualiche, Study of the characteristics of 1.55 μm quantum dash/dot semiconductor
lasers on InP substrate, Appl. Phys. Lett. 93, 161104 (2008).
28. D. Zhou, R. Piron, M. Dontabactouny, O. Dehaese, F. Grillot, and T. Batte, Low threshold current
density of InAs quantum dash laser on InP(100) through optimizing double cap technique, Appl. Phys.
Lett. 94, 081107 (2009).
29. V. Sichkovskyi, M. Waniczek, and J. Reithmaier, High-gain wavelength-stabilized 1.55 μm InAs/InP(100)
based lasers with reduced number of quantum dot active layers, Appl. Phys. Lett. 102, 221117 (2013).
30. C. Paranthoën et al., Height dispersion control of InAs/InP(113)B quantum dots emitting at 1.55 μm,
Appl. Phys. Lett. 78, 1751 (2001).
31. J. Kotani, P. J. van Veldhoven, T. de Vries, B. Smalbrugge, E. A. J. M. Bente, M. K. Smit, and R. Notzel,
First demonstration of single-layer InAs/InP (100) quantum-dot laser: Continuous wave, room tempera-
ture, ground state, Electron. Lett. 45, 1317 (2009).
32. E. Homeyer, R. Piron, F. Grillot, O. Dehaese, K. Tavernier, E. Macé, A. Le Corre, and S. Loualiche,
First demonstration of a 1.52 μm RT InAs/InP (311)B laser with an active zone based on a single QD
layer, Semicond. Sci. Technol. 22, 827 (2007).
33. C. Cornet et al., Electronic and optical properties of InAs/InP quantum dots on InP(100) and InP(311)B
substrates: Theory and experiment, Phys. Rev. B 74, 035312 (2006).
34. V. Ustinov, A. Zhukov, A. Y. Egorov, A. Kovsh, S. Zaitsev, and N. Y. Gordeev, Low threshold quantum
dot injection laser emitting at 1.9 μm, Electron. Lett. 34, 670 (1998).
35. V. Ustinov, A. Kovsh, A. Zhukov, A. Y. Egorov, N. N. Ledentsov, and A. V. Lunev, Low-threshold
quantum-dot injection heterolaser emitting at 1.84 μm, Tech. Phys. Lett. 24, 22 (1998).
36. K. Nishi, M. Yamada, T. Anan, A. Gomyo, and S. Sugou, Long-wavelength lasing from InAs self-
assembled quantum dots on (311)B InP, Appl. Phys. Lett. 73, 526 (1998).
37. N. Bertru et al., Two-dimensional ordering of self-assembled InAs quantum dots grown on (311)B InP
substrate, in Proceedings of SPIE, QD laser on InP substrate for 1.55 um emission and beyond, San
Francisco, CA, Vol. 7608, p. 76081B (2010).
38. P. Miska, J. Even, C. Paranthoën, O. Dehaese, H. Folliot, S. Loualiche, M. Senes, and X. Marie, Optical
properties and carrier dynamics of InAs/InP(113)B quantum dots emitting between 1.3 and 1.55 μm for
laser applications, Physica E 17, 56 (2003).
122 Semiconductor Nanocrystals and Metal Nanoparticles

39. C. Paranthoën et al., Growth and optical characterizations on InAs quantum dots on InP substrate:
Toward 1.55 μm quantum dot laser, J. Cryst. Growth 251, 230 (2003).
40. P. Caroff, C. Paranthoen, C. Platz, O. Dehaese, H. Folliot, and N. Bertru, High-gain and low-threshold
InAs quantum-dot lasers on InP, Appl. Phys. Lett. 87, 243107 (2005).
41. E. Homeyer, R. Piron, F. Grillot, O. Dehaese, K. Tavernier, and E. Macé, Demonstration of a low thresh-
old current in 1.54 μm InAs/InP (311)B quantum dot laser with reduced quantum dot stacks, Jpn. J.
Appl. Phys. 46, 6903 (2007).
42. K. Klaime, C. Clo, R. Piron, C. Paranthoen, D. Thiam, and T. Batte, 23 and 39 GHz low phase noise
monosection InAs/InP (113)B quantum dots mode-locked lasers, Opt. Express 21, 29000 (2013).
43. H. Saito, K. Nishi, and S. Sugou, Ground-state lasing at room temperature in long-wavelength InAs
quantum-dot lasers on InP(311)B substrates, Appl. Phys. Lett. 78, 267 (2001).
44. K. Akahane, N. Yamamoto, and T. Kawanishi, Wavelength tunability of highly stacked quantum dot
laser fabricated by a strain compensation technique, in Proceedings of the 22nd IEEE International
Semiconductor Laser Conference (ISLC), Cardiff, U.K., Vol. 37 (2010).
45. K. Akahane, N. Yamamoto, and T. Kawanishi, The dependence of the characteristic temperature of
highly stacked InAs quantum dot laser diodes fabricated using a strain-compensation technique on
stacking layer number, in Proceedings of the 22nd IEEE International Semiconductor Laser Conference
(ISLC), Cardiff, U.K., cc Vol. 82 (2012).
46. C. N. Allen, P. Poole, P. Barrios, P. Marshall, G. Pakulski, and S. Raymond, External cavity quantum
dot tunable laser through 1.55 μm, Physica E 26, 372 (2005).
47. F. Lelarge, B. Rousseau, B. Dagens, F. Poingt, F. Pommereau, and A. Accard, Room temperature
continuous-wave operation of buried ridge stripe lasers using InAs-InP(100) quantum dots as active
core, IEEE Photon. Technol. Lett. 17, 1369 (2005).
48. S. Anantathanasarn et al., Lasing of wavelength-tunable (1.55 μm region) InAs/InGaAsP/InP (100)
quantum dots grown by metal organic vapor-phase epitaxy, Appl. Phys. Lett. 89, 073115 (2006).
49. J. S. Kim, J. H. Lee, S. U. Hong, W. S. Han, H. S. Kwack, and C. W. Lee, Long-wavelength laser based
on self-assembled InAs quantum dots in InAlGaAs on InP(001), Appl. Phys. Lett. 85, 1033 (2004).
50. J. S. Kim, J. H. Lee, S. U. Hong, W. S. Han, H. S. Kwack, and C. W. Lee, Room-temperature operation
of InP-based quantum dot laser, IEEE Photon. Technol. Lett. 16, 1607 (2004).
51. C. Gilfert, V. Ivanov, N. Oehl, M. Yacob, and J. Reithmaier, High gain 1.55 μm diode lasers based on
InAs quantum dot like active regions, Appl. Phys. Lett. 98, 201102 (2011).
52. O. Mollet, A. Martinez, K. Merghem, S. Joshi, J.-G. Provost, F. Lelarge, and A. Ramdane, Dynamic
characteristics of undoped and p-doped Fabry-Perot InAs/InP quantum dash based ridge waveguide
lasers for access/metro networks, Appl. Phys. Lett. 105, 141113 (2014).
53. C. Peucheret, Direct and External Modulation of Light, Technical University of Denmark, Kongens
Lyngby, Denmark, 2009.
54. B. O. Seraphin and N. Bottka, Franz-Keldysh effect of the refractive index in semiconductors, Phys.
Rev. 139, A560 (1965).
55. K. Kechaou, T. Anfray, K. Merghem, C. Aupetit-Berthelemot, G. Aubin, C. Kazmierski, C. Jany,
P. Chanclou, and D. Erasme, Improved NRZ transmission distance at 20 Gbit/s using dual electro­
absorption modulated laser, Electron. Lett. 48, 335 (2012).
56. D. Erasme et al., The dual-electroabsorption modulated laser, a flexible solution for amplified and
dispersion uncompensated networks over standard fiber, J. Lightwave Technol. 32, 4068 (2014).
57. A. Martinez et al., Dynamic properties of InAs/InP(311B) quantum dot Fabry-Perot lasers emitting at
1.52-μm, Appl. Phys. Lett. 93, 021101 (2008).
58. D. Gready, G. Eisenstein, C. Gilfert, V. Ivanov, and J. P. Reithmaier, High-speed low-noise InAs/
InAlGaAs/InP 1.55-μm quantum-dot lasers, IEEE Photon. Technol. Lett. 24, 809 (2012).
59. D. Gready, G. Eisenstein, V. Ivanov, C. Gilfert, F. Schnabel, A. Rippien, J. P. Reithmaier, and
C. Bornholdt, High speed 1.55 μm InAs/InAlGaAs/InP quantum dot laser, IEEE Photon. Technol. Lett.
26, 11 (2014).
60. S. Bhowmick, M. Z. Baten, T. Frost, B. S. Ooi, and P. Bhattacharya, High performance InAs/
In0.53Ga0.23Al0.24As/InP quantum dot 1.55 μm tunnel injection laser, IEEE J. Quantum Electron.
50, 7 (2014).
61. W. Kaiser, K. Mathwig, S. Deubert, J. P. Reithmaier, F. Forchel, O. Parillaud, M. Krakowski, D. Hadass,
V. Mikhelashvili, and G. Eisenstein, Static and dynamic properties of laterally coupled DFB lasers
based on InAs/InP Qdash structures, Electron. Lett. 41 (2005).
62. Z. Mi and P. Bhattacharya, DC and dynamic characteristics of p-doped and tunnel injection 1.65 μm
InAs quantum-dash lasers grown on InP(001), IEEE J. Quantum Electron. 42, 1224 (2006).
From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art Lasers 123

63. S. Hein, V. von Hinten, W. Kaiser, S. Hofling, and A. Forchel, Dynamic properties of 1.5 μm quantum
dash lasers on (100) InP, Electron. Lett. 43 (2007).
64. F. Lelarge et al., Recent advances on InAs/InP quantum dash based semiconductor lasers and optical
amplifiers operating at 1.55 μm, IEEE J. Sel. Top. Quantum Electron. 13, 111 (2007).
65. Q. Zou, K. Merghem, S. Azouigui, A. Martinez, A. Accard, N. Chimot, F. Lelarge, and A. Ramdane,
Feedback-resistant p-type doped InAs/InP quantum-dash distributed feedback lasers for isolator-free
10 Gb/s transmission at 1.55 μm, Appl. Phys. Lett. 97, 231115 (1010).
66. N. Chimot, S. Joshi, G. Aubin, K. Merghem, S. barbet, A. Accard, A. Ramdane, and F. Lelarge, 1550 nm
InAs/InP quantum dash based directly modulated lasers for next generation passive optical network,
IEEE, Int. Conf. Indium Phosphide and Related Materials (IPRM), Santa Barbara, CA, p. 177 (2012).
67. S. Joshi, N. Chimot, L. A. Neto, A. Accard, and J. G. Provost, Quantum dash based directly modulated
lasers for long-reach access networks, Electron. Lett. 50, 534 (2014).
68. R. S. Tucker, Green optical communications—Part I: Energy limitations in transport, IEEE J. Sel. Top.
Quantum Electron. 17, 245 (2011).
69. D. Gready, G. Eisenstein, M. Gioannini, I. Montrosset, D. Arsenijevic, H. Schmeckebier, M. Stubenrauch,
and D. Bimberg, On the relationship between small and large signal modulation capabilities in highly
nonlinear quantum dot lasers, Appl. Phys. Lett. 102, 101107 (2013).
70. S. Joshi, N. Chimot, A. Ramdane, and F. Lelarge, On the nature of the linewidth enhancement factor in
p-doped quantum dash based lasers, Appl. Phys. Lett. 105, 241117 (2014).
71. M. Gioannini and I. Montrosset, Numerical analysis of the frequency chirp in quantum-dot semicon-
ductor lasers, IEEE J. Quantum Electron. 43, 941 (2007).
72. C. Cornet, C. Platz, P. Caroff, J. Even, C. Labbé, H. Folliot, and A. Le Corre, Approach to wetting-layer-
assisted lateral coupling of InAs/InP quantum dots, Phys. Rev. B 72, 035342 (2005).
73. R. Heitz, F. Guffarth, K. Poetschke, A. Schliwa, D. Bimberg, N. D. Zakharov, and P. Werner, Shell-like
formation of self-organized InAs/GaAs quantum dots, Phys. Rev. B 71, 045325 (2005).
74. A. J. Nozik, Multiple exciton generation in semiconductor quantum dots, Chem. Phys. Lett. 457, 3
(2008).
75. P. Miska, J. Even, O. Dehaese, and X. Marie, Carrier relaxation dynamics in InAs/InP quantum dots,
Appl. Phys. Lett. 92, 191103 (2008).
76. K. Veselinov, F. Grillot, C. Cornet, J. Even, A. Bekiarski, M. Gioannini, and S. Loualiche, Analysis of
the double laser emission occurring in 1.55-μm InAs-InP(113)B quantum-dot lasers, IEEE J. Quantum
Electron. 43, 810 (2007).
77. R. Nagarajan, M. Ishikawa, T. Fukushima, R. Geels, and J. Bowers, High speed quantum well lasers and
carrier transport effects, IEEE J. Quantum Electron. 28, 1990 (1992).
78. L. A. Coldren and S. W. Corzine, Diode Lasers and Photonic Integrated Circuits, New York: Wiley,
1995.
79. J. Siegert, S. Marcinkevicius, and Q. X. Zhao, Carrier dynamics in modulation-doped InAs/GaAs quan-
tum dots, Phys. Rev. B 72, 085316 (2005).
80. S. Marcinkevicius and R. Leon, Carrier capture and escape in InxGa1-xAs/GaAs quantum dots: Effects
of intermixing, Phys. Rev. B 59, 4630 (1999).
81. T. R. Nielsen, P. Gartner, and F. Jahnke, Many-body theory of carrier capture and relaxation in semicon-
ductor quantum-dot lasers, Phys. Rev. B 69, 235314 (2004).
82. B. Ohnesorge, M. Albrecht, J. Oshinowo, A. Forchel, and Y. Arakawa, Rapid carrier relaxation in self-
assembled InxGa1-xAs/GaAs quantum dots, Phys. Rev. B 54, 11532 (1996).
83. I. V. Ignatiev, I. E. Kozin, S. V. Nair, H. W. Ren, S. Sugou, and Y. Masumoto, Carrier relaxation dynam-
ics in InP quantum dots studied by artificial control of nonradiative losses, Phys. Rev. B 61, 15633
(2000).
84. M. Lorke, T. R. Nielsen, J. Seebeck, P. Gartner, and F. Jahnke, Influence of carrier-carrier and carrier-
phonon correlations on optical absorption and gain in quantum-dot systems, Phys. Rev. B 73, 085324
(2006).
85. I. Magnusdottir, S. Bischoff, A. V. Uskov, and J. Mork, Geometry dependence of Auger carrier capture
rates into cone-shaped self-assembled quantum dots, Phys. Rev. B 67, 205326 (2003).
86. U. Bockelmann and T. Egeler, Electron relaxation in quantum dots by means of Auger processes, Phys.
Rev. B 46, 15574 (1992).
87. P. Ferreira and G. Bastard, Phonon assisted capture and intra-dot Auger relaxation in quantum dots,
Appl. Phys. Lett. 74, 2818 (1999).
88. P. Miska, J. Even, X. Marie, and O. Dehaese, Electronic structure and carrier dynamics in InAs/InP
double-cap quantum dots, Appl. Phys. Lett. 94, 061916 (2009).
124 Semiconductor Nanocrystals and Metal Nanoparticles

89. A. V. Uskov, J. McInerney, F. Adler, H. Schweizer, and M. H. Pilkuhn, Auger carrier capture kinetics in
self-assembled quantum dot structures, Appl. Phys. Lett. 72, 58 (1998).
90. A. V. Uskov, F. Adler, H. Schweizer, and M. H. Pilkuhn, Auger carrier relaxation in self-assembled
quantum dots by collisions with two-dimensional carriers, J. Appl. Phys. 81, 7895 (1997).
91. W. W. Chow and S. W. Koch, Theory of semiconductor quantum-dot laser dynamics, IEEE J. Quantum
Electron. 41, 495 (2005).
92. W. W. Chow and S. W. Koch, Semiconductor-Laser Fundamentals, Berlin, Germany: Springer, 1999.
93. T. W. Berg, S. Bischoff, I. Magnusdottir, and J. Mørk, Ultrafast gain recovery and modulation limita-
tions in self-assembled quantum-dot devices, IEEE Photon. Technol. Lett. 13, 541 (2001).
94. W. W. Chow, M. Lorke, and F. Jahnke, Will quantum dots replace quantum wells as the active medium
of choice in future semiconductor lasers, IEEE J. Sel. Top. Quantum Electron. 17, 1349 (2011).
95. T. Takahashi and Y. Arakawa, Nonlinear gain effects in quantum well, quantum well wire, and quantum
well box lasers, IEEE J. Quantum Electron. 27, 1824 (1991).
96. M. Willatzen, A. Uskov, J. Mork, H. Olesen, B. Tromborg, and A. P. Jauho, Nonlinear gain suppression
in semiconductor lasers due to carrier heating, IEEE Photon. Technol. Lett. 3, 606 (1991).
97. D. J. Klotzkin, Introduction to Semiconductor Lasers for Optical Communications, Springer, New York, 2014.
98. Y. Lam and J. Singh, Monte Carlo simulation of gain compression effects in GRINSCH quantum well
laser structures, IEEE J. Quantum Electron, New York, 30, 2435 (1994).
99. H. C. Schneider, W. W. Chow, and S. W. Koch, Anomalous carrier-induced dispersion in quantum-dot
active media, Phys. Rev. B 66, 041310(R) (2002).
100. A. V. Uskov, E. P. O’Reilly, D. McPeake, N. N. Ledentsov, D. Bimberg, and G. Huyet, Carrier-induced
refractive index in quantum dot structures due to transitions from discrete quantum dot levels to con-
tinuum states, Appl. Phys. Lett. 84, 272 (2004).
101. S. P. Hegarty, B. Corbett, J. G. McInerney, and G. Huyet, Free-carrier effect on index change in 1.3 µm
quantum-dot lasers, Electron. Lett. 41, 416 (2005).
102. M. Osiński and J. Buus, Linewidth broadening factor in semiconductor lasers—An overview, IEEE
J. Quantum Electron. QE-23, 9 (1987).
103. G. H. Duan, P. Gallion, and G. Debarge, Analysis of the phase-amplitude coupling factor and spec-
tral linewidth of distributed feedback and composite-cavity semiconductor lasers, IEEE J. Quantum
Electron. 26, 32 (1990).
104. G. Duan, P. Gallion, and G. Gebarge, Analysis of frequency chirping of semiconductor lasers in
presence of optical feedback, Opt. Lett. 12, 800 (1987).
105. G. P. Agrawal, Intensity dependence of the linewidth enhancement factor and its implications for
­semiconductor lasers, IEEE Photon. Technol. Lett. 1, 212 (1989).
106. S. Wieczorek, B. Krauskopf, and D. Lenstra, Multipulse excitability in a semiconductor laser with
­optical injection, Phys. Rev. Lett. 88, 063901 (2002).
107. B. Haegeman, K. Engelborghs, D. Roose, D. Pierous, and T. Erneux, Stability and rupture of bifurcation
bridges in semiconductor lasers subject to optical feedback, Phys. Rev. E 66, 046216 (2002).
108. M. Sciamanna, P. Mégret, and M. Blondel, Hopft bifurcation cascade in small-α laser diodes subject to
optical feedback, Phys. Rev. E 69, 046209 (2004).
109. K. Panajotov, M. Sciamanna, M. Arteaga, and H. Thienpont, Optical feedback in vertical-cavity
­surface-emitting lasers, IEEE J. Sel. Top. Quantum Electron. 19, 1700312 (2012).
110. T. C. Newell, D. J. Bossert, A. Stintz, B. Fuchs, K. J. Malloy, and L. F. Lester, Gain and linewidth
enhancement factor in InAs quantum-dot laser diodes, IEEE Photon. Technol. Lett. 11, 1527(1999).
111. Z. Mi, P. Bhattacharya, and S. Fathpour, High-speed 1.3 μm tunnel injection quantum-dot lasers, Appl.
Phys. Lett. 86, 153109 (2005).
112. B. Dagens, A. Markus, J. X. Chen, J. G. Provost, D. Make, O. Le Goueziou, J. Landreau, A. Foire, and
B. Thedrez, Gaint linewidth enhancement factor and purely frequency modulated emission from quan-
tum dot laser, Electron. Lett. 41, 323 (2005).
113. F. Grillot, K. Veselinov, M. Gioannini, I. Montrosset, J. Even, R. Piron, E. Homeyer, and S. Loualiche,
Spectral analysis of 1.55 μm InAs–InP(113)B quantum-dot lasers based on a multipopulation rate equa-
tions model, IEEE J. Quantum Electron. 45, 872 (2009).
114. C. Wang, F. Grillot, and J. Even, Impacts of wetting layer and excited state on the modulation response
of quantum-dot lasers, IEEE J. Quantum Electron. 48, 1144 (2012).
115. M. Kuntz, Modulated InGaAs/GaAs quantum dot lasers, PhD thesis, Berlin, Germany, 2006.
116. N. A. Naderi, External control of semiconductor nanostructure lasers, PhD thesis, University of
New Mexico, Albuquerque, NM, 2011.
From Basic Physical Properties of InAs/InP Quantum Dots to State-of-the-Art Lasers 125

117. A. E. Zhukov, M. V. Maximov, A. V. Savelyev, Yu. M. Shernyakov, F. I. Zubov, V. V. Korenev,

A. Martinez, A. Ramdane, J.-G. Provost, and D. A. Livshits, Gain compression and its dependence
on output power in quantum dot lasers, J. Appl. Phys. 113, 233103 (2013).
118. J. Oksanen and J. Tulkki, Linewidth enhancement factor and chirp in quantum dot lasers, J. Appl. Phys.
94, 1983 (2003).
119. S. Melnik, G. Huyet, and A. V. Uskov, The linewidth enhancement factor α of quantum dot semiconduc-
tor lasers, Opt. Express 14, 2950 (2006).
120. S. Wieczorek, B. Krauskopf, T. B. Simpson, and D. Lenstra, The dynamical complexity of optically
injected semiconductor lasers, Phys. Rep. 416, 1 (2005).
121. W. W. Chow and F. Jahnke, On the physics of semiconductor quantum dots for applications in lasers and
quantum optics, Prog. Quantum Electron. 37, 109 (2013).
122. A. Markus, J. X. Chen, O. Gauthier-Lafaye, J. G. Provost, C. Paranthoen, and A. Fiore, Impact of intra-
band relaxation on the performance of a quantum-dot laser, IEEE J. Sel. Top. Quantum Electron. 9,
1308 (2003).
123. B. Lingnau, W. W. Chow, E. Schöll, and K. Lüdge, Feedback and injection locking instabilities in
quantum-dot lasers: A microscopically based bifurcation analysis, New J. Phys. 15, 093031 (2013).
124. F. Grillot, B. Dagens, J. G. Provost, H. Su, and L. F. Lester, Gain compression and above-threshold
linewidth enhancement factor in 1.3 µm InAs-GaAs quantum-dot lasers, IEEE J. Quantum Electron.
44, 946 (2008).
125. J.-G. Provost and F. Grillot, Measuring the chirp and the linewidth enhancement factor of optoelectronic
devices with a Mach-Zehnder interferometer, IEEE Photon. J. 3, 476 (2011).
126. C. Cornet, C. Labbé, H. Folliot, N. Bertru, O. Dehaese, J. Even, A. Le Corre, C. Paranthoën, C. Platz,
and S. Loualiche, Quantitative investigations of optical absorption in InAs/InP (311)B quantum dots
emitting at 1.55 μm wavelength, Appl. Phys. Lett. 85, 5685 (2004).
127. Z. Mi and P. Bhattacharya, Analysis of the linewidth-enhancement factor of long-wavelength tunnel-
injection quantum-dot lasers, IEEE J. Quantum Electron. 43, 363 (2007).
128. C. Wang, M. Osiński, J. Even, and F. Grillot, Phase-amplitude coupling characteristics in directly
­modulated quantum dot lasers, Appl. Phys. Lett. 105, 221114 (2014).
129. S. Gerhard, C. Schilling, F. Gerschutz, M. Fischer, J. Koeth, I. Krestnikov, A. Kovsh, M. Kamp,
S. Hofling, and A. Forchel, Frequency-dependent linewidth enhancement factor of quantum-dot lasers,
IEEE Photon. Technol. Lett. 20, 1736 (2008).
130. K. Lüdge and H. G. Schuster, Nonlinear Laser Dynamics: From Quantum Dots to Cryptography,
New York: Wiley, 2011.
131. B. Lingnau, K. Lüdge, W. W. Chow, and E. Schöll, Influencing modulation properties of quantum-dot
semiconductor lasers by carrier lifetime engineering, Appl. Phys. Lett. 101, 131107 (2012).
132. C. Wang, B. Lingnau, K. Lüdge, J. Even, and F. Grillot, Enhanced dynamic performance of quantum dot
semiconductor lasers operating on the excited state, IEEE J. Quantum Electron. 50, 723 (2014).
133. T. B. Simpon, J. M. Liu, and A. Gavrielides, Bandwidth enhancement and broadband noise reduction in
injection-locked semiconductor lasers, IEEE Photon. Technol. Lett. 7, 709 (1995).
134. A. Murakami, K. Kawashima, and K. Atsuki, Cavity resonance shift and bandwidth enhancement in
semiconductor lasers with strong light injection, IEEE J. Quantum Electron. 39, 1196 (2003).
135. F. Grillot, C. Wang, N. A. Naderi, and J. Even, Modulation properties of self-injected quantum-dot
semiconductor diode lasers, IEEE J. Sel. Top. Quantum Electron. 19, 1900812 (2013).
136. J.-P. Zhuang and S.-C. Chan, Phase noise characteristics of microwave signals generated by semicon-
ductor laser dynamics, Opt. Express 23, 2777 (2015).
 4 Optical, Photoluminescence,
and Vibrational Spectroscopy
of Metal Nanoparticles
P. Gangopadhyay

CONTENTS
4.1 Introduction........................................................................................................................... 127
4.2 Nano-Basics........................................................................................................................... 130
4.2.1 Particle Sizes versus Physical Properties................................................................... 130
4.2.2 Theoretical Framework.............................................................................................. 131
4.2.2.1 Optical Response of Metal Nanoparticles.................................................. 131
4.2.2.2 Vibrational Spectroscopy of Nanoparticles................................................ 140
4.3 Synthesis of Metal Nanoparticles.......................................................................................... 146
4.3.1 Bottom-Up Methods.................................................................................................. 147
4.3.1.1 Ion Implantation.......................................................................................... 147
4.3.1.2 Ion-Exchange Processing............................................................................ 152
4.3.2 Top-Down Methods................................................................................................... 174
4.3.2.1 Ion-Beam Sputtering Phenomena............................................................... 176
4.3.2.2 Ion-Beam Sputtering Experiments............................................................. 177
4.3.2.3 Rutherford Backscattering Results............................................................. 178
4.3.2.4 Electron Microscopy Results...................................................................... 179
4.3.2.5 Optical Absorption Results......................................................................... 181
4.4 Summary............................................................................................................................... 183
Acknowledgments........................................................................................................................... 183
References....................................................................................................................................... 183

4.1 INTRODUCTION
Historically, centuries ago fine precipitates of noble metals (gold, silver, and copper) were used in
tinted glasses to produce beautiful colors of windowpanes used for decorations and in Lycurgus
cups (developed in the fourth century AD by Roman glass workers; visit http://www.bmimages.
com for the images). The first insight into the beautiful colors of stained glasses was provided
by Michael Faraday in 1857, and he concluded thus: “I think that in all these cases the ruby tint
is due simply to the presence of diffused finely-divided gold” (Faraday 1857). In 1908, about
50 years later, Gustav Mie theoretically explained the appearance of the striking colors as due to
the resonant absorption of light in these fine metal particles (Mie 1908). This resonant absorp-
tion of light, which arises from the collective oscillations of the conduction electrons in metal
nanoparticles, is being described as localized surface-plasmon resonance (SPR) in the current
literature (Kreibig and Vollmer 1995, Quinten 2011). Plasmons are defined as quanta of collec-
tive oscillations of conduction electrons in metal nanoparticles vis-à-vis metal nanostructures
(Ghosh and Pal 2007) and also in doped semiconductors (Luther et al. 2011). In this chapter,

127
128 Semiconductor Nanocrystals and Metal Nanoparticles

we discuss the interesting features of the optical responses of various metal nanoparticles and
metal–semiconductor hybrid nanostructures in detail.
Scientifically interesting and enriching phenomena, for instance, strong optical absorption and
photoluminescence (PL) responses in the visible range of wavelength (Bohren and Huffman 1983,
Gangopadhyay et al. 2005, Xu and Suslick 2010), local field enhancements and Raman spectroscopy
(Moskovits 1985, Kneipp et al. 2002, Haynes and Van Duyne 2003), sub-wavelength optics (Salerno
et al. 2002), fast optical switching (Maier et al. 2001), plasmonics (Maier 2007, Murray and Barnes
2007), and so on, are fundamentally associated with various nanoscale metal particles and metal
nanostructures. These exciting physical properties have been driving experimental and theoretical
research for a long time (Roco et al. 1999, Allegrini et al. 2001, Hosokawa et al. 2012). Quantum-
size behavior of metal nanoparticles, for example, has opened new directions in many fields of cur-
rent research and modern technologies, such as electronic logic operations (Lee and Dickson 2003),
catalytic and photocatalytic reactions (Bell 2003, Prieto et al. 2013), nanotweezers (Boggild et al.
2001), nanosensors (Luo et al. 2006, Saha et al. 2012), integrated biosensors (Hoa et al. 2007), and
others. The field has witnessed great momentum of research studies and innovative developments
during the last two decades. With the help of the transmission electron microscope, and particularly
after the advent of sophisticated instruments like scanning tunneling microscope (Binnig et al.
1982a,b), the synthesis (e.g., nanolithography) and characterization (e.g., high-resolution surface
microscopy and spectroscopy) of various nanomaterials have advanced. Topographic images of
surfaces with atomic-scale resolution have been obtained with the tunneling microscope.
In fact, prior to the inventions of the scanning tunneling microscope, the lecture “There is
plenty of room at the bottom—An Invitation to Enter a New Field of Physics” in 1959 by R. P. Feynman
at Caltech had motivated and thrown many challenges at contemporary scientists and engineers
(Feynman 1992, 1993). The bottom-room (i.e., the scope of science and technology with nanometer-
scale materials) has really expanded with time globally and is continually being established
by the multidisciplinary researchers across the boundaries of various scientific disciplines and
technologists of diverse interests. These innovations and perspective developments, in effect,
have touched human lives with all the benefits that could be envisaged (Xiao et al. 2003, Atwater
2007, Jie et al. 2007, Pradeep and Anshup 2009). Almost everyday, the present world is observing
advanced and breakthrough discoveries that lead to miniaturization of devices. Technological
advancements are being realized through the knowledge of nanoscience and nanotechnology.
The science of very small things is catching up in very big ways. Nanotechnology is providing the
handle to manipulate materials at the nanoscale regimes. Persistent global research activities and the
large number of articles appearing in important international journals are the real testimony on
this subject. “Nano” is a common buzzword in this context. According to the definition, nanoma-
terials (metals, semiconductors, or insulators) are no different from the parent phase of the bulk
materials but for their physical dimensions, which range between 1 and 100 nm (1 nm = 10 −9 m),
at least in one direction. At nanoscale dimensions, the physical, chemical, and biological prop-
erties of materials differ in fundamental and favorable ways from those of individual atoms
and molecules or bulk matter. That brings up spiraling applications of various nanomaterials in
almost every sphere of life. Hence, understanding the sciences of various synthetic methods and
novel physical properties of matter at the nanoscale is one of the most active areas of research
today (Nalwa 2004, Hosokawa et al. 2012). Extensions of these scientific studies are again very
relevant to significant developments in multidisciplinary fields spanning physics, chemistry, and
biology. It brings out many unexpected results to ponder over. In this perspective, nanoscale metal
particles, quantum dots (for semiconductors), and metal–semiconductor hybrid nanostructures,
which form important classes in the world of nanomaterials, deserve special attention for com-
prehensive research studies (Chang et al. 2006, Akimov et al. 2007).
It is well known now that nanoscale metal structures and metal nanoparticles support local-
ized surface-plasmon modes. Additionally, the planar surface of a metal (e.g., smooth surface of
metal nanowires, Sanders et al. 2006) also supports plasmon modes, known as surface-plasmon
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 129

polaritons (SPPs). These are the propagating excitations of charge-density waves. They propagate,
typically, over a few tens of micrometers (Murray and Barnes 2007). The propagation length
depends on the diameter of the metal nanowires or the width of nanostripes (Salerno et al. 2002).
In close proximity of an optical emitter (e.g., quantum dot) with a metal nanowire, emission from the
quantum dot can be guided in the nanowire because of the strong coupling of the emitted photon with
the SPP modes (Chang et al. 2007, Fedutik et al. 2007). It is like seeing lights at the end of a normal
dielectric-based optical fiber, making the way for highly efficient quantum optoelectronic devices. In a
suitably designed nanostructured system of hybrid materials, it may be interesting to study the photon
correlations among the quantum dots and the guiding SPP modes of a metal nanowire.
The prospects of a wide range of synthesis methods, basic research, and related applications of
the novel nanomaterials are great. In recent times, researchers have demonstrated that semiconduc-
tor quantum dots are also useful in quantum communication and computation devices (Luther et al.
2011). These novel materials can facilitate exchanging information at the maximum communica-
tion speed (nearly the speed of light), overcoming the bottleneck of electronic transport of data in
present-day conventional computers. Their scope is unlimited, but to mention a few in a limited
domain, they may include, for example, high-sensitivity nanosensors for chemical and biological
detectors (Alivisatos 2005), enhanced chemical and photocatalysis using plasmonic nanostruc-
tures (Moser 1996, Campbell et al. 2002, Christopher et al. 2012), surface-enhanced Raman spec-
troscopy (Moskovits 1985, Xu et al. 1999, Tao et al. 2003), efficiency improvements of solar cells
(Nakayama et al. 2008, Atwater and Polman 2010, Bora et al. 2013, Gao et al. 2013), photothermal
effects (Xiong et al. 2012), optical cloaking (invisibility) (Atwater 2007, Alù and Engheta 2008),
tumor imaging and therapy (Loo et al. 2005, Atwater 2007), and so on. Plain silicate glasses con-
taining metallic particles of few nanometers in size are also attractive materials for optical switch-
ing applications in integrated-optical devices and fast optical computers (Hughes et al. 2014). Large
enhancement (a few orders of magnitudes higher than in their bulk counterpart) of the nonlinear
optical susceptibility (χ(3)), for instance, near the SPR frequency, with a fast optical response time
(on the order of a few picoseconds) makes the metal–glass nanocomposite materials technologically
attractive for plasmonic optical devices (Faccio et al. 1998, Allegrini et al. 2001, Hamanaka et al.
2004, Bozhevolnyi et al. 2006).
For the most part, this chapter is designed to provide a review of the growth of metal–
semiconductor nanoparticles, as well as various synthesis processes and experimental techniques
to study their optical, vibrational (Raman), and PL properties at the nanoscale. Optical proper-
ties of noble metal nanoparticles, for instance, of sizes ~10 nm, show expected results that could
be explained using Drude’s theory for free-electron metals (Kreibig and Vollmer 1995). However,
different and interesting optical responses have been observed in the case of silver particles while
reducing their size to the range of ~2 nm (Srivastava et al. 2014). Quantum mechanical models have
been employed to explain these results. Surface-confined acoustic vibration modes in metal or semi-
conductor nanoparticles are usually observed in the vibrational (phonon) spectra of these materials
(Saviot et al. 2011). They reveal the nature of spheroidal (radial and quadrupolar) oscillations of
these nanoparticles. A thorough understanding of the vibrations of nanoparticles is necessary to
elucidate some of the novel properties that these nanomaterials exhibit, such as melting (Bottani
et al. 2001), electron–phonon coupling (Bachelier and Mlayah 2004), and so on. Surface-plasmon-
enhanced Raman scattering of metal nanoparticles will be discussed here in detail to explain the
plasmon–phonon coupling in such cases (Gangopadhyay et al. 2007). Examples of metal–semi-
conductor hybrid nanostructures will be discussed to elaborate quasi-particle interactions, possible
energy transfers, and interesting PL behaviors of these novel materials. Compared to the bandgap
emission, PL of semiconductors may be quenched or enhanced in metal–semiconductor hybrid
nanostructures (Wang et al. 2009, Walters et al. 2010, Haridas et al. 2011). Different views on the
mechanisms of these processes, such as resonant coupling of surface plasmons with excitons, local
field enhancements in metal nanoparticles, and electron transfers from semiconductors to metal
nanoparticles, will also be discussed. Understanding and control of such interactions in weak and
130 Semiconductor Nanocrystals and Metal Nanoparticles

strong coupling regimes among different metal–semiconductor hybrid nanomaterials are impera-
tive. The field has attracted much interest among the researchers for fundamental understanding of
the phenomena as well as for their potential applications, though ambiguities exist among the differ-
ent results and published reports (Hosoki et al. 2008, Zhou et al. 2011, Nahm et al. 2013).

4.2 NANO-BASICS
Compared to that of bulk materials, the science of nanomaterials differs considerably. Accordingly,
the properties of materials vary at the nanoscale. For example, the optical, vibrational, chemical,
and electrical behaviors of nanomaterials are significantly different. This is due to the following
reasons: (1) for the same quantity of material, nanomaterials possess a larger surface area in com-
parison to their bulk counterparts because of the large surface-to-volume ratio of the nanoscale
geometry (consequently, nanomaterials are more chemically reactive); (2) due to absence of long-
range translational invariance compared to the bulk crystals, nanoparticles and nanostructures may
display noncrystalline structures; and (3) effects of quantum confinement (i.e., a reduction in the
degrees of freedom of free electrons or charge carriers in metal nanoparticles or quantum dots,
respectively) are predominant at the nanoscale dimensions of materials. These variations arise
entirely from the imposed size restrictions. The size limit renders nanomaterials attractive for
fundamental understanding and for technological applications.

4.2.1 Particle Sizes versus Physical Properties

Physical properties mentioned above depend largely on the size of the nanoscale material and the
size-dependent material parameters. Standard parameters such as the particle radius (R) or diameter
(d) are used to compare the sizes of the nanoscale particles of various length scales. For exam-
ple, while studying the characteristic changes of optical properties with particle growth, the very
first length scale to be compared is the mean free path (average distance traveled by the electrons
between two successive collisions) of electrons in the metal. In bulk noble metals (silver or gold),
the mean free path (lf) of electrons is ~30 nm at room temperature (Kittel 1985). We shall classify
the nanoparticles as small when d  l f . The second characteristic length scale is the wavelength
of light (λ); this scale becomes very important while exploring optical spectroscopy of relatively
large metal nanoparticles in the UV–visible to near-infrared wavelength range. Accordingly, in this
size regime, where l f  d  l, the particles are called large nanoparticles. However, for complete-
ness, the finest length scale also needs to be mentioned: that is, the Fermi wavelength (de Broglie’s
wavelength of an electron at Fermi energy ~0.5 nm for silver or gold, for instance) of an electron.
In this finest size range, particles behave as molecular types (Link et al. 2002, Zheng et al. 2004).
The physical properties of such metal species are significantly different from those of the other two
types of metal nanoparticles (Peyser et al. 2001, Thomas et al. 2002). Going by the above descrip-
tions, two different kinds of particle-size effects are defined: intrinsic and extrinsic. Intrinsic effects
are concerned with specific changes in volume and surface properties of nanomaterials. Extrinsic
effects are size-dependent responses to external fields irrespective of their intrinsic variations.
Experiments on intrinsic effects particularly focus on the question of how electronic and structural
particle properties such as chemical reactivity, binding energy, crystallographic structure, melting
temperature, and optical material property (ε (ω, R)) vary as a function of particle size, composi-
tion, and geometry. Examples of intrinsic particle size effects that will be studied in detail are the
collective electronic excitations and vibrational properties. The size-dependent optical response
(ε (ω, R)) of nanoscale metal particles is an intrinsic property, for instance, which is governed only
by the dimension and volume fraction (defined as the ratio of the particle volume to the sample vol-
ume) of the particle materials. For the experimental studies discussed here, two length scales are of
importance while elaborating the intrinsic size effects: the mean free path of electrons (lf) and the
wavelength of light (λ).
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 131

4.2.2 Theoretical Framework
To understand a number of unusual experimental results of nanomaterials (e.g., optical and vibra-
tional spectra), development of some theoretical concepts is necessary. In this section, we out-
line these to understand the fundamentals at the nanoscale and phenomena of nanoscale materials.
Although the responses of materials are different at the nanoscale and in the bulk, in many cases
experimental data measured on bulk samples provide significant clues to explain the behaviors of
nanomaterials. We first discuss the optical response of metal nanoparticles.

4.2.2.1 Optical Response of Metal Nanoparticles

Before going on to detailed discussions on the optical response of nanoscale metal particles, it is
imperative to revisit the experimental results on respective metals. For example, a “good” metal like
silver (Ag) is chosen here for the discussions on collective excitations. Among the three metals Cu,
Ag, and Au, which display SPRs in the visible wavelength of light, Ag exhibits the highest efficiency
of plasmon excitation (Kreibig and Vollmer 1995). Also, light–matter interaction cross-sections for
Ag can be ~10 times the geometric cross-section, implying that nanoparticles may capture much
more light than is physically incident on them (Evanoff and Chumanov 2004). During optical pro-
cesses, intraband (which includes Drude or free-electron absorptions) and interband transitions aris-
ing from collective excitations of an electron gas are excited in the nanoscale metal particles. In the
optical wavelength ranges, the dielectric functions ε(ω,0) ~ ε(ω) describe the collective oscillations
or responses of free electrons against the positive ion core in metals, driven by the applied electric
field (schematically shown in Figure 4.1).

4.2.2.1.1 Mie Theory

Spherical and well-isolated metal nanoparticles of uniform size in a dielectric matrix form the basis
of the simplest model system. It facilitates understanding the major optical phenomena in practical
cases. It is ideal to express the optical properties in terms of the absorption (σabs) and scattering (σsca)
cross-sections. These quantities are related to the intensity loss ΔI(z) of a parallel beam of incident
light due to absorption (generation of heat) or elastic scattering (changes of propagation direction).
According to the Lambert–Beer law

DI abs ( z ) = I o (1 - e - N a sabs z )

DI sca ( z ) = I o (1 - e - N a ssca z )

for a purely absorbing and purely scattering system of nanoparticles (Na being the number density
of the nanoparticles), respectively. In practice, the resulting extinction cross-section is given by,
sext = sabs + ssca . The extinction, absorption, and scattering cross-sections are calculated from the

Light Light
Field, E Field, E Metal
particle

FIGURE 4.1 Schematic diagram illustrating the excitation of a dipolar surface plasmon by the electric field
of the incident light at some instant of time t and after time t + T/2 for a spherical metal particle in the quasi-
static size limit. T is the time period of the oscillating field.
132 Semiconductor Nanocrystals and Metal Nanoparticles

Mie theory (following the common notations, see Kreibig and Vollmer 1995) by series expansion
of the involved fields as

å (2L + 1) Re[a
2p
sext = 2 L + bL ] (4.1)
k L =1

å (2L + 1) (|a | +|b | )

2p
ssca = L
2
L
2
(4.2)
| k |2 L =1

and the resulting absorption cross-section is given as, sabs = sext - ssca . Here, k is the wave vec-
tor (k = (2p /l) = (w/c)) and the summation index L gives the order of multipole excitations in the
nanoparticle materials. For example, L = 1, 2, 3,... corresponds to the dipole, quadrupole, octupole
fields, and so on, for higher-order optical excitations. The amplitudes aL and bL in Equations 4.1
and 4.2 are proportional to (| k | R)2 L +1. The total Mie extinction spectra consist of dipolar, quad-
rupolar, and higher modes of excitations. Each multipole is contributed by electric and magnetic
modes, that is, plasmons and eddy currents, respectively. Each mode consists of absorption and
scattering losses. For the lowest order term (L = 1), the simplified Mie formula for the dipolar
absorption is given by

w e2 (w)
sext (w) = 9 c e3m/ 2Vo (4.3)
[e1 (w) + 2e m ]2 + e2 (w)2

where
ε1(ω) and ε2(ω) are the frequency-dependent real and imaginary dielectric functions of the metal
nanoparticles
εm is the dielectric constant of the embedding medium
Vo = (4 / 3)pR 3 is the particle volume
c is the velocity of light in vacuum

The Mie formula (Equation 4.3) shows that the extinction is directly proportional to the volume
of the nanoparticles present. However, it indirectly provides the frequency of resonance through
the condition |e + 2e m | to be minimum. To know the exact resonance position, one needs to have
knowledge about the frequency dependence of ε1(ω) and ε2(ω). We shall discuss this shortly
with an example. However, it is important to know that in the quasi-static size regime (where
2R ≪ λ), the dipolar scattering cross-section which is proportional to (|k|R)6 /|k|2 and higher mul-
tipolar contributions (e.g., the quadrupolar extinction (|k|R)5 /|k|2 and the quadrupolar scattering
(|k|R)10 /|k|2 ) are strongly suppressed due to the size restrictions. Thus, in the quasi-static size
range, sext » sabs.

4.2.2.1.2 Drude–Lorentz–Sommerfeld (DLS) Model

According to free-electron theory of metals, most of the electronic and optical properties in “good”
metals (particularly, alkali and noble metals) are due to the free electrons in conduction band alone.
These metals have completely filled valence bands and partially filled conduction bands (e.g., the
electronic configuration of Ag is [Kr]364d105s1). Linear response of the metal to the interacting elec-
tromagnetic waves is described by the dielectric function ε(ω) of the free valence electron gas of
metals. For alkali metals, the role of ε(ω) is primarily governed by the transitions within the conduc-
tion band alone. In noble metals, substantial contribution of interband transitions from lower lying
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 133

energy levels into the conduction band or from the conduction band into higher energy unoccupied
levels is quite possible. As a result, the role played by the dielectric function ε(ω) of the electron gas
in noble metals in controlling the optical responses is rather complex.
This model of free-electron optical responses assumes that the response of a metal can be
found out by first considering the influence of external forces on one free conduction electron
alone. The macroscopic effective response to the external electromagnetic field can be found by
multiplying the effect on a single electron by the number of total electrons in the given system.
Thus, according to this model, the response of a free electron of mass (m) and charge (e) to an
external electric field (E) is described by the equation of drift motion superimposed on the motion
of an electron in the field-free case. Therefore, the kinematics of the electron in such a case can
be expressed as

d 2r dr
m 2
+ mG = eEoe - iwt (4.4)
dt dt

where Γ denotes the phenomenological damping constant. Equation 4.4 is solved to calculate the
induced dipole moment p (p = ero) and the polarization density P (P = np), (n is the number of elec-
trons per unit volume in the system).
Now, let us define e(w) = 1 + ( P /eo E ) and P = naE , where εo is the permittivity of free space
and α is the polarizability tensor. Assuming a solution for r as r = roe -iwt, the above equation
(Equation 4.4) may be solved. Using the definitions of ε(ω) and P (as above), the dielectric func-
tion ε(ω) of the free-electron gas may be obtained, which can be written as

w2P w2P G
e(w) = 1 - +i (4.5)
w +G
2 2
w(w2 + G 2 )

Expressing e(w) = e1 (w) + ie2 (w), the frequency-dependent real and imaginary parts of the dielec-
tric function may be written as

w2P w2P G
e1 (w) = 1 - and e2 (w) = (4.6)
w + G2
2
w(w2 + G 2 )

where wP = ne2 /eo m is the Drude frequency of plasma oscillation in metals (Kittel 1985). The
damping constant Γ (also known as scattering frequency of electrons) is related to the mean free
path (lf) of the conduction electrons by G = VF /l f , where VF is the Fermi velocity of electrons in
metals. In real metals, VF ~ 106 m s−1 and Γ ~ 1013 s−1. Now, in the UV–visible wavelength range of
light, for which w  G, the real and imaginary parts of the dielectric function of the free-electron
gas of metals can be written in simplified form as e1 (w) » 1 - (w2P /w2 ) and e2 (w) » w2P G/w3 (from
Equation 4.6). Here, ε1(ω) and ε2(ω) of the dielectric function describe the polarization and energy
dissipation of light in matter, respectively.

4.2.2.1.3 Quasi-Static Responses

The laws of electrostatics facilitates the understanding of the optical response of metal nanopar-
ticles in the size range 2R  l (λ being the wavelength of light). In this size regime, the field applied
is assumed to be nearly static. It is known as the quasi-static response. The positive charges in the
metal nanoparticles are assumed to be immobile, and the negative charges, that is, the conduction
electrons, are allowed to move under the influence of an external electric field. A schematic diagram
illustrates the collective oscillations of the conduction electrons against the positive ionic core in
134 Semiconductor Nanocrystals and Metal Nanoparticles

a metal nanoparticle (see Figure 4.1). Using the electrostatic boundary conditions at the particle–
dielectric interface, one can calculate the internal electric field as

3e m
Ei = Eo (4.7)
e + 2e m

where
εm is the dielectric constant of the embedding medium
Eo is the applied field
ε is the dielectric function of the free electrons in the metal

In the quasi-static size regime, solution of electrostatics applies as well to small metal nanoparticles
in oscillating electromagnetic fields. Because of their small size, the nanoparticles experience a field
that is spatially uniform. Under this special condition of size restriction, the internal electric field
(Ei) shows resonance whenever |e + 2e m | is minimum (from Equation 4.7); to say mathematically,
(e1 + 2e m )2 + e22 » minimum. (It may be noted that an identical condition for resonance was used
while discussing the Mie theory for the dipolar absorption [see Equation 4.3] of light.) This further
shows that a negative ε1 is necessary (e1 » -2e m) for the plasmon resonance condition to be satisfied.
Using Equation 4.6 for ε1, we may further write (neglecting Г2 in Equation 4.6)

w2P
e1 » 1 - = -2e m (4.8)
w2PR

wP
Þ wPR = (4.9)
1 + 2e m

where ωPR is the plasmon resonance frequency. For metal nanoparticles, it is also equivalently
known as the SPR frequency. The word “surface” implies that polarization of charges is mostly
taking place on the surface of metal nanoparticles. Also, one can see from the resonance frequency
(Equation 4.9) that the optical response of surface plasmons in metal nanoparticles is very sensitive
to the local environment through the dielectric constant of the medium, εm. Therefore, a change in
the value of εm alters the position of ε1 at which the resonance occurs (see Equation 4.8).
Let us examine a practical case of knowing the localized SPR frequency of spherical metal
nanoparticles in a dielectric medium, say finding the SPR frequency if silver nanoparticles are
embedded in a silica glass matrix. As we know, the refractive index (μ) of silica glass in the visible
wavelength of light is 1.5, and the dielectric constant εm is 2.25 (as e m = m2). So, the resonance con-
dition (i.e., e1 = -2e m ) would be satisfied only when e1 = -4.5. For silver, this occurs (as shown in
Figure 4.2a) at ~3 eV (415 nm). Experimental result for silver nanoparticles in a silica glass matrix is
shown in Figure 4.2b. Optical response peak at 3 eV is clearly observed. Similarly, for gold nanopar-
ticles embedded in a silica glass matrix, the SPR occurs at ~520 nm.
In addition, the size of the metal nanoparticles may also be found out from the full-width at
half-maximum (FWHM) of the SPR absorption band. Assuming Drude-like free-particle behavior
for the conduction electrons in the nanoscale particles within the quasi-static size range, one may
write R = VF t, where R is the average radius of the metal nanoparticles, VF is the Fermi velocity
of electrons in the metal, and τ is the time between two successive collisions (i.e., mean free time)
of conduction electrons in the metal nanoparticles. Quantum confinement and frequent scatter-
ing of these electrons from the nanoscale particles in the dielectric matrix can lead to quantum
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 135

2 ε2

ε1
0

–2

–4
ε1 (Drude’s free electron model)

–6

–8

–10

–12
2 3 4 5 6 7 8 9 10
(a) Energy (eV)

Ag nanoparticles
in a silica glass matrix
Optical density (arb. units)

2 3 4
(b) Energy (eV)

FIGURE 4.2 (a) Dielectric functions ε1(ђω), ε2(ђω) (real, imaginary parts) and the calculated free-electron
contribution e1free (w) (Drude) for bulk Ag. Optical data were collected from the handbook. (From Palik, E.D.
ed., Handbook of Optical Constants of Solids, Academic Press, Orlando, FL, 1985.) (b) Optical absorption
spectrum of Ag nanoparticles (average size ~2.4 nm) in a silica glass matrix revealing the SPR absorption peak
at ~3 eV. (Reprinted from Gangopadhyay, P. et al., J. Appl. Phys., 88, 4975, 2000. With permission.)
136 Semiconductor Nanocrystals and Metal Nanoparticles

fluctuations (ΔE) of the average energy of the free electrons around the SPR energy. Next, applying
the Heisenberg’s principle of uncertainty relation (DEDt = ) and using the expression for R = VF t,
we may further write

VF
R= (4.10)
DE

where  is the reduced Planck’s constant. This equation may be used to calculate the dimen-
sions of metallic nanoparticles as long as the size (d) of the nanoparticles is much less than the
mean free path (lf) of electrons in the metal and the nanoparticles are of uniform size. In the
dilute limit of noninteracting nanoparticles and under the quasi-static size-limiting conditions,
size of nanoscale metal particles may be determined from the spectral width of the SPR absorp-
tion band. However, simple explanations like this may become increasingly inapplicable in
some cases, for example, if the number density of the metal nanoparticles is beyond the dilute
limit. Interactions among the nanoscale metal particles are important in such cases. In addi-
tion, simultaneous influence of other parameters (e.g., size of the nanoparticles) may become
important while elaborating the optical response behaviors of an assembly of nanoscale metal
particles embedded in a dielectric matrix. We shall come back to this discussion later while
elaborating on the experimental results of silver nanoparticles (with a high number density) in
a soda glass matrix.

4.2.2.1.4 Interband Transitions and Core Effects

Intraband transitions like the Drude or free-carrier absorption arising from collective excitations
of an electron gas in nanoscale metal particles are excited by optical means. Apart from the free
conduction electrons, other electrons in the core levels also contribute to the optical response of
metal nanoparticles through the dielectric function ε(ω). It is common to express the dielectric
function in terms of the electric susceptibility χ as e(w) = 1 + c DS (w), where χDS denotes the free
electron Drude–Sommerfeld susceptibility. Based on the optical data (Palik 1985), the calculated
dielectric functions ε1(ђω), ε2(ђω) (real and imaginary parts) and the calculated free-electron con-
tribution e1free (w) (Drude) for bulk silver are displayed in Figure 4.2a. The figure shows that below
~4 eV, dielectric functions are dominated by the Drude’s free-electron behavior, but above 4 eV
the interband transitions become dominant. Thus, from the data and the graph, we arrive at the
important result that the influence of electrons that undergo interband transitions gives an additive
complex contribution c IB = c1IB + ic2IB to the Drude–Sommerfeld (DS) susceptibility. The complex
dielectric function ε(ω), incorporating all optical material properties around the visible region, is
thus given as e(w) = 1 + c DS (w) + c IB (w) . The imaginary part c2IB, describing the direct energy dis-
sipation, becomes large only for frequencies where interband transitions occur. The real part c1 is
IB

sometimes replaced by an average frequency independent value resembling the polarization of the
ion core, χcore. An alternative way to define χcore is to separate the total of all interband excitations
into contributions of lower and higher lying electrons; the former is then denoted as χcore.
Energetically deeper lying electrons (e.g., 3d electrons in Ag) and the ion cores contribute to the
dielectric function ε(ω) through this χcore. Noble metal atoms such as Cu, Ag, or Au have completely
filled 3d, 4d, and 5d shells, and just have one electron in the 4s, 5s, and 6s bands, respectively. It has
been shown that there is an intermediate region between the pure Drude and the interband transition
behavior in these noble metals (Hollstein et al. 1977). For example, whereas direct excitation of the
4s electrons (in Cu) to higher levels requires an energy of 4 eV, the threshold for direct excitation
of 3d-band electrons into the conduction bands (4sp) is only about 2 eV. Similarly, the threshold for
interband transition in other noble metals such as Ag and Au is found to be ~3.9 eV (Ehrenreich and
Philipp 1962) and 2.38 eV (Christensen and Seraphin 1971), respectively.
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 137

4.2.2.1.5 Size-Dependent Dielectric Functions

Following the Drude’s formalism of free-electron plasma oscillation in metals, the dielectric func-
tions for nanoscale metallic particles (of radius R) may be calculated (from Equation 4.6) as

æ 1 1 ö
e1 (w) = w2P ç 2 - 2 2 ÷
(4.11)
è w + G ¥ w + G( R ) ø
2

w2P æ G( R) G ö
e2 (w) = ç 2 - 2 ¥ 2 ÷ (4.12)
w è w + G( R ) w + G ¥ ø
2

where
G ¥ = VF /l¥
G( R) = G ¥ + A(VF /R)

In Equations 4.11 and 4.12, Γ∞ refers to the damping constant in bulk metals. Γ(R) assumes sur-
face scattering as the only size-dependent contribution to the overall damping. This simplified
approach may hold approximately since the electron–phonon relaxation is known to be different in
metal nanoparticles. Restriction arises because, first, the phonon spectrum of nanoparticles differs
from the bulk spectrum (Dickey and Paskin 1968), and, second, the external-field-induced surface
charges give rise to strong electron–phonon coupling. In metal nanoparticles, this effect would be
larger than that of bulk surfaces, due to the high surface-to-volume ratio. Electromagnetic field
enhancement effects in surface-enhanced resonance Raman scattering from the plasmonic prop-
erties of metal nanoparticles or nanostructures have been observed (Bachelier and Mlayah 2004,
Yoshida et al. 2010). The field enhancement in nanoscale metal particles becomes an important
point of discussion, and will be discussed further in this chapter with a few illustrative examples.
Thus, the above-mentioned effects change the magnitude of the dielectric functions ε1(ω) and
ε2(ω) of the nanoparticle materials. So, while describing the optical responses of such materials, it
becomes essential to define a size-dependent optical material function ε(ω) = ε(ω, R). Thus, using
Equations 4.11 and 4.12, defined above, the size-dependent dielectric function, ε(ω, R), may be
expressed as

æ 1 1 ö w2P æ G( R) G ö
e(w, R) = eb + w2P ç 2 - 2 ÷+i ç 2 - 2 ¥ 2 ÷ (4.13)
è w + G ¥ w + G( R ) w è w + G( R ) w + G ¥ ø
2 2 2
ø

where εb is the dielectric function of the bulk metal. Calculations of the size-dependent optical
responses (i.e., optical absorbance) for nanoscale silver particles have been performed using the
ε(ω, R) (Equation 4.13) and compared with the experimental results (discussed later in this chapter).

4.2.2.1.6 Quantum Surface Plasmon Resonances

According to the quasi-static response of Mie theory for small metal nanoparticles, as discussed,
the optical resonance takes place whenever the condition |e + 2e m | ≈ minimum is satisfied. So, the
optical response of the nanoparticles is strongly dependent on dielectric functions of metal and
the embedding medium. Although the Drude–Sommerfeld model accurately predicts the surface-
plasmon spectra for larger metal nanoparticles (2R > 10 nm), the error increases for nanoparticles
in the quantum-size (~2 nm) regimes. Because of the quantum-confinement effects and increased
surface scattering/collisions of electrons with phonons or other electrons or with defects, motions of
free electrons are affected. Thus, following the Mathiessen rule, a term that is inversely proportional
138 Semiconductor Nanocrystals and Metal Nanoparticles

to the size of a metal particle has been added to the surface scattering or damping constant of
bulk metals. For example, mathematically, this is expressed as G( R) = G ¥ + A(VF /R), where A is
an empirical dimensional parameter usually close to 1, which used to account for factors affecting
the width of the SPR in specific cases. The 1/R term in the expression for Γ(R) actually reflects the
ratio of the surface scattering probability (which is proportional to the surface area, ~πR2) and the
number of electrons (being proportional to the volume of the particle, ~πR3).
To understand the optical responses of quantum-size metal nanoparticles, different models have
been proposed. For example, based on the quasi-static optical polarizability of a sphere embedded
in a homogeneous medium, Raza et al. proposed a nonlocal hydrodynamic model and a general-
ized local model incorporating the inhomogeneity of the electron density induced by the quantum
wave nature of the electrons (Raza et al. 2013). However, here we discuss more about the well-
established analytic model (Thomas–Reiche–Kuhn energy-weighted f-sum rule) that is commonly
used for absorption processes. In very fine nanoparticles, due to quantum-confinement effects,
the energy levels in the conduction band are discrete and only specific electronic transitions are
allowed. Thus, considering quantum mechanical effects, Scholl et al. started with the modified
Drude model to account for these transitions (Scholl et al. 2012). They modeled the conduction
electrons as a free-electron gas constrained by infinite potential barriers at the physical boundar-
ies of the particle. The transition frequencies ωif correspond to the allowed quantum energies of
transitions of conduction electrons from occupied states i within the k-space Fermi sphere to unoc-
cupied states f immediately outside it. Here, the technique is employed in the form of oscillator
strengths (Sif), which correspond to each transition frequency. The oscillator strength terms, as
dictated by the f-sum rule, are described using the standard harmonic oscillator quantum mechani-
cal definition

2mwif 2
Sif = á f |z|iñ (4.14)
N

where N is the number of conduction electrons in the nanoparticle. The matrix element term is
determined by the allowed wave functions of the spherical-well model. In total, the particle’s dielec-
tric function can be expressed as

åå w
Sif
e(w) = e IB + w2P (4.15)
i f
2
if - w2 - iGw

where the sum is taken over all initial and final states of the electrons. The frequency of the transition
from the occupied to excited states can be described as wif = E f - Ei . The energy levels Ef and Ei
depend on the geometry and potential of the system. Treating the conduction electrons as particles in
an infinite spherical well, the energy eigenvalue may be calculated as E = ( 2 p2 /8mR 2 )(2n + l + 2)2 ,
where m is the mass of the electron, and n and l are the principal and azimuthal quantum numbers,
respectively. Figure 4.3a shows the real and imaginary parts of the dielectric function, accounting
for both the quantum-confinement effects and the surface scattering of free electrons. Here, the
damping term was allowed to vary in accordance with the expression G = G ¥ + A(VF /R). The calcu-
lation assumed the value of A to be 0.25. Using the dielectric function values (shown in Figure 4.3a)
as input to the electromagnetic Mie theory, the absorption efficiencies (absorption cross-sections
divided by physical cross-sections) of silver nanoparticles with diameters 2, 4, 6, and 8 nm were
calculated. Figure 4.3b shows the results, revealing decreasing absorption efficiency and blue-
shifting of the SPR peaks as the silver particle size decreases. Afterward, while discussing the
optical responses of fine silver nanoparticles in a silica glass matrix, we will compare the observed
experimental results with the theoretical calculations and comment on it.
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 139

×1000

8 nm
1 6 nm
4 nm
2 nm
0
ε1

–1

2
ε2

0
0.5 1 1.5 2 2.5 3 3.5 4
(a) Energy (eV)

8 nm
6 nm
4 nm
3
2 nm
Absorption efficiency

0
2.8 3 3.2 3.4 3.6 3.8 4 4.2
(b) Energy (eV)

FIGURE 4.3 Analytic quantum theory of particle dielectric functions and optical spectra. (a) Real (ε1) and
imaginary (ε2) components of the dielectric functions of Ag nanoparticles with diameters of 2, 4, 6, and 8 nm,
calculated using the analytic quantum model (Equation 4.15). (b) Corresponding absorption spectra of the
particles generated by Mie theory. The refractive index of the surrounding medium is set to 1.3. As the particle
size decreases, absorption efficiencies reduce and surface-plasmon peak energies get blue-shifted. (Reprinted
by permission from Macmillan Publishers Ltd. Nature, Scholl, J.A. et al., Quantum plasmon resonances of
individual metallic nanoparticles, 483, 421, © 2012.)
140 Semiconductor Nanocrystals and Metal Nanoparticles

4.2.2.2 Vibrational Spectroscopy of Nanoparticles

Solids in the form of tiny particles or consisting of nanoscale materials have particular interest-
ing properties originating from confinements of lattice vibrations (phonons) and surface effects.
A considerable amount of research (theory and experiments) has been reported during the past
years regarding the surface acoustic confinements of phonons in nanomaterials (Bottani et al. 2001,
Saviot et al. 2011). Although other vibrational spectroscopic studies, for example, infrared absorp-
tion (Murray et al. 2006) or neutron scattering (Saviot et al. 2008), have been carried out with a
limited success, inelastic low-frequency Raman scattering (LFRS) measurements for such nano-
materials embedded in different amorphous matrices have provided good understanding about the
confinement of acoustic phonons in the nanomaterials. The calculated sizes of nanoparticles using
this nondestructive experimental technique have shown good agreement with direct and standard
measurement procedures (e.g., transmission electron microscopy). The LFRS technique will be
elaborated in the next section with illustrative examples of nanoscale metal particles embedded in
different dielectric matrices. A theoretical description on the vibration of spherical bodies (arising
out of Lamb’s theory) will be provided here, though briefly. Nanoscale materials are generally con-
sidered as elastic bodies of spherical or ellipsoidal shape. According to Lamb’s theory, two types of
vibrational modes may exist for such elastic bodies: torsional and spheroidal (Lamb 1882). For the
low-energy modes, torsional motion of a spherical surface can be visualized as two different hemi-
spheres rotating in opposite directions (as shown in Figure 4.4). The torsional modes correspond
to movement without any volume change. Torsional displacements induce shear. On the contrary,
spheroidal oscillations (e.g., breathing modes, shown in Figure 4.5) are with dilation (i.e., with
volume change). According to group-theoretical analysis, low-frequency Raman scattering from
spherical nanoparticles are possible in the quasi-static size limit. Further, the theory also confirms
that the observable Raman transitions are possible in this class of materials for radial (l = 0) as well
as quadrupolar (l = 2) modes (Duval 1992). The characteristic vibrational frequencies associated
with the spheroidal modes in nanoscale particles are in the range of few cm−1 to a few tens of cm−1.
In the vibrational spectra of nanomaterials, low-frequency Raman modes are usually observed in
this range of frequencies (Duval et al. 1986).

FIGURE 4.4 Relative displacements of low-energy torsional modes in a spherical body. Torsional motion
of a spherical surface is visualized as two different hemispheres rotating in opposite directions without any
volume change. (Reprinted from Duval, E., Phys. Rev. B, 46, R5795, 1992. With permission.)
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 141

(a) (b)

FIGURE 4.5 Different types of spheroidal oscillation modes confined to the surface of nanoscale particles.
(a) Radial mode. (b) Quadrupolar mode. The spheroidal oscillations correspond to movement with volume
change of materials.

4.2.2.2.1 Calculations of Vibrational Frequencies

Lamb (1882) and Tamura et al. (1982) have calculated the vibrational frequencies (Ω) of a homo-
geneous, elastic spherical particle under no-stress condition. The particle is free to vibrate. These
calculations provide the size dependence of the peak frequency of vibrational eigen frequencies for
an elastic spherical body. Following the notations used by Tamura et al., the eigenvalue equation can
be derived for the spheroidal modes (l ≥ 0) as

é h jl +1 (h) ù zj (z ) h4
2 êh2 + (l - 1)(l + 2) - (l + 1) ú l +1 - + (l - 1)(2l + 1)h2
ë jl (h) û jl (z ) 2

h jl +1 (h)
+ [h2 - 2l (l - 1)(l + 2)] =0 (4.16)
jl (h)

where
h = WR /Vt and z = WR /Vl are the nondimensional eigen frequencies
l is an angular quantum number
jl (h) is the spherical Bessel function of the first kind

The eigenvalue equation (Equation 4.16) is solved by setting the material parameter Vt /V l, where
V l and Vt are the longitudinal and transverse sound velocities, respectively. The nondimensionalized
eigen frequencies (hsln and z sln ) of the spheroidal modes strongly depend on the material through
the ratio Vt/Vl. For any set of l and n, it may be noted here that the eigen frequencies hsln and z sln
are not independent but are related to each other by the relation h/z = Vl /Vt . Only the lowest vibra-
tion modes (n = 0) are the surface modes, whereas the higher modes (n ≥ 1) are the inner modes
of a vibrating elastic spherical particle. Here, hsln are the (n + 1)th eigenvalues corresponding to
the angular momentum l. Group-theoretical analysis predicts that spheroidal modes with l = 0,
2 alone are Raman active. These surface acoustic vibrations are known as spheroidal quadrupolar
(l = 2) modes and spheroidal breathing or radial (l = 0) modes. In agreement with the Raman selec-
tion rules, quadrupolar (l = 2) modes are to be observed when the polarizations of the incident and
Raman scattered light are crossed. For breathing or radial (l = 0) modes, Raman intensity vanishes
in cross-polarized conditions.
142 Semiconductor Nanocrystals and Metal Nanoparticles

The eigen frequencies for surface modes (n = 0) for various metals (with suitable values of Vt /Vl)
have been calculated numerically (using MATLAB®) by employing Equation 4.16. Knowing the
sound velocity ratio (for example, Vl /Vt = 2.19 and 1.89 for silver and cobalt metals, respectively,
Simmons and Wang 1971), the vibrational eigen frequencies (hsln and z sln ) of these metal nanopar-
ticles have been calculated. The size of nanoparticles can be calculated from these eigen frequency
equations:

WR WR
h= , z= (4.17)
Vt Vl

The experimentally measured Raman frequencies have been used in Equation 4.17 to determine
average sizes of nanoparticles of interest. This method has been used to study coarsening or thermal
growth of nanoparticles in different dielectric matrices. Let us discuss the experimental Raman
results with a few examples in this context.

4.2.2.2.2 Illustrative LFRS Studies

4.2.2.2.2.1  Silver Nanoparticles For example, dissolution of pristine Ag nanoparticles in a
silica glass matrix due to the Si ion irradiations (see Srivastava et al. 2014 for details) has been stud-
ied by the inelastic LFRS measurements. Vibrational spectra of nanomaterials appear in the low-
frequency region (from few cm−1 to few tens of cm−1). The recorded low-frequency Raman spectra
for Ag nanoparticles in a silica glass matrix are displayed in Figure 4.6. In this study, the surface
acoustic vibrations corresponding to quadrupolar (l = 2) spheroidal modes have been observed.
Te eigen frequency (h2s ) corresponding to the quadrupolar mode is calculated from Equation 4.16,
which works out to be ~2.65. Te vibrational mode frequency has been obtained after substitut-
ing the value for η in Equation 4.17. Experimentally, this vibration frequency is measured from
Intensity (counts)

(b)

(a)

10 15 20 25 30 35 40 45
Raman shift (cm–1)

FIGURE 4.6 Polarized low-frequency Raman spectra of Ag nanoparticles of different sizes in the SiO2
matrix. (a) Average size of nanoparticles 7.7 nm. (b) Average size 2.7 nm. Solid lines are the fittings through the
data points (shown as symbols). Compared to spectra (a), shifts of Raman modes to higher frequencies (spectra
(b)) verify the decrease in size of Ag nanoparticles. The estimated sizes agree with those from direct HRTEM
measurements. (Reprinted from Srivastava, S.K. et al., Chem. Phys. Lett., 607, 100, 2014. With permission.)
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 143

low-frequency Raman scattering measurements. The Raman peak frequency (n 2s , in cm−1) of the
spheroidal mode corresponding to the quadrupolar vibration mode may thus be expressed as

Vt
n 2s = 0.85 (4.18)
dc

where
d is the diameter of silver nanoparticles
Vt is the transverse sound velocity (1660 m s−1) in bulk silver
c is the velocity of light in vacuum

By measuring the Raman peak frequencies, the sizes of the silver nanoparticles have been estimated
using Equation 4.18. In this case, the experimental data have been fitted to a Lorentzian line-shape
function with an exponential Rayleigh background to estimate the Raman peak frequencies. The
directly measured values (from the high-resolution transmission electron microscopy) agree well
with the results obtained from the LFRS measurements.
Equation 4.18 further demonstrates that the vibration frequency is inversely proportional to
the diameter of the nanoparticles. As a result, with the growth of the nanoparticles, the Raman
mode frequency would further shift to lower frequencies. This condition, in particular, makes
it difficult to measure sizes of larger nanoparticles through the LFRS technique. For example,
typically, nanoparticles of sizes <10 nm may be measured. For larger nanoparticles, the Raman
frequency shift usually merges with the huge background of Rayleigh scattering (elastic), which
makes LFRS measurements, and thus obtaining the vibrational spectra of larger nanomaterials,
difficult.

4.2.2.2.2.2  Cobalt Nanoparticles LFRS studies have been carried out on cobalt nanoparticles
in a silica glass matrix. For more details, see Gangopadhyay et al. (2007). Apart from finding the
size of cobalt nanoparticles, the study also focused on elucidating the coupling of electrons and
phonons in nanoscale metals. Raman spectra were recorded using different excitation wavelengths
of the argon ion laser. As shown in Figure 4.7, intensity of the low-frequency vibrational modes
depends significantly on the wavelength of laser. The strong local field due to the SPR in a cobalt
nanoparticle with 351.1 nm excitation is expected to be the reason for the observed enhancement of
Raman scattering intensity. As the characteristic SPR absorption for cobalt nanoparticles in a silica
matrix is at ~350 nm (see Figure 4.8), the increase in the intensity of vibrational mode at 26.5 cm−1
with the excitation wavelength in the vicinity of SPR is thought to be due to an enhanced electric
field in the nanoparticles of cobalt atoms. The results thus obtained from the low-frequency Raman
scattering spectroscopy experiments emphasize the importance of surface plasmon–phonon cou-
pling in metallic nanoparticles.
As shown previously, strong light scattering is achieved with excitation close to the SPR wave-
length. With the resonance wavelength, mainly spherical nanoparticles are excited. The dependence
of Raman scattering on the incident and scattered photon polarizations can be used to characterize
the observed mode according to the Raman selection rule. Since the scattering due to pure radial
vibrations is totally polarized, the Raman mode observed at 26.5 cm−1 should disappear in depo-
larized spectra. To corroborate this, cross-polarized Raman measurements have been carried out,
and the polarization dependence of the Raman spectra is displayed in Figure 4.9. The Raman mode
observed in the polarized spectra has disappeared in the depolarized configuration. Measurements
thus confirm the mode of vibration to be of radial (l = 0) type in accordance with the vibrational
selection rule (Duval 1992). Assuming a homogeneous elastic sphere with a free surface, the nondi-
mensionalized eigen frequency corresponding to the radial mode has been worked out to be ~2.46.
The assumption of free vibration is justified because sound velocity and mass density values are
144 Semiconductor Nanocrystals and Metal Nanoparticles

Excitation lines:
351.1 nm 457.9 nm
488.8 nm 514.5 nm

Intensity (arb. units)

10 20 30 40
Raman shift (cm–1)

FIGURE 4.7 Enhancement of intensity of the surface-acoustic vibrational modes confined to the Co nano­
particles in the SiO2 matrix as the excitation wavelength from the argon-ion laser source is varied. Spectra are
offset in the intensity axis for a better clarity. (Reprinted from Gangopadhyay, P. et al., Appl. Phys. Lett., 90,
063108, 2007. With permission.)

very much different in cobalt compared to those in the silica glass matrix. Thus, the Raman peak
frequency (n os ) corresponding to the radial surface modes may be expressed as

Vl
n os = 0.78 (4.19)
dc

where
d is the diameter of the cobalt nanoparticles
Vl is the longitudinal sound velocity (5856 m s−1) in bulk cobalt metal
c is the velocity of light in vacuum

From the measured Raman mode frequencies, sizes of the nanoscale cobalt particles in the silica
glass matrix have been estimated using Equation 4.19.
Although confined surface-acoustic phonons in nanoscale metal particles can give rise to
radial (symmetric) and quadrupolar Raman modes, only quadrupolar modes are usually observed
in low-frequency Raman experiments. Quadrupolar modes are easily detected in Raman scatter-
ing experiments. This is probably because the radial modes (l = 0), for which size changes pre-
serve the shape, produce weak modulations of the electric dipole (Gersten et al. 1980). However,
in addition to the quadrupolar mode, radial modes have also been observed by other groups of
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 145

0.56
Co nanoparticles
in a silica glass

Optical density (arb. units)

0.54

(b)

(a)
0.52
300 350 400
Wavelength (nm)

FIGURE 4.8 Optical absorption spectra of Co ion–implanted (fluence 2 × 1016 ions cm−2) silica glass
­samples. (a) Before and (b) after annealing in high vacuum (for 573 K, 90 min). Absorbance has increased
(spectra b) due to thermal growth of the Co nanoparticles in the matrix. (Reprinted with permission from
Gangopadhyay, P., AIP Conference Proceedings, Bikaner, Rajasthan, India, Vol. 1536, p. 7. © 2013, American
Institute of Physics.)

λex: 351.1 nm
Polarized
Depolarized
Intensity (arb. units)

20 30 40
Raman shift (cm–1)

FIGURE 4.9 Polarization dependence of low-frequency Raman spectra of Co nanoparticles embedded in the
silica glass matrix, confirming the vibration mode of the nanoparticles to be of radial type. (Reprinted from
Gangopadhyay, P. et al., Appl. Phys. Lett., 90, 063108, 2007. With permission.)
146 Semiconductor Nanocrystals and Metal Nanoparticles

Extinction
Intensity (arb. units)
350 400 450 500
Wavelength (nm)

–40 –20 0 20 40 60
Raman shift (cm–1)

FIGURE 4.10 Raman spectra measured on Ag particles with 1.4 nm mean radius and excited at 413, 458,
476, 488, and 514 nm (from top to bottom). The corresponding extinction spectrum showing the surface-
plasmon resonance is plotted in the inset. (Reprinted from Bachelier, G. et al., Phys. Rev. B, 76, 235419, 2007.
With permission.)

researchers in their resonant Raman scattering experiments (Portales et al. 2001, Courty et al.
2002). During these measurements, the quadrupolar mode appeared very close to the direct
Rayleigh (elastic scattering) line. Compared to this, radial modes were observed with lower inten-
sities and relatively far away from the Rayleigh line. Bachelier and Mlayah have explained these
experimental observations after taking into account of various coupling mechanisms between
confined acoustic vibrations and surface plasmon–polariton states (Bachelier and Mlayah 2004).
In another experimental study, the maximum Raman scattering intensity was observed from fine
silver nanoparticles while exciting with a line close to the SPR absorption wavelength of the metal
nanoparticles (see Figure 4.10).

4.3 SYNTHESIS OF METAL NANOPARTICLES

Nanomaterials may be prepared in one dimension (e.g., very thin surface coatings), in two dimen-
sions (nanowires and nanotubes), or in three dimensions (quantum dots and nanoparticles). Some
selected methods for the synthesis of metal nanoparticles are enlisted in Table 4.1. Depending on the
processing conditions, preparations are generally classified into bottom-up or top-down approaches.
As the name suggests, in bottom-up methods, the nanomaterials are grown from the state of atoms
or ions of the materials. Growth behaviors differ depending on the processing conditions. We can
observe this while dealing with different methods of preparations. On the contrary, top-down pro-
cessing starts with bulk materials. The top-down techniques (mechanical attrition, lithography, or
ion-beam sputtering) gradually transform the bulk into different forms of nanostructures. We will
elaborate both synthesis approaches later on with a few examples. It may be restated that precise
control of the synthesis of nanomaterials is very important as many of the interesting properties
depend crucially on their size, shape, and uniformity. Controlling the level of impurities is again
vital all through the processing using chemical methods, for example. Necessarily, this requires a
good understanding of the science of the synthesis of nanomaterials.
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 147

TABLE 4.1
Selected Methods for the Synthesis of Metal Nanoparticles
Methods of Synthesis Materials References
Mechanical attrition (e.g., ball milling) Fe nanoparticles Muñoz et al. (2007)
Nanosphere lithography Ag nanostructures on various Haynes and Van Duyne (2001)
substrates
Dip-pen nanolithography Supported Au nanostructures Zhang et al. (2003)
Ion exchange followed by thermal Ag nanoparticles in a soda glass Gangopadhyay et al. (2005), Miotello
annealing, laser irradiation, light ion matrix et al. (2001), and Magudapathy et al.
irradiation (2001)
Ion implantation (direct and recoil) Ag nanoparticles in a SiO2 glass Srivastava et al. (2014), Gangopadhyay
techniques matrix; Au nanoparticles in et al. (2000), and Zhou et al. (2012)
sapphire
Ion-beam sputtering Pd nanostructures on SiO2; Ag Ruffino et al. (2009b) and
and Au nanoparticles on SiO2 Gangopadhyay et al. (2010)
glass samples
Melt-quenching technique Cu nanoparticles in barium Sendova et al. (2015)
phosphate glass
RF co-sputtering Ag nanoparticles in a SiO2 glass Fujii et al. (1991)
matrix
Ion-beam mixing Alloy nanoparticles (Co–Pt) in Balaji et al. (2014)
a-SiO2
Pulsed-laser deposition Au nanoparticles in Al2O3, Co Gonzalo et al. (2005) and Clavero et al.
nanoparticles in ZrO2 and Al2O3 (2007)
Atomic layer deposition Alloy (Ru–Pt) nanoparticles Kim et al. (2009)
Gas-phase preparation Ag–PbS nanocomposites Maisels et al. (2000)
Microemulsion technique Alloy (Co–Pt) nanoparticles Kumbhar et al. (2001)
Spark discharge technique Au–Ge nanocomposites Kala et al. (2013)

4.3.1 Bottom-Up Methods

Bottom-up methods are more prevalent for the preparation of nanomaterials (see Table 4.1).
Precise control on the nucleation and growth of metal nanoparticles through the bottom-up syn-
thesis is of prime importance. Most of the novel optoelectronic properties of nanocomposite
materials depend on physical dimensions and morphologies of these entities (Stamplecoskie and
Scaian 2010). To achieve a good control on the preparation, a basic understanding of mechanisms
governing the formation and evolution of metal nanoparticles while processing these materials
is a must. Hence, attention is drawn first to elucidating the nucleation and growth characteris-
tics of metal nanoparticles through thermal annealing of ion-implanted samples, for example.
In addition, the thermochemical stability of optical materials of this kind is an important and
pertinent issue to tailor the nanomaterials to suit technological demands. Ion implantation and ion
exchange (chemical route) will be discussed thoroughly to elucidate the nucleation and growth of
nanoparticles in different matrices. These methods are taken as illustrative examples of bottom-
up approaches.

4.3.1.1 Ion Implantation

Metal particles nucleate when there is a supersaturation of the metal atoms in a given matrix. The
only prerequisite is that the metal should be insoluble in the host matrix. By ion implantation,
148 Semiconductor Nanocrystals and Metal Nanoparticles

for example, metal atoms are introduced into a dielectric matrix at concentrations well above
the solubility limit. Depending on the energy of metal ions, particles may be formed at different
depths in the given matrix. Being a nonequilibrium process, good combinations of ion projectiles
and target matrices are feasible. The bottom-up technique provides chemically clean nanopar-
ticles with reasonable sizes and depth distributions in a matrix. Many researchers have used ion
implantation/irradiation to control and tailor the size and spatial distributions of metal nanopar-
ticles (Ren et al. 2006, Rizza et al. 2007, Bernas 2010, Ramjauny et al. 2010). Ion implantation
techniques are thus highly suitable to study the nucleation and growth of metal nanoparticles in
insoluble matrices (Mazzoldi and Mattei 2005, Bernas 2009). In fact, ion beams are also used
as a top-down approach to prepare plasmonic nanostructures on a substrate through sputtering
(Gangopadhyay et al. 2010). The basic processes behind the nucleation and growth of nanoparti-
cles formed through the ion implantation are shown schematically in Figure 4.11. Supersaturation
of metal atoms in the glass matrix leads to the nucleation of metal particles in the form of fine pre-
cipitates. According to the theory of precipitate growth (Lifshitz and Slyozov 1961, Wagner 1961),
particles with radii exceeding a critical radius can grow during further implantation of the ions or
during subsequent thermal annealing, and lead to the growth or coarsening of the nanoparticles.
It may take place following three different stages. In the first stage, nuclei of the precipitates are
formed with an average radius equal to the critical value. Homogeneous or heterogeneous nucle-
ation may occur, but in an ion-implanted system, where copious defects are produced, it is likely
that the implantation-produced defects may act as nucleating centers (i.e., heterogeneous nucle-
ation). In the second stage, diffusional growth of the precipitates occurs by solute depletion of the
surrounding matrix. Here the growth is governed by the diffusion coefficient of the precipitating
element, and, more importantly, the average radius of the particle scales as Dt , where D is the
diffusion constant and t is the time of diffusion. In the third stage, the coarsening regime, the
particle distribution established during the nucleation coarsens by the growth of larger nanopar-
ticles. In this process, growth of larger nanoparticles occurs at the expense of smaller ones. It is
a slow process. Here, the average radius of the growing particle scales as ( Dt )1/ 3. This coarsening
stage is known as Ostwald ripening, where the mass of the material is conserved. The nucleation

Ion implantation
Formation of an ensemble
of separated nanoclusters

Precipitates Rp

d
Supersaturation Growth Coalescence
Buried
Ion stopping Nucleation Ostwald ripening layer

Time of ion implantation Time of annealing

FIGURE 4.11 Ion-beam synthesis of nanostructures shown schematically. High fluence of ion implantation
into a solid leads to supersaturation of the impurity atoms. Nanoscale particles nucleate and grow during ion
implantation or during the subsequent thermal annealing. Average size of the nanoparticles as well as their
spatial and size distributions change during the Ostwald ripening. At very high fluences, buried layers can
form by coalescence of the embedded nanoparticles. (Reprinted from Reiss, S. and Heinig, K., Nucl. Instrum.
Methods Phys. Res. B, 102, 256, 1995. With permission.)
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 149

and growth of nanoparticles of the implanted atoms take place through controlled post-annealing
processes. The size of the embedded nanoparticles can be altered by adjusting the annealing
parameters and the implantation fluences (concentration of the implanted atoms). In fact, one may
end up with a continuous buried layer (see Figure 4.11) with very a high fluence of implanted ions
or at a high annealing temperature.
Now let us discuss a few examples. First, we consider the case where Ag atoms are recoil-
implanted in a silica glass matrix. During this study, the top Ag layer (~45 nm thick) of the Ag/SiO2
sample was irradiated at room temperature by argon ions (energy 105 keV, total fluence 1 × 1016
ions cm−2). Because of the energetic impacts of the argon ions, Ag atoms get recoil-implanted in the
silica glass. After the ion irradiation, the unmixed Ag layer is removed from the surface of the silica
substrates by dipping it for 2 min in a hydrochloric acid bath. The silica matrix embedded with Ag
atoms was post-annealed at different temperatures between 333 and 523 K for 1 h at each annealing
step. Optical absorption and Raman scattering spectroscopy carried out on these samples revealed
thermal growth of Ag nanoparticles in the matrix (Gangopadhyay et al. 2000). According to Mie
theory, optical absorption increases with increase in the volume of the nanoparticles (Equation 4.3).
Precisely, this is what has been observed (see Figure 4.12). The Raman scattering results are shown
in Figure 4.13. The systematic shift of the Raman mode positions to lower frequencies with the
increase of post-annealing temperature may be noted. As discussed earlier, vibration frequency
scales inversely with the diameter (Equation 4.18) of nanoparticles. Raman measurements also con-
firm the continual thermal growth of the embedded Ag nanoparticles in the matrix. Thus, from
the low-frequency Raman scattering spectroscopy experiments, the size of Ag nanoparticles was
measured, which was in the range 1–4 nm. The sizes of Ag nanoparticles are seen to increase with
increase of the annealing temperature (see Figure 4.13).
Here, two stages are identified with the nucleation and growth of Ag nanoparticles in the silica
glass matrix. First, because of concentration fluctuations, nucleation of Ag atoms takes place, and

523 K
Optical density (arb. units)

473 K

As implanted

2 3 4
Energy (eV)

FIGURE 4.12 Optical absorption spectra for Ag nanoparticles embedded in a silica glass matrix, prior to
thermal annealing and after isochronal annealing for 1 h at 473 and 523 K. Coarsening of Ag nanoparticles
increases the optical absorption intensity. (Reprinted from Gangopadhyay, P. et al., J. Appl. Phys., 88, 4975,
2000. With permission.)
150 Semiconductor Nanocrystals and Metal Nanoparticles

Post-annealed at
333 K
373 K
423 K
473 K
523 K
Rayleigh background
Lorentzian fitting
Intensity (arb. units)

10 20 30
Raman shift (cm–1)

FIGURE 4.13 Polarized low-frequency Raman spectra of Ag nanoparticles in the SiO2 glass matrix. Post-
annealing temperature is indicated for each spectrum. Symbols are data, and solid lines are the Lorentzian
fitting of data with the Rayleigh background (shown as dashed lines). (Reprinted from Gangopadhyay, P. et al.,
J. Appl. Phys., 88, 4975, 2000. With permission.)

these nuclei grow directly as a result of the supersaturation. In the second stage of the process,
nuclei that have reached a critical size grow according to the coarsening or Oswald ripening mecha-
nism. In the coarsening regime, nucleation of new particles is negligible. According to the theory
of ripening (Lifshitz and Slyozov 1961, Wagner 1961), if the total mass of clustering matter is con-
served, the following growth behavior may be observed at asymptotic times:

4
Rt3 - Ro3 = Dat (4.20)
9
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 151

where
Rt (after annealing for time t) and Ro (prior to annealing) are the average radii of the nanoparticles
D is the diffusion constant of clustering atoms in the matrix
t is the time of annealing
α characterizes the kinetics of precipitation

By applying this growth formula (Equation 4.20), the activation energy for the diffusion of Ag
atoms in the silica glass matrix was determined from these experimental results (Gangopadhyay
et al. 2000).
Moving on to the study of the higher-order modes of optical responses of metal nanoparticles,
next we consider an experimental case where Ag nanoparticles are formed in a plain soda glass
matrix. Implantation of Ag+ ions of 1.0 MeV energy with various fluences (maximum fluence
being 5 × 1016 ions cm−2) was carried out using the Tandetron (1.7 MV, High Voltage Engineering
Europe (HVEE)) accelerator at Materials Science Group (MSG), Indira Gandhi Centre for Atomic
Research (IGCAR), for the synthesis of Ag nanoparticles in a soda glass matrix. Although dipolar
SPR is predominant in the recorded optical absorption spectra (see Figure 4.14), sample with the
maximum fluence of 5 × 1016 Ag+ ions cm−2 shows some interesting features (spectra “d”, Figure
4.14). In this particular sample, Ag nanoparticles of sizes ~100–200 nm have been observed (SEM
image, Figure 4.15). In the sample with the highest fluences of Ag+ ions, defect-enhanced mobil-
ity of Ag atoms in the soda glass led to segregation of Ag atoms close to the glass surface. Higher
mobility of Ag atoms on the glass surface further helps the growth of Ag nanoparticles. Large
metal nanoparticles play a dominant role in modifying their optical properties. As observed here,
compared to other samples, the intensity and width of the SPR absorption at 420 nm are greatly
modified for the sample with the maximum ion fluence. Phase retardation of light and additional

1.2

(c)
0.6
Optical density
Optical density (arb. units)

0.8 0.4

(d) 0.2
300 325 350 375
Wavelength
0.4

(b)

0.0 (a)

300 400 500 600 700 800 900

Wavelength (nm)

FIGURE 4.14 Optical absorption spectra for Ag+ ion–implanted soda glass samples with various fluences:
(a) 5 × 1015, (b) 1 × 1016, (c) 3 × 1016, and (d) 5 × 1016 ions cm−2. Quadrupolar response for larger Ag nanopar-
ticles is shown in the inset. (Reprinted with permission from Gangopadhyay, P., AIP Conference Proceedings,
Bikaner, Rajasthan, India, Vol. 1536, p. 7. © 2013, American Institute of Physics.)
152 Semiconductor Nanocrystals and Metal Nanoparticles

1000 nm

FIGURE 4.15 Scanning electron microscope (SEM) image showing large Ag nanoparticles that appear
close to the surface of soda glass. Various sizes of Ag nanoparticles appear as bright dots in the image.
(Reprinted with permission from Gangopadhyay, P., AIP Conference Proceedings, Bikaner, Rajasthan,
India, Vol. 1536, p. 7. © 2013, American Institute of Physics.)

higher-order optical resonances have become particularly important in this sample. Because of the
nonzero field gradient across large Ag nanoparticles in the glass matrix, phase retardation may
lead to larger widths of the optical resonances. Further, the optical spectra reveal a shoulder on the
low-wavelength side (~340 nm) of the dipolar peak (inset, Figure 4.14) for the sample with larger
Ag nanoparticles. This is assigned to the quadrupolar SPR absorption in these Ag nanoparticles
(Kreibig and Vollmer 1995).
Next we discuss the optical absorption and growth of Au nanoparticles in a silica glass matrix
(De Marchi et al. 2002). Interestingly, the study has attempted to identify the difference between
a pure diffusional and a coarsening growth stage of the precipitation process. From the analysis of
experimental results, two distinct growth regimes have been observed, namely t1/2 and t1/3 (as shown
in Figure 4.16), which correspond to growth due to precipitation of supersaturated solution and
coarsening, respectively. Optical absorption spectra of Au ion–implanted silica samples at different
annealing times in air atmosphere are shown in Figure 4.17. The characteristic dipolar SPR near
530 nm, due to Au nanoparticles in the matrix, is evident. A gradual increase of optical absorption
intensity around the SPR wavelength from the as-implanted to the 12 h-annealed sample has been
observed. This is expected with the thermal growth of the Au nanoparticles (according to Mie
theory, Equation 4.3).

4.3.1.2 Ion-Exchange Processing

Next we describe the experimental observations on the nucleation and growth of Ag nanoparticles,
for instance, in plain soda glass samples. Ion exchange of various metal ions (silver, copper, gold)
in soda glasses followed by light ion irradiation is an established method to modify the linear and
nonlinear optical properties of glasses. The method is advantageous because it is easy to prepare
and commercially viable for applications in optoelectronic devices. Silver ion-exchange processing
(Ag+ ↔ Na+) in the soda glass occurs because the difference in ionic radii of Ag+ and Na+ ions is ~29%.
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 153

40 Exp. 200 Exp.

Linear fit Linear fit

30 150
R2–R02 (nm2)

R3–R03 (nm3)
20 100

10 50

0 0

0 2 4 6 8 10 12 6 8 10 12
(a) Annealing time (h) (b) Annealing time (h)

FIGURE 4.16 (a) [ R 2 (t ) - Ro2 ] and (b) [R 3 (t ) - Ro3 ] evolutions in Au-implanted silica glass samples annealed
in air at 1173 K for different time intervals. Solid lines are the linear fits to the experimental data (filled circles)
from TEM and optical fit results. (Reprinted with permission from De Marchi, G. et al., Two stages in the
kinetics of gold cluster growth in ion-implanted silica during isothermal annealing in oxidizing atmosphere,
J. Appl. Phys., 92, 4249. © 2002, American Institute of Physics.)

7h 12 h
0.5
6h
5h
3h
0.4 2h

1h
Optical density

0.3

0.5 h

0.2

As-implanted

0.1

400 500 600 700

Wavelength (nm)

FIGURE 4.17 Optical absorption spectra of Au-implanted silica glass samples (at fluence 4 × 1016 Au+1 cm –2,
energy 190 keV), annealed in air at 1173 K for different time intervals. (Reprinted with ­p ermission
from De Marchi, G. et al., Two stages in the kinetics of gold cluster growth in ion-implanted silica during
isothermal annealing in oxidizing atmosphere, J. Appl. Phys., 92, 4249. © 2002, American Institute of
Physics.)
154 Semiconductor Nanocrystals and Metal Nanoparticles

Composition of the glass is Si 21.49%, Na 7.1%, Ca 5.78%, Mg 0.34%, and Al 0.15% by weight.
Silver-exchanged glass samples are prepared by immersing preheated soda glass slides for 2 min in
a molten-salt bath of AgNO3 and NaNO3 (1:4 weight ratio) mixture at 593 K.
Subsequent evolution of Ag nanoparticles in the ion-exchanged soda glass samples occur under
different processing conditions: (1) irradiation with light ions (Arnold et al. 1996), (2) thermal
annealing (Wang 1996), (3) laser irradiation (Miotello et al. 2001), and (4) gamma irradiation (Farah
et al. 2014). In this chapter, most of our discussions would be based on the first two processes.
Let us first discuss the case of light-ion irradiation, where 100 keV He+ ions bombard the silver-
exchanged soda glass samples with a low beam current density (~1 µA cm−2) to minimize the ion
beam’s heating effect. The ion-irradiated samples exhibit an optical absorption peak at ~415 nm
(Figure 4.18a). This characteristic absorption is due to the SPR in silver nanoparticles in the glass
matrix. In this case, defects due to the ion irradiation are expected to mediate the diffusion of
silver ions in the ion-exchanged soda glass samples and control the formation and growth of sil-
ver nanoparticles. The effect due to the ion-beam heating might be less significant. Here, optical
absorption results are primarily looked into, as they show interesting features. For example, optical
absorbance increases and, more importantly, the absorbance band broadens considerably with the
increase of helium ion fluence (Figure 4.18). The resonance peak at 415 nm apparently vanishes and
the spectra appear flat over a range of optical wavelengths with higher fluences. It may imply that
SPR occurs over a broad range of photon energies in the samples irradiated with higher fluences
(Figure 4.18c and d). The systematic increase of optical absorbance with the increase of helium ion
fluences in samples confirms that the volume fractions of the nanoparticles have increased. It is
inferred from detailed studies, reported elsewhere (Magudapathy et al. 2001), that the sizes of silver
nanoparticles remained unchanged but number densities have increased with the increase of helium

3.0

2.5
(d)
Absorbance (arb. units)

2.0

1.5 (c)

1.0
(b)

0.5 (a)

300 400 500 600 700 800

Wavelength (nm)

FIGURE 4.18 Optical absorption spectra for the nanoscale Ag particles embedded in the ion-exchanged
soda glass samples irradiated with the 100 keV He+ ions. Irradiation was carried out on Ag ion–exchanged
soda glass samples with various fluences of He+ ions: (a) 1 × 1016, (b) 5 × 1016, (c) 5 × 1017, and (d) 1 × 1018
ions cm−2. Optical absorbance has increased with the increase of ion fluence. Optical band-broadening for the
samples with higher fluences is seen.
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 155

ion fluence. The reason for this additional broadening cannot be explained within the framework
of Drude’s free-electron model or from Mie scattering theory. These theories explain the optical
absorption results for noninteracting and a low volume fraction of nonoverlapping metal nanopar-
ticles. However, now we understand that with increase in the number density of silver nanoparticles,
multiple scattering events among the silver nanoparticles play an important role with respect to their
optical response functions. In this regard, a recent study based on the effective dielectric function
formalism (Sancho-Parramon 2009) may be of relevance here. The author has calculated the optical
absorption spectra by taking into account varying concentrations and multiple scattering events
of silver nanoparticles in a silica glass matrix. Based on this computational study, the observed
additional broadening of the SPR absorption is attributed to an increase of the local field fluctua-
tions due to multiple interactions among the high-density silver nanoparticles that are embedded
randomly in the soda glass matrix. As a result of these interactions, silver nanoparticles present in
the glass may support SPR absorptions over a broad range of incident photon energies.
Next we discuss the nucleation and thermal growth of Ag nanoparticles in the ion-exchanged
soda glass samples. Rutherford backscattering spectrometry (RBS) and various precise spec-
troscopy measurements have elucidated these studies. Silver ion–exchanged soda glass samples,
annealed for 1 h in vacuum (1 × 10 −6 mbar) up to 873 K, were studied. We would like to mention
here that ion exchange and post-annealing leave the silver with different charge states in the glass
matrix. This will be confirmed in subsequent sections through a range of spectroscopy studies.
Relevant experimental results would now be analyzed and discussed in detail.

4.3.1.2.1 RBS Measurements

Rutherford backscattering experiments have been carried out on silver ion–exchanged soda glass
samples to measure the concentration profiles of Ag before and after the thermal annealing (iso-
chronal for 1 h in vacuum) at various temperatures. An analyzing beam of He+ ions at the energy
of 2.0 MeV, backscattered by an angle of 160°, was used during the backscattering measurements.
RBS spectra for the silver ion–exchanged soda glass samples before and after annealing are dis-
played in Figure 4.19. A complete spectrum (inset, Figure 4.19) of the silver ion–exchanged soda
glass sample (prior to annealing) reveals other elements (e.g., Ca, Si, Na, and O) present in the
glass matrix. In the silver ion–exchanged soda glass sample, silver has the highest atomic mass
among the other elements present. So, energetic He+ ions transfer minimum energy to the Ag
atoms during elastic collisions and backscatter almost with the same incident energy (Chu et al.
1978). Thus, the He+ ions detected with the highest energy correspond to Ag-edge in the obtained
backscattering spectra, and elements with lower atomic masses appear at respective lower energy
channels (inset, Figure 4.19). For the annealed samples, partial spectra (high-energy channels)
have been shown for better clarity and to emphasize the backscattering from silver atoms only.
It is observed from the backscattering spectra that silver atoms have got accumulated near the
surface of the glass during annealing. Accumulation of silver atoms is higher for the samples
annealed at higher temperatures (Figure 4.19). Near-surface accumulation may be due to the ther-
mal diffusion of silver ions in the silver-exchanged soda-glass matrix. Outward diffusion of silver
ions may help to relax the stress (possibly arising from size difference of Ag+ and Na+ ions; ratio
of ionic sizes of Ag to Na is ~1.29; Wang 1996), thus minimizing the total energy of the system
(Dubiel et al. 2003). Out-diffusion of silver ions at elevated temperatures (723–873 K) explains
the observed increase in the backscattering intensity from Ag atoms near the glass surface. (One
may wonder how these Ag+ ions are generated during annealing at elevated temperatures. To
get an insight into this, better surface-sensitive experiments (e.g., X-ray photoelectron spectros-
copy) have been performed on these samples. So, this would be answered or get clarified a little
later when we analyze XPS as well as PL spectroscopy results.) The obtained backscattering
data have been analyzed with the help of RUMP software program (Doolittle 1986) for calculat-
ing the depth distribution of silver atoms, starting near the surface up to a certain depth inside
the silver-exchanged glass samples. A depth of about 100 nm over which the silver atoms got
156 Semiconductor Nanocrystals and Metal Nanoparticles

3000
10,000

RBS yield (counts)

O
Na Si
2500 5,000 Ca
Ag

2000 0 Ag
0 250 500 750
RBS yield (counts)

Channel no.
1500 (d)

(c)

1000
(b)

500 (a)

0
400 500 600 700 800 900
Channel no.

FIGURE 4.19 RBS spectra of Ag ion–exchanged soda glass samples after isochronal annealing for 1 h
at various temperatures: (a) 593, (b) 723, (c) 823, and (d) 873 K. Accumulation of Ag atoms is higher for
samples annealed at higher temperatures. Inset shows the complete spectrum for Ag ion–exchanged soda
glass sample prior to annealing. (Reprinted with permission from Gangopadhyay, P. et al., Chem. Phys. Lett.,
388, 416, 2004.)

accumulated during annealing has been calculated from the backscattering data. This calculated
depth is significantly smaller than the depth (~5000 nm) to which silver diffused during the silver
ion–exchange process (De Marchi et al. 1996). Thus, the observed phenomenon may be due to the
short-range diffusion of silver ions during the annealing the silver-exchanged soda glass samples.
This result allows considering the case as a semi-infinite diffusion system with the bulk acting
as a constant source of silver ions. By applying diffusion theory for a semi-infinite system, the
accumulated mass of silver atoms per unit area (say, m) near the surface may be calculated from
the following equation (Jacobs 1967):

Dt
m = co (4.21)
p

where
co is the source density
t is the time of diffusion in the sample held at temperature T (i.e., 1 h of annealing time)
D is the diffusion coefficient of silver ions in the soda glass matrix

For the samples annealed at different temperatures (Figure 4.19), the accumulated mass of silver
atoms per unit area is estimated using the expression m = N A M /N Avo (Chu et al. 1978), where NA
is the number of Ag atoms per cm2, obtained from the backscattering data through the RUMP
analysis, M is the atomic weight of silver (~108 atomic mass unit), and NAvo is the Avogadro num-
ber (6.023 × 1023). Assuming the diffusion coefficient (D) to vary as D » e -ea / kBT , where e a is
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 157

the activation energy for thermal diffusion of silver ions in the soda glass matrix, and kB is the
Boltzmann constant, Equation 4.21 may be simplified as
ea
ln(m) = K o - (4.22)
2kBT

where K o = ln(co t /p) is a constant under the present experimental conditions of isochronal annealing
(t = 1 h) as co remains constant for the fixed time of the ion-exchange (2 min here) process. Equation
4.22 is in the Arrhenius form. From the experimentally calculated values of m (using m = N A M /N Avo)
and corresponding temperature of annealing, the data (with symbols ♦) has been plotted in Figure 4.20.
From the Arrhenius plot (the best linear fit to the data) and comparing with Equation 4.22, the slope of
the fitted line is calculated. From the slope of the line, the activation energy for thermal diffusion of sil-
ver ions in the soda glass samples is estimated, which works out to be ~0.74 eV. Note that the activation
energy obtained is much higher than that of diffusion of silver atoms in the soda glass matrix (Miotello
et al. 2000). Diffusion of silver ions would require a higher activation energy than silver atoms. Hence,
it is more than likely that thermal diffusion of silver ions has been predominant in the present study.
4.3.1.2.2 XPS and Auger Spectroscopy
4.3.1.2.2.1  Principle X-ray photoelectron spectroscopy (XPS) is based on a single photon in–electron
out process (as shown in Figure 4.21). The kinetic energy of the emitted photoelectrons is given by
Ek = hn - Eb - j (4.23)

where
Ek is the measured kinetic energy
hν is the energy of the X-ray source
Eb is the binding energy of the electrons in the solid
φ is the work function of the spectrometer

5
In (mass per unit area)

2
12 14 16 18 20
1/kT (eV–1)

FIGURE 4.20 Temperature dependence of the mass of Ag atoms per unit area near the soda glass surface,
showing an Arrhenius behavior. Symbols (♦) are the data estimated from RBS spectra (see Figure 4.19), and
the solid line is the best linear fit to the analyzed backscattering data. Activation energy for thermal diffu-
sion of Ag ions in the soda glass matrix is estimated from the slope of the line. It works out to be ~0.74 eV.
(Reprinted with permission from Gangopadhyay, P. et al., Chem. Phys. Lett., 388, 416, 2004.)
158 Semiconductor Nanocrystals and Metal Nanoparticles

Ek

hν hν
L2,3 Auger emission

L1 L1

K K

FIGURE 4.21 Schematic of photoelectron (XPS) and Auger electron emission processes. The material is
being excited by the X-ray photons of energy hν.

The spectrometer work function value (~4.5 eV) and the X-ray source energy are incorporated in
the data acquisition software. In commercial spectrometers, X-ray sources Al Kα (1486.6 eV) and
Mg Kα (1253.6 eV) are normally used. The kinetic energy distribution (or the binding energy)
of the emitted photoelectrons is measured using a hemispherical electron energy analyzer, and a
photoelectron spectrum is recorded. Binding energy values of photoelectrons are characteristic of
the elements. The quantification of the elements is possible by taking their peak area and the sen-
sitivity of the peaks. Atomic concentration of an element A, for instance, in an alloy composition
AB is given by

I A /S A
c( A) = (4.24)
( I A /S A ) + ( I B /S B )

where
I is the peak intensity (area under the peak)
S stands for sensitivity of the respective element

The binding energy and shape of the photoelectron peak can be analyzed precisely to find the
chemical states of elements and their quantitative composition within the electron penetration depth
of the solid (Briggs and Seah 1990).

4.3.1.2.2.2   Modified Auger Parameters In a photoionization process, Auger electron emission

also takes place (Figure 4.21). A change in the chemical state giving rise to a chemical shift in pho-
toelectron lines will also produce a chemical shift in the Auger lines. However, often the magnitude
of the Auger chemical shift is significantly greater than that of the photoelectron chemical shift. The
modified Auger parameter (MAP, ά) is defined as the sum of the binding energy of the prominent
photoelectrons and the kinetic energy of the prominent Auger electrons of an element. So, by defini-
tion, ά is always a positive number. It may be noted here that MAP can be experimentally found out
only through the XPS system. This parameter is a good characteristic of the chemical state of the
element concerned. Use of this parameter while identifying chemical phases of materials is advanta-
geous, because (1) any systematic error in the photoelectron peak shift (arising as a result of sample
charging) is removed, and (2) because of a higher Auger chemical shift (compared to the photoelec-
tron binding energy shift), the modified Auger parameter (ά) values are distinctly different.
XPS measurements have provided chemical phase information of silver in the ion-exchanged
glass samples before and after different annealing treatments. The binding energy peak of the
Ag 3d5/2 photoelectrons is at 367.5 eV in the as-exchanged glass sample (Figure 4.22). However,
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 159

75

k (counts) 70

65

370 368 366

Binding energy (eV)

FIGURE 4.22 Photoelectron spectra of Ag 3d5/2 electrons from Ag in an ion-exchanged soda glass sample.
Data are fitted with Gaussian–Lorentzian spectral function and Shirley background. Binding energy peak of
Ag 3d5/2 photoelectrons is at 367.5 eV. Photoelectron spectroscopy analysis reveals the chemical phase of Ag
in ion-exchanged soda glass sample as AgO. Modified Auger parameter (MAP) was calculated to confirm the
chemical phase of the Ag in this sample (details are in Table 4.2). (Reprinted from Gangopadhyay, P. et al.,
Phys. Rev. Lett., 94, 047403, 2005. With permission)

chemical analysis using XPS alone is insufficient, particularly in case of silver, where the shift in
the electronic binding energies due to chemical binding is comparable to the resolution of instru-
ment. In order to confirm further the chemical state of silver in the as-exchanged sample, the
modified Auger parameter (ά) has been calculated analytically taking into account the Ag 3d5/2
photoelectron-binding energy and the Ag M4N45N45 Auger transition energy (Figure 4.23). Values
of binding energies, ά, are listed in Table 4.2. The modified Auger parameter is estimated to be
724.9 eV in the as-exchanged sample. This value agrees well with the reported value (724.8 eV)
in the literature for the AgO phase (Hoflund et al. 2000). Thus, the chemical phase has been con-
firmed as AgO in the as-exchanged sample.
For comparison, the parameter ά for silver has been calculated to be 726.2 and 724.1 eV in pure
bulk silver and pure Ag2O powders, respectively (Table 4.2). As seen from this table, the parameter
values (ά) are well separated for different chemical phases of the element (silver here). For example,
with reference to that in bulk silver, shift in the Auger parameter values for silver in AgO and pure
Ag2O powder is 1.3 and 2.1 eV, respectively. These shifts are much higher compared to the bind-
ing energy shift alone (~0.4 eV, Table 4.2), providing much confidence in identifying the chemical
phases of these compounds. Next, in thermally annealed samples, similar analyses confirm the exis-
tence of a mixture of two chemical phases: pure Ag and Ag2O. Details may be found in Table 4.2
and the corresponding XPS spectra are shown in Figure 4.24. Near the surface, the ratio of Ag2O
to Ag remains comparable as the annealing temperature increases. However, this ratio reduces
drastically inside the matrix (Table 4.2 and Figure 4.25). It clearly shows a significant increase in
the volume fractions of pure Ag atoms in the glass samples with increasing annealing tempera-
tures. Because of post-annealing, silver ions may be released in the matrix as a result of thermal
160 Semiconductor Nanocrystals and Metal Nanoparticles

500 M4NN 80
3d5/2
60
450
k (cps)

40
400
20
350
300 320 340 360 400 395 390 385 380 375 370 365 360
(a)

670
75 3d5/2
M4NN
660 70
k (cps)

650 65

640 60
300 320 340 360 400 395 390 385 380 375 370 365 360
(b)

900
3d5/2
M4NN 70
850
k (cps)

60
800
50

750
290 300 310 320 330 340 350 360 400 395 390 385 380 375 370 365 360
(c) Kinetic energy (eV) Binding energy (eV)

FIGURE 4.23 Ag M4N45N45 Auger spectra and photoelectron spectra of the Ag 3d electrons. (a) Pure bulk
Ag sample. (b) Pure Ag2O standard powders. (c) Ag ion-exchanged soda glass sample.

TABLE 4.2
Binding Energies of Ag 3d5/2 Photoelectrons at Different Stages of Sample Treatment
and the Corresponding Modified Auger Parameters (MAP, ά)
Sample BE (FWHM) of Deconvoluted BE Sample
Treatments Ag 3d5/2 (eV) (FWHM) (eV) MAP, ά (eV) Composition Ag2O/Ag Ratio
Bulk Ag 368.4 (1.1) — 726.2 — —
Ag2O powders 367.8 (1.2) — 724.1 — —
As-exchanged 367.5 (1.3) — 724.9 AgO —
Annealed at 723 K 367.9 (1.6) 367.8 (1.3) — Ag2O 2.7
368.4 (1.2) Ag
Annealed at 873 K 367.8 (1.6) 367.9 (1.3) 725.6 Ag2O 2.4
368.4 (1.3) Ag
Annealed at 823 K 368.3 (1.9) 367.9 (1.3) 726.0 Ag2O 0.6
and polished 368.5 (1.5) Ag

Note: Experimentally obtained binding energy of Ag 3d5/2 photoelectrons and ά values for bulk Ag and Ag2O powders
are also shown.
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 161

70

68

66 Ag2O
k (counts)

64 Ag

62

60
370 368 366
(a) Binding energy (eV)

76

74

Ag2O
k (counts)

72

70
Ag

68

370 368 366

(b) Binding energy (eV)

FIGURE 4.24 Deconvoluted photoelectron spectra of Ag 3d5/2 electrons from Ag in the ion-exchanged soda
glass samples post-annealed at (a) 723 and (b) 873 K. Data are fitted with Gaussian–Lorentzian spectral func-
tion and Shirley background. Low (367.8 eV) and high (368.4 eV) energy peaks in the spectra correspond to
Ag2O and Ag materials, respectively. Modified Auger parameters and the intensity ratios of Ag2O/Ag in these
samples are listed in Table 4.2. (Reprinted from Gangopadhyay, P. et al., Phys. Rev. Lett., 94, 047403, 2005.
With permission.)
162 Semiconductor Nanocrystals and Metal Nanoparticles

60

Ag

k (counts)
55

Ag2O

50

370 368 366

Binding energy (eV)

FIGURE 4.25 Deconvoluted photoelectron spectra of Ag 3d5/2 electrons from Ag in the ion-exchanged soda
glass samples post-annealed at 823 K followed by mechanical polishing. XPS data are fitted with Gaussian–
Lorentzian function and Shirley background. Low (367.9 eV) and high (368.5 eV) energy peaks in the spectra
correspond to Ag2O and Ag materials, respectively. Modified Auger parameter and the intensity ratio of
Ag2O/Ag in this sample are listed in Table 4.2. Polishing reduced the intensity ratio of Ag2O/Ag drastically.

decomposition of oxides of silver (AgO and Ag2O). Decomposition reactions of the oxides due to
the thermal annealing may proceed as follows:

3 AgO → Ag + Ag2O + O2 (for annealing temperatures ≤723 K) (Reaction 4.1)

2 Ag2O → 4 Ag + O2 (for annealing temperatures ≥723 K) (Reaction 4.2)

Because of these thermochemical reactions in vacuum, copious amounts of silver ions are released
into the glass matrix. Nucleated silver particles formed during the ion-exchange process can act as
further nucleation sites. Thus, during the thermal annealing, growth of these nucleated Ag particles
may be expected due to diffusion of the silver species (atoms/ions) in the glass matrix. In fact, the
conjecture about the growth of Ag nanoparticles is indeed true and has been confirmed with optical
absorption, LFRS and PL spectroscopy experiments. As we will see in subsequent sections, thermal
growth of the nanoparticles has significantly altered the optical properties of theses samples.
4.3.1.2.3 Optical Absorption Studies
Before discussing the optical absorption spectroscopy results, a visual inspection of these ion-
exchanged glass samples before and after the thermal annealing may be interesting. A simple pho-
tographic image is shown in Figure 4.26 to demonstrate how the colors have changed as a result of
the heat treatment. Glass samples are changing hue towards deeper shades of yellowish-brown with
increase of the annealing temperature. This change has increased opacity of the samples. In other
words, it implies that optical color density has increased consistently with the increase of annealing
temperature. This qualitative visual observation is in conformity with the precise experimental results
of optical absorption spectroscopy.
Experimental optical absorption spectrum for the silver ion-exchanged soda glass sample taken
at room temperature is shown in Figure 4.27. The absorption peak structures observed in this sam-
ple are assigned to excitonic absorptions in silver monoxide (AgO) nanostructures. The chemical
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 163

Opacity increases

As-exch. 593 723 823 K

Temperature of annealing

FIGURE 4.26 Photographic images showing Ag ion-exchanged soda glass samples after different stages of
vacuum annealing for 1 h. Darkening is due to increase of optical absorbance in these samples. (Reprinted
with permission from Gangopadhyay, P., AIP Conference Proceedings, Bikaner, India, Vol. 1536, p. 7. © 2013,
American Institute of Physics.)

0.044

Exciton
0.040 absorptions
Absorbance (arb. units)

0.036
2.28 eV
2.15 eV

0.032

2.0 2.4 2.8 3.2

Photon energy (eV)

FIGURE 4.27 Room-temperature optical absorption spectrum of the Ag ion-exchanged soda glass sample
showing peaks at 2.15 and 2.28 eV (marked with arrows) due to the formation of excitons in AgO. It is to be
noted that PL emissions are also observed at these energies in this sample.

identification of this phase was performed by XPS studies, which were discussed in the previous
section. Silver monoxide is known to be an unstable semiconductor compound. Not many related
studies have been performed with nanostructured AgO materials. However, assuming a two-band
(valence and conduction band) model, the observed peak structures in the optical absorption spec-
trum (Figure 4.27) are explained as follows: photoexcitation of electrons may induce selective
transitions from the valence band (filled states) of the AgO nanostructures to higher energy states
(unoccupied), leaving holes behind, and the excited electrons thus occupy different energy lev-
els very close to the bottom of the conduction band of the material. The excited electron and the
hole may be bound together by their attractive Coulomb interaction, just as an electron bound to
a proton in an atom. The bound electron–hole pair, which is electrically neutral, is known as the
exciton (Kittel 1985). Excitons are quasi-particle excitations. They can mediate optical absorptions
and PL processes in semiconductor materials. The signature of the formation of excitons in AgO
164 Semiconductor Nanocrystals and Metal Nanoparticles

nanostructures has been experimentally observed here. Further confirmation regarding the p­ ossible
existence of the exciton would be supplemented by the experimental results of PL studies in this
sample. Excitons are usually unstable and decay into free charge carriers (electrons and holes).
Ultimately, the electrons recombine with the holes, and in the recombination process there may be
a radiative emission of photons at particular wavelength or energy. The emission is measured in PL
spectroscopy experiments. To call these optical processes as excitonic in nature, the optical absorp-
tion and emission energies ought to match. In the present sample, the exciton absorption peaks
are observed at 2.15 and 2.28 eV, respectively (as shown in Figure 4.27). The absorption peaks are
fairly broadened. This may be because the optical absorption measurements are carried out on the
samples at room temperature. Apart from normal lifetime broadening, quasi-continuous exciton
energy bands just below the conduction band of the AgO nanostructures might be a possible reason
for the large absorption widths at room temperature in this material.
The recorded optical absorption spectra of thermally annealed samples are displayed in
Figure 4.28. Metal nanoparticles are characterized through the characteristic SPR absorp-
tions. This is performed with optical absorption spectroscopy experiments. As discussed before,

4
873 K

823 K
Absorbance (arb. units)

773 K

1
723 K

593 K

300 400 500 600

Wavelength (nm)

FIGURE 4.28 Recorded optical absorption spectra of Ag nanoparticles in the ion-exchanged soda glass
samples before and after isochronal annealing for 1 h at different temperatures (indicated in the figure).
Observed increase of optical absorbance and concomitant reduction in the width of the SPR band with the
rise of annealing temperature indicates thermal growth of Ag nanoparticles in the matrix. (Reprinted from
Gangopadhyay, P. et al., Chem. Phys. Lett., 388, 416, 2004. With permission)
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 165

the position of the resonance wavelength in the optical absorption spectra is specific to the metal,
and it depends on the dielectric constants of the metal and the host matrix. As an example, for Ag
metal nanoparticles in a glass matrix the SPR wavelength position is ~415 nm. Silver is known to
form a supersaturated solid solution because of its low solubility in the soda glass matrix (Pretorious
et al. 1978). So, nucleation of silver atoms to form very minute silver particles can take place because
of the fluctuations in the local concentrations of diffusing silver ions at the onset of the ion-exchange
(Ag+ ↔ Na+) process. Nucleated silver particles formed during the ion-exchange process can act
as further nucleation sites. During annealing, growth of these nucleated particles may be expected
because of thermal diffusion of silver ions in the glass matrix. Growth of the silver nanoparticles may
take place directly from the supersaturation. Assuming the metal nanoparticles to be of spherical
type, average size may increase with the annealing time t as (Miotello et al. 2000, Christian 2002)

cs - ce
d 2 (t ) = do2 + 8 Dt (4.25)
c p - ce

where
do is the value of d (diameter of metal nanoparticles) at t = 0 (i.e., prior to growth)
cp is the concentration of limiting reactants in the particle
D is the diffusion coefficient for the limiting reactants

The role of temperature is contained in D, the diffusion coefficient. cs and ce are the concentration of
the limiting reactant prior to growth and the equilibrium concentration in the matrix, respectively.
The degree of supersaturation (cs - ce ) decreases during this stage of growth. In the present experi-
mental studies, the system is still in the early stage of growth because of the possible diffusion and
reduction processes of the available silver ions. During annealing, the electron capture reaction may
proceed as Ag+ + e – → Ago. Thermal decomposition of silver oxides (according to Reactions 4.1
and 4.2) provides the required silver ions in the glass medium. As a result, the supply of cs would
maintain the degree of supersaturation in the system, and this could also prevent coarsening (where
larger nanoparticles grow at the expense of the smaller ones) to be effective during the time of heat
treatments. Growth of the silver nanoparticles observed here occurs as a result of temperature-
mediated, reaction-controlled supersaturation processes, as prescribed by Equation 4.25. Thus, the
observed increase in the absorption intensity (Figure 4.28) with the increase in the annealing tem-
perature is attributed to the supersaturation growth of the silver nanoparticles in the glass matrix.
The observed growth may also be explained from the Mie formula of light scattering. As can be
seen from Equation 4.3, the SPR absorption intensity is proportional to the volume of the metal
nanoparticles. So, with the growth of the silver nanoparticles, because of thermal annealing, the
resonance optical absorption intensity is expected to increase. In particular, for samples annealed
beyond 723 K, a drastic increase in the optical absorption intensity at 420 nm has been observed
(Figure 4.28). This is because, according to the thermochemical reaction (Reaction 4.2), a large
number of silver ions get released into the soda glass matrix through the thermal decomposition
of Ag2O. (Ag2O is a semiconducting material with a direct bandgap of ~2.23 eV (Varkey and Fort
1993) and decomposes at ~698 K.)
The observed optical absorption features have been discussed further based on theoretical calcu-
lations. Details of the calculation have been discussed earlier. In brief, taking into account of size-
dependent optical material functions (e(w, R)) and applying the Mie’s formula for dipolar absorption,
the optical absorption spectra for nanoscale silver particles in a soda glass matrix have been calcu-
lated. For comparison, the calculated and experimental spectra are shown in Figure 4.29 for different
average sizes of silver nanoparticles. The average sizes of silver nanoparticles are obtained from
LFRS experiments. Details of the Raman measurements will be discussed shortly. The compari-
son between the calculated and experimental absorption spectra shows good agreement for the SPR
position for the silver nanoparticles in the soda glass matrix (as shown in Figure 4.29). The result
166 Semiconductor Nanocrystals and Metal Nanoparticles

2.5

2.0

Absorbance (arb. units)

1.5

1.0

0.5
823 K annealed

723 K annealed
0.0
300 400 500 600 700 800 900
Wavelength (nm)

FIGURE 4.29 Calculated (broken lines) and experimental (continuous lines) optical absorption spectra of
Ag nanoparticles for two different sizes (3.6 and 6.4 nm) in the soda glass matrix. Vertical line indicates the
SPR position. (Reprinted with permission from Gangopadhyay, P., AIP Conference Proceedings, Bikaner,
India, Vol. 1536, p. 7. © 2013, American Institute of Physics.)

shows that the resonance position does not vary with the size of the silver nanoparticles. Further, it is
observed that the agreement is not very good for the spectral width of the resonance band. This dis-
agreement arises because of the presence of an abrupt interface between the metal nanoparticles and
the dielectric matrix. The damping (Γ) is certainly different at the interface compared to that at the
pure nanoparticles. This has been discussed as due to chemical interface damping (Hovel et al. 1993).
The damping depends on the chemical properties of the interface and the energy transfer between
the metal nanoparticles and the surrounding matrix by temporary charge-transfer reactions. Chemical
interface damping has not been properly accounted for in this simplified calculation. However, calcu-
lations properly estimate the trend of line broadening with the decrease of particle sizes.
As the increase in optical absorbance is directly related to the growth of the metal nanoparticles,
it is possible to estimate the activation energy for the diffusion of Ag ions in the glass matrix. From
the experimental data, the logarithm of the peak absorbance versus (kBT )-1 is plotted (Figure 4.30)
to estimate the diffusion parameter. The graph shows an Arrhenius behavior for the growth of silver
nanoparticles in this case. The slope of the line directly gives the activation energy for the diffusion
of silver ions in the soda glass matrix, and it works out to be about 0.76 eV. It may be noted here that
the diffusion activation energy of silver ions estimated from the optical absorption study compares
well with values obtained from RBS experiments.

4.3.1.2.4 Low-Frequency Raman Scattering Studies

LFRS measurements have been carried out on silver ion–exchanged glass samples before and after
thermal annealing. The recorded Raman spectra are shown in Figure 4.31. Details about the Raman
measurements are provided in Gangopadhyay et al. (2004). As discussed earlier, confined surface-
acoustic phonons in metal nanoparticles give rise to low-frequency Raman modes in the vibrational
spectra of the nanomaterials. In this case, as the Raman modes appeared in both polarized and
depolarized geometries, the observed surface acoustic vibrations is assigned to the quadrupolar
(l = 2) Raman mode. By knowing the Raman peak frequencies from experiments, the size of the
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 167

Slope of the line = 0.76

ln (A)
0

–1

13 14 15 16
1/kT (eV–1)

FIGURE 4.30 Arrhenius graph showing the dependence of optical peak absorbance of Ag nanoparticles on
the annealing temperature. Line through the experimental data points is the best linear fit. Slope of the line
gives the activation energy (~0.76 eV) for the diffusion of Ag ions in the soda glass matrix.
Intensity (arb. units)

653 K

Ag-exchanged

10 20 30
Raman shift (cm–1)

FIGURE 4.31 Polarized low-frequency Raman spectra for the Ag nanoparticles in the ion-exchanged soda
glass samples and after vacuum annealing at 653 K. Symbols are the experimental data, and solid lines are
the best fit to data points with the Lorentzian line-shape function and an exponential background. (Spectra are
shifted along the intensity axis for better clarity.)
168 Semiconductor Nanocrystals and Metal Nanoparticles

silver nanoparticles is estimated using the Equation 4.18. The sample details and average sizes
of the silver nanoparticles are given in Table 4.3. The measurements show thermal growth of the
silver nanoparticles in the soda glass matrix. Thermal growth of the silver nanoparticles observed
here is related to the reaction-controlled processes. The growth dynamics here is best described by
Equation 4.25. From the growth of silver nanoparticles and applying Equation 4.25 to the Raman
scattering data (from Table 4.3), the activation energy for the diffusion of silver ions in the soda
glass matrix is estimated. The slope of the Arrhenius plot of ln(d 2 - do2 ) versus (kBT )-1 (Figure 4.32)
is a measure of the activation energy, and it works out to be ~0.72 eV. The value obtained compares

TABLE 4.3
Sizes of the Silver Nanoparticles in the Silver Ion–Exchanged Soda Glass
Samples Before and After Thermal Annealing at Various Temperatures
Annealing Temperature (K) of the Measured Raman Peak Average Sizes of Ag
Ion-Exchanged Glass Frequency (cm−1) Nanoparticles, d (nm)
As-exchanged 20.0 2.35
593 18.8 2.49
653 18.2 2.58
723 12.9 3.63
773 10.2 4.62
823 7.4 6.35
873 6.0 7.82

Note: Sizes have been calculated (using Equation 4.18) from the LFRS measurements. Growth of
the silver nanoparticles with increase of annealing temperature is observed here.

4
ln (d2–d02)

–2
12 14 16 18 20
1/kT (eV–1)

FIGURE 4.32 Arrhenius plot for the calculation of the activation energy for the diffusion of Ag ions in
the soda glass matrix. do and d are the sizes of the Ag nanoparticles in the as-exchanged and in the post-
annealed soda glass samples, respectively. Solid line through the experimental data points is the best linear
fit. Slope (~0.72 eV) of the fitted line estimates the activation energy for the diffusion of Ag ions in the matrix.
(Reprinted from Gangopadhyay, P. et al., Chem. Phys. Lett., 388, 416, 2004. With permission.)
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 169

well with those estimated from other studies (RBS and optical absorption spectroscopy). According
to a previous study, the activation energy for diffusion of silver atoms in the soda glass matrix is
~0.28 eV (Miotello et al. 2000). This value is quite low compared to the value obtained here. Since
thermal diffusion of silver ions in the soda glass matrix is expected to take place with higher activa-
tion energy, it is quite likely that the diffusing species here are silver ions.

4.3.1.2.5 Photoluminescence Studies

The PL spectra of the silver-exchanged soda glass sample and after annealing the samples in high
vacuum at different temperatures are displayed in Figure 4.33. Drastic changes of the PL intensity
are observed on annealing (Figure 4.33). For instance, the PL intensity reaches a maximum and
sharply falls to a minimum value after annealing at 723 and 873 K, respectively. Through a proper
analysis of the PL data, explanations for the observed quenching of PL intensity are provided with
the help of optical absorption and XPS results. A dominant PL peak at 2.15 eV (577 nm) along with
the other less intense peak at 2.28 eV (545 nm) have been observed in the silver ion–exchanged soda
glass sample (Figure 4.34). It may be noticed here that optical absorptions are also observed exactly
at the same energy values (as shown in Figure 4.27). Thus, these emissions may be ascribed to the
excitonic PL in the AgO nanostructures in this sample. Photoexcited (by the 488 nm laser line)
electrons may undergo nonradiative thermal scattering processes before forming bound excitons

160,000

723 K

120,000
PL intensity (counts)

653 K
80,000

823 K

40,000 As-exchanged

Glass

873 K
0
1.6 1.8 2.0 2.2 2.4
Photon energy (eV)

FIGURE 4.33 Room-temperature PL spectra of the Ag ion-exchanged soda glass samples before and after
annealing at temperatures 653, 723, 823, and 873 K. The soda-glass substrate is also shown. Laser excita-
tion is at 488 nm (~2.54 eV). Drastic quenching of the PL intensity is observed after annealing samples
beyond 723 K. This happens because of the thermal growth of Ag nanoparticles in the matrix. (Reprinted from
Gangopadhyay, P. et al., Phys. Rev. Lett., 94, 047403, 2005. With permission.)
170 Semiconductor Nanocrystals and Metal Nanoparticles

653 K 723 K

PL intensity (counts)

As- 823 K
exchanged

1.5 2.0 2.5 1.5 2.0 2.5

Photon energy (eV)

FIGURE 4.34 Experimental PL spectra as a convolution of two Gaussian functions. Symbols are the experi-
mental data, and lines are the fitted functions to the data. Excitonic PL peaks in AgO nanostructures are shown in
the as-exchanged sample. In annealed samples (653 to 823 K), the lower and higher energy PL peaks are from the
Ag nanoparticles and Ag2O materials, respectively. Observed PL bands centered around 1.95 eV are attributed
to Ag nanoparticles, and the emission at 2.23 eV is assigned to the band-to-band radiative transition in Ag2O due
to annealing of the ion-exchanged soda glass samples. (Reprinted from Gangopadhyay, P. et al., Phys. Rev. Lett.,
94, 047403, 2005. With permission.)

(e–h pairs). Excitons are confined in space by Coulomb force and have enhanced probability to
recombine radiatively. Broad PL bands centered around 1.95 eV (637 nm) and 2.23 eV (557 nm) are
displayed for the samples annealed at 653 K (Figure 4.34b) and 723 K (Figure 4.34c). While the
emissions lower than 2.0 eV are attributed to the presence of silver nanoparticles in the soda glass
samples, the emission at 2.23 eV is ascribed to the band-to-band radiative transition in Ag2O nano-
structures (Peyser et al. 2001). The measured PL energy (at 2.23 eV) agrees well with the known
optical bandgap (2.25 eV) of Ag2O materials (Varkey and Fort 1993). It may be noted that observed
PL is from the hybrid nanostructures of Ag and Ag2O in all the annealed samples.
A similar work, in which a time-resolved PL study was carried out, was reported for Ag nanopar-
ticles in a soda lime glass matrix (Karthikeyan 2008). Ag ion–exchanged soda glass slides were
heated in air for 1 h at different temperatures (maximum 823 K) for the Ag nanoparticles to grow.
With the growth of Ag nanoparticles, reduction or quenching of PL intensity and clear red shift of
the emission peak were observed. The shift and quenching of PL intensity was explained as due
to the Kubo effect. According to this model, mean level spacing δ(EF) near the Fermi level (EF) is
given by d( EF ) » EF /zN At , where NAt is the number of atoms in the nanoparticle and z is the valence
of the atom (Kubo 1962). The number of Ag atoms are increased in Ag nanoparticles due to ther-
mal growth, which reduces the energy gap δ(EF) between the sp band and d band of the metallic
nanoparticles. This causes the observed shift and quenching of PL intensity emanating from Ag
nanoparticles (see Figure 4 of Karthikeyan 2008). Nonradiative decay is favored when d  kBT . So,
electron–hole recombination is mostly nonradiative in larger Ag nanoparticles, whereas radiative
emissions are more probable for smaller nanoparticles.
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 171

Time-resolved PL decay at 530 nm from the annealed silver-exchanged soda glasses with 400 nm
excitation is shown in Figure 4.35. The excitation wavelength is chosen to be close to the SPR of
Ag nanoparticles in the matrix. One observes a sharp rise in the time-resolved PL, followed by a
fast decay and a long-lived component. In metal nanoparticles, the excited-state electrons undergo
electron–electron and electron–phonon scattering and get de-excited through the radiative recom-
bination of electrons and holes (Adelt et al. 1998). The scattering events are estimated to be fast
(a few picoseconds). Radiative recombination is long lived and lasts up to 40 ns (Figure 4.35).
Measurements suggest that the optical excitation excites the valence band d electrons to the sp con-
duction band through the interband transition, and, after initial fast electronic scattering processes,
radiative recombination follows, giving rise to the visible PL.
Metal particles of very fine sizes (e.g., ~0.7 nm, the Fermi wavelength of an electron) behave
like multielectron artificial atoms (Link et al. 2002). These particles do exhibit PL; however, the
PL in such cases arises from the intraband transitions of the free electrons in the metal particles
(Zheng et al. 2004). PL of noble metals was first observed and explained by Mooradian (1969). In
the case of noble metals, in particular for silver ([Kr]4d105s1) and gold ([Xe] 4f145d106s1), the optical
properties are due to the d (valence) and the sp (conduction) electrons. As for the origin of PL in
silver nanoparticles in thermally annealed samples, the explanation provided by Mooradian is more
pertinent. Schematic of the optically excited radiative interband recombination of sp electrons and
holes in the d band is shown in Figure 4.36. The band structure of a typical noble metal has been
represented by a simple model, which includes an sp conduction band and two sets of d bands (illus-
trated with the hatches). Excitation takes place from filled states in the upper d bands to levels at and
above the Fermi energy. It is quite likely in this case that PL occurs from the direct recombination
of conduction band electrons below the Fermi energy with d band holes that have been scattered to
momentum states less than the Fermi momentum kF. Band structure calculations of noble metals
suggest that the d bands can be relatively flat at some regions in the momentum space less than
the kF. This might allow photoexcited holes to be scattered within a sufficient range of momentum
values to account for the observed broadening of the PL spectra (as shown in Figure 4.34).

10,000
Ag-350
Log (photon count)

1,000
8,000
100

10
Photon counts

0 20 40
Decay time (ns)
4,000

Instrument response
Ag-350
Ag-450

0 20 40
Decay time (ns)

FIGURE 4.35 Time-resolved PL spectra of heat-treated Ag ion-exchanged soda glasses revealing fast
electronic scattering and long-lived radiative recombination processes in Ag nanoparticles. (Reprinted from
Karthikeyan, B., J. Appl. Phys., 103, 114313, 2008. With permission.)
172 Semiconductor Nanocrystals and Metal Nanoparticles

Conduction
band
Fermi energy

Holes
Energy

d-bands

kF k

FIGURE 4.36 Schematic band structure of a noble metal showing the excitation and radiative interband
recombination between the sp electrons and d-band holes. (Reprinted from Mooradian, A., Phys. Rev. Lett.,
22, 185, 1969. With permission.)

With increase in the annealing temperature, silver may be released in the matrix as a result of
thermal decomposition of AgO. The species AgO formed in the ion-exchanged soda glass sample
is unstable against heating, and decomposes into Ag2O and Ag when annealed under high vac-
uum. The reaction may proceed as 3AgO → Ag + Ag2O + O2 (Reaction 4.1). As a result of this
thermochemical reaction, volume fractions of both nanomaterials (Ag and Ag2O) increase. The
presence of these chemicals or moieties are confirmed by XPS and Auger spectroscopy measure-
ments (as shown in Figure 4.23). Ag2O phase is thermodynamically more stable than AgO (Najafi
1992); further, being a direct bandgap material, it is highly photoluminescent. Thus, increased
amount of the Ag2O causes the PL intensity to reach the highest level at ~723 K. However, around
this temperature, Ag2O thermally decomposes. Decomposition reaction of Ag2O due to thermal
annealing at higher temperatures may proceed as follows: 2Ag2O → 4Ag + O2 (Reaction 4.2).
Further increase of the annealing temperature leads to the rapid growth of the silver nanopar-
ticles (as a result of thermal decomposition of Ag2O) within a depth scale of about 100 nm from
the soda glass surface (backscattering results, Figure 4.19). Thus, thermal decomposition of Ag2O
and rapid growth of silver nanoparticles simultaneously may have resulted in the drastic quench-
ing of the PL intensity observed for the samples post-annealed at 823 and 873 K (as shown in
Figure 4.33). Photoelectron spectroscopy (XPS) results reveal that the Ag2O phase is dominant near
the surface of the post-annealed samples (see Figure 4.24). However, the estimated ratio of Ag2O/Ag
(within a sensitive depth of ~5 nm in the photoelectron spectroscopy) has not really changed with
the increase of annealing temperature (shown in Table 4.2). So, to explain the observed quenching
phenomena further, photoelectron spectroscopy measurements have been carried out on a specially
prepared sample: the exchanged soda glass sample annealed at 823 K was mechanically polished
using 0.25 µm diamond lapping for 5 min to remove a few layers from the surface of the material.
This was performed to avoid the unwanted sputtering effects in the XPS system for obtaining depth
information. Removal of few layers in this sample reasonably ensured that PL and XPS measure-
ments are carried out deep inside and on identical materials in the soda glass sample. Interestingly,
the recorded photoelectron spectra of this sample show significant changes in the peak profiles
(Figure 4.25). Most markedly, the estimated ratio of Ag2O/Ag in the polished sample reduced dras-
tically to 0.6 from 2.7 (see Table 4.2) when compared to the most photoluminescent sample
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 173

(723 K annealed, as shown in Figure 4.33) in the present study. Particularly, the photoelectron
spectroscopy results confirm that high-temperature-annealed samples have a higher volume frac-
tion of Ag than Ag2O. Following the same line of argument, successive quenching of the PL inten-
sity for the 873 K post-annealed sample has been explained. Thus, an important correlation between
temperature-induced changes of the PL intensity and thermal growth of the nanoscale silver par-
ticles in the soda glass matrix has been established through various precise spectroscopic studies.

4.3.1.2.6 Exciton and Surface-Plasmon Interactions

Studying quasi-particle interactions such as between surface plasmons and optical excitons in
metal–semiconductor hybrid nanostructures have attracted great research interest in the last decade
(Achermann 2010, and references therein). Nanoscale silicon structures, for instance, have been
combined with metal nanostructures to explore exciton–plasmon interactions (Walters et al. 2010).
Silicon nanostructures are photoluminescent (though they are indirect bandgap materials) and
have attracted much attention as they can be integrated into existing semiconductor processing
technologies (Heitmann et al. 2005). Hybridized quantum states of plasmons–excitons (named
“­plexcitons”) in the core–shell geometry of hybrid nanomaterials have been obtained and studied
by many researchers (Fofang et al. 2008, Antosiewicz et al. 2014). Naturally, optical emission and
absorption properties of such a complex system depend on interactions of the quasi-particles, sur-
face plasmons, and excitons. Core–shell nanostructures of Ag or Au metal nanoparticles, for exam-
ple, coated with J-aggregates (a class of organic semiconductors, Jelley 1936) exhibit absolutely
different optical absorption properties due to the interactions of surface plasmons with excitons
(Wiederrecht et al. 2008). Variations in their interactions are attributed to several reasons: spectral
overlap of semiconductor emission and surface plasmon energy of metal nanoparticles, their hybrid
geometry, and the interparticle spacing between quantum dots and metal nanoparticles in hybrid
nanomaterials (Hosoki et al. 2008, Haridas et al. 2011, Zhou et al. 2011). In close proximities of
metal and semiconductor nanostructures, effects of weak and strong coupling may be discussed.
In a weak coupling situation, exciton wave functions and electromagnetic modes of surface plas-
mons are considered unperturbed, and exciton–plasmon interactions are described by the coupling
of the exciton’s dipole with the field of the surface plasmons. Phenomena like enhanced absorp-
tion cross-sections, increased radiative rates, and exciton–plasmon energy transfer are described
in the weak coupling regime. In the strong coupling regime, resonant exciton–plasmon interactions
modify exciton wave functions and the surface-plasmon modes. This can also lead to change of
exciton and surface-plasmon resonance energies. In this range of coupling, the excitation energy is
shared, and it oscillates between the plasmonic and excitonic systems (known as Rabi oscillations,
Bellessa et al. 2004). Investigating the coupling mechanisms of surface plasmons with excitons
in hybrid nanomaterials is important due to a wide range of possible applications such as light-
emitting diodes, quantum information and communication technology, and biosensing (Okamoto
et al. 2005, Lee et al. 2007, Zhang et al. 2010, Belacel et al. 2013) among others. Theoretical studies,
advancing the basic understanding of the underlying physical processes such as the resonant energy
transfer between excitons and plasmons, strong electromagnetic coupling interactions with coherent
exciton–plasmon, and so on, play vital roles in designing and modifying the optical properties of
the semiconductor–metal coupled nanosystems of interest (Vasa et al. 2008, Marinica et al. 2013).
With this short introduction, let us now move on to discuss an experimental observation of exci-
ton–surface plasmon interactions in a hybrid nanomaterial. Here, for example, the core–shell hybrid
nanostructure of Ag─Ag2O (Ag as core and Ag2O as shell) is considered. The core–shell hybrid
nanomaterial is formed during the thermal annealing of silver ion–exchanged soda glass samples.
From the relative intensities of Ag 3d5/2 photoelectron peaks of Ag and Ag2O (see Figure 4.25), the
thickness of the Ag2O layer has been estimated. Assuming photoelectrons are emitted from the
Ag2O layer of uniform thickness ς over the metallic Ag nanoparticle, the thickness of the oxide
layer is calculated from V = -r ln(1 - %Ag2O) cos j (Laibinis and Whitesides 1992), where ρ is the
mean free path and φ is the take-off angle of photoelectrons. Using this expression, the thickness
174 Semiconductor Nanocrystals and Metal Nanoparticles

of the Ag2O layer works out to be ~2.2 nm for the sample annealed at 823 K. The shell thickness
is small compared to the average core size of Ag nanoparticles of diameter ~6.4 nm (Table 4.3).
Schematically, the core–shell nanostructure is shown in Figure 4.37. Thus, on the surface of Ag
nanoparticles, a shell of Ag2O has formed during thermal annealing of the samples in a high-vac-
uum atmosphere (Bera et al. 2006). This observation plays an important role in understanding the
optical results shown in Figure 4.38. For example, signatures of excitonic absorptions and optical
emissions from the silver oxide (Ag2O) nanomaterials are absent in the annealed soda glass samples.
This may be explained as follows: because of the incoherent and out-of-phase dipolar interactions
of the surface plasmons with the excitons, in close proximity of Ag nanoparticles and Ag2O shells,
exciton absorptions and emissions in the Ag2O nanomaterial may be suppressed. Possible exci-
tonic optical absorptions in the semiconducting Ag2O have not been observed in the annealed sam-
ples (Figure 4.38). This is probably due to the destructive interferences of excitons (having dipole
 
moment pexi ) with the surface-plasmon oscillations (with a dipole moment pSP) in the Ag2O/Ag hybrid
nanostructures during optical excitations (see Figure 4.37).

4.3.2 Top-Down Methods
Alternative fabrications of nanoparticles and nanostructures have become vitally important as
sizes of devices reach nanoscale dimensions and the resolution of conventional optical lithog-
raphy approaches its physical limit. A variety of alternative approaches, such as thermal dewet-
ting, laser beam irradiation, ion-beam-induced dewetting, and direct writing by e-beam (Bischof
et al. 1996, Lian et al. 2006, Vorobyev et al. 2007, Elbadawi et al. 2013), have been developed for
the synthesis of desired supported nanostructures. Apart from these different successful sample
processing techniques, synthesis (top-down approach) by ion-beam sputtering is also a promising

pexi
Ag

Light
pSP
Ag2O

SPs
Eex
Energy

Coupling

hv
EF
EO
SC
MNP

Electron positions

FIGURE 4.37 Schematic diagram illustrating the surface plasmon–exciton coupling in the core–shell
metal–semiconductor hybrid nanostructure of Ag/Ag2O. In close proximity of the nanoparticles, possible
excitonic optical absorption in Ag2O has not been observed in the annealed samples (also see Figure 4.38).
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 175

0.04 (e)

Exciton
absorptions
2.28 eV

2.15 eV
0.03
Optical absorbance (arb. units)

(d)

0.8
(c)

(b)

0.4

(a)

2.0 2.2 2.4 2.6

Photon energy (eV)

FIGURE 4.38 Optical absorption spectrum revealing excitonic signatures in AgO (e) in the silver ion–
exchanged soda glass sample. Optical absorption spectra of samples are shown after isochronal annealing
at temperatures (a) 653 (b) 723 (c) 773, and (d) 873 K. On annealing between 653 and 873 K, no significant
feature between 2 and 2.6 eV is observed in the absorption spectra except for the known excitation energy
dependence (see spectra (a)–(d)). (Reprinted from Gangopadhyay, P. et al., Phys. Rev. Lett., 94, 047403, 2005.
With permission.) Thermal growth of Ag nanoparticles in the soda-glass matrix has been noticed due to the
annealing (also see Figure 4.28).
176 Semiconductor Nanocrystals and Metal Nanoparticles

method leading to the fabrication of tailored nanostructures on given substrates (Facsko et al. 1999,
Datta and Bhattacharyya 2003, Ruffino et al. 2009a). In fact, the production of submicrometric
and nanometric features on the surfaces of solid metal targets using argon ion beam etching (sput-
tering) at low energy was reported first by Cunningham et al. (1960). In recent times, there have
been several attempts to develop experimental techniques that would allow low-cost and large-
area fabrication of nanoscale structures on supported functional substrates (Li et al. 2001, Chan
and Chason 2007). During ion-beam-induced sputtering, self-organized nanostructures are often
observed. Nanostructuring occurs spontaneously from the interaction of ions with the surface and
near-­surface target atoms. The main underlying processes are ion-induced sputtering, generation
of point defects, and diffusion of surface defects (including adatoms). Nonequilibrium processing
based on ion-beam sputtering is also of interest because it is relatively simple and cost effective com-
pared to direct-write techniques, such as optical or electron beam lithography. The preparation of
nanostructure materials on different substrates is possible using the ion-beam sputtering on various
thin films (e.g., metals, semiconductors, insulators). The process can be suitably controlled through
the selection of the mass of the projectile ions, fluences, ion-beam and target materials parameters,
and the substrate temperature. Ion-beam-induced surface diffusion of the atoms of interest plays
a key role in determining the ultimate topography of ion-irradiated samples. Formation of such
supported novel structures by ion-beam irradiation techniques is thus a technology of choice for
the benefit of seamless integration with the fabrication lines. Nanomaterials consisting of metallic
nanoparticles (particularly noble metals) on insulating dielectric or glass surfaces are technologi-
cally important materials in many areas, ranging from the field of plasmonics (Smolyaninov et al.
2002, Maier and Atwater 2005, Maier 2007) to chemical reaction catalysis (Moser 1996, Moriarty
2001, Cortie and van der Lingen 2002). It has been demonstrated that noble metal nanostructures
of fairly uniform size distribution could be produced, for example, on a silica glass substrate by
the ion-beam sputtering (Gangopadhyay et al. 2010). Morphological differences between different
target metal films do arise because of various effects of the ion-beam-induced sputtering. Factors
that lead to morphological differences are (1) variations of sputtering yields of targets, (2) activation
energies for atomic surface diffusion on the substrate being different for different target metals, and
(3) anisotropic growth.

4.3.2.1 Ion-Beam Sputtering Phenomena

Sputtering or erosion from the surface occurs when energetic ions impinge on a solid surface.
Sputtering is a physical process whereby atoms are ejected from a solid surface as a result of bom-
bardment of the target by energetic particles. It is driven by the momentum transfer between the
ions and atoms in the materials during ion–atom collisions (Sigmund 1969, Behrisch and Eckstein
2007). The incident ions initiate collision cascades in the target. Moving atoms in the cascade may
eventually reach the surface and, if kinetic energy of the atoms is large enough to overcome the
attractive potential of the solid, they may escape from the solid and get sputtered (Thompson 1969).
Two distinct scenarios occur depending on the density of the moving atoms: When the density of the
moving atoms is small, so that the probability of collision between two moving atoms is negligible,
the sputtering is said to be linear. Figure 4.39 schematically describes the sputtering from a linear
collision cascade. Consequence of such linearity is that the sputtering yield, that is, the average
number of ejected atoms per incoming particle, is a linear function of the energy deposited at the
surface. The sputtering yield depends on the ion’s incident angle, the energy of the ion, the masses
of the ion and target atoms, and the surface binding energy of the atoms in the target.
Since the positions of entry of the ions into the solid are statistically distributed, individual ion
stopping positions and details of the collision volume are also statistically distributed. In the lin-
ear cascade regime, the collision volume is approximately of ellipsoidal shape. Contours of equal
energy deposition are ellipsoids of revolution, with the major axis along the incident beam direction
and centered at a depth below the surface, which increases with increasing ion energy. As a result
of this distribution of energy, atoms on/and close to the surface may receive sufficient energy and
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 177

Secondary and
Sputtered atoms higher-order
knock-on atoms

Ar+ ions

Primary knock-on
atoms
Target surface

FIGURE 4.39 Schematic of the Ar+ ion beam–induced sputtering of target atoms due to a linear collision
cascade close to the surface of the target.

momentum, so they are likely to be ejected or sputtered from the surface of a target. A schematic
diagram is shown in Figure 4.39 to show the ion-beam-induced sputtering of target atoms due to a
linear displacement cascade process near the surface of the target.
Point defects (vacancies and interstitials) are created on/and below the target surface as a result
of ion irradiation. Some of these point defects may recombine and annihilate or cluster to form
extended defects; others may migrate to the surface. Radiation-induced release of interstitials that
survive on the surface form the adatoms. Because of the statistical nature of incident ion beam and
the nonoverlapping character of linear collision cascades in the interaction timescale (≈10 −13 s),
these defects are expected to be randomly distributed on the surface, resulting in a rough surface
at the early stage of ion bombardment or prior to any structure formation. Ion-beam-induced migra-
tion or diffusion of these defects plays a significant role in developing the surface morphologies of
ion-irradiated samples.

4.3.2.2 Ion-Beam Sputtering Experiments

Examples of Ar+ ion-beam sputtering experiments on gold and silver metal films on silica glass
substrates are discussed here. A thermal evaporation setup was used to deposit gold and silver
metallic films separately on cleaned silica glass substrates. Subsequently, the as-deposited metal
films were irradiated at room temperature using 100 keV Ar+ ions with various fluences in the
range of 5 × 1015 and 1 × 1017 ions cm−2. Mass-analyzed Ar+ ion beam impinged at normal inci-
dence and scanned over the metallic films on an area ≈1 cm2 for uniform irradiation. The ion
beam current density was kept at ~1 µA cm−2 on the samples during the irradiation to minimize
heating due to the ion beam. The evaporated thicknesses of the silver and gold films were ~60 and
~40 nm, respectively. The thickness of the silver and gold metal films and energy of the Ar+ ions
were chosen (based on the TRIM simulation results, Ziegler et al. 1985) primarily to ensure the
stopping of the ions within the metal/glass interface. The TRIM-calculated projected range and
straggling values of 100 keV Ar+ ions in gold are ~34 and ~20 nm, respectively. Thicknesses of
the as-deposited metal films on silica glass substrates were more than the projected ranges of the
energetic Ar+ ions in the respective metals. Thus, the probability of ballistic transport or mixing
178 Semiconductor Nanocrystals and Metal Nanoparticles

of the metal atoms in the underlying glass substrate could be minimized. This further ensured that
the observed experimental results are only due to the metal nanoparticles and the nanostructured
materials supported on the silica glass substrates. The metallic nanostructures were observed here
after a significant fraction of the metals was ejected or sputtered out as a result of the impact of the
Ar+ ions (Gangopadhyay et al. 2010).
In the next sections, we discuss various experimental (e.g., optical absorption spectroscopy, RBS
spectrometry, scanning electron microscopy) results to highlight the effects of the ion-beam sputtering
on metallic films. Development of various types of nanostructures and nanoparticles of noble metals
on the silica glass substrates has been observed with the increase of fluence of the Ar+ ions. These
morphological transitions are solely due to the ion-beam-induced sputtering of the thin metallic films.

4.3.2.3 Rutherford Backscattering Results

RBS experiments were carried out to measure the areal density of gold atoms in the as-deposited
film (thickness ~40 nm) as well as in the 100 keV Ar+ ion–irradiated samples with various
ion fluences. An analyzing 4He2+ ion beam of energy 2.0 MeV was used during RBS experi-
ments. A surface barrier detector kept at a scattering angle of 165° was utilized for detection of
the backscattered particles. The experimentally obtained backscattering spectra are displayed in
Figure 4.40. The figure displays the RBS spectra of the Au film and of films sputtered by 100 keV
Ar+ ions with various fluences ranging from 5 × 1015 to 7 × 1016 Ar+ ions cm−2. The spectra
(Figure 4.40) show the systematic decrease of the backscattering yield from the Au atoms with the
increase of the Ar+ ion fluence. The decrease in the area under the Au peak (which is proportional
to the Au areal density) with increase of Ar+ ion fluence confirms the loss of Au atoms from the
films due to the ion-beam sputtering.

Gold film Gold film

8000 8000
5 × 1015 ions cm–2
1 × 1016 ions cm–2
RBS yield (arb. units)

6000
3 × 1016 ions cm–2
5 × 1016 ions cm–2 4000
6000
7 × 1016 ions cm–2
2000
RBS yield (arb. units)

0
500 750 1000 1250 1500 1750 2000
4000 Backscattered energy (keV)

2000

0
1750 1800 1850 1900 1950
Backscattered energy (keV)

FIGURE 4.40 Rutherford backscattering spectra of the Au film and Ar+ ion–irradiated Au films as a func-
tion of the ion fluence. Partial spectra to emphasize the backscattering from Au atoms are shown. Systematic
decrease of the area under the Au peak with increase of the ion fluence confirms the ion-beam-induced sput-
tering of the Au film. The inset shows the complete backscattering spectra of the Au film on the silica-glass
substrate. (Reprinted from Gangopadhyay, P. et al., Vacuum, 84, 1411, 2010. With permission.)
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 179

4.3.2.4 Electron Microscopy Results

Scanning electron microscopy measurements were carried out using a Philips XL 30 scanning
electron microscope (SEM). Microscopy experiments were carried out primarily to observe the
morphology of the top surface of the as-deposited gold and silver metallic films in real space, as
well as to study the morphological changes in the post-irradiated samples as a function of the Ar+
ion fluence. Also, the approximate sizes of nanoparticles of gold and silver atoms were estimated
using electron microscopy in the ion-beam-sputtered samples with higher fluences of Ar+ ions. As
discussed before, ion-beam sputtering is known to create many hills and valleys on the surface
of the metals. According to the sputtering model due to Bradley and Harper, atoms in the valley
regions are sputtered preferentially compared to those on the hills (Bradley and Harper 1988). So,
compared to the valleys, sputtering is low from the tops of the hills. Because of the differences in
sputtering efficiencies, hills are more stable against the ion-beam-induced sputtering. This may lead
to morphological contrasts among the hills and the valleys at the microscopic level. This contrast
has been actually observed in the electron micrographs (displayed in Figure 4.41). As discussed in
the earlier section, argon ion-beam-induced sputtering and radiation-induced surface diffusion may
be identified as the underlying process of the observed microscopic features on the surface of the
glass substrate. The scanning electron micrographs (Figure 4.41) display surface morphologies of
the as-deposited gold film on the silica glass substrate as well as the variation of surface morpholo-
gies of the gold film as a function of the fluence of the Ar+ ions.
Interestingly, as shown in electron micrographs, an initial continuous and smooth film of gold
(Figure 4.41a) turns into a discontinuous or network-like film (Figure 4.41b) structure (a kind of
labyrinth nanostructures) because of the ion-beam sputtering. On continuing the sputtering experi-
ments with higher fluences (e.g., 5 × 1016 and 7 × 1016 ions cm−2) of Ar+ ions, drastic changes in
the surface morphology of the gold film were observed (see Figure 4.41c and d). The micrographs
clearly reveal that the gold metal film has been transformed into various sizes of gold nanoparticles

500 nm
500 nm
(a) (b)

500 nm 500 nm
(c) (d)

FIGURE 4.41 Scanning electron micrographs comparing the surface morphologies of the (a) as-depos-
ited Au film (thickness ~40 nm) on the silica glass substrate, and the evolution of the Au nanostructures
as a function of the Ar+ ion fluence of (b) 1 × 1016, (c) 5 × 1016, and (d) 7 × 1016 ions cm−2. (Reprinted from
Gangopadhyay, P. et al., Vacuum, 84, 1411, 2010. With permission.)
180 Semiconductor Nanocrystals and Metal Nanoparticles

on the surface of the glass substrate as a result of the extended ion-beam sputtering in association
with the radiation-induced surface diffusion of the gold atoms.
Because of the ion-beam-induced sputtering of Au atoms, the electrical conductivity of the sam-
ple decreases at higher ion fluences. This has made the SEM images blurred or of poorer quality
(as seen in Figure 4.41d) compared to Figure 4.41c. In corroboration with the RBS results (shown
in Figure 4.40), it only means that the area occupied by the gold atoms on the silica glass surface
has decreased with the increase of fluence of the Ar+ ions. In these ion-beam-sputtered samples, the
average sizes of the Au nanoparticles are measured to be in the range of 50–60 nm (estimated from
Figure 4.41c and d). Nearly uniform sizes of Au nanoparticles are observed in the samples sput-
tered with higher fluences of the Ar+ ions. Also, as shown, the Au nanoparticles are quite isolated.
Although data on radiation-induced adatom diffusion of Au on the surface of the substrate is not
available, maybe the low surface diffusivity of Au atoms on the substrate forces the nanoparticles
to be isolated. This also possibly explains why the Au nanoparticles do not coarsen during the
ion-beam-induced sputtering. Generally speaking, nanoparticles are expected to be isolated and of
uniform size provided diffusion-controlled coarsening is inhibited. This is particularly observed
during the irradiation of Au films with higher fluences of Ar+ ions.
Next, in the case of ion-beam sputtering of Ag metallic films with Ar+ ions of energy 100 keV,
different surface morphological results are observed. For example, a qualitative comparison of
the surface morphologies between the ion-beam-sputtered samples of Au and Ag films can be
made. As shown in the micrographs in Figures 4.41 and 4.42, Ar+ ions of same energy and
with similar ion fluences have sputtered the metal films quite differently. This difference may
be expected. Ion simulation (TRIM) calculations show higher sputtering yield for the Au film

1000 nm 1000 nm

(a) (b)

1000 nm 1000 nm

(c) (d)

FIGURE 4.42 Scanning electron micrographs displaying (a) as-deposited Ag film (thickness ~60 nm) on
silica glass, and Ar+ ion–irradiated Ag films with ion fluence (b) 1 × 1016, (c) 3 × 1016, and (d) 5 × 1016 ions cm−2.
Network structures of Ag at lower ion fluences (see image (b)) give rise to various sizes of Ag nanoparticles
at higher ion fluences (see images (c) and (d)) due to the ion-beam sputtering and radiation-induced surface
diffusion of the Ag atoms on the substrate. (Reprinted from Gangopadhyay, P. et al., Vacuum, 84, 1411, 2010.
With permission.)
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 181

(about nine atoms per ion) compared to the Ag film (about seven atoms per ion) upon Ar+ ion
irradiation (Ziegler et al. 1985). According to Sigmund’s theory of linear cascade sputtering, the
yield is more for heavier atoms (Nastasi et al. 1996). The surface morphologies of the Au and Ag
films sputtered with higher ion fluences are shown in these micrographs (Figures 4.41c and 4.42d)
for comparison. For equal fluences of the Ar+ ions, these micrographs reveal completely different
surface morphologies. As a consequence of higher sputtering yield of Au compared to Ag, and also
due to a low surface diffusivity of Au atoms on the substrate (Levine et al. 1991), nearly uniform
sizes (average size ~50 nm) of Au nanoparticles are produced on the surface (see Figure 4.41c
and d). On the contrary, Ag nanoparticles are found to be of various sizes (the bright speckles in
Figure 4.42d), the maximum size being ~200 nm. The following points need to be considered while
explaining the observed differences in the surface morphologies (shown in Figures 4.41 and 4.42) of
the ion-beam-sputtered samples. First, according to the TRIM calculations and Sigmund’s theory,
sputtering yields are different. Second, according to thermal diffusion studies, the experimentally
found activation energy for the diffusion of Ag atoms was lower than that of Au atoms on silica glass
(Levine et al. 1991, Kim and Alford 2002). These differences are likely to enhance the ripening of
Ag nanoparticles on the silica glass substrate compared to Au nanoparticles.

4.3.2.5 Optical Absorption Results

The presence of the nanoscale metal particles and nanostructures of Au and Ag atoms in the ion-
beam-sputtered metallic films is elaborated further through the optical absorption results. The opti-
cal absorption spectra of the supported Au nanoparticles on the silica glass substrates are displayed
in Figure 4.43. As seen in these spectra, the characteristic optical absorption peak of Au nanopar-
ticles on the substrate has developed with the increase of Ar+ ion fluence. The observed optical
absorption peak is in the range 550–570 nm following the increase of Ar+ ion fluences from 5 × 1016
to 1 × 1017 ions cm−2, respectively. The response peak is attributed to the SPR absorption in Au
nanoparticles (Kreibig and Vollmer 1995). Moreover, as shown in the optical absorption spectra
for the supported Au nanoparticles on the substrate, the peak position of the SPR absorption band
is shifted by ~20 nm toward longer wavelengths (shown by the arrow in Figure 4.43) in the sample

1 × 1017 ions cm–2

0.15 7 × 1016 ions cm–2

5 × 1016 ions cm–2
Absorbance (arb. units)

0.10

0.05

400 500 600 700 800

Wavelength (nm)

FIGURE 4.43 Optical absorption spectra of Au nanoparticles on silica glass substrates. Au nanoparticles
are formed during the ion beam sputtering of Au films with various Ar+ ion fluences (indicated in the figure).
The optical absorption peak around 550 nm is due to the surface-plasmon resonance in Au nanoparticles. The
arrow indicates the shift of the absorption band. (Reprinted from Gangopadhyay, P. et al., Vacuum, 84, 1411,
2010. With permission.)
182 Semiconductor Nanocrystals and Metal Nanoparticles

sputtered with the maximum Ar+ ion fluence of 1 × 1017 ions cm−2. The observed results must be
related to the size, shape, and morphology of the ion-beam-sputtered Au samples on the silica
substrate. The observed red shift (~20 nm) may be due to the reduction in the average sizes of the
Au nanoparticles. For fully embedded Au nanoparticles in a dielectric matrix, size reduction shifts
the absorption band toward lower wavelengths (blue shift) (Kreibig and Genzel 1985, Kreibig and
Vollmer 1995). However, using the discrete dipole approximation calculations, the red shift of SPR
has been theoretically calculated for 10 nm Ag nanoparticles on a mica substrate to highlight the
substrate effects (Kelly et al. 2003). In the present case of supported Au nanoparticles on the sub-
strate, it adds a new level of complexity to the electromagnetic effects on metal nanoparticles due
to dissimilar environments. In the present case, a plausible explanation of the observed red shift
with the decrease in size of the Au nanoparticles may be understood in the following way: Because
of size effects, the number of polarized electrons on the surface of the Au nanoparticles is reduced
(as polarizability is proportional to volume of nanoparticles). Additionally, there could be a finite
polarization imbalance at the metal–dielectric interface for supported metal nanoparticles because
of interaction of surface charges with the glass substrate (Persson 1993). This may lead to further
reduction of the overall charge density. The decreased surface charge corresponds to a smaller
restoring force, which in turn may shift the resonance toward higher wavelengths.
Experimental optical absorption spectrum of supported Ag nanoparticles on the silica glass sub-
strate is displayed (solid line) in Figure 4.44. The sample is obtained after Ag thin film samples
were sputtered with the maximum fluence of 5 × 1016 Ar+ ions cm−2 (energy 100 keV). The recorded
spectrum was deconvoluted to show the optical absorption peaks of Ag nanoparticles on the sub-
strate. The observed peak at 410 nm is attributed to the characteristic dipolar SPR absorption in
nanoscale Ag particles: dipolar absorption is the most dominant. In addition to the dipolar absorp-
tion peak, a low-intensity peak on lower wavelength side (~375 nm) is also observed. Earlier studies
have reported on the appearance of such a shoulder for metal nanoparticles on LiF crystal and silica
glass substrates (Hoheisel et al. 1993, Kreibig and Vollmer 1995, Kelly et al. 2003). This peak is
assigned to the quadrupole resonance absorption in large-sized Ag nanoparticles on the substrate.

0.04 5 × 1016 ions cm–2

Absorbance (arb. units)

0.02

0.00

300 400 500 600 700

Wavelength (nm)

FIGURE 4.44 Optical absorption spectra of the supported Ag nanoparticles on the silica glass substrate.
Various sizes of Ag nanoparticles are formed during the ion beam sputtering of the Ag metallic film with high
fluence (5 × 1016 ions cm−2) of Ar+ ions. In addition to the dipolar absorption peak at 410 nm, a low-intensity
peak on the lower wavelength side (~375 nm) is also observed. This peak is assigned to the quadrupole reso-
nance absorption in large Ag nanoparticles on the silica glass substrate. (Reprinted from Gangopadhyay, P.
et al., Vacuum, 84, 1411, 2010. With permission.)
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 183

In the ion-beam-sputtered sample, large-scale Ag nanoparticles are observed (see the SEM micro-
graph, Figure 4.42d). Related effects, such as phase retardation, field becoming inhomogeneous
across the nanoparticles, and presence of higher multipole optical resonances, are usually observed
in relatively larger metal nanoparticles (Kreibig and Vollmer 1995, Ovchinnikov 2008).

4.4 SUMMARY
Tremendous advances have taken place during last few decades in synthesis of materials of nanoscale
dimensions. In this chapter, we discussed various metal nanoparticles and metal–semiconductor
hybrid nanomaterials with a few examples of some interesting optical (absorption and emission)
and vibrational properties. For instance, collective oscillations of electrons in nanoscale metals give
rise to SPR absorptions. Further, electromagnetic interactions of surface plasmons with excitons
in hybrid nanomaterials show interesting optical features. Because of the phonon confinement on
nanomaterials, their vibrational properties are drastically different from those of bulk materials.
Vibrational modes measured through LFRS can quantify the size of nanomaterials very fast and
in a nondestructive way. Estimation of the thermal growth of metal nanoparticles and activation
energy for the diffusion of metal species in different glass matrices are feasible. Vibrational spec-
troscopy has established the existence of electron–phonon coupling in nanoscale metal particles.
These studies help us to explain plasmon-field-induced enhancements of Raman intensity in metal
nanoparticles.
Ion-beam-induced sputtering (a top-down approach) has ample potential for the synthesis of
nanoparticles and nanostructures of different materials on functional substrates. Depending on the
ion-beam parameters, very interesting surface morphologies may be obtained. For example, uni-
form size of Au nanoparticles has been observed at relatively high fluences of argon ions. The study
also suggests that the higher mobility of Ag atoms compared to Au atoms on the glass surface can
lead to the synthesis of large Ag nanoparticles. Quadrupolar optical absorption is possible in such
large-sized metal nanoparticles.

ACKNOWLEDGMENTS
The author thanks the collaborators and colleagues for many scientific discussions and their help in
performing various experiments that were discussed here. The author is particularly grateful to the
management of IGCAR for the support in pursuing this research study.

REFERENCES
Achermann, M. 2010. Exciton–plasmon interactions in metal-semiconductor nanostructures. J. Phys. Chem.
Lett. 1: 2837.
Adelt, M., S. Nepijko, W. Drachsel, and H.-J. Freund. 1998. Size-dependent luminescence of small palladium
particles. Chem. Phys. Lett. 291: 425.
Akimov, A. V., A. Mukherjee, C. L. Yu et al. 2007. Generation of single optical plasmons in metallic nanow-
ires coupled to quantum dots. Nature 450: 402.
Allegrini, M., N. Garci, and O. Marti (eds.). 2001. Nanometer Scale Science and Technology. Amsterdam,
the Netherlands: IOS Press.
Alivisatos, A. P. 2005. The use of nanocrystals in biological detection. Nat. Biotechnol. 22: 47.
Alù, A. and N. Engheta. 2008. Multifrequency optical invisibility cloak with layered plasmonic shells. Phys.
Rev. Lett. 100: 113901.
Antosiewicz, T. J., S. P. Apell, and T. Shegai. 2014. Plasmon-exciton interactions in a core–shell geometry:
From enhanced absorption to strong coupling. ACS Photon. 1: 454.
Arnold, G. W., G. De Marchi, F. Gonella et al. 1996. Formation of nonlinear optical waveguides by using
ion-exchange and implantation techniques. Nucl. Instrum. Methods Phys. Res. B 116: 507.
Atwater, H. A. 2007. The promise of plasmonics. Sci. Am. 296: 56.
Atwater, H. A. and A. Polman. 2010. Plasmonics for improved photovoltaic devices. Nat. Mater. 9: 205.
184 Semiconductor Nanocrystals and Metal Nanoparticles

Bachelier, G., J. Margueritat, A. Mlayah et al. 2007. Size dispersion effects on the low-frequency Raman
scattering of quasispherical silver nanoparticles: Experiment and Theory. Phys. Rev. B 76: 235419.
Bachelier, G. and A. Mlayah. 2004. Surface plasmon mediated Raman scattering in metal nanoparticles.
Phys. Rev. B 69: 205408.
Balaji, S., B. K. Panigrahi, K. Saravanan et al. 2014. Ion beam shaping of embedded metal nanoparticles by
Si+ ion irradiation. Appl. Phys. A 116: 1595.
Behrisch, R. and W. Eckstein. 2007. Sputtering by Particle Bombardment: Experiments and Computer
Calculations from Threshold to MeV Energies. Berlin, Germany: Springer-Verlag.
Belacel, C., B. Habert, F. Bigourdan et al. 2013. Controlling spontaneous emission with plasmonic optical
patch antennas. Nano Lett. 13: 1516.
Bell, A. T. 2003. The impact of nanoscience on heterogeneous catalysis. Science 299: 1688.
Bellessa, J., C. Bonnand, J. C. Plenet, and J. Mugnier. 2004. Strong coupling between surface plasmons and
excitons in an organic semiconductor. Phys. Rev. Lett. 93: 036404.
Bera, S., P. Gangopadhyay, K. G. M. Nair, B. K. Panigrahi, and S. V. Narasimhan. 2006. Electron spectro-
scopic analysis of silver nanoparticles in a soda-glass matrix. J. Electron Spectrosc. Relat. Phenom.
152: 91.
Bernas, H. (ed.). 2009. Materials Science with Ion Beams. Berlin, Germany: Springer.
Bernas, H. 2010. Can ion beams control nanostructures in insulators? Nucl. Instrum. Methods Phys. Res. B
268: 3171.
Binnig, G., H. Rohrer, Ch. Gerber, and E. Weibel. 1982a. Tunneling through a controllable vacuum gap. Appl.
Phys. Lett. 40: 178.
Binnig, G., H. Rohrer, Ch. Gerber, and E. Weibel. 1982b. Surface studies by scanning tunneling microscopy.
Phys. Rev. Lett. 49: 57.
Bischof, J., D. Scherer, S. Herminghaus, and P. Leiderer. 1996. Dewetting modes of thin metallic films:
Nucleation of holes and spinodal dewetting. Phys. Rev. Lett. 77: 1536.
Boggild, P., T. M. Hansen, C. Tanasa, and F. Grey. 2001. Fabrication and actuation of customized nanotwee-
zers with a 25 nm gap. Nanotechnology 12: 331.
Bohren, C. F. and D. R. Huffman. 1983. Absorption and Scattering of Light by Small Particles. New York:
Wiley.
Bora, M., E. M. Behymer, D. A. Dehlinger et al. 2013. Plasmonic black metals in resonant nanocavities. Appl.
Phys. Lett.102: 251105.
Bottani, C. E., A. Li Bassi, A. Stella, P. Cheyssac, and R. Kofman. 2001. Investigation of confined acoustic
phonons of tin nanoparticles during melting. Europhys. Lett. 56: 386.
Bozhevolnyi, S. I., V. S. Volkov, E. Devaux, J.-Y. Laluet, and T. W. Ebbesen. 2006. Channel plasmon subwave-
length waveguide components including interferometers and ring resonators. Nature 440: 508.
Bradley, R. M. and J. M. E. Harper. 1988. Theory of ripple topography induced by ion bombardment. J. Vac.
Sci. Technol. A 6: 2390.
Briggs, D. and M. P. Seah (eds.). 1990. Practical Surface Analysis—Auger and X-Ray Photoelectron
Spectroscopy, 2nd edn. Chichester, U.K.: Wiley Interscience.
Campbell, C. T., S. C. Parker, and D. E. Starr. 2002. The effect of size-dependent nanoparticle energetics on
catalyst sintering. Science 298: 811.
Chan, W. L. and E. Chason. 2007. Making waves: Kinetic processes controlling surface evolution during low
energy ion sputtering. J. Appl. Phys. 101: 121301.
Chang, D. E., A. S. Sørensen, P. R. Hemmer, and M. D. Lukin. 2006. Quantum optics with surface plasmons.
Phys. Rev. Lett. 97: 053002.
Chang, D. E., A. S. Sørensen, P. R. Hemmer, and M. D. Lukin. 2007. Strong coupling of single emitters to
surface plasmons. Phys. Rev. B 76: 035420.
Christensen, N. E. and B. O. Seraphin. 1971. Relativistic band calculation and the optical properties of gold.
Phys. Rev. B 4: 3321.
Christian, J. W. 2002. The Theory of Transformations in Metals and Alloys. Amsterdam, the Netherlands:
Pergamon.
Christopher, P., H. Xin, A. Marimuthu, and S. Linic. 2012. Singular characteristics and unique chemical bond
activation mechanisms of photocatalytic reactions on plasmonic nanostructures. Nat. Mater. 12: 1044.
Chu, W. K., J. W. Mayer, and M.-A. Nicolet. 1978. Backscattering Spectrometry. New York: Academic Press.
Clavero, C., G. Armelles, J. Margueritat et al. 2007. Interface effects in the magneto-optical properties of Co
nanoparticles in dielectric matrix. Appl. Phys. Lett. 90: 182506.
Cortie, M. B. and E. van der Lingen. 2002. Catalytic gold nano-particles. Mater. Forum 26: 1.
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 185

Courty, A., I. Lisiecki, and M. P. Pileni. 2002. Vibration of self-organized silver nanocrystals. J. Chem. Phys.
116: 8074.
Cunningham, R. L., P. Haymann, C. Lecomte, W. J. Moore, and J. J. Trillat. 1960. Etching of surfaces with
8-keV Argon ions. J. Appl. Phys. 31: 839.
Datta, D. and S. R. Bhattacharyya. 2003. Role of interface modifications in ion-sputtering mechanism of gold
thin films. Nucl. Instrum. Methods Phys. Res. B 212: 201.
De Marchi, G., F. Caccavale, F. Gonella et al. 1996. Silver nanoclusters formation in ion-exchanged wave-
guides by annealing in hydrogen atmosphere. Appl. Phys. A 63: 403.
De Marchi, G., G. Mattei, P. Mazzoldi, C. Sada, and A. Miotello. 2002. Two stages in the kinetics of gold
cluster growth in ion-implanted silica during isothermal annealing in oxidizing atmosphere. J. Appl.
Phys. 92: 4249.
Dickey, J. M. and A. Paskin. 1968. Phonon spectrum changes in small particles and their implications for
superconductivity. Phys. Rev. Lett. 21: 1441.
Doolittle, L. R. 1986. A semiautomatic algorithm for Rutherford backscattering analysis. Nucl. Instrum.
Methods Phys. Res. B 15: 227.
Dubiel, M., H. Hofmeister, G. L. Tan, K.-D. Schicke, and E. Wendler. 2003. Silver diffusion and precipitation
of nanoparticles in glass by ion implantation. Eur. Phys. J. D 24: 361.
Duval, E. 1992. Far-infrared and Raman vibrational transitions of a solid sphere: Selection rules. Phys. Rev.
B 46: 5795.
Duval, E., A. Boukenter, and B. Champagnon. 1986. Vibration eigenmodes and size of microcrystallites in
glass: Observation by very-low-frequency Raman scattering. Phys. Rev. Lett. 56: 2052.
Ehrenreich, H. and H. R. Philipp. 1962. Optical properties of Ag and Cu. Phys. Rev. 128: 1622.
Elbadawi, C., M. Toth, and C. J. Lobo. 2013. Pure platinum nanostructures grown by electron beam induced
deposition. ACS Appl. Mater. Interfaces 5: 9372.
Evanoff, D. D. and Jr. G. Chumanov. 2004. Size-controlled synthesis of nanoparticles. Measurement of extinc-
tion, scattering, and absorption cross sections. J. Phys. Chem. B 108: 13957.
Faccio, D., P. Di Trapani, E. Borsella, F. Gonella, P. Mazzoldi, and A. M. Malvezzi. 1998. Measurement of
the third-order nonlinear susceptibility of Ag nanoparticles in glass in a wide spectral range. Europhys.
Lett. 43: 213.
Facsko, S., T. Dekorsy, C. Koerdt et al. 1999. Formation of ordered nanoscale semiconductor dots by ion sput-
tering. Science 285: 1551.
Faraday, M. 1857. Experimental relations of gold (and other metals) to light. Trans. R. Soc. Lond. 147: 145.
Farah, K., F. Hosni, A. Mejri, B. Boizot, A. H. Hamzaoui, and H. B. Ouada. 2014. Effect of gamma rays
absorbed doses and heat treatment on the optical absorption spectra of silver ion-exchanged silicate
glass. Nucl. Instrum. Methods Phys. Res. B 323: 36.
Fedutik, Y., V. V. Temnov, O. Schops, and U. Woggon. 2007. Exciton-plasmon-photon conversion in plas-
monic nanostructures. Phys. Rev. Lett. 99: 136802.
Feynman, R. P. 1992. There’s plenty of room at the bottom. J. Microelectromech. Syst. 1: 60.
Feynman, R. P. 1993. Infinitesimal machinery. J. Microelectromech. Syst. 2: 4.
Fofang, N. T., T.-H. Park, O. Neumann, N. A. Mirin, P. Nordlander, and N. J. Halas. 2008. Plexcitonic nanopar-
ticles: Plasmon-exciton coupling in nanoshell-J aggregate complexes. Nano Lett. 8: 3481.
Fujii, M., T. Nagareda, S. Hayashi, and K. Yamamoto. 1991. Low-frequency Raman scattering from small
silver particles embedded in SiO2 thin films. Phys. Rev. B 44: 6243.
Gangopadhyay, P. 2013. AIP Conference Proceedings, Bikaner, Rajasthan, India, Vol. 1536, p. 7.
Gangopadhyay, P., R. Kesavamoorthy, S. Bera et al. 2005. Optical absorption and photoluminescence spec-
troscopy of the growth of silver nanoparticles. Phys. Rev. Lett. 94: 047403.
Gangopadhyay, P., R. Kesavamoorthy, K. G. M. Nair, and R. Dhandapani. 2000. Raman scattering studies on
silver nanoclusters in a silica matrix formed by ion-beam mixing. J. Appl. Phys. 88: 4975.
Gangopadhyay, P., P. Magudapathy, R. Kesavamoorthy, B. K. Panigrahi, K. G. M. Nair, and P. V. Satyam.
2004. Growth of silver nanoclusters embedded in soda glass matrix. Chem. Phys. Lett. 388: 416.
Gangopadhyay, P., T. R. Ravindran, K. G. M. Nair, S. Kalavathi, B. Sundaravel, and B. K. Panigrahi. 2007.
Raman scattering studies of cobalt nanoclusters formed during high energy implantation of cobalt ions
in a silica matrix. Appl. Phys. Lett. 90: 063108.
Gangopadhyay, P., S. K. Srivastava, P. Magudapathy, T. N. Sairam, K. G. M. Nair, and B. K. Panigrahi. 2010.
Ion-beam sputtering and nanostructures of noble metals. Vacuum 84: 1411.
Gao, H. L., X. W. Zhang, Z. G. Yin, S. G. Zhang, J. H. Meng, and X. Liu. 2013. Efficiency enhancement of
polymer solar cells by localized surface plasmon of Au nanoparticles. J. Appl. Phys. 114: 163102.
186 Semiconductor Nanocrystals and Metal Nanoparticles

Gersten, J. I., D. A. Weitz, T. J. Gramila, and A. Z. Genack. 1980. Inelastic Mie scattering from rough metal
surfaces: Theory and experiment. Phys. Rev. B 22: 4562.
Ghosh, S. K. and T. Pal. 2007. Interparticle coupling effect on the surface plasmon resonance of gold nanopar-
ticles: From theory to applications. Chem. Rev. 107: 4797.
Gonzalo, J., A. Perea, D. Babonneau et al. 2005. Competing processes during the production of metal nanopar-
ticles by pulsed laser deposition. Phys. Rev. B 71: 125420.
Hamanaka, Y., K. Fukuta, A. Nakamura, L. M. Liz-Marzán, and P. Mulvaney. 2004. Enhancement of third-
order nonlinear optical susceptibilities in silica-capped Au nanoparticle films with very high concentra-
tions. Appl. Phys. Lett. 84: 4938.
Haridas, M., L. N. Tripathi, and J. K. Basu. 2011. Photoluminescence enhancement and quenching in metal-
semiconductor quantum dot hybrid arrays. Appl. Phys. Lett. 98: 063305.
Haynes, C. L. and R. P. Van Duyne. 2001. Nanosphere lithography: A versatile nanofabrication tool for stud-
ies of size-dependent nanoparticle optics. J. Phys. Chem. B 105: 5599.
Haynes, C. L. and R. P. Van Duyne. 2003. Plasmon-sampled surface-enhanced Raman excitation spectros-
copy. J. Phys. Chem. B 107: 7426.
Heitmann, J., F. Muller, M. Zacharias, and U. Gosele. 2005. Silicon nanocrystals: Size matters. Adv. Mater.
17: 795.
Hoa, X. D., A. G. Kirk, and M. Tabrizian. 2007. Towards integrated and sensitive surface plasmon resonance
biosensors: A review of recent progress. Biosens. Bioelectron. 23: 151.
Hoflund, G. B., Z. F. Hazos, and G. N. Salaita. 2000. Surface characterization study of Ag, AgO, and Ag2O
using x-ray photoelectron spectroscopy and electron energy-loss spectroscopy. Phys. Rev. B 62: 11126.
Hoheisel, W., M. Vollmer, and F. Trager. 1993. Desorption of metal atoms with laser light: Mechanistic
studies. Phys. Rev. B 48: 17463.
Hollstein, T., U. Kreibig, and F. Leis. 1977. Optical properties of Cu and Ag in the intermediate region between
pure Drude and interband absorption. Phys. Status Solidi (b) 82: 545.
Hosokawa, M., K. Nogi, M. Naito, and T. Yokoyama (eds.). 2012. Nanoparticle Technology Handbook,
2nd edn. Oxford, U.K.: Elsevier.
Hosoki, K., T. Tayagaki, S. Yamamoto, K. Matsuda, and Y. Kanemitsu. 2008. Direct and stepwise energy
transfer from excitons to plasmons in close-packed metal and semiconductor nanoparticle monolayer
films. Phys. Rev. Lett. 100: 207404.
Hovel, H., S. Fritz, A. Hilger, U. Kreibig, and M. Vollmer. 1993. Width of cluster plasmon resonances: Bulk
dielectric functions and chemical interface damping. Phys. Rev. B 48: 18178.
Hughes, M. A., Y. Fedorenko, B. Gholipour et al. 2014. n-Type chalcogenides by ion implantation. Nat.
Commun. 5: 5346.
Jacobs, M. H. 1967. Diffusion Processes. New York: Springer-Verlag.
Jelley, E. E. 1936. Spectral absorption and fluorescence of dyes in the molecular state. Nature 138: 1009.
Jie, G., B. Liu, H. Pan, J.-J. Zhu, and H.-Y. Chen. 2007. CdS nanocrystal-based electrochemiluminescence
biosensor for the detection of low-density lipoprotein by increasing sensitivity with gold nanoparticle
amplification. Anal. Chem. 79: 5574.
Kala, S., R. Theissmann, and F. E. Kruis. 2013. Generation of AuGe nanocomposites by co-sparking tech-
nique and their photoluminescence properties. J. Nanopart. Res. 15: 1963.
Karthikeyan, B. 2008. Fluorescent glass embedded silver nanoclusters: An optical study. J. Appl. Phys. 103:
114313.
Kelly, K. L., E. Coronado, L. L. Zhao, and G. C. Schatz. 2003. The optical properties of metal nanoparticles:
The influence of size, shape, and dielectric environment. J. Phys. Chem. B 107: 668.
Kim, H., H.-B.-R. Lee, and W.-J. Maeng. 2009. Applications of atomic layer deposition to nanofabrication and
emerging nanodevices. Thin Solid Films 517: 2563 (a review article).
Kim, H. C. and T. L. Alford. 2002. Thickness dependence on the thermal stability of silver thin films. Appl.
Phys. Lett. 81: 4287.
Kittel, C. 1985. Introduction to Solid State Physics. New Delhi, India: Wiley Eastern Limited.
Kneipp, K., H. Kneipp, I. Itzkan, R. R. Dasari, and M. S. Feld. 2002. Surface-enhanced Raman scattering and
biophysics. J. Phys.: Condens. Matter 14: R597.
Kreibig, U. and L. Genzel. 1985. Optical absorption of small metallic particles. Surf. Sci. 156: 678.
Kreibig, U. and M. Vollmer. 1995. Optical Properties of Metal Clusters. Berlin, Germany: Springer.
Kubo, R. 1962. Electronic properties of metallic fine particles. J. Phys. Soc. Jpn. 17: 975.
Kumbhar, A., L. Spinu, F. Agnoli, K.-Y. Wang, W. Zhou, and C. J. O’Connor. 2001. Magnetic properties of
cobalt and cobalt-platinum alloy nanoparticles synthesized via microemulsion technique. IEEE Trans.
Magnet. 37: 2216.
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 187

Lamb, H. 1882. On the vibrations of an elastic sphere. Proc. Lond. Math. Soc.13: 187.
Laibinis, P. E. and G. M. Whitesides. 1992. Self-assembled monolayers of n-alkanethiolates on copper are
barrier films that protect the metal against oxidation by air. J. Am. Chem. Soc. 114: 9022.
Lee, J., P. Hernandez, J. Lee, A. O. Govorov, and N. A. Kotov. 2007. Exciton-plasmon interactions in molecu-
lar spring assemblies of nanowires and wavelength-based protein detection. Nat. Mater. 6: 291.
Lee, T.-H. and R. M. Dickson. 2003. Discrete two-terminal single nanocluster quantum optoelectronic logic
operations at room temperature. Proc. Natl. Acad. Sci. U.S.A. 100: 3043.
Levine, J. R., J. B. Cohen, and Y. W. Chung. 1991. Thin film island growth kinetics: A grazing incidence small
angle x-ray scattering study of gold on glass. Surf. Sci. 248: 215.
Li, J., D. Stein, C. McMullan, D. Branton, M. J. Aziz, and J. A. Golovchenko. 2001. Ion-beam sculpting at
nanometre length scales. Nature 412: 166.
Lian, J., L. Wang, X. Sun, Q. Yu, and R. C. Ewing. 2006. Patterning metallic nanostructures by ion-
beam-induced dewetting and Rayleigh instability. Nano Lett. 6: 1047.
Lifshitz, I. M. and V. V. Slyozov. 1961. The kinetics of precipitation from supersaturated solid solutions.
J. Phys. Chem. Solids 19: 35.
Link, S., A. Beeby, S. FitzGerald, M. A. El-Sayed, T. G. Schaaff, and R. L. Whetten. 2002. Visible to infrared
luminescence from a 28-atom gold cluster. J. Phys. Chem. B 106: 3410.
Loo, C., A. Lowery, N. Halas, J. West, and R. Drezek. 2005. Immunotargeted nanoshells for integrated cancer
imaging and therapy. Nano Lett. 5: 709.
Luo, X., A. Morrin, A. J. Killard, and M. R. Smyth. 2006. Application of nanoparticles in electrochemical
sensors and biosensors. Electroanalysis 18: 319.
Luther, J. L., P. K. Jain, T. Ewers, and A. P. Alivisatos. 2011. Localized surface plasmon resonances arising
from free carriers in doped quantum dots. Nat. Mater. 10: 361.
Magudapathy, P., P. Gangopadhyay, B. K. Panigrahi, K. G. M. Nair, and S. Dhara. 2001. Electrical transport
studies of Ag nanoclusters embedded in a glass matrix. Physica B 299: 142.
Maier, S. A. 2007. Plasmonics: Fundamentals and Applications. Berlin, Germany: Springer-Verlag.
Maier, S. A. and H. A. Atwater. 2005. Plasmonics: Localization and guiding of electromagnetic energy in
metal/dielectric structures. J. Appl. Phys. 98: 011101.
Maier, S. A., M. L. Brongersma, P. G. Kik, S. Meltzer, A. A. G. Requicha, and H. A. Atwater. 2001.
Plasmonics—A route to nanoscale optical devices. Adv. Mater. 13: 1501.
Maisels, A., F. E. Kruis, H. Fissan, B. Rellinghaus, and H. Zähres. 2000. Synthesis of tailored composite
nanoparticles in the gas phase. Appl. Phys. Lett. 77: 4431.
Marinica, D. C., H. Lourenco-Martins, J. Aizpurua, and A. G. Borisov. 2013. Plexciton quenching by resonant
electron transfer from quantum emitter to metallic nanoantenna. Nano Lett. 13: 5972.
Mazzoldi, P. and G. Mattei. 2005. Potentialities of ion implantation for the synthesis and modification of metal
nanoclusters. Riv Nuovo Cimento 28: 1.
Mie, G. 1908. Beitrage zur optik truber medien speziell kolloidaler metallosungen. Ann. Phys. Lpz. 25: 377.
Miotello, A., M. Bonelli, G. De Marchi, G. Mattei, P. Mazzoldi, and C. Sada. 2001. Formation of sil-
ver nanoclusters by excimer laser interaction in silver-exchanged soda-lime glass. Appl. Phys. Lett.
79: 2456.
Miotello, A., G. De Marchi, G. Mattei, P. Mazzoldi, and A. Quaranta. 2000. Clustering of silver atoms in
hydrogenated silver-sodium exchanged glasses. Appl. Phys. A 70: 415.
Mooradian, A. 1969. Photoluminescence of metals. Phys. Rev. Lett. 22: 185.
Moriarty, P. 2001. Nanostructured materials. Rep. Prog. Phys. 64: 297.
Moser, W. R. 1996. Advanced Catalysts and Nanostructured Materials. San Diego, CA: Academic Press.
Moskovits, M. 1985. Surface-enhanced spectroscopy. Rev. Mod. Phys. 57: 783.
Muñoz, J. E., J. Cervantes, R. Esparza, and G. Rosas. 2007. Iron nanoparticles produced by high-energy ball
milling. J. Nanopart. Res. 9: 945.
Murray, D. B., C. H. Netting, L. Saviot et al. 2006. Far-infrared absorption by acoustic phonons in titanium
dioxide nanopowders. J. Nanoelectron. Optoelectron. 1: 92.
Murray, W. A. and W. L. Barnes. 2007. Plasmonic materials. Adv. Mater. 19: 3771.
Nahm, C., D.-R. Jung, J. Kim et al. 2013. Photoluminescence enhancement by surface-plasmon resonance:
Recombination-rate theory and experiments. Appl. Phys. Exp. 6: 052001.
Najafi, S. I. (ed.). 1992. Introduction to Glass Integrated Optics. Norwood, MA: Artech House.
Nakayama, K., K. Tanabe, and H. A. Atwater. 2008. Plasmonic nanoparticle enhanced light absorption in
GaAs solar cells. Appl. Phys. Lett. 93: 121904.
Nalwa, H. S. (ed.). 2004. Encyclopedia of Nanoscience and Nanotechnology. Los Angeles, CA: American
Scientific Publishers.
188 Semiconductor Nanocrystals and Metal Nanoparticles

Nastasi, M., J. W. Mayer, and J. K. Hirvonen. 1996. Ion-Solid Interactions: Fundamentals and Applications.
Cambridge, U.K.: Cambridge University Press.
Okamoto, K., I. Niki, A. Scherer, Y. Narukawa, T. Mukai, and Y. Kawakami. 2005. Surface plasmon enhanced
spontaneous emission rate of InGaN/GaN quantum wells probed by time-resolved photoluminescence
spectroscopy. Appl. Phys. Lett. 87: 071102.
Ovchinnikov, V. 2008. Formation and characterization of surface metal nanostructures with tunable optical
properties. Microelectron. J. 39: 664.
Palik, E. D. (ed.). 1985. Handbook of Optical Constants of Solids. Orlando, FL: Academic Press Inc.
Persson, B. N. J. 1993. Polarizability of small spherical metal particles: Influence of the matrix environment.
Surf. Sci. 281: 153.
Peyser, L. A., A. E. Vinson, A. P. Bartko, and R. M. Dickson. 2001. Photoactivated fluorescence from individual
silver nanoclusters. Science 291: 103.
Portales, H., L. Saviot, E. Duval et al. 2001. Resonant Raman scattering by breathing modes of metal nanopar-
ticles. J. Chem. Phys. 115: 3444.
Pradeep, T. and Anshup. 2009. Noble metal nanoparticles for water purification: A critical review. Thin Solid
Films 517: 6441.
Pretorius, R., J. M. Harris, and M.-A. Nicolet. 1978. Reaction of thin metal films with SiO2 substrates. Solid-
State Electron. 21: 667.
Prieto, G., J. Zecevic, H. Friedrich, K. P. de Jong, and P. E. de Jongh. 2013. Towards stable catalysts by control-
ling collective properties of supported metal nanoparticles. Nat. Mater. 12: 34.
Quinten, M. 2011. Optical Properties of Nanoparticle Systems: Mie and Beyond. Weinheim, Germany:
Wiley-VCH.
Ramjauny, Y., G. Rizza, S. Perruchas, T. Gacoin, and R. Botha. 2010. Controlling the size distribution of
embedded Au nanoparticles using ion irradiation. J. Appl. Phys.107: 104303.
Raza, S., N. Stenger, S. Kadkhodazadeh et al. 2013. Blueshift of the surface plasmon resonance in silver
nanoparticles studied with EELS. Nanophotonics 2: 131.
Reiss, S. and K. Heinig. 1995. Computer simulation of mechanisms of the SIMOX process. Nucl. Instrum.
Methods Phys. Res. B 102: 256.
Ren F., C. Jiang, C. Liu, J. Wang, and T. Oku. 2006. Controlling the morphology of Ag nanoclusters by ion
implantation to different doses and subsequent annealing. Phys. Rev. Lett. 97: 165501.
Rizza, G., H. Cheverry, T. Gacoin, A. Lamasson, and S. Henry. 2007. Ion beam irradiation of embedded
nanoparticles: Toward an in situ control of size and spatial distribution. J. Appl. Phys. 101: 014321.
Roco, M. C., R. S. Williams, and A. P. Alivisatos (eds.). 1999. Nanotechnology Research Directions: IWGN
Workshop Report. Dordrecht, the Netherlands: Kluwer Academic Publishers.
Ruffino, F., M. G. Grimaldi, C. Bongiorno et al. 2009a. Normal and abnormal grain growth in nanostructured
gold film. J. Appl. Phys. 105: 054311.
Ruffino, F., A. Irrera, R. De Bastiani, and M. G. Grimaldi. 2009b. Room-temperature grain growth in sputtered
nanoscale Pd thin films: Dynamic scaling behaviour on SiO2. J. Appl. Phys. 106: 084309.
Saha, K., S. S. Agasti, C. Kim, X. Li, and V. M. Rotello. 2012. Gold nanoparticles in chemical and biological
sensing. Chem. Rev. 112: 2739.
Salerno, M., J. R. Krenn, B. Lamprecht et al. 2002. Plasmon polaritons in metal nanostructures: The optoelec-
tronic route to nanotechnology. Optoelectron. Rev. 10: 217.
Sancho-Parramon, J. 2009. Surface plasmon resonance broadening of metallic particles in the quasi-
static approximation: A numerical study of size confinement and interparticle interaction effects.
Nanotechnology 20: 235706.
Sanders, A. W., D. A. Routenberg, B. J. Wiley, Y. Xia, E. R. Dufresne, and M. A. Reed. 2006. Observation of
plasmon propagation, redirection, and fan-out in silver nanowires. Nano Lett. 6: 1822.
Saviot, L., A. Mermet, and E. Duval. 2011. Acoustic vibrations in nanoparticles. In Handbook of Nanophysics,
Vol. 3, K. D. Sattler (ed.). Boca Raton, FL: Taylor & Francis.
Saviot, L., C. H. Netting, and D. B. Murray. 2008. Inelastic neutron scattering due to acoustic vibrations con-
fined in nanoparticles: Theory and experiment. Phys. Rev. B 78: 245426.
Scholl, J. A., A. L. Koh, and J. A. Dionne. 2012. Quantum plasmon resonances of individual metallic nanopar-
ticles. Nature 483: 421.
Sendova, M., J. A. Jiménez, R. Smitha, and N. Rudawskic. 2015. Kinetics of copper nanoparticle precipitation
in phosphate glass: an isothermal plasmonic approach. Phys. Chem. Chem. Phys. 17: 1241.
Sigmund, P. 1969. Theory of sputtering. I. Sputtering yield of amorphous and polycrystalline targets. Phys.
Rev. 184: 383.
Optical, Photoluminescence, and Vibrational Spectroscopy of Metal Nanoparticles 189

Simmons, G. and H. Wang (eds.). 1971. Single Crystal Elastic Constants and Calculated Aggregate Properties:
A Handbook, 2nd edn. Cambridge, U.K.: MIT.
Smolyaninov, I. I., A. V. Zayats, A. Gungor, and C. C. Davis. 2002. Single-photon tunneling via localized
surface plasmons. Phys. Rev. Lett. 88: 187402.
Srivastava, S. K., P. Gangopadhyay, S. Amirthapandian et al. 2014. Effects of high-energy Si ion-irradiations
on optical responses of Ag metal nanoparticles in a SiO2 matrix. Chem. Phys. Lett. 607: 100.
Stamplecoskie, K. G. and J. C. Scaian. 2010. Light emitting diode irradiation can control the morphology and
optical properties of silver nanoparticles. J. Am. Chem. Soc. 132: 1825.
Tao, A., F. Kim, C. Hess et al. 2003. Langmuir-Blodgett silver nanowire monolayers for molecular sensing
using surface-enhanced Raman spectroscopy. Nano Lett. 3: 1229.
Tamura, A., K. Higeta, and T. Ichinokawa. 1982. Lattice vibrations and specific heat of a small particle.
J. Phys. C 15: 4975.
Thomas, O. C., W. Zheng, S. Xu, and Jr. K. H. Bowen. 2002. Onset of metallic behavior in magnesium clus-
ters. Phys. Rev. Lett. 89: 213403.
Thompson, M. W. 1969. Defects and Radiation Damage in Metals. Cambridge, U.K.: Cambridge University
Press.
Varkey, A. J. and A. F. Fort. 1993. Some optical properties of silver peroxide (AgO) and silver oxide (Ag2O)
films produced by chemical-bath deposition. Sol. Energy Mater. Sol. Cells 29: 253.
Vasa, P., R. Pomraenke, S. Schwieger et al. 2008. Coherent exciton-surface-plasmon-polariton interaction in
hybrid metal-semiconductor nanostructures. Phys. Rev. Lett. 101: 116801.
Vorobyev, A. Y., V. S. Makin, and C. Guo. 2007. Periodic ordering of random surface nanostructures induced
by femtosecond laser pulses on metals. J. Appl. Phys. 101: 034903.
Wagner, C. 1961. Theory of transformation in sludge throw the resolution. Z. Elektrochemie. 65: 581.
Walters, R. J., R. V. A. van Loon, I. Brunets, J. Schmitz, and A. Polman. 2010. A silicon-based electrical
source of surface plasmon polaritons. Nat. Mater. 9: 21.
Wang, P. W. 1996. Thermal stability of silver in ion-exchanged soda lime glasses. J. Vac. Sci. Technol. A 14: 465.
Wang, Y., T. Yang, M. T. Tuominen, and M. Achermann. 2009. Radiative rate enhancements in ensembles of
hybrid metal-semiconductor nanostructures. Phys. Rev. Lett. 102: 163001.
Wiederrecht, G. P., G. A. Wurtz, and A. Bouhelier. 2008. Ultrafast hybrid plasmonics. Chem. Phys. Lett.
461: 171.
Xiao, Y., F. Patolsky, E. Katz, J. F. Hainfeld, and I. Willner. 2003. Plugging into enzymes: Nanowiring of
redox enzymes by a gold nanoparticle. Science 299: 1877.
Xiong, Y., R. Long, D. Liu et al. 2012. Solar energy conversion with tunable plasmonic nanostructures for
thermoelectric devices. Nanoscale 4: 4416.
Xu, H., E. J. Bjerneld, M. Kall, and L. Borjesson. 1999. Spectroscopy of single hemoglobin molecules by
surface enhanced Raman scattering. Phys. Rev. Lett. 83: 4357.
Xu, H. and K. S. Suslick. 2010. Water-soluble fluorescent silver nanoclusters. Adv. Mater. 22: 1078.
Yoshida, K., T. Itoh, H. Tamaru, V. Biju, M. Ishikawa, and Y. Ozaki. 2010. Quantitative evaluation of elec-
tromagnetic enhancement in surface-enhanced resonance Raman scattering from plasmonic properties
and morphologies of individual Ag nanostructures. Phys. Rev. B 81: 115406.
Zhang, H., K.-B. Lee, Z. Li, and C. A Mirkin. 2003. Biofunctionalized nanoarrays of inorganic structures
prepared by dip-pen nanolithography. Nanotechnology 14: 1113.
Zhang, J. T., Y. Tang, K. Lee, and M. Ouyang. 2010. Tailoring light-matter-spin interactions in colloidal
hetero-nanostructures. Nature 466: 91.
Zheng, J., C. Zhang, and R. M. Dickson. 2004. Highly fluorescent, water-soluble, size-tunable gold quantum
dots. Phys. Rev. Lett. 93: 077402.
Zhou, L., C. Zhang, Y. Yang, B. Li, and L. Zhang. 2012. Narrow size distribution of Au nanocrystals formed
in sapphire by utilizing Ar ion irradiation and thermal annealing. Nucl. Instrum. Methods Phys. Res.
B 278: 42.
Zhou, X. D., X. H. Xiao, J. X. Xu, G. X. Cai, F. Ren, and C. Z. Jiang. 2011. Mechanism of the enhancement
and quenching of ZnO photoluminescence by ZnO–Ag coupling. Europhys. Lett. 93: 57009.
Ziegler, J. F., J. P. Biersack, and U. Littmark. 1985. The Stopping and Range of Ions in Solids. New York:
Pergamon.
 5 Properties and Potential Applications
Silicon Nanocrystals

Spiros Gardelis

CONTENTS
5.1 Overview................................................................................................................................ 191
5.2 Introduction........................................................................................................................... 192
5.3 A Historic Overview.............................................................................................................. 193
5.4 Methods of Fabrication of Silicon Nanocrystals................................................................... 193
5.4.1 Electrochemistry: Porous Silicon and Stain Etching................................................ 193
5.4.2 Silicon Nanocrystals Embedded in Dielectric Matrices........................................... 194
5.4.2.1 Oxidation of Porous Silicon........................................................................ 194
5.4.2.2 Silicon Nanocrystals Embedded in Silicon Dioxide................................... 194
5.4.2.3 Silicon Nanocrystals Embedded in Other Dielectric Matrices
(Silicon Nitride–Silicon Oxynitride).......................................................... 195
5.4.2.4 Growth of Silicon Nanocrystals by Laser Ablation.................................... 196
5.4.2.5 Colloidal Chemistry.................................................................................... 196
5.5 Optical Properties of Silicon Nanocrystals........................................................................... 196
5.5.1 Light Emission from Silicon Nanocrystals—General.............................................. 196
5.5.2 Light Emission from Silicon Nanocrystals Embedded in Silicon Dioxide
and Silicon Nitride..................................................................................................... 198
5.5.3 Photoluminescence (PL) Dynamics........................................................................... 199
5.5.4 Other Issues Regarding Light Emission....................................................................202
5.5.4.1 Optical Gain................................................................................................202
5.5.4.2 Multiple Exciton Generation (MEG) in Silicon Nanocrystals....................202
5.5.4.3 Silicon Nanocrystals as Sensitizers............................................................203
5.5.5 Electroluminescence.................................................................................................. 203
5.6 Charging Effects in Silicon Nanocrystals—Nonvolatile Memories..................................... 205
5.7 Applications of Silicon Nanocrystals....................................................................................205
5.8 Conclusions............................................................................................................................206
References.......................................................................................................................................207

5.1 OVERVIEW
Silicon nanocrystals exhibit unique optical and electrical properties due to their quantum size and
charging effects, respectively. In the 1990s, for the first time, it was observed that nanostructured
silicon in the form of porous silicon emitted efficiently visible light even at room temperature. Later,
it was also observed that silicon nanocrystals with sizes smaller than the Bohr radius of the exciton
in bulk silicon could also emit visible light. These observations triggered enormous efforts by the
scientific community to use silicon nanostructures for optoelectronic applications, as this material
could be integrated in an all-silicon chip containing optoelectronic components. Beyond their appli-
cations in optoelectronics, silicon nanocrystals showed potential use in sensors, photonics, and solar
cells. The most recent report on 100% internal quantum efficiency of light-emitting silicon nano-
crystals has revived interest in silicon nanocrystals as an efficient light-emitting material that could

191
192 Semiconductor Nanocrystals and Metal Nanoparticles

replace the toxic Cd- and Pb-based quantum dots in many biological and medical applications. On
the other end, the charging effects observed in silicon nanocrystals have already been exploited
in embedded nonvolatile memories in microcontrollers. In this chapter, we give a brief historical
account on this field, present methods of growing silicon nanocrystals, and review their optical and
electrical properties as well as the prospects for their applications in electronics, sensors, photonics,
photovoltaic devices, biology, and medicine.

5.2 INTRODUCTION
After the discovery of the transistor in 1948, there was a revolution in the electronics industry,
and semiconductors have been the key materials used extensively in microelectronics, optoelec-
tronics, and solar cells. An important property of a semiconductor that makes it useful for the
aforementioned applications is its energy bandgap. The existence of a bandgap poses a barrier
in the carrier transport. At energies above this barrier, carriers can flow as in the case of a metal
resistor, whereas below it the carriers cannot flow and the material acts as an insulator. The
presence of the bandgap results in the existence of two types of carriers in the semiconductor:
electrons and holes. Suitable doping of the semiconductor with acceptors or donors can make the
material p-type or n-type, respectively. Depending on the doping concentration, the resistivity
can be tailored over a broad range. This makes possible the fabrication of diodes and transistors
for various applications. Thus, the rectifying behavior of a diode and the property of switching
in a transistor—the two properties microelectronics is based on—are due to the existence of the
energy bandgap.
Another important consequence of the existence of energy bandgap in semiconductors is light
emission and photoconductivity, which are essential in optoelectronics and solar cell applications,
respectively. Miniaturization and integration are two essential properties of semiconductors that
have helped the semiconductor industry contribute significantly to the world economic growth
steadily in the last 40 years with significant consequences in technology and society. Both these
properties have driven the race for increasing densities of semiconductor devices that can be incor-
porated in a single chip in order to make processors faster and carry out more complicated tasks.
Of course, the leading semiconducting material for this race is silicon. The reason for this is that
silicon is abundant in nature and thus cheaper to fabricate. Although the first transistor was fabri-
cated using germanium, this material was quickly abandoned in the microelectronics industry as
it has a lower cutoff current with smaller temperature variation compared to silicon and a poorer
ability to withstand heat than silicon. Besides, silicon is more abundant and thus cheaper than
germanium.
Although silicon is a semiconductor material with very good electrical properties, it is an
indirect bandgap semiconductor, and as such it is not very useful for optoelectronic applications.
The idea of all-silicon optoelectronics, that is, the fabrication of a silicon chip where lasers,
photodetectors, and transistors are fully integrated, has been the driving force for finding materi-
als with good optoelectronic properties that could be integrated with silicon. Many significant
efforts have been made in this direction. SiGe, erbium-doped silicon, and silicon nanocrystals
were the best candidates to fulfill this idea of an all-silicon superchip. Prompted by this impor-
tant potential application of silicon nanocrystals, we will review the methods of their fabrication,
their optoelectronic properties, and, finally, their applications that span from optoelectronics to
biology and medicine.
This chapter is structured as follows: First, we will give a historic review of this subject. Then
we will present the fabrication methods of silicon nanocrystals. Then we will review the optical,
electrical, and optoelectronic properties of silicon nanocrystals, which are different from those of
bulk silicon as a result of quantum size effects. Finally, we will review the possible applications of
silicon nanocrystals.
Silicon Nanocrystals 193

5.3 A HISTORIC OVERVIEW

In 1990, Canham observed efficient room-temperature photoluminescence in the visible region
from porous silicon [1]. This was strange, as silicon is an indirect energy bandgap material and as
such it cannot emit efficiently. This was literally the starting point of rigorous research on the origin
of light emission from porous silicon. After many contradicting theories put forward to explain the
origin of the effect, there is now consensus that light emission is a result of quantum and spatial con-
finement of electrons and holes, which recombine radiatively within the nanostructured silicon that
exists between the pores of porous silicon. For light emission to occur, the silicon nanostructures
must have dimensions smaller than the Bohr radius of exciton in bulk silicon (~4.3 nm) [2]. The idea
that quantum confinement within the silicon nanostructures in porous silicon is the origin of light
emission led the researchers to fabricate silicon nanocrystals with controllable sizes able to emit in
different energies within the visible and near-infrared region [3–5].
In addition to the light emission properties of the silicon nanocrystals, unique electrical proper-
ties were also observed because of their charging effects [6–8]. This initiated rigorous research to
use this property in nonvolatile memory structures [6,9–11].

5.4 METHODS OF FABRICATION OF SILICON NANOCRYSTALS

Various routes of fabrication have been used to produce silicon nanocrystals such as electrochemi-
cal, sputtering, ablation, and colloidal chemistry. We review briefly each of these methods.

5.4.1 Electrochemistry: Porous Silicon and Stain Etching

Electrochemical anodization of crystalline silicon in aqueous or ethanolic solutions of hydro-
fluoric acid (HF) can cause the selective etching of silicon and modify it into a porous material.
Between the pores, the remaining silicon retains its crystallinity. Depending on the conditions of
the anodic reaction, the material can be nanoporous to microporous (pore size < 2 nm), mesopo-
rous (pore size 2–50 nm), or macroporous (pore size > 50 nm). Figure 5.1 shows an example of

10 nm

FIGURE 5.1 Transmission electron microscopy (TEM) image of a microporous silicon layer. (Reproduced
with permission from Cullis, A.G., Canham, L.T., and Calcott, P.D.J., The structural and luminescence prop-
erties of porous silicon, J. Appl. Phys., 82, 909. © 1997, American Institute of Physics.)
194 Semiconductor Nanocrystals and Metal Nanoparticles

microporous silicon [12]. Specifically, the morphology of porous silicon depends on the doping
and type of the starting bulk silicon, the concentration of hydrofluoric acid in the electrolyte, and
the anodic current density [12,13]. The etching takes place with the help of holes. Thus, if the
starting material is p-type, the anodic reaction is carried out in the dark. In the case where the
starting material is of n-type and the majority carriers are electrons, etching can occur only under
illumination so that the necessary holes can be generated for the reaction. The thickness of the
porous silicon layer depends on the anodization reaction time.
Stain etching is another method to fabricate nanocrystalline silicon. In this method, crystalline
Si is dipped in aqueous solutions containing HF and HNO3 [14]. Selective etching of Si occurs via
its oxidation by hole injection from HNO3 and subsequent etching of the oxide by the HF. The thick-
ness of the film is limited by the diffusion of the ionic species and not by the reaction time [15].
Increasing the impurity doping of the substrate results in thicker films [16].
Porous Si produced by both electrochemical and chemical etching of silicon is hydrogen-
terminated and very vulnerable to environmental oxidation, which affects its stability with regard
to its light emission properties. In order to stabilize the porous structure against degradation in light
emission, controlled oxidation was carried out by chemical, electrochemical, or thermal oxida-
tion processes. The next logic step was to consider fabricating silicon nanocrystals embedded in a
dielectric matrix.

5.4.2 Silicon Nanocrystals Embedded in Dielectric Matrices

Two dielectric matrices compatible with Si technology were investigated in order to produce Si
nanocrystals of well-controlled sizes and to stabilize mainly their light emission properties: they
were stoichiometric silicon dioxide and silicon nitride. In the following, we review the fabrication
methods of these structures.

5.4.2.1 Oxidation of Porous Silicon

Oxidation of porous silicon can produce silicon nanocrystals that are interconnected and covered
by a silicon dioxide layer. We now review the different methods of oxidation of porous silicon. Four
different methods were used to oxidize the pore walls intentionally and stabilize them against fur-
ther oxidation due to exposure to the ambient air. This can be done chemically, electrochemically,
by thermal oxidation, or by wet oxidation. Chemical oxidation can be done in oxidants such as
hydrogen peroxide, nitric acid, and dimethylsulfoxide [17,18]. Electrochemical oxidation takes place
in electrolytes such as KNO3 with the use of current in an electrochemical cell [19]. Thermal oxida-
tion is performed in dry oxygen ambient in a furnace at temperatures between 300°C and 900°C.
At moderate temperatures between 300°C and 600°C, oxidation causes loss of hydrogen from the
pore walls, resulting in poor electronic and light emission properties. To resolve this problem,
a subsequent high-temperature oxidation (between 800°C and 900°C)—usually a rapid one in order
to avoid total oxidation of porous silicon—must be performed. This process leads to a good passiv-
ation of the pore walls with stoichiometric silicon dioxide and consequently improves and stabilizes
the light emission and electronic properties of the material [20–22]. Finally, wet oxidation at high
pressures [23] or at ambient pressure [24] has also been used to passivate porous silicon with a good
quality layer of silicon dioxide.

5.4.2.2 Silicon Nanocrystals Embedded in Silicon Dioxide

The luminescent properties of porous silicon led various research groups to develop new meth-
ods of producing silicon nanostructures. In these methods, the silicon nanocrystals are literally
embedded in silicon dioxide. We will review these methods, which produce silicon nanocrystals
with controllable sizes. One approach is ion implantation of silicon in a silicon dioxide matrix. Ion
doses above the equilibrium solid solubility limit results in phase separation during ion implantation
Silicon Nanocrystals 195

and aggregation of silicon ions, which eventually form clusters. Post-annealing, depending on the
temperature, and the ion dose determine the final size and the crystallinity of the resulting silicon
clusters. Also, acceleration energies determine the depth to which these ions can penetrate and even-
tually form the silicon clusters [25–27]. Another approach of silicon nanocrystal growth in silicon
dioxide is by sputtering. One example is sputtering of silicon suboxides and subsequent annealing.
Annealing at temperatures as high as 1100°C in a nitrogen ambient separates the silicon suboxide
into two different phases of silicon and silicon dioxide, forming silicon nanocrystals embedded in
a silicon dioxide matrix [3]. Cosputtering of crystalline silicon and silicon dioxide targets and
subsequent annealing at similar temperatures in nitrogen also leads to aggregation of silicon in
silicon nanocrystals within a silicon dioxide matrix [7].
Growth of silicon nanocrystals embedded in silicon dioxide out of the gas phase and subsequent
thermal oxidation is another well-known method. The method has two variations. The first one is
called low-pressure chemical vapor deposition (LPCVD) and the other one is plasma-enhanced
chemical vapor deposition (PECVD). Silane is used in both techniques as a precursor. In LPCVD,
the temperature of the reaction can be as high as 600°C, whereas in PECVD the deposition can be
done at lower temperatures. For example, the growth of silicon on a layer of silicon dioxide grown
by LPCVD at a growth temperature 580°C is amorphous with nanocrystalline nucleation sites [4,5],
whereas at growth temperatures above 600°C the silicon layers grown are nanocrystalline [28,29].
In PECVD, where the deposition is performed at a much lower growth temperature (e.g., 250°C),
the deposited silicon layer is amorphous. In this case, the crystallization temperature needs to be as
high as 1000°C [30]. In either LPCVD or PECVD, to grow well-separated silicon nanocrystals by
silicon dioxide, a subsequent step of thermal oxidation at elevated temperatures, usually between
800°C and 900°C, is necessary. The oxidation time and the initial thickness of the nanocrystalline
silicon layer determine the final size of the silicon nanocrystals. In the PECVD method, an anneal-
ing step is necessary before thermal oxidation to crystallize the initially amorphous silicon layer.
The size of the silicon nanocrystals grown by these CVD techniques after thermal oxidation is
well controlled and the size dispersion narrow. However, in order to fabricate silicon nanocrystals
well separated by silicon dioxide barriers, the initial nanocrystalline layer thickness must be small
enough, that is, a few nanometers. Silicon nanocrystalline layers have also been grown on quartz
substrates by these techniques. In the case of LPCVD growth, the nanocrystalline layer is columnar,
consisting of silicon nanocrystals with a certain lateral size dispersion, and their dimension in the
growth direction is equal to the thickness of the nanocrystalline layer [28,29]. In this case, silicon
nanocrystals embedded in silicon dioxide can be fabricated after thermal oxidation but the initial
nanocrystalline layer must be thin enough so that the nanocrystals become separated by a silicon
dioxide barrier [28,29]. Figure 5.2 shows two examples of multiple layers of silicon nanocrystals
embedded in silicon dioxide [3,5].

5.4.2.3 Silicon Nanocrystals Embedded in Other Dielectric

Matrices (Silicon Nitride–Silicon Oxynitride)
One drawback of silicon dioxide is its high barrier, which makes carrier injection and tunneling
through it more difficult and thus electroluminescence poor, and the prospect of using this system
as a light-emitting diode (LED) is challenging [31,32]. Another important drawback is that the color
tunability of such a structure is limited by defects at the silicon dioxide–silicon nanocrystal inter-
face, which introduce energy states within the energy bandgap of the silicon nanocrystals [33,34].
To resolve these problems and make silicon nanocrystals more useful for optoelectronic applica-
tion, silicon dioxide has been replaced by silicon nitride, which has a lower energy bandgap barrier.
This has been realized by growing silicon-rich silicon nitride layers either by PECVD or LPCVD
and subsequently annealing them at elevated temperatures so that silicon within these layers can
agglomerate forming silicon nanocrystals within a stoichiometric silicon nitride matrix [35–39].
196 Semiconductor Nanocrystals and Metal Nanoparticles

Glue

4 nm 3

5 nm 1
Quartz
(a) (b)

FIGURE 5.2 Silicon nanocrystals/silicon dioxide superlattices fabricated (a) by silicon suboxide sputter-
ing and subsequent annealing, and (b) by LPCVD deposition of silicon and subsequent thermal oxidation.
(a: Reproduced with permission from Zacharias, M., Heitmann, J., Scholz, R., Kahler, U., Schmidt, M., and
Bläsing, J., Size-controlled highly luminescent silicon nanocrystals: A SiO/SiO2 superlattice approach, Appl.
Phys. Lett., 80, 661. © 2002, American Institute of Physics; b: Reproduced with permission from Gardelis,
S., Nassiopoulou, A.G., Vouroutzis, N., and Frangis, N., Effect of exciton migration on the light emission
properties in silicon nanocrystal ensembles, J. Appl. Phys., 105, 113509. © 2009, American Institute of
Physics.)

5.4.2.4 Growth of Silicon Nanocrystals by Laser Ablation

Another technique of silicon nanocrystal growth is laser ablation. In this technique, the target can
be crystalline silicon onto which a pulsed laser is focused. The laser heats locally the silicon target,
generating a plasma, which nucleates on the substrate and forms silicon nanocrystals that are unpas-
sivated. Controlled passivation with either silicon dioxide or hydrogen makes the silicon nanocrys-
tals to emit visible light [40,41].

5.4.2.5 Colloidal Chemistry

Chemistry offers versatile routes to produce silicon nanocrystals with suitable surface function-
alization for various applications, mainly biological ones [42]. These chemical routes of silicon
nanocrystal fabrication are based on the reduction of halide salts [43,44], reaction of alkali silicides
[45–47], or thermal treatment of sol–gel precursors [48,49]. Finally, it has been demonstrated that
silicon nanocrystals can be doped. Such an example is phosphorous doping using PCl3 [50].

5.5 OPTICAL PROPERTIES OF SILICON NANOCRYSTALS

5.5.1 Light Emission from Silicon Nanocrystals—General
The first observation of light emission from silicon nanostructures was reported by Canham in 1990
and was from porous silicon. This was attributed from the very beginning to the quantum confine-
ment of the excitons within the silicon nanocrystals or nanowires of porous silicon [1,51]. Other
groups also confirmed efficient room-temperature luminescence from other silicon nanostructures
such as porous SiGe [51], silicon nanocrystals embedded in silicon dioxide [3,4], or silicon nitride
[36–39]. There has been a very vivid discussion on the origin of the light emission, which was
Silicon Nanocrystals 197

sparked by the fact that luminescence from porous silicon was vulnerable to the chemical termina-
tion of the surface. For example, the fact that luminescence was quenched after desorption of hydro-
gen in the freshly anodized porous silicon and recovered after its immersion in hydrofluoric acid
solutions led some researchers to claim that silicon hydride [52] or siloxane [53] was the origin of the
light emission. However, observations complying with the quantum confinement of excitons within
the silicon nanocrystals in the porous silicon such as the blue shift of the luminescence [1,51,54] and
of the absorption edge with decreasing size of the silicon nanocrystals, which were also confirmed
by theoretical calculations [55], led to the firm conclusion that quantum confinement of excitons is
the origin of luminescence. Surface termination and passivation of silicon nanostructures determine
the emission properties, but are not the origin of luminescence [56]. For example, hydrogen pas-
sivation does not limit blue shift of the luminescence with decreasing size, and the luminescence
follows the bandgap opening. On the contrary, silicon oxide termination limits the blue shift, as it
introduces energy states within the energy bandgap of the silicon nanocrystals that are involved in
the emission [33]. Different chemicals on the surface can quench luminescence as they introduce
nonradiative centers. Specifically, a number of organic and inorganic molecules adsorbed on the
surface of the silicon nanocrystals exhibited a quenching behavior for the light emission. This effect
has been mainly explained by the energy or electron transfer between the intrinsic energy states of
the nanocrystals and those of the quencher [57–59]. For example, in the presence of aromatic nitro
compounds, luminescence is quenched because of the transfer of an electron from the conduction
band of the nanocrystal to a vacant orbital of the quencher [60,61]. Moreover, it was reported that
suitable surface passivation of the silicon nanocrystals can result in quantum yields as high as
40%–70% for silicon nanocrystals [62–64]. More recently, colloidal silicon nanocrystals with suit-
able ligand passivation showed quantum yields near 100% [65].
Why then can silicon nanocrystals be so bright when silicon itself is an indirect bandgap semi-
conductor? Two preconditions must be fulfilled in order to observe light emission from a silicon
nanocrystal. The most important one is that the size of the silicon nanocrystal must be smaller than
the Bohr radius of exciton in bulk silicon, which has been calculated to be 4.3 nm [2,66]. The second
one is the passivation of any dangling bonds, which could act as nonradiative centers and quench
the luminescence, with suitable atoms such as hydrogen. We will examine each of these conditions
separately. The spatial confinement of the generated excitons within a silicon nanocrystal smaller
than 4.3 nm results in a significant overlap, which increases the oscillator strength of the radia-
tive emission even though silicon is an indirect bandgap semiconductor and as such any transition
needs phonons to be fulfilled. Because of the Heisenberg uncertainty principle, spatial confinement
enhances the uncertainty in the momentum space k. Thus, the selection rule for phonon-assisted
transitions breaks down, resulting in a much larger number of phonon modes fulfilling the radia-
tive recombination of the excitons within the silicon nanocrystal [67]. The second precondition
for efficient luminescence is the passivation of the dangling bonds. This can be done by hydrogen
passivation, which protects silicon nanocrystals from oxidation that results in the degradation of
luminescence efficiency and also in the change in the energetics of the luminescence [33,68]. Other
than hydrogen passivation, additional passivation is needed in the case where silicon nanocrystals
are to be used as luminescent labels in biological applications where cadmium-based quantum
dots can be toxic. Thus, the need for a suitable functionalization of their surface, so that silicon
nanocrystals could be linked to biomolecules, has led to the alkylation of their hydrogen-terminated
surface [69–71]. This makes them chemically stable, able to retain the efficient luminescence, and
also suitably functionalized for such applications.
Another important consequence of the confinement of the carriers in the silicon nanocrystals is
quantization of the energy states where the carrier can be found. This occurs when the size of the
nanocrystal is smaller than the de Broglie length, or in other words, the wavelength that corresponds
to the carrier. In the case of silicon, the wavelength of the electron is 8 nm. When the size of the
nanocrystal decreases below the electron wavelength, then quantum size effects and a widening of
the energy bandgap occur. Thus, for nanocrystal sizes smaller than the exciton Bohr radius, where
198 Semiconductor Nanocrystals and Metal Nanoparticles

efficient light emission is observed, the energy states of the nanocrystal become quantized and the
energy bandgap increases with decreasing nanocrystal size.

5.5.2 Light Emission from Silicon Nanocrystals Embedded

in Silicon Dioxide and Silicon Nitride

From the very beginning of the investigation of light emission from silicon nanostructures, the role
of oxygen in the degradation and the modification of the energetics of the luminescence from these
structures were acknowledged. This was first observed in luminescent porous silicon. Freshly pre-
pared porous silicon is very reactive to the oxygen of the atmosphere, and as it ages in atmosphere,
hydrogen is replaced gradually by silicon oxides. This causes quenching and spectral shift of the
initial efficient luminescence of the freshly prepared porous silicon. It has been observed that for
sizes of silicon nanocrystals smaller than 3 nm, although oxidation would cause decrease of the
silicon size and thus a blue shift to the luminescence spectrum, the luminescence peak could not
shift to energies higher than 2.1 eV. Instead, after air exposure, a red shift of the luminescence as
large as 1 eV was observed for sizes less than 2 nm. Wolkin et al. [33] were the first to resolve this
controversy by introducing a model in which silicon oxide was assumed to introduce energy states
in the energy bandgap of the smaller silicon nanocrystals (smaller than 3 nm), which are involved
in the radiative recombination process, thus determining the energetics of the luminescence.
Figure 5.3 shows the evolution of the electronic states in oxidized silicon nanocrystals as a function
of their size, as proposed by Wolkin et al. [33].
Theoretical work supported this idea and explained further the role of oxidation of the silicon
nanostructures in light emission. It was confirmed that silicon oxides introduce interface states
within the energy bandgap of the silicon nanocrystals or even strain, which modify luminescence
spectrum from the case of the hydrogen termination of the nanocrystals [72–75].
However, a deliberate passivation of the surface of porous silicon with a good-quality thermal
silicon dioxide formed at elevated temperatures (800°C–900°C) has been proved to decrease the
number of nonradiative centers and stabilize the photoluminescence properties of porous silicon

1
Energy (eV)

Zone III Zone II Zone I

C.B. energy
–1 V.B. energy
Trapped electron
Trapped hole
–2
0 1 2 3 4 5
Diameter (nm)

FIGURE 5.3 Electronic states in oxidized silicon nanocrystals as a function of their size. The trapped electron
state is a p-state localized on the Si atom of the Si=O bond, and the trapped hole state is a p-state localized
on the oxygen atom. (Reprinted with permission from Wolkin, M.V., Jorne, J., Fauchet, P.M., Allan, G., and
Delerue, C., Phys. Rev. Lett., 82, 197, 1999. © 1999 by the American Physical Society.)
Silicon Nanocrystals 199

against degradation in atmosphere. However, oxidation results in a large Stokes shift (~1 eV).
The same occurs in the light emission of silicon nanocrystals embedded in silicon dioxide.
An example of this Stokes shift of the photoluminescence peak relative to the absorption edge is
shown in Figure 5.4 [29]. Regardless of the large Stokes shift, tunability of the energy bandgap
with the size of the silicon nanocrystals has been shown. Figure 5.5 shows the blue shift of the
photoluminescence spectra of silicon nanocrystals embedded in a silicon dioxide matrix with
decreasing nanocrystal size [3].
Efficient luminescence has also been observed in silicon nanocrystals embedded in silicon
nitride and attributed to quantum size effects within the silicon nanocrystals [32,35–39]. Colloidal
silicon nanocrystals have also shown efficient luminescence, which can be tailored from the near-
infrared to the whole visible spectrum depending on the size [42].

5.5.3 Photoluminescence (PL) Dynamics

PL decay measurements have shown a stretched exponential time dependence of the PL intensity,
IPL(t), expressed by
b
I PL (t ) = I 0e -( t / tPL ) (5.1)

where
I0 is the photoluminescence intensity at t = 0
τPL is a characteristic decay time, or carrier recombination time
β is a dispersion factor that takes values between 0 and 1 and expresses the deviation of the
photoluminescence time decay from the single exponential [5,76,77]

Figure 5.6 shows PL decay times extracted from stretched exponential fittings (insert) as a function
of temperature for different films of silicon nanocrystals [5].

1 2 3 4
1200
60
Photoluminescence (PL) intensity (a.u.)

1000
(ahv)1/2 (cm–1/2eV1/2)

800

600
30

400

200

0 0
1 2 3 4
Energy (eV)

FIGURE 5.4 Example of Stokes shift of photoluminescence peak relative to the absorption edge in a layer of
silicon nanocrystals embedded in silicon dioxide. The size of the nanocrystals is less than 2 nm. (Reproduced with
permission from Gardelis, S., Nassiopoulou, A.G., Manousiadis, P., Milita, S., Gkanatsiou, A., Frangis, N., and
Lioutas, Ch.B., Structural and optical characterization of two-dimensional arrays of Si nanocrystals embedded in
SiO2 for photovoltaic applications, J. Appl. Phys., 111, 083536. © 2012, American Institute of Physics.)
200 Semiconductor Nanocrystals and Metal Nanoparticles

Diameter equal or below

3.8 nm 2.0 nm
1

PL intensity (a.u.)

0
1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8
Energy (eV)

FIGURE 5.5 Normalized PL spectra from silicon nanocrystals embedded in silicon dioxide as a func-
tion of their sizes. (Reproduced with permission from Zacharias, M., Heitmann, J., Scholz, R., Kahler, U.,
Schmidt, M., and Bläsing, J., Size-controlled highly luminescent silicon nanocrystals: A SiO/SiO2 superlattice
approach, Appl. Phys. Lett., 80, 661. © 2002, American Institute of Physics.)

The measured PL recombination rate rPL is the sum of the radiative (rr) and nonradiative (rnr)
recombination rates according to the following expression:

1 1 1
rPL = rr + rnr = = + (5.2)
tPL tr tnr

where τr and τnr are the radiative and nonradiative recombination times, respectively. This stretched
exponential behavior of PL decay is generally expected in disordered systems in which more than
one pathway of carrier recombination exists. Different models have been proposed to explain this
behavior. Some of these models suggest that this distribution of different times is a result of the
interaction between the silicon nanocrystals in an ensemble of such crystals, which allows exciton
migration from smaller to larger ones [78–82]. Other models suggest that migration of carriers is
not necessary for the observed behavior of the PL decay. Instead, the excitons contributing to the PL
are strongly localized, and recombination dynamics, which is a competition between radiative and
nonradiative recombination of the localized excitons, is characterized by different decay times due
to the dispersion in the sizes of the silicon nanocrystals in an ensemble. That is to say, the stretched
exponential behavior of the PL decay is not due to the interaction between the silicon nanocrystals
in an ensemble [83]. Specifically, this behavior is a result of the indirect nature of the energy band-
gap of the silicon nanocrystals [84,85].
PL recombination times of up to a few microseconds have been reported for silicon nanocrys-
tals. These long recombination times are due to the indirect nature of the energy bandgap of the
Silicon Nanocrystals 201

A
B
C
D
100
PL decay times (μs)

1
Normalized PL intensity

0.1

10

0.01
0 100 200 300 400 500
Time (μs)
50 100 150 200 250 300
Temperature (K)

FIGURE 5.6 PL decay times extracted from stretched exponential fitting (insert) as a function of tempera-
ture. Samples A and B are porous silicon layers with porosities 85% and 62%, respectively. Sample C is
a heavily oxidized porous silicon layer containing silicon nanocrystals within a silicon dioxide matrix.
Sample D contains silicon nanocrystals embedded in a silicon dioxide matrix, grown by LPCVD deposition
of silicon and subsequent thermal oxidation (its structure is shown in Figure 5.2b). (Reproduced with permis-
sion from Gardelis, S., Nassiopoulou, A.G., Vouroutzis, N., and Frangis, N., Effect of exciton migration on
the light emission properties in silicon nanocrystal ensembles, J. Appl. Phys., 105, 113509. © 2009, American
Institute of Physics.)

silicon nanocrystals. It has been observed that the recombination time increases with decreasing
temperature (see, e.g., Figure 5.6). Also, the PL intensity increases with decreasing tempera-
ture. This is due to the competition between radiative and nonradiative recombination. At higher
temperatures, nonradiative transitions come into play, resulting in the quenching of PL and in
a decrease of the measured recombination time according to Equation 5.2. Another interesting
observation is that the PL intensity increases at lower temperatures up to a certain temperature
and then decreases, whereas the recombination time increases. This effect was attributed to the
fact that, due to quantum confinement, the orbital degeneracy of the bulk Si valence band is
lifted by quantum confinement, resulting in splitting of a localized exciton state by the exchange
interaction into a spin singlet and a spin triplet, with the triplet lower in energy. At low enough
temperatures (e.g., 2 K), the generated excitons should be in the lower energy triplet state. Optical
transitions from the triplet state are forbidden due to spin selection rules. On the contrary, transi-
tions from the singlet state are allowed [12,86–88]. Thus, the radiative lifetime corresponding
to transitions form triplet state should be infinite. However, due to spin–orbit scattering, there
is some mixing between triplet and singlet states, and thus the transition from the triplet state
becomes weakly allowed. At higher temperatures, singlet states become thermally populated and
transitions take place from these states. Thus, the radiative lifetime becomes smaller, and the PL
intensity increases with temperature. At some temperature, the PL intensity reaches its maximum
value and at higher temperatures it decreases again due to the nonradiative transitions, as we have
already mentioned.
202 Semiconductor Nanocrystals and Metal Nanoparticles

Recombination lifetime has also a spectral dependence. It increases with increasing wavelength
or decreasing detection energy. PL at lower wavelengths originates from the smaller nanocrystals
within the ensemble, and thus the oscillator strength of the recombination transitions is larger. This
is a consequence of the breakdown of the k-conservation rule and explains the faster recombination
times at the lower wavelengths in the PL spectrum [89].
In an ensemble of silicon nanocrystals, recombination rates can be affected by the degree of
electrical interconnection between the silicon nanocrystals. For example, in mesoporous and nano-
porous silicon films, one measures recombination rates faster than in ensembles of silicon nanocrys-
tals embedded in a silicon dioxide matrix. For example, in the first case, excitons can migrate and
recombine away from their generation locations, whereas in the latter case the excitons are more
localized. This can be understood if we introduce in the measured recombination rates a rate rm,
which is due to this exciton migration. Thus, the recombination rate rPL of equation becomes
rPL = rr + rnr + rm [5,78–82]. Exciton migration in porous silicon results in a red shift of the lumi-
nescence with increasing temperature, which is larger than that expected from the temperature
reduction of the energy bandgap of the nanocrystals. This occurs because at higher temperatures
excitons are less localized and can recombine via lower energy states, that is, within larger silicon
nanocrystals, than those in which they initially were generated. On the contrary, in the case of
silicon nanocrystals embedded in the silicon dioxide, the excitons are more localized and there is
almost no shift of the PL with the temperature [5].

5.5.4 Other Issues Regarding Light Emission

5.5.4.1 Optical Gain
Optical gain is necessary for laser applications. Pavesi et al. claimed that they observed optical
gain in silicon nanocrystals embedded in a silicon dioxide matrix [90]. They explained this effect
by introducing a three-level model. Specifically, they considered that the generated carriers, before
their recombination, populate intermediate energy states, within the energy bandgap of the silicon
nanocrystals, associated with interface states of the silicon nanocrystals with the surrounding sili-
con nanocrystals. This created a population inversion, which resulted in stimulated emission and
thus optical gain. Other researchers have also reported optical gain in similar systems of silicon
nanocrystals [91,92].
However, when population inversion is needed to achieve lasing, high excitation is needed. Under
these conditions, fast nonradiative processes, such as Auger recombination (participation of three
particles in nonradiative processes or free carrier absorption), come into play. These processes
­provide loss mechanisms, as they deplete the excited population. Thus, gain competes with these
loss mechanisms. Due to these competing loss mechanisms, optical gain from silicon nanocrystals
still remains an open question. Suppression of these two loss mechanisms would result in the
exploitation of gain for lasing applications [93].

5.5.4.2 Multiple Exciton Generation (MEG) in Silicon Nanocrystals

MEG was first demonstrated in colloidal silicon nanocrystals. It was found that the threshold for
MEG in silicon nanocrystals with a size of 9.5 nm, which corresponds in an energy bandgap of
1.2 eV, is 2.4 times the energy bandgap of the silicon nanocrystal. Also, a quantum yield of 2.6 per
absorbed photon was found in silicon at 3.4 times the energy bandgap [94]. Later, the effect was dem-
onstrated in silicon nanocrystals embedded in silicon dioxide. The multiple excitons are generated by
the absorption of one photon of energy larger than twice the energy bandgap of the nanocrystals, not
in the same nanocrystal but in adjacent ones, so that multiple excitons do not undergo nonradiative
Auger recombination. This effect is very important for solar cell applications, as the generation of
multiple excitons could improve the conversion efficiencies of silicon-based solar cells [95].
Silicon Nanocrystals 203

5.5.4.3 Silicon Nanocrystals as Sensitizers

Silicon nanostructures have been used as optical amplifiers in order to enhance erbium lumines-
cence (1.54 μm), which is used in optical communications due to the low attenuation of this emis-
sion in long-haul optical fibers. Also, this optical amplification provided by the silicon nanocrystals
is promising for the development of erbium lasers. The first efforts were made in porous silicon
[96,97]. For example, an order of magnitude more intense luminescence from erbium was observed
in porous silicon compared to that of erbium in crystalline bulk silicon. This was one of the first
observations that silicon nanostructures could provide amplification of the erbium luminescence
[96]. More recently, erbium-doped silicon nanocrystals showed remarkable enhancement of the
erbium luminescence [98–102]. This enhancement originates from the higher absorption cross-
section of silicon nanocrystals in combination with the efficient energy transfer to the adjacent
erbium ions.

5.5.5 Electroluminescence
The significant result of light emission from silicon nanostructures initiated significant effort to
demonstrate the effect electrically, that is, with electrical injection of electrons and holes within the
silicon nanostructures. Electroluminescence was first demonstrated by a redox reaction in a solution
during the anodic oxidation of porous silicon and then in various types of solid-state devices, such as
Schottky [103–109], p–n, or p–i–n diodes [110–116]. The interest and the effort to fabricate efficient
electroluminescent devices of silicon nanostructures passed from porous silicon to other types of
silicon nanostructures. A demonstration of electroluminescence from silicon nanostructures other
than porous silicon was reported in 1996 for silicon nanopillars (nanowires) fabricated in a vertical
direction on a crystalline silicon wafer. The device was fabricated as a Schottky diode [117]. The main
interest now for silicon nanocrystal–based electroluminescent devices has shifted to silicon nano-
crystals embedded in dielectric matrices. Although these structures show very stable and efficient
photoluminescence, the presence of the electrically insulating dielectric matrix is an obstacle for
efficient carrier injection into the silicon nanocrystals, making electroluminescence very inefficient.
Silicon nitride or silicon carbide has been suggested as a dielectric matrix alternative to silicon
dioxide because of their smaller energy bandgap compared to that of silicon dioxide. This would
ease tunneling or injection of the carriers into the silicon nanocrystals and consequently improve
the efficiency of the electroluminescent device. Also, thinner dielectric matrices would make car-
rier injection into the silicon nanocrystals easy [118]. Earlier works showed electroluminescence in
MOS devices in which the silicon nanocrystals were embedded in the silicon dioxide of the gate.
These devices work under high voltages, higher than 5 V. Under these voltages, unipolar field-
enhanced Fowler–Nordheim tunneling is the main mechanism of carrier injection. Under these
conditions, charge transport occurs in the oxide via hot electrons, whereas the electron–hole pairs
are generated within the silicon nanocrystals by impact ionization [119–127]. Quantum efficiencies
of these devices were generally small, 10 −3% to 10 −4%. A record high of 0.16% under pulsed opera-
tion has been reported [112]. A previous record high was 0.01% [114]. Figure 5.7 shows an example
of electroluminescence from silicon nanocrystals embedded in the silicon dioxide of the gate of an
MOS device [122].
Hybrid devices with silicon nanocrystals and organic polymers have shown efficient white-light
electroluminescence or tunable electroluminescence depending on the size of the silicon nano-
crystals. External quantum efficiencies as high as 1.1% or even 3.6% have been reported [128,129].
A significant drawback of the silicon nanocrystal–based electroluminescent devices, apart from
their low efficiencies, is that they suffer from long emission decays, which limit the modulation
frequency of the electroluminescence only to applications for light-emitting display diodes, thus
making them unsuitable for laser applications.
204 Semiconductor Nanocrystals and Metal Nanoparticles

Transparent Al

Si nanocrystals
SiO2

SiO2

p-type Si (100)

(a) Al ohmic back contact

Sample C

+
PL 15 mW, 457.9 nm Ar
EL 6.7 V, 2.7 mA
0 ×5 EL 5 V, 1.4 mA
1

Sample B

EL 5.5 V, 5.8 mA
EL 9 V, 13.3 mA
+
×5 PL 15 mW, 457.9 nm Ar
0
1

Sample A
EL 5.5 V, 3.7 mA
EL 9.5 V, 9.1 mA
0 ×5 PL 15 mW, 457.9 nm Ar
+

500 550 600 650 700 750 800

(b) Wavelength (nm)

FIGURE 5.7 (a) Light-emitting diode based on silicon nanocrystals embedded in a silicon dioxide matrix.
Hot electrons are injected into the silicon nanocrystals generating electron–hole pairs by impact ionization,
which eventually recombine giving rise to the observed electroluminescence (EL). (b) Comparison of PL and
EL from samples containing different sizes of silicon nanocrystals (A = 2 nm, B = 1.5 nm, and C = 1 nm).
(Reproduced with permission from Photopoulos, P. and Nassiopoulou, A.G., Room- and low-temperature volt-
age tunable electroluminescence from a single layer of silicon quantum dots in between two thin SiO2 layers,
Appl. Phys. Lett., 77, 1816. © 2000, American Institute of Physics.)
Silicon Nanocrystals 205

5.6 CHARGING EFFECTS IN SILICON

NANOCRYSTALS—NONVOLATILE MEMORIES
Silicon nanocrystals show charging effects when their sizes are of the order of a few nanometers
[6–9]. Specifically, if the silicon nanocrystal is charged, extra energy is necessary to inject an extra
electron into it because of Coulomb interactions. This extra energy is called the “charging energy”
of the nanocrystal. This effect can be exploited in nonvolatile memories. The use of embedded flash
nonvolatile memories in microcontrollers, which have found their use in consumer, automotive, and
other industrial markets, is the driving force for more scalable and thus less energy-consuming flash
memories. However, the existing technology of flash memories has its limitations in scalability and
energy consumption. For example, floating memories, which dominate the industry, suffer from
scalability as the tunnel oxide cannot be made thinner than 10 nm, as stress-induced leakage cur-
rents come into play and reduce the retention times of the memory. Other types of memories, such
as resistive, ferroelectric RAM, or phase change memories, are not as mature as the silicon-based
ones. Silicon nanocrystal–based nonvolatile memories can solve these problems of flash memo-
ries. The silicon nanocrystal nonvolatile memory is fabricated as a field-effect transistor with its
gate formed by silicon nanocrystals embedded in silicon dioxide. This memory is better than the
standard floating-gate flash memory as the gate is not a continuous nanocrystalline silicon layer
but consists of discrete nanocrystals that can be charged separately. In the standard flash memory,
the continuity of the gate results in leakages, which destroy the memory effect. In the case of the
silicon nanocrystal, if any defect at the silicon nanocrystal–silicon dioxide interface causes any
leakage of the charge out of the nanocrystal, this causes only a minimal shift in the threshold volt-
age without destroying the nonvolatile behavior of the memory. Also, this allows scaling the bottom
tunnel oxide by 40%–50% compared to a floating gate and consequently for lower energy consump-
tion [11]. The concept of using discrete silicon nanocrystals for storing charge was introduced in
1996 [130]. Since then, different approaches of a single layer of silicon nanocrystals embedded in
the gate oxide of a field-effect transistor have been implemented. For example, LPCVD [4,28,29]
and PECVD [131] growth of nanocrystalline or amorphous silicon and subsequent thermal oxida-
tion results in the formation of a single layer of silicon nanocrystals embedded in a silicon dioxide
matrix. Another method is by implantation of silicon ions in the gate oxide of a field-effect transistor
at low silicon ion energies and the subsequent annealing, which results in the formation of a single
layer of silicon nanocrystals in the gate oxide [10,132]. Today, microcontrollers with embedded
silicon nanocrystal nonvolatile memories are commercially available [11,133].

5.7 APPLICATIONS OF SILICON NANOCRYSTALS

Silicon has been the dominant material in the microelectronics industry for over half a century now.
Advancements in silicon CMOS (complementary metal–oxide–semiconductor) technology, integra-
tion, and miniaturization have enabled the fabrication of complex, fast, and low-power-consumption
devices. However, silicon as an indirect bandgap semiconductor cannot be used for optical ampli-
fiers and light sources. It would be ideal if such devices could be fabricated and integrated with the
existing CMOS technology in integrated circuits where all components are made out of silicon, that
is, transistors, optical amplifiers, light sources, and photodetectors. Efficient light emission from
silicon nanostructures was the driving force to investigate this material as a potential light source
that could be integrated with the existing CMOS technology. Unfortunately, the slow switching and
the low electroluminescence efficiency of the silicon nanocrystals make this material unsuitable
for laser applications. On the other hand, the observed optical gain that is achieved at high pump-
ing powers must compete with the nonradiative Auger recombination, and thus optical gain is not
sufficient to support lasing. Nevertheless, efficient light-emitting diodes for display purposes could
be produced using silicon nanocrystals and could replace other toxic materials such as Cd or Pb
[128,129]. However, silicon nanocrystals could be used as sensitizers for erbium light emission,
206 Semiconductor Nanocrystals and Metal Nanoparticles

resulting in a considerable enhancement of erbium emission, which could be used for lasing appli-
cations or optical amplifiers [96,101]. Thus, silicon nanocrystals indirectly can be very useful for
realizing a silicon-based laser source or an optical amplifier that could be integrated.
The recently observed 100% internal quantum efficiency of suitably passivated silicon nano-
crystals shows that silicon nanocrystals, which are biocompatible, can replace the toxic Cd- and
Pb-based quantum dots in biomedical imaging and diagnostic applications [65]. Another interesting
application of such highly efficient light-emitting silicon nanocrystals could be their use in silicon-
based solar cell application for downshifting [134,135]. As silicon is an indirect bandgap material,
the photons of the solar spectrum with energies much higher than the energy bandgap of silicon
can be lost as heat before being collected by the electrodes of the solar cell. By depositing a suitable
layer of light-emitting silicon nanocrystals on the surface of the solar cell, the high-energy photons
of the solar spectrum can be converted efficiently into photons of lower energies, which could be
absorbed from the silicon absorber underneath and converted into electricity more efficiently, thus
improving the conversion efficiency of the solar cell. This effect of downshifting could be enhanced
even further by the use of metal nanoparticles deposited in the vicinity of the silicon nanocrystals,
which can couple with the plasmon modes of the metal nanoparticles.
Silicon nanocrystals, as a result of their energy bandgap tailoring, could be used in tandem solar
cell configurations [136]. These cells can be fabricated as multiple p–i–n junctions in a tandem
configuration, where i is a layer of silicon nanocrystals of a certain size that are embedded in a
dielectric matrix. Such solar cells are expected to show higher conversion efficiencies compared to
the conventional silicon-based solar cells. Apart from silicon dioxide, other dielectric matrices such
as silicon nitride and silicon carbide have been suggested, as these dielectrics form smaller barriers
with the embedded silicon nanocrystals, thus making it easier for the carriers generated within the
silicon nanocrystals to overcome the barrier and be collected by the electrodes of the cell [137].
Much effort has been made to produce suitable solar cell absorber layers of silicon nanocrystals
embedded in a dielectric matrix [138–140]. In another approach, a hybrid silicon nanocrystal–organic
solar cell has been realized [141]. Also a Schottky-type solar cell containing a layer of silicon nano-
crystals has been demonstrated. The conversion efficiency, though, was very low, only 0.02% [142].
All these approaches have resulted in very small conversion efficiencies compared to conventional
silicon-based cells, although silicon nanocrystals show higher absorbance than silicon [29,143].
One drawback of using layers of nanocrystals is the difficulty to extract the generated carriers out
of the dots and eventually to collect them by the electrodes of the cell. The cell is resistive due to
the separation of the silicon nanocrystals. One way to reduce this effect is by embedding higher
densities of silicon nanocrystals that are very close to each other, thus increasing conductivity and
absorption. Finally, multiple exciton generation in silicon nanocrystals, as already described previ-
ously, could be exploited in silicon nanocrystal–based solar cells to improve conversion efficiencies
in comparison with bulk silicon solar cells.

5.8 CONCLUSIONS
In conclusion, silicon nanocrystals show unique optical and electronic properties due to quantum
(for sizes less than the electron wavelength of ~8 nm in silicon) and spatial confinement of carri-
ers (for sizes less than the Bohr radius of excitons in bulk silicon of ~5 nm). The result of quan-
tum confinement is the increase of the energy bandgap of the silicon nanocrystals as their size
decreases. On the other hand, the result of the spatial confinement is the increase of the absorption
and radiative emission oscillator strength due to the increasing overlap of the electron and hole
wave functions of the generated excitons within the silicon nanocrystals after the absorption of
photons with energies larger than the energy bandgap. Both these effects result in the efficient
light emission from silicon nanocrystals for sizes less than 5 nm, which shifts to higher ener-
gies with decreasing size. With suitable surface passivation, the internal quantum efficiency can
reach almost 100%, although silicon nanocrystals retain the indirect nature of the energy bandgap.
Silicon Nanocrystals 207

However, the indirectness of the bandgap results in slow recombination mechanisms that govern
the light emission, making silicon nanocrystals unsuitable for fast-switching optical applications.
Also, the optical gain, which is necessary for laser applications, and multiple exciton generation,
which have been observed in silicon nanocrystals, compete with nonradiative Auger recombi-
nation mechanisms, thus making these effects very inefficient. However, light emission can be
exploited in sensing and particularly in biological and medical applications, as silicon nanocrystals
are compatible with living organisms and can replace the toxic Cd- and Pb-based quantum dots in
many bioapplications. Also, they can be used for downshifting in solar cell applications. Finally,
spatial confinement results in charging effects within the silicon nanocrystals, thus making them
useful in nonvolatile memory applications.

REFERENCES
1. L. T. Canham, Silicon quantum wire array fabrication by electrochemical and chemical dissolution of
wafers, Appl. Phys. Lett. 57, 1046 (1990).
2. A. D. Yoffe, Low-dimensional systems: Quantum size effects and electronic properties of semiconduc-
tor microcrystallites (zero-dimensional systems) and some quasi-two-dimensional systems, Adv. Phys.
42, 173 (1993).
3. M. Zacharias, J. Heitmann, R. Scholz, U. Kahler, M. Schmidt, and J. Bläsing, Size-controlled highly
luminescent silicon nanocrystals: A SiO/SiO2 superlattice approach, Appl. Phys. Lett. 80, 661 (2002).
4. P. Photopoulos, A. G. Nassiopoulou, D. N. Kouvatsos, and A. Travlos, Photoluminescence from nano-
crystalline silicon in Si/SiO2 superlattices, Appl. Phys. Lett. 76, 3588 (2000).
5. S. Gardelis, A. G. Nassiopoulou, N. Vouroutzis, and N. Frangis, Effect of exciton migration on the light
emission properties in silicon nanocrystal ensembles, J. Appl. Phys. 105, 113509 (2009).
6. A. G. Nassiopoulou, Silicon nanocrystals in SiO2 thin layers, in Encyclopedia of Nanosciences and
Nanotechnology, H. S. Nalwa (ed.), Vol. 9, pp. 793–813 (2004).
7. I. Balberg, E. Savir, J. Jedrzejewski, A. G. Nassiopoulou, and S. Gardelis, Fundamental transport
processes in ensembles of silicon quantum dots, Phys. Rev. B 75, 235329 (2007).
8. P. Manousiadis, S. Gardelis, and A. G. Nassiopoulou, Lateral electronic transport in 2D arrays of oxi-
dized Si nanocrystals on quartz: Coulomb blockade effect and role of hydrogen passivation, J. Appl.
Phys. 109, 083718 (2011).
9. D. N. Kouvatsos, V. Ioannou-Sougleridis, and A. G. Nassiopoulou, Charging effects in silicon nano-
crystals within SiO2 layers, fabricated by chemical vapor deposition, oxidation and annealing, Appl.
Phys. Lett. 82(3), 397 (2003).
10. P. Dimitrakis, E. Kapetanakis, D. Tsoukalas, D. Skarlatos, C. Bonafos, G. Ben Asssayag, A. Claverie
et al., Silicon nanocrystal memory devices obtained by ultra-low-energy ion-beam-synthesis, Solid
State Electron. 48, 1511 (2004).
11. J. A. Yater, Implementation of Si nanocrystals in non-volatile memory devices, Phys. Status Solidi A
210(8), 1505–1511 (2013).
12. A. G. Cullis, L. T. Canham, and P. D. J. Calcott, The structural and luminescence properties of porous
silicon, J. Appl. Phys. 82, 909 (1997).
13. S. Gardelis, U. Bangert, A. J. Harvey, and B. Hamilton, Double crystal x-ray diffraction, electron dif-
fraction and high resolution electron microscopy of luminescent porous silicon, J. Electrochem. Soc.
142, 2094 (1995).
14. É. Váazsonyi, E. Szilágyi, P. Petrik, Z. E. Horvátha, T. Lohner, M. Fried, and G. Jalsovszky, Porous
silicon formation by stain etching, Thin Solid Films 388, 295 (2001).
15. C. J. M. Eukel, J. Branebjerg, M. Elwenspoek, and F. C. M. Van De Pol, A new technology for microma-
chining of silicon: Dopant selective HF anodic etching for the realization of low-doped monocrystalline
silicon structures, IEEE Electron Dev. Lett. 11, 588 (1990).
16. M. T. Kelly, J. K. M. Chun, and A. B. Bocarsly, High efficiency chemical etchant for the formation of
luminescent porous silicon, Appl. Phys. Lett. 64, 1693 (1994).
17. C. A. Caras, J. M. Reynard, and F. V. Bright, An in-depth study linking the infrared spectroscopy and
photoluminescence of porous silicon during ambient hydrogen peroxide oxidation, Appl. Spectrosc.,
67, 570 (2013).
18. M. Steinert, J. Acker, A. Henssge, and K. Wetzig, Experimental studies on the mechanism of wet
chemical etching of silicon in HF/HNO3 mixtures, J. Electrochem. Soc. 152, C843 (2005).
208 Semiconductor Nanocrystals and Metal Nanoparticles

19. M. A. Hory, R. Hkrino, M. Ligeon, F. Muller, F. Gaspard, I. Mihalcescu, and J. C. Vial, Fourier trans-
form IR monitoring of porous silicon passivation during post-treatments such as anodic oxidation and
contact with organic solvents, Thin Solid Films 255, 200 (1995).
20. V. Petrova‐Koch, T. Muschik, A. Kux, B. K. Meyer, F. Koch, and V. Lehmann, Rapid‐thermal‐oxidized
porous Si—The superior photoluminescent Si, Appl. Phys. Lett. 61, 943 (1992).
21. S. Gardelis and B. Hamilton, The effect of surface modification on the luminescence of porous silicon,
J. Appl. Phys. 76, 5327 (1994).
22. L. Debarge, J. P. Stoquert, A Slaoui, L. Stalmans, and J. Poortmans, Rapid thermal oxidation of porous
silicon for surface passivation, Mater. Sci. Semicond. Process. 1, 281 (1998).
23. B. Gelloz and N. Koshida, Highly enhanced photoluminescence of as-anodized and electrochemically
oxidized nanocrystalline p-type porous silicon treated by high-pressure water vapor annealing, Thin
Solid Films 508, 406 (2006).
24. Y. Ishikawa, A. V. Vasin, J. Salonen, S. Muto, V. S. Lysenko, A. N. Nazarov, N. Shibata, and V.-P. Lehto,
Color control of white photoluminescence from carbon-incorporated silicon oxide, J. Appl. Phys. 104,
083522 (2008).
25. T. Shimizu-Iwayama, N. Kurumado, D. E. Hole, and P. D. Townsend, Optical properties of silicon
nanoclusters fabricated by ion implantation, J. Appl. Phys. 83, 6018 (1998).
26. H. Z. Song and X. M. Bao, Visible photoluminescence from silicon-ion-implanted SiO2 film and its
multiple mechanisms, Phys. Rev. B 55, 6988 (1997).
27. T. Shimizu‐Iwayama, S. Nakao, and K. Saitoh, Visible photoluminescence in Si+‐implanted thermal
oxide films on crystalline Si, Appl. Phys. Lett. 65, 1814 (1994).
28. Ch. B. Lioutas, N. Vouroutzis, I. Tsiaoussis, N. Frangis, S. Gardelis, and A. G. Nassiopoulou, Columnar
growth of ultra-thin nanocrystalline Si films on quartz by low pressure chemical vapor deposition:
Accurate control of vertical size, Phys. Status Solidi A 205, 2615 (2008).
29. S. Gardelis, A. G. Nassiopoulou, P. Manousiadis, A. Gkanatsiou, N. Frangis, and Ch. B. Lioutas,
Structural and optical characterization of two-dimensional arrays of Si nanocrystals embedded in SiO2
for photovoltaic applications, J. Appl. Phys. 111, 083536 (2012).
30. L. Tsybeskov, K. D. Hirschman, S. P. Duttagupta, M. Zacharias, P. M. Fauchet, J. P. McCaffrey, and
D. J. Lockwood, Nanocrystalline-silicon superlattice produced by controlled recrystallization, Appl.
Phys. Lett. 72, 43 (1998).
31. L.-Y. Chen, W.-H. Chen, and F. C.-N. Hong, Visible electroluminescence from silicon nanocrystals
embedded in amorphous silicon nitride matrix, Appl. Phys. Lett. 86, 193506 (2005).
32. C. Liu, C. Li, A. Ji, L. Ma, Y. Wang, and Z. Cao, Intense blue photoluminescence from Si-in-SiNx thin
film with high-density nanoparticles, Nanotechnology 16, 940 (2005).
33. M. V. Wolkin, J. Jorne, P. M. Fauchet, G. Allan, and C. Delerue, Electronic states and luminescence in
porous silicon quantum dots: The role of oxygen, Phys. Rev. Lett. 82, 197 (1999).
34. F. Iacona, G. Franzò, and C. Spinella, Correlation between luminescence and structural properties of
Si nanocrystals, J. Appl. Phys. 87, 1295 (2000).
35. D. S. Chao and J. H. Liang, Annealing temperature dependence of photoluminescent characteristics of
silicon nanocrystals embedded in silicon-rich silicon nitride films grown by PECVD, Nucl. Instrum.
Methods Phys. Res. B: Beam Interact. Mater. Atoms 307, 344 (2013).
36. L. Dal Negro, J. H. Yi, L. C. Kimerling, S. Hamel, A. Williamson, and G. Galli, Light emission from
silicon-rich nitride nanostructures, Appl. Phys. Lett. 88, 183103 (2006).
37. T.-Y. Kim, N.-M. Park, K.-H. Kim, G. Y. Sung, Y.-W. Ok, T.-Y. Seong, and C.-J. Choi, Quantum con-
finement effect of silicon nanocrystals in situ grown in silicon nitride films, Appl. Phys. Lett. 85, 5355
(2004).
38. V. Em. Vamvakas, N. Vourdas, and S. Gardelis, Optical characterization of Si-rich silicon nitride films
prepared by low pressure chemical vapor deposition, Microelectron. Reliab. 47, 794 (2007).
39. V. Em. Vamvakas and S. Gardelis, FTIR characterization of light emitting Si-rich nitride films prepared
by low pressure chemical vapor deposition, Surf. Coat. Technol. 201, 9359 (2007).
40. E. Werwa, A. A. Seraphin, L. A. Chiu, C. Zhou, and K. D. Kolenbrander, Synthesis and processing
of silicon nanocrystallites using a pulsed laser ablation supersonic expansion method, Appl. Phys. Lett.
64, 1821 (1994).
41. T. A. Burr, A. A. Seraphin, E. Werwa, and K. D. Kolenbrander, Carrier transport in thin films of silicon
nanoparticles, Phys. Rev. B 56, 4818 (1997).
42. B. F. P. McVey and R. D. Tilley, Solution synthesis, optical properties, and bioimaging applications of
silicon nanocrystals, Acc. Chem. Res. 47, 3045 (2014).
Silicon Nanocrystals 209

43. J. H. Warner, A. Hoshino, K. Yamamoto, and R. D. Tilley, Water-soluble photoluminescent silicon

quantum dots, Angew. Chem., Int. Ed. 44, 4550 (2005).
44. J. P. Wilcoxon, G. A. Samara, and P. N. Provencio, Optical and electronic properties of Si nanoclusters
synthesized in inverse micelles, Phys. Rev. B 60, 2704 (1999).
45. C.-S. Yang, R. A. Bley, S. M. Kauzlarich, H. W. H. Lee, and G. R. Delgado, Synthesis of alkyl-­terminated
silicon nanoclusters by a solution route, J. Am. Chem. Soc. 121, 5191 (1999).
46. M. P. Singh, T. M. Atkins, E. Muthuswamy, S. Kamali, C. Tu, A. Y. Louie, and S. M. Kauzlarich,
Development of iron-doped silicon nanoparticles as bimodal imaging agents, ACS Nano 6, 5596 (2012).
47. C. Tu, X. Ma, P. Pantazis, S. M. Kauzlarich, and A. Y. Louie, Paramagnetic, silicon quantum dots for
magnetic resonance and two-photon imaging of macrophages, J. Am. Chem. Soc. 132, 2016 (2010).
48. C. M. Hessel, E. J. Henderson, and J. G. C. Veinot, Hydrogen silsesquioxane: A molecular precursor for
nanocrystalline Si−SiO2 composites and freestanding hydride-surface-terminated silicon nanoparticles,
Chem. Mater. 18, 6139 (2006).
49. E. J. Henderson, J. A. Kelly, and J. G. C. Veinot, Influence of HSiO1.5 sol−gel polymer structure and
composition on the size and luminescent properties of silicon nanocrystals, Chem. Mater. 21, 5426
(2009).
50. R. K. Baldwin, J. Zou, K. A. Pettigrew, G. J. Yeagle, R. D. Britt, and S. M. Kauzlarich, The preparation
of a phosphorus doped silicon film from phosphorus containing silicon nanoparticles, Chem. Commun.
6, 658–660 (2006).
51. S. Gardelis, J. S. Rimmer, P. Dawson, B. Hamilton, R. A. Kubiak, T. E. Whall, and E. H. C. Parker,
Evidence for quantum confinement in the photoluminescence of porous Si and SiGe, Appl. Phys. Lett.
59, 2118 (1991).
52. S. M. Prokes, O. J. Glembocki, V. M. Bermudez, R. Kaplan, L. E. Friedersdorf, and P. C. Searson,
SiHx excitation: An alternate mechanism for porous Si photoluminescence, Phys. Rev. B 45, 13788(R)
(1992).
53. H. D. Fuchs, M. Stutzmann, M. S. Brandt, M. Rosenbauer, J. Weber, A. Breitschwerdt, P. Deák, and
M. Cardona, Porous silicon and siloxene: Vibrational and structural properties, Phys. Rev. B 48, 8172
(1993).
54. D. J. Lockwood and A. G. Wang, Quantum confinement induced photoluminescence in porous silicon,
Solid State Commun. 94, 905 (1995).
55. C. Delerue, G. Allan, and M. Lannoo, Theoretical aspects of the luminescence of porous silicon, Phys.
Rev. B 48, 11024 (1993).
56. H. Mizuno, H. Koyama, and N. Koshida, Oxide‐free blue photoluminescence from photochemically
etched porous silicon, Appl. Phys. Lett. 69, 3779 (1996).
57. J. M. Lauerhaas and M. J. Sailor, Chemical modification of the photoluminescence quenching of porous
silicon, Science 261, 1567 (1993).
58. M. T. Kelly, J. K. Chun, and A. B. Bocarsly, General Brönsted acid behavior of porous silicon: A mecha-
nistic evaluation of proton-gated quenching of photoemission from oxide-coated porous silicon, J. Phys.
Chem. B 101, 2702 (1997).
59. J. L. Coffer, S. C. Lilley, R. A. Martin, and L. Files-Sesler, Surface reactivity of luminescent porous
silicon, J. Appl. Phys. 74, 2094 (1993).
60. J. M. Rehm, G. L. McLendon, and P. M. Fauchet, Conduction and valence band edges of porous silicon
determined by electron transfer, J. Am. Chem. Soc. 118, 4490 (1996).
61. I. N. Germanenko, S. Li, and M. S. El-Shall, Decay dynamics and quenching of photoluminescence
from silicon nanocrystals by aromatic nitro compounds, J. Phys. Chem. B 105, 59 (2001).
62. D. Jurbergs, E. Rogojina, L. Mangolini, and U. Kortshagen, Silicon nanocrystals with ensemble quan-
tum yields exceeding 60%, Appl. Phys. Lett. 88, 233116 (2006).
63. R. J. Anthony, D. J. Rowe, M. Stein, J. Yang, and U. Kortshagen, Routes to achieving high quantum yield
luminescence from gas-phase-produced silicon nanocrystals, Adv. Funct. Mater. 21, 4042 (2011).
64. M. L. Mastronardi, F. Maier-Flaig, D. Faulkner, E. J. Henderson, C. Kübel, U. Lemmer, and G. A. Ozin,
Size-dependent absolute quantum yields for size-separated colloidally-stable silicon nanocrystals, Nano
Lett. 12, 337 (2012).
65. F. Sangghaleh, I. Sychugov, Z. Yang, J. G. C. Veinot, and J. Linnros, Near-unity internal quantum effi-
ciency of luminescent silicon nanocrystals with ligand passivation, ACS Nano 9(7), 7097–7104 (2015).
66. D. J. Lockwood and L. Pavesi, Silicon fundamentals for photonic applications, in Silicon Photonics,
Topics in Applied Physics, Vol. 94, p. 13, L. Pavesi and D. J. Lockwood (eds.), Springer-Verlag, Berlin,
Germany (2004).
210 Semiconductor Nanocrystals and Metal Nanoparticles

67. D. Kovalev, H. Heckler, M. Ben-Chorin, G. Polisski, M. Schwartzkopff, and F. Koch, Breakdown of the
k-conservation rule in Si nanocrystals, Phys. Rev. Lett. 81, 2803 (1998).
68. S. Godefroo, M. Hayne, M. Jivanescu, A. Stesmans, M. Zacharias, O. I. Lebedev, G. Van Tendeloo, and
V. V. Moshchalkov, Classification and control of the origin of photoluminescence from Si nanocrystals,
Nat. Nanotechnol. 2, 486–489 (2007).
69. L. H. Lie, M. Duerdin, E. M. Tuite, A. Houlton, and B. R. Horrocks, Preparation and characterization of
luminescent alkylated silicon quantum dots, J. Electroanal. Chem. 538, 183 (2002).
70. L. H. Lie, S. N. Patole, A. R. Pike, L. C. Ryder, B. A. Connolly, A. D. Ward, E. M. Tuite, A. Houlton,
and B. R. Horrocks, Immobilisation and synthesis of DNA on Si(111), nanocrystalline porous silicon
and silicon nanoparticles, Faraday Discuss. 125, 235 (2004).
71. Y. Chao, A. Houlton, B. R. Horrocks, M. R. C. Hunt, N. R. J. Poolton, J. Yang, and L. Šillera, Optical
luminescence from alkyl-passivated Si nanocrystals under vacuum ultraviolet excitation: Origin and
temperature dependence of the blue and orange emissions, Appl. Phys. Lett. 88, 263119 (2006).
72. M. Luppi and S. Ossicini, Ab initio study on oxidized silicon clusters and silicon nanocrystals embed-
ded in SiO2: Beyond the quantum confinement effect, Phys. Rev. B 71, 035340 (2005).
73. L. E. Ramos, J. Furthmüller, and F. Bechstedt, Effect of backbond oxidation on silicon nanocrystallites,
Phys. Rev. B 70, 033311 (2004).
74. I. Vasiliev, J. R. Chelikowsky, and R. M. Martin, Surface oxidation effects on the optical properties of
silicon nanocrystals, Phys. Rev. B 65, 121302(R) (2002).
75. G. Hadjisavvas and P. C. Kelires, Structure and energetics of Si nanocrystals embedded in a-SiO2, Phys.
Rev. Lett. 93, 226104 (2004).
76. L. Pavesi and M. Ceschini, Stretched-exponential decay of the luminescence in porous silicon, Phys.
Rev. B 48, 17625(R) (1993).
77. P. J. Ventura, M. C. do Carmo, and K. P. O’Donnell, Excitation dynamics of luminescence from porous
silicon, J. Appl. Phys. 77, 323 (1995).
78. L. Pavesi, Influence of dispersive exciton motion on the recombination dynamics in porous silicon,
J. Appl. Phys. 80, 216 (1996).
79. F. Priolo, G. Franzò, D. Pacifici, V. Vinciguerra, F. Iacona, and A. Irrera, Role of the energy transfer
in the optical properties of undoped and Er-doped interacting Si nanocrystals, J. Appl. Phys. 89, 264
(2001).
80. J. Linnros, N. Lalic, A. Galeckas, and V. Grivickas, Analysis of the stretched exponential photolumines-
cence decay from nanometer-sized silicon crystals in SiO2, J. Appl. Phys. 86, 6128 (1999).
81. J. Heitmann, F. Müller, L. Yi, M. Zacharias, D. Kovalev, and F. Eichhorn, Excitons in Si nanocrystals:
Confinement and migration effects, Phys. Rev. B 69, 195309 (2004).
82. I. V. Antonova, M. Gulyaev, E. Savir, J. Jedrzejewski, and I. Balberg, Charge storage, photolumines-
cence, and cluster statistics in ensembles of Si quantum dots, Phys. Rev. B 77, 125318 (2008).
83. I. Mihalcescu, J. C. Vial, and R. Romestain, Absence of carrier hopping in porous silicon, Phys. Rev.
Lett. 80, 3392 (1998).
84. C. Delerue, G. Allan, C. Reynaud, O. Guillois, G. Ledoux, and F. Huisken, Multiexponential photo­
luminescence decay in indirect-gap semiconductor nanocrystals, Phys. Rev. B 73, 235318 (2006).
85. X. Zianni and A. G. Nassiopoulou, Photoluminescence lifetimes of Si quantum dots, J. Appl. Phys. 100,
074312 (2006).
86. P. D. J. Calcott, K. J. Nash, L. T. Canham, M. J. Kane, and D. Brumhead, Identification of radiative
transitions in highly porous silicon, J. Phys.: Condens. Matter 5, L91 (1993).
87. O. Bisi, S. Ossicini, and L. Pavesi, Porous silicon: A quantum sponge structure for silicon based
optoelectronics, Surf. Sci. Rep. 9, 126 (2000).
88. N. Arad-Vosk and A. Sa’ar, Radiative and nonradiative relaxation phenomena in hydrogen- and oxygen-
terminated porous silicon, Nanoscale Res. Lett. 9(1), 47 (2014).
89. D. Kovalev, J. Diener, H. Heckler, G. Polisski, N. Künzner, and F. Koch, Optical absorption cross
­sections of Si nanocrystals, Phys. Rev. B 61, 4485 (2000).
90. L. Pavesi, L. Dal Negro, C. Mazzoleni, G. Franzó, and F. Priolo, Optical gain in silicon nanocrystals,
Nature 408, 440 (2000).
91. L. Khriachtchev, M. Räsänen, S. Novikov, and J. Sinkkonen, Optical gain in SiO/SiO2 lattice:
Experimental evidence with nanosecond pulses, Appl. Phys. Lett. 79, 1249 (2001).
92. K. Luterová, I. Pelant, I. Mikulskas, R. Tomasiunas, D. Muller, J.-J. Grob, J.-L. Rehspringer, and
B. Hönerlage, Stimulated emission in blue-emitting Si+-implanted SiO2 films, J. Appl. Phys. 91, 2896
(2002).
Silicon Nanocrystals 211

93. I. Pelant, Optical gain in silicon nanocrystals: Current status and perspectives, Phys. Status Solidi A
208, 625 (2011).
94. M. C. Beard, K. P. Knutsen, P. Yu, J. M. Luther, Q. Song, W. K. Metzger, R. J. Ellingson, and A. J. Nozik,
Multiple exciton generation in colloidal silicon nanocrystals, Nano Lett. 7(8), 2506 (2007).
95. M. Tuan Trinh, R. Limpens, W. D. A. M. de Boer, J. M. Schins, L. D. A. Siebbeles, and T. Gregorkiewicz,
Direct generation of multiple excitons in adjacent silicon nanocrystals revealed by induced absorption,
Nat. Photon. 6, 316 (2012).
96. T. Taskin, S. Gardelis, J. H. Evans, B. Hamilton, and A. R. Peaker, Sharp 1.54 μm luminescence of
porous, erbium doped silicon, Electron. Lett. 31, 2132 (1995).
97. L. Gu, Z. Xiong, G. Chen, Z. Xiao, D. Gong, X. Hou, and X. Wang, Luminescent erbium-doped porous
silicon bilayer structures, Adv. Mater. 13, 1402 (2001).
98. M. Fujii, M. Yoshida, S. Hayashi, and K. Yamamoto, Photoluminescence from SiO2 films containing
Si nanocrystals and Er: Effects of nanocrystalline size on the photoluminescence efficiency of Er3+,
J. Appl. Phys. 84, 4525 (1998).
99. C. E. Chryssou, A. J. Kenyon, T. S. Iwayama, C. W. Pitt, and D. E. Hole, Evidence of energy
coupling between Si nanocrystals and Er3+ in ion-implanted silica thin films, Appl. Phys. Lett. 75,
2011 (1999).
100. G. Franzò, D. Pacifici, V. Vinciguerra, F. Priolo, and F. Iacona, Er3+ ions–Si nanocrystals interactions
and their effects on the luminescence properties, Appl. Phys. Lett. 76, 2167 (2000).
101. M. Makarova, V. Sih, J. Warga, R. Li, L. Dal Negro, and J. Vuckovic, Enhanced light emission in
photonic crystal nanocavities with Erbium-doped silicon nanocrystals, Appl. Phys. Lett. 92, 161107
(2008).
102. R. Hoffmann, J. Beyer, V. Klemm, D. Rafaja, B. C. Johnson, J. C. McCallum, and J. Heitmann,
Erbium-doped slot waveguides containing size-controlled silicon nanocrystals, J. Appl. Phys. 17,
163106 (2015).
103. N. Koshida and H. Koyama, Visible electroluminescence from porous silicon, Appl. Phys. Lett. 60, 347
(1992).
104. H. Shi, Y. Zheng, Y. Wang, and R. Yuan, Electrically induced light emission and novel photocurrent
response of a porous silicon device, Appl. Phys. Lett. 63, 770 (1993).
105. A. Richter, P. Steiner, F. Kozlowski, and W. Lang, Current induced light-emission from a porous silicon
device, IEEE Electron. Device Lett. 12, 691 (1991).
106. V. A. Kuznetsov, I. Andrienko, and D. Haneman, High efficiency blue–green electroluminescence and
scanning tunneling microscopy studies of porous silicon, Appl. Phys. Lett. 72, 3323 (1998).
107. T. Oguro, H. Koyama, T. Ozaki, and N. Koshida, Mechanism of the visible electroluminescence from
metal/porous silicon/n-Si devices, J. Appl. Phys. 81, 1407 (1997).
108. N. Koshida, H. Koyama, Y. Yamamoto, and G. J. Collins, Visible electroluminescence from porous
silicon diodes with an electropolymerized contact, Appl. Phys. Lett. 63, 2655 (1993).
109. Y. Yang, Q. Li, and X. Liu, Study on electroluminescence from porous silicon light-emitting diode,
Chin. Opt. Lett. 4, 297 (2006).
110. F. Namavar, H. P. Maruska, and N. M. Kalkhoran, Visible electroluminescence from porous silicon np
heterojunction diodes, Appl. Phys. Lett. 60, 2514 (1992).
111. H. Li, B. Huang, D. Yi, H. Cui, Y. He, and J. Peng, Efficient visible electroluminescence from porous
silicon diodes with low driven voltage, Chin. Opt. Lett. 2, 171 (2004).
112. J. Linnros and N. Lalic, High quantum efficiency for a porous silicon light emitting diode under pulsed
operation, Appl. Phys. Lett. 66, 3048 (1995).
113. K. Nishimura, Y. Nagao, and N. Ikeda, High external quantum efficiency of electroluminescence from
photoanodized porous silicon, Jpn. J. Appl. Phys. 37, L303 (1998).
114. P. Steiner, F. Kozlowski, and W. Lang, Light‐emitting porous silicon diode with an increased electro­
luminescence quantum efficiency, Appl. Phys. Lett. 62, 2700 (1993).
115. G. Barillaro, A. Diligent, F. Pieri, F. Fuso, and M. Allegrini, Integrated porous-silicon light-emitting
diodes: A fabrication process using graded doping profiles, Appl. Phys. Lett. 78, 4154 (2001).
116. F. Kozlowski, P. Steiner, W. Lang, and H. Sandmaier, Light-emitting diodes in porous silicon, Sens.
Actuators A 43, 153 (1994).
117. A. G. Nassiopoulos, S. Grigoropoulos, and D. Papadimitriou, Electroluminescent device based on
silicon nanopillars, Appl. Phys. Lett. 69, 2267 (1996).
118. G. Conibeer, M. Green, E.-C. Cho, D. König, Y.-H. Cho, T. Fangsuwannarak, G. Scardera et al., Silicon
quantum dot nanostructures for tandem photovoltaic cells, Thin Solid Films 516, 6748 (2008).
212 Semiconductor Nanocrystals and Metal Nanoparticles

119. H.-Z. Song, X.-M. Bao, N.-S. Li, and J.-Y. Zhang, Relation between electroluminescence and photo­
luminescence of Si+-implanted SiO2, J. Appl. Phys. 82, 4028 (1997).
120. A. G. Nassiopoulou, V. Ioannou-Sougleridis, P. Photopoulos, A. Travlos, V. Tsakiri, and D. Papadimitriou,
Stable visible photo- and electroluminescence from nanocrystalline silicon thin films fabricated on thin
SiO2 layers by low pressure chemical vapour deposition, Phys. Status Solidi A 165, 79 (1998).
121. P. Photopoulos, A. G. Nassiopoulou, D. N. Kouvatsos, and A. Travlos, Photo- and electrolumines-
cence from nanocrystalline silicon single and multilayer structures, Mater. Sci. Eng. B69–B70, 345
(2000).
122. P. Photopoulos and A. G. Nassiopoulou, Room- and low-temperature voltage tunable electrolumines-
cence from a single layer of silicon quantum dots in between two thin SiO2 layers, Appl. Phys. Lett. 77,
1816 (2000).
123. J. Valenta, N. Lalic, and J. Linnros, Electroluminescence microscopy and spectroscopy of silicon nano-
crystals in thin SiO2 layers, Opt. Mater. 17, 45 (2001).
124. G. Franzò, A. Irrera, E. C. Moreira, M. Miritello, F. Iacona, D. Sanfilippo, G. Di Stefano, P. G. Fallica,
and F. Priolo, Electroluminescence of silicon nanocrystals in MOS structures, Appl. Phys. A 74, 1–5
(2002).
125. R. J. Walters, G. I. Bourianoff, and H. A. Atwater, Field-effect electroluminescence in silicon nanocrys-
tals, Nat. Mater. 4, 143 (2005).
126. T. Creazzo, B. Redding, E. Marchena, J. Murakowski, and D. W. Prather, Tunable photoluminescence
and electroluminescence of size-controlled silicon nanocrystals in nanocrystalline-Si/SiO2 superlat-
tices, J. Lumin. 130, 631 (2010).
127. J. López-Vidrier, Y. Berencén, S. Hernández, B. Mundet, S. Gutsch, J. Laube, D. Hiller et al., Structural
parameters effect on the electrical and electroluminescence properties of silicon nanocrystals/SiO2
superlattices, Nanotechnology 26, 185704 (2015).
128. B. Ghosh, Y. Masuda, Y. Wakayama, Y. Imanaka, J.-I. Inoue, K. Hashi, K. Deguchi et al., Hybrid white
light emitting diode based on silicon nanocrystals, Adv. Funct. Mater. 24, 7151 (2014).
129. F. Maier-Flaig, J. Rinck, M. Stephan, T. Bocksrocker, M. Bruns, C. Kübel, A. K. Powell, G. A. Ozin, and
U. Lemmer, Multicolor silicon light-emitting diodes (SiLEDs), Nano Lett. 13, 475 (2013).
130. S. Tiwari, F. Rana, H. Hanafi, A. Hartstein, E. F. Crabbé, and K. Chan, A silicon nanocrystals based
memory, Appl. Phys. Lett. 68, 1377 (1996).
131. H. Qin, X. Gu, H. Lu, J. Liu, X. Huang, and K. Chen, Observation of Coulomb-blockade in a field-effect
transistor with silicon nanocrystal floating gate at room temperature, Solid State Commun. 111, 171
(1999).
132. E. Kapetanakis, P. Normand, D. Tsoukalas, and K. Beltsios, Room-temperature single-electron charg-
ing phenomena in large-area nanocrystal memory obtained by low-energy ion beam synthesis, Appl.
Phys. Lett. 80, 2794 (2002).
133. S.-T. Kang, B. Winstead, J. Yater, M. Suhail, G. Zhang, C.-M. Hong, H. Gasquet et al., High perfor-
mance nanocrystal based embedded flash microcontrollers with exceptional endurance and nanocrystal
scaling capability, in 2012 Fourth IEEE International Memory Workshop (IMW ), Milan, Italy, pp. 1–4,
doi:10.1109/IMW.2012.6213668.
134. W. R. Taube, A. Kumar, R. Saravanan, P. B. Agarwal, P. Kothari, B. C. Joshi, and D. Kumar, Efficiency
enhancement of silicon solar cells with silicon nanocrystals embedded in PECVD silicon nitride matrix,
Solar Energy Mater. Solar Cells 101, 32 (2012).
135. F. Sgrignuoli, P. Ingenhoven, G. Pucker, V. D. Mihailetchi, E. Froner, Y. Jestin, E. Moser, G. Sànchez,
and L. Pavesi, Purcell effect and luminescent down shifting in silicon nanocrystals coated back-contact
solar cells, Solar Energy Mater. Solar Cells 132, 267 (2015).
136. E.-C. Cho, M. A. Green, G. Conibeer, D. Song, Y.-H. Cho, G. Scardera, S. Huang et al., Silicon quantum
dots in a dielectric matrix for all-silicon tandem solar cells, Adv. OptoElectron. 2007, Article ID 69578
(2007).
137. G. Conibeer, M. Green, R. Corkish, Y. Cho, E.-C. Cho, C.-W. Jiang, T. Fangsuwannarak et al., Silicon
nanostructures for third generation photovoltaic solar cells, Thin Solid Films 511–512, 654 (2006).
138. W. R. Taube, A. Kumar, R. Saravanan, P. B. Agarwal, P. Kothari, B. C. Joshi, and D. Kumar, Efficiency
enhancement of silicon solar cells with silicon nanocrystals embedded in PECVD silicon nitride matrix,
Solar Energy Mater. Solar Cells 101, 32–35 (2012).
139. P.-J. Wu, Y.-C. Wang, and I-C. Chen, Fabrication of Si heterojunction solar cells using P-doped Si nano-
crystals embedded in SiNx films as emitters, Nanoscale Res. Lett. 8, 457 (2013).
Silicon Nanocrystals 213

140. P. Löper, A. Witzky, A. Hartel, S. Gutsch, D. Hiller, J. C. Goldschmidt, S. Janz, S. W. Glunz, and
M. Zacharias, Photovoltaic properties of silicon nanocrystals in silicon carbide, in Physics, Simulation, and
Photonic Engineering of Photovoltaic Devices, Proceedings of SPIE, Vol. 8256, p. 82560G, A. Freundlich
and J.-F. F. Guillemoles (eds.) (February 9, 2012), San Francisco, CA, doi: 10.1117/12.906669.
141. Y. Ding, R. Gresback, R. Yamada, K. Okazaki, and T. Nozaki, Hybrid silicon nanocrystal/
poly(3-hexylthiophene-2,5-diyl) solar cells from a chlorinated silicon precursor, Jpn. J. Appl. Phys. 52,
11NM04(2013).
142. C.-Y. Liu and U. R. Kortshagen, A silicon nanocrystal Schottky junction solar cell produced from
colloidal silicon nanocrystals, Nanoscale Res. Lett. 5(8), 1253 (2010).
143. S.-K. Kim, C.-H. Cho, B.-H. Kim, S.-J. Park, and J. W. Lee, Electrical and optical characteristics of
silicon nanocrystal solar cells, Appl. Phys. Lett. 95, 143120 (2009).
 6 Electronic and Optical
Properties of Si and
Ge Nanocrystals
Tupei Chen

CONTENTS
6.1 Introduction........................................................................................................................... 215
6.2 Theoretical Studies................................................................................................................ 216
6.2.1 Hydrogenated Si Nanocrystals (nc-Si)....................................................................... 216
6.2.2 Oxidized nc-Si........................................................................................................... 229
6.2.3 Ge Nanocrystals........................................................................................................ 231
6.3 Experimental Studies of Optical Properties.......................................................................... 238
6.3.1 nc-Si Embedded in SiO2 Thin Films......................................................................... 239
6.3.2 nc-Si Embedded in Si3N4 Thin Films........................................................................ 243
6.3.3 Self-Assembled Ge Nanocrystals.............................................................................. 245
6.3.4 Ge Nanocrystals Embedded in SiO2 Matrices........................................................... 249
Acknowledgment............................................................................................................................ 251
References....................................................................................................................................... 251

6.1 INTRODUCTION
Bulk crystalline silicon (Si) and germanium (Ge) are the two most important indirect bandgap
semiconductors. Bulk Si crystal is the main material of today’s microelectronic/nanoelectronic,
photovoltaic, and microelectromechanical system (MEMS) technologies. In particular, the ubiqui-
tous “Si chip” is taken for granted in today’s society. Germanium is a semiconductor material that
formed the basis for the development of transistor technology (the first solid-state transistor was
made with Ge). But by the late 1950s, silicon had emerged as the favored semiconductor, and it has
remained ever since because of the breakthrough of planar technology and integrated circuit (IC)
technology. Nevertheless, Ge is an important material for optoelectronic devices such as photode-
tectors and solar cells, and in recent years there has been renewed interest in Ge, which has been
triggered by its strong potential for deep submicrometer (sub-45 nm) IC technologies.
Optical properties of a semiconductor can be any of the properties that involve the interac-
tion between light and the semiconductor, including absorption, diffraction, polarization, reflec-
tion, refraction, and so on. Most optical properties of semiconductors are related to the particular
nature of the electronic band structures of the semiconductors, which are in turn determined by the
crystallographic structure, the particular atoms, and their bonding. The full symmetry of the space
groups is also essential in determining the structure of the energy bands. From the macroscopic
viewpoint, the interaction of matter with light is described by Maxwell’s equations, in which the
optical properties of matter are introduced as the constants characterizing the medium, such as
the dielectric constant. The dispersion of the dielectric constant over a particular range of photon
energy is called the “dielectric function” (ε) ⋅ ε is a complex number, which is often described by its
real (ε1) and imaginary (ε2) parts, that is, ε = ε1−iε2. The dielectric function can be expressed also
in terms of the complex refractive index (N) with ε = N2 = (n–ik)2 where n and k are the refractive

215
216 Semiconductor Nanocrystals and Metal Nanoparticles

index and extinction coefficient of the semiconductor, respectively. Both n and k are important opti-
cal constants; they are real and positive numbers.
The band structure of a semiconductor can be significantly modified when its dimensions are
reduced to the scale of the exciton Bohr radius (e.g., ~4.9 nm for bulk crystalline silicon) of the
bulk crystalline semiconductor, which is due to the quantum confinement effect. Therefore, the
electronic and optical properties of a semiconductor in the nanoscale (e.g., Si and Ge nanocrystals)
would be significantly different from those of its bulk counterpart. In this chapter, we review the
electronic and optical properties of Si and Ge nanocrystals. The organization of this chapter is based
on the following two main sections: in Section 6.2, some theoretical calculations of the electronic
and optical properties of small Si and Ge nanocrystals with up to several hundreds of atoms are
introduced; in Section 6.3, some experimental studies of the optical properties (with a focus on the
dielectric functions and optical constants) of Si nanocrystals embedded in a dielectric matrix (SiO2
and Si3N4) and of self-assembled Ge nanocrystals, which are based on spectroscopic ellipsometry
(SE), are presented.

6.2 THEORETICAL STUDIES

The calculation of the electronic and optical properties of Si and Ge nanocrystals is a difficult task.
The nanocrystals that can be observed in experiments are usually too small to be described by those
theoretical techniques (e.g., the k-space techniques) that are used for large crystals. On the other
hand, they are too big to be described by real-space, high-level ab initio quantum chemistry tech-
niques that are used for small molecules. For example, a small Si nanocrystal with a diameter of
2.5 nm consists of 281 Si atoms and 172 H atoms on the surface including 4106 electrons [1]. This is
a huge many-body system. The solution to the time-independent Schrödinger equation for the system
is a formidable task. It is therefore necessary to resort to approximations, or to different approaches.
The theoretical ab initio investigation of the optical properties demands an accurate account
of electron correlation for both the ground and the excited states and involves high computational
cost, which scales at least as the fifth or sixth power of the diameter d of the nanocrystal [2]. The
ab initio approaches are limited by their applicability to systems of less than 1000 atoms with the
current computer power. Calculations of the optical properties can be performed using an ab initio
technique based on density functional theory (DFT) and local density approximation (LDA) [3–5]
or using time-dependent LDA [6,7]. Many-body effects such as self-energy corrections and exci-
tonic effects have been taken into account [6,8–10]. However, the correct treatment of the many-
body effects in the calculation of the electron–hole pair excitation energies is a concern [6,11,12].
Many theoretical calculations on the electronic and optical properties of Si and Ge nanocrystals
are semiempirical. For example, semiempirical tight-binding approaches [13,14] and the empirical
pseudopotential approach [15,16] allow the treatment of much larger nanocrystalline systems. Such
approaches are based on the knowledge of the electronic structure of bulk crystals, but the trans-
ferability of bulk electronic interaction parameters to a nanocrystalline environment is question-
able. This holds true for both the tight-binding approximation and the empirical-pseudopotential
approach [5]. Various approaches have been used to calculate the electronic and optical properties
of small nanocrystals, such as hydrogenated Si and Ge nanocrystals [2–4,6–8,10,14,17–20], oxi-
dized Si nanocrystals [2,18,21–23], and Si nanocrystals embedded in a SiO2 matrix [22,24–27].
In this section, we present theoretical analysis of the structural and electronic properties
and optical gaps of various nanocrystal systems including hydrogenated Si and Ge nanocrystals,
­oxidized Si nanocrystals, and Si nanocrystals embedded in SiO2 matrix.

6.2.1 Hydrogenated Si Nanocrystals (nc-Si)

Simple cubic supercells can be used to model the nc-Si surrounded by vacuum and with differ-
ent passivation techniques (e.g., hydrogen passivation) of the surface dangling bonds. For hydrogen
Electronic and Optical Properties of Si and Ge Nanocrystals 217

passivation, with the relaxed SimHn structures, Degoli et al. performed an ab initio calculation of the
electronic and structural properties of small hydrogenated nc-Si as a function of dimension [17]. The
calculation was done within DFT using a pseudopotential plane-wave approach. They first examined
how the ground-state structural properties of the nc-Si change as a function of the cluster dimension
and how the creation of an electron–hole pair modifies the overall structures. Then they showed that
the structural modifications are immediately reflected into the electronic properties of nc-Si.
In the study reported by Degoli et al. [17], the calculations for each cluster were performed both
in the ground and the excited states, and the structural properties were determined by allowing full
relaxation of each hydrogenated nc-Si. The starting configuration for the clusters was fixed with
all Si atoms occupying the same position as in the bulk crystal, and passivation of the surface was
done with H atoms placed along the bulk crystal directions. The formation of an electron–hole pair
under excitation was taken into account by forcing one electron to occupy the the lowest unoccu-
pied single-particle state (LUMO), thus leaving a hole in the highest occupied single-particle state
(HOMO) [17].
Figure 6.1 shows the average Si–Si bond lengths of the relaxed Si5H12, Si10H16, Si29H36, and
Si35H36 clusters in their ground- and excited state configurations [17]. It can be clearly seen in the
figure that there is a contraction of the hydrogenated nc-Si in its ground-state configuration with
respect to bulk silicon. The contraction becomes smaller and the average interatomic distances tend
to the calculated bulk value as the size of the cluster increases (the average bond lengths approach
the bulk value for hydrogenated nc-Si with diameter of about 20 Å). However, no clear trend can be
identified for the excited-state configuration because of heavy clusters distortions.
Figure 6.2 demonstrates that the cluster relaxation in the excited configuration causes struc-
tural distortions [17]. As shown in Figure 6.2a for the Si10H16 cluster, the first-, second-, third-, and
fourth-neighbor Si–Si distances are almost unchanged, with respect to bulk values in the ground-
state configuration, whereas significant deviations are induced by the excitation. The same situation
is observed for the Si29H36 cluster (see Figure 6.2b). It has been concluded that the presence of an
electron–hole pair in the clusters causes a strong deformation of the structures with respect to their
ground-state configuration, and this is more evident for smaller clusters [17].
The structural modifications due to the cluster dimension effect or to the creation of an elec-
tron–hole pair lead to changes in the electronic structures. Figure 6.3 shows the Kohn–Sham levels

2.95

2.93
Interatomic distances (Å)

2.91

2.89

2.37

2.35

2.33

2.31
2.3
0 10 20 30 40
Number of silicon atoms

FIGURE 6.1 Average Si–Si interatomic distances for the relaxed Si5H12, Si10H16, Si29H36, and Si35H36 clusters
in their ground-state (circle) and excited-state (square) configurations. The horizontal line indicates the calcu-
lated interatomic distances in bulk Si. (From Degoli, E. et al., Phys. Rev. B, 69, 155411, 2004. With permission.)
218 Semiconductor Nanocrystals and Metal Nanoparticles

Si–Si distances (Å)

4

2
0 10 20 30 40 50
(a) Si–Si pair index

6
Distance of each Si from the center (Å)

2
0 10 20 30
(b) Index of the Si atoms in the cluster

FIGURE 6.2 Calculated Si–Si distances (a) for each pair of Si atoms for the Si10H16 cluster, and (b) of
each Si atom with respect to the central Si of the Si29H36 cluster. Circles and squares represent ground- and
­excited-state configurations, respectively. Straight lines represent the calculated bulk silicon values for the
first-, second-, third-, and fourth-neighbor distances. (From Degoli, E. et al., Phys. Rev. B, 69, 155411, 2004.
With permission.)

for the Si5H12, Si10H16, Si29H36, and Si35H36 clusters in both the ground- (left panel) and excited-state
(right panel) configurations [17]. It can be concluded from the figure that the ground-state energy
gap increases as the cluster dimension decreases. In addition, the excitation of the electron–hole
pair causes a reduction of the energy gap, which is more significant for smaller clusters. For the
excited clusters, the HOMO and LUMO become strongly localized in correspondence to the struc-
tural distortion, giving rise to defect-like states that reduce the gap; and the effect is stronger for a
smaller cluster [17].
The nanocrystal size influences both the energetic positions of the optical transitions and their
oscillator strengths. This has been demonstrated by many calculations [5,18]. Figure 6.4 shows
one such calculation of the transition energies and the oscillator strengths of the HOMO–LUMO
Electronic and Optical Properties of Si and Ge Nanocrystals 219

6 6
5 5
4 4
3 3
2 2

Energy (eV)
Energy (eV)

1 1
0 0
–1 –1
–2 –2
–3 –3
–4 –4
–5 –5
–6 –6
(a) Γ Γ

6 6
5 5
4 4
3 3
2 2
Energy (eV)
Energy (eV)

1 1
0 0
–1 –1
–2 –2
–3 –3
–4 –4
–5 –5
–6 –6
(b) Γ Γ

6 6
5 5
4 4
3 3
2 2
Energy (eV)

Energy (eV)

1 1
0 0
–1 –1
–2 –2
–3 –3
–4 –4
–5 –5
–6 –6
(c) Γ Γ

6 6
5
4 4
3
2 2
Energy (eV)
Energy (eV)

1
0 0
–1
–2 –2
–3
–4 –4
–5
–6 –6
(d) Γ Γ

FIGURE 6.3 Calculated energy levels at the Γ point for (a) Si5H12, (b) Si10H16, (c) Si29H36, and (d) Si35H36
clusters in ground-state (left panel) and excited-state (right panel) configuration. The energies are referred to
the highest valence level. (From Degoli, E. et al., Phys. Rev. B, 69, 155411, 2004. With permission.)
220 Semiconductor Nanocrystals and Metal Nanoparticles

4
Transition energy (eV)
3

0.0005
1
0.42 0.05 0.005 0.001 0.000002

–1
5 17 41 83 147 239

FIGURE 6.4 Energy-level schemes for the Kohn–Sham eigenvalues of Si nanocrystals passivated with
hydrogen. The number of Si atoms is indicated by integers. The arrows indicate the HOMO–LUMO transition
with the oscillator strength given by the real numbers. (From Ramos, L.E. et al., Phys. Stat. Sol. (b), 242, 3053,
2005. With permission.)

transitions for nc-Si passivated with hydrogen [18]. In this figure, the quantum confinement effects
on both the energy levels and the oscillator strengths are clearly visible.
Öğüt et al. performed the calculation on the quantum size effect on the quasiparticle gaps, self-
energy corrections, exciton Coulomb energies, and optical gaps in Si nanocrystals (i.e., quantum
dots) from first principles using a real-space pseudopotential method [8]. The calculations were
performed on hydrogen-passivated spherical Si clusters with diameters up to 27.2 Å (~800 Si
and H atoms). For an n-electron system, the quasiparticle gap Egqp can be expressed in terms of the
ground-state total energies E of the (n + 1)-, (n – 1)‐, and n-electron systems as [8]

Egqp = E (n + 1) + E (n - 1) - 2 E (n) = Egband + S (6.1)

where
Egband is the usual single-particle LDA (local density approximation) band gap (defined as the
eigenvalue difference between the lowest unoccupied and the highest occupied orbitals)
Σ is the self-energy correction

Figure 6.5 shows the size dependence of the quasiparticle and LDA band gaps, and the self-energy
corrections [8]. The quasiparticle bandgap ( Egqp (d )), the LDA bandgap ( Egband (d )), and the self-energy
correction (Σ(d)) for a nanocrystal of size d are enhanced substantially with respect to their bulk
values ( Egqp (bulk ), Egband (bulk )), and (Σ(bulk)), respectively) as a result of quantum confinement.
Power-law fitting to the calculated data in Figure 6.5 shows that Egqp (d ) - Egqp (bulk ), Egqp (d ) - Egband
(bulk ), and S(d ) - S(bulk ) scale as d−1.2, d−1.1, and d−1.5, respectively [8].
For direct comparison with experimental absorption data, the Coulomb and exchange-­
correlation energies of the exciton need to be included. Compared to the Coulomb energy, exciton
Electronic and Optical Properties of Si and Ge Nanocrystals 221

6 Quasiparticle
LDA band
Self-energy
5
Energy (eV)
4

10 14 18 22 26 30
Quantum dot diameter (Å)

FIGURE 6.5 Calculated quasiparticle and LDA bandgaps and self-energy corrections as a function of the
nanocrystal (quantum dot) diameter d (in Å). The solid lines are power-law fits to the calculated data. The
horizontal dotted line is the bulk limit of the self-energy correction (0.68 eV). (From Öğüt, S. et al., Phys. Rev.
Lett., 79, 1770, 1997. With permission.)

exchange-correlation energies are much smaller for the nanocrystals, and can therefore be neglected [8].
Quantum confinement (QC) in nanostructures enhances the bare exciton Coulomb interaction and
also reduces electronic screening, so that the exciton Coulomb energy ECoul can be comparable to
the quasiparticle gap. ECoul can be approximately calculated from the effective mass approximation
(EMA), which yields (in atomic units) [8]

3.572
ECoul = (6.2)
ed

where ε and d are the dielectric constant and diameter of the nanocrystal, respectively [8]. Though
EMA is not able to yield accurate exciton Coulomb energies, it can provide a simple picture of
the quantum size dependence of the exciton Coulomb energy. Figure 6.6 shows the unscreened
exciton Coulomb energies obtained from various calculations, including the direct ab initio pseudo-
potential calculation, the EMA calculation, and the semiempirical pseudopotential calculation [8].
A strong size dependence of the unscreened Coulomb energy can be observed from Figure 6.6.
The unscreened exciton Coulomb energies are inversely proportional to the nanocrystal diameter d,
that is, there is a power-law dependence d−α with α = 0.7 for the ab initio pseudopotential calcula-
tion, 0.8 for the semiempirical calculation, and 1 for the EMA calculation.
An accurate calculation of the exciton Coulomb energy requires the knowledge of the dielectric
constant of the nanocrystal. If ECoul can be calculated accurately, the optical gap can be extracted as
Egopt = Egqp – ECoul [8]. Figure 6.7 shows the calculated optical gaps along with the quasiparticle gaps
and experimental absorption data as a function of the nanocrystal diameter [8]. A strong quantum
size dependence of the optical gap can be observed from the figure.
222 Semiconductor Nanocrystals and Metal Nanoparticles

Ab initio
Semiempirical
5
EMA

Unscreened Coulomb energy (eV)

4

1
10 14 18 22 26 30
Quantum dot diameter (Å)

FIGURE 6.6 Unscreened exciton Coulomb energies as a function of the nanocrystal (quantum dot) diameter d
(in Å) calculated by (1) effective mass approximation (EMA) (dot line), (2) direct semiempirical pseudopoten-
tial calculations [28], and (3) direct ab initio pseudopotential calculations. The solid lines are power-law fits to
the calculated data. (From Öğüt, S. et al., Phys. Rev. Lett., 79, 1770, 1997. With permission.)

5
Energy (eV)

10 14 18 22 26 30
Quantum dot diameter (Å)

FIGURE 6.7 Calculated quasiparticle gaps (dotted line), optical gaps (shown by × fitted to the solid line),
and experimental absorption data from Si nanocrystals (◽ and ♢) as a function of the nanocrystal (quantum
dot) diameter d. (From Öğüt, S. et al., Phys. Rev. Lett., 79, 1770, 1997. With permission.)
Electronic and Optical Properties of Si and Ge Nanocrystals 223

A simple expression for the size dependence of the optical band gap can be obtained from the
power-law fitting to the calculated optical gaps. The expression can be written as

c
Egopt (d ) = Egopt (¥) + (6.3)
dm

where
Egopt (d ) is the optical gap (in eV)
Egopt (¥) is the bandgap of the bulk material (here Egopt (¥) is 1.12 eV for bulk crystalline silicon)
d is the nanocrystal diameter (in nm)

The constants c and m are usually different for different calculations. Weissker et al. reported the
size dependence with c = 2.96 and m = 1.0 [6]. The calculation of Delerue et al. yields c = 3.73 and
m = 1.39 [9]. Lehtonen and Sundholm proposed c = 3.00 and m = 1.575 [29].
The optical absorption spectra of SinHm nanoclusters up to ~250 atoms were computed by
Vasiliev et al. using a linear response theory within the time-dependent local density approximation
(TDLDA) [7]. The TDLDA formalism incorporates the electronic screening and correlation effects,
which determine exciton binding energies, and thus it represents a full ab initio formalism for the
excited states. Figure 6.8 shows the calculated absorption spectra of SinHm clusters with various
sizes (up to 147 Si atoms and 100 H atoms) [7]. It can be observed from the figure that, as the size
of the cluster increases, the absorption gaps gradually decrease, and the discrete spectra for small
clusters evolve into quasi-continuous spectra for silicon nanocrystals. At the same time, the oscilla-
tor strength of dipole-allowed transitions near the absorption edge decreases with increasing cluster
size. This fact is consistent with the formation of an indirect bandgap in the limit of bulk silicon [7].
Delerue et al. calculated the absorption coefficients of bigger silicon nanocrystals [9]. Figure 6.9
shows the absorption coefficient of a silicon nanocrystal with the size of 3.86 nm, which has a calcu-
lated band gap of 1.67 eV (without exciton binding energy) [9]. As shown in Figure 6.9a, the absorp-
tion edge is shifted to near 3.5 eV, which corresponds to the direct-gap absorption of bulk silicon.
Nothing is visible in the figure between 1.67 and ~3.0 eV at that scale. The reason is that the optical
matrix element for transitions with energy between 1.67 and 3.0 eV is several orders of magnitude
lower than for transitions above 3.0 eV. The optical absorption becomes very close to that of bulk
silicon—but with a blue shift of the absorption edge, that is, it is very close to the absorption of an
indirect semiconductor. In order to have a close view of the absorption near the band edge (1.67 eV),
the absorption coefficient in the energy range 1.4–2.4 eV is plotted in Figure 6.9b at a different
scale and in the form of a bar chart (the amplitudes of the bars represent the integrated absorption
coefficient over the width of the bar). Figure 6.9b shows that the absorption threshold is at the band-
gap energy (1.67 eV) because the transition is dipole-allowed even if this is only with fairly weak
oscillator strength. The optical threshold is also subject to a blue shift depending on the size of the
nanocrystals. It is interesting to note that the absorption coefficient follows approximately a square-
law variation with photon energy [9].
Weissker et al. performed parameter-free calculations of the frequency-dependent dielectric func-
tions of Si nanocrystals [6]. The calculations are based upon the independent-particle approximation
and a pseudopotential plane wave method. The nanocrystals are described by clusters of up to 363
atoms. Their surfaces are passivated by hydrogen atoms. The electronic structure calculations are
based on the DFT-LDA. Each of the Si nanocrystals is situated at the center of a supercell. The super-
cells form an artificial simple cubic (sc) crystal. They considered large sc supercells with nominally
1000 atoms in the bulk limit. The supercells allow the treatment of nearly spherical nanocrystals with
5, 17, 41, 83, 147, 239, and 363 Si atoms and a corresponding number of passivating hydrogen atoms.
The atoms are assumed to be tetrahedrally coordinated with a distance 2.34 Å, which was taken from
the bulk crystal. Some of the calculation results are briefly described in the following.
224 Semiconductor Nanocrystals and Metal Nanoparticles

SiH4 LDA
TDLDA
Si2H6

Si5H12

Si10H16

Photoabsorption (a.u.) Si14H20

Si29H36

Si35H36

Si47H60

Si71H84

Si87H76

Si99H100

Si123H100

Si147H100

2 4 6 8 10
Photon energy (eV)

FIGURE 6.8 Calculated TDLDA absorption spectra of SinHm clusters (solid lines). Spectra of time-­
independent Kohn–Sham LDA eigenvalues (dotted lines) are shown for comparison. (From Vasiliev, I. et al.,
Phys. Rev. Lett., 86, 1813, 2001. With permission.)

Figure 6.10 shows the calculated dielectric functions of spherical nanocrystals with 5, 17, 41, 83,
147, and 239 Si atoms [6]. Although the model atomic structure used is somewhat unrealistic, and
important many-body effects are not included, the absorption spectrum for the smallest Si cluster
of five atoms shows the same basic features as obtained in more sophisticated calculations. The
spectra in Figure 6.10, in particular, those representing the imaginary part of the dielectric function,
are strongly influenced by quantum-confinement effects. The magnitude of the imaginary part of
the dielectric function decreases with decreasing nanocrystal size. On the other hand, the absorp-
tion threshold moves to lower energies with increasing nanocrystal size. In the imaginary part of
the dielectric function in Figure 6.10b, only one broad structure appears with a maximum between
4 and 6 eV. It exhibits a small shift of the main peak toward smaller photon energies with increasing
nanocrystal size. The structure seems to develop into the bulk E2 peak. In the case of the largest
nanocrystals considered here, the almost complete absence of the E1 structure may be related to
neglecting the excitonic effects [6].
It has been mentioned previously that the magnitude of the imaginary part of the dielectric
function decreases with decreasing nanocrystal size. Actually, an important finding from many
Electronic and Optical Properties of Si and Ge Nanocrystals 225

Optical absorption (a.u.) 0.8

0.4

0.0
0.0 2.0 4.0 6.0
(a) E (eV)

0.8
Optical absorption (a.u.)

0.4

0.0
1.4 1.6 1.8 2.0 2.2 2.4
(b) E (eV)

FIGURE 6.9 (a) Calculated optical absorption coefficient with respect to the photon energy E for a silicon
nanocrystal with a diameter of 3.86 nm (the bandgap is calculated at 1.67 eV). (b) Same as (a) but only the
energy region near the bandgap is plotted (amplitudes of the bars represent the integrated absorption coef-
ficient over the width of the bar). (Adapted from Delerue, C. et al., Phys. Rev. B, 48, 11024, 1993. With
permission.)
226 Semiconductor Nanocrystals and Metal Nanoparticles

4
Re ε(ω)

0
(a)

6
Im ε(ω)

0
0 2 4 6 8
(b) Photon energy (eV)

FIGURE 6.10 (a) Real and (b) imaginary part of dielectric functions of Si nanocrystallites with a vary-
ing number N of atoms. N = 5, solid line; N = 17, dotted line; N = 41, dashed line; N = 83, long-dashed line;
N = 147, dot-dashed line; and N = 239, solid line. The vertical arrows in the absorption spectra indicate the
single-particle HOMO–LUMO gaps. (Adapted from Weissker, H.-Ch. et al., Phys. Rev. B, 65, 155328-7, 2002.
With permission.)

investigations on semiconductor nanocrystallites is the drastic reduction in the static dielectric con-
stant when the size of nanocrystals approaches less than 10 nm. Reduction in the static dielectric
constant causes an increase in the Coulomb interaction energy between electrons, holes, and ionized
shallow impurities, and therefore can significantly modify the optical absorption and the transport
phenomenon of a nanometer-sized device [30]. The existing theoretical investigations on the size
dependence of static dielectric constant can be divided into two categories: (1) calculations per-
formed with the use of the semiempirical version of the Penn model, referred to as the modified or
generalized Penn model [31,32], and (2) numerically computed results employing empirical pseudo-
potentials and the semiempirical linear combination of atomic orbitals [15,33,34].
Electronic and Optical Properties of Si and Ge Nanocrystals 227

Tsu et al. presented a simple single-oscillator model for the size-dependent reduction of the static
dielectric constant of silicon nanocrystals [32]. The modified Penn model, taking into account the
quantum confinement–induced discrete-energy states, leads to the size-dependent dielectric constant

eb - 1
e s ( R) = 1 + (6.4)
1 + (DE / DEg )2

where
R is the radius of a spherical silicon nanocrystal
εb is the bulk dielectric constant
ΔE is an energy separation
Eg is an energy gap, as shown in Figure 6.11 [32]

Quantum confinement increases the separations of the discrete states, resulting in an increase in
the energy denominator and a subsequent reduction in the static dielectric constant. Referring to
Figure 6.11, the energy separation is given by [32]

pEF
DE = (6.5)
kF R

Taking the parameters for Si, εb = 12 (or 11.3), Eg = 4 eV (note that the fundamental Γ–Δ gap at 1.1 eV
plays almost no role in the dielectric function), and filling the energy bands up to EF (=12.6 eV),
the computed εs(R) according to the modified Penn model is shown in Figure 6.12 [32].

є2

0 4
E (eV)

E+

Eg

E–

KF 0 KF
K΄ K

FIGURE 6.11 Electron energy versus k for an isotropic, three-dimensional, nearly free electron model.
The inset shows the absorption spectra (ε2) versus photon energy, giving justification for setting Eg = 4 eV
for Si. Round dots indicate the discrete energies and momenta. E+ and E_ define the positions of the new
gap. (From Tsu, R. et al., J. Appl. Phys., 82, 1327, 1997. With permission.)
228 Semiconductor Nanocrystals and Metal Nanoparticles

12

10

8
εs(R)

0
0 10 20 30 40 50 60
R (Å)

FIGURE 6.12 Size dependence of the static dielectric constants εs(R) of silicon nanocrystals. (a) The solid
line delineates εs(R) from the modified Penn model using εb = 12. (b) The dash-dot line delineates ε(R) from
the modified Penn model using εb = 11.3. (c) The crosses depict εs (q) from [35] converted into εs (R) using
q = π/R. (d) The dashed line with shorter dashes delineates the screening dielectric constant from [15].
(e) The dashed line with longer dashes delineates the dielectric constant from [36]. (From Tsu, R. et al.,
J. Appl. Phys., 82, 1327, 1997. With permission.)

It has been suggested that the static dielectric constant εs(R) for a spherical silicon nanocrystal
with radius R can approximately be represented by

eb - 1
e s ( R) = 1 + (6.6)
1 + ( a /R ) m

where m = 2, α = 10.93 Å in Si, and εb = 11.4 is the bulk dielectric constant [15]. This general-
ized Penn model (GPM) is simple, but may be of questionable validity. It has been often used to
explain experimental results on photoluminescence, photoabsorbance, and other optical properties.
However, the size dependence of the static dielectric constant is still not fully understood. The
existing calculations based on GPM are semiempirical in nature, in which the size dependence of
the static dielectric constant has mainly been introduced through the size-dependent energy separa-
tion between two discrete energy levels. The oscillator strength and the energy bandgap have been
treated as independent of the size in most of the existing calculations [30].
On the other hand, Wang et al. performed a “microscopic” calculation of the static dielectric
constant of Si nanocrystals based on the quantum mechanical pseudopotential calculation of the
absorption spectra ε2(E) [15]. The absorption spectra ε2(E) of Si nanocrystals containing up to
~1300 Si atoms were calculated fully quantum mechanically using an empirical pseudopotential
plane wave representation and a novel moments method. The static dielectric constant εs is given by
the integral of the absorption spectra ε2(E):

¥
2
es = 1 +
p ò
e2 ( E )/E dE
0
(6.7)
Electronic and Optical Properties of Si and Ge Nanocrystals 229

The calculation according to Equation 6.7 gives the total polarization dielectric constant depicted in
Figure 6.13 [15]. Fitting the calculation result to the analytic form of Equation 6.6 (see the solid line
shown in Figure 6.13) yields α = 4.25 Å, m = 1.25 for the total polarization dielectric constant εs. Both
α and m are significantly smaller than the results of the GPM, which gives α = 10.93 Å and m = 2.

6.2.2 Oxidized nc-Si
It should be clear from the above discussions that the electronic and optical properties of nc-Si are
determined by the nc-Si size. In addition to the size, the surface and interface properties of nc-Si
should also play an important role in the electronic and optical properties. For instance, it is gener-
ally accepted that the quantum confinement, caused by the restricted space of the nanometer sizes,
is essential for visible light emission in Si nanostructures, but some controversial interpretations of
the photoluminescence (PL) properties of low-dimensional Si structures still exist [22]. In particu-
lar, the effect of oxidation of nc-Si surface is a question that needs to be addressed. Both theoretical
calculations and experiments have been carried out to examine the role of the surface/interface on
the electronic and optical properties. Wolkin et al. observed that oxidation introduces defects in the
nc-Si bandgap, which pin the transition energy [37]. They claimed the formation of a Si=O double
bond as the pinning state. However, Vasiliev et al. pointed out that similar results could also be
obtained for O connecting two Si atoms (single bond) at the nc-Si surface [23]. The assistance of
Si–O vibrations at the interface was also proposed as the dominant path for recombination [38]. On
the other hand, interface radiative states have been suggested to play a key role in the mechanism of
population inversion at the origin of the optical gain observed in nc-Si [39,40].
Garoufalis and Zdetsis reported accurate high-level calculations of the optical gap and absorp-
tion spectrum of small Si nanocrystals, with hydrogen and oxygen at the surface [2]. The calcu-
lations were performed in the framework of time-dependent density functional theory (TDDFT)
using the hybrid nonlocal exchange and correlation functional of Becke and Lee, Yang and Parr
(B3LYP). The oxygen contamination is introduced with several different bonding configurations
(e.g., Si=O double bonds, Si–O–Si bridging bonds, and hydroxyl passivation Si–OH) [2].
Table 6.1 shows the calculated optical gaps of some oxygen-free Si nanocrystals and oxygen-
contaminated Si nanocrystals containing Si=O bonds, Si–O–Si bridging bonds, and both bridg-
ing and double bonds [2]. As can be seen in the table, oxygen contamination leads to a significant
reduction in the optical gaps. There is a red shift of 0.94 eV in the optical gap when the nanocrystal
structure is changed from the oxygen-free Si99H100 to oxygen-rich Si99H76O12 with Si=O bonds;

14
12 εb = 11.4
Dielectric constant

10
εs
8
6
~
εs GPM
4
2
0
4 6 8 10 12 14 16 18 20
R (Å)

FIGURE 6.13 Dielectric constants as a function of the radius R of Si nanocrystals. Here, ε s is for the total
polarization and e s is for exciton screening. The symbols denote the results calculated with Equation 6.7,
while the solid lines are the fitted curves. The dashed curve corresponds to the GPM of Equation 6.6.
(From Wang, L.-W. and Zunger, A., Phys. Rev. Lett., 73, 1039, 1994. With permission.)
230 Semiconductor Nanocrystals and Metal Nanoparticles

TABLE 6.1
Optical Gap of Oxygen-Free and Oxygen-Rich Si Nanocrystals
Oxygen-Free Nanocrystal Optical Gap (eV) Oxygenated Nanocrystal Optical Gap (eV)
Si99H100 3.39 Si99H76O12 2.45
Si147H100 3.19 Si147H52O24 1.85
Si47H48O6 2.70
Si47H36O12 3.7
Si47H24O18 2.5

Source: Adapted from Garoufalis, C.S. and Zdetsis, A.D., Phys. Chem. Chem. Phys., 8, 808, 2006. With permission.

while the red shift is 1.34 eV when the structure is changed from the larger oxygen-free Si147H100 to
oxygen-rich Si99H52O24 with Si=O bonds. Figure 6.14 shows a comparison of the nanocrystal’s size
dependence of the optical gap between oxygen-free Si nanocrystal and oxygen-containing Si nano-
crystal [2]. The calculated red shift ranges from ~2 eV for the smaller nanocrystals down to ~1 eV
for the larger nanocrystals considered here, which is in agreement with Wolkin et al., who have
shown that even a 3 min exposure of the samples in air produces a red shift as large as 1 eV (for rela-
tively large nanocrystals) due to the formation of oxygen bonds [37]. It seems that the oxygen effect
is at least an important factor responsible for the conflicting experimental results reported in the
literature about the exact dependence of the optical gap on the size of the nanocrystals and in par-
ticular the critical size for visible PL. Wolkin et al. obtained the optical gaps as small as 2.2 eV, for
nanoclusters with a diameter of 18 Å [37]. For nanoclusters of about the same size, Wilcoxon et al.
showed a much larger gap (larger than 3.2 eV) for highly purified samples of the same diameter [41].
The critical parameter for the change of the optical gap is not just the amount of oxygen but the
bonding environment and the way the oxygen atoms are accumulated on the surface [2]. The role
of the bonding type of the surface can be highlighted by the comparison of optical gap among the

10

9 Theory { Present TDDFT/B3LYP oxygen free

Present TDDFT/B3LYP oxygen rich
8

7
Experiment
{ Wilcoxon et al. oxygen free
Wilcoxon et al. oxygen rich
6
Energy (eV)

0
0 5 10 15 20 25 30 35 40
Diameter (Å)

FIGURE 6.14 Variation of the optical gap as a function of the Si nanocrystal diameter (in Å). The shaded
area corresponds to oxygen-contaminated nanocrystals containing Si=O bonds. The experimental result is
from Wilcoxon et al. [41]. For diameters 10 < d < 20 Å, the average calculated optical gap for oxygenated
nanocrystals ranges from 2.5 to 1.85 eV. (From Garoufalis, C.S. and Zdetsis, A.D., Phys. Chem. Chem. Phys.,
8, 808, 2006. With permission.)
Electronic and Optical Properties of Si and Ge Nanocrystals 231

following similar nanocrystals with different bonding structures: Si47H48O6 (with Si=O double bonds),
Si47H36O12 (with Si–O–Si bridging oxygen bonds), and Si47H24O18 nanocrystal (with both Si=O and
Si–O–Si bonds). As can be seen in Table 6.1, the optical gaps of these nanoparticles are different. The
Si47H48O6 nanocrystal containing Si=O double bonds has an optical gap of 2.70 eV; but the optical gap
of the Si47H36O12 nanocrystal containing Si–O–Si bridging oxygen bonds is blue-shifted up to 3.7 eV.
However, in the Si47H24O18 nanocrystal, which contains both double (Si=O) and bridging (Si–O–Si)
oxygen bonds, the fundamental optical gap is red-shifted back to the lower energy side (2.5 eV, which
is close to the optical gap of the Si47H48O6 nanocrystal containing only Si=O double bonds).
To examine the possible role of single-bonded surface oxygen atoms, Garoufalis and Zdetsis
also performed calculations on the Si35 nanocrystals with the surface dangling bonds passivated by
both hydrogen atoms and hydroxyl groups [2]. For the fully hydroxyl-passivated Si35(OH)36 nano-
crystal, the HOMO–LUMO gap obtained is 2.6 eV. However, for the similar partially “hydroxyl-
ated” nanocrystals such as Si35H 24(OH)12 and Si35H34(OH)2, the gap is increased by as much as
1.8 eV for Si35H24(OH)12 and 2.2 eV for Si35H34(OH)2 (the HOMO–LUMO gaps of Si35H24(OH)12 and
Si35H34(OH)2 are 4.4 and 4.8 eV, respectively). This indicates that the effect of hydroxyl passivation
is largely dependent on the number of hydroxyl groups deposited on the surface and probably on
their spatial distribution. This observation highlights the role of the relative positions of the oxygen
atoms. The large difference in the behavior between the hydroxyl and double-bonded oxygen is
that, although a few (even two or three) double-bonded oxygen atoms can significantly reduce the
HOMO–LUMO gap, to accomplish the same reduction of the gap by hydroxyl groups, a much larger
number of hydroxyls is needed. Therefore, the presence of double-bonded oxygen “contaminants”
is most effective to reduce the gap [2].
Garoufalis and Zdetsis also calculated the absorption spectra of oxygen-free and oxygen-rich Si
nanocrystals with emphasis on the lower part of the absorption spectra [2]. Figure 6.15 shows the
comparison of the calculated absorption spectrum between the oxygen-free nanocrystals (Si35H36 and
Si71H84) and oxygen-rich nanocrystals (Si35H24O6 and Si71H72O6) containing Si=O double bonds [2].
The optical gaps of the oxygen-free Si35H36 and Si71H84 nanocrystals are 4.42 and 3.64 eV, respectively;
the optical gaps of the oxygen-rich Si35H24O6 and Si71H72O6 nanocrystals are red-shifted to 2.82 and
2.21 eV, respectively. As we can see in Figure 6.15, the presence of surface oxygen not only affects the
value of the fundamental optical gap but also produces a significant enhancement of the absorbance
(large oscillator strengths). The result shown in Figure 6.15 suggests that, in addition to reducing the
optical gap, the Si=O bonds could also enhance the light absorption in the tail of the spectrum.

6.2.3 Ge Nanocrystals
The extension of research interest to Ge nanocrystals (nc-Ge) is rather straightforward, as both
Si and Ge are group IV semiconductors. PL has been experimentally observed in the wavelength
range 350–700 nm from Ge nanocrystals with the sizes of 2–5 nm [42], while there have also been
studies concerning much larger nanoparticles [43]. Various researchers have observed evidence for
quantum confinement effects [42,44–46] and size-dependent PL in the near-infrared region, which
could possibly be due to a radiative recombination of excitons confined in the nanoclusters [44].
Theoretical studies on nc-Ge are relatively limited as compared to those on nc-Si. There have
been calculations reported using various techniques including tight-binding [47], empirical pseu-
dopotentials [48,49], DFT in the local density approximation (LDA) [50,51], a combination of DFT
in the LDA or local spin density approximation (LSDA) with self-consistent field theory (SCF)
[5,6,52,53], and TDDFT [54,55]. The qualitative trend of all of these reports is a significant increase
in the optical gap as the cluster diameter decreases; however, there are disagreements between the
theoretical results. Some results of the theoretical calculations are briefly described in the following.
Garoufalis et al. conducted ab initio calculations of the electronic structures and optical gaps
of Ge nanocrystals based on TDDFT employing the hybrid nonlocal exchange-correlation func-
tional of Becke, Lee, Yang, and Parr (B3LYP) [56]. The size of the nanocrystals considered in the
232 Semiconductor Nanocrystals and Metal Nanoparticles

Si35H36
Si35H24O6

Photoabsorption (a.u.)

2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0

(a) Energy (eV)

Si71H84
Si71H72O6
Photoabsorption (a.u.)

2.0 2.5 3.0 3.5 4.0 4.5 5.0

(b) Energy (eV)

FIGURE 6.15 Comparison of the absorption spectrum of oxygen-free and oxygen-rich Si nanocrystals. The
small arrows indicate a weak peak at ~2.8 eV. (From Garoufalis, C.S. and Zdetsis, A.D., Phys. Chem. Chem.
Phys., 8, 808, 2006. With permission.)

calculations is rather small, ranging from 5 to 99 Ge atoms, with 12–100 H atoms (the diameter of
the largest nanocrystal with Ge99H100 is 19 Å). Figure 6.16 shows the calculated DOS (density of
states) for two Ge nanocrystals with different sizes (Ge47H60 and Ge99H100) [56]. As can be observed
from the figure, the gap of the nanocrystals increases with decreasing size of the crystals, which
is practically “symmetrical” with respect to the conduction and valence band edges. On the other
hand, the similarity of the electronic structure of Si and Ge nanocrystals can be demonstrated by
Figure 6.17, in which the DOS curves for Ge47:H60 and Si47:H60 are shown [56]. The electronic struc-
tures of the two nanocrystals are fully homologous. This could imply that their optical properties
are homologous also. However, Garoufalis et al. showed that, due to the smaller bandgaps of Ge
nanocrystals, the diameter of the smallest Ge nanocrystal that can emit in the visible region is ~19–20 Å,
which is smaller than the corresponding “critical” diameter of Si nanocrystals (22 Å) [56].
Electronic and Optical Properties of Si and Ge Nanocrystals 233

100
90
Ge47:H60
80 Ge99:H100
70
Arbitrary units
60
50
40
30
20
10
0
–20 –18 –16 –14 –12 –10 –8 –6 –4 –2 0 2 4 6 8 10
Energy (eV)

FIGURE 6.16 Comparison of the density of states of Ge47H60 and Ge99H100 nanocrystals. (From Garoufalis, C.S.
et al., J. Phys.: Conf. Ser., 10, 97, 2005. With permission.)

80

Ge47:H60
70
Si47:H60
60

50
Arbitrary units

40

30

20

10

0
–20 –18 –16 –14 –12 –10 –8 –6 –4 –2 0 2 4 6 8 10
Energy (eV)

FIGURE 6.17 Comparison of density of states of Ge 47H60 and Si47H60. (From Garoufalis, C.S. et al.,
J. Phys.: Conf. Ser., 10, 97, 2005. With permission.)

Weissker et al. calculated the energy levels and optical transitions of Ge and Si nanocrystals using
the DFT-LDA technique [6]. In their calculation, the nanocrystals are described by clusters of up to
363 atoms, and their surfaces are passivated by hydrogen atoms. Figure 6.18 shows the energy level
schemes and optical transitions for Ge nanocrystals with 41, 147, and 239 atoms obtained from the
calculation [6]. In this figure, the quantum confinement effects on the energy levels are clearly vis-
ible, and the effect of confinement is also obvious for the oscillator strengths of the optical transitions.
Note that the HOMO–LUMO transition is forbidden by symmetry; in contrast, the transition from the
threefold degenerate second-highest state into the nondegenerate LUMO state possesses extremely
large oscillator strength. Very strong transitions occur close to the absorption edge, even for clusters
234 Semiconductor Nanocrystals and Metal Nanoparticles

Ge 41 Ge 147 Ge 239

2
Energy (eV)
0.16 LUMO
0.13
1 0.25
0.30
0.30 0.13

0.42
HOMO
0

FIGURE 6.18 Level scheme for Ge nanocrystals with 41, 147, and 239 Ge atoms. The HOMO level defines
zero energy. The allowed optical transitions are indicated by vertical arrows. The oscillator strength of a given
transition is indicated by the number at the corresponding arrow. (From Weissker, H.-Ch. et al., Phys. Rev. B,
65, 155328-7, 2002. With permission.)

with a diameter of ~2.2 nm (239 Ge atoms). These transitions could be related to the formation of an
E0-like absorption feature in the more extended Ge nanocrystallites. The behavior of the oscillator
strengths near the absorption edges of Ge nanocrystals is very different from that of Si nanocrystals.
In the case of Si nanocrystals with diameters above 1.5 nm (more than 83 Si atoms), a tail of weak
transitions appears just above the HOMO–LUMO gap. The oscillator strengths of these transitions
are much smaller than the maximum oscillator strengths of ~0.4. The occurrence of the tail can be
interpreted as an indication of the development of bulk properties with increasing nanocrystal size [6].
On the other hand, Nesher et al. also conducted a TDLDA (time-dependent local density approx-
imation) calculation of average oscillator strengths of near-gap transitions for Ge nanocrystals [55].
They pointed out that, because of the large number of near-gap transitions, theoretical predictions
for near-gap optical activity are meaningful only if one averages the oscillator strengths over a
relatively narrow energy window around the optical gap. The size dependence of the calculated
average oscillator strengths for Ge nanocrystals is shown in Figure 6.19 [55]. It can be clearly seen
from the figure that the average oscillator strength of the optical transitions near the absorption edge
decreases strongly with increasing nanocrystal size. The diminishing optical activity with increas-
ing size is consistent with a shift from direct absorption for the smallest molecules and clusters to
the formation of an indirect bandgap in the limit of bulk Ge.
As can be seen in Figure 6.18, the nanocrystal size influences not only the oscillator strengths
but also the energetic positions of the optical transitions. Garoufalis et al. showed that both the
HOMO–LUMO gap and the fundamental optical gap increase as the diameter of the Ge nanocrys-
tals decreases, exhibiting the quantum-size dependence similar to that of Si nanocrystals [56]. They
also showed that, as compared to the Si nanocrystals with the same sizes, the Ge nanocrystals have a
smaller HOMO–LUMO gap and optical gap. The optical gap for the nanocrystal of Ge99H100 (about
19 Å in diameter) obtained from their calculation is 2.95 eV (420 nm in wavelength). Weissker
et al. also calculated the lowest electron–hole pair excitation energies of Ge and Si nanocrystals
with various sizes, and the result is shown in Figure 6.20 [6]. The pair excitation energies are not
much larger than the single-particle HOMO–LUMO gaps estimated by means of the Kohn–Sham
eigenvalues. For the considered crystallite sizes, the pair excitation energy varies between 5 (6) and
1 (2) eV for Ge (Si) nanocrystals. The decrease of the transition energies with the nanocrystal size
is rather rapid. The pair excitation energies are smaller in Ge nanocrystals than in Si nanocrystals.
Electronic and Optical Properties of Si and Ge Nanocrystals 235

10–3
Oscillator strength (a.u.)

10–4

8 10 12 14 16
Diameter (Å)

FIGURE 6.19 Average oscillator strengths for near-edge optical transitions in Ge nanocrystals as a function
of nanocrystal size. The dashed line is a linear fit. (From Nesher, G. et al., Phys. Rev. B, 71, 035344-5, 2005.
With permission.)

5
Pair excitation energy (eV)

1
0.2 0.4 0.6 0.8 1.0 1.2
Radius (nm)

FIGURE 6.20 Lowest electron–hole pair excitation energies of Ge (dots) and Si (triangles) nanocrystals.
The solid lines are the fitted curves according to Eg(R) = Eg(bulk) + α(Å/R)l with α = 14.7 eV and l = 1.0 for
Ge and α = 14.8 eV and l = 1.0 for Si. (From Weissker, H.-Ch. et al., Phys. Rev. B, 65, 155328-7, 2002. With
permission.)
236 Semiconductor Nanocrystals and Metal Nanoparticles

In the interested size range, the Si energies are larger by about 0.5 eV. This value corresponds
roughly to the difference of the fundamental energy gaps in the bulk limit. It has been reported that
the size dependence of the bandgap of Ge nanocrystals also follows Equation 6.3. The fitting to the
calculated pair excitation energies of Ge nanocrystals shown in Figure 6.20 with Equation 6.3 yields
c = 2.94 and m = 1.0 [6]. This means that with m = 1.0 the gap energy varies approximately like the
inverse nanocrystalline diameter. This is much weaker than expected from the quantum mechanics
of the three-dimensional spherical potential well.
Garoufalis et al. also estimated the binding energy (EB) of the exciton formed as a result of the
electronic excitation, which is approximately the difference between the HOMO–LUMO gap and
the fundamental optical gap, and the result is shown in Figure 6.21 [56]. As expected by the quan-
tum confinement hypothesis, the binding energy increases as the diameter of the Ge nanocrystals
decreases. For diameters of ~20 Å, the value of EB is about 0.45 eV. It is interesting to note that,
although the Si and Ge nanocrystals with the diameter of ~20 Å contain different numbers of atoms,
their EB values are about the same.
Nesher et al. calculated ab initio absorption spectra for hydrogen-passivated Ge nanocrystals
using TDDFT within the adiabatic local density approximation [55]. The computed time-dependent
LDA (TDLDA) spectra are shown in Figure 6.22. Each spectrum is compared with a time-­
independent LDA spectrum obtained by considering filled and empty Kohn–Sham orbitals as true
one-electron wave functions. As can be observed in the figure, with increasing nanocrystal size,
both LDA and TDLDA spectra feature a gradual decrease in the absorption and a gradual evolution
from discrete spectra for the smaller nanocrystals to quasicontinuous spectra for the larger ones.
The confinement effect is found to have a strong influence on the frequency-dependent dielectric
function for Ge nanocrystals. Figure 6.23 shows the dielectric function as a function of the number
of Ge atoms calculated by Ramos et al. in the framework of the independent-particle approach [18].
As can be observed in the figure, the effect of quantum confinement on the HOMO–LUMO (cor-
responding to E 0 in bulk Ge) transition is clearly visible in the imaginary part, though it becomes
weaker for high-energy optical transitions. The peak corresponding to the E1 peak in the bulk spec-
trum shows a red shift with increasing nanocrystal size. However, the size effect is practically not
observed in the case of the E2-like transitions near 4.2 eV.

0.8

0.7
Exciton binding energy (eV)

0.6

0.5

0.4

0.3
10 12 14 16 18 20
Diameter (Å)

FIGURE 6.21 Variation of the exciton binding energy of Ge nanocrystals with the crystal size. (From
Garoufalis, C.S. et al., J. Phys.: Conf. Ser., 10, 97, 2005. With permission.)
Electronic and Optical Properties of Si and Ge Nanocrystals 237

GeH4 LDA
TDLDA

Ge2H6

Ge5H12

Ge10H16

Absorption cross section (a.u.) Ge14H20

Ge29H36

Ge35H36

Ge47H60

Ge71H84

Ge87H76

Ge99H100

Ge123H100

Ge147H100

2 4 6 8 10
Photon energy (eV)

FIGURE 6.22 Absorption spectra of Ge nanocrystals calculated using TDLDA (solid lines) and LDA
(dashed lines). (From Nesher, G. et al., Phys. Rev. B, 71, 035344-5, 2005. With permission.)

The optical gap can be extracted from the low-energy optical spectrum. Figure 6.24 shows the
size dependence of the optical gap of spherical Ge nanocrystals calculated using tight binding [47]
and TDLDA [54]. In the same figure, the PL peak as function of the cluster diameter, measured by
Takeoka et al. [44] and Kanemitsu et al. [57], is included also. As can be seen in Figure 6.24, the
optical gap increases with decreasing nanocrystal size. The dependence of the optical gap on the
size of the nanocrystals can be explained with the help of the quantum confinement model [54].
According to this model, PL comes from the recombination of electron–hole pairs confined in the
nanocrystals. By considering the simplified picture of an exciton confined into an infinite spherical
potential, the lowest energy of the electron–hole pair Eg is given by [54]

p2 2
Eg = Eg 0 + (6.8)
2mR 2

where
Eg0 is the optical gap for the bulk crystalline material
μ is the reduced mass of the exciton
R is the radius of the spherical potential
238 Semiconductor Nanocrystals and Metal Nanoparticles

8 8
E2

E1 6
6

Re ε(ω)
4
Im ε(ω)

E0 2
2

0
0 2 4 6 8 0 2 4 6 8
Photon energy (eV) Photon energy (eV)

FIGURE 6.23 (a) Imaginary and (b) real part of the frequency-dependent dielectric functions of Ge nanocrys-
tals passivated with H, containing 363 (dotted line), 239 (solid line), 147 (dash-dotted line), 83 (long-dashed
line), 41 (short-dashed line), 17 (dotted line), and 5 (solid line) Ge atoms. The arrows indicate the single-particle
HOMO–LUMO gaps. (From Ramos, L.E. et al., Phys. Stat. Sol. (b), 242, 3053, 2005. With permission.)

3.5

3
Optical gap (eV)

2.5

1.5

0.5

0
0 2 4 6 8 10 12
Diameter (nm)

FIGURE 6.24 Optical gap of spherical Ge nanocrystals calculated using tight binding by Niquet et al.
(dotted line) [47] and TDLDA by Tsolakidis et al. (asterisks) [54]. Experimental photoluminescence energies
are from Kanemitsu et al. [57] (×) and Takeoka et al. [44] (+). (From Tsolakidis, A. and Martin, R.M., Phys. Rev.
B, 71, 125319, 2005. With permission.)

This simplified model can provide a qualitative explanation for the size dependence of the optical
gap. However, as can be seen in Figure 6.24, the optical gap does not have the simple size depen-
dence of 1/R2 predicted by Equation 6.8.

6.3 EXPERIMENTAL STUDIES OF OPTICAL PROPERTIES

As discussed previously, there have been many theoretical calculations on the optical properties
of Si and Ge nanocrystals. Experimental works on the optical properties were mainly focused on
the luminescence properties of Si and Ge nanocrystals. Although some experimental studies of
Electronic and Optical Properties of Si and Ge Nanocrystals 239

the optical properties of a continuous Si nanocrystal thin film [58] and SiO2/nanocrystalline Si
multilayers [59] have been reported recently, there have been relatively few optical studies so far to
experimentally determine the dielectric functions or optical constants of Si and Ge nanocrystals. In
particular, a comprehensive experimental study of the optical properties of Si and Ge nanocrystals
embedded in various dielectric matrixes over a wider photon energy range is still lacking. In the
material system of dielectric films embedded with silicon or Ge nanocrystals, the dielectric func-
tions or optical constants of the embedded nanocrystals should be different from those of the bulk
crystalline semiconductor due to the size effect, and should be also different from those of a con-
tinuous nanocrystal thin film. In this section, we present some experimental studies of the dielec-
tric functions/optical constants of both Si nanocrystals embedded in SiO2 and Si3N4 matrixes and
self-assembled Ge nanocrystals using spectroscopic ellipsometry (SE) [60–69]; in addition, we also
describe briefly an experimental study of size-dependent near-infrared photoluminescence from Ge
nanocrystals embedded in SiO2 matrices [44].

6.3.1 nc-Si Embedded in SiO2 Thin Films

SE has been used to study the optical properties of silicon nanocrystals (nc-Si) embedded in SiO2
matrix [63–67]. For example, Ding et al. reported such a study on the determination of the optical
properties of nc-Si in the photon energy range 1.1–5.0 eV [63]. nc-Si with a mean size of ~4 nm
embedded in a SiO2 matrix was synthesized by Si+ implantation with a dose of 1 × 1017 atoms/cm2
at the energy of 100 keV into a 550-nm-thick SiO2 film thermally grown on a p-type Si substrate.
Stable nc-Si was formed after thermal annealing at 1000°C for 30 min in nitrogen gas. The dielec-
tric function of the nc-Si is found to be well described by both the Lorentz oscillator model [70] and
the Forouhi–Bloomer (FB) model [71].
The dielectric functions of the nc-Si obtained from the SE analysis is shown in Figure 6.25 [63].
As can be observed in the figure, the overall spectral features of optical properties of the nc-Si
are similar to those of bulk crystalline Si. However, the nc-Si shows a significant reduction in the
dielectric function compared to bulk crystalline Si.
Using the same SE methodology, Ding et al. obtained the dielectric functions and optical con-
stants of the nc-Si with various sizes embedded in SiO2, which are shown in Figures 6.26 and 6.27,
respectively [66]. As can be observed in the two figures, nc-Si exhibits a significant reduction in
the dielectric functions and optical constants with respect to bulk crystalline silicon, and the size
of nc-Si has a large influence on both the magnitude and shape of the spectra of the dielectric func-
tions and optical constants. It is well known that dielectric functions of a crystalline material are
closely associated with its electronic band structure, which is often described by the joint DOS [66].
The critical points observed in the dielectric spectra of crystalline material are believed to originate
from singularities in the joint DOS. As can be observed in Figure 6.26, the imaginary part of the
dielectric function of bulk crystalline silicon has main peaks at the transition energies E1 (~3.4 eV)
and E2 (~4.3 eV) as its critical points. The main peaks are responsible for the high absorption of the
light wave by the material. In the case of nc-Si embedded in SiO2, the magnitude of the imaginary
part of the dielectric functions decreases with decreasing nc-Si size. There is not much change in
the transition energy of E1 for different nc-Si sizes, but a large red shift (~0.3 eV) in the transition
energy of E2 is observed for the nc-Si sizes of 4.6 and 5.3 nm [66].
The static dielectric constant of the widely distributed nc-Si embedded in SiO2 matrix is found to
be 9.7, which is obtained from the SE analysis based on the four-term FB model by setting the photon
energy to zero [60]. Compared to the static dielectric constant (11.4) of bulk crystalline silicon, there
is a significant reduction in that of nc-Si. It has been well established that reduction of the static
dielectric constant becomes significant as the size of the quantum-confined physical systems, such as
quantum dots and wires, approaches the nanometer range [15,32,33,66,72]. However, the origin of the
reduction in static dielectric constant with the size is still not fully understood. It is often attributed
to the opening of the gap, which should lower the polarizability, but it has also been shown that the
240 Semiconductor Nanocrystals and Metal Nanoparticles

60

40

20
ε1

–20
Bull crystalline Si
Si nanocrystals (Lorentz oscillator nodel)
50 Si nanocrystals (Forouhi–Bloomer model)

40

30
ε2

20

10

1 2 3 4 5
Energy (eV)

FIGURE 6.25 Real (ε1) and imaginary (ε2) part of the complex dielectric function of the nc-Si obtained from the
spectral fittings based on the Lorentz oscillator model and the FB model. The dielectric function of bulk crystalline
silicon is also included for comparison. (From Ding, L. et al., Phys. Rev. B, 72, 125419, 2005. With permission.)

reduction is due to the breaking of polarizable bonds at the surface and not to the opening of the
bandgap induced by the confinement [34]. Taking the screening effect by the medium into account,
Equation 6.6 can be expressed for the screening static dielectric constant of the nc-Si [15,60,66], as

e r (¥ ) - 1
er ( D ) = 1 + (6.9)
1 + (6.9 /R)1.37

where εr(∞) (=11.4) is the static dielectric constant of bulk crystalline silicon, and R is the radius
of nc-Si (in Å). The static dielectric constant of nc-Si with the diameter of 4.5 nm calculated with
Equation 6.9 is 9.7, which is equal to the value obtained from the SE analysis based on the FB model
[60]. Figure 6.28 shows the size dependence of the static dielectric constant of silicon nanocrystal
calculated with Equation 6.9.
The FB model can also yield the bandgap of the isolated nc-Si embedded in the SiO2 matrix.
The bandgap of the nc-Si obtained from the SE spectral fitting based on the FB model is 1.74 eV,
which is ~0.6 eV larger than the bandgap of bulk crystalline Si [63]. On the other hand, based on the
plot of (αE)γ versus E, where α = 4πk/λ (where k is the extinction coefficient of the nc-Si obtained
Electronic and Optical Properties of Si and Ge Nanocrystals 241

40

Real part of dielectric functions

20

Sample 1, Dno-Si = 4.6 nm

Sample 2, Dno-Si = 5.3 nm
–20 Sample 3, Dno-Si = 5.8 nm
Sample 4, Dno-Si = 6.3 nm
Sample 5, Dno-Si = 7.6 nm
Imaginary part of dielectric functions

40 Bulk crystalline silicon

30

20

10

0
1 2 3 4 5
Photon energy (eV)

FIGURE 6.26 Real (εr) and imaginary (εi) part of the complex dielectric function of the nc-Si with various
sizes obtained from the spectral fittings. The dielectric function of bulk crystalline silicon is also included for
comparison. (From Ding, L. et al., J. Appl. Phys., 101, 103525, 2007. With permission.)

from the SE analysis and λ is the wavelength) is the absorption coefficient of the nc-Si and E is the
photon energy, one can examine whether the nc-Si is a direct ( g = 2) or indirect (γ = 1/2) bandgap
semiconductor [60,73]. The Tauc plots shown in [73] indicate that nc-Si has an indirect bandgap
structure with the value of 1.75 eV, which is almost the same as the value (1.74 eV) obtained from
the SE analysis based on the FB model. The bandgap obtained here is in good agreement with that
obtained from the first-principles calculation of the optical gap of Si nanocrystals based on quantum
confinement [8]. A fit using Equation 6.3 to the calculation result (Figure 6.7) obtained by Öğüt
et al. [8] yields the band gap expansion [66]
C
Eg (D) - Eg 0 = (6.10)
Dn
where
D is the nanocrystal size (in nm)
Eg(D) is the bandgap (in eV) of the nanocrystal
Eg0 = 1.12 eV is the bandgap of bulk crystalline Si
C = 3.9
n = 1.22
242 Semiconductor Nanocrystals and Metal Nanoparticles

5
Refractive index
4

2 Sample 1, Dno-Si = 4.6 nm

Sample 2, Dno-Si = 5.3 nm
Sample 3, Dno-Si = 5.8 nm
Sample 4, Dno-Si = 6.3 nm
5 Sample 5, Dno-Si = 7.6 nm
Bulk crystalline silicon
Extinction coefficient

0
400 600 800 1000
Wavelength (nm)

FIGURE 6.27 Refractive index (n) and extinction coefficient (k) of nc-Si of various sizes. The optical
constants of bulk crystalline silicon are also included for comparison. (From Ding, L. et al., J. Appl. Phys.,
101, 103525, 2007. With permission.)

12

10
εr

6
0 20 40 60 80 100 120
R (Å)

FIGURE 6.28 Size dependence of the static dielectric constant of silicon nanocrystal calculated with Equation 6.9.
Electronic and Optical Properties of Si and Ge Nanocrystals 243

For the nc-Si size of ~4.2 nm, Equation 6.10 gives a bandgap expansion of ~0.67 eV, which agrees
with the values mentioned previously [73].
From the SE spectral fitting based on the FB model, Ding et al. obtained the bandgaps of
nc-Si with various sizes embedded in SiO2 [66]. nc-Si exhibits a large expansion in the band gap
as compared to that of the bulk crystalline silicon, and the bandgap of the nc-Si increases when
the nc-Si size is reduced. The bandgap expansion as a function of the nc-Si size is shown in
Figure 6.29 [66]. For comparison, a calculation of the bandgap expansion with Equation 6.10 is
also shown in the figure. A good agreement can be seen in the figure. The bandgap expansion is
the most direct evidence of quantum confinement effect of nc-Si. The bandgap expansion of nc-Si
and its dependence on the nanocrystal size have been demonstrated by many theoretical calcula-
tions, as discussed in Section 6.2.

6.3.2 nc-Si Embedded in Si3N4 Thin Films

Cen et al. reported an SE study on the optical properties of Si nanocrystals dispersed in a
silicon nitride (Si3N4) thin film synthesized with the Si ion implantation [68]. The study is sum-
marized here.
A Si3N4 thin film with thickness of ~120 nm was deposited by low-pressure chemical vapor depo-
sition (LPCVD) onto a 30 nm SiO2 thin film (i.e., a stress-relief oxide layer) thermally grown on a
Si (100) substrate. The Si3N4 film was subsequently implanted with Si ions with a dose of 3.5 × 1016
atoms/cm2 at the energy of 30 keV. Afterward, thermal annealing was carried out in nitrogen
ambient at different temperatures (i.e., 800°C, 900°C, 1000°C, and 1100°C) for 1 h. The multi-
layer optical model shown in Figure 6.30a was used in the SE analysis [68]. Transmission electron
microscopy (TEM) measurement showed that the nc-Si size was in the order of 2 nm. Figure 6.30b
shows a typical TEM image.

0.8 Experiment
Calculation
Band-gap expansion (eV)

0.6

0.4

0.2
4 5 6 7 8
nc-Si size (nm)

FIGURE 6.29 Bandgap expansion of nc-Si as a function of the crystal size. The symbols are the data
obtained from the SE spectral fitting based on the FB model, and the line represents the calculation with
Equation 6.10. (From Ding, L. et al., J. Appl. Phys., 101, 103525, 2007. With permission.)
244 Semiconductor Nanocrystals and Metal Nanoparticles

20
Volume fraction (%) of implanted Si

Pure silicon nitride

15

Si substrate
SiO2
10

0
0 20 40 60 80 100 120
(a) Depth (nm)

5 nm

(b)

FIGURE 6.30 (a) Multilayer model used in the SE analysis and (b) TEM image of the Si nanoparticles or
nanoclusters embedded in the silicon nitride thin film for the sample annealed at 1100°C. (From Cen, Z.H.
et al., Appl. Phys. Lett., 93, 023122, 2008. With permission.)

In the SE analysis, the Kramers–Kronig-consistent Tauc–Lorentz (TL) model was used [68]. In
the model, the imaginary part (ε2) of dielectric function was written as
ì0, for 0 < E £ Eg
ïï
e2 ( E ) = í é AE C ( E - E )2 1ù (6.11)
0 g
ïê 2 × ú , for E > Eg
ïî ë ( E - E0 ) + C E E û
2 2 2 2

where
A is the amplitude of a Lorentz oscillator
E0 is the peak transition energy
C is the broadening term
Eg is the bandgap energy
Electronic and Optical Properties of Si and Ge Nanocrystals 245

The real part (ε1) of dielectric function is obtained by Kramers–Kronig integration of ε2 with
a fitting constant ε1(∞). In the SE spectral fitting, A, E 0, C, and Eg as well as the thicknesses of
both the silicon nitride layer and the SiO2 layer are the fitting parameters. Figure 6.31 shows the
spectral fittings for the sample annealed at 1100°C [68]. The dielectric functions of the nc-Si dis-
tributed in Si3N4 obtained from the SE fittings are shown in Figure 6.32 for different annealing
temperatures [68].
It is observed from Figure 6.32 that the dielectric functions for all the samples are generally
similar. The spectral features of the dielectric functions are also similar to those of amorphous Si.
The broadened peak structures, particularly the broad peaks of the as-implanted sample, suggest
that the implanted Si is in an amorphous or disordered state to a certain extent. The nc-Si distributed
in Si3N4 shows a significant reduction in the dielectric functions as compared to bulk crystalline
silicon, and the situation is similar to that of nc-Si distributed in SiO2 [68].

6.3.3 Self-Assembled Ge Nanocrystals

Goh et al. reported a study on the bandgap and optical properties (dielectric functions and optical
constants) of self-assembled Ge nanocrystals (nc-Ge) using SE based on the FB optical dispersion
model [69]. They showed that, as compared to bulk crystalline Ge, the nc-Ge exhibited a large
bandgap expansion and a significant reduction in the dielectric function. The information presented
in this section is adapted from [69].
The synthesis procedure of the self-assembled nc-Ge is as follows. A 3-nm-thick Ge layer
was deposited onto a 3 nm SiO2 layer, which was thermally grown on p-type 〈100〉 Si substrate,
using electron beam evaporation under the base pressure of 5 × 10 −6 mbar with the deposition
rate of 0.07 nm/s. The sample underwent a rapid thermal annealing (RTA) in nitrogen ambient
at 450°C for 50 s to form nc-Ge. With the existence of the SiO2 layer, possible reaction between
the deposited Ge layer and the Si substrate during the annealing process could be avoided. The
formation of the nc-Ge is clearly demonstrated by the comparison of the atomic force microscopy
(AFM) image (planar view) of the Ge layer before the RTA with that after the RTA, as shown in
Figure 6.33a and b [69]. In addition, as revealed by the TEM image shown in Figure 6.33c, the

60
Measured Ψ at 75° 180
Measured Ψ at 70°
50 160
Measured Δ at 75°
Measured Δ at 70° 140
Model
40
120

100
Ψ (°)

Δ (°)

30
80
20 60

40
10
20

0 0
300 400 500 600 700 800 900 1000 1100
Wavelength (nm)

FIGURE 6.31 SE spectral fittings for the sample annealed at 1100°C. (From Cen, Z.H. et al., Appl. Phys.
Lett., 93, 023122, 2008. With permission.)
246 Semiconductor Nanocrystals and Metal Nanoparticles

20

10
ε1

As-implanted
800°C
900°C
1000°C
1100°C

0
(a)

As-implanted
20 800°C
900°C
1000°C
1100°C
ε2

10

1 2 3 4 5
(b) E (eV)

FIGURE 6.32 Annealing effect on the dielectric functions of the implanted Si. (a) Real part (ε1) and (b)
imaginary part (ε2) of the dielectric functions. (From Cen, Z.H. et al., Appl. Phys. Lett., 93, 023122, 2008.
With permission.)

self-assembled nc-Ge has a dome shape with average size of ~6 nm in height and ~13 nm in diam-
eter [69]. The formation of nc-Ge is explained by stress relaxation, dispersion force, and surface
energy minimization during the annealing process [69].
In the SE analysis for the nc-Ge on the SiO2 layer thermally grown on the Si substrate, the four-phase
model, that is, air/nc-Ge layer/SiO2 layer/Si substrate, which is shown in the inset of Figure 6.34, was
used. The nc-Ge layer comprised of voids and nc-Ge. Its effective dielectric function εi (= N i2, where
Ni is the complex refractive index of the layer) was modeled with the Bruggeman effective medium
approximation (EMA) [69]:

e nc-Ge - ei e -e
f + void i (1 - f ) = 0 (6.12)
e nc-Ge + 2ei e void + 2ei
Electronic and Optical Properties of Si and Ge Nanocrystals 247

5.0 nm 20.0 nm

2.5 nm 10.0 nm

0.0 nm 0.0 nm

(a) (b)

20.0 nm

20 nm

Capping layer 10.0 nm

SiO2 buffer
0.0 nm
layer
Ge
nanocrystals

Si substrate

(c) (d)

FIGURE 6.33 AFM images (plane view) of the 3 nm Ge thin film sample on Si substrate (a) before anneal-
ing and (b) after the annealing to form nc-Ge. (c) TEM image of the nc-Ge formed on Si substrate (note that
a SiO2 capping layer was used for the TEM experiment). (d) Planar view AFM image of the nc-Ge formed on
fused silica substrate. (From Goh, E.S.M. et al., J. Appl. Phys., 109, 064307, 2011. With permission.)

where εnc‐Ge ( = (nnc-Ge - knc-Ge )2 , nnc-Ge and knc-Ge are the refractive index and extinction coefficient
of the nc-Ge, respectively) is the effective dielectric function of the nc-Ge, f is the volume fraction
of the voids in the layer, and the dielectric function (εvoid) of the voids is 1. It was found that the FB
model satisfactorily describes the optical constants of the nc-Ge [69].
Figure 6.34 shows the fittings based on the Bruggeman EMA and the FB model to the SE
spectra measured at three angles of incidence [69]. As can be seen in this figure, the spectra
of both Ψ and Δ in the whole wavelength range for all the three angles of incidence fit excel-
lently. The volume fraction of nc-Ge in the nc-Ge layer and the effective thickness of the
layer obtained from the fittings are 55.79% and ~6 nm, respectively. The result agrees with
the TEM measurement, from which the effective thickness and the volume fraction were esti-
mated based on the average diameter and height of nc-Ge. This indicates that the FB model
describes the dispersion of the optical constants of the nc-Ge well and the fitting procedure
is effective [69].
248 Semiconductor Nanocrystals and Metal Nanoparticles

30
Θo
25

3 nm SiO2
Ψ (degree) 20
p-Si substrate
15

10

5 70°
75°
80°
160 Fitting

140

120
Δ (degree)

100

80

60

40

20

0
400 600 800 1000 1200
Wavelength (nm)

FIGURE 6.34 Spectral fittings of (a) Ψ and (b) Δ at the angles of incidence of 70°, 75°, and 80°. The inset
shows the four-phase model used in the SE analysis. (From Goh, E.S.M. et al., J. Appl. Phys., 109, 064307,
2011. With permission.)

Figure 6.35 shows the real and imaginary parts of the complex dielectric function of the nc-Ge
[69]. For comparison, the dielectric function of bulk crystalline Ge is also included in the figure.
A strong reduction in the dielectric function is observed for the nc-Ge with respect to bulk crystal-
line Ge. For example, at 400 nm, the real and the imaginary parts of dielectric function of nc-Ge
are lower than those of bulk crystalline Ge by 30% and 76%, respectively.
Perhaps the most important parameter obtained from the spectral fitting based on the FB model
is the bandgap of nc-Ge. This value of the nc-Ge formed on Si substrate obtained is 0.92 eV, which
is consistent with the absorption measurement of the nc-Ge formed on a fused silica substrate [69].
Figure 6.36 shows the Tauc plot of (αE)1/2 versus E obtained from the absorption measurement [69].
The bandgap of the nc-Ge formed on fused silica substrate obtained from the linear extrapolation
of the Tauc plot is 0.86 eV. Therefore, it could be concluded that the nc-Ge has a bandgap expan-
sion of ~0.2 eV compared to bulk crystalline Ge (note that the bandgap of bulk crystalline Ge is
0.66 eV). On the other hand, the good linearity of the Tauc plot of (αE)1/2 versus E could suggests
that nc-Ge has an indirect bandgap structure.
Electronic and Optical Properties of Si and Ge Nanocrystals 249

30

Real part of dielectric function

25

20

15

10

5 Bulk Ge
25 nc-Ge
Imaginary part of dielectric function

20

15

10

1 2 3
Energy (eV)

FIGURE 6.35 Real (ε1) and imaginary (ε2) part of the dielectric function of nc-Ge, obtained from the spec-
tral fittings based on the FB model. The dielectric function of bulk crystalline Ge is included for comparison.
(From Goh, E.S.M. et al., J. Appl. Phys., 109, 064307, 2011. With permission.)

6.3.4 Ge Nanocrystals Embedded in SiO2 Matrices

Takeoka et al. reported an experimental study of the PL of nc-Ge with 0.9–5.3 nm average diam-
eter (D) in the near-infrared region [44]. Ge nanocrystals were fabricated by rf cosputtering of Ge
and SiO2 and post-annealing at 800°C for 30 min. The size nc-Ge was controlled by changing the
volume fraction (fGe) of Ge in the films. Figure 6.37 shows a typical cross-sectional high-resolution
TEM (HRTEM) image of the sample with fGe = 3.6%. Figure 6.38 shows the PL spectra for vari-
ous nc-Ge sizes. As can be observed in the figure, the sample with D = 5.3 nm shows a PL peak at
~0.88 eV, which is slightly larger than the bandgap of bulk Ge crystal. As D decreases, the PL peak
shifts monotonously to higher energies and reaches ~1.54 eV as D decreases to 0.9 nm [44]. The
strong size dependence of the PL peak energy indicates that the bandgap of the nc-Ge widens when
the nc-Ge size decreases as a result of quantum confinement and that the light emission originates
from the recombination of electron–hole pairs confined in nc-Ge.
250 Semiconductor Nanocrystals and Metal Nanoparticles

500 Experiment
Linear extrapolation

400

(αE)1/2(eV1/2cm–1/2)
300

200

100

0
1.0 1.5 2.0 2.5
Energy (eV)

FIGURE 6.36 Tauc plot of (αE)1/2 versus E yielded from the absorption measurement for the nc-Ge formed
on a fused silica substrate. The bandgap of the nc-Ge obtained from the linear extrapolation of the Tauc plot
is 0.86 eV. (From Goh, E.S.M. et al., J. Appl. Phys., 109, 064307, 2011. With permission.)

m
3n
0.3
m
3n
0.3

5 nm

FIGURE 6.37 Cross-sectional HRTEM image of the sample with fGe = 3.6%. Lattice fringes correspond
to {111} planes of Ge. The diamond structure can clearly be seen. (From Takeoka, S. et al., Phys. Rev. B, 58,
7921, 1998. With permission.)
Electronic and Optical Properties of Si and Ge Nanocrystals 251

nc-Ge in SiO2
Average
300 K diameter
5.3 nm X 16

4.7 nm X 8.3

4.3 nm X 4.8

PL intensity (a.u.)
3.8 nm X 4.2
3.2 nm X 2.4

2.3 nm X1
1.7 nm X 1.2

1.2 nm
X 1.5
0.9 nm
X 2.1

SiO2

1.0 1.5 2.0 2.5

Photon energy (eV)

FIGURE 6.38 Dependence of PL spectra on the average diameter of nc-Ge. A PL spectrum of a SiO2 film is
also shown. (From Takeoka, S. et al., Phys. Rev. B, 58, 7921, 1998. With permission.)

ACKNOWLEDGMENT
The author wishes to thank Y. Liu, L. Ding, Z. H. Cen, E. S. M. Goh, and others for their contributions
to the results from the group at Nanyang Technological University presented here.

REFERENCES
1. Zdetsis, A. D. 2006. Optical and electronic properties of small size semiconductor nanocrystals and
nanoclusters. Rev. Adv. Mater. Sci. 11: 56–78.
2. Garoufalis, C. S. and Zdetsis, A. D. 2006. High accuracy calculations of the optical gap and absorption
spectrum of oxygen contaminated Si nanocrystals. Phys. Chem. Chem. Phys. 8: 808–813.
3. Delley, B. and Steigmeier, E. F. 1995. Size dependence of band gaps in silicon nanostructures. Appl.
Phys. Lett. 67: 2370–2372.
4. Delley, B. and Steigmeier, E. F. 1993. Quantum confinement in Si nanocrystals. Phys. Rev. B 47:
1397–1400.
5. Weissker, H.-Ch., Furthmüller, J., and Bechstedt, F. 2002. Optical properties of Ge and Si nanocrystal-
lites from ab initio calculations. I. Embedded nanocrystallites. Phys. Rev. B 65: 155327(1-9).
6. Weissker, H.-Ch., Furthmüller, J., and Bechstedt, F. 2002. Optical properties of Ge and Si nanocrystal-
lites from ab initio calculations. II. Hydrogenated nanocrystallites. Phys. Rev. B 65: 155328(1-7).
7. Vasiliev, I., Öğüt, S., and Chelikowsky, J. R. 2001. Ab initio absorption spectra and optical gaps in
­nanocrystalline silicon. Phys. Rev. Lett. 86: 1813–1816.
8. Öğüt, S., Chelikowsky, J. R., and Louie, S. G. 1997. Quantum confinement and optical gaps in Si
­nanocrystals. Phys. Rev. Lett. 79: 1770–1773.
9. Delerue, C., Allan, G., and Lannoo M. 1993. Theoretical aspects of the luminescence of porous silicon.
Phys. Rev. B 48: 11024–11036.
10. Delerue, C., Lannoo, M., and Allan, G. 2000. Excitonic and quasiparticle gaps in Si nanocrystals. Phys.
Rev. Lett. 84: 2457–2460.
252 Semiconductor Nanocrystals and Metal Nanoparticles

11. Godby, R. W. and White, I. D. 1998. Density-relaxation part of the self-energy. Phys. Rev. Lett. 80:
3161.
12. Franceschetti, A., Wang, L. W., and Zunger, A. 1999. Comment on “quantum confinement and optical
gaps in Si nanocrystals. Phys. Rev. Lett. 83: 1269.
13. Baierle, R. J., Caldas, M. J., Molinari, E., and Ossicini, S. 1997. Optical emission from small Si particles.
Solid State Commun. 102: 545–549.
14. Palummo, M., Onida, G., and Sole, R. D. 1999. Optical properties of germanium nanocrystals. Phys.
Stat. Sol. (a) 175: 23–31.
15. Wang, L.-W. and Zunger, A. 1994. Dielectric constants of silicon quantum dots. Phys. Rev. Lett. 73:
1039–1042.
16. Zunger, A. 2001. Pseudopotential theory of semiconductor quantum dots. Phys. Stat. Sol. (b) 224:
727–734.
17. Degoli, E., Cantele, G., Luppi, E. et al. 2004. Ab initio structural and electronic properties of hydroge-
nated silicon nanoclusters in the ground and excited state. Phys. Rev. B 69: 155411(1-10).
18. Ramos, L. E., Weissker, H.-Ch., Furthmüller, J., and Bechstedt, F. 2005. Optical properties of Si and Ge
nanocrystals: Parameter-free calculations. Phys. Stat. Sol. (b) 242: 3053–3063.
19. Vincent, J. E., Kim, J., and Martin, R. M. 2007. Quantum Monte Carlo calculations of the optical gaps
of Ge nanoclusters using core-polarization potentials. Phys. Rev. B 75: 045302(1-10).
20. Weissker, H.-Ch., Furthmüller, J., and Bechstedt, F. 2003. Oscillator strengths and excitation energies of
Ge and Si nanocrystals from ab initio supercell calculations. Mater. Sci. Eng. B 101: 39–42.
21. Luppi, E., Iori, F., Magri, R. et al. 2007. Excitons in silicon nanocrystallites: The nature of lumines-
cence. Phys. Rev. B 75: 033303(1-4).
22. Luppi, M. and Ossicini, S. 2005. Ab initio study on oxidized silicon clusters and silicon nanocrystals
embedded in SiO2: Beyond the quantum confinement effect. Phys. Rev. B 71: 035340(1-15).
23. Vasiliev, I., Chelikowsky, J. R., and Martin, R. M. 2002. Surface oxidation effects on the optical proper-
ties of silicon nanocrystals. Phys. Rev. B 65: 121302(1-4).
24. Guerra, R. and Ossicini, S. 2010. High luminescence in small Si/SiO2 nanocrystals: A theoretical study.
Phys. Rev. B 81: 245307(1-6).
25. Guerra, R., Degoli, E., and Ossicini, S. 2009. Size, oxidation, and strain in small Si/SiO2 nanocrystals.
Phys. Rev. B 80: 155332(1-5).
26. Guerra, R., Marri, I., Magri, R. et al. 2009. Silicon nanocrystallites in a SiO2 matrix: Role of disorder
and size. Phys. Rev. B 79: 155320(1-9).
27. Guerra, R., Marri, I., Magri, R. et al., 2009. Optical properties of silicon nanocrystallites in SiO2 matrix:
Crystalline vs. amorphous case. Superlatt. Microstruct. 46: 246–252.
28. Francheschetti, A. and Zunger, A. 1997. Direct pseudopotential calculation of exciton coulomb and
exchange energies in semiconductor quantum dots. Phys. Rev. Lett. 78: 915–918.
29. Lehtonen, O. and Sundholm, D. 2009. Computational studies of free-standing silicon nanoclusters.
In Silicon Nanophotonics. Khriachtchev, L. (ed.), pp. 61–88. World Scientific Publishing, Singapore.
30. Sharma, A. C. 2006. Size-dependent energy band gap and dielectric constant within the generalized
Penn model applied to a semiconductor nanocrystallite. J. Appl. Phys. 100: 084301(1-8).
31. Tsu, R. and Babić, D. 1994. Doping of a quantum dot. Appl. Phys. Lett. 64: 1806–1808.
32. Tsu, R., Babić, D., and Ioriatti, Jr. L. 1997. Simple model for the dielectric constant of nanoscale silicon
particle. J. Appl. Phys. 82: 1327–1329.
33. Wang, L. W. and Zunger, A. 1996. Pseudopotential calculations of nanoscale CdSe quantum dots. Phys.
Rev. B 53: 9579–9582.
34. Delerue, C., Lannoo, M., and Allen, G. 2003. Concept of dielectric constant for nanosized systems.
Phys. Rev. B 68: 15411(1-4).
35. Walter, J. P. and Cohen, M. L. 1970. Wave-vector-dependent dielectric function for Si, Ge, GaAs, and
ZnSe. Phys. Rev. B 2: 1821–1826.
36. Lannoo, M., Delerue, C., and Allan, G. 1995. Screening in semiconductor nanocrystallites and its
consequences for porous silicon. Phys. Rev. Lett. 74: 3415–3418.
37. Wolkin, M. V., Jorne, J., Fauchet, P. M., Allan, G., and Delerue, C. 1999. Electronic states and lumines-
cence in porous silicon quantum dots: The role of oxygen. Phys. Rev. Lett. 82: 197–200.
38. Garrido, B., Lopez, M., Gonzalez, O., Perez-Rodriguez, A., Morante, J. R., and Bonafos, C. 2000.
Correlation between structural and optical properties of Si nanocrystals embedded in SiO2: The mecha-
nism of visible light emission. Appl. Phys. Lett. 77: 3143–3145.
39. Pavesi, L., Negro, L. D., Mazzoleni, C., Franzò, G., and Priolo, F. 2000. Optical gain in silicon nano-
crystals. Nature (London) 408: 440–444.
Electronic and Optical Properties of Si and Ge Nanocrystals 253

40. Negro, L. D., Cazzanelli, M., Pavesi, L., Ossicini, S., Pacifici, D., Franzò, G., Priolo, F., and Iacona, F.
2003. Dynamics of stimulated emission in silicon nanocrystals. Appl. Phys. Lett. 82: 4636–4638.
41. Wilcoxon, J. P., Samara, G. A., and Provencio, P. N. 1999. Optical and electronic properties of Si nano-
clusters synthesized in inverse micelles. Phys. Rev. B 60: 2704–2714.
42. Wilcoxon, J. P., Provencio, P. P., and Samara, G. A. 2001. Synthesis and optical properties of colloidal
germanium nanocrystals. Phys. Rev. B 64: 035417(1-9).
43. Heath, J. R., Shiang, J. J., and Alivisatos, A. P. 1994. Germanium quantum dots: Optical properties and
synthesis. J. Chem. Phys. 101: 1607–1615.
44. Takeoka, S., Fujii, M., Hayashi, S., and Yamamoto, K. 1998. Size-dependent near-infrared photolumi-
nescence from Ge nanocrystals embedded in SiO2 matrices. Phys. Rev. B 58: 7921–7925.
45. Taraschi, G., Saini, S., Fan, W. W., Kimerling, L. C., and Fitzgerald, E. A. 2003. Nanostructure and
infrared photoluminescence of nanocrystalline Ge formed by reduction of Si0.75Ge0.25O2/Si0.75Ge0.25
using various H2 pressures. J. Appl. Phys. 93: 9988–9996.
46. Konchenko, A., Nakayama, Y., Matsuda, I., Hasegawa, S., Nakamura, Y., and Ichikawa, M. 2006.
Quantum confinement observed in Ge nanodots on an oxidized Si surface. Phys. Rev. B 73: 113311(1-4).
47. Niquet, Y., Allan, G., Delerue, C., and Lannoo, M. 2000. Quantum confinement in germanium nano-
crystals. Appl. Phys. Lett. 77: 1182–1183.
48. Reboredo, F. A. and Zunger, A. 2000. L-to-X crossover in the conduction-band minimum of Ge quan-
tum dots. Phys. Rev. B 62: R2275–R2278.
49. Reboredo, F. A. and Zunger, A. 2001. Surface-passivation-induced optical changes in Ge quantum dots.
Phys. Rev. B 63: 235314(1-7).
50. Melnikov, D. V. and Chelikowsky, J. R. 2003. Absorption spectra of germanium nanocrystals. Solid
State Commun. 127: 361–365.
51. Melnikov, D. V. and Chelikowsky, J. R. 2004. Electron affinities and ionization energies in Si and
Ge nanocrystals. Phys. Rev. B 69: 113305(1-4).
52. Weissker, H.-Ch., Furthmüller, J., and Bechstedt, F. 2003. Structural relaxation in Si and Ge nanocrys-
tallites: Influence on the electronic and optical properties. Phys. Rev. B 67: 245304(1-7).
53. Weissker, H.-Ch., Furthmüller, J., and Bechstedt, F. 2004. Structure- and spin-dependent excitation
energies and lifetimes of Si and Ge nanocrystals from ab initio calculations. Phys. Rev. B 69: 15310(1-8).
54. Tsolakidis, A. and Martin, R. M. 2005. Comparison of the optical response of hydrogen-passivated
germanium and silicon clusters. Phys. Rev. B 71: 125319(1-8).
55. Nesher, G., Kronik, L., and Chelikowsky, J. R. 2005. Ab initio absorption spectra of Ge nanocrystals.
Phys. Rev. B 71: 035344(1-5).
56. Garoufalis, C. S., Skaperda, M. S., and Zdetsis, A. D. 2005. The optical gap of small Ge nanocrystals.
J. Phys.: Conf. Ser. 10: 97–100.
57. Kanemitsu, Y., Uto, H., Masumoto, Y., and Maeda, Y. 1992. On the origin of visible photoluminescence
in nanometer size Ge crystallites. Appl. Phys. Lett. 61: 2187–2189.
58. Amans, D., Callard, S., Gagnaire, A. et al. 2003. Ellipsometric study of silicon nanocrystal optical
constants. J. Appl. Phys. 93: 4173–4179.
59. Lee, K.-J., Kang, T.-D., Lee, H. et al. 2005. Optical properties of SiO2/nanocrystalline Si multilayers
studied using spectroscopic ellipsometry. Thin Solid Films 476: 196–200.
60. Chen T. P. and Ding, L. 2010. Optical and optoelectronic properties of silicon nanocrystals embedded
in SiO2 matrix. In Nanostructured Thin Films and Coatings: Functional Properties, Zhang, S. (ed.),
pp. 113–165. CRC Press, Boca Raton, FL.
61. Cen Z. H. 2011. Studies of optical properties and light emission of silicon-rich silicon nitride thin films.
PhD thesis, Nanyang Technological University, Singapore.
62. Ding L. 2009. Optical and optoelectronic properties of Si nanocrystals embedded in dielectric matrix.
PhD thesis, Nanyang Technological University, Singapore.
63. Ding, L., Chen, T. P., Liu, Y., Ng, C. Y., and Fung, S. 2005. Optical properties of silicon nanocrystals
embedded in a SiO2 matrix. Phys. Rev. B 72: 125419(1-7).
64. Ding, L., Chen, T. P., Liu, Y. et al. 2005. Thermal annealing effect on the band gap and dielectric
functions of silicon nanocrystals embedded in SiO2 matrix. Appl. Phys. Lett. 87: 121903(1-3).
65. Chen, T. P., Liu, Y., Tse, M. S. et al. 2003. Dielectric functions of Si nanocrystals embedded in a SiO2
matrix. Phys. Rev. B 68: 153301(1-4).
66. Ding, L., Chen, T. P., Liu, Y. et al. 2007. Influence of nanocrystal size on optical properties of Si
nanocrystals embedded in SiO2 synthesized by Si ion implantation. J. Appl. Phys. 101: 103525(1-6).
67. Ding, L., Chen, T. P., Wong, J. I. et al. 2006. Dielectric functions of densely stacked Si nanocrystal layer
embedded in SiO2 thin films. Appl. Phys. Lett. 89: 251910(1-3).
254 Semiconductor Nanocrystals and Metal Nanoparticles

68. Cen, Z. H., Chen, T. P., Ding, L. et al. 2008. Annealing effect on the optical properties of implanted
silicon in a silicon nitride matrix. Appl. Phys. Lett. 93: 023122(1-3).
69. Goh, E. S. M., Chen, T. P., Huang, S. F., Liu, Y. C., and Sun, C. Q. 2011. Band gap expansion and dielec-
tric suppression of self-assembled Ge nanocrystals. J. Appl. Phys. 109: 064307(1-4).
70. Azzam, R. M. A. and Basharra, N. M. 1977. Ellipsometry and Polarized Light. Amsterdam, the
Netherlands: North-Holland.
71. Forouhi, A. R. and Bloomer, I. 1988. Optical properties of crystalline semiconductors and dielectrics.
Phys. Rev. B 38: 1865–1874.
72. Penn, D. R. 1962. Wave-number-dependent dielectric function of semiconductors. Phys. Rev. 128:
2093–2097.
73. Ding, L., Chen, T. P., Liu, Y. et al. 2008. Evolution of photoluminescence mechanisms of Si+-implanted
SiO2 films with thermal annealing. J. Nanosci. Nanotechnol. 8: 3555–3560.
 7 Light Emission Properties of
Si Nanocrystals Embedded
in a Dielectric Matrix
Tupei Chen

CONTENTS
7.1 Introduction........................................................................................................................... 255
7.2 Fundamental Luminescence Properties of Silicon Nanocrystals.......................................... 257
7.3 Photoluminescence and Electroluminescence from Si+-Implanted SiO2 Thin Films........... 263
7.3.1 Photoluminescence.................................................................................................... 263
7.3.2 Electroluminescence..................................................................................................264
7.4 Photoluminescence and Electroluminescence from Si+-Implanted Si3N4 Thin Films..........266
7.4.1 Photoluminescence....................................................................................................266
7.4.2 Electroluminescence.................................................................................................. 269
7.4.2.1 Fabrication of EL Structures....................................................................... 269
7.4.2.2 EL Properties and Mechanisms.................................................................. 270
7.4.2.3 Annealing Effect......................................................................................... 273
7.4.2.4 Electrically Tunable EL.............................................................................. 277
Acknowledgments........................................................................................................................... 279
References....................................................................................................................................... 279

7.1 INTRODUCTION
A semiconductor could emit photons under either optical or electrical pumping. Photoluminescence
(PL) and electroluminescence (EL) refer to light emission due to the excitation by optical pumping
(i.e., optical radiation) and electrical pumping (i.e., application of electric field or current injection),
respectively. PL and EL are usually concerned with the following optical transitions: (1) interband
transition with intrinsic emission corresponding very close in energy to the bandgap of the semi-
conductor, where phonons and excitons may be involved, and (2) the transition involving states
in the bandgap due to impurities or defects. Not all transitions can occur in the same material or
under the same conditions, and not all transitions are radiative. The light emission process consists
of two optical transitions: the excitation transition, and the recombination transition. In the excita-
tion transition, the electron is excited from the ground state (e.g., a state in the valence band of
the semiconductor) to the excited state (e.g., a state in the conduction band of the semiconductor)
by absorption of a photon under optical pumping, or is electrically injected into the excited state
under electrical pumping. In the radiative recombination transition, the electron in the excited state
recombines with the hole at the ground state to emit a photon. For a given input excitation energy,
the radiative recombination process competes with the nonradiative process (such as the Auger
recombination). To have efficient light emission, the radiative transitions must predominate over
the nonradiative ones [1–3].
The excitation and recombination optical transitions are different in the direct and indirect band-
gap semiconductors. If the conduction band minimum and the valence band maximum have the

255
256 Semiconductor Nanocrystals and Metal Nanoparticles

same k-vectors (wave vectors) in the Brillouin zone, the semiconductor is called a “direct bandgap”
material; and if the k-vectors are different, it is called an “indirect bandgap” material [1].
In the direct bandgap semiconductor, an electron in the valence band can be excited into the
conduction band after the absorption of a photon with an energy equal to or larger than the energy
bandgap (Eg). However, such a vertical absorption cannot happen in the indirect bandgap materials.
To conserve momentum, the photon absorption process must be assisted by either the absorption
or emission of a phonon, because the change in the electron momentum cannot be provided by
the absorption of a photon [2,3]. The minimum photon energy required for the indirect transition
involving absorption of a phonon with energy of Ep is (Eg − Ep), and it is (Eg + Ep) for the indirect
transition involving phonon emission. In a direct bandgap material, any photon with energy equal
to or larger than the bandgap energy can participate in a direct transition and excite an electron
from the top of the filled valence band to one of the states at the bottom of the conduction band.
However, nonvertical absorptions in an indirect bandgap material occur via an intermediate virtual
state, whereby photon absorption is accompanied by either the creation or annihilation of a phonon
to achieve momentum conservation [1]. In a perfect semiconductor, electron–hole pairs thermalize
and accumulate at the conduction and valence band extremes. Fundamental radiative transitions in
a semiconductor are those occurring at or near the band edges, namely, band-to-band or excitonic
transitions. In a direct bandgap material, momentum-conserving transitions connect states with the
same k-value. Accordingly, the emitted photon has a low-energy threshold at hν = Eg [1]. Indirect
band-to-band radiative transitions could be realized by the assistance of phonon emission or absorp-
tion. With the involvement of a phonon, both the requirements of energy and momentum conserva-
tions are met for the band-to-band transitions in the indirect bandgap semiconductor [1]. However,
the light emission efficiency of the indirect bandgap semiconductor would be much lower due to
the need for the involvement of phonons as compared to that of the direct bandgap semiconductor.
Silicon is the most important semiconductor material in microelectronics industry, because
nowadays, silicon devices constitute over 95% of the market of semiconductor devices [1,3]. On the
other hand, as electronic device dimensions become smaller and smaller, the traditional electrical
interconnects that are used for intra- and inter-chip communication become increasingly impracti-
cal due to the RC delay and heat dissipation associated with the metal interconnects. When the
device dimensions decrease to the nanoscale, the device will suffer a lot from the traditional elec-
trical interconnects due to propagation delays, high power consumption, and low bandwidth [4].
Fortunately, optical interconnects would probably provide us with a promising alternative strat-
egy for overcoming these challenges. Since communication via optical interconnects requires an
on-chip emitter and detector, an important challenge on the materials and the integration of pho-
tonic devices into the main-stream Si process has triggered a new research subject, namely, Si
photonics, recently [1,3]. Most of the photonic components fabricated with Si technology have been
demonstrated, such as optical modulators [5,6], switches [7,8], detectors [9,10], and low-loss wave-
guides [11,12]. However, bulk crystalline silicon cannot be used for light-emitting diodes (LEDs)
because it is an indirect bandgap semiconductor and thus the probability for a radiative transition is
very low. In this regard, the most challenging task is to look for Si-based materials with high emis-
sion efficiency for realizing Si-based LEDs.
Many strategies have been proposed to realize Si-based LEDs. The most successful ones are
based on the exploitation of low-dimensional silicon, where the silicon material is nanostrucured
such that its optical and electronic properties are modified by quantum confinement effects [1,3].
Of all the types of low-dimensional silicon, porous Si has received the most intensive research atten-
tion in the early years of exploitation of nanoscaled Si. The discovery of the visible PL from porous
Si in 1990 [13] has triggered large research efforts in realizing luminescence from porous silicon.
However, it is not reliable to utilize porous silicon in optoelectronic devices due to various issues
such as its instability in light emission, structural fragility, and incompatibility with conventional
complementary metal–oxide–semiconductor (CMOS) technology [14–16]. Another nanostructured
silicon material, namely, the silicon nanocrystal (nc-Si), is considered to be a promising material
Light Emission Properties of Si Nanocrystals Embedded in a Dielectric Matrix 257

system for application in Si LEDs [15–17]. In particular, nc-Si embedded in a dielectric matrix such
as SiO2 or Si3N4 thin films is a promising candidate for realizing Si-based LEDs with the advantages
of chemical stability and full compatibility with the mainstream silicon technology that is used in
the microelectronics industry.

7.2 FUNDAMENTAL LUMINESCENCE PROPERTIES OF SILICON NANOCRYSTALS

The band structure of crystalline silicon can be significantly modified when its dimensions are
reduced to the scale of the exciton Bohr radius (~4.9 nm) of bulk crystalline silicon, which is due
to the quantum confinement effect. In bulk crystalline silicon, since the radiative time of indirect
transitions is extremely long and the transport of excitons is efficient, the main decay channel for
free excitons is their capture in bound exciton states or nonradiative recombination. This results in
a very low quantum yield of free exciton emission. The situation changes in a silicon nanocrystal
with a dimension comparable to the exciton Bohr radius [1]. The following changes significantly
affect the luminescence properties of silicon: the spatial confinement in the nanocrystal results
in shifts of both the absorbing and luminescing states to higher energies; the geometrical change
in the nanocrystal causes delocalization of carriers in k-space, thus allowing zero phonon optical
transitions and significantly enhancing the oscillator strength of the zero phonon optical transitions
in extremely small silicon nanocrystals; and due to the better overlap of electron and hole envelope
wavefunctions, a strong enhancement of the electron–hole (e–h) exchange interaction would be
expected [2].
Spatial localization of the electron and hole wavefunctions in a Si nanocrystal of nanometer scale
increases their spread in the momentum space, thus increasing their overlap in the Brillouin zone.
Due to the spatial confinement of the electrons and holes in the nanocrystal, their motion and energy
states will be quantized in three dimensions. As a result of the broadening of the electron–hole pair
state in momentum space, the momentum conservation rule is relaxed for the radiative transitions
in Si nanocrystals, and more phonons with satisfied momentum can be involved in the radiative
transitions. Thus, the probability of radiative recombination is increased with the lifetime reduced
to microseconds or even nanoseconds. Moreover, the bandgap of Si nanocrystals expands (i.e., the
nc-Si bandgap is larger than that of bulk crystalline Si) as a result of the quantum confinement
effect. Therefore, the luminescence wavelength can be manipulated from near-infrared (NIR) (the
energy close to 1.12 eV, the bandgap of bulk crystalline Si) to the visible range by controlling the
size of the Si nanocrystal [2,3].
Light emission from Si nanocrystal is frequently concerned with the system of Si nanocrystal
dispersed in a dielectric matrix such as SiO2 or Si3N4. For such system, besides the quantum confine-
ment effect of the Si nanocrystal, the interfaces between the Si nanocrystals and the host dielectric
matrix play a significant role in the light emission because of the large surface-to-volume ratio of
Si nanocrystal and the existence of interface states. Therefore, effects related to both the interfaces
and the matrix, such as surface passivation and the luminescence centers at the interfaces or even
in the matrix, should be taken into account for the light emission from Si nanocrystal embedded in
a dielectric matrix.
In spite of the intensive research progress made toward understanding the nature of efficient
light emission from a silicon nanocrystal, the mechanism is still controversial. This is because of
the extreme complexity of the material system, as silicon nanocrystal ensembles are complicated
heterogeneous systems that include various parameters such as crystalline network properties,
matrix properties, surface passivation conditions, nanocrystal size and shape distributions, and
so on, affecting the optical properties of the nanocrystal [2,3]. Figure 7.1 shows some possible
radiative electron–hole recombination processes in the system of nc-Si embedded in a dielec-
tric matrix [3]. The quantum confinement effect of nc-Si, luminescent surface/interface states in
the interface regions between nc-Si and the dielectric matrix, and luminescent defect states in the
dielectric matrix could be involved in the light emission [3].
258 Semiconductor Nanocrystals and Metal Nanoparticles

Interface region

Conduction band edge

nc-Si core
Surface/interface state

Defect state in
dielectric matrix

Valence band edge

Dielectric matrix

FIGURE 7.1 Schematic of some possible radiative electron–hole recombination processes in the system of
nc-Si embedded in a dielectric matrix.

As pointed out previously, quantum confinement effect is likely to be a source of photon emission
in structures with one or more dimensions smaller than the Bohr radius of an exciton (excited elec-
tron–hole pair). When injected electrons and holes by electrical or optical stimulation are confined
into an extremely small Si nanocrystal, the potential barrier formed around each silicon nanocrystal
inhibits the diffusion of electrons and holes. This effect significantly increases the radiative recom-
bination possibility. If quantum confinement effect is responsible for the light emission, a blue shift
in the luminescence wavelength with reduction of nc-Si size should be observed. This is because the
quantum confinement effect causes an expansion of the nc-Si bandgap when the nc-Si size is reduced.
Figure 7.2 shows the bandgap of nc-Si as a function of the nc-Si diameter [16]. As shown in the figure,
the bandgap increases with decreasing nc-Si size. The expanded bandgap results in a blue shift of the
emitted wavelength. The blue shift of the luminescence peak wavelength with reduction of nc-Si size
has been frequently observed. Figure 7.3 shows PL spectra of nc-Si embedded in SiO2 prepared by
cosputtering Si and SiO2 and annealing at temperatures higher than 1100°C [18]. The sample with

4.0
5 nm

3.5
Bandgap (eV)

3.0

2.5

2.0

1.5
1 2 3 4 5
Dot size (nm)

FIGURE 7.2 Bandgap (i.e., PL peak energy) of amorphous silicon quantum dot as a function of the dot size.
The dashed line was obtained from the effective mass theory for three-dimensionally confined Si dots (i.e., nc-Si).
(From Park, N.-M. et al., Appl. Phys. Lett., 78, 2575, 2001. With permission.)
Light Emission Properties of Si Nanocrystals Embedded in a Dielectric Matrix 259

1.4 1.4
nc-Si in SiO2 T = 295 K nc-Si in SiO2 T=4 K
1.2 λ = 488 nm 4 mW/cm2 1.2 λ = 488 nm 4 mW/cm2
9.0 nm Diameter 4.2 nm 9.0 nm Diameter 4.2 nm
1.0 1.0

Intensity (normalized)
Intensity (normalized)

0.8 0.8

0.6 0.6

0.4 0.4

0.2 0.2

0.0 0.0
0.8 1.0 1.2 1.4 1.6 1.8 0.8 1.0 1.2 1.4 1.6 1.8
(a) Photon energy (eV) (b) Photon energy (eV)

FIGURE 7.3 PL spectra of nc-Si embedded in SiO2 thin films (a) at room temperature and (b) at 4 K.
(From Takeoka, S. et al., Phys. Rev. B, 62, 16820, 2000. With permission.)

the average nc-Si diameter of 59.0 nm exhibits PL at ~1.19 eV. The peak energy is slightly larger than
the bandgap energy of bulk Si crystal (1.12 eV). With decreasing average nc-Si diameter from 9.0 to
4.2 nm, the PL peak shifts monotonously to higher energy and reaches 1.42 eV [18]. Figure 7.4 shows
the PL peak energy at room temperature as a function of the nc-Si size [18]. The PL energy varies
from the vicinity of the bulk bandgap to 1.9 eV when decreasing the size from 9 to 1 nm.

2.0

nc-Si
1.8
Peak energy (eV)

1.6

1.4

1.2

Band gap of bulk Si

1.0
0 2 4 6 8 10
Diameter (nm)

FIGURE 7.4 PL peak energy versus average diameter of nc-Si embedded in SiO2 thin films. (From Takeoka, S.
et al., Phys. Rev. B, 62, 16820, 2000. With permission.)
260 Semiconductor Nanocrystals and Metal Nanoparticles

Figure 7.5 shows PL lifetime for two samples of nc-Si embedded in SiO2 thin films with different
nc-Si sizes (average diameters of 4.2 and 9.0 nm) [18]. We can see that the temperature dependence of
the PL lifetime is much different between the two samples. For the sample with the average diameter of
4.2 nm, the PL lifetime changes abruptly at 10–20 K. On the other hand, for the average diameter
of 9.0 nm, the PL lifetime gradually decreases until 150 K. The temperature dependence can be well
explained by the model proposed by Calcott et al. [19]; the exciton state is split into a singlet state and
a triplet state due to the electron–hole exchange interaction. In this model (see the inset of Figure 7.5),
the excitonic levels are split by the energy (ΔE) due to the exchange interaction of an electron and
a hole. The lower (upper) level is a triplet (singlet) state with a radiative lifetime τt (τs). The overall
temperature dependence of the PL lifetime can be calculated on the basis of Boltzmann statistics
with a weight factor of 3 for the triplet state [18]. The broken curves in Figure 7.5 represent the
results of the fitting based on the model. We can see that this model can well reproduce the observed
temperature dependence of the PL lifetime below 150 K. Both the lifetimes (τt and τs,) of the triplet
and singlet states and the splitting energy (ΔE) can be obtained from the fitting. As the PL energy
increases, τt and τs become shorter and ΔE becomes larger. The values of τt, τs, and ΔE represent
the degree of carrier confinement; τt and τs depend on the optical transition oscillator strength, and
ΔE is determined by the spatial separation between an electron and a hole in nc-Si. The continuous
change of these parameters suggests that the PL mechanism is the same in the wide PL energy range.
Furthermore, the PL peak energy and the splitting energy of excitons change continuously from the
bulk values. This continuity from the bulk values indicates that the PL is due to the recombination of
electron–hole pairs confined in nc-Si, that is, a quantum confinement model [18].
The presence of luminescent surface states in the interface between nc-Si and the dielectric
matrix can result in photon emission also. Figure 7.3b shows the PL spectra of nc-Si embedded
in SiO2 at 4 K [18]. At this temperature, in addition to the PL peak at ~1.2–1.4 eV, which can be
observed at room temperature, another peak appears at ~0.8–0.9 eV. The low-energy peak position
shows a size dependence similar to that of the high-energy one, although the dependence is weaker.

10 nc-Si in SiO2
8
6

4
PL lifetime (m)

dave = 9.0 nm
2

1
8 ΔE
6

4
τs τt
dave = 4.2 nm
2

0.1
0 100 200 300
1000/T (K–1)

FIGURE 7.5 Temperature dependence of PL lifetime for two samples of nc-Si embedded in SiO2 thin films
with different nc-Si sizes [18]. The inset shows a schematic illustration of singlet and triplet splitting of an
exciton state. τs and τt represent singlet and triplet lifetimes, respectively, and ΔE the splitting energy. (From
Takeoka, S. et al., Phys. Rev. B, 62, 16820, 2000. With permission.)
Light Emission Properties of Si Nanocrystals Embedded in a Dielectric Matrix 261

The low-energy peak is commonly observed for oxidized nc-Si, for example, nc-Si embedded in
SiO2 matrices and surface-oxidized porous Si. This peak is assigned to the recombination of carri-
ers trapped at Pb centers at nc-Si/SiO2 interfaces [18], or to the recombination of a conduction band
electron with a hole in a deep surface trap at nc-Si/SiO2 interfaces. The size-dependent shift of the
low-energy peak is considered to reflect that of the conduction band edge.
The component of defect luminescence can be diminished or removed with passivating tech-
niques. It can also be distinguished by its rapid decay: in this indirect bandgap semiconductor
system, a long lifetime (longer than ~1 μs) is usually expected for excitonic recombination, while
defect-related luminescence could have a lifetime of less than 100 ns. The defects can locate at the
interface between nc-Si and the dielectric matrix. However, the recombination of carriers trapped
at radiative recombination centers that form in the defects in the dielectric matrix itself can result in
luminescence also [20–23]. Defect-related PL from SiO2 has been observed; for example, Figure 7.6
shows the two PL bands at 4.3 and 2.6 eV due to oxygen-deficiency-type defects in a boron-implanted
thermal SiO2 film and oxygen deficiency-type bulk SiO2 [22].
Figures 7.7 through 7.9 present a study of PL from Si-implanted SiO2 layers [24]. Figure 7.7
shows the PL spectra for a sample implanted at a dose of 2 × 1017 cm−2 and after vacuum annealing
at 500°C and 700°C for 30 min. On the as-implanted sample, a broad band around 550–600 nm
(a green band) is observed. The intensity of this band at ~560 nm increases as a function of the
annealing temperature, up to 700°C. No NIR band around 750–850 nm is detected. An NIR band is
observed after 1000°C annealing, as shown in Figure 7.8. The PL intensity for the NIR band grows
with increasing annealing time (Figure 7.9). As can be observed in the three figures, the green band
exists in the as-implanted sample and also after 1000°C annealing and its intensity increases as a
function of the annealing temperature. The intensity reaches its maximum after 1000°C annealing.
For increasing annealing time at 1000°C, the green band intensity decreases (Figure 7.9). The emis-
sion times for the green band are in the nanosecond range, while the NIR emission is characterized
by times from 1 to 0.3 ms. The short emission time and the annealing behavior suggest that the
green emission is related to extended defects like silicon chains in the oxide matrix while the NIR
band is due to the emission from the nc-Si (the quantum confinement effect) [24].

6000
RT 4.3 eV
5000
2.6 eV B-implanted
thermal SiO2
4000
PL intensity (a.u.)

3000

2000

Bulk SiO2
1000

0
2 3 4 5 6
Photon energy (eV)

FIGURE 7.6 PL spectrum at room temperature at a B-implanted thermal SiO2 film (acceleration energy:
30 keV, dose: 1016 cm−2) and oxygen-deficient bulk SiO2 when excited at 5.0 eV photons from a KrF excimer
laser. (From Nishikawa, H. et al., J. Appl. Phys., 78, 842, 1995. With permission.)
262 Semiconductor Nanocrystals and Metal Nanoparticles

0.4

(a) as-implanted
(c)
(b) 500°C vacuum
0.3 (c) 700°C vacuum

Intensity (a.u.)
(b)
0.2

0.1 (a)

0.0
500 600 700 800
Wavelength (nm)

FIGURE 7.7 (a) PL spectra for a sample implanted at the Si+ dose of 2 × 1017 cm−2. (b) Spectra after 500°C and
(c) 700°C 30 min vacuum annealing. (From Ghislotti, G. et al., J. Appl. Phys., 79, 8660, 1996. With permission.)

0.6
(a) 1000°C 30 min Ar
(b) 1000°C 30 min N2
(b)

0.4
Intensity (a.u.)

(a)

0.2

0.0
500 600 700 800 900
Wavelength (nm)

FIGURE 7.8 PL spectra for the sample already annealed in vacuum (reported in Figure 7.7) after 1000°C
annealing in (a) argon and (b) nitrogen annealing. (From Ghislotti, G. et al., J. Appl. Phys., 79, 8660, 1996.
With permission.)

For silicon-rich silicon nitride (SRN) materials (the excess Si can be in the form of nanoparticle
or nanocrystal distributed in the nitride matrix), various localized states in the band tail of the
SRN materials can assist the radiative recombination [2]. In an amorphous SRN film, the static
disorder of the structural randomness can cause potential fluctuations within the exponential band
tails [25–28]. According to the Anderson localization principle, if the average additional disorder-
induced potential is strong enough to suppress the interaction between atoms, frequent electron
Light Emission Properties of Si Nanocrystals Embedded in a Dielectric Matrix 263

0.7

(a) 30 min 1000°C

0.6 (b) 5 h 1000°C
(b)

0.5

Intensity (a.u.) 0.4 (a)

0.3

0.2

0.1

0.0
500 600 700 800 900
Wavelength (nm)

FIGURE 7.9 PL emission from a sample implanted at 3 × 1017 cm−2 fluence and annealed in nitrogen at
1000°C for different times: (a) 30 min and (b) 5 h. (From Ghislotti, G. et al., J. Appl. Phys., 79, 8660, 1996.
With permission.)

scattering results in the loss of the phase coherence in the electron wavefunction and the formation
of strong localization of the scattered electrons [2]. In this case, a large uncertainty in the electron
momentum is caused by the strong scattering, which leads to the loss of the momentum conserva-
tion in the electronic transitions. As a result, an optical transition is allowed between two states for
which energy conservation applies. Thus, the radiative probability is increased in the amorphous
SRN system [2]. Therefore, the light emission from a SRN material may be related to the quantum
confinement effect of the Si nanocrystal or the localized bandtail states.

7.3 PHOTOLUMINESCENCE AND ELECTROLUMINESCENCE

FROM Si+-IMPLANTED SiO2 THIN FILMS
7.3.1 Photoluminescence
Efficient room-temperature PL from SiO2 thin films embedded with nc-Si synthesized by various
techniques has been frequently reported, and various mechanisms for the PL have been pro-
posed. For the nc-Si/SiO2 system synthesized with Si ion implantation into the SiO2 thin films,
the visible-light (in particular the red light) luminescence could be due to the quantum confine-
ment effect of the nc-Si [15,29–33]. However, it was also suggested that the short-wavelength
luminescence was related to the presence of the oxygen vacancies in the Si+-implanted dielectric
matrix [24,34]. It is now generally agreed that there are multiple mechanisms responsible for the
luminescence. As an example, the study of PL properties of silicon nanocrystals embedded in
SiO2 films synthesized by the Si ion implantation technique reported by Ding et al. [35] is sum-
marized in the following.
In their study, Si ion implantation into a 550 nm SiO2 thin film thermally grown on p-type Si
wafer was carried out with the dose of 1 × 1017 cm−2 at the energy of 100 keV. Thermal annealing
was carried out in nitrogen ambient at the temperatures of 500°C, 600°C, 700°C, 850°C, 900°C,
1000°C, and 1100°C for 20 min. The PL spectra of the samples consist of six PL bands peaked at the
wavelengths of ~415, ~460, ~520, ~630, ~760, and ~845 nm. Such PL bands have been frequently
264 Semiconductor Nanocrystals and Metal Nanoparticles

reported and widely accepted in the literature. The PL bands do not show a significant change in the
full-width at half-maximum (FWHM) with the annealing temperature. The PL bands at ~415, ~460,
and ~520 nm have a similar dependence on the annealing temperature. The integrated intensity of the
three bands first increases slightly and then decreases dramatically with the increase of the anneal-
ing temperature. The three bands almost disappear when the annealing temperature is higher than
900°C. The ~630 nm band shows no change with annealing temperatures up to 1000°C but disap-
pears at the annealing temperature of 1100°C. On the other hand, the integrated PL intensity of both
the ~760 and ~845 nm bands increases slightly with the annealing temperature up to 1000°C and
then shows a dramatic increase by ~30 times when the annealing temperature reaches 1100°C [35].
The above results suggest that not all the PL bands originate from the quantum confinement
effect of the nc-Si and some PL bands can be ascribed to the oxide matrix. It is well known that
various defects in the oxide matrix can serve as visible luminescent centers. The weak oxygen bond
(WOB) defects in silicon oxide could be responsible for the 415 nm (~3 eV) PL band [1,36–38]. Ion
implantation produces oxygen vacancies in the SiO2 network, and oxygen interstitials (the precur-
sors of the WOB defects) are created at the same time. The oxygen interstitials change into the
WOB defects immediately after annealing, and the reaction is reversible subject to an excessive
thermal annealing energy. The reverse reaction can explain the decrease in the intensity of the
415 nm PL band with the increase of the annealing temperature when the temperature is higher than
700°C. Pure silica glass [39] and Si-rich silicon oxides [34] have been reported to have a PL band at
~460 nm (~2.7 eV), which is associated with the neutral oxygen vacancy (NOV) defect represented
by O3≡Si–Si≡O3 [34,36]. The NOV defects are actually generated in the same process as the WOB
defects in the Si+-implanted SiO2 film. This explains why the ~415 (WOB) and ~460 nm (NOV) PL
bands have a similar dependence on the annealing temperature. The ~520 nm band is attributed to
the Ed¢ defects [23,35,36], which can be generated by extrinsic ion-implantation-induced dissocia-
tion process [36,37] and/or intrinsic UV photon absorption during the PL measurement [41]. An
NOV defect can be transformed to an Ed¢ center [23,36]. The ~630 nm (~2.0 eV) PL band could
be due to another radiative defect, the non-bridging oxygen hole center (NBOHC) that has been
observed in both pure silica glass and ion-irradiated SiO2 [37,42–45].
The drastic increase in the peak intensity of the ~760 and ~845 nm bands when the annealing
temperature approaches 1100°C indicates that the luminescence mechanisms of the two bands are
different from that of the defects in the oxide matrix discussed above. The ~760 nm band is attrib-
uted to the transitions in the nc-Si due to the quantum confinement (QC) effect [1,35]. However,
there is an energy difference of ~0.13 eV (i.e., the Stokes shift) between the energy of the ~760 nm
PL peak (~1.63 eV) and the nc-Si bandgap (~1.76 eV). As the nc-Si has an indirect band structure
[1,35], the probability of direct transitions is very low. Here the nc-Si/SiO2 interface is thought to
play an important role in the emission of the ~760 nm PL band. The Stokes shift is about the same
as the energy (~0.13 eV) of the Si–O vibration with a stretching frequency of ~1083 cm−1 in
the system of nc-Si embedded in SiO2 [46]. It was shown in a previous study that the coupling of the
confined excitons and the Si–O stretching vibrations dominate the PL process in nc-Si embedded in
SiO2 [47]. The phonons (i.e., the Si–O vibration at the nc-Si/SiO2 interface) can provide the means
required for both the energy dissipation due to the energy conservation requirement and the momen-
tum conservation in the PL process [1,35]. Therefore, the ~760 nm band can be attributed to the
band-to-band transition of the nc-Si assisted by the emission of a phonon (i.e., the involvement of
the Si–O vibration at the nc-Si/SiO2 interface). On the other hand, the 845 nm band could originate
from the localized luminescent centers at the nc-Si/SiO2 interface [1,35,48,49].

7.3.2 Electroluminescence
Visible EL from Si+-implanted SiO2 films has been frequently reported [20,50–59]. As an example,
the EL from nc-Si light-emitting structures with a 30 nm SiO2 thin film embedded with nc-Si fabri-
cated with low-energy Si ion implantation is briefly discussed in the following [1,57].
Light Emission Properties of Si Nanocrystals Embedded in a Dielectric Matrix 265

In the study reported in [57], the nc-Si-distributed SiO2 light-emitting layer was fabricated by the
implantation of Si ions with various doses into thermally grown 30 nm SiO2 thin films on p-type
Si wafers at low implantation energies (2–8 keV). A post-implantation thermal annealing was con-
ducted at 1000°C or 1100°C in N2 for 1 h to induce nanocrystallization of excess Si in SiO2. Visible
and NIR EL has been observed from the EL structures under a negative gate voltage (i.e., a negative
bias is applied to the indium–tin–oxide (ITO) electrode).
It was shown that the EL spectrum consisted of the four EL bands at ~460, ~610, ~740, and
~1260 nm [57]. The ~610 nm band dominates the major EL peak, and the ~460 and ~740 nm bands
also have a large contribution to the major EL peak; and the ~1260 nm band is obviously respon-
sible for the weak luminescence in the NIR region. The peak wavelengths of the four EL bands are
independent of the gate voltage. However, the contribution of each EL band (i.e., the ratio of the
integrated intensity of each EL band to the integrated intensity of the EL spectrum) changes with
the gate voltage. As the gate voltage increases, the contribution of the ~610 nm band decreases, but
the contributions of the ~460 and ~740 nm bands increase and the contribution of the ~1260 nm
band remains small [57].
Both the implantation dose and energy have a large impact on the EL. The EL intensity increases
with the implantation dose; the EL intensity also increases with the implantation energy [57]. The
increase in the EL intensity with either the implanted Si ion dose or the implantation energy can
be explained by the enhancement in the current conduction of the Si+-implanted oxide. Although
the integrated EL intensity increases with both implanted Si ion dose and implantation energy, the
contribution of each EL band shows a different picture of its dependence on the implanted Si ion dose
and implantation energy. The contribution of each EL band is almost independent of the Si ion
dose. However, the contribution of each EL band changes with the implantation energy. As the
implantation energy increases from 2 to 8 keV, the contribution of the ~600 nm band significantly
increases from ~55% to ~61%, the contributions of both the ~740 nm band and the ~1260 nm band
remain unchanged, and the contribution of the ~460 nm EL band significantly decreases from ~23%
to ~13% [1,57].
Similar EL bands have been reported by different groups (e.g., see Song et al. [56]). Various
luminescent defects, such as WOB (O–O) defects [20,36], neutral oxygen vacancy (NOV) (O3≡Si–
Si≡O3) [20,34,36,40,45,53,55,56,61], non-bridging oxygen hole center (NBOHC) (O3≡Si–Si–O∙)
[20,36,45,56,60,61], D center [61], and E′ center [15,20,31,36,56,61] have been proposed as the
mechanisms of luminescence from Si+-implanted SiO2 films. It has been well accepted that the
defect luminescence centers play a major role in EL from Si+-implanted SiO2 films. Although the
origins for the above-mentioned EL bands are still under debate, the ~460 nm (~2.7 eV) band
and the ~600 nm (~2 eV) band have been ascribed to the NOV defect and the NBOHC defect,
respectively [1,57].
As discussed earlier, both PL and EL bands at ~740 nm are observed, suggesting that they have
the same emission mechanism. It has been suggested that the ~740 nm PL band (~1.7 eV) is due
to the band-to-band transition in the nc-Si assisted by the emission of a phonon, which is actually
the Si–O vibration at the nc-Si/SiO2 interface. Therefore, the ~740 nm EL band can be attributed
to the recombination of carries within the nc-Si assisted by the Si–O vibration at the nc-Si/SiO2
interface [1,57]. Here the phonons (i.e., the Si–O vibration at the nc-Si/SiO2 interface) provide the
means required for both the energy dissipation due to the energy conservation requirement and the
momentum conservation in the EL process [1]. As for the infrared EL band centered at ∼1260 nm,
it most possibly originates from the Si substrate because it is also observed from the pure, thermally
grown SiO2 thin films and the SiO2 films implanted with different elements such as Ge and Al. It is
a result of electron–hole recombination in the accumulation region formed beneath the oxide layer
near the interface between the Si substrate and SiO2 [57]. Such recombination is assisted by the
Si–O vibration at the interface between SiO2 and the Si substrate [1,57].
The excitation of the NOV defect acting as a luminescence center under electrical pumping can
be enhanced by the electric field in the oxide [20,56,62]. This explains the result that the contribution
266 Semiconductor Nanocrystals and Metal Nanoparticles

of the ~460 nm band, which is attributed to the NOV defect, increases with the applied voltage
[1,57]. On the other hand, for higher implantation energy, the implanted-Si-distributed region is
extended, leading to a reduction in the tunnel oxide (i.e., the pure SiO2 region) and thus an increase
in the injection current. As a result, for a given applied voltage, the voltage drops in the ITO/oxide
junction and the oxide/substrate junction increase, and thus the voltage drop in the implanted-Si-
distributed oxide region decreases. Thus, the electric field in the implanted-Si-distributed region
decreases, weakening the excitation of the NOV defect. Therefore, the contribution of the ~460 nm
band decreases with the implantation energy [1,57].

7.4 PHOTOLUMINESCENCE AND ELECTROLUMINESCENCE

FROM Si+-IMPLANTED Si3N4 THIN FILMS
7.4.1 Photoluminescence
In this section, we present a study on the PL from Si+-implanted Si3N4 thin films reported by Cen
et al. [63]. Silicon nitride films with the thickness of 120 nm were deposited by low-pressure chemi-
cal vapor deposition (LPCVD) on a 30 nm thick SiO2 thin film thermally grown on Si(100) wafers.
Si ions were implanted into the silicon nitride films at the energy of 30 keV with a dose of 3.5 × 1016
atoms/cm2. The implanted films were thermally annealed in N2 ambient for 1 h at temperatures
ranging from 800°C to 1100°C. The PL spectrum of the Si-implanted sample annealed at 1100°C
is shown in Figure 7.10 [63]. The PL from the control sample (i.e., without the Si ion implantation)
is also included in the figure for comparison. Two emission bands located at ~435 nm (2.8 eV, blue
PL band) and ~680 nm (1.8 eV, red PL band) can be observed for both samples, but the red light
emission from pure silicon nitride is much weaker compared to that of the Si-implanted sample. The
blue PL band is attributed to the radiative transition between the defect states related to the nitrogen
dangling bonds [63]. Although this blue emission is believed to originate from the silicon nitride
host matrix, it is affected by the Si ion implantation also. As shown in Figure 7.10, compared to the
blue PL band from the pure silicon nitride, the blue PL band from the Si-implanted sample shows
a large reduction in the peak intensity and is broadened and red-shifted. This is somewhat similar
to the situation of Ar+ implantation into silicon nitride [64]. This reduced intensity of the blue PL
band from the Si-implanted silicon nitride film could be due to the nonradiative defect centers

0.08

Pure silicon nitride

0.06
PL intensity (a.u.)

0.04
Implanted sample

0.02

0.00 Both annealed at 1100°C

400 500 600 700 800 900 1000

Wavelength (nm)

FIGURE 7.10 PL spectra of the Si-implanted silicon nitride and the pure silicon nitride thin film (i.e., the
control sample), both annealed at 1100°C. (From Cen, Z.H. et al., Electrochem. Solid-State Lett., 12, H38, 2009.
With permission.)
Light Emission Properties of Si Nanocrystals Embedded in a Dielectric Matrix 267

introduced during the Si ion implantation [2,64,65]. In contrast to the blue PL band, the red PL
band is enhanced by the Si ion implantation. As shown in Figure 7.10, the red PL band at ~680 nm
is much stronger due to the Si implantation. The origin of the red PL band is still controversial. It
is often attributed to the recombination via the defect states, such as the Si–Si defects in the silicon
nitride matrix (note that Si implantation can generate defects); however, the radiative transition due
to the quantum confinement in the nc-Si formed in the Si-implanted silicon nitride could be respon-
sible for the red PL band also [63].
More experiments are carried out to understand the red PL band. It can be observed from the
PL measurements shown in Figure 7.11 [63] that the red PL band from the as-implanted sample is
insignificant, but it emerges after a thermal annealing. The nonradiative centers introduced by the
ion implantation in the as-implanted sample, which suppress the light emission, can explain the
absence of the red PL band. With the elimination of these nonradiative centers by thermal anneal-
ing, the intensity of the red and blue PL bands increases. When the annealing temperature is higher
than 800°C, the light emission of the two bands is quenched gradually, in general, as shown in the
inset of Figure 7.11. The evolution of the PL intensity with annealing temperature is similar for the
two bands, and such evolution can also be observed for the control sample. It can be interpreted in
terms of the nonradiative defects due to the hydrogen desorption [63]. On the other hand, with the
1100°C annealing, the red PL band is as strong as or even stronger than the blue PL band, as shown
in Figure 7.10. Indeed, for a sample with a thinner Si-implanted silicon nitride film and annealed at
a high temperature, the red PL band is the major contributor to the light emission, because the con-
tribution of the blue PL band originating from the silicon nitride host matrix is relatively small due
to the thin silicon nitride film. There is no significant shift in the red PL band for different annealing
temperatures, which is a characteristic of defect-related PL [63].

0.20 Blue PL band

1100°C (X1.5)
Red PL band
PL intensity (a.u.)

0.16
0.06
0.12

0.08

0.04
PL intensity (a.u.)

0.04 0.00
As-implanted 800 900 1000 1100
Annealing temperature (°C)

0.02 800°C (X0.3)

As-implanted (X0.4)
0.00

400 500 600 700 800 900 1000

Wavelength (nm)

FIGURE 7.11 PL spectra of the Si-implanted silicon nitride thin films annealed at different temperatures.
The inset shows the evolution of the PL intensity with the annealing temperature for the blue (435 nm) and red
(680 nm) PL bands. (From Cen, Z.H. et al., Electrochem. Solid-State Lett., 12, H38, 2009. With permission.)
268 Semiconductor Nanocrystals and Metal Nanoparticles

The emission mechanism of the red PL band can be further investigated with the PL excitation
(PLE) measurement. The PLE spectra monitored at 680 nm (the red PL band) for the as-implanted
and the annealed Si-implanted samples as well as the pure silicon nitride are shown in Figure 7.12
[63]. As the red PL band is weak for the as-implanted sample and the control sample, there is no
obvious feature or structure in the PLE spectra of these samples. As shown in Figure 7.12a, for the
short excitation wavelengths, a broad PLE peak at around 290 nm exists for the annealed samples.
The excitation transition process begins at the excitation wavelength of ~310 nm (4 eV), which is
close to the Tauc bandgaps (4.1 eV) of the annealed Si-implanted samples derived from the optical
transmissions [63]. Therefore, it is suggested that the major contribution for the red band emission is
the excitation transition between the Si-bond-related states of σ and σ*, which are near the valence
and conduction band edges of silicon nitride, respectively [63].
On the other hand, another PLE peak at around 425 nm (2.9 eV) emerges after the annealing
at 1100°C, and it is not observed in other Si-implanted samples annealed at lower temperatures, as
shown in Figure 7.12b [63]. The emergence of this PLE peak after the annealing at 1100°C indi-
cates that annealing introduces another excitation transition for the red PL band. It has been well
established that annealing at 1100°C leads to the formation of stable nc-Si or Si nanoclusters in
the silicon nitride layer. Therefore, the PLE peak at 425 nm could originate from the formation of
stable nc-Si or Si nanoclusters. To further understand the PLE peak at 425 nm, PLE was conducted
at different monitored wavelengths of the red PL band for the sample annealed at 1100°C, and the
result is shown in Figure 7.13 [63]. It can be seen from the figure that the PLE peak at ~425 nm

80,000
Monitored at 680 nm
1100°C

60,000
PLE intensity (a.u.)

800°C
40,000
As-implanted

20,000
Silicon nitride

0
250 260 270 280 290 300 310 320
(a) Excitation wavelength (nm)

Monitored at 680 nm
20,000
PLE intensity (a.u.)

15,000

1100°C 10,000
800°C
As-implanted 5,000
Silicon nitride
0
400 450 500 550 600 650
(b) Excitation wavelength (nm)

FIGURE 7.12 (a) PLE spectra monitored at 680 nm of the Si-implanted samples annealed at different tempera-
tures. (b) Individual PLE spectra are shifted along the vertical axis for a clear comparison. (From Cen, Z.H. et al.,
Electrochem. Solid-State Lett., 12, H38, 2009. With permission.)
Light Emission Properties of Si Nanocrystals Embedded in a Dielectric Matrix 269

30,000
Monitored at
25,000 610 nm
630 nm

PLE intensity (a.u.)

20,000 670 nm
710 nm
15,000

10,000

5,000

0 425 nm

–5,000
350 400 450 500 550 600 650
Excitation wavelength (nm)

FIGURE 7.13 PLE spectra monitored at different wavelengths for the Si-implanted sample annealed at
1100°C. Individual PLE spectra are shifted along the vertical axis for a clear comparison. (From Cen, Z.H.
et al., Electrochem. Solid-State Lett., 12, H38, 2009. With permission.)

is independent of the monitored wavelength. It is expected that the PLE peak shifts toward lower
energy as the monitored wavelength is increased for the transitions associated with the quantum
confinement effect [63]. Therefore, it is believed that the quantum confinement effect does not play
a major role for the red PL band. As the red PL band is strongly influenced by the Si implantation
and cannot be observed from the Ge-implanted silicon nitride films [65], it is reasonable to argue
that the emission of the red PL band originates from the radiative recombination associated with the
defect states related to the nc-Si or Si nanoclusters embedded in the silicon nitride matrix [63]. The
large Stokes shifts observed for the two PLE peaks (i.e., at 290 and 425 nm, respectively) suggest
that the PL primarily occurs via the relaxed radiative recombination states [63].

7.4.2 Electroluminescence
In this section, we present a systematic study of the EL from Si+-implanted Si3N4 thin films reported
by Cen et al. [66–70]. The EL mechanisms, thermal annealing effect, and EL color tunability are
discussed.

7.4.2.1 Fabrication of EL Structures

A silicon nitride thin film with the thickness of 30 nm was deposited on a p-type (100) Si wafer
by the LPCVD technique at 800°C. The reactant gases were dichlorosilane (DCS) and ammo-
nia (NH3) with flow rates of 50 and 200 sccm, respectively, and the total deposition pressure was
300 mTorr. In order to enhance the injection current and improve the light-emission performance,
a nearly uniform distribution of the implanted Si ions was achieved by multiple implantations with
the following implantation recipe [2,66]: the first implantation at an energy of 25 keV with a dose
of 4 × 1016 atoms/cm2; the second implantation at 8 keV with a dose of 8 × 1015 atoms/cm2; and
the third implantation at 2 keV with a dose of 3 × 1015 atoms/cm2. As revealed by the stopping
and range of ions in matter (SRIM) simulation, a nearly uniform distribution of the implanted Si
in the silicon nitride thin film is achieved, and the average implanted Si concentration throughout
the Si-implanted film is ~1.2 × 1022 cm−3 (the corresponding volume fraction of the implanted Si is
~25%), except for the area close to the film surface. The multiple implantations were used for the
studies on the EL mechanism and the annealing effect. For other studies discussed later, a single
270 Semiconductor Nanocrystals and Metal Nanoparticles

implantation was used, and the details will be given in the relevant discussions. After the implan-
tations, thermal annealing was conducted in nitrogen ambient at various temperatures for 1 h. To
fabricate the EL structure, a 200 nm thick Al layer was deposited onto the backside of the Si sub-
strate to form the ohmic contact, and circular transparent ITO electrodes with a thickness of 130 nm
were deposited on the Si-implanted silicon nitride thin film through a hard shadow mask with a pad
radius of 1.2 mm by the sputtering technique [66].

7.4.2.2 EL Properties and Mechanisms

Figure 7.14 shows the EL spectra from the 1100°C-annealed Si-implanted silicon nitride thin film
measured at various gate biases [66]. A primary EL peak at ~415 nm can be observed in the EL
spectra. In addition to the main EL peak, there is an apparent shoulder at ~560 nm in the EL spec-
tra. For low applied voltages, the EL intensity increases with the applied voltage, but large voltages
cause EL degradation. Despite the variation of the EL intensity, no significant changes in the spec-
tral shape can be observed under different voltages.
The current conduction of the silicon nitride thin film is greatly enhanced by the nearly uniform
distribution of the implanted Si in the silicon nitride thin film. This means that a large injection cur-
rent can be easily achieved with a relatively low voltage. Figure 7.15 shows the current–voltage (I–V)
characteristic of the EL structure [66]. As can be observed in the figure, the current conduction in
the Si+-implanted nitride film follows a power law that can be described by the equation shown in
Figure 7.15. The power-law fitting to the experimental I–V data yielded a ζ value of 2.06, suggest-
ing that the current conduction is close to 2D transport. The power-law behavior indicates that the

Gate voltage: –16 V Violet band

Green-yellow band
EL intensity (a.u.)

Near IR
band
UV band
EL intensity (a.u.)

–16 V

1.5 2.0 2.5 3.0 3.5 4.0

–19 V Photon energy (eV)

–14 V

–12 V
–10 V

300 400 500 600 700 800 900

Wavelength (nm)

FIGURE 7.14 EL spectra measured at different gate voltages. The inset shows the deconvolution of the EL
spectrum measured at the gate voltage of −16 V. (From Cen, Z.H. et al., Appl. Phys. Lett., 94, 041102, 2009.
With permission.)
Light Emission Properties of Si Nanocrystals Embedded in a Dielectric Matrix 271

–40 Current
Power-law fitting I = α0(V – VT)ζ

Integrated EL intensity (a.u.)

–30 Integrated EL intensity

Current (mA)
–20

–10

0
10 5 0 –5 –10 –15 –20
Gate voltage (V)

FIGURE 7.15 Gate current and integrated EL intensity as funtions of the gate voltage. The power-law fitting
yields ζ = 2.06, α0 = 0.23 mAV−ζ, and VT = −8.1 V. (From Cen, Z.H. et al., Appl. Phys. Lett., 94, 041102, 2009.
With permission.)

enhancement in the current conduction can be explained in terms of the formation of conductive
percolation paths by the defects, that is, nc-Si and Si nanoparticles distributed in the nitride matrix
due to the Si implantation [66].
The dependence of the integrated EL intensity on the voltage is also shown in Figure 7.15 [66].
There is no light emission under a positive voltage due to insufficient hole injection from the ITO
electrode. EL is observed under a negative voltage with the voltage magnitude larger than ~10 V.
As can be seen in Figure 7.15, in the voltage range of −10 to −16 V, the integrated EL intensity is
proportional to the injection current, implying that the EL originates from the recombination of the
electrons and holes injected from the ITO gate and the p-type Si substrate, respectively. However,
when the magnitude of the voltage is larger than ~16 V, the integrated EL intensity is quenched,
and it decreases gradually as the magnitude of the voltage further increases. It has been found that
such an EL quenching is related to the charge trapping in the Si-implanted nitride thin film, which
reduces both the electron injection from the ITO gate and the hole injection from the p-Si substrate [2].
Moreover, the high electric field in the silicon nitride film, which is estimated to be ~6 MV/cm
for the voltages of −17 to −20 V, could have an impact on the radiative lifetime of the luminescent
centers, leading to a quenching of the radiative recombination [71]. In addition, nonradiative Auger
recombination may also affect the light emission when multiple carriers are present in the same
nanostructures [72].
The EL spectra can be deconvolved into two primary Gaussian-shaped EL bands, that is, the
violet EL band (~3.0 eV or 415 nm) and the green-yellow EL band (~2.2 eV or 560 nm). When the
magnitude of the voltage is larger than 15 V, an ultraviolet (UV) band (~3.8 eV or 327 nm) and an
NIR band (~1.45 eV or 850 nm) emerge, but their contributions to the total EL intensity are much
smaller compared to the two main EL bands. As an example, the deconvolution of the EL spectrum
measured at an applied voltage of −16 V is shown in the inset of Figure 7.14 [66]. The two primary
EL bands, that is, the violet and green-yellow bands, together contribute up to 90% of the total EL
intensity. The intensity of these two EL bands (in particular, the violet band) decreases with the volt-
age when its magnitude is larger than ~16 V, which is responsible for the quenching of the total EL
intensity at high voltage. Some of the possible radiative recombination processes responsible for the
four EL bands (i.e., the violet band, green-yellow band, UV band, and NIR band) are schematically
shown in Figure 7.16 and discussed below [2,67].
The strong violet EL band at 3.0 eV, which can also be observed from the Ge-implanted SiO2
layer [55], has seldom been reported for silicon nitride–based materials, although violet PL has been
272 Semiconductor Nanocrystals and Metal Nanoparticles

Vg < 0 Electron
Hole

Ef (Green-yellow band)
ITO gate
Long-wavelength major EL band

Near IR band
Band tail

Short-wavelength UV band
major EL band
(violet band)
Si0

p-type Si
Silicon
nanocrystal Ef
=N–
Silicon Silicon
nitride nitride σ

FIGURE 7.16 Schematic band diagram of the EL structure under forward bias. Carrier transport, defect
states, and Si nanocrystal in the nitride film, as well as various recombination processes are also shown.
(From Cen, Z.H. et al., J. Appl. Phys., 105, 123101, 2009. With permission.)

obtained from Si nanostructures in silicon nitride films [16,73]. PL with energies close to this violet
EL band from thick Si-implanted silicon nitride films, which are fabricated with the implantation
recipe similar to that of the EL structures, has been observed. One of the mechanisms usually used
to explain the short wavelength light emission from silicon nitride films is radiative recombination
through the defect states within the silicon nitride bandgap [74,75]. The defect-related mechanism
should particularly play an important role in the case of Si-implanted silicon nitride thin films because
various point defects can be introduced into the dielectric matrix during the Si implantation [61].
This can explain the difference in light emissions between the Si-rich silicon nitride materials fab-
ricated by ion implantation and those by CVD. The violet emission at ~3.0 eV could be attributed
to the radiative transition from the defect state related to the silicon dangling bonds (≡Si0) located
at 3.1 eV above the valence band maximum (VBM) of silicon nitride to the bonding state (σ) of
the ≡Si–Si≡ unit that is close to the valence band [75–77] or the radiative transition from the defect
state (≡Si0) to the VBM [66]. On the other hand, similar defect-related luminescence mechanisms
could be also used to explain another primary band, the green-yellow EL band. Band tail radiative
recombination has been used to account for the emission of the green-yellow band [78]; nevertheless
the radiative recombination assisted by defect states could also be the origin [74,79]. The green-yel-
low band could be due to the transition between the ≡Si0 state and the state related to the nitrogen
dangling bonds (=N–) in the valence band tail, which is located at 0.8 eV above the VBM [2,75,77].
As mentioned earlier, the two minor EL bands, that is, the UV band and the NIR band, appear
at a sufficiently high voltage (high-level injection). The mechanism responsible for the UV band
peak at ~3.8 eV is still not clear. Two possible mechanisms are explained in the following [2,66].
Compared to the stoichiometric silicon nitride, Si-implanted silicon nitride thin film has been found
to have a large trap density of shallow energy levels below the silicon nitride conduction band
Light Emission Properties of Si Nanocrystals Embedded in a Dielectric Matrix 273

minimum (CBM). A significant number of electrons can be captured by these shallow traps during
the electron injection from the ITO gate under a high voltage. The radiative recombination between
the electrons captured by the shallow traps and the holes in the valence band of silicon nitride
injected from the p-type Si substrate produces a UV emission [66]. It is also possible that under
a high bias voltage, electrons can be injected into the conduction band of silicon nitride and then
recombine with the holes in the band tail states above the silicon nitride VBM, leading to the UV
band emission [2,67], as shown in Figure 7.16. It is worth mentioning that PL near the UV region
has been observed from silicon-rich silicon nitride thin films prepared by LPCVD also [80]. On the
other hand, at a high voltage, significant electron or hole injection into the conduction or valence
bands of the nc-Si embedded in the silicon nitride matrix, respectively, could also occur. The radia-
tive recombination of the injected electrons and holes via nc-Si can produce EL in the NIR range
as a result of the quantum confinement effect. This argument is supported by the fact that the peak
energy of the NIR band is close to the bandgap of the nc-Si (~1.5 eV) [66].

7.4.2.3 Annealing Effect

Figure 7.17 shows the effect of annealing on the I–V characteristics of the EL structure [67]. Compared
to the as-implanted film, the film thermally annealed at 800°C shows a reduction in the current con-
duction at gate voltages larger than ~12 V; and the current reduction becomes more significant and
over a wider voltage range for annealing at 1100°C. The annealing-induced current reduction can be
attributed to the recovery of the silicon nitride matrix during the thermal annealing process [67].
As shown in Figure 7.17, the I–V characteristics also follow a power law, which can be explained in
terms of the formation of conductive percolation tunneling paths by the Si nanocrystals as well as of
the defect states created by the ion implantation [67]. With a large number of nc-Si or Si nanoclusters
and defect states distributed in the thin film, the injected carriers can be transported through tunnel-
ing via the Si nanostructures and the defect states. As the defect states can be removed by annealing,
the tunneling via the defect states will decrease after annealing. This explains the increase of the
global threshold voltage or the reduction of the conduction current with increase in annealing tem-
perature [67]. The annealing effect on the current conduction will, of course, affect the EL properties.
The integrated EL intensity for the as-implanted film and the films annealed at 800°C and
1100°C as a function of the gate voltage is shown in Figure 7.18 [67]. For both the as-implanted film

ζ α (A/cm2V–ζ) VT (V)
–70
1.9 0.29 –2.4
–60 1.5 0.84 –2.8
2.1 0.23 –8.1
–50
As-implanted
Current (mA)

–40 800°C annealing

1100°C annealing
–30 Power law fitting

–20

–10

0
0 –5 –10 –15 –20
Gate voltage (V)

FIGURE 7.17 Annealing effect on the I–V characteristics. (From Cen, Z.H. et al., J. Appl. Phys., 105, 123101,
2009. With permission.)
274 Semiconductor Nanocrystals and Metal Nanoparticles

5 × 105
1100°C annealing

4 × 105 800°C annealing

Integrated EL intensity (a.u.)

As-implanted

3 × 105

2 × 105

1 × 105

0
–10 –12 –14 –16 –18
Gate voltage (V)

FIGURE 7.18 Integrated EL intensity for the as-implanted and annealed films as a function of the gate
­voltages. (From Cen, Z.H. et al., J. Appl. Phys., 105, 123101, 2009. With permission.)

and the film annealed at 800°C, the integrated EL intensity increases with the voltage, and there is
a linear relationship between the integrated EL intensity and the injection current, implying that the
EL originates from the recombination of electrons and holes injected from the ITO gate and the Si
substrate, respectively. However, for the sample annealed at 1100°C, as can be seen in Figure 7.18,
the integrated EL intensity decreases when the voltage is larger than ~16 V. The quenching phe-
nomenon can be explained by different mechanisms, as pointed out earlier. On the other hand, the
EL quantum efficiency can be enhanced largely by annealing because annealing can lead to a large
increase in the EL intensity (Figure 7.18) and a large reduction in the injection current (Figure 7.17).
In addition, as can be seen in Figure 7.18, the EL turn-on voltage decreases significantly as a result of
annealing. The nonradiative defects can be removed by annealing, and nc-Si and other luminescent
centers can also be formed during annealing, particularly at a high temperature. Thus, annealing
can increase the light emission quantum efficiency [67].
In addition to the EL intensity, the EL spectrum is also significantly affected by annealing.
The EL spectrum of the as-implanted film shows a broad peak located at an orange wavelength
of ~600 nm (~2.0 eV). However, in the films annealed at 800°C or 1100°C, a dominant peak at
~400 nm (i.e., the violet band) and a shoulder band at ~550 nm (i.e., the green-yellow band), which
make up a nearly white EL emission, dominate the EL spectra. Figure 7.19 shows the comparison
of the EL deconvolution between annealing at 800°C and 1100°C [67]. All the EL spectra of the
as-implanted and annealed films consist of two major bands including the violet band (named the
“short-wavelength major band” in [67]) and the green-yellow band (i.e., the “long-wavelength major
band” in [67]). The minor UV band emerges after annealing at both 800°C and 1100°C, and the
minor NIR band appears after annealing at 1100°C, as shown in Figure 7.19b.
Figure 7.20 shows the evolution of the peak position, intensity, and FWHM of the four EL bands
with annealing temperature [67]. Annealing leads to blue shifts of the two major bands (i.e., violet
and green-yellow). As the annealing temperature is increased from 800°C to 1100°C, the UV band
becomes remarkable, and the NIR band is detectable after annealing at 1100°C [67]. The blue shift
of the two major bands with increase in annealing temperature may be due to the evolution of the
band-tail states with annealing. Due to annealing, the implantation-induced damage is recovered
and the Si-implanted film becomes more ordered, thus reducing the band tail. Therefore, for a
higher annealing temperature, the radiative recombination of the electrons from the luminescence
Light Emission Properties of Si Nanocrystals Embedded in a Dielectric Matrix 275

Wavelength (nm)
900 800 700 600 500 400 300
Short-wavelength major band

Long-wavelength
major band
EL intensity (a.u.)

UV band

1.5 2.0 2.5 3.0 3.5 4.0

(a) Photon energy (eV)

Wavelength (nm)
900 800 700 600 500 400 300

Short-wavelength major band

Long-wavelength
major band
EL intensity (a.u.)

UV band

Near IR
band

1.5 2.0 2.5 3.0 3.5 4.0

(b) Photon energy (eV)

FIGURE 7.19 Deconvolution of the EL spectrum at the gate voltage of −17 V for annealing at (a) 800°C and
(b) 1100°C. (From Cen, Z.H. et al., J. Appl. Phys., 105, 123101, 2009. With permission.)

centers with the holes in the states in the reduced band tail can produce light emission with higher
energies [2,67]. If the UV band is due to the recombination of the injected electrons in the conduc-
tion band of silicon nitride with the holes in the band tail above the silicon nitride valence band, its
slight blue shift with annealing temperature shown in Figure 7.20a can be explained by the reduced
valence band tail at a higher annealing temperature [2,67]. On the other hand, as the NIR band is
attributed to the band-to-band transition in the nc-Si, its emergence should be related to the forma-
tion of stable nc-Si as a result of annealing at the high temperature [2,67].
276 Semiconductor Nanocrystals and Metal Nanoparticles

900
Near IR band
1.5
800 Long-wavelength major EL band
EL band wavelength (nm) Short-wavelength major EL band

EL band energy (eV)

700 UV band

2
600

500 2.5

400 3
3.5
300 4
As-implanted 800 1100
(a) Annealing temperature (°C)

1800 Near IR band

1600 Long-wavelength major EL band
Short-wavelength major EL band
EL band intensity (a.u.)

1400
UV band
1200
1000
800
600
400
200
0
As-implanted 800 1100
(b) Annealing temperature (°C)

0.9

0.8

0.7
FWHM (eV)

0.6

0.5

0.4 Near IR band

Long-wavelength major EL band
0.3 Short-wavelength major EL band
UV band
0.2
As-implanted 800 1100
(c) Annealing temperature (°C)

FIGURE 7.20 Evolution of each EL band with annealing temperature at the gate voltage of −17 V. (a) Peak
energy. (b) Intensity. (c) FWHM. (From Cen, Z.H. et al., J. Appl. Phys., 105, 123101, 2009. With permission.)
Light Emission Properties of Si Nanocrystals Embedded in a Dielectric Matrix 277

7.4.2.4 Electrically Tunable EL

It has been shown that the EL wavelength can be tuned by varying the injection current [69]. In
the study reported in [69], a single Si ion implantation into a 30 nm Si3N4 thin film at the energy of
12 keV with the implantation dose of 4 × 1016 atoms/cm2 was carried out, and the implanted film
was annealed at 1100°C in N2 atmosphere for 1 h. Figure 7.21 shows EL spectra and the photographs
of the corresponding light emissions under various injection currents [69]. As can be seen in the
figure, the spectral shape or the emission color changes with the injection current, and even bright
white light emission can be achieved at a relatively high injection current. The change in the emis-
sion color is actually due to the changes in both the intensities and FWHM of the two major bands
(i.e., the violet band and the green-yellow band).
The peak positions (i.e., the peak wavelength λpeak or the peak energy Epeak), the FWHM of the
two major bands, the ratio (Imax-GY/Imax-V) of the peak intensity of the green-yellow band to that of the
violet band, and the integrated EL intensity ratio (SGY/SV) of the green-yellow band to the violet band
as a function of the injection current are shown in Figure 7.22a through d, respectively [69]. As can
be seen in Figure 7.22a, the peak positions of the two EL bands do not shift with the injection c­ urrent.
As shown in Figure 7.22b, the FWHM of the green-yellow band is larger than that of the violet band,
and as the injection current is increased from −26 mA/cm2, the FWHM of the green-yellow band
decreases significantly but that of the violet band increases gently. The peak intensities of both bands
increase with injection current simultaneously. However, the peak intensity of the green-yellow
band increases faster than that of the violet band, leading to an increase in the peak intensity ratio
(Imax-GY/Imax-V) with injection current, as shown in Figure 7.22c. For low injection currents (e.g., less
than −61 mA cm–2), Imax-GY/Imax-V < 0.55, the violet band is dominant. However, at high injection cur-
rent of −265 mA/cm2, Imax-GY/Imax-V = 1.1, showing that the peak intensity of the green-yellow band

–9 mA/cm2 –9 mA/cm2
300 –11.84 V
(a)
200
2500 –177 mA/cm2 –177 mA/cm2
–15.26 V
100 2000 (d)
1500
–44 mA/cm2 –44 mA/cm2 1000
EL intensity (a.u.)

1000 –13.84 V
EL intensity (a.u.)

(b) 500
0
500 4000 –265 mA/cm2 –265 mA/cm2
(e) –16.07 V
3000
0
2000
–88 mA/cm2 –88 mA/cm2
1500 –14.12 V
1000
(c) 1.2 mm
1000 0
1.5 2.0 2.5 3.0 3.5 4.0
500 Photon energy (eV)

0
1.5 2.0 2.5 3.0 3.5 4.0
Photon energy (eV)

FIGURE 7.21 (a–c) EL spectra and photographs of the corresponding light emissions under different injection
currents. (d,e) The EL spectra consist of the two major bands, namely, the violet band and the green-yellow band.
(From Cen, Z.H. et al., Opt. Express, 18, 20439, 2010. With permission.)
278 Semiconductor Nanocrystals and Metal Nanoparticles

600
550 2.2

Epeak (eV)
(a) 2.4

λpeak (nm)
Violet band
500 2.6
Green-yellow band
450 2.8
400 3
0.7
FWHM (eV) (b)
0.6
0.5
0.4
1.2
(c)
Imax0-GY/Imax-V

0.8

0.4

1.2
(d)
0.8
SGY /SV

0.4

0 –50 –100 –150 –200 –250

Current density (mA/cm2)

FIGURE 7.22 (a) Peak positions (λ peak or Epeak), (b) peak FWHM of the violet and green-yellow bands,
(c) peak-intensity ratio Imax-GY/Imax-V of the green-yellow band to the violet band, and (d) the integrated-EL-
intensity ratio S GY/S V of the green-yellow band to the violet band as a function of the injection current.
(From Cen, Z.H. et al., Opt. Express, 18, 20439, 2010. With permission.)

is even slightly greater than that of the violet band. The integrated-EL-intensity ratio (SGY/SV) of the
green-yellow band to the violet band, which includes the effects of both the peak intensity and the
FWHM, also increases as the injection current is increased, as shown in Figure 7.22d. The changes in
Imax-GY/Imax-V and FWHM with injection current are responsible for the observed changes in the spec-
tral shape or the color shown in Figure 7.21. In fact, the color transition from violet at low injection
currents to white at high injection currents is mainly due to the changes in Imax-GY/Imax-V [69].
The evolution of the dominant EL bands (the violet and green-yellow) with injection current of the
Si-implanted film with a single implantation cannot be observed in the multiply Si-implanted films with
a uniform distribution of the implanted Si ions [69]. The difference in EL evolution between a single
implantation (most of the implanted Si atoms are located in the middle of the silicon nitride film) and
multiple implantations (the implanted Si is distributed almost uniformly throughout the entire film) sug-
gests that the evolution of the two major EL bands could be related to the distribution of the implanted Si
in the silicon nitride thin film. In the case of the single implantation, the major radiative recombination
can occur in the region near the Si substrate for low negative voltages (small injection current) and in
middle of the Si-implanted film for high negative voltages (large injection current) [69]. Therefore, the
luminescence centers from the Si3N4 host determine the light emission under small injection currents,
which is due to the low implanted Si concentration near the Si substrate, while the large concentration
of implanted Si in the middle of the film should play a more important role in EL for large injection cur-
rents. As discussed previously, the violet band originates from the silicon nitride matrix and the green-
yellow band is related to the implanted Si. Therefore, if the major radiative recombination region shifts
from the lightly implanted Si areas to the highly implanted Si areas due the increase in the voltage,
Light Emission Properties of Si Nanocrystals Embedded in a Dielectric Matrix 279

a transition in the dominant EL band from violet to green-yellow could occur. On the other hand, the
EL quenching of Si-implanted Si3N4 films can also affect the evolution of the intensity ratio of the violet
band to the green-yellow band with increasing injection current [69].

ACKNOWLEDGMENTS
The author wishes to thank Y. Liu, L. Ding, Z. H. Cen, and others at the Nanyang Technological
University for their contribution to the results presented here.

REFERENCES
1. Chen, T. P. and Ding, L. 2010. Optical and optoelectronic properties of silicon nanocrystals embedded
in SiO2 matrix. In Nanostructured Thin Films and Coatings: Functional Properties, Zhang, S. (Ed.),
pp. 113–165. CRC Press, Boca Raton, FL.
2. Cen, Z. H. 2011. Studies of optical properties and light emission of silicon-rich silicon nitride thin films.
PhD thesis, Nanyang Technological University, Singapore.
3. Ding, L. 2009. Optical and optoelectronic properties of Si nanocrystals embedded in dielectric matrix.
PhD thesis, Nanyang Technological University, Singapore.
4. Pavesi, L. and Lockwood, D. J. (Eds.) 2004. Silicon Photonics. Springer, Berlin, Germany.
5. Cutolo, A., Iodice, M., Spirito, P., and Zeni, L. 1997. Silicon electro-optic modulator based on a three
terminal device integrated in a low-loss single-mode SOI waveguide. Journal of Lightwave Technology
15:505–518.
6. Liu, A., Jones, R., Liao, L. et al. 2004. A high-speed silicon optical modulator based on a metal-oxide-
semiconductor capacitor. Nature 427:615–618.
7. Liu, Y., Liu, E., Li, G. et al. 1994. Novel silicon waveguide switch based on total internal reflection.
Applied Physics Letters 64:2079–2080.
8. Zhao, C. Z., Chen, A. H., Liu, E. K., and Li, G. Z. 1997. Silicon-on-insulator asymmetric optical switch
based on total internal reflection. IEEE Photonics Technology Letters 9:1113–1115.
9. Ghioni, M., Zappa, F., Kesan, V. P., and Warnock, J. 1996. A VLSI-compatible high-speed silicon
­photodetector for optical data link applications. IEEE Transactions on Electron Devices 43:1054–1060.
10. Hawkins, A. R., Wu, W., Abraham, P., Streubel, K., and Bowers, J. E. 1997. High gain-bandwidth-
product silicon heterointerface photodetector. Applied Physics Letters 70:303–305.
11. Pellegrino, P., Garrido, B., Garcia, C. et al. 2005. Low-loss rib waveguides containing Si nanocrystals
embedded in SiO2. Journal of Applied Physics 97:074312(1-8).
12. Valenta, J., Pelant, I., and Linnros, J. 2002. Waveguiding effects in the measurement of optical gain in a
layer of Si nanocrystals. Applied Physics Letters 81:1396–1398.
13. Canham, L. T. 1990. Silicon quantum wire array fabrication by electrochemical and chemical dissolu-
tion of wafers. Applied Physics Letters 57:1046–1048.
14. Tischler, M. A., Collins, R. T., Stathis, J. H., and Tsang, J. C. 1992. Luminescence degradation in porous
silicon. Applied Physics Letters 60:639–641.
15. Mutti, P., Ghislotti, G., Bertoni, S. et al. 1995. Room-temperature visible luminescence from silicon
nanocrystals in silicon implanted SiO2 layers. Applied Physics Letters 66:851–853.
16. Park, N.-M., Kim, T.-S., and Park, S.-J. 2001. Band gap engineering of amorphous silicon quantum dots
for light-emitting diodes. Applied Physics Letters 78:2575–2577.
17. Belyakov, V. A., Burdov, V. A., Lockwood, R., and Meldrum, A. 2008. Silicon nanocrystals: Fundamental
theory and implications for stimulated emission. Advances in Optical Technologies 2008: Article ID
279502, 32pp.
18. Takeoka, S., Fujii, M., and Hayashi, S. 2000. Size-dependent photoluminescence from surface-oxidized
Si nanocrystals in a weak confinement regime. Physical Review B 62:16820–16825.
19. Calcott, P. D. J., Nash, K. J., Canham, L. T., Kane, M. J., and Brumhead, D. 1993 Identification of
radiative transitions in highly porous silicon. Journal of Physics: Condensed Matter 5:L91–L98.
20. Lin, C. J. and Lin, G. R. 2005. Defect-enhanced visible electroluminescence of multi-energy silicon-
implanted silicon dioxide film. IEEE Journal of Quantum Electronics 41:441–447.
21. Guha, S. 1998. Characterization of Si+ ion-implanted SiO2 films and silica glasses. Journal of Applied
Physics 84:5210–5217.
22. Nishikawa, H., Watanabe, E., Ito, D. et al. 1995. Photoluminescence study of defects in ion-implanted
thermal SiO2 films. Journal of Applied Physics 78:842–846.
280 Semiconductor Nanocrystals and Metal Nanoparticles

23. Nishikawa, H., Stahlbush, R. E., and Stathis, J. H. 1999. Oxygen-deficient centers and excess Si in
buried oxide using photoluminescence spectroscopy. Physical Review B 60:15910–15918.
24. Ghislotti, G., Nielsen, B., Asoka-Kumar, P. et al. 1996. Effect of different preparation conditions on light
emission from silicon implanted SiO2 layers. Journal of Applied Physics 79:8660–8663.
25. Giorgis, F., Vinegoni, C., and Pavesi, L. 2000. Optical absorption and photoluminescence properties of
a-Si1−xNx:H films deposited by plasma-enhanced CVD. Physical Review B 61:4693–4698.
26. Yerci, S., Li, R., Kucheyev, S. O. et al. 2010. Visible and 1.54 µm emission from amorphous silicon nitride
films by reactive cosputtering. IEEE Journal of Selected Topics in Quantum Electronics 16:114–123.
27. Kato, H., Kashio, N., Ohki, Y., Seol, K. S., and Noma, T. 2003. Band-tail photoluminescence in hydro-
genated amorphous silicon oxynitride and silicon nitride films. Journal of Applied Physics 93:239–244.
28. Shen, J. M., Palsule, C., Gangopadhyay, S., Naseem, H. A., Kizzar, S., and Goh, F. H. C. 1994.
Characterization of fluorinated hydrogenated amorphous-silicon nitride (a-SiNx:H) alloys. Journal of
Applied Physics 76:1055–1061.
29. Guha, S., Pace, M. D., Dunn, D. N., and Singer, I. L. 1997. Visible light emission from Si nanocrystals
grown by ion implantation and subsequent annealing. Applied Physics Letters 70:1207–1209.
30. Wilkinson, A. R. and Elliman, R. G. 2004. The effect of annealing environment on the luminescence of
silicon nanocrystals in silica. Journal of Applied Physics 96:4018–4020.
31. Shimizu-Iwayama, T., Fujita, K., Nakao, S. et al. 1994. Visible photoluminescence in Si+-implanted
silica glass. Journal of Applied Physics 75:7779–7783.
32. Guha, S., Qadri, S. B., Musket, R. G., Wall, M. A., and Shimizu-Iwayama, T. 2000. Characterization of
Si nanocrystals grown by annealing SiO2 films with uniform concentrations of implanted Si. Journal of
Applied Physics 88:3954–3961.
33. Cheylan, S. and Elliman, R. G. 2001. Effect of particle size on the photoluminescence from hydrogen
passivated Si nanocrystals in SiO2. Applied Physics Letters 78:1912–1914.
34. Liao, L.-S., Bao, X.-M., Zheng, X.-Q., Li, N.-S., and Min, N.-B. 1996. Blue luminescence from
Si+ -implanted SiO2 films thermally grown on crystalline silicon. Applied Physics Letters 68:850–852.
35. Ding, L., Chen, T. P., Liu, Y. et al. 2008. Evolution of photoluminescence mechanisms of Si+-implanted
SiO2 films with thermal annealing. Journal of Nanoscience and Nanotechnology 8:3555–3560.
36. Lin, C.-J., Lee, C.-K., Diau, E. W.-G., and Lin, G.-R. 2006. Time-resolved photoluminescence analysis
of multidose Si-ion-implanted SiO2. Journal of the Electrochemical Society 153:E25–E32.
37. Fischer, T., Petrova-Koch, V., Shcheglov, K., Brandt, M. S., and Koch, F. 1996. Continuously tunable pho-
toluminescence from Si+-implanted and thermally annealed SiO2 films. Thin Solid Films 276:100–103.
38. Cheang-Wong, J. C., Oliver, A., Roiz, J. et al. 2001. Optical properties of Ir2+-implanted silica glass.
Nuclear Instruments and Methods in Physics Research B 175–177:490–494.
39. Tohmon, R., Shimogaichi, Y., Mizuno, H. et al. 1989. 2.7-eV luminescence in as-manufactured
high-purity silica glass. Physical Review Letters 62:1388–1391.
40. Bae, H. S., Kim, T. G., Whang, C. N. et al. 2002. Electroluminescence mechanism in SiOx layers
containing radiative centers. Journal of Applied Physics 91:4078–4081.
41. Stathis, J. H. and Kastner, M. A. 1984. Photoinduced paramagnetic defects in amorphous silicon
dioxide. Physical Review B 29:7079–7081.
42. Skuja, L. 1992. Time-resolved low temperature luminescence of non-bridging oxygen hole centers in
silica glass. Solid State Communications 84:613–616.
43. Munekuni, S., Yamanaka, T., Shimogaichi, Y. et al. 1990. Various types of nonbridging oxygen hole
center in high-purity silica glass. Journal of Applied Physics 68:1212–1217.
44. Bakos, T., Rashkeev, S. N., and Pantelides, S. T. 2002. The origin of photoluminescence lines in irradi-
ated amorphous SiO2. IEEE Transactions on Nuclear Science 49:2713–2717.
45. Valakh, M. Y., Yukhimchuk, V. A., Bratus, V. Y. et al. 1999. Optical and electron paramagnetic res-
onance study of light-emitting Si+ ion implanted silicon dioxide layers. Journal of Applied Physics
85:168–173.
46. Liu, Y., Chen, T. P., Fu, Y. Q. et al. 2003. A study on Si nanocrystal formation in Si-implanted SiO2 films
by x-ray photoelectron spectroscopy. Journal of Physics D: Applied Physics 36:L97–L100.
47. Kanemitsu, Y., Shimizu, N., Komoda, T., Hemment, P. L. F., and Sealy, B. J. 1996. Photoluminescent
spectrum and dynamics of Si+-ion-implanted and thermally annealed SiO2 glasses. Physical Review B
54:14329–14332.
48. Allan, G., Delerue, C., and Lannoo, M. 1996. Nature of luminescent surface states of semiconductor
nanocrystallites. Physical Review Letters 76:2961–2964.
49. Zhuravlev, K. S., Gilinsky, A. M., and Kobitsky, A. Y. 1998. Mechanism of photoluminescence of Si
nanocrystals fabricated in a SiO2 matrix. Applied Physics Letters 73:2962–2964.
Light Emission Properties of Si Nanocrystals Embedded in a Dielectric Matrix 281

50. Kulakci, M., Serincan, U., and Turan, R. 2006. Electroluminescence generated by a metal oxide semi-
conductor light emitting diode (MOS-LED) with Si nanocrystals embedded in SiO2 layers by ion
implantation. Semiconductor Science and Technology 21:1527–1532.
51. Lalic, N. and Linnros, J. 1999. Light emitting diode structure based on Si nanocrystals formed by
implantation into thermal oxide. Journal of Luminescence 80:263–267.
52. Luterova, K., Pelant, I., Valenta, J. et al. 2000. Red electroluminescence in Si+-implanted sol–gel-derived
SiO2 films. Applied Physics Letters 77:2952–2954.
53. Matsuda, T., Nishihara, K., Kawabe, M. et al. 2004. Blue electroluminescence from MOS capacitors
with Si-implanted SiO2. Solid-State Electronics 48:1933–1941.
54. Muller, D., Knapek, P., Faure, J. et al. 1999. Blue electroluminescence from high dose Si+ implantation
in SiO2. Nuclear Instruments and Methods in Physics Research B 148:997–1001.
55. Rebohle, L., von Borany, J., Yankov, R. A. et al. 1997. Strong blue and violet photoluminescence and
electroluminescence from germanium-implanted and silicon-implanted silicon-dioxide layers. Applied
Physics Letters 71:2809–2811.
56. Song, H.-Z., Bao, X.-M., Li, N.-S., and Zhang, J.-Y. 1997. Relation between electroluminescence and
photoluminescence of Si+-implanted SiO2. Journal of Applied Physics 82:4028–4032.
57. Ding, L., Chen, T. P., Liu, Y. et al. 2007. The influence of the implantation dose and energy on the
electroluminescence of Si+-implanted amorphous SiO2 thin films. Nanotechnology 18:455306(1-6).
58. Ding, L., Chen, T. P., Yang, M. et al. 2009. Relationship between current transport and electrolumines-
cence in Si+-implanted SiO2 thin films. IEEE Transactions on Electron Devices 56:2785–2791.
59. Garrido, B., Lopez, M., Perez-Rodriguez, A. et al. 2004. Optical and electrical properties of
Si-nanocrystals ion beam synthesized in SiO2. Nuclear Instruments and Methods in Physics Research
B 216:213–221.
60. Jeong, J. Y., Im, S., Oh, M. S., Kim, H. B., Chae, K. H., Whang, C. N., and Song, J. H. 1999. Defect ver-
sus nanocrystal luminescence emitted from room temperature and hot-implanted SiO2 layers. Journal
of Luminescence 80:285–289.
61. Song, H. Z. and Bao, X. M. 1997. Visible photoluminescence from silicon-ion-implanted SiO2 film and
its multiple mechanisms. Physical Review B 55:6988–6993.
62. Sahoo, P. K., Gasiorek, S., Dhar, S., Lieb, K. P., and Schaaf, P. 2006. Cathodoluminescence and epitaxy
after laser annealing of Cs+-irradiated α-quartz. Applied Surface Science 252:4477–4480.
63. Cen, Z. H., Chen, T. P., Ding, L. et al. 2009. Optical transmission and photoluminescence of silicon
nitride thin films implanted with Si ions. Electrochemical and Solid-State Letters 12:H38–H40.
64. Seol, K. S., Futami, T., Watanabe, T., Ohki, Y., and Takiyama, M. 1999. Effects of ion implantation and
thermal annealing on the photoluminescence in amorphous silicon nitride. Journal of Applied Physics
85:6746–6750.
65. Tyschenko, I. E., Volodin, V. A., Rebohle, L., Voelskov, M., and Skorupa, V. 1999. Photoluminescence
of Si3N4 films implanted with Ge+ and Ar+ ions. Semiconductors 33:523–528.
66. Cen, Z. H., Chen, T. P., Ding, L. et al. 2009. Strong violet and green-yellow electroluminescence from
silicon nitride thin films multiply implanted with Si ions. Applied Physics Letters 94:041102(1-3).
67. Cen, Z. H., Chen, T. P., Ding, L. et al. 2009. Evolution of electroluminescence from multiple Si-implanted
silicon nitride films with thermal annealing. Journal of Applied Physics 105:123101(1-5).
68. Cen, Z. H., Chen, T. P., Ding, L. et al. 2011. Influence of implantation dose on electroluminescence
from Si-implanted silicon nitride thin films. Applied Physics A: Materials Science and Processing
104:239–245.
69. Cen, Z. H., Chen, T. P., Liu, Z. et al. 2010. Electrically tunable white-color electroluminescence from
Si-implanted Silicon nitride thin film. Optics Express 18:20439–20444.
70. Cen, Z. H., Chen, T. P., Ding, L. et al. 2009. Quenching and reactivation of electroluminescence by
charge trapping and detrapping in Si-implanted silicon nitride thin film. IEEE Transactions on Electron
Devices 56:3212–3217.
71. Photopoulos, P. and Nassiopoulou, A. G. 2000. Room- and low-temperature voltage tunable electrolu-
minescence from a single layer of silicon quantum dots in between two thin SiO2 layers. Applied Physics
Letters 77:1816–1818.
72. Delerue, C., Lannoo, M., Allan, G. et al. 1995 Auger and Coulomb charging effects in semiconductor
nanocrystallites. Physical Review Letters 75:2228–2231.
73. Ma, L. B., Song, R., Miao, Y. M. et al. 2006. Blue-violet photoluminescence from amorphous Si-in-SiNx
thin films with external quantum efficiency in percentages. Applied Physics Letters 88:093102(1-3).
74. Mo, C. M., Zhang, L. D., Xie, C. Y., and Wang, T. 1993. Luminescence of nanometer-sized amorphous
silicon nitride solids. Journal of Applied Physics 73:5185–5188.
282 Semiconductor Nanocrystals and Metal Nanoparticles

75. Robertson, J. and Powell, M. J. 1984. Gap states in silicon nitride. Applied Physics Letters 44:415–417.
76. Chen, L. Y., Chen, W. H., and Hong, F. C. N. 2005. Visible electroluminescence from silicon nanocrys-
tals embedded in amorphous silicon nitride matrix. Applied Physics Letters 86:193506(1-3).
77. Deshpande, S. V., Gulari, E., Brown, S. W., and Rand, S. C. 1995. Optical properties of silicon nitride
films deposited by hot filament chemical vapor deposition. Journal of Applied Physics 77:6534–6541.
78. Hao, H. L., Wu, L. K., Shen, W. Z., and Dekkers, H. F. W. 2007. Origin of visible luminescence in
hydrogenated amorphous silicon nitride. Applied Physics Letters 91:201922(1-3).
79. Pei, Z. W., Chang, Y. R., and Hwang, H. L. 2002. White electroluminescence from hydrogenated
amorphous-SiNx thin films. Applied Physics Letters 80:2839–2841.
80. Gritsenko, V. A., Zhuravlev, K. S., Milov, A. D. et al. 1999. Silicon dots/clusters in silicon nitride:
Photoluminescence and electron spin resonance. Thin Solid Films 353:20–24.
 8 Optical Properties of
Semiconductor Nanoparticles
in Photoelectrochemical Cells
M.H. Buraidah, S.N.F. Yusuf, I.M. Noor, and A.K. Arof

CONTENTS
8.1 Introduction........................................................................................................................... 283
8.2 Structure and Mechanism of DSSC and QDSSC..................................................................284
8.2.1 Structure and Mechanism of a DSSC........................................................................284
8.2.2 Structure and Mechanism of QDSSC........................................................................ 286
8.3 Semiconducting Metal Oxides for DSSCs and QDSSCs and Their Optical Properties......... 288
8.3.1 TiO2............................................................................................................................ 288
8.3.2 SnO2........................................................................................................................... 290
8.3.3 ZnO............................................................................................................................ 291
8.3.4 Nb2O5......................................................................................................................... 291
8.3.5 SrTiO3......................................................................................................................... 293
8.4 Optical Properties of Metal Chalcolgenides for Quantum Dots (QDs)................................. 293
8.4.1 Lead Chalcogenide Quantum Dots........................................................................... 293
8.4.2 Cadmium Chalcogenide Quantum Dots.................................................................... 295
8.4.3 Zinc Chalcogenide Quantum Dots............................................................................ 296
8.4.4 Cadmium Selenium Telluride.................................................................................... 296
8.5 Nanostructures....................................................................................................................... 297
8.5.1 1D Nanostructure....................................................................................................... 297
8.5.2 2D Nanostructure...................................................................................................... 298
8.5.3 3D Nanostructure....................................................................................................... 298
8.5.4 1D–2D–3D Nanostructure......................................................................................... 299
8.6 Semiconductor Assembly Process.........................................................................................300
8.6.1 Self-Assembly Technique..........................................................................................300
8.6.2 Layer-by-Layer Technique.........................................................................................300
8.6.3 Electrophoretic Deposition Technique...................................................................... 301
8.7 Summary............................................................................................................................... 301
References....................................................................................................................................... 301

8.1 INTRODUCTION
Photoelectrochemical cells are devices that can harness energy from the sun and do not need mov-
ing parts. These devices are environmentally friendly, noiseless, and nonpolluting. One of the
components of a photoelectrochemical cell is a semiconductor layer. The dye-sensitized solar cell
(DSSC) and the quantum dot–sensitized solar cell (QDSSC) are photoelectrochemical cells that
convert energy from sunlight into electricity. In DSSCs and QDSSCs, the semiconducting layer
is usually a metal oxide (MO). Apart from the MO semiconductor, which has its own role to play,

283
284 Semiconductor Nanocrystals and Metal Nanoparticles

metal chalcogenide semiconductors are used as sensitizers in QDSSCs. Examples of MO semicon-

ductors used in DSSCs and QDSSCs include TiO2 (Gratzel 2001), SnO2 (Chappel and Zaban 2002),
and ZnO (Zhang et al. 2009), whereas examples of sensitizers or quantum dots used in QDSSCs
are CdS (Schaller and Klimov 2004), CdSe (Fuke et al. 2010), CdTe (Xiaoyan et al. 2014), and PbS
(Lee et al. 2009, Ju et al. 2010) metal chalcolgenide semiconductors. Knowledge of the semicon-
ductors is important especially for the design of photoelectrochemical cells with good performance.
In this chapter, we focus on these materials and their application in DSSCs and QDSSCs.

8.2 STRUCTURE AND MECHANISM OF DSSC AND QDSSC

The mesoporous MO film is an important component of DSSC and QDSSC devices. This oxide
film, which consists of an MO network, is coated over a dense blocking layer deposited on a trans-
parent conducting oxide (TCO) substrate. The two layers are sintered to establish electronic conduc-
tion (Hagfeldt et al. 2010). Typically, the mesoporous layer thickness is ~20 μm (Arof et al. 2013,
2014, Aziz et al. 2014, Bandara et al. 2015a, Yusuf et al. 2014) and the nanoparticle is ~20 nm (Arof
et al. 2013, 2014, Aziz et al. 2014, Bandara et al. 2015b, Yusuf et al. 2014). The TCO substrate can
be made of glass or plastic. The most common TCO substrate is fluorine-doped tin oxide (FTO),
although indium-doped tin oxide (ITO) substrate is also used. The charge-transfer dye is attached
to the surface of the mesoporous film.

8.2.1 Structure and Mechanism of a DSSC

Figure 8.1 depicts the DSSC structure. It consists of a transparent substrate that has been coated
with FTO or ITO, an MO semiconductor layer, a dye, an electrolyte or ionic conductor, and a coun-
ter electrode (CE) coated on another substrate.
The mesoporous MO film has pores that will provide additional surface area for dye loading so
that more of the incident light can be absorbed. The MO can be, but not limited to, TiO2 (Kim et al.
2013), ZnO (Wang et al. 2013), SnO2 (Lee et al. 2011), Nb2O5 (Ghosh et al. 2011), and SrTiO3
(Yang et al. 2010) only. The mesoporous layer or film ensures direct contact of the dye mol-
ecule with the electrolyte, which also fills the pores in the film. The CE is usually a platinum-
coated transparent substrate. Apart from platinum, binary and ternary metal chalcogenides such
as NiSe2 (Gong et al. 2013), CoSe2 (Dong et al. 2015), MoS2 (Kim et al. 2015, Wu et al. 2011a),
CoS (Huo et al. 2015, Wang et al. 2009), CuS (Kim et al. 2015, Savariraj et al. 2014), Cu2S (Feng
et al. 2015), WS2 (Wu et al. 2011b), and NiCo2S4 (Xiao et al. 2013); MOs, for example, V2O3 (Wu
et al. 2012), ZnO (Wang et al. 2013), SnO2 (Bu and Zheng 2015), and MoO 2 (Wu et al. 2012):
metal carbides such as TiC (Wu et al. 2013), VC (Wu et al. 2012), WC (Vijayakumar et al. 2015),
and MoC (Wu et al. 2011a); and metal nitrides such as MoN (Song et al. 2012), Mo2N (Wu et al.
2011c), Nb4N5 (Cui et al. 2015), and TiN (Li et al. 2010) can also be used as CE. In addition, Au
has also been reported as a CE (Lee and Lo 2009, Seol et al. 2010). An electrolyte is sandwiched
between the two electrodes. The electrolyte can be solid, quasi-solid, or liquid containing a redox
mediator. For liquid electrolytes, a thermoplastic film is placed in between the photoanode and
the CE to act as a spacer so that the space created can be filled by the electrolyte. The film also
serves as a sealant.
The working of a DSSC is also shown in Figure 8.1 and has been discussed by many researchers
(Aziz et al. 2014, Bandara et al. 2014, Buraidah et al. 2011, Gratzel 2001, Hagfeldt et al. 2010, Le
Bahers et al. 2013, Maçaira et al. 2014, Nazeeruddi et al. 2004, Noor et al. 2011, Peter 2011, Yusuf
et al. 2014). A photocurrent is generated when the dye sensitizer absorbs the incident light. The
MO semiconductor nanoparticles play an important role in allowing the light to pass through to the
dye. The dye should strongly adhere to the mesoporous semiconductor. Thus, MO semiconductor
nanoparticles with a large energy gap are needed for DSSC. This shows that the optical properties
of the semiconducting MO nanoparticles are important for photoelectrochemical cells. Solar energy
Optical Properties of Semiconductor Nanoparticles in Photoelectrochemical Cells 285

E vs
vacuum
(eV)
MO dye electrolyte
hv
–4.0

Conducting substrate

Counter electrode
Ox

–4.5
Re

–5.0

–
e e–
Load –5.5

E vs NHE
– (V)
e
S*
LUMO
–0.5
Conduction band
Fermi level

0
hv
n+
[Co(Phen)3]
– –
l3 /l 0.5
ΔV [Co(bpy)3]
n+
n+
[Co(Cl-Phen)3]
n+
[Co(bpy-pz)2]
Redox potential 1.0
HOMO
S°

FIGURE 8.1 Operating principles of dye-sensitized solar cells.

absorbed by the dye should be more than or equal to the energy gap of the dye. This results in the
excitation of the dye molecules. This excitation process is represented by Equation 8.1:
S + hv ® S * light absorption by the dye (8.1)

According to Nazeeruddin et al. (2004), the dye to be used in DSSCs should have the following
properties: it should harvest light in the visible spectrum (i.e., from 400 to 700 nm); it must have
good adhesion to the semiconductor MO surface; and the O − anions of the dye should be electro-
statically bound to the H+ on the MO surface (Figure 8.2).
The energy level of the dye’s excited state must be above the semiconductor’s conduction band
(CB) edge. For rapid dye regeneration, the redox potential of the mediators must be sufficiently less
positive (with reference to the E vs vacuum scale) than the potential of the dye in the ground state.
The dye should be stable over a long period (i.e., between 20 and 25 years). The excited dye mol-
ecule (S*) releases an electron into the semiconducting MO. The dye molecule is now in an oxidized
state (S+) (Equation 8.2):

S* ® e -(MO) + S+ electron injection (8.2)

286 Semiconductor Nanocrystals and Metal Nanoparticles

SCN NCS O
O
N Ru N
HO OH
N N

O O
O– O–
H+ H+

Metal oxide semiconductor

FIGURE 8.2 Attachment of the dye to the TiO2 surface.

The transport of electrons to the TCO occurs through the dye/MO interface. This can be attrib-
uted to the coupling between the dye and the MO (Thomas et al. 2014), as shown in Figure 8.2.
The electron pathway begins from the highest occupied molecular orbital (HOMO) of the dye, for
example, the Ru-SCN moieties, to the lowest unoccupied molecular orbital (LUMO) at the carbox-
ylated bipyridyl ligand into the CB of the MO semiconductor within a few femtoseconds (Gratzel
1991). The transferred electrons percolate through the interconnected MOs till they reach the TCO
substrate. The electrons then pass through a load, and finally reach the CE. The circuit or pathway
of the electrons is completed when they are returned to the holes in the dye molecules via the redox
couple in the electrolyte. If the redox mediator is an I−/I3− couple, then it is the I3− ion that accepts
the electron from the CE and is reduced to the I− ion. The I− ions release the electrons to the holes
in the dye. The reduction of I3− is shown in Equation 8.3:

1 - - 3
I3 + e ( CE ) ® I - overall charge transfer reaction (8.3)
2 2

Upon regeneration of the dye molecules, the I− ion is oxidized to I3− ion. The process is shown in
Equation 8.4. This process occurs continuously.

3 - 1
S+ + I ® S + I3- dye regeneration (8.4)
2 2

8.2.2 Structure and Mechanism of QDSSC

The structure and working principle of a QDSSC (Figure 8.3) are similar to those of the DSSC. The
difference between QDSSC and DSSC is in their sensitizer (quantum dots and dyes, respectively)
and the electrolyte component. For QDSSCs, instead of the dye, the mesoporous layer is covered
by quantum dot (QD) semiconductors such as CdS (Schaller and Klimov 2004), CdSe (Fuke et al.
2010), CdTe (Xiaoyan et al. 2014), PbS (Lee et al. 2009), and PbSe (Schaller and Klimov 2004).
A typical redox mediator is the I−/I3− couple and is used in DSSC. However, I−/I3− mediators can cor-
rode the QDs. Hence, a polysulfide electrolyte consisting of S2‒ /Sx2‒ redox couple (x = 2–5) is used
in QDSSCs (Jun et al. 2013). The semiconductor nanocrystals mentioned above are able to absorb
visible light from the solar spectrum. This is why these QDs can serve as sensitizers. The structure
of a QDSSC is illustrated in Figure 8.3. The working principles of QDSSCs have been discussed by
many researchers (Brown and Kamat 2008, Choi et al. 2014, Fuke et al. 2010, Jabbour and Doderer
2010, Jun et al. 2013, Kamat 2008, 2013, Kim et al. 2015, Lee and Lo 2009, Lee et al. 2008, 2009,
Optical Properties of Semiconductor Nanoparticles in Photoelectrochemical Cells 287

E vs
vacuum
(eV)
–3.5
MO QD electrolyte
hv

Conducting substrate

Counter electrode
Ox –4.5

Re –5.5

–6.5
– –
e e
Load

E vs
NHE (V)
–
e –1.0
CB of QD
Conduction band

Fermi level 0
ΔV
2– 2–
S /Sx
hv
Redox potential
1.0

VB of QD

2.0

QD

FIGURE 8.3 Structure and working of a QDSSC.

Parsi Benehkohal et al. 2012, Rhee et al. 2013, Salant et al. 2010, Santra and Kamat 2012, Santra
et al. 2013, Savariraj et al. 2014, Tian and Cao 2013, Zhao et al. 2014).
Under illumination, photons will be absorbed by the QDs, and electron–hole (e–h) pairs (elec-
trons in the CB and holes in the valence band (VB)) will be produced.

QD + photons ® QD (e + h) (8.5)

The electrons in the CB of QD will then be injected into the CB of the MO semiconductor (Jun
et al. 2013) such as TiO2, ZnO, and SnO2 (Figure 8.4).

QD (e + h) + MO ® QD(h) + MO(e) (8.6)

The injected electrons then reach the transparent conducting glass through the semiconductor MO net-
work and travel to the CE through the external circuit (Hassan et al. 2014). At the electrolyte/CE inter-
face, Sx2− is reduced by the electrons to S2−, which then releases the electrons to the holes in the VB of
the QDs (Jun et al. 2013) to complete the circuit, as shown in Equations 8.7 through 8.9 and Figure 8.4.
288 Semiconductor Nanocrystals and Metal Nanoparticles

e– CB

Metal oxide Quantum

S2–/Sx2–
semiconductor dot
VB
h+

FIGURE 8.4 Electron injection in TiO2 and CdS QD regeneration by the redox mediator S2−/Sx2−.

Sx 2 - + 2e ® Sx -12 - + S2 - (8.7)

CdX(h) + S2 - ® CdX + S (8.8)

S + Sx -12 - ® Sx 2 - (8.9)

The voltage produced depends on the difference between the Fermi level of the MO semiconductor
and the redox potential of the polysulfide electrolyte.
The issues of concern in this chapter with respect to DSSCs and QDSSCs are (1) light harvesting
and (2) electron transfer from the MO layer to the TCO and finally to the external circuit. Due to
the size of the TiO2 nanoparticles, which is ~20 nm in diameter, scattering of the incident light is
negligible. Electron transfer through the network of nanoparticles in the mesoporous and blocking
layers is quite slow and may lead to electron recombination to the electrolyte and/or dye. We will
show in the following section what researchers have done to address these issues.

8.3 SEMICONDUCTING METAL OXIDES FOR DSSCs AND

QDSSCs AND THEIR OPTICAL PROPERTIES
Many types of MOs have been used as the blocking layer between the TCO and the electrolyte.
MOs with wide bandgaps have been widely used because these materials are able to control the
transfer of electrons from the dye’s excited state to the substrate (Krüger et al. 2011). These MOs
include TiO2, ZnO, and Nb2O5 (Sangiorgi et al. 2014). A blocking layer acts as a barrier at the
interface between the conducting substrate and electrolyte in order to prevent electron recombina-
tion during the transport and collection process of electrons in DSSCs. Electron recombination
that occurs at the conducting substrate/electrolyte interface is a factor responsible for the limita-
tion of DSSC’s performance.

8.3.1 TiO2
There are three forms of TiO2, namely, anatase, rutile, and brookite (Tang et al. 1994). Although
these are isostructural, the open-circuit voltage of the DSSC using anatase TiO2 is higher than that
using rutile TiO2. This is due to the smaller difference in energy between the CB and the redox
potential in rutile TiO2. The bandgap for rutile is 3.0 eV and for anatase it is 3.2 eV (Tang et al.
1994). As it is thermodynamically less stable, the brookite form of TiO2 can be transformed into
Optical Properties of Semiconductor Nanoparticles in Photoelectrochemical Cells 289

other phases at high temperatures (Koyama et al. 2006). According to Zallen and Moret (2006),
the absorption spectrum of brookite extends throughout the visible region with a broad and gradual
edge. These authors have reported that the lowest bandgap for brookite should be at least 3.54 eV.
Among the MOs shown in Figure 8.5, TiO2 has the highest refractive index in the wavelength region
from 430 to 1530 nm (Bond 1965, Devore 1951, Dodge 1986, Gao et al. 2012, Pan et al. 2008). The
anatase form of TiO2 is usually used in DSSCs and QDSSCs.
Table 8.1 lists the refractive index of TiO2 films versus wavelength at several oxidation tempera-
tures (Ting et al. 2000). The values of the temperature and refractive index are estimated from the
original results. It can be seen that the refractive index increases with the oxidation temperature
for all wavelengths in the visible region. However, as the wavelength increases, the refractive index
decreases. This is true at all temperatures.
The absorption coefficient of TiO2 films does not show much change in the visible range.
However, the absorption coefficient, α (cm−1), increases abruptly as the wavelength moves toward
the UV region (Devore 1951). In the visible region, the absorption coefficient shows noticeable
changes, and at 400 nm at 900°C, the absorption is 2.53 × 104 cm−1. As the temperature increases
from 700°C to 900°C, the energy gap decreases by 0.80 eV (Ting et al. 2000).
TiO2 is said to be the most effective electrolyte blocker among the many MOs tried. It blocks
electron recombination to the electrolyte (Barea and Bisquert 2013). The blocking layer can be
prepared by grinding TiO2 powder (particle size ~15 nm) with 0.1 M nitric acid, spin-coating the

2.8
TiO2
Refrective index, n

2.6
SrTiO3
2.4

2.2 Nb2O5
ZnO
2
SnO2
1.8
430 530 630 730 830 930 1030 1130 1230 1330 1430 1530
Wavelength, λ (nm)

FIGURE 8.5 Refractive index for TiO2, ZnO, SrTiO3, Nb2O5, and SnO2 metal oxide semiconductor. (From
Gao, L. et al., Opt. Express, 20(14), 15734, 2012; Pan, S.S. et al., J. Appl. Phys., 103(9), 093103, 2008; Devore,
J.R., J. Opt. Soc. Am., 41(6), 416, 1951; Bond, W.L., J. Appl. Phys., 36(5), 1674, 1965; Dodge, M.J., Refractive
index, in M.J. Weber, ed., Handbook of Laser Science and Technology, CRC Press, Boca Raton, FL, 1986.)

TABLE 8.1
Effect of Oxidation Temperature on Refractive Index
of the TiO2 Films in Terms of Wavelength
Wavelength (nm) →
Temperature (°C) ↓ 400 500 600 700 800
700 2.70 2.61 2.55 2.51 2.47
800 2.76 2.64 2.57 2.53 2.49
900 2.77 2.68 2.62 2.58 2.57

Source: Devore, J.R., J. Opt. Soc. Am., 41(6), 416, 1951.

290 Semiconductor Nanocrystals and Metal Nanoparticles

mixture on the conducting glass substrate, and sintering at ~450°C for 30 min (Arof et al. 2014,
Aziz et al. 2013, Bandara et al. 2013, Hassan et al. 2014, Yusuf et al. 2014). The blocking layer is
dense. The dense blocking layer generates effective electron pathways to the FTO and finally to the
external circuit. The blocking layer should not exceed a certain thickness because this will lead to
the presence of more trap states in the thicker layer, which can hinder the passage of electrons to the
external circuit (Berger et al. 2007, Choi et al. 2012). According to Seo et al. (2011), the performance
of the DSSC with the blocking layer is 30% higher than that of the cell without the blocking layer.
A mesoporous TiO2 film is then applied over the blocking layer, which can be prepared by grinding
TiO2 powder (particle size ~20 nm) with 0.1 M nitric acid, a low molecular weight polymer, and a
few drops of a surfactant (Arof et al. 2014, Aziz et al. 2014, Bandara et al. 2013, Hassan et al. 2014,
Yusuf et al. 2014). Using the doctor-blade technique, the TiO2 layer is then deposited on the block-
ing layer and heated at ~450°C for 30 min. The evaporation of the low molecular weight polymer in
the composition produces mesopores in the MO layer, which is necessary for sufficient dye loading.
The MO layer in Figures 8.1 and 8.3 comprises a blocking layer and a mesoporous layer. This layout
was reported in various works (Arof et al. 2013, 2014, Aziz et al. 2013, Bandara et al. 2012, 2013,
2014, 2015a,b, Yusuf et al. 2014).
According to Lee et al. (2012), since the size of the TiO2 crystallites in the mesoporous layer is
~20 nm, light cannot be scattered, which results in photons not being totally absorbed by the dye-
sensitized photoanode. The photons will be either lost through the counter electrode or partially
absorbed by the electrolyte solution. Hence part of the light is wasted. In order to optimize harness-
ing of the incoming light, the escaping photons must be collected. To achieve this, a highly diffusive
reflecting layer consisting of submicrometer-sized TiO2 spheres is applied over the mesoporous film.
The scattering intensity depends on the size and refractive index of the particles. High-refractive-index,
submicrometer TiO2 particles of size between 0.3 and 1 μm can be used as efficient scatters.
In an effort to enhance DSSC performance, the mesoporous TiO2 electrodes have been coated
with MOs and used in DSSCs (Diamant et al. 2004). These MOs have wide a bandgap, and include
Nb2O5, ZnO, SrTiO3, ZrO2, Al2O3, SnO2, and V2O5, which act as the shell materials. The Nb2O5
coating delays or prevents electron recombination by forming a surface energy barrier between the
transparent conducting oxide and the electrolyte. The other shell materials, namely, ZnO, SrTiO3,
ZrO2, Al2O3, and SnO2, shift the conduction band of the TiO2 by forming a dipole layer at the core–
shell interface. For example (Diamant et al. 2004), surface dipole generation at the interface of the
TiO2/SrTiO3 core/shell shifts the CB to a more negative level. This is due to quantum confinement.
The surface dipole is created by the difference in electron affinity or isolectic point of the SrTiO3
and TiO2 semiconductors. Utilizing the core/shell structure in DSSCs, the open-circuit voltage (VOC)
increased but the short-circuit current (JSC) decreased. However, the overall conversion efficiency
increased by ~15%. Elbohy et al. (2015) proved that DSSCs with mesoporous TiO2 coated with V2O5
exhibited enhanced efficiency by ~10%. This indicates that V2O5 functions effectively as a blocking
layer at the TiO2/electrolyte interface, thus preventing reactions.

8.3.2 SnO2
SnO2 is another MO semiconductor that has been used as a blocking or compact layer. However,
among the many MOs presented in Figure 8.5, SnO2 has the lowest refractive index between 1.8
and 2 in the wavelength range 430–530 nm. The bandgap of SnO2 is 3.6 eV (Dou et al. 2011, Okuya
et al. 2001, Ramasamy and Lee 2011). SnO2 is transparent, chemically stable, and has high electron
mobility between 100 and 200 cm2/V s (Dou et al. 2011).
Yong et al. (2014) introduced an ultrathin SnO2 blocking layer in DSSCs in the effort to improve
the cell’s performance. The SnO2 blocking layer was deposited on FTO-coated glass before TiO2
was pasted over it using the doctor-blade method. The SnO2 blocking layer reduced the recombina-
tion of charges at the FTO/electrolyte interface. The cascading band structure of TiO2 and SnO2,
shown in Figure 8.6, increased the electron lifetime, resulting in an effective charge collection.
Optical Properties of Semiconductor Nanoparticles in Photoelectrochemical Cells 291

E vs
FTO SnO2 TiO2 E vs NHE
vacuum Dye
glass (V)
(eV)

–3.9 eV e– –0.6 V
–4 e– –0.5
–4.2 eV –0.3 V
–4.7 eV – e– hv 0.2 V
e
–5 –4.85 eV 0.35 V 0.5

3.2 eV
–6 1.5

3.6 eV
–7 2.5

–8 3.5

FIGURE 8.6 Schematic of the cascading band structure formed by introducing an SnO2 blocking layer
between FTO glass and TiO2 photoelectrode. (From Yong, S.-M. et al., Thin Solid Films, 556(0), 503, 2014.)

The bottom of the CB and the top of the VB of TiO2 were both higher than those of SnO2, enabling
the photogenerated electrons to easily migrate from TiO2 into SnO2, with fewer photogenerated
electrons undergoing recombination.
Duong et al. (2013) have also employed SnO2 thin films to reduce the number of electrons
entering the electrolyte and recombining with the mediator. The authors observed that the electron
lifetime is longer in samples with SnO2 blocking layers. The performance of the DSSC with the
SnO2 blocking layer improved as compared to the DSSC without the SnO2 blocking layer.

8.3.3 ZnO
ZnO is another wide bandgap semiconductor used in DSSCs (Qiu et al. 2011, Wong et al. 2012).
Among the MOs shown in Figure 8.5, only ZnO has a refractive index larger than that of SnO2.
Since the refractive index n > 1, the refractive angle is always smaller than the incident angle and
the refracted light is closer to the normal, implying that the incident light will mostly be directed
into the solar cell.
Al-Kahlout (2015) has used ZnO nanoparticles in the photoanode of the DSSC. The ZnO
nanopowder was wetted to make it into a paste. The ZnO paste was deposited on a conducting sub-
strate using the doctor-blade technique to form several films. The films were dried at 100°C for 1 h,
followed by sintering in air at temperatures ranging from 200°C to 500°C for a specified length of
time. DSSCs using the ZnO films exhibited different performances, as shown in Table 8.2.
It can be seen that the DSSC using ZnO layers sintered at 400°C exhibited the best cell perfor-
mance with an efficiency of 3.01% and a high photocurrent density of 15.6 mA/cm2. The improved
performance may be attributed to increased light harvesting and better charge transport channel with
less charge recombination at FTO/ZnO/electrolyte interface. These results indicate that the overall
performance of the DSSC can be optimized by sintering the ZnO film at a suitable temperature.

8.3.4 Nb2O5
Xia et al. (2007) have used Nb2O5 as the blocking layer or a compact film between the FTO glass
and mesoporous MO film for application in DSSC. The use of Nb2O5 as blocking layer improved
the fill factor (FF) and open-circuit voltage (VOC). The probable reason for the improvement is the
292 Semiconductor Nanocrystals and Metal Nanoparticles

TABLE 8.2
ZnO-DSSC Performance with Photoelectrodes Annealed at Different
Temperatures
Temperature (°C) VOC (V) JSC (mA/cm2) FF (%) Efficiency, η (%)
200 0.545 4.42 44.42 1.07
300 0.544 12.51 29.68 2.02
350 0.550 12.87 33.62 2.38
400 0.552 15.66 34.82 3.01
450 0.554 13.38 30.49 2.26
500 0.573 8.64 29.69 1.47

formation of a barrier at the interface between the FTO conducting oxide and the mesoporous TiO2
layer, which effectively suppresses electron leakage (Figure 8.7). This results in the increase of VOC.
From the figure, it is seen that the Nb2O5 blocking layer potential is higher than the CB edge of
TiO2. This will impose difficulty to the electrons and prevent the charge carriers from crossing over
to the electrolyte. However, electron injection can take place via electron tunneling because of the
nanometer thickness of the Nb2O5 compact layer.
Sacco et al. (2015) have also shown that a Nb2O5 blocking layer can increase the photovoltaic
efficiency by reducing electron leakage or recombination at substrate/MO semiconductor interface.
Nb2O5 has a permittivity of ~53. Its bandgap is 3.3 eV (Hashemzadeh et al. 2014, Hu and Liu 2015).
Since it is a good insulator, it is expected to act as a barrier to block back electron reaction from the
TCO to the redox couple without lowering the mobility of the injected electrons. This makes Nb2O5
very suitable as a blocking layer. DSSC with a Nb2O5 blocking layer has shown a short-circuit cur-
rent density of 124 A/m2, which is more than 40% increase compared to that of the cell without
Nb2O5 blocking layer (Sacco et al. 2015).

E vs
FTO Nb2O5 TiO2 Redox E vs
vacuum Dye
glass mediator NHE (V)
(eV)

–3.9 eV e– –0.6 V
–4.0 –4.1 eV –0.4 V –0.5
–4.2 eV –0.3 V
–4.5 eV 0V
–4.5 0
e–
–4.95 eV 0.45 V
–5.0 0.5
3.3 eV

–5.5 1.0
3.2 eV

–6.0 1.5

–6.5 2.0

–7.0 2.5

–7.5 3.0

FIGURE 8.7 Schematic view of the electron transfer of the new structured electrode. (From Xia, J. et al.,
J. Photochem. Photobiol. A: Chem., 188(1), 120, 2007.)
Optical Properties of Semiconductor Nanoparticles in Photoelectrochemical Cells 293

8.3.5 SrTiO3
SrTiO3 has a bandgap of ~3.2 eV (Burnside et al. 1999). Its CB is ~0.2 eV higher than the CB of
the anatase form of TiO2. SrTiO3, which has structural similarities with anatase TiO2, is expected to
produce good photovoltage, thereby making it suitable for the development of DSSCs. According
to Jayabal et al. (2014), porous SrTiO3 can be produced via hydrothermal interaction of strontium
acetate and titanium isopropoxide. SrTiO3 is a perovskite material with a cubic structure. Although
SrTiO3 and TiO2 are electronically isostructural, the flat-band potential of SrTiO3 is larger than that
of anatase TiO2. SrTiO3 absorbs more at wavelengths less than 400 nm and shows a maximum at
350 nm. SrTiO3 has high refractive index, below that of TiO2 as shown in Figure 8.5. Using the syn-
thesized SrTiO3 and organic Eosin yellow dye as a sensitizer, a DSSC has been fabricated. The cell
showed a VOC of 730 mV. The value of JSC, FF, and η were 44 A/m2, 55%, and 0.51%, respectively.

8.4 OPTICAL PROPERTIES OF METAL CHALCOLGENIDES

FOR QUANTUM DOTS (QDs)
QDs of different sizes can emit light of different wavelengths or colors although made up of the
same material. This is attributed to quantum confinement. Large QDs exhibit redder fluorescence
as their size increases and bluer as their size decreases. Thus, larger QDs emit light of lower energy
compared to smaller QDs. The coloration produced is directly related to the bandgap energy. This is
because, as the particle size decreases (and may become too small to be comparable to the electron
or exciton wavelength), the decreased confining dimension makes the energy levels discrete and this
widens the bandgap, resulting in an increase in the bandgap energy. Larger QDs have more closely
spaced energy levels, allowing the QDs to absorb low-energy photons. The closely spaced energy
levels of large QDs can also trap the excitons or electron–hole quasiparticles, and therefore elec-
tron–hole pairs in larger QDs have longer life. QDs can increase the efficiency of QDSSCs through
multiple exciton generation (Schaller and Klimov 2004), as shown in Figure 8.8.

8.4.1 Lead Chalcogenide Quantum Dots

As we already mentioned, the optical properties of QDs are dependent on their size. For a QD of size d,
its molar extinction coefficient ε increases according to d1.3 (Moreels et al. 2009). For PbS QDs, the
exciton lifetime τ is between 1 and 1.18 μs. Bulk lead chalcogenide semiconductors such as PbS,

e–

e– e–

Impact ionization
Eg hv > 2Eg or multiple exciton
generation

+ +

FIGURE 8.8 Schematic diagram of multiple exciton generation (MEG) or impact ionization.
294 Semiconductor Nanocrystals and Metal Nanoparticles

PbSe, and PbTe have energy bandgaps of 0.41, 0.28, and 0.31 eV, respectively, and are suitable as
sensitizers for QDSSCs (Onicha et al. 2012). The exciton Bohr radius of PbS, PbSe, and PbTe is 18,
46, and 150 nm, respectively (Fu and Tsang 2012, Zhao et al. 2014). The energy bandgaps of PbS,
PbSe, and PbTe QDs can be tailored to be in the range 0.9–1.1, 0.7–1.7, and 0.6–1.1 eV, respectively
(Murphy et al. 2006, Rhee et al. 2013). Therefore, for PbS, PbSe, and PbTe, the optical absorption
edge can be extended to ~1300, ~1500, and ~2000 nm, respectively, which is in the infrared region.
As mentioned earlier, QDs can produce multiple excitons from one high-energy photon through
impact ionization. The threshold photon energy hvth for impact ionization or multiple exciton
generation (MEG) to occur is given by

æ m* ö
hvth = Eg ç 2 + e ÷ (8.10)
è m*h ø

where
m*e is the effective electron mass
m*h is the effective mass of the hole

The impact ionization threshold for PbSe QD was observed to be close to 3Eg, whereas the thresh-
old of MEG was found to be ~2.5Eg and ~2Eg for PbS and PbTe, respectively (Hardman et al. 2011,
Murphy et al. 2006, Schaller and Klimov 2004). As an example, when PbS QDs with the energy
bandgap of 1.1 eV is irradiated with light, energy from wavelength ~1300 nm to the UV region
will be absorbed. Each photon absorbed by the PbS QD produces one exciton until the threshold
wavelength ~451 nm is reached, when more than one exciton will be produced. The threshold
wavelength of 451 nm is based on the PbS bandgap of 1.1 eV. Since the energy bandgap of PbS can
be tailored from 0.9 to 1.1 eV, the smaller tailored bandgap MEG will occur at the longer wavelength
and more excitons or electron–hole quasiparticles will be produced. The increase in the number of
such exciton in the visible region will lead to an increase in the electron injection rate, current den-
sity, and, finally, the solar conversion efficiency.
These also apply to PbSe and PbTe QDs, and therefore lead chalcogenide can help enhance the
performance of solar cells.
Figure 8.9 shows the dependence of the energy gap and CB edge on the size of the PbS QDs. The
figure suggests that the electrons in the CB of bulk PbS cannot easily cross over into the CB of TiO2
due to the CB edge of bulk PbS being at a lower level than that of TiO2. The CB edge of bulk PbS
is at −4.7 eV. The CB edge of bulk TiO2 is 0.5 eV above that of PbS (Tian and Cao 2013). To enable
electrons to jump into the CB of TiO2, the energy gap has to be increased, and this can be done by
decreasing the size of PbS. From Figure 8.9, it can be seen that electron injection will increase with
decreasing PbS size in the PbS-TiO2 photoelectrode.
Doping in QDs with transition metals such as Cu2+ and Mn 2+, which are optically active, cre-
ates electronic states in the bandgap of the QDs, enabling the tuning of their properties. This can
alter the dynamics of charge separation and recombination. The CB of the PbS QDs was found
to be shifted upward with Hg2+ doping. This can lead to higher JSC. The efficiency of 2.01% and
JSC of 210 A/m 2 have been achieved when using a Cu-doped PbS/CdS photoelectrode (Rhee
et al. 2013).
There are several factors limiting the use of lead chalcogenides in DSSCs. The structure of lead
chalcogenide QDs is unstable at temperatures above 100°C. The optical properties of PbSe QDs are
also sensitive to how they are dispersed in a medium. There are also other factors that can reduce the
quantum yield and lead to band shift in the photoluminescence (PL) spectrum (Zhao et al. 2014).
By coating the MO semiconductor with suitable QDs to form a core/shell structure, stability can be
Optical Properties of Semiconductor Nanoparticles in Photoelectrochemical Cells 295

1.4 –3

1.2 –3.2

Conduction band edge (eV)

–3.4
1
Energy gap (eV) –3.6
0.8
–3.8
0.6
–4
0.4 –4.2
0.2 –4.4

0 –4.6
0 2 4 6 8 10 12
Diameter (nm)

FIGURE 8.9 Energy gap and conduction band edge of PbS QDs with different diameters. (From Tian, J. and
Cao, G., Nano Rev., 4, 8, 2013.)

improved and maximization of fluorescence of the QD core can be achieved. The core can be pro-
tected by the shell against oxidation by acting as a barrier and minimize surface defects by providing
better passivation. At the same time, thermal and photostability of QDs can also be improved.

8.4.2 Cadmium Chalcogenide Quantum Dots

Over the years, cadmium chalcogenide QDs have served as sensitizers in photoelectrochemical
cells to enhance device performance and stability. The sensitizers include CdS, CdSe, and CdTe.
The optical bandgap of CdS of 5 nm size is 2.3 eV, enabling light absorption until ~540 nm (Rhee
et al. 2013). For CdSe with 3 nm size, the bandgap is 1.7 eV; and thus light can be absorbed
up to ~731 nm wavelength. CdTe with 3.8 nm size has a bandgap of 1.45 eV, leading to light
absorption at the wavelengths up to ~887 nm. Fluorescent intensity and Stokes shift are known
to increase and decrease, respectively, with temperature when CdS QDs are embedded in a silica
matrix (Reda 2008). These results indicate that low temperature is favored for the preparation
high-efficiency QDs. The size of CdS-doped silica crystallites increases with the annealing tem-
perature from 373 to 673 K. Absorption and emission maxima are also red-shifted. Photostability
studies have shown that QDs prepared at low temperatures have high fluorescent intensity and
Stokes shift, indicating that the QDs are efficient for solar cells. In another study, Manna et al.
(2002) showed that the size of CdS and CdS/ZnS semiconducting QDs is independent of the
growth time. CdS/ZnS core–shell QDs exhibit more intense PL than that of CdS QDs without
the ZnS shell. According to Lee and Lo (2009), the efficiency of QDSSCs using CdS and CdSe
is 1.15% and 1.24%, respectively. QDSSCs with CdSe exhibited a higher efficiency due to the
smaller energy gap of CdSe compared to CdS. However, since the CB edge of CdSe is almost at
the same level as that of TiO2 (−4.2 eV), electron injection occurs less forcefully or at a slower
rate. The CB edge position of CdS is higher than that of both TiO2 and CdSe (−3.9 eV). Thus,
by co-sensitizing CdS/CdSe QDs on a TiO2 mesoporous layer, adjustment of the Fermi level
resulted in the increase in injection rate from CdSe to TiO2, as proven by the efficiency increase
to 2.9%. Figure 8.10 shows the cascadal potential of TiO2, CdS, and CdSe after the Fermi level
adjustment.
The efficiency of QDSSCs using Mn-doped CdS/CdSe electrode was observed to increase up
to 5.4% (Kamat 2013). The midgap states created by Mn doping prevent electron leakage so that
electron–hole and electron–oxidized polysulfide electrolyte recombination could not occur (Santra
and Kamat 2012).
296 Semiconductor Nanocrystals and Metal Nanoparticles

e–

Conducting glass
e–
e– CdSeCB
CdSCB
TiO2 CB Fermi
level

FIGURE 8.10 Possible band edge structure after the adjustment of Fermi level for TiO2/CdS/CdSe electrode.
(From Lee, Y.-L. and Lo, Y.-S., Advanced Functional Materials, 19, 604, 2009.)

8.4.3 Zinc Chalcogenide Quantum Dots

Among the II–VI semiconductor QDs, ZnS, ZnSe, and ZnTe are the safe choices compared to toxic
cadmium and lead. However, the larger energy bandgap may limit their performance in QDSSCs.
The optical bandgap for ZnS, ZnSe, and ZnTe is 3.6, 2.7, and 2.4 eV, respectively. Zinc chalcogenide,
particularly ZnS, has been extensively studied as a surface passivation layer for photoelectrodes in
QDSSCs. ZnS can protect QD materials from photocorrosion. In addition, ZnS can prevent elec-
trons recombining at the electrode/electrolyte interface. An increase in QDSSC’s efficiency was
observed from 1.5% using TiO2/CdSe electrode to 1.9% using TiO2/CdSe/ZnS electrode. A similar
observation was made by Lee and Lo (2009) when ZnS was coated on TiO2/CdS/CdSe photoelec-
trode. The efficiency increased from 2.9% to 3.7%. A schematic diagram of surface passivation layer
is shown in Figure 8.11.

8.4.4 Cadmium Selenium Telluride

Alloyed semiconducting QDs (cadmium selenium telluride) have been prepared to tailor the opti-
cal properties of individual binary QDs and maintain the particle size (Bailey and Nie 2003). The
absorption/emission energies and composition of these alloyed semiconducting QDs have a nonlin-
ear relationship. This leads to new properties different from those of the parent binary systems. This
new QDs opens more avenues for bandgap tailoring.
Figure 8.12 shows the bandgaps and energy levels of some semiconductors (metal oxides, CdS,
and CdSe) used in DSSCs and QDSSCs.

TiO2
FTO

CdS CdSe

ZnS

FIGURE 8.11 Surface passivation of ZnS on TiO2/CdS/CdSe photoelectrode.

Optical Properties of Semiconductor Nanoparticles in Photoelectrochemical Cells 297

E vs
vacuum E vs
(eV) NHE (V)
–3 –1.5

TiO2 SrTiO3 CdS

–4 ZnO CdSe –0.5
WO3
SnO2

2.25 eV

1.7 eV
–5 0.5
3.2 eV

3.2 eV
2.6 eV

3.2 eV
–6 1.5

3.8 eV
–7 2.5

–8 3.5

FIGURE 8.12 Energy diagram for some metal oxides, CdS, and CdSe.

8.5 NANOSTRUCTURES
In the attempt to address the issues concerning DSSCs and QDSSCs, for example, preventing elec-
tron recombination at the interface by improving the electron transport from the MO nanoparticle
network to the TCO, nanostructures have been synthesized. Structures with at least one dimension
less than 100 nm are called “nanostructures.” Nanostructures can be classified into zero-dimensional
(0D nanoparticle), one-dimensional (1D, nanotube or nanowire), two-dimensional (2D, nanofilm),
and three-dimensional (3D) spheres and helix structures.

8.5.1 1D Nanostructure
One of the strategies to address the transport-limiting process is to use 1D nanostructures (Besra
and Liu 2007). Nanowires (Bendall et al. 2011, Fan et al. 2013a), nanorodes (Jeng et al. 2013, Lai
et al. 2010, 2011, Shaikh et al. 2013, Wang et al. 2015), and nanotubes (Dembele et al. 2013,
Jingbin et al. 2010, Pugliese et al. 2014, Song et al. 2014) have been prepared. According to Lin
et al. (2003), 1D nanostructures can be composed from various MOs such as TiO2, ZnO, or SnO.
Diffusion coefficients of these materials are higher than those of unordered nanostructures. The
electron diffusion lengths of these films are also longer than their thickness. These nanostructures
can serve as photoelectrodes in DSSCs and QDSSCs. The bandgap of TiO2 nanowires embedded in
anodic alumina membranes annealed at 500°C is ~3.35 eV. Because of the quantum size effect, the
optical absorption band edge of these nanowire arrays shows a blue shift compared to that of bulk
anatase TiO2, with a bandgap of 3.2 eV.
Using ZnO in the form of nanowires and the cobalt complex [Co(bpy)3]2+/3+ as redox mediators,
the fabricated DSSCs exhibited a VOC of ~0.2 V higher than that of the DSSC with I−/I3− redox couple
(Fan et al. 2013b), see Figure 8.13. Barpuzary et al. (2014) investigated the performance of a DSSC
consisting of 1D ZnO nanowire, donor−π−acceptor type carbazole dye, or more specifically 2-cyano-
3-(4-(2-(9-p-tolyl-9H-fluoren-6-yl)vinyl)phenyl)acrylic acid, and cobalt tris(2,2′-bipyridyl) redox
mediator. A solar conversion efficiency of 5.7% with JSC of 122 A/m2, VOC of 0.72 V, and FF of 0.65
were achieved. The higher VOC with the cobalt complex mediator is due to the more positive redox
potential (E vs NHE scale) of the cobalt complex compared to that of I−/I3−, as shown in Figure 8.13.
Although DSSCs with these 1D nanostructures showed improved electron transport properties
and assist electrolyte diffusion throughout the photoelectrode, the performance of DSSCs using
298 Semiconductor Nanocrystals and Metal Nanoparticles

FTO ZnO NWs Dye E vs NHE FTO ZnO NWs Dye

(V)
S* S*
CB –0.5 CB
Fermi level Fermi level

0
l3–/l–
0.32 V
[Co(bpy)3]2+/3+
0.56 V 0.5

1.0

S/S+ S/S+

FIGURE 8.13 Schematic energy diagram of DSSC using ZnO nanowire with I−/I3− or cobalt complex
redox couples. (From Fan et al., ACS Applied Materials & Interfaces, 5, 1902, 2013b.)

TiO2 nanoparticles is still better. This is because of the low surface area of the 1D nanostructures,
which causes less absorption of the dye and the QDs. In other words, the light harvesting issue is
still not addressed. Hence, although vertically aligned nanostructures do help to address the trans-
port-limiting problems and assist the electrolyte diffusion throughout the photoelectrode, they still
could not exhibit high device performance (Lee et al. 2014a). Therefore, further efforts have to be
made to improve the performance of solar cells by using 2D and 3D nanostructures.

8.5.2 2D Nanostructure
A 2D semiconductor nanostructure is a type of natural semiconductor with thicknesses on the atomic
scale. An example is the 2D semiconducting graphene. This material is composed of one layer of C
atoms arranged in a honeycomb lattice. Because of their efficient stacking, the 2D nanostructures
can provide good light reflecting ability and good electron diffusion paths (Chen et al. 2012). The
organized 2D nanostructure should be a suitable candidate for scattering layers in DSSCs.
As we now know, although nanoparticles can provide high surface area for dye and QD adsorp-
tion, such nanoparticle-based cells still exhibit poor particle–particle interconnectivity, leading to
high electron leakage and low fill factor. Hence the nanoparticle networks should be replaced by
nanostructures with improved interconnectivity, better electron pathways, and good light harvesting
capability. Xu et al. (2014) have synthesized porous SnO2 nanosheets with surface area ~5 times that
of a single-crystalline SnO2 nanosheet, which led to improved conversion efficiency. The efficiency
enhancement could be attributed to the porous architecture of the 2D nanostructure, which provides
an efficient electron pathway and good light scattering ability.

8.5.3 3D Nanostructure
According to Maçaira et al. (2013), 3D photoelectrode nanostructure should have a large surface
area and large interconnected pores for efficient electrolyte diffusion. Apart from that, hindrance to
electron transport by defect levels, particle–particle boundaries, and recombination losses should be
minimized. Lee et al. (2014b) fabricated a CdSe QDSSC with a 3D TiO2 nanohelix to address the
light scattering issue. The absorbance of nanohelix array was greater than that of TiO2 nanoparticles
in the reference DSSC. This work has shown that the 3D TiO2 nanohelix array has improved light
harnessing, leading to increased absorbance. The electron transport time also decreased.
Optical Properties of Semiconductor Nanoparticles in Photoelectrochemical Cells 299

8.5.4 1D–2D–3D Nanostructure

Kim and Yong (2013) have developed 1D ZnO nanowires and 3D ZnO nanostructures. They
compared the diffuse reflectance spectra of a ZnO sputtered film, a 1D ZnO nanorod, and a 3D
nanosheet-branched ZnO nanorods. The 3D nanosheet-branched ZnO nanorods demonstrated
good light trapping or capturing properties and efficient light scattering capacity in the wavelength
region 400–750 nm. Table 8.3 shows the comparison of the characteristics between the QDSSCs
with nanosheet-branched ZnO nanorods and with ZnO nanorods.
Feng et al. (2015) have developed TiO2 nanowires (TNWs), TiO2 nanowire-TiO2 nanosheet
(TNW-TNS) arrays, and hyperbranched TiO2 nanowire-TiO2 nanorod-ZnO nanorod heterostruc-
tured array (TNW-TNS-ZNR). Light scattering increased in the order TNW < TNW-TNS < TNW-
TNS-ZNR in the wavelength range 380–800 nm. The high light scattering intensity for the 3D
heterostructure will be an advantage for photocurrent enhancement, which will increase the solar
conversion efficiency. Impedance characteristic studies revealed that the Nyquist plot of the nano-
structures consists of two depressed semicircles, which can be represented by an equivalent circuit
as shown in Figure 8.14.

TABLE 8.3
Characteristics of QDSSC with 1D and 3D Nanostructures
ZnO Nanostructure JSC (mA/cm2) VOC (mV) Fill Factor, FF (%) Efficiency, η (%)
QDSSCs with nanosheet- 14.2 631 48 4.4
branched ZnO nanorods
QDSSCs with ZnO nanorods 12.3 643 45 3.5

Source: Feng, H.-L. et al., J. Mater. Chem. A, 3, 7, 2015.

R1 R2
Rs
CPE1 CPE2
Z˝ (Ω)

R1 R2
Rs

Z΄ (Ω)

FIGURE 8.14 Nyquist plot of TiO2 nanowire.

300 Semiconductor Nanocrystals and Metal Nanoparticles

TABLE 8.4
R2 and τ Values
Nanostructure R2 (Ω) τ (ms)
1D 52.11 66
1D–2D 94.13 88
3D 103.9 99

Source: Feng, H.-L. et al., J. Mater. Chem. A, 3, 7, 2015.

In Figure 8.14, R2 represents electron-recombination resistance at the semiconductor array/QD/

electrolyte interface. The electrolyte lifetime (τ) within the cells is obtained as 2π/ω, where ω is
the minimum frequency at the low-frequency side of each semicircle. Table 8.4 shows the R2 and τ
values for the DSSCs using the 1D, 1D–2D, and 3D nanostructures.
The increase in the lifetime indicates the effect of charge extraction and collection due to the
branch in the nanostructure.

8.6 SEMICONDUCTOR ASSEMBLY PROCESS

8.6.1 Self-Assembly Technique
Self-assembly is a technique to form organized structures from a disordered system of pre-existing
components without external direction. In DSSCs, self-assembly forms a compact TiO2 under-
layer film to enhance optical transmission and electron extraction properties, which could lead to
improved DSSC performance. Self-assembly of TiO2 layer results in good contact with TCO glass,
increased optical transmission, and reduced charge recombination. In the self-assembly technique,
the MO solution is poured on the desired substrate, and the solution is allowed to spread over the
substrate (Xie et al. 2014). In the self-assembly process, the MO solution extends outward for the
MO nanocrystals to gradually pack as a film. By varying the concentration and coating volume of
nanocrystals forming the film, the thickness can be controlled. According to Sun et al. (2012), self-
assembled TiO2 layers when used in DSSCs exhibit better efficiency compared to DSSCs without
self-assembled TiO2.

8.6.2 Layer-by-Layer Technique
The layer-by-layer (LbL) technique is used for preparing thin films. This can be accomplished by
solution immersion, spinning, or spraying. LbL is simple and can be inexpensive. Film thickness
can be controlled to as fine as 1 nm resolution.
The LbL self-assembly technique has been used to deposit SnO2 layers on an FTO substrate
(Kim et al. 2012). The DSSC exhibited an increase in JSC from 89.6 to 109.7 A/m2. The efficiency
increased from 5.43% to 6.57%. The enhancement can be attributed to the improved TiO2-FTO
adhesion by the ultrathin SnO2 layer. The ultrathin SnO2 layer also reduced the electron percolation
time in the TiO2 layer. The observed photovoltaic properties were attributed to the unique CB loca-
tion of the LbL-assembled SnO2, which is higher than the CB edge of the FTO and lower than that
of TiO2. The bandgap of the FTO and nano SnO2-based film formed via the LbL-assembly technique
was calculated to be 3.92 eV and that of SnO2 was 4.23 eV (Kim et al. 2012). Hence, there is good
band-match so that electrons can be transported from the dye to TiO2, SnO2, and the FTO collecting
substrate, as shown in Figure 8.6, for the electrons to exit to the external circuit and do work. This
band alignment formed upon introducing a SnO2 ultrathin interfacial layer increases the electron
diffusion from TiO2 to the TCO, thus shortening the transit time of the electrons in TiO2.
Optical Properties of Semiconductor Nanoparticles in Photoelectrochemical Cells 301

8.6.3 Electrophoretic Deposition Technique

Electrophoretic deposition (EPD) is a method of coating an electrode with a layer of solid particles.
In the initial steps of EPD, the solid particles to be deposited are dispersed in an ion-containing liquid.
Ions in the solution are selectively adsorbed on the surface of the solid particles. Hence, the surface
of the solid particles will be charged (Santhanagopalan et al. 2010). The charged solid particles will
move toward the corresponding electrodes and will be deposited when a voltage is applied across the
electrodes. Santhanagopalan et al. (2010) have deposited 1D carbon nanostructures on TCOs at room
temperature. All colloidal particles that can carry charge and form stable suspensions can be used in
EDP to coat ceramics, organics, and even metals to any electrically conductive surface.

8.7 SUMMARY
We attempted to address two DSSC- and QDSSC-related issues, namely, optimization of light
harvesting, and electron transport from the mesoporous MO layer to the glass or plastic collecting
substrate that has been coated with fluorine-doped or indium-doped tin oxide. The blocking or com-
pact MO layer can be TiO2, SnO2, ZnO, Nb2O5, or SrTiO3. The best blocker is TiO2. All these MOs
have a wide bandgap. These blocking layers are dense and have shown the ability to reduce electron
leakage, ensuring smooth electron flow to the external circuit. The optical properties of QDs were
also discussed in this chapter. Several approaches have been adopted in the efforts to enhance the
photoelectrochemical cell performance. TiO2 nanoparticles were coated with blocking oxides,
the size of QDs was reduced, bandgaps were tuned, and QD/MO structures were passivated. Since
the dimensional geometry of MO conductors are also important, various 1D, 2D, and 3D nanostruc-
tures and heterostructures have been fabricated. In the 1D nanostructure, the smaller surface area
can limit the photoelectrochemical cell performance. Even the 3D helix TiO2 nanostructure exhibits
a smaller surface area than conventional nanoparticles, but it is a good light harvester. Finally, we
described briefly how nanostructures can be prepared.

REFERENCES
Al-Kahlout, A. 2015. Thermal treatment optimization of ZnO nanoparticles-photoelectrodes for high photo-
voltaic performance of dye-sensitized solar cells. Journal of the Association of Arab Universities for
Basic and Applied Sciences 7:66.
Arof, A.K., Aziz, M.F., Noor, M.M. et al. 2014. Efficiency enhancement by mixed cation effect in dye-sensitized
solar cells with a PVdF based gel polymer electrolyte. International Journal of Hydrogen Energy
39(6):2929–2935.
Arof, A.K., Naeem, M., Hameed, F. et al. 2013. Quasi solid state dye-sensitized solar cells based on poly-
vinyl alcohol (PVA) electrolytes containing I−/I3− redox couple. Optical and Quantum Electronics
46(1):143–154.
Aziz, M.F., Noor, I.M., Sahraoui, B. et al. 2013. Dye-sensitized solar cells with PVA-KI-EC-PC gel electro-
lytes. Optical and Quantum Electronics 46(1):133–141.
Aziz, M.F., Noor, I.M., Sahraoui, B. et al. 2014. Dye-sensitized solar cells with PVA-KI-EC-PC gel electrolytes.
Optical and Quantum Electronics 46(1):133–141.
Bailey, R.E. and Nie, S. 2003. Alloyed semiconductor quantum dots: Tuning the optical properties without
changing the particle size. Journal of the American Chemical Society 125:7.
Bandara, T.M.W.J., Aziz, M.F., Fernando, H.D.N.S. et al. 2015a. Efficiency enhancement in dye-sensitized
solar cells with a novel PAN-based gel polymer electrolyte with ternary iodides. Journal of Solid State
Electrochemistry 19:7.
Bandara, T.M.W.J., Jayasundara, W.J.M.J.S.R., Dissanayake, M.A.K.L. et al. 2013. Effect of cation size on the
performance of dye sensitized nanocrystalline TiO2 solar cells based on quasi-solid state PAN electro-
lytes containing quaternary ammonium iodides. Electrochimica Acta 109(0):609–616.
Bandara, T.M.W.J., Jayasundara, W.J.M.J.S.R., Fernado, H.D.N.S. et al. 2014. Efficiency enhancement of
dye-sensitized solar cells with PAN:CsI:LiI quasi-solid state (gel) electrolytes. Journal of Applied
Electrochemistry 44(8):917–926.
302 Semiconductor Nanocrystals and Metal Nanoparticles

Bandara, T.M.W.J., Jayasundara, W.J.M.J.S.R., Fernado, H.D.N.S. et al. 2015b. Efficiency of 10% for quasi-
solid state dye-sensitized solar cells under low light irradiance. Journal of Applied Electrochemistry
45(4):289–298.
Bandara, T.M.W.J., Svensson, T., Dissanayake, M.A.K.L. et al. 2012. Tetrahexylammonium iodide contain-
ing solid and gel polymer electrolytes for dye sensitized solar cells. Energy Procedia 14(0):1607–1612.
Barea, E.M. and Bisquert, J. 2013. Properties of chromophores determining recombination at the TiO2–dye–
electrolyte interface. Langmuir 29(28):8773–8781.
Barpuzary, D., Patra, A.S., Vaghasiya, J.V. et al. 2014. Highly efficient one-dimensional ZnO nanowire-based
dye-sensitized solar cell using a metal-free, D−π−A-type, carbazole derivative with more than 5%
power conversion. ACS Applied Materials & Interfaces 6(15):12629–12639.
Bendall, J.S., Etgar, L., Tan, S.C. et al. 2011. An efficient DSSC based on ZnO nanowire photo-anodes and a
new D-[small pi]-A organic dye. Energy & Environmental Science 4(8):2903–2908.
Berger, T., Lana-Villarreal, T., Monllor-Satoca, D. et al. 2007. An electrochemical study on the nature of trap
states in nanocrystalline rutile thin films. Journal of Physical Chemistry C 111(27):9936–9942.
Besra, L. and Liu, M. 2007. A review on fundamentals and applications of electrophoretic deposition (EPD).
Progress in Materials Science 52(1):1–61.
Bond, W.L. 1965. Measurement of the refractive indices of several crystals. Journal of Applied Physics
36(5):1674–1677.
Brown, P. and Kamat, P.V. 2008. Quantum dot solar cells. Electrophoretic deposition of CdSe−C60 com-
posite films and capture of photogenerated electrons with nC60 cluster shell. Journal of the American
Chemical Society 130(28):8890–8891.
Bu, I.Y.Y. and Zheng, J. 2015. A new type of counter electrode for dye sensitized solar cells based on solution
processed SnO2 and activated carbon. Materials Science in Semiconductor Processing 39:223–228.
Buraidah, M.H., Teo, L.P., Yusuf, S.N.F. et al. 2011. TiO2/Chitosan-NH4I(+I2)-BMII-based dye-sensitized
solar cells with anthocyanin dyes extracted from black rice and red cabbage. International Journal of
Photoenergy 10(1155):11.
Burnside, S., Moser, J.-E., Brooks, K. et al. 1999. Nanocrystalline mesoporous strontium titanate as photo-
electrode material for photosensitized solar devices: Increasing photovoltage through flatband potential
engineering. Journal of Physical Chemistry B 103(43):9328–9332.
Chappel, S. and Zaban, A. 2002. Nanoporous SnO2 electrodes for dye-sensitized solar cells: Improved cell per-
formance by the synthesis of 18 nm SnO2 colloids. Solar Energy Materials and Solar Cells 71(2):141–152.
Chen, H.-Y., Kuang, D.-B., and Su, C.-Y. 2012. Hierarchically micro/nanostructured photoanode materials for
dye-sensitized solar cells. Journal of Materials Chemistry 22:15.
Choi, H., Nahm, C., Kim, J. et al. 2012. The effect of TiCl4-treated TiO2 compact layer on the performance of
dye-sensitized solar cell. Current Applied Physics 12(3):737–741.
Choi, H.M., Ji, I.A., and Bang, J.H. 2014. Metal selenides as a new class of electrocatalysts for quantum dot-
sensitized solar cells: A tale of Cu1.8Se and PbSe. ACS Applied Materials and Interfaces 6(4):2335–2343.
Cui, H., Zhu, G., Liu, X. et al. 2015. Niobium nitride Nb4N5 as a new high-performance electrode material for
supercapacitors. Advanced Science 2: 1500126. doi:10.1002/advs.201500126.
Dembele, K.T., Selopal, G.S., Soldano, C. et al. 2013. Hybrid carbon nanotubes–TiO2 photoanodes for high
efficiency dye-sensitized solar cells. Journal of Physical Chemistry C 117(28):14510–14517.
Devore, J.R. 1951. Refractive indices of rutile and sphalerite. Journal of the Optical Society of America
41(6):416–417.
Diamant, Y., Chappel, S., Chen, S.G. et al. 2004. Core–shell nanoporous electrode for dye sensitized solar
cells: The effect of shell characteristics on the electronic properties of the electrode. Coordination
Chemistry Reviews 248(13–14):1271–1276.
Dodge, M.J. 1986. Refractive index. In M.J. Weber (ed.), Handbook of Laser Science and Technology, Vol. 4.
Boca Raton, FL: CRC Press.
Dong, J., Wu, J., Jia, J. et al. 2015. Cobalt selenide nanorods used as a high efficient counter electrode for
dye-sensitized solar cells. Electrochimica Acta 168:69–75.
Dou, X., Sabba, D., Mathews, N. et al. 2011. Hydrothermal synthesis of high electron mobility Zn-doped
SnO2 nanoflowers as photoanode material for efficient dye-sensitized solar cells. Chemistry of Materials
23(17):3938–3945.
Duong, T.-T., Choi, H.-J., He, Q.-J. et al. 2013. Enhancing the efficiency of dye sensitized solar cells
with an SnO2 blocking layer grown by nanocluster deposition. Journal of Alloys and Compounds
561(0):206–210.
Elbohy, H., Thapa, A., Poudel, P. et al. 2015. Vanadium oxide as new charge recombination blocking layer for
high efficiency dye-sensitized solar cells. Nano Energy 13:368–375.
Optical Properties of Semiconductor Nanoparticles in Photoelectrochemical Cells 303

Fan, J., Fàbrega, C., Zamani, R.R. et al. 2013a. Enhanced photovoltaic performance of nanowire dye-sensi-
tized solar cells based on coaxial TiO2@TiO heterostructures with a cobalt(II/III) redox electrolyte.
ACS Applied Materials & Interfaces 5(20):9872–9877.
Fan, J., Hao, Y., Cabot, A. et al. 2013b. Cobalt(II/III) redox electrolyte in ZnO nanowire-based dye-sensitized
solar cells. ACS Applied Materials & Interfaces 5(6):1902–1906.
Feng, H.-L., Wu, W.-Q., Rao, H.-S. et al. 2015. Three-dimensional hyperbranched TiO2/ZnO heterostructured
arrays for efficient quantum dot-sensitized solar cells. Journal of Materials Chemistry A 3:7.
Fu, H. and Tsang, S.-W. 2012. Infrared colloidal lead chalcogenide nanocrystals: Synthesis, properties, and
photovoltaic applications. Nanoscale 4(7):2187–2201.
Fuke, N., Hoch, L.B., Koposov, A.Y. et al. 2010. CdSe quantum-dot-sensitized solar cell with ∼100% internal
quantum efficiency. ACS Nano 4(11):6377–6386.
Gao, L., Lemarchand, F., and Lequime, M. 2012. Exploitation of multiple incidences spectrometric measure-
ments for thin film reverse engineering. Optics Express 20(14):15734–15751.
Ghosh, R., Brennaman, M.K., Uher, T. et al. 2011. Nanoforest Nb2O5 photoanodes for dye-sensitized solar
cells by pulsed laser deposition. ACS Applied Materials & Interfaces 3(10):3929–3935.
Gong, F., Xu, X., Li, Z. et al. 2013. NiSe2 as an efficient electrocatalyst for a Pt-free counter electrode of
dye-sensitized solar cells. Chemical Communications 49(14):1437–1439.
Gratzel, M. 1991. Mesoporous oxide junctions and nanostructured solar cells. Current Opinion in Colloid &
Interface Science 4:8.
Gratzel, M. 2001. Photoelectrochemical cells. Nature 414(6861):338–344.
Hagfeldt, A., Boschloo, G., Sun, L. et al. 2010. Dye-sensitized solar cells. Chemical Reviews
110(11):6595–6663.
Hardman, S.J.O., Graham, D.M., Stubbs, S.K. et al. 2011. Electronic and surface properties of PbS nanopar-
ticles exhibiting efficient multiple exciton generation. Physical Chemistry Chemical Physics 13:9.
Hashemzadeh, F., Rahimi, R., and Ghaffarinejad, A. 2014. Mesoporous nanostructures of Nb2O5 obtained by
an EISA route for the treatment of malachite green dye-contaminated aqueous solution under UV and
visible light irradiation. Ceramics International 40(7, Part A):9817–9829.
Hassan, H.C., Abidin, Z.H.Z., Careem, M.A. et al. 2014. Chlorophyll as sensitizer in I−/I3−-based solar cells
with quasi-solid-state electrolytes. High Performance Polymers 26(6):647–652.
Hu, B. and Liu, Y. 2015. Nitrogen-doped Nb2O5 nanobelt quasi-arrays for visible light photocatalysis. Journal
of Alloys and Compounds 635:1–4.
Huo, J., Zheng, M., Tu, Y. et al. 2015. A high performance cobalt sulfide counter electrode for dye-sensitized
solar cells. Electrochimica Acta 159:166–173.
Jabbour, G.E. and Doderer, D. 2010. Quantum dot solar cells: The best of both worlds. Nature Photonics
4(9):604–605.
Jayabal, P., Sasirekha, V., Mayandi, J. et al. 2014. A facile hydrothermal synthesis of SrTiO3 for dye sensitized
solar cell application. Journal of Alloys and Compounds 586:456–461.
Jeng, M.-J., Wung, Y.-L., Chang, L.-B. et al. 2013. Dye-sensitized solar cells with anatase TiO2 nanorods pre-
pared by hydrothermal method. International Journal of Photoenergy Article ID 280253, 2013(8 pp).
Jingbin, H., Fengru, F., Chen, X. et al. 2010. ZnO nanotube-based dye-sensitized solar cell and its application
in self-powered devices. Nanotechnology 21(40):405203.
Ju, T., Graham, R.L., Zhai, G. et al. 2010. High efficiency mesoporous titanium oxide PbS quantum dot solar
cells at low temperature. Applied Physics Letters 97(4):043106.
Jun, H.K., Careem, M.A., and Arof, A.K. 2013. Quantum dot-sensitized solar cells—Perspective and recent
developments: A review of Cd chalcogenide quantum dots as sensitizers. Renewable and Sustainable
Energy Reviews 22:148–167.
Kamat, P.V. 2008. Quantum dot solar cells. Semiconductor nanocrystals as light harvesters. Journal of
Physical Chemistry C 112(48):18737–18753.
Kamat, P.V. 2013. Quantum dot solar cells. The next big thing in photovoltaics. The Journal of Physical
Chemistry Letters 4(6):908–918.
Kim, B., Park, S.W., Kim, J.Y. et al. 2013. Rapid dye adsorption via surface modification of TiO2 photoanodes
for dye-sensitized solar cells. ACS Applied Materials and Interfaces 5(11):5201–5207.
Kim, H. and Yong, K. 2013. A highly efficient light capturing 2D (nanosheet)-1D (nanorod) combined hierar-
chical ZnO nanostructure for efficient quantum dot sensitized solar cells. Physical Chemistry Chemical
Physics 15(6):2109–2116.
Kim, H.-J., Kim, J.-H., Pavan Kumar, C.S.S. et al. 2015. Facile chemical bath deposition of CuS nano peas
like structure as a high efficient counter electrode for quantum-dot sensitized solar cells. Journal of
Electroanalytical Chemistry 739:20–27.
304 Semiconductor Nanocrystals and Metal Nanoparticles

Kim, Y.J., Kim, K.H., Kang, P. et al. 2012. Effect of layer-by-layer assembled SnO2 interfacial layers in photo-
voltaic properties of dye-sensitized solar cells. Langmuir 28(28):10620–10626.
Koyama, H., Fujimoto, M., Ohno, T. et al. 2006. Effects of thermal annealing on formation of micro porous
titanium oxide by the sol–gel method. Journal of the American Ceramic Society 89(11):3536–3540.
Krüger, S., Hickey, S.G., Tscharntke, S. et al. 2011. Study of the attachment of linker molecules and their
effects on the charge carrier transfer at lead sulfide nanoparticle sensitized ZnO substrates. Journal of
Physical Chemistry C 115(26):13047–13055.
Lai, M.H., Lee, M.W., Wang, G.-J. et al. 2011. Photovoltaic performance of new-structure ZnO-nanorod
dye-sensitized solar cells. International Journal of Electrochemical Science 6:9.
Lai, M.-H., Tubtimtae, A., Lee, M.-W. et al. 2010. ZnO-nanorod dye-sensitized solar cells: New structure
without a transparent conducting oxide layer. International Journal of Photoenergy 2010, Article ID
497095, 5 pp.
Le Bahers, T., Pauporté, T., Lainé, P.P. et al. 2013. Modeling dye-sensitized solar cells: From theory to experi-
ment. The Journal of Physical Chemistry Letters 4(6):1044–1050.
Lee, G., Lee, H., Um, M.-H. et al. 2012. Light scattering amplification on dye sensitized solar cells assembled
by Hollyhock-shaped CdS-TiO2 composites. Bulletin of the Korean Chemical Society 33(9):5.
Lee, H.J., Chen, P., Moon, S.-J. et al. 2009. Regenerative PbS and CdS quantum dot sensitized solar cells with
a cobalt complex as hole mediator. Langmuir 25(13):7602–7608.
Lee, J.-H., Park, N.-G., and Shin, Y.-J. 2011. Nano-grain SnO2 electrodes for high conversion efficiency SnO2–
DSSC. Solar Energy Materials and Solar Cells 95(1):179–183.
Lee, K., Mazare, A., and Schmuki, P. 2014a. One-dimensional titanium dioxide nanomaterials: Nanotubes.
Chemical Reviews 114(19):9385–9454.
Lee, S.H., Jin, H., Kim, D.-Y. et al. 2014b. Enhanced power conversion efficiency of quantum dot sensi-
tized solar cells with near single-crystalline TiO2 nanohelixes used as photoanodes. Optics Express
22(S3):A867–A879.
Lee, Y.-L., Huang, B.-M., and Chien, H.-T. 2008. Highly efficient CdSe-sensitized TiO2 photoelectrode for
quantum-dot-sensitized solar cell applications. Chemistry of Materials 20(22):6903–6905.
Lee, Y.-L. and Lo, Y.-S. 2009. Highly efficient quantum-dot-sensitized solar cell based on co-sensitization of
CdS/CdSe. Advanced Functional Materials 19(4):604–609.
Li, G.-r., Wang, F., Jiang, Q.-W. et al. 2010. Carbon nanotubes with titanium nitride as a low-cost coun-
ter-electrode material for dye-sensitized solar cells. Angewandte Chemie International Edition
49(21):3653–3656.
Lin, Y., Wu, G.S., Yuan, X.Y. et al. 2003. Fabrication and optical properties of TiO2 nanowire arrays made by
sol–gel electrophoresis deposition into anodic alumina membranes. Journal of Physics: Condensed Matter
15:2917–2922.
Maçaira, J., Andrade, L., and Mendes, A. 2013. Review on nanostructured photoelectrodes for next generation
dye-sensitized solar cells. Renewable and Sustainable Energy Reviews 27:334–349.
MaçAira, J., Andrade, L., and Mendes, A. 2014. Modeling, simulation and design of dye sensitized solar cells.
RSC Advances 4(6):2830–2844.
Manna, L., Scher, E.C., Li, L.-S. et al. 2002. Epitaxial growth and photochemical annealing of graded CdS/
ZnS shells on colloidal CdSe nanorods. Journal of the American Chemical Society 124:10.
Moreels, I., Lambert, K., Smeets, D. et al. 2009. Size-dependent optical properties of colloidal PbS quantum
dots. ACS Nano 3(10):3023–3030.
Murphy, J.E., Beard, M.C., Norman, A.G. et al. 2006. PbTe colloidal nanocrystals: Synthesis, characteriza-
tion, and multiple exciton generation. Journal of the American Chemical Society 128:7.
Nazeeruddin Md, K., Zakeeruddin, S.M., Lagref, J.J. et al. 2004. Stepwise assembly of amphiphilic
ruthenim sensitizers and their applications in dye-sensitized solar cell. Coordination Chemistry
Reviews 248:12.
Noor, M.M., Buraidah, M.H., Yusuf, S.N.F. et al. 2011. Performance of dye-sensitized solar cells with
(PVDF-HFP)-KI-EC-PC electrolyte and different dyematerials. International Journal of Photoenergy
10(1155):5.
Okuya, M., Kaneko, S., Hiroshima, K. et al. 2001. Low temperature deposition of SnO2 thin films as trans-
parent electrodes by spray pyrolysis of tetra-n-butyltin(IV). Journal of the European Ceramic Society
21(10–11):2099–2102.
Onicha, A.C., Petchsang, N., Kosel, T.H. et al. 2012. Controlled synthesis of compositionally tunable ternary
PbSexS1–x as well as binary PbSe and PbS nanowires. ACS Nano 6(3):2833–2843.
Pan, S.S., Zhang, Y.X., Teng, X.M. et al. 2008. Optical properties of nitrogen-doped SnO2 films: Effect of the
electronegativity on refractive index and bandgap. Journal of Applied Physics 103(9):093103.
Optical Properties of Semiconductor Nanoparticles in Photoelectrochemical Cells 305

Parsi Benehkohal, N., González-Pedro, V., Boix, P.P. et al. 2012. Colloidal PbS and PbSeS quantum dot
sensitized solar cells prepared by electrophoretic deposition. Journal of Physical Chemistry C
116(31):16391–16397.
Peter, L.M. 2011. The Grätzel cell: Where next? The Journal of Physical Chemistry Letters 2(15):1861–1867.
Pugliese, D., Lamberti, A., Bella, F. et al. 2014. TiO2 nanotubes as flexible photoanode for back-illuminated
dye-sensitized solar cells with hemi-squaraine organic dye and iodine-free transparent electrolyte.
Organic Electronics 15(12):3715–3722.
Qiu, J., Guo, M., and Wang, X. 2011. Electrodeposition of hierarchical ZnO nanorod-nanosheet structures and
their applications in dye-sensitized solar cells. ACS Applied Materials & Interfaces 3(7):2358–2367.
Ramasamy, E. and Lee, J. 2011. Ordered mesoporous Zn-doped SnO2 synthesized by exotemplating for
efficient dye-sensitized solar cells. Energy & Environmental Science 4(7):2529–2536.
Reda, S.M. 2008. Synthesis and optical properties of CdS quantum dots embedded in silica matrix thin films
and their applications as luminescent solar concentrators. Acta Materialia 56(2):259–264.
Rhee, J.H., Chung, C.-C., and Diau, E.W.-G. 2013. A perspective of mesoscopic solar cells based on metal
chalcogenide quantum dots and organometal-halide perovskites. NPG Asia Materials 5:17.
Sacco, A., Di Bella, M.S., Gerosa, M. et al. 2015. Enhancement of photoconversion efficiency in dye-sensi-
tized solar cells exploiting pulsed laser deposited niobium pentoxide blocking layers. Thin Solid Films
574(0):38–42.
Salant, A., Shalom, M., Hod, I. et al. 2010. Quantum dot sensitized solar cells with improved efficiency pre-
pared using electrophoretic deposition. ACS Nano 4(10):5962–5968.
Sangiorgi, A., Bendoni, R., Sangiorgi, N. et al. 2014. Optimized TiO2 blocking layer for dye-sensitized solar
cells. Ceramics International 40(7, Part B):10727–10735.
Santhanagopalan, S., Teng, F., and Meng, D.D. 2010. High-voltage electrophoretic deposition for vertically
aligned forests of one-dimensional nanoparticles. Langmuir 27(2):561–569.
Santra, P.K. and Kamat, P.V. 2012. Mn-doped quantum dot sensitized solar cells: A strategy to boost efficiency
over 5%. Journal of the American Chemical Society 134(5):2508–2511.
Santra, P.K., Nair, P.V., George Thomas, K. et al. 2013. CuInS2-sensitized quantum dot solar cell. Electrophoretic
deposition, excited-state dynamics, and photovoltaic performance. The Journal of Physical Chemistry
Letters 4(5):722–729.
Savariraj, A.D., Viswanathan, K.K., and Prabakar, K. 2014. CuS nano flakes and nano platelets as counter
electrode for quantum dots sensitized solar cells. Electrochimica Acta 149:6.
Schaller, R.D. and Klimov, V.I. 2004. High efficiency carrier multiplication in PbSe nanocrystals: Implications
for solar energy conversion. Physical Review Letters 92(18):4.
Seo, H., Son, M.-K., Kim, J.-K. et al. 2011. Method for fabricating the compact layer in dye-sensitized solar
cells by titanium sputter deposition and acid-treatments. Solar Energy Materials and Solar Cells
95(1):340–343.
Seol, M., Kim, H., Tak, Y. et al. 2010. Novel nanowire array based highly efficient quantum dot sensitized
solar cell. Chemical Communications 46(30):5521–5523.
Shaikh, S.F., Kalanur, S.S., Mane, R.S. et al. 2013. Monoclinic WO3 nanorods-rutile TiO2 nanoparticles core–
shell interface for efficient DSSCs. Dalton Transactions 42(28):10085–10088.
Song, C.B., Zhao, Y.L., Song, D.M. et al. 2014. Dye-sensitized solar cells based on TiO2 nanotube/nanoparticle
composite as photoanode and Cu2SnSe3 as counter electrode. International Journal of Electrochemical
Science 9:8.
Song, J., Li, G.R., Xiong, F.Y. et al. 2012. Synergistic effect of molybdenum nitride and carbon nanotubes
on electrocatalysis for dye-sensitized solar cells. Journal of Materials Chemistry 22(38):20580–20585.
Sun, Z., Kim, J.H., Zhou, Y. et al. 2012. Improved photovoltaic performance of dye-sensitized solar cells
with modifiedself-assembling highly ordered mesoporous TiO2 photoanodes. Journal of Materials
Chemistry 22:9.
Tang, H., Prasad, K., Sanjinbs, R. et al. 1994. Electrical and optical properties of TiO2 anatase thin films.
Journal of Applied Physics 75(4):6.
Thomas, S., Deepak, T.G., Anjusree, G.S. et al. 2014. A review on counter electrode materials in dye-sensi-
tized solar cells. Journal of Materials Chemistry A 2:17.
Tian, J. and Cao, G. 2013. Semiconductor quantum dot-sensitized solar cells. Nano Reviews 4:8.
Ting, C.-C., Chen, S.-Y., and Liu, D.-M. 2000. Structural evolution and optical properties of TiO2 thin films
prepared by thermal oxidation of sputtered Ti films. Journal of Applied Physics 88:7.
Vijayakumar, P., Senthil Pandian, M., Lim, S.P. et al. 2015. Investigations of tungsten carbide nanostructures
treated with different temperatures as counter electrodes for dye sensitized solar cells (DSSC) applica-
tions. Journal of Materials Science: Materials in Electronics 26(10):7977–7986.
306 Semiconductor Nanocrystals and Metal Nanoparticles

Wang, H., Wang, B., Yu, J. et al. 2015. Significant enhancement of power conversion efficiency for dye sen-
sitized solar cell using 1D/3D network nanostructures as photoanodes. Scientific Reports 5:9305, 9 pp.
Wang, H., Wei, W., and Hu, Y.H. 2013. Efficient ZnO-based counter electrodes for dye-sensitized solar cells.
Journal of Materials Chemistry A 1(22):6622–6628.
Wang, M., Anghel, A.M., Marsan, B. et al. 2009. CoS supersedes pt as efficient electrocatalyst for triiodide
reduction in dye-sensitized solar cells. Journal of the American Chemical Society 131(44):15976–15977.
Wong, K.K., Ng, A., Chen, X.Y. et al. 2012. Effect of ZnO nanoparticle properties on dye-sensitized solar cell
performance. ACS Applied Materials & Interfaces 4(3):1254–1261.
Wu, M., Lin, X., Hagfeldt, A. et al. 2011a. Low-cost molybdenum carbide and tungsten carbide counter elec-
trodes for dye-sensitized solar cells. Angewandte Chemie International Edition 50(15):3520–3524.
Wu, M., Lin, X., Wang, Y. et al. 2012. Economical Pt-free catalysts for counter electrodes of dye-sensitized
solar cells. Journal of the American Chemical Society 134(7):3419–3428.
Wu, M., Wang, Y., Lin, X. et al. 2011b. Economical and effective sulfide catalysts for dye-sensitized solar cells
as counter electrodes. Physical Chemistry Chemical Physics 13:4.
Wu, M., Wang, Y., Lin, X. et al. 2013. TiC/Pt composite catalyst as counter electrode for dye-sensitized solar
cells with long-term stability and high efficiency. Journal of Materials Chemistry A 1(34):9672–9679.
Wu, M., Zhang, Q., Xiao, J. et al. 2011c. Two flexible counter electrodes based on molybdenum and tungsten
nitrides for dye-sensitized solar cells. Journal of Materials Chemistry 21(29):10761–10766.
Xia, J., Masaki, N., Jiang, K. et al. 2007. Fabrication and characterization of thin Nb2O5 blocking layers for
ionic liquid-based dye-sensitized solar cells. Journal of Photochemistry and Photobiology A: Chemistry
188(1):120–127.
Xiao, J., Zeng, X., Chen, W. et al. 2013. High electrocatalytic activity of self-standing hollow NiCo2S4 single
crystalline nanorod arrays towards sulfide redox shuttles in quantum dot-sensitized solar cells. Chemical
Communications 49(100):11734–11736.
Xiaoyan, L., Chunlei, W., Shuhong, X. et al. 2014. Manipulation of inter-particle interactions between TiO2
and CdTe: An effective method to enhance the performance of quantum dot sensitized solar cells.
Journal of Physics D: Applied Physics 47(1):015103.
Xie, F., Cherng, S.-J., Lu, S. et al. 2014. Functions of self-assembled ultrafine TiO2 nanocrystals for high
efficient dye-sensitized solar cells. ACS Applied Materials & Interfaces 6(8):5367–5373.
Xu, X., Qiao, F., Dang, L., and Gao, F. 2014. Porous tin oxide nanosheets with enhanced conversion efficiency
as dye-sensitized solar cell electrode. Journal of Physical Chemistry C118:16856.
Yang, S., Kou, H., Wang, J. et al. 2010. Tunability of the band energetics of nanostructured SrTiO3 electrodes
for dye-sensitized solar cells. Journal of Physical Chemistry C 114(9):4245–4249.
Yong, S.-M., Tsvetkov, N., Larina, L. et al. 2014. Ultrathin SnO2 layer for efficient carrier collection in
dye-sensitized solar cells. Thin Solid Films 556(0):503–508.
Yusuf, S.N.F., Aziz, M.F., Hassan, H.C. et al. 2014. Phthaloylchitosan-based gel polymer electrolytes for
efficient dye-sensitized solar cells. Journal of Chemistry Article ID 783023, 2014 (8 pp).
Zallen, R. and Moret, M.P. 2006. The optical absorption edge of brookite TiO2. Solid State Communications
137:4.
Zhang, Q., Dandeneau, C.S., Zhou, X. et al. 2009. ZnO nanostructures for dye-sensitized solar cells. Advanced
Materials 21:22.
Zhao, H., Liang, H., Vidal, F. et al. 2014. Size dependence of temperature-related optical properties of PbS and
PbS/CdS core/shell quantum dots. Journal of Physical Chemistry C 118:9.
 9 Second-Order Nonlinear
Susceptibility in Quantum
Dot Structures

M. Abdullah, Farah T. Mohammed Noori,

and Amin H. Al-Khursan

CONTENTS
9.1 Introduction...........................................................................................................................307
9.2 Semiconductor Nanostructures..............................................................................................308
9.3 Quantum Dots........................................................................................................................308
9.4 Semiconductor Nonlinearities...............................................................................................309
9.5 Nonlinear Optical Susceptibility........................................................................................... 310
9.6 Second-Order Nonlinearity................................................................................................... 311
9.7 Quantum Dot Structure under Study..................................................................................... 314
9.8 Quantum Disk Model under Applied Electric Field.............................................................. 315
9.9 Calculated QD Subbands under Applied Electric Field........................................................ 319
9.10 Density Matrix Formulation of Optical Susceptibility.......................................................... 320
9.11 Results on SONS................................................................................................................... 324
9.12 SHG, SFG, DFG, and OR...................................................................................................... 328
9.12.1 Sum-Frequency Generation....................................................................................... 330
9.12.2 Difference-Frequency Generation............................................................................. 330
9.12.3 Optical Rectification.................................................................................................. 331
9.13 Inhomogeneity in QDs........................................................................................................... 331
9.14 Results and Discussion of SHG, SFG, DFG, and OR in QDs............................................... 332
References....................................................................................................................................... 339

9.1 INTRODUCTION
With the development of the fabrication technique of nanostructures, multiform semiconductor
structures are fabricated [1]. The twenty-first century will see a dramatic change in lighting tech-
nologies. By 2025, fluorescent and incandescent illumination sources should be replaced by more
efficient, long-lasting, and versatile light sources, offering more lumens per cm2 and decreasing the
consumption of energy for lighting by 29% [2]. The core of such lighting devices, in its simplest
form, is a junction, a relatively simple multilayered structure formed by a semiconductor crystal
between two higher bandgap semiconductors, which emits light when an electric current passes
through it. The localization of carriers in all three dimensions breaks down the classical band struc-
ture of the continuous dispersion of energy as a function of momentum. Unlike quantum wells
(QWs) and quantum wires (QWi’s), the energy-level structure of quantum dots (QDs) is quite dis-
crete. This unique structure of QDs opens a new chapter both in fundamental physics in which they

307
308 Semiconductor Nanocrystals and Metal Nanoparticles

can be regarded as artificial atoms and in potential applications as devices [3,4]. The density of
states for a bulk material is a function of energy (~E1/2), while in a zero-dimensional (QD) crystal,
the density of states is described by a discrete δ-function, (δ(E)) [5], due to the quantum confine-
ment effect.

9.2 SEMICONDUCTOR NANOSTRUCTURES

One of the important features of semiconductor nanostructures is the flexibility of controlling and
designing the properties of such materials [5]. In nanostructures, ideally 1–50 nm in scale, the
dimensions commensurate with the de Broglie wavelength of the charge carriers so that quantum
confinement effects become important and the properties of the semiconductors are significantly
modified [5].
Advanced semiconductor growth techniques, such as molecular beam epitaxy (MBE) and metal
organic chemical vapor deposition (MOCVD), allow the fabrication of various semiconductor
nanostructures or low-dimensional structures. Such low-dimensional structures include (1) QWs,
where the charge carriers are confined along the growth direction but are free in other two direc-
tions; (2) QWi’s, where the charge carriers are confined in two directions and allow free carrier
motion in only one dimension; and (3) QDs, where the charge carriers are confined in all the three
directions (see Figure 9.1). QW or QWi confinement gains the electron at some degrees of free-
dom, but it still gives the electrons at least one direction to propagate. On the other hand, today’s
technology allows us to create QD nanostructures where electron subbands were quantized in all
the degrees of freedom [6]. Strong interband (IB) transitions are possible in these low-dimensional
structures because there is a probability of a strong overlap between the wave functions of elec-
trons and holes.

9.3 QUANTUM DOTS

The term “quantum dot” is usually used to describe a semiconductor nanocrystal. QDs are
zero-dimensional semiconductor systems created at nanoscale. This confinement results in prop-
erties that are not found in bulk materials; see Figure 9.2. One of the main differences between
QDs and traditional semiconductors is the tunability by both the dot size and composition [7].

ρDOS ρDOS

E–1/2 δ(E)

E E
(c) (d)

ρDOS ρDOS
z

E1/2 E0 = const.
y
E E
x
(a) (b)

FIGURE 9.1 Electronic density of states of semiconductor crystals for (a) bulk, (b) quantum well, (c) quantum
wire, and (d) quantum dot.
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 309

1D(e)
Energy
1P(e)
CB
CB 1S(e)

Eg (Bulk) Eg (QD)

vB 1S(h)
vB
1P(h)
1D(h)
Bulk QDs

FIGURE 9.2 Comparison of energy levels between a bulk and a quantum dot semiconductor.

The importance of QDs is originated from the fact that their electrical conductivity can be altered
by an external stimulus such as voltage or photon flux. QDs are very sensitive to the surface proper-
ties due to a high surface to volume ratio. Surface states play a dominant role in these systems, act-
ing as efficient traps for electrons and holes, and thus surface passivation is essential for fabricating
practical semiconductor devices based on such low-dimensional structures [5].

9.4 SEMICONDUCTOR NONLINEARITIES

Semiconductor materials play an important role in nonlinear optics because they both produce a
large nonlinear optical response and lend themselves to the construction of integrated devices in
which electronic devices, semiconductor lasers, and nonlinear optical components are all fabricated
on a single semiconductor substrate [8].
Modern technology has allowed scientists to fabricate high-precision semiconductor QD nano-
structures. In these structures, the precise engineering would enable to confine the motion of
charge carriers in three dimensions. So, a great deal of work has been performed in this area [9,10].
Electronic and optical properties of QWs with an applied external electric field are of increasing
interest [11]. An analytical relation for energy subband calculations is stated earlier and gives a good
result compared with experimental measurements [10–12] and numerical calculations [13,14].
By adding an additional distortion to the energy subbands through the application of an external
electric field, one can tune QD nonlinear properties and change the emission spectrum of photo-
luminescence, both in intensity and wavelength [15]. This results in a precise control of oscillator
strength, which opens the way for the development of practical devices such as optical filters and
color-tunable sources [16]. Thus, studying of nonlinearity in QD nanostructures under an applied
electric field is important for a large number of device applications.
Dane et al. [17] studied the effect of an electric field on the binding energy of a shallow donor
impurity in a spherical QD with an infinite barrier. Xie [18] studied the second-order nonlinear
susceptibility (SONS) in QDs with a quantum disk shape under an applied electric field. Baskoutas
et al. [19] studied nonlinear optical rectification (OR) in semiparabolic QDs, where they found that
SONS depends on the type of quantum confinement. Vaseghi et al. [20] studied OR in cubic QDs
with an infinite potential barrier, where they found that both the dot size and electric field strength
increase with OR. Second-harmonic generation (SHG) in cubic QDs with infinite potential under an
applied electric field is considered in the work of Shao et al. [21], where they found a nonmonotonic
behavior of SONS with cubic length and applied field. The aforementioned works either consider
parabolic confinement or infinite potential. The parabolic confinement gives an equidistant energy
subband that is far from the experiment. Also, the use of an infinite barrier limits the accuracy
310 Semiconductor Nanocrystals and Metal Nanoparticles

compared with experimental data. Although this type of calculations to get an overview of the
physical problem is important, it is impractical to consider the finite barrier for real calculations.
The accurate description of QD states requires a multiband k · p calculation, which is limited by
the knowledge of the exact shape and composition of QDs [22]. Beyond this tedious calculation, it
is required to build a model to calculate the QD energy subbands under the applied electric field.
No work deals with QDs in the shape of quantum disks with the finite potential barrier under the
applied electric field. This shape of dots was considered in a large number of literatures and self-
assembled QDs can be approximated to it, for example, see [22–29]. Thus, starting from the quan-
tum disk model [12], we introduced the electric field effect on QD subbands, where the subbands are
shifted to higher energies with the field.

9.5 NONLINEAR OPTICAL SUSCEPTIBILITY

Nonlinear optical properties of semiconductors have received much attention in recent years. The
very large optical nonlinearities in semiconductors have offered promise for practical applications
in low-power, high-speed, room-temperature optical switching, and signal processing devices [30].
Nonlinear optics has been rapidly growing as a scientific field in recent decades. It is based on
the phenomena related to the interaction of intense coherent light radiation with matter. Nonlinear
optics is the study of interactions of light with matter under conditions in which the nonlinear
response plays an important role. During the past three decades, optics has secured a good place
in application areas previously dominated by electronics. Developments in the field of nonlinear
optics promise for important applications in optical information processing, telecommunications,
and integrated optics. Because of the emergence of this field from solid-state physics in which inor-
ganic semiconductors, insulators, and crystals have constituted a major part of the scientific base,
the early experimental and theoretical investigations were primarily concerned with materials from
these classes [31].
Nonlinear optical phenomena are “nonlinear” in the sense that they occur when the response
of a material system to an applied optical field depends in a nonlinear manner on the strength of
the optical field. For example, SHG occurs as a result of the part of the atomic response that scales
quadratically with the strength of the applied optical field. Consequently, the intensity of light
generated at the second-harmonic frequency tends to increase as the square of the intensity of the
applied field [8].
Second-order nonlinear optical interactions can occur only in noncentrosymmetric crystals, that
is, in crystals that do not display inversion symmetry. Since liquids, gases, amorphous solids (such
as glass), and even many crystals display inversion symmetry, SONS (χ(2)) vanishes identically for
such media, and consequently such materials cannot produce second-order nonlinear optical inter-
actions. On the other hand, third-order nonlinear optical interactions (i.e., those described by a
third-order nonlinear susceptibility χ(3)) can occur for both centrosymmetric and noncentrosym-
metric media [8].
A linear dielectric medium is characterized by the linear relation P = ε0 · χE, between the polar-
ization density, P, and the electric field, E, where ε0 is the permittivity of free space and χ is the opti-
cal susceptibility of the medium. A nonlinear dielectric medium, on the other hand, is characterized
by a nonlinear relation between P and E; see Figure 9.3. The nonlinearity may be of microscopic or
macroscopic origin. The polarization density P = Nρ is a product of the individual dipole moment ρ
induced by the applied electric field E and the number density of dipole moments N. The nonlinear
behavior may reside either in ρ or in N. The relation between P and E is linear when E is small but
becomes nonlinear when E acquires values comparable to interatomic electric fields. Since exter-
nally applied optical electric fields are small in comparison with characteristic interatomic or crys-
talline fields, even when focused laser light is used, the nonlinearity is usually weak. The relation
between P and E is then approximately linear for small E [32].
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 311

FIGURE 9.3 The P–E relation for a linear dielectric medium (blue line) and a nonlinear medium
(red line).

9.6 SECOND-ORDER NONLINEARITY

In the regime of conventional optics, the electric polarization vector P is simply assumed to be linearly
proportional to the electric field strength E of an applied optical wave, that is [33]

P = e0cE (9.1)

where
ε0 is the free-space permittivity
χ is the susceptibility of a given medium

A plot of P versus E is a straight line. Equation 9.1 is valid for field strengths of conventional
sources. The quantity χ is a constant only in the sense of being independent of E; its magnitude is a
function of the frequency. With sufficiently intense laser radiation, this relation does not hold good
and has to be generalized to Equation 9.2, which can be written in the following vector form, as by
a power series [33]:

P(t ) = e0 éëc(1) E(t ) + c( 2 ) E 2 (t ) + c(3) E 3 (t ) + ùû (9.2)

The second-order nonlinear effects can occur in the noncentrosymmetrical crystals only. In the
dielectric dipole approximation, isotropic media and centrosymmetrical crystals cannot be used to
generate second-order nonlinear effects. Therefore, the media for SHG should be the crystals hav-
ing no inversion symmetry. This requirement is the same as that for the piezoelectric effect; thus,
all SHG crystals are piezoelectric crystals although the physical mechanisms for these two effects
are different [33].
In SHG, the combination (addition) of two photons of the same frequency was considered to
produce a single photon of twice the original frequency. Generalization of this process allows the
case in which the two photons have different frequencies. SHG is a nonlinear process where a pho-
ton at frequency 2ω is generated from the interaction between intense light at frequency ω and a
nonlinear medium. The interaction of weak light field with matter is dominated by a linear process.
312 Semiconductor Nanocrystals and Metal Nanoparticles

With a very intense light field, nonlinear process such as SHG becomes observable [34]. OR in
anisotropic media is well known and was first observed in a potassium dihydrogen phosphate
crystal [8]. The magnitude of the induced polarization was found to be proportional to the square
of the optical electric field amplitude. For the effect to arise, the medium, isotropic or anisotropic,
needs to be of sufficiently low symmetry such that the optically induced electric polarization does
not reverse exactly with the optical field [35]. The second-order nonlinear polarization created in
the crystal is given by [36]

P ( 2 ) (t ) = e0c( 2 ) E 2 (t ) = 2c( 2 ) EE * + (c( 2 ) E 2e -2iwt ) + c.c. (9.3)

The second-order polarization combines two parts: The first part is with zero frequency that leads to
the generation of a static electric field within the nonlinear crystal and is known as OR. The second
part contains e−2iwt frequency that also leads to the generation of the radiation at the SHG [36].
Let us consider the process of SHG, which is illustrated schematically in Figure 9.4. Here, a
laser beam is incident upon a crystal for which the second-order susceptibility χ(2) is nonzero.
The nonlinear polarization that is created in such a crystal is given according to Equation 9.3.
Under proper experimental conditions, the process of SHG can be so efficient that nearly all of the
power in the incident beam at frequency ω is converted into radiation at the second-harmonic fre-
quency 2ω. One common use of SHG is to convert the output of a fixed-frequency laser to a differ-
ent spectral region. SHG can be visualized by considering the interaction in terms of the exchange
of photons between the various frequency components of the field. According to this picture, which
is illustrated in part (b) of Figure 9.4, two photons of frequency ω are destroyed and a photon of
frequency 2ω is simultaneously created in a single quantum-mechanical process [8].
Let the optical field incident upon a second-order nonlinear optical medium consists of two
­distinct frequency components; it can be represented by

E(t ) = E1e -iw1t + E2e -iw2 t + c.c. (9.4)

The nonlinear polarization, using Equation 9.2, is given by

P 2 (t ) = e0c( 2 ) é E12e -2iw1t + E22e -2iw2 t + 2 E1E2e -i ( w1 + w2 )t + 2 E1E2*e -i ( w1 - w2 )t + c.c.ù

ëê ûú

+ 2eoc( 2 )[ E1E1* + E2 E2* ] (9.5)

ω
ω
ω 2ω
χ(2) 2ω

(a) (b)

FIGURE 9.4 (a) Geometry of second-harmonic generation. (b) Energy-level diagram describing second-
harmonic generation.
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 313

Each expression in this equation refers to a nonlinear physical process. The first two terms with
e−2iwt dependence are related to SHG, the third term powers to (w1 + w2) refers to sum-frequency
generation (SFG), the fourth term powers to (w1 − w2) is the difference-frequency generation
(DFG), while the last two terms refer to the OR process. There is also a response to the negative
of each of nonzero frequencies. It is not necessary to take explicit account of both the positive
and negative frequency components [8]. SHG, SFG, DFG, and OR are four different nonzero
frequency components present in the nonlinear polarization. However, typically no more than
one of these frequency components will be present with any appreciable intensity in the radiation
generated by the nonlinear optical interaction. The reason for this behavior is that the nonlinear
polarization can efficiently produce an output signal only if a certain phase-matching condition
is satisfied, and usually this condition cannot be satisfied for more than one frequency component
of the nonlinear polarization. Operationally, one often chooses which frequency component will
be radiated by properly selecting the polarization of the input radiation and the orientation of the
nonlinear crystal [8].
The process of SFG is illustrated in Figure 9.5. In many ways, it is analogous to that of SHG,
except that in SFG the two input waves are at different frequencies. The process of DFG is illus-
trated in Figure 9.6. Superficially, DFG and SFG appear to be a very similar process [8].
Nonlinear optics has been growing rapidly as an important scientific field in recent decades. It
is based on a phenomenon related to the interaction of intense coherent light radiation with matter.
Development in this field is promised for important applications in the optical information process-
ing, telecommunications, and integrated optics. Because of the emergence of this field from solid-
state physics, in which inorganic semiconductors, insulators, and crystals have constituted a major
part of the scientific base, the early experimental and theoretical investigations were primarily con-
cerned with the materials from these classes [31]. Modern technology has allowed scientists to
fabricate high-precision semiconductor nanostructures, QDs. Therefore, they tremendously attract
the attention due to their unique physical properties and their potential applications in micro- and
nano-optoelectronic devices [9,10].

ω2
ω1
ω3 = ω1 + ω2 ω3
ω2 χ(2)

ω1

(a) (b)

FIGURE 9.5 Sum-frequency generation. (a) Geometry of the interaction. (b) Energy-level description.

ω2
ω1
ω 3 = ω 1 – ω2 ω1
ω2 χ(2)
ω3

(a) (b)

FIGURE 9.6 Difference-frequency generation. (a) Geometry of the interaction. (b) Energy-level description.
314 Semiconductor Nanocrystals and Metal Nanoparticles

One of the most basic nonlinear processes is the SONS. It requires a symmetry breaking. Thus,
it can be explored in QWs having asymmetry resulted from the modulation doping or from the
application of an external electric field. QDs are considered as asymmetric structures due to their
inhomogeneity, and thus SONS in QDs is calculated in some articles without consideration of an
applied electric field [37]. Third-order nonlinearity in QDs was studied in detail; as an example,
see [23–26,38], considering zero-diagonal matrix elements (symmetry is found).
QDs are promising candidates to achieve large nonlinear susceptibilities. In nanostructures,
nonlinear susceptibility increases due to two factors: (1) large dipole matrix elements associated
with intersubband (ISB) transitions and (2) these ISB transitions that can be adjusted by changing
the QW size [37]. Additionally, nonlinear phenomena at nanoscale are different from that at bulk
(conventional structures). The conventional and second-harmonic waves must propagate in phase to
fulfill the phase-matching condition to create constructive interference and, then, increase conver-
sion efficiency. For small photonic cavities, the classical phase-matching conditions are replaced by
a spatial overlap of localized modes [39].
Earlier studies of SONS in QDs depend on the asymmetrical shape of the self-assembled QDs.
In [37], a lens-shaped QD with infinite barrier potential was considered by Brunhes et al. to study
SONS. They predicted a giant SONS compared with the bulk and QW response due to the achieve-
ment of resonance conditions with intraband transitions. A deviation from experimental results
was observed, which can be reasoned to the infinite potential considered. The research work in [37]
was developed in [40] using the finite potential barrier, and a good agreement was obtained between
theoretical predictions and experimental measurements of SONS.
In addition to the asymmetrical shape of spontaneously formed QDs during the growth ­process,
most of the works dealing with SONS in QDs under applied electric field consider an infinite
­barrier or a simple harmonic oscillator type. This is resultant from the complexity associated with
the numerical calculations and then only small number of studies dealt with QDs under the applied
electric field [15].
The purpose of this chapter is to show that a record value of SONS can be achieved using QD
IB and ISB transitions. To demonstrate this effect, the chosen model system was InAs/InGaAs
self-assembled QDs, which corresponds to the standard Stranski–Krastanov growth mode [37].
Although the presented results are directly connected to the shape and composition of the QDs, sim-
ilar features are expected to occur for other types or shapes of QDs. The QD energy subbands are
first calculated in the effective mass approximation by solving the three-dimensional Schrodinger
equation. Then, SONS is computed in both the conduction and valence bands from the calculated
energy dependence of the confined subbands.
Recent development in the field of nonlinear optics has been pushing nonlinear optical materials
into practical applications. Nonlinear optical materials are those in which light waves can interact
with each other [41]. To measure the nonlinear response of matter to electromagnetic waves in the
optical region, in general, high fields are necessary, starting at about 1 kV/cm. The corresponding
light intensities of some kW/cm2 necessitate laser beams. As laser physics started with the ruby laser
with its high pulse intensities, it took only few years after the invention of the laser [42] that many
classical experiments in nonlinear optics were successfully performed.

9.7 QUANTUM DOT STRUCTURE UNDER STUDY

The QD structure simulated in this theoretical study is a 10-fold InAs QD layer grown by
MBE at NanoSemiconductor GmbH in Germany [26,28]. Each QD array is covered with an
InGaAs wetting layer (WL) and a 33 nm thick GaAs barrier layer. Each QD active layer is sand-
wiched between 1.5 µm thick AlGaAs cladding layers. An InAs QD is treated as a quantum disk
with a radius of a and a height of h. The corresponding material parameters used are listed in
Tables 9.1 and 9.2.
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 315

TABLE 9.1
Experimental Parameters
Parameter InAs (Dot) InxGa1−xAs (Barrier)
Eg (eV) 0.354 0.75 at x = 0.47
m*e 0.023mo 0.041mo
m*h 0.4mo 0.46mo
ζ 41 —
N 5 × 1010 cm−2 —

TABLE 9.2
Experimental Parameters of Relaxations
and Frequencies [26]
Parameter InAs (Dot) InGaAs (Barrier)
γ21 (1/ps) 6.25 —
γ20 (1/ps) 1 —
γ32 (1/ps) 0.333 —
γ31 (1/ps) 0.2 —
γ10 (1/ps) 0.3 —
γ12 (1/ps) 0.833 —
γ02 (1/ps) 0.33 —
γ23 (1/ps) 1 —
γ13 (1/ps) 0.1 —
γ01 (1/ps) 0.25 —
Ωs (eV) 0.1 —
Ωp (eV) 1 —

9.8 QUANTUM DISK MODEL UNDER APPLIED ELECTRIC FIELD

Dots here are considered as a quantum disk with a radius of a and a height of h grown on WL in
the form of a QW, with a finite constant potential assumed for both quantum disk and WL. The
Hamiltonian in the cylindrical coordinates (ρ, ϕ, z) is given by

 2 é 1 ¶ æ 1 ¶ ö 1 ¶2 ¶2 ù
H =- ê ç ÷ + + ú +V (9.6)
2m* ë r ¶r è r ¶r ø r ¶f ¶z 2 û
2 2

where the effective mass is m* = m*d inside the disk and m* = m*b in the barrier. Similarly, the elec-
tric potential is V = Vd inside the disk and V = V b in the barrier. For strongly confined nanostruc-
tures, the electron–hole Coulomb interaction is surpassed by the quantization energy [43]. This is
used in a large number of works that deal with QDs under an applied electric field. For example,
in [21], the Hamiltonian is the same, but the researchers considered infinite potential that is far
from practice although it has theoretical importance in viewing the problem. In [44], the similar
Hamiltonian is used, but the researchers considered a harmonic oscillator type of confinement
potential that makes energy subbands as that of the simpler problem of an electron in a box where
316 Semiconductor Nanocrystals and Metal Nanoparticles

energy subbands become multipliers of ћw, with the addition of the part due to the applied field.
Thus, our Hamiltonian is acceptable and more practical. Solving the Schrodinger equation under
the parabolic band model gives the wave function of the quantum disk. Each state can be charac-
terized by three integral quantum numbers (nml), where nm and l correspond to ρ–ϕ (transverse)
and z-dependence, respectively. The wave function of the state (nml) at the position r = (ρ, ϕ, z) is
expressed as [12]

ì J m ( pr) cos(kz z)
ï -a( z -h / 2)
ï J m ( pr) cos(kz h / 2)e r £ a and z > h/2
imf ï
e ï J m ( pr)
fd (r ) = Cnm í K m (qr) cos(kz z ) r > a and z £ h/2 (9.7)
2p ï K m (qr)
ï J ( pr) -a( z -h / 2)
ï m K m (qr) cos(kz h / 2)e Otherwise
ïî K m (qr)

where
Jm(pρ) and Km(qρ) are the Bessel function of the first kind and the modified Bessel function of
the second kind, respectively
Cnm is the normalization constant
p, q, kz, and α are constants that are determined from the boundary conditions at the interface
between the quantum disk and the surrounding material

If the separation of variables is assumed in the solution of the Hamiltonian, an approximate wave
function of the quantum disk can be obtained [42] by solving the well-known problems of the two-
dimensional circular potential well in the ρ–ϕ direction and the one-dimensional square potential
well in the z-direction. In the ρ–ϕ direction, we have a solution of the form

eimf ìïC1J m ( pr) r £ a

Y(r, f) = í (9.8)
2p ïîC2 K m (qr) r > a

where

2m*d ( Er - Vd ) 2m*b (Vb - Er )

p= and q =
 

Using the boundary condition in which the wave function Ψ and its first derivative (dΨ/dt) are
divided by the effective mass (i.e., (1/m*)(dΨ/dρ)) and are continuous at the interface between the
barrier and the dot, one can obtain the eigenequation. The ­procedure of derivation is described
well in [12]. In [27], the results of the model are compared with that obtained from tight-binding
calculations and are found convenient with it. For convenience, if the potential in the disk is taken
as Vd = 0, the transverse eigenenergy E ρ is given by [12]

 2 ( pr)2
Er = (9.9)
2m*d a
2
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 317

The solution for the wave function for the z-dependence in a finite QW of width L and depth Vo in the
presence of a constant electric field F lies along the positive direction of the well z [13]. Generally,
when an electric field is applied to a QW structure as schematically illustrated in Figure 9.7, the
profile of the potential will be changed. The total potential is given by [45]

V ( z, F ) = V ( z, 0) - ezF (9.10)

where
V(z, 0) is the potential profile of the QW
F is the applied electric field in (kV/cm)
e is the electronic charge
z is the associated spatial coordinate

We choose the origin to be at the center of the well.

Substituting Equation 9.10 into the Schrodinger equation, we arrive at the following formula:

2 d2
- y( z) + e Fzy( z) = Ey( z) z £ L /2
2m* dz
2

(9.11)
2 d2
- y( z) + (Vo + e Fz)y( z ) = Ey( z ) z £ L /2
2m* dz
2

where the potential profile of the QD in the z-direction is given by

ìï 0 for z £ L / 2
V ( z) = í (9.12)
îïVo for z ³ L / 2

Potential energy
V(z)

Vo

–L/2 0 L/2 z

FIGURE 9.7 Potential energy profile V(z) for a quantum dot in the z-direction, with depth Vo and width
L subjected to an external electric field. Note that the origin is chosen at the center of the well.
318 Semiconductor Nanocrystals and Metal Nanoparticles

where Vo = Beff [Egw − Egd], in which Beff is the band offset and Egw and Egd are the bandgaps of WL
and QD, respectively. The wave function in the well and barrier regions is described by [13]

ìC3 Ai(h2 ) z > -L /2

ï
Y( z ) = íC4 Ai(h1 ) + D2 Bi(h1 ) z £ L /2 (9.13)
ïC é Bi(h ) + iAi(h ) ù z < - L/2
î 5ë 2 2 û

where
C3, C4, C5, and D2 are constants
Ai and Bi are the homogeneous Airy function

From the properties of Airy function, it is clear that Bi(η2) increases with increasing η2 and becomes
infinity when η2 is at infinity. In order to make the wave function well behaved in the entire region,
this part is not added in the wave function in the region z > −L/2. Note that
1/3
é 2m* ù
h1 = - ê 2
ú ( E - e Fz)
êë (eF ) úû

1/3
(9.14)
é 2m* ù
h2 = - ê 2
ú ( E - Vo - e Fz )
êë (eF ) úû

The required boundary conditions for the coefficients are obtained from the current continuity
­conditions at the heterojunction as

Y (h1 ) z = z = Y (h2 ) z = z
o o

(9.15)
1 dY (h1 ) 1 dY (h2 )
=
mb* dz z = zo m*d dz z = zo

This results in the determinant

Ai(h1+ ) Bi(h1+ ) - Ai(h2+ ) 0

A¢i(h1+ ) B¢i(h1+ ) - A¢i(h2+ ) 0
det =0 (9.16)
Ai(h1- ) Bi(h1- ) 0 - éë Bi(h2- ) + iAi(h2- ) ùû

A¢i(h1- ) B¢i(h1- ) 0 - éë B¢i(h2- ) + iA¢i(h2- ) ùû

where h1± and h2± are the values of η1 and η2 evaluated at z = L/2 and z = −L/2, respectively. Then,
we obtain

ìé æ m*b ö ù é æ m* ö ùü
ï ÷ Ai¢(h1 ) Bi (h2 ) ú + i ê Ai (h1 ) Ai¢(h2 ) - ç b ÷ Ai¢(h1 ) Ai (h2 ) ú ïý
A11 = í ê Ai (h1 ) Bi¢(h2 ) - ç (9.17)
ïî êë ç m*d ÷ ú ê ç m*d ÷ úï
è ø û ë è ø ûþ

ìé æ m* ö ù é æ m* ö ùü
ï ï
A12 = í ê Bi (h1 ) Bi¢(h2 ) - ç b ÷ Bi¢(h1 ) Bi (h2 ) ú + i ê Bi (h1 ) Ai¢(h2 ) - ç b ÷ Bi¢(h1 ) Ai (h2 ) ú ý (9.18)
ê ç * ÷ ú ê ç * ÷ ú
ïî ë è md ø û ë è md ø û ïþ
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 319

ìé æ m*b ö ùü
ï
A21 = í ê A¢i(h1 ) Ai(h2 ) - ç ÷ Ai(h1 ) A¢i(h2 ) ú ïý (9.19)
ïî êë ç m*d ÷ úï
è ø ûþ

ìé æ m*b ö ùü
ïê ï
¢
A22 = - í A i(h2 ) Bi(h1 ) - ç ÷ Ai(h2 ) B i(h1 ) ú ý
¢ (9.20)
ê ç * ÷ ú
ïî ë è md ø û ïþ

ao A11 + bo A12 = 0
(9.21)
ao A21 + bo A22 = 0

A11 A12
=0 (9.22)
A21 A22

The eigenenergy Ez is obtained by solving Equation 9.22. The total eigenenergy of the quantum
disk Ed is, approximately, the summation of the transverse and longitudinal eigenenergies and is
expressed as

Ed = Er + Ez (9.23)

This gives the eigenenergy of the QD structure under the applied electric field.

9.9 CALCULATED QD SUBBANDS UNDER APPLIED ELECTRIC FIELD

Figure 9.8 shows the calculated QD ground (GS) and excited state (ES) conduction subbands
under an applied electric field. The solid lines represent the GS conduction subband (Ec1),
while the dashed lines represent the first ES conduction subband (Ec2). For the height h = 2 nm

1100

1050
Energy levels of CB (meV)

1000
Ec1, h = 2 nm
Ec2
950 Ec1, h = 3 nm
Ec2
900

850

800
10 20 30 40 50 60 70 80 90 100
Electric field (kV/cm)

FIGURE 9.8 Quantum dot energy conduction subbands versus the applied electric field for heights h = 2 nm
(the red line) and h = 3 nm (the blue line). The disk radius a = 13 nm. The solid lines are for ground state (Ec1),
while the dashed lines are for first excited state (Ec2).
320 Semiconductor Nanocrystals and Metal Nanoparticles

–40

–50

First energy level of VB (meV)

–60

–70

–80
Ev1, h = 2 nm
–90
Ev1, h = 3 nm
–100

–110

–120

–130
10 20 30 40 50 60 70 80 90 100
Electric field (kV/cm)

FIGURE 9.9 Quantum dot ground-state valence subband versus the applied electric field for heights
h = 2 nm (the red line) and h = 3 nm (the blue line). The disk radius is a = 13 nm.

(the red lines) and when the field changes from 10 to 100 V/m, Ec1 changes by 35.4 meV, while Ec2
changes by 78.9 meV. This is a Stark shift for QD subbands. Thus, a nonmonotonic change was
obtained with the applied field on the QD ES conduction subband. For h = 3 nm (the blue lines),
approximately similar changes were obtained. A nonmonotonic change of energy subbands with
an electric field is also shown in [45,46]. This behavior leads us to use Airy functions to describe
wave functions under an applied electric field. From Figure 9.9, it is shown that subband energy
becomes higher at some value of the applied electric field. This depends on the type of confine-
ment, QD shape, and size. This result coincides with the conclusion drawn in [19,45,46]. This can
be justified by a resonance that occurs between the applied field and the subband energy.
Figure 9.9 shows the change of QD GS valence subband for disk radius a = 13 nm and for two
heights (h = 2 nm the red line and h = 3 nm the black line) when the field changes by the same range
(10–100 V/m). For h = 2 nm, GS shifts by 89 meV. This shows that the valence subbands are more
sensitive to the applied electric field.
Additionally, as it is shown in Figure 9.10, GS conduction subband was plotted for different
quantum disk radii and at three field values. The subband energy decreased with increasing the disk
radius and increased with increasing the electric field. Electrons and holes are separated under the
applied electric field, and they are pushed apart by the field. Reducing the QD size increases the
confinement of carrier wave functions. The shift of QD energy subbands by the Stark effect com-
petes with the shift due to the quantum-size effect. Increasing the disk radius by 1 nm reduces the
subband energy by ~3 meV, while increasing electric field by 1 kV/cm increases the subband energy
by ~0.7 meV.

9.10 DENSITY MATRIX FORMULATION OF OPTICAL SUSCEPTIBILITY

To derive SONS relation, we use the density matrix approach. It has been proven very useful to
study quantum electronic processes in different material systems [47]. The density operator ρ satis-
fies the equation of motion

¶ -i
r = [ H , r] (9.24)
¶t 
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 321

h = 2 nm
0.865
F = 10 kV/cm
0.86
F = 50 kV/cm

First energy level of CB (eV)

0.855 F = 100 kV/cm

0.85

0.845

0.84

0.835

0.83

0.825
13 13.5 14 14.5 15 15.5 16
Radius (nm)

FIGURE 9.10 Quantum dot ground-state energy conduction subbands versus radius for three values of
applied electric field. The disk height is h = 2 nm.

The Hamiltonian operator consists of three parts

H = H 0 + H ¢ + Hrandom (9.25)

where H0 is the unperturbed Hamiltonian and H′ accounts for the interaction such as the electron–
photon interaction, which is given by

3
H ¢ = -m × E (t ) = - å m E (t )
j =1
j
j (9.26)

where µ is the dipole operator, µ = er, and e = −|e| for electrons and +|e| for holes. E is the electric
field. Hrandom includes the relaxation effects due to incoherent scattering processes. The equation of
motion, Equation 9.24, becomes

¶ -i æ ¶r ö
r = [ H 0 + H ¢, r] + ç ÷ (9.27)
¶t  è ¶t ørelax

where

æ ¶r ö -i
ç ¶t ÷ = éë Hrandom, r - r( 0 ) ùû (9.28)
è ørelax 

which can be written in terms of T1 and T2 time constants as follows:

¶ ruu - ruu
(0)

¶t
(
ruu - ruu
(0)
)
relax
=-
(T1 )uu
(9.29)

æ¶ ö r
ç ¶t ruu ÷ = - uu
(T2 )uv
for u = v (9.30)
è ørelax
322 Semiconductor Nanocrystals and Metal Nanoparticles

For the initial distributions, r(uv0 ) = r(uu0 )duv, which are diagonal for each state. It is convenient to use

1
g uu =
(T1 )uu
(9.31)
1
g uv = for u ¹ v
(T2 )uv

Taking the uv component of equation of motion (Equation 9.27) and using

u H 0r - rH 0 v = ( Eu - Ev )ruv (9.32)

we obtain the density matrix equation in the presence of an optical excitation

å å (m
¶ -i
¶t
ruv = ( Eu - Ev )ruv +

i

j
r
uu¢ u¢v ) (
- ruu¢muj ¢v E j (t ) - g uv ruv - r(u0v) ) (9.33)
u¢ j

In general, we may define

Eu - Ev
wuv = (9.34)


and consider the interaction term H′ = −µ ⋅ E(t) as a small perturbation. The perturbation series gives

r = r( 0 ) + r(1) ( E ) + r( 2 ) ( E 2 ) +  (9.35)

One obtains

å å (m
¶ ( n +1)
¶t
( n +1)
ruv = (-iwuv - g uv )ruv +
i

j
r (n)
uu¢ u¢u - ruu )
¢m u¢v E j (t )
(n) j
(9.36)
u¢ j

For n ≥ 0, consider an optical field given by

N
E(t ) = å E(w )e
a =1
a
- iwa t
(9.37)

The polarization per unit volume is calculated from the trace of the matrix product of the dipole
moment matrix µ and the density matrix

1
P(t ) = Tr[r(t )m]
V

år
1
= uv (t )muv
V uv

1
= (raam aa + rabm ba + rbam ab + rbbm bb ) (9.38)
V
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 323

The ith component of the polarization density, to the first order in the optical electric field, is given by

Pi( n ) (t ) =
V
(
1 i (n)
m barab (w) + miabr(ban) (w) e -iwt )
+
V
(
1 i (n)
m barab (-w) + miabr(ban) (-w) eiwt ) (9.39)

The definition of the electric susceptibility χij is obtained using

P ( n ) (t ) = e0c( n ) (w)E(w)e -iwt + e0c( n ) (-w)E(-w)e + iwt (9.40)

By calculating the density operator ρ(n) for the nth order for the two- or three-level system and substitut-
ing into Equation 9.39 and comparing with Equation 9.40, we obtain the optical susceptibility χ(n)(w).
In the case of conventional (i.e., linear) optics, the induced polarization depends linearly on the
electric field strength in a manner that can often be described by the relationship

P (1) (t ) = e0c(1) E(t ) (9.41)

where the constant of proportionality χ(1) is known as the linear susceptibility and ε0 is the permit-
tivity of free space. In nonlinear optics, the optical response can often be described by generalizing
Equation 9.41 by expressing the polarization P(t) as a power series in the field strength E(t) as

P(t ) = e0 éëc(1) E(t ) + c( 2 ) E 2 (t ) + c(3) E 3 (t ) + ùû

º P (1) (t ) + P ( 2 ) (t ) + P (3) (t ) +  (9.42)

The quantities χ(2) and χ(3) are known as the second- and third-order nonlinear optical susceptibili-
ties, respectively.
In the three-level system considered (Figure 9.11) here, only 2 → 1 and 3 → 2 transitions are
dipole allowed [29]. From the earlier relations, the IB transition, r(212), can be derived. Then, SONS
for IB transition is written as

1 ì W sm12 (r11
(0)
- r(220) )(m22 - m11 ) ü
c( 2 ) (ws ) = í ý (9.43)
e 0 î [(w21 - ws ) - ig 21 ]2 þ

QD WL

Δp |3 ≥ C2
Pump
wp M32 = intersubband dipole
|2 ≥ C1
Δs

M21 = interband dipole

Source
ws

|1 ≥ H1

FIGURE 9.11 Schematic diagram of the potential profiles and relevant parameters in a semiconductor quan-
tum dot system. (H1 = the heavy-hole ground state, C1 = the conduction band ground state, and C2 = the first
excited state of conduction band.)
324 Semiconductor Nanocrystals and Metal Nanoparticles

Similarly, for ISB transitions, ISB SONS was derived. It can be written as

1 ìï W m (r( 0 ) - r( 0 ) )(m - m ) üï
p 23 22 33 33 22
c( 2 ) ( w p ) = í ý (9.44)
e 0 ïî éë(w32 - w p ) - ig 32 ùû
2
ïþ

where ω s and ω p are their frequency separations of the signal and the pump, respectively,
Ω s = µ 21Es /ћ, w21 = (E2 − E1)/ћ and Ω p = µ 32 Ep /ћ, w32 = (E3 − E2)/ћ. The ISB dipole moment (µ 32)
between the GS (C1 or level 2) and the first ES (C2 or level 3) of the conduction band is calculated
according to [25]

m32 = fd 3 (r ) er fd 2 (r ) (9.45)

where
e is the unit charge
|r| is the distance

On the other hand, the IB dipole moment (µ21) is expressed as follows [25]:

e
m 21 = z uc pˆ uv fd 2 (r ) fd1 (r ) (9.46)
m0 w21

where
ζ is an enhancement factor due to excitonic effects
m 0 is the free electron mass

9.11 RESULTS ON SONS

The real and imaginary parts of the SONS for IB transition at 10–100 kV/cm electric field strengths
are shown in Figure 9.12. From Figure 9.12a through c, it is shown that both real and imaginary
parts of susceptibility were reduced and the peak wavelength was shifted toward the longer wave-
length by increasing the electric field strength. The susceptibility was reduced by two times and the
peak wavelength was increased by 36 nm when the electric field increases to 50 kV/cm and by 1.5
times and 89 nm when the electric field increases to 100 kV/cm. This is consistent with the experi-
mental evidence that the applied electric field quenches the photoluminescence in QDs [16]. It is a
result of the reduced overlap between electron and hole wave functions, where they are separated by
the electric field due to their opposite charges. Our results of SONS in Figure 9.12 are on the order
of 10 –2 m/V, which is in the range of He and Xie results [48].
For the case of conduction ISB transitions, SONS is shown in Figure 9.13, both real and imagi-
nary parts of SONS were increased by more than two orders of magnitude, and their peak wave-
lengths were increased more than 7000 nm when the field changes from 10 to 100 kV/cm. Note that
the peak wavelength was reduced when the field is 50 kV/cm than that at 10 and 100 kV/cm. This
relates to the position of subband energy when the field was increased to 50 kV/cm as shown in
Figures 9.8 and 9.10.
SONS, for valence ISB transitions, is shown in Figure 9.14. The peak values are of the same
order of that of conduction ISB transitions, while their wavelengths are increased by more than one
order of magnitude. Their obtained wavelengths lie between 33 and 157 µm, which are important in
terahertz applications. The subband energy difference at 10 kV/cm is 0.8, 0.14, and 0.037 eV for the
first IB transition, conduction ISB transition, and first valence ISB, respectively. These differences
can explain the ranges of wavelengths of these structures shown in Figures 9.12 through 9.14.
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 325

Interband
0.05
Re
0.04 F = 10 kV/cm Im

0.03

χ(2) (m/V) 0.02

0.01

–0.01

–0.02

–0.03
1283 1284 1285 1286 1287 1288 1289 1290 1291
(a) Wavelength (nm)

Interband
0.025
Re
0.02 F = 50 kV/cm
Im
0.015

0.01
χ(2) (m/V)

0.005

–0.005

–0.01

–0.015

–0.02
1320 1321 1322 1323 1324 1325 1326
(b) Wavelength (nm)

Interband
0.02
Re
0.015 F = 100 kV/cm Im

0.01
χ(2) (m/V)

0.005

–0.005

–0.01

–0.015
1409 1410 1411 1412 1413 1414 1415
(c) Wavelength (nm)

FIGURE 9.12 The imaginary (the blue line) and the real (the red line) second-order nonlinear susceptibil-
ity for interband transition under applied electric field with strengths of (a) 10 kV/cm, (b) 50 kV/cm, and
(c) 100 kV/cm.
326 Semiconductor Nanocrystals and Metal Nanoparticles

Intersubband
1.5
Re
F = 10 kV/cm Im
1

0.5
χ(2) (m/V)

–0.5

–1
8650 8660 8670 8680 8690 8700 8710 8720 8730 8740 8750
(a) Wavelength (nm)

Intersubband
0.8
Re
0.6 F = 50 kV/cm Im

0.4
χ(2) (m/V)

0.2

–0.2

–0.4

–0.6
7400 7420 7440 7460 7480 7500 7520 7540
(b) Wavelength (nm)

Intersubband
0.8
Re
0.6 F = 100 kV/cm Im

0.4
χ(2) (m/V)

0.2

–0.2

–0.4

–0.6
(c) 6930 6940 6950 6960 6970 6980 6990 7000 7010
Wavelength (nm)

FIGURE 9.13 The imaginary (the blue line) and the real (the red line) second-order nonlinear susceptibility
for conduction intersubband transition under applied electric field with strengths of (a) 10 kV/cm, (b) 50 kV/cm,
and (c) 100 kV/cm.
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 327

VB
5
Re
4 F = 10 kV/cm Im

χ(2) (m/V) 2

–1

–2

–3
1.4 1.45 1.5 1.55 1.6 1.65 1.7 1.75
(a) Wavelength (nm) ×105

VB
0.6
Re
0.5 F = 50 kV/cm
Im
0.4
0.3
0.2
χ(2) (m/V)

0.1
0
–0.1
–0.2
–0.3
–0.4
5.15 5.2 5.25 5.3 5.35 5.4 5.45 5.5
(b) Wavelength (nm) ×104

VB
0.15
Re
Im
0.1 F=100 kV/cm

0.05
χ(2) (m/V)

–0.05

–0.1
3.3 3.32 3.34 3.36 3.38 3.4 3.42
(c) Wavelength (nm) ×104

FIGURE 9.14 The imaginary (the blue line) and the real (the red line) second-order nonlinear susceptibility
for valence intersubband transition under applied electric field with strengths of (a) 10 kV/cm, (b) 50 kV/cm,
and (c) 100 kV/cm.
328 Semiconductor Nanocrystals and Metal Nanoparticles

85

80

75

(abs((M22 – M11)/M21))2
70

65

60

55

50

45
10 20 30 40 50 60 70 80 90 100
Electric field (kV/cm)

FIGURE 9.15 Ratio of momentum matrix elements for quantum dot structure shown in Figure 9.11 versus
the applied electric field.

Our result of SONS for both conduction and valence ISB transitions in Figures 9.13 and
9.14 is higher by one order of magnitude than that of IB transitions shown in Figure 9.12. The main
factor that controls SONS in QDs, here, is the momentum matrix element. For conduction ISB
transitions, the factor µ23(µ 33 − µ22) in Equation 9.44 is higher by one order of magnitude than the
factor µ12(µ22 − µ11) in Equation 9.43 for IB transitions. This can explain why SONS of ISB transi-
tions is higher than that of IB structure. Figure 9.15 shows the ratio |(M22 − M11)/M12| for the IB
structure where a nonmonotonic change is shown.

9.12 SHG, SFG, DFG, AND OR

The second-order nonlinear processes SHG, SFG, DFG, and OR are the simplest examples of non-
linear molecular spectroscopes and will serve as our example of diagrams that describe nonlinear
spectroscopy. Second-order spectroscopes vanish for isotropic samples and are therefore surface
selective. There are various nonlinear optical processes. Some of them involve the generation of a
new frequency that is different from that of the pump source, while some others do not. The former
includes SHG, SFG, and DFG. They are also the topics that will be involved in this chapter. The
processes of SHG, SFG, and DFG are all due to the second-order optical nonlinearity [49].
An electromagnetic field incident on a medium induces bound electrons to oscillate about their
equilibrium position. In the linear regime, the resulting dielectric polarization is proportional to the
applied electric field [8]. The polarization is described by Equation 9.42, where

E(t ) = E1e -iw1t + E2e -iw2 t + c.c. (9.47)

It can be shown that the nonlinear polarization is given by

P 2 (t ) = e0c( 2 ) é E12e -2iw1t + E22e -2iw2 t + 2 E1E2e -i ( w1 + w2 )t + 2 E1E2*e -i ( w1 - w2 )t + c.c.ù

êë úû

+ 2e0c( 2 ) é E1E1* + E2 E2* ù (9.48)

êë úû
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 329

It is convenient to express this result using the notation

P ( 2 ) (t ) = å P(w )e
n
n
- iwn t
(9.49)

where the summation extends over positive and negative ωn frequencies. The complex amplitudes of
various frequency components of the nonlinear polarization are hence given by [8]

P(2w1 ) = e0c( 2 ) E12 (SHG)

P(2w2 ) = e0c( 2 ) E22 (SHG)

P(w1 + w2 ) = 2e0c( 2 ) E1E2 (SFG) (9.50)

P(w1 - w2 ) = 2e0c( 2 ) E1E2* (DFG)

P(0) = 2e0c( 2 ) ( E1E1* + E2 E2* ) (OR )

Note that, in accordance with our complex notation, there is also a response at the negative of each
of the nonzero frequencies just given by

P(-2w1 ) = e0c( 2 ) E1*2 (SHG)

P(-2w2 ) = e0c( 2 ) E2*2 (SHG)

(9.51)
P(-w1 - w2 ) = 2e0c E1*E2* (SFG)
(2)

P(w2 - w1 ) = 2eoc( 2 ) E2 E1* (DFG)

However, since each of these quantities is simply the complex conjugate of one of the quantities
given in Equation 9.50, it is not necessary to take explicit account of both the positive and negative
frequency components.
For SFG, according to Equation 9.51, the complex amplitude of the nonlinear polarization
describing this process is given by the expression [8]

P(w1 + w2 ) = 2e0c( 2 ) E1E2 (9.52)

In many ways, the process of SFG is analogous to that of SHG, except that in SFG where the two
input waves are at different frequencies.
For DFG, the process of DFG is described by a nonlinear polarization of the form [8]

P(w1 - w2 ) = 2e0c( 2 ) E1E2* (9.53)

Here, the frequency of the generated wave is the difference of those of the applied fields, and DFG
and SFG appear to be of very similar processes. We see that the conservation of energy requires that
for every photon that is created at the difference frequency ω3 = ω1 − ω2, a photon at the higher input
frequency (ω1) must be destroyed, and a photon at the lower input frequency (ω2) must be created.
330 Semiconductor Nanocrystals and Metal Nanoparticles

Thus, the lower frequency input field is amplified by the process of DFG. For this reason, the pro-
cess of DFG is also known as optical parametric amplification. According to the photon energy–
level description of DFG, the atom, first, absorbs a photon of frequency ω1 and jumps to the highest
virtual level. This level decays by a two-photon emission process that is stimulated by the presence
of the ω2 field, which is already present. Two-photon emission can occur even if the ω2 field is not
applied. The generated fields in such a case are very much weaker, since they are created by spon-
taneous two-photon emission from a virtual level [8].
The definition of the electric susceptibility χij is obtained using

P ( n ) (2ws - 2ws ) = e0c( n ) (2ws )E(2ws )e -i ( 2 ws )t

+ e0c( n ) (-2ws )E(-2ws )e + i ( 2 ws )t (9.54)

It is not necessary to take explicit account of both the positive and negative frequency components [8].
Using the three-level system shown in Figure 9.11, in this configuration, only 2 → 1 and 3 → 2
dipole transitions are allowed, where ωs and ωp are their frequency separations of the signal and the
pump, respectively, and the optical susceptibility χ(n)(2ws) can be written as

Nm21m12 ìï (r( 0 ) - r( 0 ) )(m - m ) üï

c( 2 ) (2ws ) = í
11 22 22 11
2 ý
(9.55)
 2e0 ïî éë(w21 - 2ws ) - ig 21 ùû ïþ

Similarly, the SONS for ISB transitions can be written as

Nm32m 23 ìï (r(220) - r(330) )(m33 - m 22 ) üï

c ( 2 ) (2 w p ) = í 2 ý
(9.56)
e 0 îï [(w32 - 2w p ) - ig 32 ] þï

Similarly, calculated SFG, DFG, and OR for IB and ISB transitions are derived by the same method.

9.12.1 Sum-Frequency Generation

The SFG relation for our system can be written as

Nm13m31 ïì W sm32 (r11(0)

- r(220) )
c( 2 ) (ws + w p ) = í
2 2e0 ïî [(w21 - ws ) - ig 21 ][(w31 - ws - w p ) - ig 31 ]

W pm 21 (r(220) - r33
(0)
) üï
- ý (9.57)
[(w32 - w p ) - ig 32 ][(w31 - ws - w p ) - ig 31 ] þï

where Ω s = µ21Es/ћ, w21 = (E2 − E1)/ћ and Ω p = µ 32Ep/ћ, w32 = (E3 − E2)/ћ.

9.12.2 Difference-Frequency Generation
The DFG relation corresponding to our system can be written as

Nm12m32 ìï W W*(r( 0 ) - r( 0 ) )(r( 0 ) - r( 0 ) )(m - m )(m - m ) ïü

s p 11 22 33 22 22 11 22 33
c( 2 ) (ws - w p ) = í ý (9.58)
2 2e0 ïî [( w21 - w s ) - i g 21 ]2
[( w 23 + w p ) + i g 23 ]2
ïþ

where W*p = m32 E*p /.

Second-Order Nonlinear Susceptibility in Quantum Dot Structures 331

9.12.3 Optical Rectification
The OR for IB transitions (see Figure 9.11) is

Nm12m 21 ïì W s W*s (r11

(0)
- r(220) )(r(220) - r11
(0)
)(m22 - m11 )(m 22 - m11 ) ïü
c ( 2 ) (0 ) = í ý (9.59)
2 e0 îï
2
[ w21 - ig 21 ] [ w21 + ig12 ]2
2
þï

The OR for ISB transitions is

Nm32m23 ìï W W*(r( 0 ) - r( 0 ) )(r( 0 ) - r( 0 ) )(m - m )(m - m ) üï

p p 22 33 33 22 33 22 22 33
c ( 2 ) (0 ) = í ý (9.60)
2 2e0 ïî [ w 32 - i g 32 ]2
[ w 23 + i g 23 ]2
ïþ

9.13 INHOMOGENEITY IN QDs

The aforementioned works consider either parabolic confinement or infinite potential. The para-
bolic confinement gives an equidistant energy subband that is far from experiment. Also, the use
of an infinite barrier limits the accuracy comparison with experimental data. Although this type of
calculations to get an overview of the physical problem is important, it is impractical to consider a
finite barrier for real calculations.
Consider that QD inhomogeneity is important from the practical point of view. The inhomogene-
ity was included in the study of SONS in QDs in the work of Brunhes et al. [37,40], but they took a
QD distribution normalized to the dot height and not as energy distribution. The QD height is not
the only factor that is causing inhomogeneous distribution. Many other factors, such as size, dot
distribution, and imperfections in the QD shape, contribute to the QD inhomogeneity. Thus, one
must cover the spectrum of the dots through the emission spectrum of each one. The energy is the
adequate factor to represent. This is what the authors have done here. We need to deal with SONS
in QDs and derive that their relations depend on the structure consideration. This is preferred than
the present forms applied for any structure atom, QW, QD, or any other types of electronic transition
system studied. This is done also here.
The inhomogeneous SONS in QDs is then given by

ò
c( 2 ) = c( 2 ) (ws + w p , w p , ws ) D(w)dw (9.61)

Including QD inhomogeneity makes our formula different from all other researches that deal with
SONS calculations in QDs. It is done here via the convolution over the inhomogeneous density of
states, which is given by [22]

(
æ - w - E i
) ö
2
si 1
exp ç ÷
max
D(w) = eff (9.62)
Vdot 2ps 2 ç
ç 2 s 2 ÷÷
è ø

where
si is the degeneracy number at the QD state (si = 2) in the quantum disk model used here
σ is the spectral variance of QDs
eff
Vdot (= h /N D ) is the effective volume of QDs, h is the dot height, and ND is the areal density of QDs
332 Semiconductor Nanocrystals and Metal Nanoparticles

i
The transition energy at the QD maximum distribution of the ith optical transition is Emax . Note
that although the inhomogeneity is included in the study of SONS in QDs in the work of Brunhes
et al. [37], they take a QD distribution normalized to the dot height not as energy distribution as in
Equation 9.62.

9.14 RESULTS AND DISCUSSION OF SHG, SFG, DFG, AND OR IN QDs

Figure 9.16 shows SONS for the first structure versus SHG wavelength at a 10 kV/cm applied field
(Figure 9.16a), where 0.4 m/V SONS peak is obtained at 5095.5 nm peak wavelength. Increasing
the field to 100 kV/cm reduces SONS peak to 0.12 m/V, and its peak wavelength is red shifted by
approximately 200 nm as in Figure 9.16b.

F = 10 kV/cm
0.8

0.6 Re homo
Im
Re inhomo
0.4 Im
SHG (m/V)

0.2

–0.2

–0.4

–0.6
5088 5090 5092 5094 5096 5098 5100 5102 5104
(a) λSHG (nm)

F = 100 kV/cm
0.25

0.2 Re homo
Im
0.15 Re inhomo
Im
0.1
SHG (m/V)

0.05

–0.05

–0.1

–0.15
5280 5285 5290 5295 5300 5305
(b) λSHG (nm)

FIGURE 9.16 Second-order nonlinear susceptibility real (red lines) and imaginary (blue lines) parts by
second-harmonic generation (SHG) versus SHG wavelength for homogeneous (solid lines) and inhomoge-
neous (dotted lines) of the first quantum dot structure at (a) 10 kV/cm and (b) 100 kV/cm electric field strengths.
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 333

Including the QD inhomogeneity in the calculations is important to cover synthesis imper-

fections. The SONS calculated with inhomogeneous density of states is shown as dashed curves in
Figure 9.16a and b. It reduces the peak value of SONS to approximately its half value, this result
is shown for both real and imaginary parts of SONS, and this is with our recent result in [24] for
the third-order nonlinear susceptibility. Thus, it is important to include inhomogeneity in SONS
calculations in QDs.
The ISB case (second structure) is shown in Figure 9.17. When the applied field is 10 kV/m,
the spectrum is peaked at 1.18 mm as shown in Figure 9.17a, that is, it is extended by 1.175 mm as
compared with Figure 9.16a, while at a 100 kV/m applied field, it is peaked at 0.28 mm as shown
in Figure 9.17b, and at this latter case, SONS is reduced by more than three times (in both real and
imaginary parts).

VB, F = 10 kV/cm
0.25
0.2 Re homo
Im
0.15 Re inhomo
Im
0.1
SHG (m/V)

0.05
0
–0.05
–0.1
–0.15
–0.2
0.9 1 1.1 1.2 1.3 1.4 1.5
(a) λSHG (mm)

×10–3 VB, F = 100 kV/cm

2

1.5 Re homo
Im
1 Re inhomo
Im
SHG (m/V)

0.5

–0.5

–1

–1.5
0.23 0.235 0.24 0.245 0.25 0.255 0.26 0.265 0.27
(b) λSHG (mm)

FIGURE 9.17 Second-order nonlinear susceptibility real (red lines) and imaginary (blue lines) parts by sec-
ond-harmonic generation (SHG) versus SHG wavelength for homogeneous (solid lines) and inhomogeneous
(dotted lines) of the second quantum dot structure at (a) 10 kV/cm and (b) 100 kV/cm electric field strengths.
334 Semiconductor Nanocrystals and Metal Nanoparticles

SFG in the first structure is shown in Figure 9.18a, where SONS is peaked at 10 µm wavelength
under the 10 kV/cm applied field and its value (imaginary SONS) is increased to 1.1 m/V. Figure 9.18b
shows SONS when the applied field is increased to 100 kV/cm, where SONS peak becomes 0.35 m/V
at 8 µm wavelength peak.
SFG for the second structure (valence ISB case) under 10 kV/cm applied electric field is shown
in Figure 9.19a, where SONS peaks at 1.363 mm and its value is reduced by three times compared
with IB case. Increasing the field to 100 kV/cm peaks SONS at 0.293 mm, as shown in Figure 9.19b,
and its value was reduced by two orders compared with Figure 9.19a.
DFG in the first structure is shown in Figure 9.20a, where SONS is peaked at 7.4 µm under
10 kV/cm applied field and its value is reduced by seven orders compared with SONS of SHG and
SFG of the first structure (Figures 9.16a and 9.18a). Figure 9.20b shows the case when the field is

F = 10 kV/cm
1.2

1 Re inhomo
Im
0.8

0.6
SFG (m/V)

0.4

0.2

–0.2

–0.4

–0.6
0.997 0.998 0.999 1 1.001 1.002 1.003
λSFG (nm) ×104
(a)

F = 100 kV/cm
0.5
Re inhomo
0.4 Im

0.3
SFG (m/V)

0.2

0.1

–0.1

–0.2
7970 7980 7990 8000 8010 8020 8030 8040 8050
(b) λSFG (nm)

FIGURE 9.18 Second-order nonlinear susceptibility real (red lines) and imaginary (blue lines) parts by
sum-frequency generation (SFG) versus SFG wavelength for inhomogeneous broadening of the first quantum
dot structure at (a) 10 kV/cm and (b) 100 kV/cm electric field strengths.
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 335

×10–4 VB, F = 10 kV/cm

2
Re inhomo
1.5 Im

1
SFG (m/V)
0.5

–0.5

–1

–1.5
1.3633 1.3634 1.3635 1.3636 1.3637 1.3638 1.3639 1.364
(a) λSFG (nm)

×10–6 VB, F = 100 kV/cm

12
10 Re inhomo
Im
8
6
4
SFG (m/V)

2
0
–2
–4
–6
–8
0.2936 0.2936 0.2936 0.2937 0.2937 0.2937 0.2937 0.2937
(b) λSFG (nm)

FIGURE 9.19 Second-order nonlinear susceptibility real (red lines) and imaginary (blue lines) parts by sum-
frequency generation (SFG) versus SFG wavelength for inhomogeneous broadening of the second quantum
dot structure at (a) 10 kV/cm and (b) 100 kV/cm electric field strengths.

increased to 100 kV/cm, where SONS is peaked at 5.3 µm and its value is reduced by one order of
magnitude compared with Figure 9.20a.
Figure 9.21a shows DFG for the second structure at 10 kV/m applied electric field, where a high
SONS is obtained, and it is peaked at 0.468 mm wavelength. Increasing the field to 100 kV/cm,
Figure 9.21b, SONS is peaked at 0.1 mm and its value is reduced by two orders compared with
Figure 9.21a.
Figure 9.22a shows OR for the first structure under the applied electric field 10 kV/m, where it is
peaked at 1275 nm and its value is 7.8 × 10 –5 m/V. Increasing the field to 100 kV/cm reduces SONS
peak by more than two times and the peak wavelength is red shifted to 1323 nm.
For OR in the second structure, Figure 9.23a, SONS is reduced by two times compared with
Figure 9.22a, while the peak wavelength is 0.285 mm. Reducing the field to 100 kV/cm as in
336 Semiconductor Nanocrystals and Metal Nanoparticles

×10–8 F = 10 kV/cm
6
5 Re inhomo
Im
4
3
2
DFG (m/V)

1
0
–1
–2
–3
–4
7435 7440 7445 7450 7455 7460 7465 7470
(a) λDFG (nm)

×10–9 F = 100 kV/cm

4
Re inhomo
3 Im

2
DFG (m/V)

–1

–2

–3
5340 5345 5350 5355 5360 5365 5370 5375 5380
(b) λDFG (nm)

FIGURE 9.20 Second-order nonlinear susceptibility real (red lines) and imaginary (blue lines) parts by
difference-frequency generation (DFG) versus DFG wavelength for inhomogeneous broadening of the first
quantum dot structure at (a) 10 kV/cm and (b) 100 kV/cm electric field strengths.

Figure 9.23b reduces SONS by four time, while the peak wavelength is 64 µm. For both struc-
tures, SONS was reduced with increasing field in all SHG, SFG, DFG, and OR, which refers to a
higher asymmetry at a low applied field.
This reduction in SONS is due to the Stark shift of both conduction and valence subband ener-
gies under the application of the electric field. For the first (IB) structure, the difference between
the first two conduction subbands is 186.3 meV for 100 kV/cm, while this difference is reduced to
142 meV for 10 kV/cm applied field. Additionally, the momentum matrix element µ12 is reduced
when the field was increased from 10 to 100 kV/cm. These factors cause the reduction of SONS
when the field increases to 100 kV/cm. A nonmonotonic behavior of SHG in QDs with field is also
noticed by Shao et al. [21]. Chen et al. [44] show the reduction of OR in QDs with an increasing
electric field and their results are in the range of 10 –6 m/V, which is on the order of our results of IB
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 337

VB, F = 10 kV/cm
0.3
0.25 Re inhomo
Im
0.2
0.15
DFG (m/V)
0.1
0.05
0
–0.05
–0.1
–0.15
–0.2
0.4686 0.4687 0.4688 0.4689
(a) λDFG (mm)

×10–3 VB, F = 100 kV/cm

5
Re inhomo
4 Im
3

2
DFG (m/V)

–1

–2

–3
0.1002 0.1002 0.1002 0.1002 0.1002 0.1002 0.1002 0.1002 0.1002
(b) λDFG (mm)

FIGURE 9.21 Second-order nonlinear susceptibility real (red lines) and imaginary (blue lines) parts by
difference-frequency generation (DFG) versus DFG wavelength for inhomogeneous broadening of the second
quantum dot structure at (a) 10 kV/cm and (b) 100 kV/cm electric field strengths.

case (first structure) under 100 kV/cm. Results of Baskoutas et al. range between 10 –8 and 10 –3 m/V
depending on the type of confinement of QDs [18]. Our results here for peak OR range between
6.5 × 10 –11 m/V, for the second structure under 100 kV/cm (Figure 9.23b), and 7.5 × 10 –5 m/V, for the
first structure under 10 kV/cm (Figure 9.22a). This shows the difference between the confinements
of the two structures. The momentum matrix elements of the IB structure are higher than that of
the ISB structure.
The difference between the first two valence subbands at 10 kV/cm applied field was 4.2 meV,
while this difference increases to 36.7 meV when the field increases to 100 kV/cm, that is, the
subbands of the second structure shift by approximately one order of magnitude, while the first
­(subband) structure increases by 1.3 times with an increasing field. This is with the wavelength
338 Semiconductor Nanocrystals and Metal Nanoparticles

×10–5 F = 10 kV/cm
8
Re inhomo
6 Im

4
OR (m/V)
2

–2

–4

–6
1250 1255 1260 1265 1270 1275 1280 1285 1290 1295 1300
(a) λS (nm)

×10–7 F = 100 kV/cm

10

8 Re inhomo
Im
6

4
OR (m/V)

–2

–4

–6
1300 1305 1310 1315 1320 1325 1330 1335 1340 1345
(b) λS (nm)

FIGURE 9.22 Second-order nonlinear susceptibility real (red lines) and imaginary (blue lines) parts by opti-
cal rectification (OR) versus OR wavelength for inhomogeneous broadening of the first quantum dot structure
at (a) 10 kV/cm and (b) 100 kV/cm electric field strengths.

shift of the two structures. When the field increases from 10 to 100 kV/cm, the peak wavelength is
shifted by 1–2 µm (or less) for the IB structure, while it is shifted by one order of magnitude for the
ISB structure.
One of the main results that must be discussed in the study is the very long wavelengths obtained
in the second structures, millimeter wavelengths. Energy subbands of the structures can also explain
it. While the peak wavelength ranges between 1.2 and 10 µm for the first structure, it ranges between
0.064 and 1.363 mm for the second structure; the reason for this also lies in the values of energy
difference between energy states of the two structures. This difference is only few milli–electron
volts for the ISB (second) structure, while it is few electron volts for IB (first) structure. Thus, ISB
structures can be used in the application of millimeter wavelength.
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 339

×10–7 VB, F = 10 kV/cm

1.5

Re inhomo
1 Im

OR (m/V) 0.5

–0.5

–1

–1.5
0.24 0.26 0.28 0.3 0.32 0.34
(a) λS (mm)

×10–11 VB, F = 100 kV/cm

8
Re inhomo
6 Im

4
OR (m/V)

–2

–4

0.059 0.06 0.061 0.062 0.063 0.064 0.065 0.066

(b) λS (mm)

FIGURE 9.23 Second-order nonlinear susceptibility real (red lines) and imaginary (blue lines) parts by
optical rectification (OR) versus OR wavelength for inhomogeneous broadening of the second quantum dot
structure at (a) 10 kV/cm and (b) 100 kV/cm electric field strengths.

REFERENCES
1. X. Li, C. Zhang, Y. Tang, and B. Wang, Nonlinear optical rectification in asymmetric quantum dots with
an external static magnetic field, Physica E, 56, 2014, 130–133.
2. A. L. Allenic, Structural, electrical and optical properties of p-type ZnO epitaxial films, PhD thesis,
University of Michigan, Ann Arbor, MI, 2008.
3. T. Inoshita and H. Sakaki, Electron relaxation in a quantum dot-significance of Multiphonon Processes,
Physical Review B, 46, 1992, 7260–7263.
4. J. Shah, Ultrafast Spectroscopy of Semiconductors and Semiconductor Nanostructures, Springer-
Verlag, Berlin, Germany, 1999.
5. X. Wen, Ultrafast spectroscopy of semiconductor nanostructures, PhD thesis, University of Technology,
Melbourne, Victoria, Australia, 2007.
340 Semiconductor Nanocrystals and Metal Nanoparticles

6. M. Abdullah, A study on electronic and optical properties of ZnO-MgZnO quantum-dot semiconductor

optical amplifiers, MSc thesis, University of Baghdad, Baghdad, Iraq, 2009.
7. Amin H. Al-Khursan, Intensity noise characteristics in quantum-dot lasers: Four-level rate equations
analysis. Journal of Luminescence, 113, 2005, 129–136.
8. R. W. Boyd, Nonlinear Optics, 3rd edn., Academic Press, London, 2003.
9. M. Sabaeian and A. K. Nasab, Size-dependent intersubband optical properties of dome-shaped InAs/
GaAs quantum dots with wetting layer, Applied Optics, 51, 2012, 4176–4185.
10. A. Enshaeian, G. Rezaei, and S. F. Taghizadeh, Investigation of an external electric field effects on
diamagnetic susceptibility of an off-center hydrogenic donor in a spherical Gaussian quantum dot, in
Proceedings of the Fourth International Conference on Nanostructures (ICNS4), March 12–14, 2012,
Kish Island, Iran.
11. D. Ahn and S. L. Chuang, Exact calculations of quasibound states of an isolated quantum well with
uniform electric field: Quantum-well Stark resonance, Physical Review B, 34, 1986, 9034–9037.
12. J. Kim and S. L. Chuang, Theoretical and experimental study of optical gain, refractive index change and
linewidth enhancement factor of p-doped quantum dot lasers, IEEE Journal of Quantum Electronics,
42, 2006, 942–952.
13. D. Ahn and S. L. Chuang, Calculation of linear and nonlinear intersubband optical absorptions in
a quantum well model with an applied electric field, IEEE Journal Quantum Electronics, 23, 1987,
2196–2204.
14. Z. Li and X. Hong-Jing, Studies on the second-order nonlinear optical properties of parabolic and semi-
parabolic quantum wells with applied electric fields, Communications in Theoretical Physics, 41, 2004,
761–766.
15. E. C. Niclescu and M. Cristea, Dielectric mismatch effect on the electronic states in ZnS/CdSe core–
shell quantum dots under applied electric field, U.P.B. The Scientific Bulletin A, 75, 2013, 195–204.
16. R. Korlacki, R. F. Saraf, and S. Ducharme, Electrical control of photoluminescence wavelength from
semiconductor quantum dots in a ferroelectric polymer matrix, Applied Physics Letter, 99, 2011, 153112.
17. C. Dane, H. Akbas, S. Minez, and A. Guleroglu, Electric field effect in GaAs/Al/As spherical quantum
dot, Physica E, 41, 2008, 278.
18. W. Xie, The nonlinear optical rectification of a confined exciton in a quantum dot, Journal of
Luminescence, 131, 2011, 943–946.
19. S. Baskoutas, E. Paspalakis, and A. F. Terzis, Effects of excitons in nonlinear optical rectification in
semiparabolic quantum dots, Physical Review B, 74, 2006, 153306.
20. B. Vaseghi, G. Rezaei, V. Azizi, and S. M. Azami, Spin–orbit interaction effects on the optical rectifica-
tion of a cubic quantum dot, Physica E, 44, 2012, 1241–1243.
21. S. Shao, K. Guo, Z. Zhang, N. Li, and C. Peng, Studies on the second-harmonic generations in cubical
quantum dots with applied electric field, Physica B, 406, 2011, 393–396.
22. D. Ahn and S.-H. Park, Engineering Quantum Mechanics, John Wiley & Sons, Hoboken, NJ, 2011.
23. B. Al-Nashy, S. M. M. Amin, and A. H. Al-Khursan, Kerr dispersion in Y-configuration quantum dot
system, Journal of Optics, 16, 2014, 105205.
24. B. Al-Nashy, S. M. M. Amin, and A. H. Al-Khursan, Kerr effect in Y-configuration double quantum dot
System, Journal of the Optical Society of America B, 31, 2014, 1991–1996.
25. A. H. Al-Khursan, M. K. Al-Khakani, and K. H. Al-Mossawi, Third-order non-linear susceptibility
in a three-level QD system, Photonics and Nanostructures—Fundamentals and Applications 7, 2009,
153–160.
26. J. Kim, M. Laemmlin, C. Meuer, D. Bimberg, and G. Eisenstein, Static gain saturation model of quan-
tum-dot semiconductor optical amplifiers, IEEE Journal of Quantum Electronics, 44, 2008, 658–666.
27. H. Al-Husaini, A. H. Al-Khursan, and S. Y. Al-Dabagh, III-N QD lasers, Open Nanoscience Journal,
3, 2009, 1–11.
28. S. S. Mikhrin, A. R. Kovsh, I. L. Krestnikov, A. V. Kozhukhov, D. A. Livshits, N. N. Ledentsov, Y. M.
Shernyakov et al., High power temperature-insensitive 1.3 µm InAs/InGaAs/GaAs quantum dot lasers,
Semiconductor Science and Technology, 20, 2005, 340–342.
29. C. J. Chang-Hasnain, P. C. Ku, J. Kim, and S. L. Chuang, Variable optical buffer using slow light in
semiconductor nanostructures, Proceedings of IEEE, 91, 2003, 1884–1896.
30. Y. C. Chang, Nonlinear optical properties of semiconductor superlattices, Journal of Applied Physics,
58, 1985, 499–509.
Second-Order Nonlinear Susceptibility in Quantum Dot Structures 341

31. S. Suresh, A. Ramanand, D. Jayaraman, and P. Mani, Review on theoretical aspect of nonlinear optics,
Reviews on Advanced Materials Science, 30, 2012, 175–183.
32. N. Bloembergen, Nonlinear optics, in Fundamentals of Photonics, Bahaa E. A. Saleh and Malvin C.
Teich (Eds.), John Wiley & Sons, New York, 1991.
33. K. Semwal and S. C. Bhatt, Tuning of wavelengths for producing eye safe laser using second order
nonlinear processes, International Journal of Optics and Applications, 2, 2012, 20–28.
34. Y. R. Shen, Optical second harmonic generation at interfaces, Annual Review of Physical Chemistry 40,
1989, 327–350.
35. P. Fischer and A. C. Albrecht, On optical rectification in isotropic media, Laser Physics, 2002, 1177–1181.
36. M. V. Farsi, Investigation of heavily doped congruent lithium tantalite for single-pass second harmonic
generation, Master thesis work, Universitat Politècnica de Catalunya (UPC), Universitat Autònoma de
Barcelona (UAB), Barcelona, Spain, 2013.
37. T. Brunhes, P. Boucaud, S. Sauvage, A. Lemaıtre, J. M. Gerard, F. Glotin, R. Prazeres, and J.-M. Ortega
Infrared second-order optical susceptibility in InAs/GaAs self-assembled quantum dots, Physical
Review B, 61, 2000, 5562.
38. Y. She, X. Zheng, D. Wang, and W. Zhang, Controllable double tunneling induced transparency and
solitons formation in a quantum dot molecule, Optics Express, 21, 2013, 17392–17403.
39. P. Ginzburg, A. Krasavin, Y. Sonnefraud, A. Murphy, R. J. Pollard, S. A. Maier, and A. V. Zayats,
Nonlinearly coupled localized plasmon resonances: Resonant second-harmonic generation, Physical
Review B, 86, 2012, 085422.
40. T. Brunhes, P. Boucaud, S. Sauvage, F. Glotin, R. Prazeres, J.-M. Ortega, A. Lemaitre, and J.-M. Gerard,
Midinfrared second-harmonic generation in p-type InAs/GaAs self-assembled quantum dots, Applied
Physics Letters, 75, 1999, 835–837.
41. P. Gter (Ed.), Nonlinear Optical Effects and Materials, Springer, Berlin, Germany, 2000.
42. M. Imlau, Nonlinear Optics, Fachbereich Physic, University at Osnabruck, Osnabruck, Germany, 2005.
43. D. Bimberg, M. Grumman, and N. Ledentsov, Quantum Dot Heterostructures, John Wiley & Sons,
Chichester, U.K., 1999.
44. T. Chen, W. Xie, and S. Liang, The nonlinear optical rectification of an ellipsoidal quantum dot with
impurity in the presence of an electric field, Physica E, 44, 2012,786–790.
45. E. Psarakis, Simulation of performance of quantum well infrared photodetection, PhD thesis, University
of California, Berkeley, CA, 2005.
46. W. Xie, Nonlinear optical rectification of a hydrogenic impurity in a disc-like quantum dot, Physica B,
404, 2009, 4142–4145.
47. Y. R. Shen, The Principles of Nonlinear Optics, John Wiley & Sons, New York, 1984.
48. L. He and W. Xie, Effects of an electric field on the confined hydrogen impurity states in a spherical
parabolic quantum dot, Superlattices and Microstructures, 47, 2010, 266–273.
49. F. Xie, Resonant optical nonlinearities in cascade and coupled Quantum well structures, PhD thesis,
Texas A&M University, College Station, TX, 2008.
 10 Metallic Nanopastes for
Power Electronic Packaging
Denzel Bridges, Ruozhou Li, Zhiming Gao,
Zhiqiang Wang, Zhiyong Wang, Anming Hu,
Zhili Feng, Leon M. Tolbert, and Ning-Cheng Lee

CONTENTS
10.1 Introduction...........................................................................................................................344
10.2 Synthesis Methods................................................................................................................. 345
10.2.1 Chemical Reduction.................................................................................................. 345
10.2.1.1 Chemical Reduction at Ambient Pressure..................................................346
10.2.1.2 Hydrothermal/Solvothermal Synthesis.......................................................346
10.2.1.3 Sonochemical and Sonoelectrochemical Methods..................................... 347
10.2.2 Electrodeposition....................................................................................................... 347
10.2.3 Laser Ablation...........................................................................................................348
10.2.4 Comparison................................................................................................................ 348
10.3 Sintering Fundamentals......................................................................................................... 349
10.3.1 Mass Transport Mechanisms..................................................................................... 350
10.3.2 Thermal Effects......................................................................................................... 351
10.3.3 Effect of Pressure....................................................................................................... 351
10.3.4 Particle Size............................................................................................................... 352
10.3.5 Shape Effect............................................................................................................... 353
10.4 Sintering Methods................................................................................................................. 353
10.4.1 Thermal Sintering...................................................................................................... 353
10.4.2 Joule Heating............................................................................................................. 353
10.4.3 Ultrasonic Sealing..................................................................................................... 354
10.4.4 Photonic Sintering..................................................................................................... 354
10.4.5 Laser Sintering........................................................................................................... 354
10.4.6 Comparison................................................................................................................ 355
10.5 Potential Application in Power Electronics........................................................................... 356
10.5.1 Electric Vehicles........................................................................................................ 356
10.5.2 Renewable Energy..................................................................................................... 357
10.5.3 Smart Grid and Microgrids....................................................................................... 357
10.5.4 Military, Aerospace, and High-Power Semiconductor Device Packaging
Applications............................................................................................................... 359
10.6 Future Trends......................................................................................................................... 359
10.6.1 Green Synthesis......................................................................................................... 359
10.6.2 Core–Shell Nanoparticles.......................................................................................... 359
10.6.3 Nanoalloys and Nanoscale Metallic Compounds......................................................360
10.7 Summary............................................................................................................................... 361
Acknowledgments........................................................................................................................... 361
References....................................................................................................................................... 361

343
344 Semiconductor Nanocrystals and Metal Nanoparticles

10.1 INTRODUCTION
Power electronics are devices designed to handle the control and conversion of electrical power.
Such devices include four types of power converters, that is, DC–DC, DC–AC, AC–DC, and AC–AC
converters. Power electronics are particularly crucial in the development of renewable energy (RE)
technologies, smart grid technologies, and electric/hybrid electric vehicles (HEVs) because these
applications (Bose 2012; Hegazy et al. 2013) operate under high power (from tens to hundreds of kW)
and/or high temperature (>200°C) (Drobnik and Jain 2013). Power semiconductor modules, the basic
elements of a power electronic converter, are specifically designed to handle high power; however,
high-temperature operation is still a challenge. As such, the constituent materials of power semicon-
ductor modules (shown in Figure 10.1) must be able to sustain such harsh conditions especially for the
packaging materials that bond the power semiconductors to the substrate and bond power electronic
modules to the thermal management modules (Sheng and Colino 2004). At high power and high tem-
peratures, these packaging materials face accelerated fatigue conditions. The development of wide-
bandgap semiconductors such as silicon carbide (SiC) and gallium nitride (GaN) has increased the
potential and feasibility of high-temperature and high-power electronics (Khazaka et al. 2015). Power
electronic packaging components including substrates, bonding materials (e.g., soldering alloys), and
functional components have to satisfy the requirements of the enhanced bonding strength, high-
temperature operation, and resistance to harsh environments.
The bonding materials currently used for die attachment and interconnecting electronic compo-
nents are generally Pb-containing solders. However, due to the well-known long-term toxic nature
of Pb, many countries have placed limitations on the use of Pb-containing materials. A Pb-free
substitute is greatly desired. Some Sn-based alternatives have been developed, such as Sn–Bi and
Sn–Ag–Cu solder alloys. However, Sn–Bi and Sn–Ag–Cu solder alloys suffer from a melting tem-
perature below 225°C and significant creep deformation at service temperatures (Ohnuma et al.
2000; Shen et al. 2013). Sn–Bi solder alloys sometimes contain nanometallic fillers to improve
its creep resistance, hardness, and other mechanical properties. However, these soldering alloys
cannot survive at higher temperatures of power electronic applications such as the conditions in
exhaust sensing in an electric vehicle (EV), where the temperatures can be as high as 850°C
(Greenwell et al. 2011).
Potential alternatives to conventional bulk soldering materials (e.g., Pb–Sn, Sn–Bi, and Sn–Ag–Cu)
are metallic nanopastes. Several studies have been done to investigate the development and suitabil-
ity of metallic nanopastes as a replacement for conventional solders (Chen et al. 2014; Siow 2014).
In this chapter, we will discuss the recent progress of synthesis and sintering methods of metallic
nanopastes and their potential applications in the power electronic industry. We will also discuss

Encapsulation
Die attach material

Power Wirebonds
semiconductors
Substrate
attach material Substrate

Heat spreader

FIGURE 10.1 Basic packaging structure of a power semiconductor module. (From Sheng, W.W. and
Colino, R.P., Power Electronic Modules Design and Manufacture, CRC Press, Boca Raton, FL, 2004.)
Metallic Nanopastes for Power Electronic Packaging 345

future trends in the development of metallic nanopastes for power electronic packaging. This chap-
ter is organized as follows: first, we will discuss the synthesis methods for Ag-based and Cu-based
metallic nanopastes, then sintering fundamentals and techniques, applications in power electronics,
and finally future work in the field. We believe that this chapter is useful for advancing the relevant
techniques and facilitate innovative power electronic applications.

10.2 SYNTHESIS METHODS

There are two primary approaches to metal nanomaterial synthesis: top-down and bottom-up. In
a top-down approach, particles are formed by breaking a bulk metal into smaller particles and
dispersed in a proper medium. These methods include mechanical grinding, laser ablation, rapid
condensation of metal vapor, thermal heating/decomposition, or plasma excitation of bulk metal
or metal powders. In a bottom-up approach, nanomaterials are formed from an ionic precursor by
reaction with a reducing agent (Kamyshny and Magdassi 2014). As seen in Figure 10.2, the two
main mechanisms of particle growth in a bottom-up approach are (1) the reduction of the precursor
and (2) particle coarsening through consuming nearby smaller particles. Bottom-up approaches are
among the most popular methods because this approach is promising for large-scale, low-cost metal
nanomaterial synthesis. However, laser ablation of bulk metal in a liquid medium is also used to
fabricate nanoparticles (NPs) (Nguyen et al. 2012).

10.2.1 Chemical Reduction

Chemical reduction methods are one of the most common methods for synthesizing metallic nano-
materials. These processes typically involve dispersing an ionic precursor and reducing agent in a
medium and heating to an elevated temperature until the reaction is complete.
In addition to an ionic precursor and reducing agent, polymers and other organic compounds can
also be used as capping agents to stabilize materials such as Cu and Fe that are prone to oxidation
(Jeong et al. 2013; Kim et al. 2014a; Mott et al. 2007; Peng et al. 2006; Zhang et al. 2012). Some
capping agents, such as polyvinylpyrrolidone (PVP), can also be used to regulate the nanomaterial
growth in certain directions to form nanomaterials with controlled shapes, such as metallic nanow-
ires (MNWs) or nanoplates (NPLs) (Li et al. 2015). Some methods use an oxygen scavenger such as
triethanolamine to prevent highly reactive NPs from oxidation (Chang et al. 2005; Yang et al. 2013).
Another way to protect reactive nanomaterials from oxidation during synthesis is to fabricate them
in an oxygen-free atmosphere filled with an inert gas such as nitrogen or argon (Ha et al. 2011).
A popular subset of chemical reduction methods is called polyol methods. Polyol methods use
polyols (e.g., ethylene glycol, propylene glycol, and diethylene glycol) as both solvents and reducing

Heat +
reducing agent

Metal Nucleation
precursor in 2
solution e
Tim
1

Growth

FIGURE 10.2 General process of nanoparticle synthesis using a bottom-up approach. During a typical
bottom-up approach, nanoparticles typically grow after nucleation via two different pathways: (1) consuming
the precursor material and (2) larger particles consume smaller particles.
346 Semiconductor Nanocrystals and Metal Nanoparticles

agents. These methods allow reactions to be carried out at a higher temperature than aqueous solu-
tion (Li et al. 2013; Siow 2012; Yan et al. 2011). Polyols are mild reducing agents and are most
suitable for obtaining nanomaterials of electropositive metals such as silver, gold, platinum, and pal-
ladium (Dzido et al. 2015; Long et al. 2010, 2011, 2013; Mottaghi et al. 2014). However, polyol meth-
ods can be used for more electronegative metals such as copper (Lee et al. 2008; Mott et al. 2007).
There are further two primary types of chemical reduction methods: standard chemical reduction
(performed at ambient pressure) and solvothermal (ST)/hydrothermal (HT) reduction (performed
at high pressure).

10.2.1.1 Chemical Reduction at Ambient Pressure

The main advantage of chemical reduction methods at ambient pressure is their relative simplicity.
It is also easy to determine and control the experimental conditions in situ (e.g., temperature and
pressure). Other methods such as ST/HT methods are more difficult to determine the experimental
conditions. The chemical reduction process at ambient pressure is also scalable and easy to imple-
ment in an industrial setting (Carroll et al. 2011; Lee et al. 2008).
The main disadvantage of wet chemical methods is that the yield is generally low. Wet chemical
methods require a significant amount of solvent, most of which is decanted when the synthesis is
completed. This is of course quite inefficient. In addition, a conventional wet chemical method is
limited to a synthesis temperature lower than or approximately equal to the boiling temperature of
the solvent.

10.2.1.2 Hydrothermal/Solvothermal Synthesis

HT/ST synthesis is a chemical reduction method performed under high pressure (several MPa). HT
synthesis is performed using an aqueous solution. ST synthesis is the counterpart of HT synthesis
that uses a nonaqueous solvent. Usually, the solvent is organic. This method enables users to synthe-
size metallic nanomaterials at temperatures as high as the supercritical temperatures of the solvent
(Kawai-Nakamura et al. 2008; Li et al. 2015). The high pressure nature of the reaction allows the
precursor solution to be heated to supercritical temperatures.
HT and ST syntheses can be performed using either a batch reactor or a flow-cell reactor
(Miyakawa et al. 2014). Using the batch reactor procedure, it is difficult to control the heat-up
time of the precursor solution, which affects the rapid and homogeneous nucleation of the result-
ing nanomaterials (Aksomaityte et al. 2013). Flow-cell synthesis is a faster HT/ST synthesis tech-
nique because the precursor solution can be constantly fed into the apparatus such as the one in
Figure 10.3 for large-scale synthesis. The other reactants that can be added via a different line in the
flow cell can either be stabilizing agents or additional ionic precursors that enable core–shell NP
synthesis (Miyakawa et al. 2014).
Most ST/HT methods use conventional heating methods; however, they can also be incorporated
with microwave (MW)-assisted heating. MW heating is typically faster than traditional heating

Microwave heating

Precursor
solution

Output

Additional reagents

FIGURE 10.3 Basic diagram of the construction of a flow cell incorporated with microwave heating.
(From Kawai-Nakamura, A. et al., Mater. Lett., 62, 3471, 2008.)
Metallic Nanopastes for Power Electronic Packaging 347

methods. Traditional heating methods take minutes to hours to complete, while MW heating can
take mere seconds to complete (Ramulifho et al. 2012). To efficiently convert MW energy into
thermal energy, the solvent must have a high dielectric constant, so polyols and water are popular
choices (Kim et al. 2014b; Yi et al. 2013).

10.2.1.3 Sonochemical and Sonoelectrochemical Methods

Sonochemical synthesis of metal nanomaterials uses ultrasonic vibrations generated in a liquid
medium to generate the necessary energy and localized hot spots for chemical reduction of the pre-
cursor solution to occur (Godínez-García et al. 2012; Xu and Suslick 2010). Sonochemical methods
can be quite slow compared to some of the previously mentioned synthesis techniques like HT and
ST techniques, sometimes taking hours to complete.
Sonoelectrochemical methods combine sonochemical and electrodeposition methods by
using ultrasonic vibrations to enhance a conventional electrodeposition process (Zin et al. 2014).
Sonoelectrochemical methods can be faster than some electrodeposition techniques but have many
of the same weaknesses as electrodeposition techniques (Chang et al. 2012). Both types of meth-
ods allow for synthesis temperatures as low as room temperature (Darroudi et al. 2012; Haas et al.
2006), while the localized heating can be as high as 5000 K (Sakkas et al. 2012). The center of the
material is exposed to extreme conditions because the ultrasonic waves from all directions focus at
the center, as seen in Figure 10.4.

10.2.2 Electrodeposition
Electrodeposition is a process that uses an electric field between two electrodes to reduce a metal
precursor solution and deposit them on a substrate. A sample setup is shown in Figure 10.5. While
this technique is frequently used to form thin films (Zhou et al. 2012), it can also be used to fabri-
cate and/or align metal nanowires (NWs) using nanoporous templates (Cui et al. 2009) or micro-
electrodes (Wang et al. 2013a). Using a nanoporous membrane regulates the growth of NWs in one
direction without the use of a capping agent (Ohgai and Hashiguchi 2013). An example of such is
shown in Figure 10.5. Anodized alumina films (AAFs) are usually used for templating for NW
growth. This method is cheap and simple to synthesize aligned NWs. After growth, the AAF can be
easily removed by a strong base solution. However, this templating method is difficult to grow NWs
directly from an arbitrary substrate.

Intermediate
conditions on the
surface

Ultrasound from all

directions

Extreme conditions in
the core

FIGURE 10.4 A metal precursor material that is exposed to ultrasonic vibrations transported through a
liquid medium. Since the ultrasound is focused from all direction, all ultrasonic waves focus at the core of the
precursor material.
348 Semiconductor Nanocrystals and Metal Nanoparticles

Porous
template
V – +

Porous template
Reference
electrode

Cathode

Anode
Growing nanowires

FIGURE 10.5 Schematic of electrodeposition that employs a porous template for growing nanowires. Inset:
a schematic of a porous template cross section, inside which nanowires are being grown.

10.2.3 Laser Ablation
Laser ablation is the process of removing a material from a block or plate using laser irradiation
(Kamyshny and Magdassi 2014; Nguyen et al. 2012). Interaction with the laser causes the material
to either sublimate or become a plasma, and then the material is rapidly quenched and condenses
to form NPs in a liquid or gas medium (Mottaghi et al. 2014). As a plasma, the target material can
reach temperatures of several thousand Kelvin (Saito et al. 2014).
Most laser ablation methods in the last few years use laser pulses as opposed to continuous
laser irradiation because laser ablation processes with millisecond and longer laser pulses as an
input energy source are often described by a thermal evaporation process. For shorter laser pulses
(nanosecond or even shorter), the ablation processes is described by additional nonequilibrium
processes such as thermal and mechanical fragmentation of superheated melted material, spinodal
decomposition, explosive boiling, and spallation (Hu et al. 2007; Itina and Voloshko 2013). The
laser wavelength, fluence, and pulse duration are chosen based on the surface plasmon resonance,
desired NP size, and particle shape (Amendola and Meneghetti 2009; Becker et al. 1998; Muniz-
Miranda et al. 2011). For example, Messina et al. reported creating cubic Ag NPs when a 532 nm
laser radiation is used and octahedral Ag NPs when using 1064 nm laser radiation.
Laser ablation is a very versatile top-down method that can be done in an ambient atmosphere,
in a vacuum or even in a liquid (Hu et al. 2006, 2007, 2008; Ossi et al. 2011; Tarasenko et al. 2005).
Laser ablation can be used to produce NPs of almost any material. One of the most important
advantages of laser ablation is that it is a green synthesis (GS) method and generally no additional
reagents are required. The disadvantages of laser ablation are the small yield of NPs and the fact
that it is difficult to use laser beams serially. These issues are particularly important for industrial
applications.

10.2.4 Comparison
The synthesis methods discussed in this section differ in maturity, speed, and complexity. These
factors play a role in whether or not they are ready for commercial or industrial implementation.
In terms of complexity, chemical reduction at ambient pressure is the most widely used and well-
understood synthesis technique. Chemical reduction at ambient pressure is also easily scalable for
Metallic Nanopastes for Power Electronic Packaging 349

Ambient pressure
High HT/ST
Electrodeposition
SC/SEC
Laser ablation

Scalability
Low
High

Speed

High
Low Cost
Low

FIGURE 10.6 Comparison of the current synthesis methods.

industrial applications, but depending on the reactants used, this can introduce health hazards to
workers and environmental contamination. HT/ST methods have similar health hazards depending
on the technique, but when the method is MW-assisted and incorporated with a flow cell, it has the
potential to be one of the fastest synthesis techniques. Electrodeposition and sonochemical methods
can potentially be used in industrial applications; however, more research is needed to make these
methods suitable for large-scale production. Laser ablation is inherently one of the greenest and
fastest synthesis methods, but the small yield of NPs and difficulty in coupling laser energy to the
target due to attenuation of plumes limits laser power efficiency. Special skills required for operat-
ing the laser also limit its application. Figure 10.6 compares five different synthesis methods based
on the current cost, speed, and scalability. The cost is judged based on the current cost of equipment
and operation for the technique. Scalability is based on the current feasibility of large-scale synthe-
sis of nanomaterials using this technique. The speed is in terms of the relative NP yields per unit
time. To satisfy a low-cost, large-scale synthesis for industrial application, a synthesis technique
must be highly scalable and fast.

10.3 SINTERING FUNDAMENTALS

Sintering differs from other joining processes (i.e., solid-state bonding, brazing, soldering) in sev-
eral ways. For instance, brazing involves the soldering/bonding material being distributed between
the two bonding surfaces using capillary action after heating to a high temperature, usually above
450°C. Soldering involves melting the soldering/bonding material directly onto the bonding surface
at a temperature lower than 450°C. Solid-state bonding does not involve any melting whatsoever
but is usually dominated by pressure-assisted solid-state diffusion. Sintering is the process of join-
ing of two particles through a solid-state diffusion (solid-state sintering) or fusing smaller par-
ticles together usually by surface melting of the particles (liquid-phase sintering [LPS]). Figure 10.7
shows sintering that occurs in solid state and in liquid state, separately. Sintering generally occurs at
a lower temperature than soldering and brazing because the particles are not necessarily completely
melted (Zhou and Hu 2011).
350 Semiconductor Nanocrystals and Metal Nanoparticles

1. Surface diffusion 2. Volume and grain 3. Densification

(a) boundary diffusion

1. Solid state 2. Diffusion 3. Densification

diffusion through the liquid
(b)

FIGURE 10.7 Simplified comparison of (a) solid-state sintering and (b) liquid-phase sintering diffusion
pathways.

The thermodynamic driving force of sintering is the reduction of surface energy. The driving
force is given by

æ 1 1 ö
s = gç + ÷ (10.1)
R
è 1 R2 ø

where
γ is the surface energy of the material
R1 and R2 are the principal radii of curvature of the neighboring particles (Fang and Wang 2008)

10.3.1 Mass Transport Mechanisms

Surface diffusion is the initial mass transport mechanism at sintering temperatures because the acti-
vation energy of surface diffusion is smaller than that of other mass transport mechanisms such as
lattice diffusion. Surfaces with high curvature typically have a high defect population that includes
ledges, kinks, vacancies, and adatoms. The typical atomic movement that occurs during surface
diffusion comes in three steps: (1) breaking an atom away from existing bonds, usually at a
kink, (2) random motion across the surface, and (3) reattachment at an available surface site. When
it comes to sintering two particles in contact with each other, this available surface site is typically
on the surface of a neighboring particle.
As sintering progresses, surface diffusion becomes less important and lattice diffusion and grain
boundary diffusion increase in importance. Lattice diffusion and grain boundary diffusion are both
dominated by the motion of vacancies. As atoms diffuse through the volume of the particle to the
grain boundary between fused particles, mass is deposited in the neck region. Grain boundary dif-
fusion is particularly important for sintering densification of metals because grain boundaries form
in the sinter bond between misaligned crystals and it is through grain boundary diffusion that mass
is deposited in the neck region like the one in Figure 10.7a (German 1996). Vacancies are particu-
larly important for these types of diffusion because generally all types of diffusion occur at a faster
rate when the concentration of available lattice sites increases.
Metallic Nanopastes for Power Electronic Packaging 351

The mass transport mechanisms and driving force of LPS differ slightly from solid-state sinter-
ing. LPS involves dispersing solid grains in a liquid. The solubility causes the liquid to wet the solid,
providing capillary forces, and pulls the grains together (German et al. 2009). Atoms in the liquid
phase also display higher mobility, which facilitates the sintering process. Compared to solid-state
diffusion, LPS provides higher diffusion rates and is potentially more time-effective because it typi-
cally occurs at a higher temperature than a solid-state diffusion procedure (Alarifi et al. 2013;
Hu et al. 2013; Marzbanrad et al. 2013).

10.3.2 Thermal Effects
Sintering is usually thermally activated, so naturally increasing the temperature increases the
sintering rate. The equilibrium vacancy concentration in a material is proportional to kT, where k
is the Boltzmann constant and T is temperature. Therefore, when there is a temperature gradient
in a material, generally a vacancy concentration gradient also exists. According to Fick’s first
law (Equation 10.2), the diffusive flux (J) of a substance (e.g., atoms or vacancies) is proportional
to the concentration gradient (dΦ/dx). When vacancies diffuse in one direction, atoms diffuse
in the other direction until a steady state is reached. Therefore, when the vacancy concentration
increases, the diffusive flux of vacancies increases. Since the diffusive flux is the rate at which
a substance is transported per unit area per unit time, a higher diffusive flux in sintering means
that atoms are diffusing to the neck region and neighboring particles faster (i.e., the sintering
rate increases):

æ dF ö
J = Diffusivity ´ ç ÷ (10.2)
è dx ø

In addition, the sintering temperature of silver nanopastes has been shown to be positively corre-
lated with the shear strength of the sintered joint (Alarifi et al. 2011; Khazaka et al. 2014). Akada
et al. report that the bonding strength of a sintered joint using silver NPs increases as the sintering
temperature increases (Akada et al. 2008).
Chemically active materials such as Cu and Fe may oxidize at high sintering temperatures.
Oxides of these materials are typically weak, brittle, and nonconductive (Hai et al. 2013). Such
oxides are detrimental to power electronic packaging. Organic shells can be used to stabilize some
metal NPs and protect them from oxidation. However, organic shells are not an optimized solution
for high-temperature power electronic packing because organic shells typically decompose at high
temperatures. In addition, removal of the organic shell is critical to sintering because a thick organic
shell prevents direct contact between particle surfaces (Manikam et al. 2012). An example of this is
shown in Figure 10.8 (Wang et al. 2013b).

10.3.3 Effect of Pressure
Applying pressure during the sintering process can increase the strength of the sintered joint by
increasing the green density. The green density is the ratio between the actual volume of the sin-
tered nanomaterial and the external volume of the sintered structure. A higher green density means
that the particles are more close-packed. When the particles are more close-packed, there are more
neighboring particles, thereby allowing a lower sintering temperature (German 1996) and stronger
bonds between particles.
One drawback to increase the pressure (especially when joining areas ≥100 mm) is that increas-
ing the green density of a nanopaste can prevent solvents and other volatile components in the
nanopaste from escaping the center of the sintered region (Fu et al. 2014; Xiao et al. 2013). If the
volatile components cannot be removed, these also hinder the sintering process because many of
352 Semiconductor Nanocrystals and Metal Nanoparticles

Organic shell

No direct
contact

Metal NP

Heat

Heat
(a) (b)

FIGURE 10.8 Example of sintering with an organic shell around the particles. (a) Schematic of sintering
metal nanoparticles with a thick organic shell. (b) Schematic of sintering metal nanoparticles with a thin
organic shell. (From Wang, S. et al., Scripta Mater., 69, 789, 2013b.)

these components are organic. In addition, it is also a challenge to sinter some power electronic
semiconductor materials under high pressure since they are quite fragile.

10.3.4 Particle Size
It is well known that the melting point of a material decreases as the particle size decreases as seen
in Figure 10.9 (Yang et al. 2013). The same is true for sintering for which the temperature can be
calculated from the following equation from Fang et al.:

é 2 S (¥ ) 1 ù
Tis (r ) = 0.3Tm (¥) exp ê - m ú (10.3)
ë 3k (r / r0 ) - 1 û

where
Tis is the sintering temperature
r is the particle radius
Tm(∞) is the melting temperature of the bulk material
Sm(∞) is the bulk melting entropy
k is the Boltzmann constant
r0 = 3h, where h is the atomic diameter (Fang and Wang 2008)

The decreasing sintering temperature of NPs is also due to the increasing surface energy. This
higher surface energy increases the driving force, allowing large particles to grow by “consuming”
smaller particles (Asoro et al. 2014). Once sintered, the nanomaterial thermal, electrical, and physi-
cal properties more closely resemble the bulk material properties (Hu et al. 2010; Lee et al. 2008;
Yang et al. 2013).
Metallic Nanopastes for Power Electronic Packaging 353

Tm (K)
m.p. bulk
1300

1000

ses
decrea
point
ting
Mel

500

300
0 50 100 150 200 D (Å)

FIGURE 10.9 Relationship between melting temperature and particle diameter of gold nanoparticles.
(From Yang, C. et al., J. Mater. Chem. C, 1(207890), 4052, 2013.)

10.3.5 Shape Effect

Sintering of convex surfaces is more energetically favorable than concave surfaces because,
according to Fang et al., the curvature of a convex surface is taken to be positive, which yields a
positive driving force according to Equation 10.1. When surfaces of different shapes come into
contact, anisotropic sintering occurs. For example, Li et al. discuss how the morphology and
anisotropic sintering in an Ag NPL/NW composite paste can change as a function of temperature
and how it is different in a pure Ag NW paste. In this study, the sides of NWs mostly sinter to the
edge of NPLs instead of the face of the NPLs. The surface energy of the edge of an NPL is greater
than the face due to the smaller radius of curvature. This increases the driving force of the sinter-
ing process according to Equation 10.1 and is thus more energetically favorable than side-to-face
sintering (Li et al. 2015).

10.4 SINTERING METHODS

10.4.1 Thermal Sintering
Thermal sintering is the most popular and well-understood sintering method. Thermal sintering
involves heating a nanopaste to its sintering temperature using a hot plate or furnace. For nano-
materials, sintering can occur at as low as room temperature, but most of the time, this method is
conducted at a temperature greater than 100°C (Alarifi et al. 2011; Hu et al. 2010) and can be done
with and without additional pressure (Fu et al. 2014; Kähler et al. 2012). This method is also fea-
sible for removing solvents and organic material from the surface of nanomaterials (Kamyshny and
Magdassi 2014; Morisada et al. 2010). If the organic shell is removed, it is possible to sinter NPs at
room temperature (Marzbanrad et al. 2013; Peng et al. 2013, 2015).
This sintering process is conceptually simple and relies on thermally activated diffusion.
Thermal sintering is relatively cheap, but the process is relatively slow, and heating certain NPs to
high temperatures can cause unwanted oxides to form.

10.4.2 Joule Heating

Joule heating (or current-assisted sintering) is a unique sintering method that employs a sufficiently
high current to melt the surface of the nanomaterials and then sinter them together. The dissipated
354 Semiconductor Nanocrystals and Metal Nanoparticles

heat due to the internal resistance of the materials in combination with electromigration in the mate-
rials causes the materials to locally join.
This joule heating process may take only a few seconds to complete if a pulsed current is applied,
and it is practical in an industrial setting. Joule heating is an attractive option for forming conductive
interconnections due to its simplicity, cleanliness, and reliability. However, the data on the applica-
tion of joule heating in nanojoining applications are still limited because high current flow through
metal nanomaterials can cause severe structural and functional damage.
Peng et al. reported that this can be offset by including sacrificial NWs to prevent the functional
NWs from being damaged (Peng et al. 2009). However, this will further increase the material costs
of utilizing this technique. When using DC joule heating, Vafaei et al. report employing a voltage
ramp to find a “surge voltage” (the voltage at which there is a sharp increase in current) and impose
a current limit to avoid damage to Ag NWs (Vafaei et al. 2014). A recent study has also investigated
replacing DC with AC to enhance the sintering process for Ag nanopaste and found that the joints
can survive much longer than conventional pressure-assisted thermal sintering (Mei et al. 2014).

10.4.3 Ultrasonic Sealing
Ultrasonic sealing uses high-frequency vibrations beyond the audible range to melt or diffuse
materials together. This is a versatile method because the vibrational energy can be transmitted
through substrates to the joint area where the mechanical energy is converted to thermal energy
through internal friction. The thermal energy is enough to induce either melting or thermally acti-
vated diffusion to join nanomaterials.
Unfortunately, the quality of the sintered joint is strongly dependent on the design of the equip-
ment, design of the components being joined, the material properties, and the energy process
(Guo 2009). Right now, ultrasonic sealing for nanomaterials has been developed to join carbon
nanotubes to titanium electrodes, but there are not a lot of data on ultrasonic sealing of metal nano-
materials (Chen et al. 2006).

10.4.4 Photonic Sintering
By utilizing the plasmonic effect, millisecond light pulses can be concentrated between two adja-
cent nano-objects to produce “hot spots” without heating their surroundings using flash lamps and
camera flash sources (Li et al. 2014). The most effective wavelength of light for photonic sintering
is based on the plasmon absorption band. For instance, Ag NPs have a plasmon absorption band
of ~430 nm, so blue light is the most effective to initiate the sintering of Ag NPs (Hösel and Krebs
2012; Li et al. 2014). Once the particles are sintered, their plasmonic resonant frequency will dra-
matically shift to the red side (a long wavelength) (Li et al. 2014; Yang et al. 2011). In this sense, a
white light source with wide bands is preferred for sintering Ag NPs.
By not heating the surroundings, the risk of damaging some of the temperature sensitive com-
ponents in the power semiconductor module is eliminated. Photonic sintering begins with surface
atoms that are the ones that absorb the majority of the photons, and then heat transfer from the
surface leads to sintering of underlying nano-objects. Unfortunately, this leads to the formation of
a higher-density surface layer compared to the internal bulk. Regardless, photonic sintering is very
fast and cheap and has already been commercialized by NovaCentrix (2015). However, photonic
sintering is not possible for die-attach applications because the nanopaste will not effectively absorb
the photons (Kamyshny and Magdassi 2014).

10.4.5 Laser Sintering
Laser sintering is a type of photonic sintering that involves using laser irradiation to selectively
excite nanomaterials based on their surface plasmon absorption band, and the absorbed photons
Metallic Nanopastes for Power Electronic Packaging 355

melt the nanomaterial, allowing atoms to diffuse between neighboring particles and fuse together
(Cui et al. 2009). Most laser sintering methods involve multiple pulsed laser interactions and
depend on the length of exposure. Figure 10.10 shows the time scale for various secondary pro-
cesses in laser–matter interactions (Hu et al. 2011). A long-pulse-width laser or a continuous wave
laser may induce thermal sintering, while an ultrashort laser, that is, the laser pulse width shorter
than the thermal coupling time, can induce a nonthermal surface melting and thereby result in an
LPS (Hu et al. 2013).
By being able to selectively excite nano-objects, laser sintering holds an advantage over normal
photonic sintering by being able to form specific structures such as nanoscale porous surfaces (Cui
et al. 2009; Huang et al. 2012; Xie et al. 2013). Studies that investigate femtosecond laser welding
focus on nonthermal laser–matter interactions. This technique is ideal for fabricating microelectro-
mechanical systems and nanoelectromechanical systems because it allows for precise welding of
electrical interconnections.
Laser sintering can also be incorporated in an industrial setting as well. However, it is difficult
to use for die-attach applications because the nanomaterials cannot effectively absorb the photons
(Kähler et al. 2012; Khazaka et al. 2015; Tan et al. 2015). In addition, femtosecond laser sinter-
ing is not fully understood. Three phase changes may occur during femtosecond laser interaction:
melting, evaporation, and resolidification. If the laser intensity is too high, the laser can cause
fragmentation of the nanomaterial, or the particle is melted completely (Hu et al. 2011).

10.4.6 Comparison
Of the methods discussed in this section, thermal sintering is one of the most well understood, easily
implementable, and conceptually simple and feasible for certain applications. For example, thermal
sintering is not ideal when depositing metallic nanopastes on temperature sensitive components.
Likewise, photonic and laser sintering is challenging for die-attach and other applications where the
nanopaste cannot effectively absorb incoming photons. As previously mentioned, photonic sintering
and ultrasonic welding have made their way into industrial applications, but laser sintering has yet
to be widely used in an industrial setting for metallic nanopastes. Joule heating has potential, but
there are not much data on joule heating of metallic nanopastes. In short, thermal sintering is the
most mature sintering technique, but there is room in industrial applications for other methods to

10–14 Electronic dephasing

Fermi level
10–13 Electronic thermalization

Electronic cooling Nonthermal

10–12 Phonon relaxation
Tsurface < Tcore
Thermal
10–11 Thermal
melting/diffusion

10–10 Ablation

Time (s)

FIGURE 10.10 Time scale for various secondary processes in laser–matter interactions. (From Hu, A. et al.,
Open Surf. Sci. J., 3, 42, 2011.)
356 Semiconductor Nanocrystals and Metal Nanoparticles

Thermal sintering
Joule heating
Ultrasonic welding
Photonic sintering
Laser sintering
High

Versatility
High

Cost

Low
High Low
Speed Low

FIGURE 10.11 Comparison of current sintering methods.

make an impact. Figure 10.11 compares sintering techniques based on cost, speed, and ­versatility.
Versatility refers to a particular technique’s capability of being used on any part of the power elec-
tronic. Currently, thermal sintering, joule heating, and ultrasonic welding are the most balanced
techniques in terms of cost, speed, and versatility. Photonic and laser sintering need improvements
in terms of cost and versatility before they can be widely used in industrial applications.

10.5 POTENTIAL APPLICATION IN POWER ELECTRONICS

Power electronics have already been heavily integrated into many fields from telecommunications
and computing power supplies to transportation to smart grid technologies (Figure 10.12). Metallic
nanopastes would mostly be used to make conductive interconnections and die-attach materials
(Ahmed et al. 2013; Hegazy et al. 2013).

10.5.1 Electric Vehicles
All EVs have yet to fully catch on as an attractive option for personal transportation, but more
countries have started to adopt stricter carbon emission standards and have begun to turn to EVs for
public transportation (Ahmed et al. 2013; Nakamura and Mufson 2014), and HEVs are becoming
increasingly popular for personal transportation. However, the increasing popularity of EVs and
HEVs comes with technical challenges because of new failure modes including inverter failure,
sensor failure, and motor overheating (Zhuoping et al. 2014).
Current EV and HEV designs incorporate a number of power electronics including DC–DC
converters, DC–AC converters, battery charging systems, and the electric motor (Drobnik and Jain
2013; Dusmez and Khaligh 2013; Madawala and Thrimawithana 2011; Onar et al. 2013). Metallic
nanopastes would most likely serve as die-attach materials for high-temperature and high-power-
density SiC power modules that currently have unproven high-temperature reliability in EV/HEV
environments (Rajashekara 2013). Using metallic nanopastes should improve the efficiency and
reliability of high-density power electronics that suffer a loss in efficiency when exposed to high
electrical power (Whitaker et al. 2014).
Many of these systems operate at a temperature at or above 200°C (Xu et al. 2013); some may
even operate at temperatures as high as 850°C (Greenwell et al. 2011). Electrically conductive metal
Metallic Nanopastes for Power Electronic Packaging 357

Smart
grid

Power Renewable
Aerospace
electronics energy

Electric
vehicles

FIGURE 10.12 Applications of power electronics. (From Zhu, H. and Liu, X., J. Chem. Pharmaceut. Res.,
6(7), 859, 2014; Khazaka, R. et al., IEEE Trans. Power Electron., 30(5), 2456, 2015; Gould, K. et al., Thermal
management of silicon carbide power module for military hybrid vehicles, in 2014 Semiconductor Thermal
Measurement and Management Symposium (SEMI-THERM), IEEE, 2014, pp. 105–108; Manojkumar, M. et al.,
Power electronics interface for hybrid renewable energy system—A survey, in 2014 International Conference on
Green Computing Communication and Electrical Engineering (ICGCCEE), Las Vegas, Nevada, 2014, pp. 1–9.)

nanopastes are also typically thermally conductive, which can transfer heat away from the ­functional
parts of the semiconductor-based components and boost efficiency (Pinkos and Guo 2013).

10.5.2 Renewable Energy
Improving the efficiency of solar cells, wind turbines, and other RE technologies is an important
research topic for meeting future energy needs. Aside from losses due to the energy conversion and
harvesting (e.g., light energy to electrical energy for photovoltaic cells), RE technologies suffer from
DC–AC conversion losses in the power electronic converters when connected to the electrical grid.
Photovoltaic cells also suffer from ohmic losses from electrical contacts.
Hybrid energy systems (HESs) and power electronic–based energy storage systems (PEBESSs)
are shown in Figures 10.13 and 10.14, respectively, and are being increasingly used and investi-
gated for the electric grid. Both systems incorporate DC–DC converters, DC–AC converters, and
other power electronics into their design. HESs and PEBESSs will require more advanced power
­electronics to be successful (Manojkumar et al. 2014; Sirisukprasert 2014).
Ag pastes and inks are currently used in commercial photovoltaic cells and have excellent
electrical properties; however, the Ag pastes and inks alone account for one quarter of the total
material cost for crystalline silicon solar cells (Hai et al. 2013; Silva et al. 2011; Yang et al. 2011).
Various materials have been considered as a replacement for this expensive material such as
Cu–Ag core–shell particles because having a copper core greatly reduces the overall material
cost (Bridges et al. 2015).

10.5.3 Smart Grid and Microgrids

A popular topic in electrical engineering is the development of the smart grid, an electric power grid
that more efficiently and reliably distributes electrical power to a region based on supply, demand,
location, and time. Also of interest to companies are microgrids, which can provide generation
locally and continue to provide power to key loads even when there is an interruption from the main
358 Semiconductor Nanocrystals and Metal Nanoparticles

DC–DC Local DC
converters loads

DC–DC
converters

DC–AC Power
Energy converters distribution
storage
system

FIGURE 10.13 Schematic of a hybrid energy system and its components. (From Manojkumar, M. et al.,
Power electronics interface for hybrid renewable energy system—A survey, in 2014 International Conference on
Green Computing Communication and Electrical Engineering (ICGCCEE), Las Vegas, Nevada, 2014, pp. 1–9.)

Energy storage DC–DC Conventional

device energy sources
C
DC–A
C AC–D
DC–DC

Electrical
power grid
AC–
AC
AC
Intermittent DC–
energy sources Power
consumption

FIGURE 10.14 Block diagram of a the flow of DC–AC power between different components of a smart grid
system. (From Kobayashi, T., Power electronics technology in smart grid projects—Applications and experi-
ences, in 2014 International Power Electronics Conference, Hiroshima, Japan, 2014, pp. 1868–1873.)

utility grid. Key components of smart grids and microgrids are intermittent RE sources such as
photovoltaic and wind power, robust large-scale energy storage, regional/microenergy management
system, demand-side energy management, and advanced metering infrastructure (Kobayashi 2014).
Due to the interactive nature of the smart grid and microgrids, power electronics are necessary for
applying varying electrical loads based on demand and location because the spatial and temporal
distribution of power in a region is unbalanced (Yue et al. 2014; Zhu and Liu 2014).
The development of the smart grid is also important for high penetration levels of intermit-
tent energy sources such as photovoltaic and wind power. The peak energy production times of
intermittent energy sources do not necessarily correlate with peak energy consumption times. In
tandem with a large-scale energy storage system, the smart grid would use data from the demand-
side energy management system to determine when the energy generated from intermittent energy
sources needs to be stored or immediately supplied to the grid. The power electronics again is the
key components to implement smart electrical management.
Metallic Nanopastes for Power Electronic Packaging 359

10.5.4 Military, Aerospace, and High-Power Semiconductor

Device Packaging Applications
More advanced power electronics are also of great interest to the military, aerospace, high-power
semiconductor laser (HPSL), and high-power light-emitting diodes (HPLEDs). For the military,
power electronics have very different requirements. The power electronic packaging must have high
resistance to pyrotechnical shocks and vibrations, thermal cycling, and corrosive atmospheres such
as salt spray and fog. HEVs are also of great interest to the military (Gould et al. 2014).
In the aerospace industry, the world is moving toward “more electric” and “all electric” aircraft
and hence replacing hydraulic systems with power electronic systems. Doing so will significantly
improve aircraft reliability and reduce complexity, weight, and maintenance (Pathak 2012).
HPSLs are important for various scientific and industrial instruments such as laser material pro-
cessing, laser projection display, and fluorescence analysis. Novel packaging materials are sought
particularly for thermal management of HPSLs. The obstacles are that current die-attach materials
suffer from a relatively low melting temperature and thermal conductivity and high thermal stresses
due to a coefficient of thermal expansion mismatch (Higurashi et al. 2015; Okumura et al. 2014).
HPLEDs are desired for various lighting applications such as automobile headlights and other
high-brightness lighting applications. Like HPSLs, novel packaging materials are desired mainly
for thermal management of HPLEDs. It is well known that the luminosity of light-emitting diodes
decreases as temperature increases. Without heat dissipation in the semiconductor, the temperature
HPLED will continue to increase and the luminosity will deteriorate (Qu et al. 2014; Thenmozhi
et al. 2014). Therefore, a thermally conductive interconnection between the high-power semicon-
ductor module and the heat spreader in Figure 10.1 is highly desirable. Li et al. report a high lumi-
nous flux when Ag nanopaste is used as the die-attach material compared to Sn–Ag–Cu solder and
silver epoxy (Li et al. 2010).

10.6 FUTURE TRENDS

Metallic nanopastes have advanced greatly in the last two decades, but there is still much room for
improvement before commercialization and various applications. We will now discuss the advance-
ments being made in metallic nanopaste technologies.

10.6.1 Green Synthesis

As previously mentioned, several methods for synthesizing metallic nanomaterials use toxic reduc-
ing agents. As a result, GS procedures are a growing interest among researchers and the industry.
GS refers to synthesis procedures that use nontoxic and environmentally friendly reactants. Most
GS procedures use aqueous solutions and call for nontoxic organic reducing agents; ascorbic acid
(a form of vitamin C) is the most popular one (Usman et al. 2013; Valodkar et al. 2011; Zhao et al.
2011). Other reducing agents such as starch, gelatin (Darroudi et al. 2012), and tartaric acid (Chen
et al. 2013) have been used as well.

10.6.2 Core–Shell Nanoparticles

Core–shell NPs are unique nanostructured materials that are formed by first forming core NP
from one metal, then coating with another metal by a transmetalation reaction such as the one in
Figure 10.15 for Cu–Ag core–shell NPs (Grouchko et al. 2009) or electrodeposition (Lin-You et al.
2002). The core material is often a cheaper or chemically less stable metal and the shell is usually
more chemically stable (Ger et al. 2014; Xia et al. 2010). In the case of Cu–Ag core–shell NPs,
Cu and Ag have comparable thermal and electrical properties; however, Cu is prone to oxidation
and pure Ag pastes are quite expensive. Therefore, a Cu core can be synthesized and then encased
360 Semiconductor Nanocrystals and Metal Nanoparticles

S+ C+

Core Core Shell Nucleation Full shell

metal metal precursor on core formation
precursor formed added surface

FIGURE 10.15 Simplified diagram of formation of core–shell nanoparticles by galvanic metal displacement
reaction. S+ refers to a shell material ion and C+ refers to a core material ion.

in a Ag shell to protect the Cu core from oxidation (Bridges et al. 2015; Kirubha and Palanisamy
2014; Miyakawa et al. 2014; Peng et al. 2012).

10.6.3 Nanoalloys and Nanoscale Metallic Compounds

Some nanomaterials are mixed to lower the overall material cost by mixing a more expensive mate-
rial such as Ag with a less expensive, but compatible material. As previously stated, the result of the
sintered nanopaste is comparable to that of a bulk alloy. Nanomaterial mixture displays properties
that are a combination of the constituent nanomaterials (Morisada et al. 2010).
Nanomaterials can also be mixed for interesting mechanical and oxidation properties. Li et al.
report mixing Ag NWs and NPLs (as seen in Figure 10.16) to bond Cu wire. The result was a joint

(a)

(b) (c)

FIGURE 10.16 The morphology of (a) Ag nanowires, (b) Ag nanoplates, and (c) Ag nanowire and nanoplate
composite paste.
Metallic Nanopastes for Power Electronic Packaging 361

that had electrical conductivity resembling a pure NW paste and increased bonding strength from
the added NPLs (Li et al. 2015). Mixing Ag and Cu NPs results in an oxidation-resistant Cu–Ag
alloy according to several studies (Tan et al. 2015; Valodkar et al. 2011; Yan et al. 2012).

10.7 SUMMARY
Packaging with nanomaterials (i.e., nanojoining) and power electronic technology are progressing
very rapidly. Both technologies will be critical in the way that we generate, transmit, store, and use
energy. A number of synthesis and sintering techniques have emerged over the years, providing
different metal nanopastes with different shapes and morphologies. Each nanopaste and sinter-
ing technique holds different advantages and disadvantages that are best suited depending on the
intended production scale and application. The unique properties of metallic nanopastes show great
potential to be a viable alternative to Pb-based solders. It is crucial that the cost of synthesizing and
sintering metallic nanopastes is greatly reduced before they become widely used in power electronic
applications. When improvements are made to reduce the cost and further advance the mechanical
properties of metallic nanopastes, it will surely propel power electronics into the future.

ACKNOWLEDGMENTS
The authors appreciate the research initiative funding provided by the University of Tennessee as a
new hire package to AH.

REFERENCES
Ahmed, S., H. Kim, T. Prohaska, T. Ronkainen, and R. Burgos. 2013. Stability study of electric vehicle power
electronics based power system. In IEEE International Electric Vehicle Conference, Santa Clara, CA,
October 23–25, 2013.
Akada, Y., H. Tatsumi, T. Yamaguchi, A. Hirose, T. Morita, and E. Ide. 2008. Interfacial bonding mecha-
nism using silver metallo-organic nanoparticles to bulk metals and observation of sintering behavior.
Materials Transactions 49(7): 1537–1545.
Aksomaityte, G., M. Poliakoff, and E. Lester. 2013. The production and formulation of silver nanoparticles
using continuous hydrothermal synthesis. Chemical Engineering Science 85(January): 2–10.
Alarifi, H., A. Hu, M. Yavuz, and Y. Norman Zhou. 2011. Silver nanoparticle paste for low-temperature
bonding of copper. Journal of Electronic Materials 40(6): 1394–1402.
Alarifi, H.A., M. Atiş, C. Özdoğan, A. Hu, M. Yavuz, and Y. Zhou. 2013. Determination of complete melting
and surface premelting points of silver nanoparticles by molecular dynamics simulation. The Journal of
Physical Chemistry C 117(23): 12289–12298.
Amendola, V. and M. Meneghetti. 2009. Laser ablation synthesis in solution and size manipulation of noble
metal nanoparticles. Physical Chemistry Chemical Physics 11: 3805–3821.
Asoro, M.A., D. Kovar, and P.J. Ferreira. 2014. Effect of surface carbon coating on sintering of silver nanopar-
ticles: In situ TEM observations. Chemical Communications 50: 4835–4838.
Becker, M.F., J.R. Brock, H. Cai, D.E. Henneke, J.W. Keto, J. Lee, W.T. Nichols, and H.D. Glicksman. 1998.
Metal nanoparticles generated by laser ablation. Nanostructured Materials 10(5): 853–863.
Bose, B. 2012. Global energy scenario and impact of power electronics in 21st century. IEEE Transactions on
Industrial Electronics 60(7): 1.
Bridges, D., R.-Z. Li, M. Floarea, and A. Hu. 2015. Cu–Ag core–shell nanopastes obtained by microwave-
assisted solvothermal method and galvanic metal displacement for Cu to Cu bonding. Unpublished work.
Carroll, K.J., J.U. Reveles, M.D. Shultz, S.N. Khanna, and E.E. Carpenter. 2011. Preparation of elemental
Cu and Ni nanoparticles by the polyol method: An experimental and theoretical approach. The Journal
of Physical Chemistry C 115(6): 2656–2664.
Chang, C.C., K.H. Yang, Y.C. Liu, and C.C. Yu. 2012. Surface-enhanced Raman scattering-active silver nano-
structures with two domains. Analytica Chimica Acta 709: 91–97.
Chang, Y., M.L. Lye, and H.C. Zeng. 2005. Large-scale synthesis of high-quality ultralong copper nanowires.
Langmuir 21(20): 3746–3748.
362 Semiconductor Nanocrystals and Metal Nanoparticles

Chen, C., L. Yan, E.S.-W. Kong, and Y. Zhang. 2006. Ultrasonic nanowelding of carbon nanotubes to metal
electrodes. Nanotechnology 17: 2192–2197.
Chen, G., L. Yu, Y.-H. Mei, X. Li, X. Chen, and G.-Q. Lu. 2014. Reliability comparison between SAC305 joint
and sintered nanosilver joint at high temperatures for power electronic packaging. Journal of Materials
Processing Technology 214(9): 1900–1908.
Chen, K.T., D. Ray, Y.H. Peng, and Y.C. Hsu. 2013. Preparation of Cu–Ag core–shell particles with their anti-
oxidation and antibacterial properties. Current Applied Physics 13(7): 1496–1501.
Cui, Q., F. Gao, S. Mukherjee, and Z. Gu. 2009. Joining and interconnect formation of nanowires and carbon
nanotubes for nanoelectronics and nanosystems. Small 5(11): 1246–1257.
Darroudi, M., A.K. Zak, M.R. Muhamad, N.M. Huang, and M. Hakimi. 2012. Green synthesis of colloidal
silver nanoparticles by sonochemical method. Materials Letters 66: 117–120.
Drobnik, J. and P. Jain. 2013. Electric and hybrid vehicle power electronics efficiency, testing and reliability.
In International Electric Vehicle Symposium and Exhibition, Barcelona, Spain, November 17–20, 2013,
pp. 1–12.
Dusmez, S. and A. Khaligh. 2013. A compact and integrated multifunctional power electronic interface for
plug-in electric vehicles. IEEE Transactions on Power Electronics 28(12): 5690–5701.
Dzido, G., P. Markowski, A. Małachowska-Jutsz, K. Prusik, and A.B. Jarzębski. 2015. Rapid continu-
ous microwave-assisted synthesis of silver nanoparticles to achieve very high productivity and full
yield: From mechanistic study to optimal fabrication strategy. Journal of Nanoparticle Research: An
Interdisciplinary Forum for Nanoscale Science and Technology 17(1): 27.
Fang, Z. and H. Wang. 2008. Densification and grain growth during sintering of nanosized particles.
International Materials Reviews 53(6): 389–400.
Fu, S., Y. Mei, G.-Q. Lu, X. Li, G. Chen, and X. Chen. 2014. Pressureless sintering of nanosilver paste at low tem-
perature to join large area (≥100 mm2) power chips for electronic packaging. Materials Letters 128: 42–45.
Ger, T.-R., H.-T. Huang, C.-Y. Huang, W.-C. Liu, J.-Y. Lai, B.-T. Liu, J.-Y. Chen, C.-W. Hong, P.-J. Chen, and
M.-F. Lai. 2014. Comparing the magnetic property of shell thickness controlled of Ag–Ni core–shell
nanoparticles. Journal of Applied Physics 115: 17B528.
German, R.M. 1996. Sintering Theory and Practice. New York: John Wiley & Sons, Inc.
German, R.M., P. Suri, and S.J. Park. 2009. Review: Liquid phase sintering. Journal of Materials Science 44:
1–39.
Godínez-García, A., J.F. Pérez-Robles, H.V. Martínez-Tejada, and O. Solorza-Feria. 2012. Characterization
and electrocatalytic properties of sonochemical synthesized PdAg nanoparticles. Materials Chemistry
and Physics 134(2–3): 1013–1019.
Gould, K., S.Q. Cai, C. Neft, and A. Bhunia. 2014. Thermal management of silicon carbide power module for
military hybrid vehicles. In 2014, 30th annual Semiconductor Thermal Measurement and Management
Symposium (SEMI-THERM), IEEE, San Jose, CA, pp. 105–108, March 9–13, 2014.
Greenwell, R.L., B.M. McCue, L. Zuo, M.A. Huque, L.M. Tolbert, B.J. Blalock, and S.K. Islam. 2011.
SOI-based integrated circuits for high-temperature power electronics applications. In Conference
Proceedings—26th annual IEEE Applied Power Electronics Conference and Exposition (APEC), Fort
Worth, TX, pp. 836–843, March 6–11, 2011.
Grouchko, M., A. Kamyshny, and S. Magdassi. 2009. Formation of air-stable copper–silver core–shell
nanoparticles for inkjet printing. Journal of Materials Chemistry 19: 3057–3062.
Guo, K.W. 2009. A review of micro/nano welding and its future developments. Recent Patents on
Nanotechnology 3: 53–60.
Ha, D.-H., L.M. Moreau, C.R. Bealing, H. Zhang, R.G. Hennig, and R.D. Robinson. 2011. The structural evo-
lution and diffusion during the chemical transformation from cobalt to cobalt phosphide nanoparticles.
Journal of Materials Chemistry 21(31): 11498–11510.
Haas, I., S. Shanmugam, and A. Gedanken. 2006. Pulsed sonoelectrochemical synthesis of size-controlled
copper nanoparticles stabilized by poly(N-vinylpyrrolidone). The Journal of Physical Chemistry B
110(34): 16947–16952.
Hai, H.T., H. Takamura, and J. Koike. 2013. Oxidation behavior of Cu–Ag core–shell particles for solar cell
applications. Journal of Alloys and Compounds 564: 71–77.
Hegazy, O., R. Barrero, J. Van Mierlo, P. Lataire, N. Omar, and T. Coosemans. 2013. An advanced power
electronics interface for electric vehicles applications. IEEE Transactions on Power Electronics 28(12):
5508–5521.
Higurashi, E., K. Okumura, K. Nakasuji, and T. Suga. 2015. Surface activated bonding of GaAs and SiC
wafers at room temperature for improved heat dissipation in high-power semiconductor lasers. Japanese
Journal of Applied Physics 54(March): 030207.
Metallic Nanopastes for Power Electronic Packaging 363

Hösel, M. and F.C. Krebs. 2012. Large-scale roll-to-roll photonic sintering of flexo printed silver nanoparticle
electrodes. Journal of Materials Chemistry 22(31): 15683.
Hu, A., I. Alkhesho, W.W. Duley, and H. Zhou. 2006. Cryogenic graphitization of submicrometer grains
embedded in nanostructured tetrahedral amorphous carbon films. Journal of Applied Physics 100:
084319.
Hu, A., G.L. Deng, S. Courvoisier, O. Reshef, C.C. Evans, E. Mazur, and Y. Zhou. 2013. Femtosecond laser
induced surface melting and nanojoining for plasmonic circuits. In Proceedings of the SPIE Plasmonics:
Metallic Nanostructures and Their Optical Properties XI, San Diego, CA, Vol. 8809, August 25, 2013.
Hu, A., J.Y. Guo, H. Alarifi, G. Patane, Y. Zhou, G. Compagnini, and C.X. Xu. 2010. Low temperature sinter-
ing of Ag nanoparticles for flexible electronics packaging. Applied Physics Letters 97: 153117.
Hu, A., M. Rybachuk, I. Alkhesho, Q.-B. Lu, and W. Duley. 2008. Nanostructure and sp1/sp2 clustering in
tetrahedral amorphous carbon thin films grown by femtosecond laser deposition. Journal of Laser
Applications 20(1): 37.
Hu, A., M. Rybachuk, Q.B. Lu, and W.W. Duley. 2007. Direct synthesis of sp-bonded carbon chains on graph-
ite surface by femtosecond laser irradiation. Applied Physics Letters 91: 131906.
Hu, A., Y. Zhou, and W.W. Duley. 2011. Femtosecond laser-induced nanowelding: Fundamentals and applica-
tions. The Open Surface Science Journal 3: 42–49.
Huang, C.-J., H.-T. Cheng, C.-N. Kuo, C.-W. Cheng, W.-L. Tsai, and Y.-H. Yang. 2012. Fabrication of porous ti
surface by femtosecond laser sintering of Ti powder. In 2012 Conference on Lasers and Elecro-Optics,
San Jose, CA, Vol. 1, pp. 1–2, May 6–11, 2012.
Itina, T.E. and A. Voloshko. 2013. Nanoparticle formation by laser ablation in air and by spark discharges at
atmospheric pressure. Applied Physics B: Lasers and Optics 113: 473–478.
Jeong, S., S.H. Lee, Y. Jo, S.S. Lee, Y.-H. Seo, B.W. Ahn, G. Kim et al. 2013. Air-stable, surface-oxide free Cu
nanoparticles for highly conductive Cu ink and their application to printed graphene transistors. Journal
of Materials Chemistry C 1: 2704–2710.
Kähler, J., N. Heuck, A. Wagner, A. Stranz, E. Peiner, and A. Waag. 2012. Sintering of copper particles
for die attach. IEEE Transactions on Components, Packaging and Manufacturing Technology 2(10):
1587–1591.
Kamyshny, A. and S. Magdassi. 2014. Conductive nanomaterials for printed electronics. Small 10(17):
3515–3535.
Kawai-Nakamura, A., T. Sato, K. Sue, S. Tanaka, K. Saitoh, K. Aida, and T. Hiaki. 2008. Rapid and continu-
ous hydrothermal synthesis of metal and metal oxide nanoparticles with a microtube-reactor at 523 K
and 30 MPa. Materials Letters 62: 3471–3473.
Khazaka, R., L. Mendizabal, and D. Henry. 2014. Review on joint shear strength of nano-silver paste and its
long-term high temperature reliability. Journal of Electronic Materials 43(7): 2459–2466.
Khazaka, R., L. Mendizabal, D. Henry, and R. Hanna. 2015. Survey of high-temperature reliability of power
electronics packaging components. IEEE Transactions on Power Electronics 30(5): 2456–2464.
Kim, J.S., H.J. Jeon, H.W. Yoo, Y.K. Baek, K.H. Kim, D.W. Kim, and H.T. Jung. 2014a. Generation of mono-
disperse, shape-controlled single and hybrid core–shell nanoparticles via a simple one-step process.
Advanced Functional Materials 24: 841–847.
Kim, M., W.-S. Son, K.H. Ahn, D.S. Kim, H.-S. Lee, and Y.-W. Lee. 2014b. Hydrothermal synthesis of
metal nanoparticles using glycerol as a reducing agent. The Journal of Supercritical Fluids 90(June):
53–59.
Kirubha, E. and P.K. Palanisamy. 2014. Green synthesis, characterization of Au–Ag core–shell nanopar-
ticles using gripe water and their applications in nonlinear optics and surface enhanced Raman studies.
Advances in Natural Sciences: Nanoscience and Nanotechnology 5: 045006.
Kobayashi, T. 2014. Power electronics technology in smart grid projects—Applications and experiences. In
2014 International Power Electronics Conference, Hiroshima, Japan, pp. 1868–1873, May 18–21, 2014.
Lee, Y., J.-R. Choi, K.J. Lee, N.E. Stott, and D. Kim. 2008. Large-scale synthesis of copper nanoparticles
by chemically controlled reduction for applications of inkjet-printed electronics. Nanotechnology 19:
415604.
Li, R.-Z., A. Hu, T. Zhang, and K.D. Oakes. 2014. Direct writing on paper of foldable capacitive touch pads
with silver nanowire inks. Applied Materials and Interfaces 6: 21721–21729.
Li, R.-Z., T. Zhang, A. Hu, and D. Bridges. 2015. Ag nanowire and nanoplate composite paste for low tempera-
ture bonding. Materials Transactions 56: 984–987.
Li, W., M. Chen, J. Wei, W. Li, and C. You. 2013. Synthesis and characterization of air-stable Cu nanopar-
ticles for conductive pattern drawing directly on paper substrates. Journal of Nanoparticle Research
15: 1949.
364 Semiconductor Nanocrystals and Metal Nanoparticles

Li, X., X. Chen, and G.Q. Lu. 2010. Effect of die-attach material on performance and reliability of high-power
light-emitting diode modules. In Proceedings of the 60th Electronic Components and Technology
Conference, Las Vegas, Nevada, pp. 1344–1346, June 1–4, 2010.
Lin-You, C., D. Peng, and L. Zhong-Fan. 2002. Preparation of Au(core)–Cu (shell) nanoparticles assembly by
electrodeposition. Acta Physico Chimica Sinica 18(12): 1062–1067.
Long, N.V., T. Asaka, T. Matsubara, and M. Nogami. 2011. Shape-controlled synthesis of Pt–Pd core–shell
nanoparticles exhibiting polyhedral morphologies by modified polyol method. Acta Materialia 59(7):
2901–2907.
Long, N.V., N.D. Chien, T. Hayakawa, H. Hirata, G. Lakshminarayana, and M. Nogami. 2010. The synthe-
sis and characterization of platinum nanoparticles: A method of controlling the size and morphology.
Nanotechnology 21(3): 035605.
Long, N.V., M. Ohtaki, M. Yuasa, S. Yoshida, T. Kuragaki, C.M. Thi, and M. Nogami. 2013. Synthesis and
self-assembly of gold nanoparticles by chemically modified polyol methods under experimental control.
Journal of Nanomaterials 2013: 793125.
Madawala, U.K. and D.J. Thrimawithana. 2011. A bidirectional inductive power interface for electric vehicles
in V2G systems. IEEE Transactions on Industrial Electronics 58(10): 4789–4796.
Manikam, V.R., K.A. Razak, and K.Y. Cheong. 2012. Sintering of silver–aluminum nanopaste with vary-
ing aluminum weight percent for use as a high-temperature die-attach material. IEEE Transactions on
Components, Packaging and Manufacturing Technology 2(12): 1940–1948.
Manojkumar, M., K. Porkumaran, and C. Kathirvel. 2014. Power electronics interface for hybrid renewable
energy system—A survey. In 2014 International Conference on Green Computing Communication and
Electrical Engineering (ICGCCEE), Coimbatore, India, pp. 1–9, March 6–8, 2014.
Marzbanrad, E., A. Hu, B. Zhao, and Y. Zhou. 2013. Room temperature nanojoining of triangular and hexago-
nal silver nanodisks. The Journal of Physical Chemistry C 117: 16665–16676.
Mei, Y.H., Y. Cao, G. Chen, X. Li, G.Q. Lu, and X. Chen. 2014. Characterization and reliability of sintered
nanosilver joints by a rapid current-assisted method for power electronics packaging. IEEE Transactions
on Device and Materials Reliability 14(1): 262–267.
Miyakawa, M., N. Hiyoshi, M. Nishioka, H. Koda, K. Sato, A. Miyazawa, and T.M. Suzuki. 2014. Continuous
syntheses of Pd@Pt and Cu@Ag core–shell nanoparticles using microwave-assisted core particle forma-
tion coupled with galvanic metal displacement. Nanoscale 6: 8720–8725.
Morisada, Y., T. Nagaoka, M. Fukusumi, Y. Kashiwagi, M. Yamamoto, and M. Nakamoto. 2010. A low-
temperature bonding process using mixed Cu–Ag nanoparticles. Journal of Electronic Materials 39(8):
1283–1288.
Mott, D., J. Galkowski, L. Wang, J. Luo, and C.J. Zhong. 2007. Synthesis of size-controlled and shaped copper
nanoparticles. Langmuir 23(10): 5740–5745.
Mottaghi, N., M. Ranjbar, H. Farrokhpour, M. Khoshouei, A. Khoshouei, P. Kameli, H. Salamati, M. Tabrizchi,
and M. Jalilian-Nosrati. 2014. Ag/Pd core–shell nanoparticles by a successive method: Pulsed laser
ablation of Ag in water and reduction reaction of PdCl2. Applied Surface Science 292: 892–897.
Muniz-Miranda, M., C. Gellini, and E. Giorgetti. 2011. Surface-enhanced Raman scattering from copper
nanoparticles obtained by laser ablation. The Journal of Physical Chemistry C 115: 5021–5027.
Nakamura, D. and S. Mufson. 2014. China, U.S. agree to limit greenhouse gases. The Washington Post,
November 12, 2014, Business section. http://www.washingtonpost.com/business/economy/china-us-
agree-to-limit-greenhouse-gases/2014/​11/11/9c768504-69e6-11e4-9fb4-a622dae742a2_story.html.
Nguyen, T.B., T.K.T. Vu, C.D. Nguyen, T.D. Nguyen, T.A. Nguyen, and T.H. Trinh. 2012. Preparation of metal
nanoparticles for surface enhanced Raman scattering by laser ablation method. Advances in Natural
Sciences: Nanoscience and Nanotechnology 3(2): 025016.
NovaCentrix. 2015. PulseForge tools for printed electronics. http://www.novacentrix.com/products/pulseforge.
Accessed on May 3, 2015.
Ohgai, T. and K. Hashiguchi. 2013. Functional nanowires array electrodeposited into nano-porous membrane
thin films. Journal of Physics: Conference Series 417: 012047.
Ohnuma, I., M. Miyashita, K. Anzai, X. J. Liu, H. Ohtani, R. Kainuma, and K. Ishida. 2000. Phase equilibria
and the related properties of Sn–Ag–Cu based Pb-free solder alloys. Journal of Electronic Materials
29: 1137–1144.
Okumura, K., E. Higurashi, T. Suga, and K. Hagiwara. 2014. Low-temperature GaAs/SiC wafer bonding with
Au thin film for high-power semiconductor lasers. In 2014 International Conference on Electronics
Packaging (ICEP), IEEE, Toyama, Japan, pp. 716–719, April 23–25, 2014.
Onar, O.C., J. Kobayashi, and A. Khaligh. 2013. A fully directional universal power electronic interface for
EV, HEV, and PHEV applications. IEEE Transactions on Power Electronics 28(12): 5489–5498.
Metallic Nanopastes for Power Electronic Packaging 365

Ossi, P.M., F. Neri, N. Santo, and S. Trusso. 2011. Noble metal nanoparticles produced by nanosecond laser
ablation. Applied Physics A: Materials Science and Processing 104: 829–837.
Pathak, A.D. 2012. High reliability power electronics. In Proceedings of the PCIM ASIA 2012 International
Conference and Exhibition for Power Electronics, Intelligent Motion, Renewable Energy and Energy
Management, Shanghai, China, pp. 29–39.
Peng, P., A. Hu, A.P. Gerlich, Y. Liu, and Y.N. Zhou. 2015. Self-generated local heating induced nanojoining
for room temperature pressureless flexible electronic packaging. Scientific Reports 5: 1–8.
Peng, P., L. Liu, A.P. Gerlich, A. Hu, and Y.N. Zhou. 2013. Self-oriented nanojoining of silver nanowires via
surface selective activation. Particle & Particle Systems Characterization 30(5): 420–426.
Peng, S., C. Wang, J. Xie, and S. Sun. 2006. Synthesis and stabilization of monodisperse Fe nanoparticles.
Journal of the American Chemical Society 128: 10676–10677.
Peng, Y., T. Cullis, and B. Inkson. 2009. Bottom-up nanoconstruction by the welding of individual metallic
nanoobjects using nanoscale solder. Nano Letters 9(1): 91–96.
Peng, Y.-H., C.-H. Yang, K.-T. Chen, S.R. Popuri, C.-H. Lee, and B.-S. Tang. 2012. Study on synthesis of ultra-
fine Cu–Ag core–shell powders with high electrical conductivity. Applied Surface Science 263: 38–44.
Pinkos, A.F. and Y. Guo. 2013. Automotive design challenges for wide-band-gap devices used in high tem-
perature capable, scalable power vehicle electronics. In 2013 IEEE Energytech, Cleveland, OH, May
21–23, 2013.
Qu, H.-M., X.-H. Yang, Q. Zheng, X.-T. Wang, and Q. Chen. 2014. Thermal management technology of high-
power light-emitting diodes for automotive headlights. IEICE Electronics Express 11(23): 1–11.
Rajashekara, K. 2013. Present status and future trends in electric vehicle propulsion technologies. IEEE
Journal of Emerging and Selected Topics in Power Electronics 1(1): 3–10.
Ramulifho, T., K.I. Ozoemena, R.M. Modibedi, C.J. Jafta, and M.K. Mathe. 2012. Fast microwave-assisted
solvothermal synthesis of metal nanoparticles (Pd, Ni, Sn) supported on sulfonated MWCNTs: Pd-based
bimetallic catalysts for ethanol oxidation in alkaline medium. Electrochimica Acta 59: 310–320.
Saito, G., Y. Nakasugi, and T. Akiyama. 2014. Excitation temperature of a solution plasma during nanopar-
ticles formation. Journal of Applied Physics 116: 083301.
Sakkas, P., O. Schneider, S. Martens, P. Thanou, G. Sourkouni, and C. Argirusis. 2012. Fundamental studies of
sonoelectrochemical nanomaterials preparation. Journal of Applied Electrochemistry 42(9): 763–777.
Shen, L., Z.Y. Tan, and Z. Chen. 2013. Nanoindentation study on the creep resistance of SnBi solder alloy with
reactive nano-metallic fillers. Materials Science and Engineering A 561: 232–238.
Sheng, W.W. and R.P. Colino. 2004. Power Electronic Modules Design and Manufacture. Boca Raton, FL:
CRC Press.
Silva, J.A., M. Gauthier, C. Boulord, C. Oliver, A. Kaminski, B. Semmache, and M. Lemiti. 2011. Improving
front contacts of N-type solar cells. Energy Procedia 8: 625–634.
Siow, K.S. 2012. Mechanical properties of nano-silver joints as die attach materials. Journal of Alloys and
Compounds 514: 6–19.
Siow, K.S. 2014. Are sintered silver joints ready for use as interconnect material in microelectronic packag-
ing? Journal of Electronic Materials 43(4): 947–961.
Sirisukprasert, S. 2014. Power electronics-based energy storages: A key component for smart grid technology.
In Proceedings of the International Electrical Engineering Congress 2014, Chonburi, Thailand, March
19–24, 2014.
Tan, K.S., Y.H. Wong, and K.Y. Cheong. 2015. Thermal characteristic of sintered Ag–Cu nanopaste for high-
temperature die-attach application. International Journal of Thermal Sciences 87: 169–177.
Tarasenko, N.V., A.V. Butsen, and E.A. Nevar. 2005. Laser-induced modification of metal nanoparticles
formed by laser ablation technique in liquids. Applied Surface Science 247(1–4): 418–422.
Thenmozhi, R., C. Sharmeela, P. Natarajan, and R. Velraj. 2014. Experimental investigation on performance
improvement in high power LED lamps. In 2014 International Conference on Green Computing
Communication and Electrical Engineering (ICGCCEE), Coimbatore, India, pp. 3–7.
Usman, M.S., M.E. El Zowalaty, K. Shameli, N. Zainuddin, M. Salama, and N.A. Ibrahim. 2013. Synthesis,
characterization, and antimicrobial properties of copper nanoparticles. International Journal of
Nanomedicine 8: 4467–4479.
Vafaei, A., A. Hu, and I.A. Goldthorpe. 2014. Joining of individual silver nanowires via electrical current.
Nano-Micro Letters 6(4): 293–300.
Valodkar, M., S. Modi, A. Pal, and S. Thakore. 2011. Synthesis and anti-bacterial activity of Cu, Ag and Cu–Ag
alloy nanoparticles: A green approach. Materials Research Bulletin 46(3): 384–389.
Wang, B., Y. Xu, K.-L. Yung, W. Chen, and C.-L. Kang. 2013a. Aligning and soldering pure-copper nanowires
for nanodevice fabrication. Journal of Microelectromechanical Systems 22(3): 519–526.
366 Semiconductor Nanocrystals and Metal Nanoparticles

Wang, S., M. Li, H. Ji, and C. Wang. 2013b. Rapid pressureless low-temperature sintering of Ag nanoparticles
for high-power density electronic packaging. Scripta Materialia 69: 789–792.
Whitaker, B., Z. Cole, B. Passmore, T. Mcnutt, M.N. Ericson, S. Shane, C.L. Britton et al. 2014. High-
temperature SiC power module with integrated SiC gate drivers for future high-density power elec-
tronics applications. In IEEE Workshop on Wide Bandgap Power Devices and Applications (WiPDA),
Knoxville, TN. IEEE, New York, pp. 36–40.
Xia, L., X. Hu, X. Kang, H. Zhao, M. Sun, and X. Cihen. 2010. A one-step facile synthesis of Ag–Ni core–shell
nanoparticles in water-in-oil microemulsions. Colloids and Surfaces A: Physiochemical Engineering
Aspects 367: 96–101.
Xiao, K., S. Luo, K. Ngo, and G.-Q. Lu. 2013. Low-temperature sintering of a nanosilver paste for attach-
ing large-area power chips. In 15th International Symposium and Exhibition on Advanced Packaging
Materials (APM 2013), Irvine, CA, Vol. 5, pp. 192–202.
Xie, F., X. He, X. Lu, S. Cao, and X. Qu. 2013. Preparation and properties of porous Ti–10Mo alloy by selec-
tive laser sintering. Materials Science and Engineering C: Materials for Biological Applications 33(3):
1085–1090.
Xu, H. and K.S. Suslick. 2010. Sonochemical synthesis of highly fluorescent Ag nanoclusters. ACS Nano 4(6):
3209–3214.
Xu, Z., M. Li, F. Wang, and Z. Liang. 2013. Investigation of Si IGBT operation at 200°C for traction applica-
tions. IEEE Transactions on Power Electronics 28: 2604–2615.
Yan, J., G. Zou, A. Hu, and Y.N. Zhou. 2011. Preparation of PVP coated Cu NPs and the application for
­low-temperature bonding. Journal of Materials Chemistry 21(1): 15981–15986.
Yan, J., G. Zou, A. Wu, J. Ren, A. Hu, and Y.N. Zhou. 2012. Polymer-protected Cu–Ag mixed NPs for
low-temperature bonding application. Journal of Electronic Materials 41(7): 1886–1892.
Yang, C., C.P. Wong, and M.M.F. Yuen. 2013. Printed electrically conductive composites: Conductive filler
designs and surface engineering. Journal of Materials Chemistry C 1(207890): 4052–4069.
Yang, Y., S. Seyedmohammadi, U. Kumar, D. Gnizak, E. Graddy, and A. Shaikh. 2011. Screen printable silver
paste for silicon solar cells with high sheet resistance emitters. Energy Procedia 8: 607–613.
Yi, Z., X. Xu, K. Zhang, X. Tan, X. Li, J. Luo, X. Ye et al. 2013. Green, one-step and template-free synthesis
of silver spongelike networks via a solvothermal method. Materials Chemistry and Physics 139(2–3):
794–801.
Yue, L., S. He, and W. Chen. 2014. On the application of advanced power electronics technology in smart grid.
Journal of Chemical and Pharmaceutical Research 6(7): 793–797.
Zhang, J., N. Ma, F. Tang, Q. Cui, F. He, and L. Li. 2012. pH- and glucose-responsive core–shell hybrid
nanoparticles with controllable metal-enhanced fluorescence effects. ACS Applied Materials and
Interfaces 4: 1747–1751.
Zhao, J., D. Zhang, and J. Zhao. 2011. Fabrication of CuAg coreshell bimetallic superfine powders by eco-
friendly reagents and structures characterization. Journal of Solid State Chemistry 184(9): 2339–2344.
Zhou, D., M. Zhou, M. Zhu, X. Yang, and M. Yue. 2012. Electrodeposition and magnetic properties of FeCo
alloy films. Journal of Applied Physics 111: 07A319.
Zhou, Y. and A. Hu. 2011. From microjoining to nanojoining. The Open Surface Science Journal 3: 32–41.
Zhu, H. and X. Liu. 2014. On the application of advanced power electronics technology in smart grid. Journal
of Chemical and Pharmaceutical Research 6(7): 859–863.
Zhuoping, Y., J. Wu, and L. Xiong. 2014. Research of stability control of distributed drive electric vehicles
under motor failure modes. In 2014 IEEE Conference and Expo Transportation Electrification Asia-
Pacific (ITEC Asia-Pacific), IEEE, Beijing, China, pp. 1–5, August 31–September 3, 2014.
Zin, V., K. Brunelli, and M. Dabalà. 2014. Characterization of Cu–Ni alloy electrodeposition and synthesis
of nanoparticles by pulsed sonoelectrochemistry. Materials Chemistry and Physics 144(3): 272–279.
 11 Applications of Metal
Nanoparticles and
Nanostructures Fabricated
Using Ultrafast Laser
Ablation in Liquids
S. Venugopal Rao, S. Hamad, and G. Krishna Podagatlapalli

CONTENTS
11.1 Introduction.......................................................................................................................... 368
11.2 Fundamentals of Ultrafast Laser Ablation........................................................................... 370
11.2.1 Ablation of Metals in Ambient Air........................................................................... 370
11.2.2 Demerits of Laser Ablation in Ambient Air............................................................. 371
11.3 Ablation of Metals in Liquids.............................................................................................. 371
11.3.1 Cavitation Bubble Dynamics.................................................................................... 372
11.4 Advantages of Pulsed Laser Ablation in Liquids................................................................. 372
11.5 Parameters Influencing the Ablation Mechanism................................................................ 373
11.5.1 Input Wavelength...................................................................................................... 373
11.5.2 Pulse Duration.......................................................................................................... 373
11.5.3 Energy per Pulse, Spot Size, and Fluence................................................................ 374
11.5.4 Number of Pulses...................................................................................................... 375
11.6 Results and Discussion......................................................................................................... 376
11.6.1 Experimental Details................................................................................................ 376
11.6.2 Results from Ag Target Ablation in Water............................................................... 377
11.6.2.1 Ag Targets Ablated with Different Fluences and Scanning Conditions..... 377
11.6.2.2 Ag Targets Ablated with Different Angles of Incidence............................ 381
11.6.2.3 Ag Targets Ablated with ps/fs Bessel Beams............................................. 388
11.6.3 Cu Targets Ablated with Different Energies and Scanning Conditions................... 389
11.6.4 Cu Targets Ablated in Chloroform........................................................................... 397
11.6.5 Cu Targets Ablated with Corroles and Surface-Enhanced Fluorescence/SERS
Studies.......................................................................................................................400
11.7 Nonlinear Optical Studies of Cu, Ag, and Al Colloids........................................................403
11.7.1 NLO Properties of Cu Colloids................................................................................403
11.7.2 NLO Properties of Ag Colloids................................................................................406
11.7.3 NLO Properties of Al Colloids.................................................................................408
11.8 Antibacterial Studies of Cu, Ag Colloids............................................................................409
11.9 Conclusions.......................................................................................................................... 410
11.10 Future Scope........................................................................................................................ 411
Acknowledgments........................................................................................................................... 412
References....................................................................................................................................... 412

367
368 Semiconductor Nanocrystals and Metal Nanoparticles

11.1 INTRODUCTION
The last few decades were dedicated for the investigation of materials that enabled scientists to under-
stand their physical, chemical, and structural properties and utilize them in applications such as plas-
monics, photonics, and biomedicine to name a few. The interrogation of materials in the nanoscale
became an imperative and inevitable task since the dimensions modify the basic behavior of the
material through alteration of boundary conditions (Kelly et al. 2003, Link et al. 2003). The materials
whose dimensions are in the 1–100 nm range are termed as nanomaterials (Kelly et al. 2003, Link
et al. 2003). The utilization of nanoscale plasmonic materials (e.g., Au, Ag, and Cu) is well known
from the ancient times of history. Nanomaterials are essential building blocks of a wide range of sci-
entific applications such as photo-induced thermal therapy, biochemical sensors, surface-enhanced
Raman spectroscopy, carriers of drug delivery, nanophotonic devices, biosensing, in vivo and in vitro
diagnostics, solar cells, optoelectronic devices, antibacterial agents, cancer treatment, catalysis, imag-
ing, sensing, biology and medicine, and sensors (Arruebo et al. 2007, Guo and Wang 2011, Jain et al.
2008, Murphy et al. 2005, Pradeep and Anshup 2009). Among the various top-down and bottom-up
lithographic fabrication techniques (Makaraov 2013, Narayanan and Sakthivel 2010, Pelton et al.
2008, Sakamoto et al. 2009, Sau and Rogach 2010, Xia et al. 1999), laser ablation in liquid (LAL)
media is an efficient and “green” technique to produce both nanoparticles (NPs) and nanostructures
(NSs) in a single experiment. This technique is deemed to be “green” since there are no chemicals
(surfactants) involved during or after ablation. Ablation is the removal of fragments when a material
is hit by an intense laser pulse. Fragments account for atoms, ions, molecular clusters, etc. The inter-
action of ultrashort laser pulses with materials conjures strong interest since their pulse duration is
much less than the time required for several relaxation processes (transfer of energy between electron
and lattice systems, heat diffusion, etc.) (Perez and Lewis 2003, Perez et al. 2008). These short pulse
durations lead to the deposition of higher energy densities (higher peak intensities) when they focus
on a bulk target surface. Consequently, higher pressures and temperatures will be attained at the
point of the impact leading to surface modification of the material and effortless fragmentation. The
transient state of the material after the irradiation of a focused pulsed laser can be approximated by a
phase transition (Radziemski and Creamers 1989) since the absorption of laser energy by electron gas
promotes the target material to a state of higher temperature by leaving the remaining target (lattice)
at the initial temperatures (Radziemski and Creamers 1989). Restoration of the system to a state of
equilibrium depends on the parameters of laser pulses used for ablation and the intrinsic nature of
the target. Ultrafast laser ablation (ULA) minimizes the random corrugation of the target since the
pulse duration is extremely short and the heat-affected zone (HAZ) provided is negligible compared
to the HAZ provided by the longer laser pulses. When the fs pulses interact with materials, the energy
is absorbed by inverse Bremsstrahlung followed by ionization processes typically in the ~10–100
fs time scales (Perez and Lewis 2003, Perez et al. 2008). The excited electrons collide with each
other or with bound electrons, resulting in the production of high conductivity. This phenomenon is
described as electron–electron scattering (or coulomb explosion), which occurs in ~1 ps time scale
(Perez and Lewis 2003, Perez et al. 2008). After that, electrons transfer their energy to lattice through
electron–lattice coupling and the corresponding characteristic time is ~10 ps (Perez and Lewis 2003,
Perez et al. 2008). Thermal diffusion occurs inside the lattice via phonon–phonon scattering, which
is typically after 10 ps and up to ~1 nanosecond (ns) time. The electron–electron, electron–phonon,
and phonon–phonon relaxations can be explained by the two-temperature model (Harilal et al. 2014,
Von Der Linde and Sokolowski-Tinten 2000). Further stages involved are plasma formation and plume
quenching (~100 ns to 1 μs), formation and collapse of cavitation bubble (1–200 μs), and nanomaterials
(NMs) formation in liquid media (~1 ms) (De Bonis et al. 2013, Harilal et al. 2014). Figure 11.1 illus-
trates the various mechanisms and time scales recorded for a typical ablation process.
LAL with ns pulses has been studied for more than two decades now including the pioneering
work by Yang and coworkers (Liang et al. 2014, Liu et al. 2010, Wang et al. 2005, Yang 2007, 2011).
There are several review articles outlining the basic physics, fabrication aspects, characterization,
Applications of Metal Nanoparticles and Nanostructures 369

Inverse Bremsstrahlung
Excitation of
carriers
Ionization process

Electron–electron relaxation
Coulomb
explosion and
lattice heating
Electron–phonon relaxation

Thermal phenomena Phonon–phonon relaxation

Plasma formation, confinement

Plasma and and shock wave propagation
cavitation
Plasma quenching
bubble
Cavitation
bubble
Growth and agglomeration
Nanomaterials
of nanomaterials

10–15 10–12 10–9 10–6 10–3

FIGURE 11.1 A typical timing diagram of fs laser ablation for various stages.

and applications of the generated NPs (Besner and Meunier 2010, Ishikawa et al. 2006, Sasaki et al.
2006, Shafeev 2008, Stratakis 2012, Yan and Chrisey 2012, Zeng et al. 2012) and NSs (Barmina et al.
2010, Tsuji et al. 2010). Several pioneering works have been published by a number of groups all over
the world, and some significant contributions emanated from Meunier’s group (Besner et al. 2005,
2007, 2008, 2009, 2010a,b, Boulais et al. 2013, Dallaire et al. 2012, Hatef and Meunier 2015, Kabashin
and Meunier 2003, 2006, Kenth et al. 2011, Lachaine et al. 2014, Lorazo et al. 2003, Papgiannouli
et al. 2015, Rioux et al. 2009, Robitaille et al. 2013, Sylvestre et al. 2004a,b, 2005, Zakharko et al.
2011), Barcikowski’s group (Asahi et al. 2015, Barcikowski and Mafune 2011, Barcikowski et al.
2007, 2010, De Giacomo et al. 2013, Gökce et al. 2015, Ibrahimkutty et al. 2012, Menéndez-Manjón
et al. 2011, 2013, Petersen et al. 2009, Rehbock et al. 2014, Sajti et al. 2010, Stratakis et al. 2009b,
Taylor et al. 2011), Shafeev’s group (Dolgaev et al. 2002, Kazakevich et al. 2004, 2006, Kuzmin
et al. 2010, Menéndez-Manjón et al. 2010, Serkov et al. 2015, Simakin et al. 2001, 2004, Stratakis
et al. 2009a,c, Tiedemann et al. 2014, Truong et al. 2007a,b, Zavedeev et al. 2006), Amendola and
Meneghetti’s group (Amendola et al. 2006, 2007, 2009a,b, 2011a,b, 2012, 2013, 2014, Bertorelle
et al. 2014), Itina’s group that includes the theoretical modeling of ablation phenomena (Delfour and
Itina 2015, Dell’Aglio et al. 2015, Itina 2011, Noël et al. 2007, Povarnitsyn et al. 2007), Brandi’s
group (Bagga et al. 2013, Intartaglia et al. 2011, 2012a–c, 2013, 2014), others (Compagnini et al.
2003, Hu et al. 2011, Lalayan 2005, Medivil et al. 2013, O’Malley et al. 2014, Salminen et al. 2012,
Tilaki et al. 2007, Usui et al. 2005, Jiang et al. 2011), and our group (Hamad et al. 2012, 2013,
2014a,b,c, 2015, Podagatlapalli et al. 2012, 2013, 2014a,b,c, 2015a,b, Saikiran et al. 2014, Syed et al.
2012, Vendamani et al. 2015, Venugopal Rao et al. 2014).
Our group efforts, over the last few years, have mainly been focused on ultrafast (ps or fs) abla-
tion in a variety of liquids and investigating different plasmonic metals such as Al, Ag, Au, and
alloys of Ag–Au. There are very few reports in literature (in comparison with ns ablation) on the
femtosecond ablation of any targets (metallic and nonmetallic) in liquid media (Alnassar et al. 2013,
Ancona et al. 2014, Arboleda et al. 2015, Barchanski et al. 2015, Chelnokov et al. 2012, De Bonis
et al. 2014, Hatanaka et al. 1999, Hu et al. 2015, Liu et al. 2007, Povarnitsyn et al. 2013, Santagata
370 Semiconductor Nanocrystals and Metal Nanoparticles

et al. 2015, Semaltianos et al. 2010, Shaheen et al. 2013, Tagami et al. 2014, Tan et al. 2011, Tsuji
et al. 2003). Some of the large number of applications envisaged from NPs created with LAL and
other lithographic techniques include photovoltaic devices (Spyropoulus et al. 2012), photocata-
lytic activity (Liu et al. 2009, Zimbone et al. 2015), antibacterial properties (Guisbiers et al. 2015,
Ismail et al. 2015, Kamat 2002, Nath et al. 2012, Zimbone et al. 2015), magnetic resonance imaging
(MRI) (Tian et al. 2014), plasmonics (Intartaglia et al. 2013), upconversion (Onodera et al. 2013),
surface-enhanced Raman scattering (SERS) studies (Fan et al. 2013, Jin et al. 2014, Nedderson et al.
1993, Nie and Emory 1997, Ou et al. 2014, Qua et al. 2013, Wang et al. 2015, Zheng et al. 2014),
and photonics/nonlinear optics (Amendola et al. 2009, Del Fatti and Vallée 2001, Fan et al. 2014,
Fazio and Neri 2013, Ganeev et al. 2004, 2005, Hajiesmaeilbaigi and Motamedi 2007, Karavanskii
et al. 2004). LAL has enabled the fabrication of a variety of materials such as alloy NPs (Chen
and Beraun 2001, Compagnini et al. 2008, Kuladeep et al. 2012, Lee et al. 2001, Olea-Mejía et al.
2015, Shah et al. 2012), semiconductor NPs (Eroshova et al. 2012, Vadavalli et al. 2014), TiO2 NPs
(Siuzdak et al. 2014), Au-poly methyl methylyacrylate (PMMA) composites (Schwenke et al. 2013),
and MoS2 NSs (Compagnini et al. 2012).
This chapter is organized as follows. We first provide a brief introduction to ultrafast ablation
in air and water and explain their advantages and disadvantages. Later we discuss, in detail, the
fabrication of Ag, Cu, and Al NPs and NSs. We also summarize the results obtained from different
parameters such as input energy, scanning speeds, and input angle during ablation of these materi-
als. Further, we elaborate on three different applications of NPs and NSs generated by ultrafast laser
ablation in liquids (ULAL) technique: (1) SERS studies for explosives (e.g., RDX, trinitrotoluene
[TNT], ANTA, FOX-7, CL-20, etc.) detection using NSs of Ag, Cu, etc., created by ULAL; (2) pho-
tonic applications of NPs created by ULAL, wherein we utilized the Z-scan and four-wave mixing
techniques to study the nonlinear optical (NLO) coefficients in the ps and fs time domains; and
(3) antibacterial applications of the Cu NPs generated using ULAL.

11.2 FUNDAMENTALS OF ULTRAFAST LASER ABLATION

11.2.1 Ablation of Metals in Ambient Air
Laser ablation of solid targets is a top-down method and attracted the attention of the scientific com-
munity for the last half century. The material cannot be fragmented from the surface immediate to
the dumping of laser energy. The electric field provided by the impact of the laser pulse is sufficient
to eject only the surface electrons rather than the fragments. The ejection of electrons, ions, molecular
­clusters, and other intermediate products is governed by two important nonlinear mechanisms, namely,
­multiphoton absorption and cascade ionization (Amoruso et al. 2005a,b, Batani 2010, Batani et al. 2014,
Chichkov et al. 1996, Nolte et al. 1999, Upadhyay et al. 2008, Zhakhovskii et al. 2009, Zhigilei et al.
2009). These mechanisms are purely dependent on the pulse duration (Pereira et al. 2004). Initially,
surface electrons of the material absorb laser energy by means of inverse Bremsstrahlung and acquire
higher kinetic energies. Consequently, they maintain a temperature very much higher than that of the
surrounding lattice. Gradient in the temperature of the hot electrons and lattice can be explained by the
two-temperature model (Chen and Yeh 2001) governed by the following set of equations:

¶Te ¶Q( z)
Ce = - g(Te - Ti ) + S
¶t ¶t
¶T
Ci i = g(Te - Ti )
¶t (11.1)
¶T
Q( z) = -ke e
¶z
S = I (t )a exp(-az )
Applications of Metal Nanoparticles and Nanostructures 371

where
Q(z) is the heat flux
I(t) is the laser intensity
α is the material absorption coefficient
Ce, Ci are the electron and lattice heat capacities per unit volume
γ is the electron–phonon coupling constant
ke is the electron thermal conductivity

In general, in the case of ULA attainable, temperature and pressure of the plasma plume are ~103 K and
~1010 Pa, respectively (Margetic 2002). The hot electrons in the conduction band transfer the excess
heat energy through the electron lattice collisions, which lasts for a few ps. After thermalization, the
entire system reaches an equilibrium state. At this stage, if the attained temperature is greater than the
melting threshold of the material, then the portion of the material at the impact of laser pulses starts
melting immediately. The entire process lasts for several ps. In a time scale of a few microseconds, NPs
are ejected from the target surface. A few articles (Amoruso et al. 2005, Gamaly et al. 2002, Ivanov
and Zhigilei 2003, König et al. 2005, Noël et al. 2007, Povarnitsyn et al. 2007, Zhigilei et al. 2009)
are reported in recent literature that discusses in detail the complete mechanisms of ultrafast ablation
of metal and metal films in ambient air. The dynamics of ablation even in ambient air are affected by
the input laser parameters to a great extent. Lorazo et al. (2006) reported that pulse duration is indeed
an important parameter producing different outcomes in the case of longer pulses compared to the
ultrashort pulses. Lorazo et al. (2006) made an attempt to explain the thermodynamic pathways of
various stages of ablation in an extensive manner. They have observed that isochoric heating and rapid
adiabatic expansion of the material (silicon) provided a natural pathway for phase explosion. This was
not observed with slower, nonadiabatic cooling with ps pulses where fragmentation of the hot metallic
fluid was the only pertinent ablation mechanism. Amoruso et al. (2005) reported the ultrafast ablation
in vacuum, and comparison was carried out between ps and fs regime dynamics. Similarly, Balling
and Schou (2006) reported the dynamics of ns laser ablation of dielectrics in air ambient with plausible
explanations. Semaltianos (2010) reported the effect of medium (both air and liquid) on the fabricated
NPs and underlying physics elaborately. Although there are several reports describing that ULAL fab-
ricates a variety of metallic and nonmetallic NPs and NSs, there are very few reviews on this subject.

11.2.2 Demerits of Laser Ablation in Ambient Air

• The generated NPs during ablation process are expelled into the environment contaminat-
ing the surroundings.
• In this process, NSs can only be obtained by placing suitable wafers in the close vicinity of
ablation volume, resulting in inhomogeneous (thickness) film of NPs.
• Plasma produced in ambient air expands into the surrounding rapidly, which slows down
the process of NP generation. Ablation in air requires higher input energies to produce NPs.

11.3 ABLATION OF METALS IN LIQUIDS

If the target is surrounded by an aqueous medium, then the dynamics of ablation happen to be
extremely complicated. When the laser beam is focused on the surface of a target material under the
liquid layer, local melting of the metal target takes place. As a result, the metal portion at which the
laser beam is impinged goes to the melt phase. The adjacent liquid layer of the metallic melt absorbs
a part of the heat energy and attains higher temperature. At this temperature, liquid cannot sustain
in its own phase and, therefore, evaporates. During the process of evaporation, the liquid layer exerts
a recoil pressure on the underlying melt. This pressure is usually much higher than the atmospheric
pressure. However, the surface tension forces those exist in the melt try to sustain it as a spherical
entity. As a result of the higher recoil pressures exerted by the evaporating liquid layer, metallic
372 Semiconductor Nanocrystals and Metal Nanoparticles

melt splashes into fragments and each fragment of nanodimension goes into the surrounding liquid
medium. At higher values of the recoil pressure, in addition to the fabrication of NPs, redistribution
of the metallic melt takes place before its condensation, forming different NSs.

11.3.1 Cavitation Bubble Dynamics

The physical processes including the generation, transformation, and condensation of plasma plume
induced at the time of LAL have been investigated by many groups (Amendola et al. 2009, Besner
and Meunier 2010, Semaltianos 2010, Shafeev 2008, Simakin et al. 2004, Stratakis 2012, Yan and
Chrisey 2012, Zeng et al. 2012). The exact dynamics of LAL are still being debated, but according to
some of the recent works, it is explained as a complicated laser–matter interaction under the liquid
layer leading to the generation of the plasma plume. The created plasma expands into the surround-
ing liquid medium resulting in the generation of a shock wave. During the process of expansion,
the plasma plume cools down and plasma transfers energy to the surrounding liquid medium. The
annihilation of the plasma plume occurs in typical time scales of 10 –8 to 10 –7 s resulting in the for-
mation of a cavitation bubble that tries to expand in the medium in a time scale of 10 –7 to 10 –6 s and
continues expanding up to ~10 –4 s. After a certain period of time (usually of the order of few hundred
microseconds) at which the inside pressure decreases compared to the surrounding liquid medium,
the cavitation bubble collapses along with the generation of a second shock wave. The exact stages at
which nanomaterials are generated are still under investigation. Many reports suggest that the nano-
materials are formed during the expansion of the cavitation bubble inside it on a time scale of ~10 –6
to 10 –4 s. The expansion of the cavitation bubble into the liquid medium exerts a recoil pressure on
the metallic melt formed under the plasma plume. As mentioned earlier, the expansion of the plasma
plume and its transient dynamics can prompt the mixing of the vaporized material with the liquid
medium surrounding it. During later stages, the vaporized material condenses into the liquid medium
resulting in solidification (generation of NPs) in a time scale of a few hundred ns (≪time taken for
the condensation of vapor plume in liquid [~100 µs]). Nanoentities inside the cavitation bubble are at
a higher temperature than the surrounding liquid environment. The gradient in temperature on both
sides of the liquid–bubble interface leads to nucleation and condensation of the fabricated nanoma-
terials after collapse of the cavitation bubble. Fabrication of nanomaterials not only depends on the
laser parameters and materials but also on the nature of the surrounding liquid. The polarity, viscos-
ity, and refractive index of the surrounding media likewise play a key role in post ablation processes
such as nucleation and growth. In general, ablation of metals in polar liquids results in the fabrication
of NPs with smaller size through the formation of electrical double layers (EDLs) on the produced
charged NP surface that prevents further growth. Viscosity of the liquid medium sustains the plume
for a longer time and enhances the probability of the second ablation at the point of plume formation.

11.4 ADVANTAGES OF PULSED LASER ABLATION IN LIQUIDS

• Confinement of the produced plasma in aqueous media supports the ablation to occur at
lower energies compared to ablation in ambient air.
• Pulsed LAL avoids the contamination of the surrounding air media since the ablated frag-
ments (NPs) from the target surface directly enter the surrounding liquid medium to form
a colloidal solution.
• Utility of surfactants is not necessary to produce dispersed NPs unlike chemical methods
of NPs synthesis. Thus, ULAL can be considered as a green method.
• Simultaneous preparation of NPs and NSs can be achieved in a single experiment within a
few minutes while generating NSs over an area of a few inches. Moreover, the fabricated
NPs and NSs can be utilized for other experimental purposes devoid of subjecting them to
a rigorous cleaning with chemicals. ULAL is a simple platform to produce not only diverse
NPs of metals but also NPs of semiconductors, alloys, oxides, magnetic materials, biaxial
heterostructures, core–shell type, etc.
Applications of Metal Nanoparticles and Nanostructures 373

11.5 PARAMETERS INFLUENCING THE ABLATION MECHANISM

The products of ablation depend on many parameters of the input laser pulses as well as the sur-
rounding liquid media. Wavelength, pulse duration, fluence, beam waist, repetition rate, and number
of pulses incident per spot size are the pulse parameters, whereas linear refractive index, viscosity,
and the polarity of the liquid along with the thickness of the liquid layer on the target surface are
the liquid parameters that influence the process of ablation to a great extent. Variation of one of the
parameters to a slightest degree affects the nature of the products.

11.5.1 Input Wavelength
Wavelength of the incident laser beam determines the skin depth and, as a consequence, ablation
depth can be determined. Nichols and coworkers (2006a–c) reported that the absorption of UV pho-
tons by the surface electrons via interband transitions was more uniform and leads to nice corrugation
on the surface, whereas near-infrared radiation is preferentially absorbed by the defects and impuri-
ties of the material, consequently resulting in the formation of random surface structures. For UV
radiation, the tendency of scattering is higher compared to NIR. Furthermore, the wavelength of the
laser beam should be chosen in such a way that it should not be absorbed by the fabricated intermedi-
ate NPs within the liquid. It has been proven that the incident laser wavelength should not be equal to
the wavelength corresponding to the surface plasmon resonance (SPR) peak of the NP colloidal solu-
tion to avoid the ambiguity of the intermediate absorption of incident light. If absorption occurs, then
nanomaterials are subjected to unusual modifications. It has also been documented that the shorter
the wavelength, the higher the photon energies through which bond breaking and ionization processes
become simpler when compared to longer wavelengths. Mortazavi and coworkers (2011) reported
the observed variations in the size distribution of the fabricated Pd NPs in deionized water fabricated
with the wavelengths 1064 nm (Nd:YAG) and 193 nm (Excimer laser) at an equal pulse duration of 6
ns. According to their data and analysis, laser ablation with longer wavelengths followed bottom-up
behavior, which supported the aggregation of synthesized NPs due to thermal effects from plasma-
induced ablation ending up with NPs of larger sizes. Laser ablation with shorter wavelengths follows
top-down behavior directly fabricating smaller NPs since the photon energy is higher. Schwenke and
coworkers (2011) also reported the effect of laser ablation of metal in the solvent tetrahydrofuran
with fundamental (1064 nm) and second harmonic (532 nm) pulses in the ps regime and observed
good rate and yield of NP fabrication with 1064 nm pulses. A linear dependence of decrease in the
hydrodynamic particle size in the case of 1064 nm ablation was observed since the absorption and
scattering loss is weak at this wavelength. In the case of 532 nm ablation, exponential decrement of
the hydrodynamic particle size could be assigned to particle fragmentation during the postirradiation
of dispersed NPs. Moreover, depending on the wavelength of the incident laser, the cross section of
inverse Bremsstrahlung varies and cascade ionization will be altered since the aforementioned cross
section varies directly with the second power of the incident laser (Batani 2010, Batani et al. 2014).

11.5.2 Pulse Duration
Developments in achieving superior pulsed laser systems from ns time scales to fs time scales have
further established practical applications resulting from the advantages of pulsed laser ablation. Since
the short pulsed lasers support the higher peak intensities, explosive boiling of the material under
consideration has become an easier task. When laser pulses of longer duration (μs to ns) are incident
on the metallic targets, following the electron–lattice collisions, thermal processes such as explosive
boiling and evaporation take place. Further to the absorption of laser pulse, heat diffuses into the
lattice from the electrons in a very short time compared to its pulse duration (μs or ns). The process
of ablation of targets with μs or ns pulses is dominated by heat conduction, melting, evaporation, and
plasma formation. For the longer pulse ablation, HAZs prevail over short pulse ablation. We per-
formed the ablation of aluminum (Al) target using ps and fs pulses, and the differences between
374 Semiconductor Nanocrystals and Metal Nanoparticles

ablation by ~7 ns and ~40 fs laser pulses in water exhibited crater formation with different properties
(Podagatlapalli et al. 2014). Ablation with nanosecond pulses consists of large HAZs, while in the
case of femtosecond ablation, very minimal or negligible HAZ was observed. The melting continues
for a long time resulting in the boundaries of the ablated zone with no sharp boundaries. In addi-
tion, a plasma-shielding effect prevails in which the trailing part of the laser pulse interacts with the
plasma generated by the leading part of the pulse, resulting in unwanted modification of the nature
of previously produced nanomaterials. Recently, Niu and coworkers (2010) reported the advantage of
even longer-duration pulses (ms) to promote the surface reactions of metal droplets and control over
the formation of diverse NSs, hollow spheres, core–shell nanospheres, heterostructures, nanocubes,
and nanowires. A vast amount of experimental work has been carried out and reported on ablation
using ns laser pulses.
In the case of ultrafast pulses, ablation dynamics are explained by multiphoton-induced absorp-
tion and cascaded photoionization (avalanche). Particularly, for fs ablation, instantaneous multi-
photon absorption prevails, while in the case of ps ablation, the produced hot electrons behave as
seed in the buildup of avalanche. In the short pulse case, since the pulse duration τL is much shorter
than the electron–phonon equilibration time τE (10 –12 to 10 –11 s) and liquid–vapor equilibration
time τLV (10 –12 to 10 –11 s), nonthermal processes prevail, while in the ablation with longer pulses,
thermal processes usually dominate. During the ablation, unwanted interaction of the pulse with
the plasma plume is avoided. ULA can be considered as an isochoric process (considerable change
of the volume of sample does not occur) since the irradiation of laser pulse causes local heating in
very short time and lasts before the expansion of metal takes place. In the ultrafast ablation, accel-
eration of the ionized entities to enormous velocities leads to the development of higher pressures
and temperatures. Since the interaction time is very short, the material cannot evaporate continu-
ously but transforms into a state of overheated liquid. As a result, NSs formed on the solid targets
after irradiation of ultrafast laser pulse exhibit sharp boundaries with nice corrugation since the
material is evaporated with a minimal HAZ. Barmina and coworkers (2009) performed the ablation
of tantalum (Ta2O5) with laser pulses of 350 ps, 5 ps, and 180 fs duration, and their corresponding
wavelengths were 1064, 248, and 800, respectively, for fabricating metallic nanostructured sub-
strates. According to them, surface nanostructuring was chiefly dependent on the laser pulse dura-
tion (defines the energy density) than the wavelength of incident laser. In the case of 180 fs pulses
(it was not observed with the other two pulse durations), they observed small-scale periodic struc-
tures (along with hillocks) of periodicity less than the laser wavelength. The fundamental differ-
ences between the longer pulse ablation of targets with ultrafast pulse ablation were also reported,
and the effect of the number of pulses on ablation along with the spot size at the focus was discussed
(Leitz et al. 2011). Barcikowski and coworkers (2007) demonstrated the efficiency of ps and fs pulse
ablation of a Ag strip in water flow. It was observed that the ablation with liquid flow improved the
reproducibility and increased the productivity of NPs by 380% compared to ablation performed
with stationary liquid. Fs ablation in water was 20% more efficient than ps laser ablation, but due to
higher ps laser power (higher repetition rate), the productivity of NPs at the same fluence was three
times higher for ps ablation. Further, they observed that fs (120 fs, 5 kHz) ablation was efficient to
generate 2 μg of NPs/J, whereas it was 1.5 μg for ps (10 ps, 50 kHz) pulses. At the same time, the
rate of generation observed was excellent for ps ablation (34 mg/h) compared to fs ablation (6 mg/h).
Recent investigations (Riabinina et al. 2012) of the ablation phenomenon in water demonstrated bet-
ter yields of Au NPs. They found that ~2 ps was the optimal pulse duration at 5 mJ energy per pulse
at the expense of photoionization to obtain better yields of Au NPs.

11.5.3 Energy per Pulse, Spot Size, and Fluence

In general, the decomposition or material removal from a solid is the consequence of an energy input
into the target, resulting in the domination of the target’s binding energy. In ULAL, the total density
of hot electrons depends upon the energy of the primarily ejected ballistic electrons since their kinetic
Applications of Metal Nanoparticles and Nanostructures 375

energy only determines the number of secondary electrons. The mentioned electrons are the supreme
initiators for the formation of metal melt through electron–phonon collisions to attain equilibrium.
Thus, the entire phenomenon depends on the energy of the incident laser beam. Several groups
(Upadhyay et al. 2008, Zhigilei et al. 2009) performed molecular dynamic simulations to explain the
effect of laser parameters on the ablation and its products. Their studies illustrated that the rate of
ablation will be different for different pulse energies, and the productivity of nanomaterial fabrication
followed a linear relationship with the pulse energy. At higher energies, the volume of melt reservoir
will be high and, consequently, a higher number of particles can be fabricated. But in the sense of NS
fabrication, up to a certain energy corrugation was observed to be good, beyond which corrugation
became random structuring. The main reason behind this is the occurrence of many other dynam-
ics such as fragmentation, phase explosion, boiling, and vaporization (Ivanov and Zhigilei 2003,
Upadhyay et al. 2008, Zhigilei et al. 2009). Similarly, laser spot diameter at the focus also plays a
crucial role in determination of the products of the ablation. If the beam waist lies exactly on the sur-
face of the target, microstructures were observed, whereas the nanoripples with a periodicity of the
order 400 nm were observed when the target is placed beyond or before the focal plane. At the central
part of the beam, waist fragmentation mechanism prevails, while at the edges, thermal mechanisms
prevail. The absorbed energy density per unit time and unit volume by the material at the focal plane
is related to the gradient of the Poynting vector. If the laser beam is very tightly focused on the target
surface, it dumps the entire energy and may increase the productivity of NPs but simultaneously
loses its control on the fabrication of structures with sharp boundaries since tight focusing leads to a
gradient of temperature and pressure and, consequently, an inhomogeneous gradient of the Poynting
vector. In the case of a moderately focused laser beam, the gradient of the pressure and temperature
is smoother, leading to fine ablation of the target surface with a better performance in fabrication of
both NPs and NSs. A combined effect of beam waist at the focus and energy per pulse determines
the laser fluence, and this combined effect ablates the material in a different way to produce NPs
and NSs. Elsayed and coworkers (2013) also investigated the fabrication of gold NPs in water for
different laser fluences using 10 ns pulses and discussed increment and decrement about the certain
fluence and sizes of the gold NPs. Along with this, ablation of the gold surface exactly at focus and
just above/below the focal plane was also investigated. Kabashin and coworkers (2003) investigated
the effect of 110 fs pulses in the fabrication of gold NPs in water. They observed that the yield of Au
NPs was good at higher laser fluences (1000 J/cm2) than at lower fluences (60 J/cm2). Moreover, NP
distribution was observed to be better at lower fluences than at higher fluences. Average size was
small in the former case than the latter case for Au NPs. At intermediate fluences, both smaller and
larger NPs were observed. Barsch and coworkers (2009) reported the fabrication of Ag, Cu, Mg, and
ZrO2 NPs with ps laser pulses. They made an attempt to explain the fabrication of the mentioned
metal and ceramic NPs at different fluences and focal positions.

11.5.4 Number of Pulses
Another important parameter is the number of laser pulses per spot, which strongly affects the abla-
tion yield, size distribution, and the structure on the metal surfaces. The threshold fluence is a func-
tion of number of pulses impinging on the point of ablation. For multishot ablation, the threshold
fluence is different from the single-shot ablation since the incubation effects diminish the ablation
threshold. In the single-shot ablation, absorption of the laser energy by the target follows the Beer–
Lambert relation. But in the case of multishot ablation, we cannot apply the Beer–Lambert relation
since the reflectivity drops during the impulse of the first few pulses. The efficiency of the input
energy coupling increases with the surface structures through the surface plasmons and, thereby,
induces losses in reflectivity of the input beam. Leitz and coworkers (2011) discussed the effect of
the number of laser pulses on the crater formation for μs, ns, ps, and fs pulses, and diameters of the
holes formed were increased to a great extent and blurred in the case of micro (1,000 pulses) and ns
(250 pulses) ablation and to a smaller extent in the case of ps pulses (500,000) and fs pulses (5,000).
376 Semiconductor Nanocrystals and Metal Nanoparticles

11.6 RESULTS AND DISCUSSION

11.6.1 Experimental Details
The schematic of a typical ablation in the ultrashort pulse regime is demonstrated in Figure 11.2a.
Ultrafast experiments ablation were carried out using a chirped pulse-amplified (CPA) Ti:sapphire
laser system (LEGEND, Coherent) delivering nearly transform-limited laser pulses of ~2 ps
(or ~40 fs) duration, 1 kHz repetition rate, at 800 nm. Initial experiments were performed with
a Ag substrate submerged in double-distilled water (DDW) (2–3 mm above the Ag sample) in a
Pyrex cell that was placed on a motorized X–Y stage. Plane polarized (s-polarization) laser pulses
were allowed to focus vertically onto the Ag substrate through a planoconvex lens of focal length
25 cm. The position of the focus was approximated to lie at the point where plasma was generated.
X–Y translation stages, interfaced to Newport ESP 300 motion controller, were utilized to draw
periodic structures on the Ag substrate with separations of ~5, ~25, ~50, ~75 µm. Typical energy

M M

Brewster λ/2
Ultrashort polarizer
pulses M

M
Lens

Target
Oscillator
15 fs, 78 MHz Amplifier,
~40 fs/~2 ps,
Nd:YLF 200 ns, 1 kHz,
ESP 300 1 kHz) 800 nm
Motion controller

(a)

0.1
μm
60 μm

50 μm 50 μm
5 μm
Single line ablation Multiple line ablation
(b)

FIGURE 11.2 (a) Schematic of the ps laser ablation experimental setup. (b) Pulses incident on target.
(Reproduced with permission from Hamad, S., Podagatlapalli, G.K., Tewari, S.P., and Venugopal Rao, S.,
Influence of picosecond multiple/single line ablation on copper nanoparticles fabricated for surface enhanced
Raman spectroscopy and photonics applications, J. Phys. D Appl. Phys., 46, 485501. © 2013 by the IOP
Publishing.)
Applications of Metal Nanoparticles and Nanostructures 377

used was ~150 µJ per pulse. For example, in the case of ~5 µm, a structure was initially drawn on
the Ag substrate through the movement of the X-stage followed by vertical movement of ~5 µm
(using Y stage), and second-line structure was drawn on the substrate through the movement of the
X-stage in the opposite direction. Thus, ablation was carried out for ~25, ~50, and ~75 µm separa-
tions. The schematic of multiple/double/single ablation carried out on the Ag substrate is explained
in Figure 11.2b. The scanning speeds of the X–Y stages were 0.4 and 0.5 mm/s, respectively. For
the case of Bessel beam ablation, the same ~2 ps/~40 fs laser pulses were utilized. Ag targets were
purchased from Alfa Aeser (1 mm thick, 99.9% pure) and were immersed in HPLC grade DDW/
acetone located in a pyrex cell, and the whole setup was placed on a motorized nanodirect X–Y–Z
stage. The typical thickness of the liquid layer on the target surface was ~6 mm. ps/fs laser pulses
were focused onto the Ag target immersed in DDW/acetone by an axicon (base angle = 25°). To
match the focal plane exactly with the target surface, initially the focus was adjusted on the target
in the absence of any liquid. Later, depending on the thickness of liquid layer, focal plane displace-
ment (with respect to the focal plane in air) was estimated and corrected accordingly. Ablation of
the Ag target in DDW was performed with ~2 ps laser pulses wherein the target has been moved
by X–Y stages (Newport) to draw periodic lines at a separation of ~60 µm (total of 120 lines). The
speeds of the X and Y stages were 0.05 and 0.5 m/s, respectively. Accelerations of X and Y stages
were 0.05 and 0.5 m/s2, respectively. Input pulse energies utilized in the ps ablation case were
~400, ~600, ~800, and ~1000 µJ. Similarly, ablation of the Ag target in acetone was carried out by
~40 fs laser pulses. In this case, the Ag target in the focal plane of the Bessel beam was moved by
NTS NanoDirect stage in which the vertical stage (Z) was utilized to adjust the focal point on the
target surface and the other two were utilized to draw periodic lines (each line length was 5 mm,
total lines −80) on Ag targets with a separation of 25 µm. In this fs ablation case, pulse energies
utilized were of ~600, ~800, and ~1000 µJ. The speeds of the X and Y stages utilized in fs ablation
were 100 and 500 µm/s, respectively. Typical duration of each sample ablation was ~40 min. After
completion of the ablation, targets were removed and cleaned properly. Similarly, the colloidal solu-
tion was taken in air-tightened vessels to prevent it from oxidation. Focusing of Gaussian pulses
using an axicon and a transverse profile at the focus is shown in Figure 11.3a. Ag colloids prepared
in DDW with ~2 ps laser pulses had different degrees of gray coloration, whereas in acetone (with
~40 fs laser pulses), they exhibited diverse contrasts of golden-yellow coloration. Cu NPs and NSs
have been prepared by ablation of a Cu target in liquid media with ~2 ps pulses. Pure Cu targets
were washed with acetone after sonication to eliminate any residual organic impurities from the
surface. The laser pulses were allowed to focus onto the Cu substrate using a planoconvex lens
( f = 25 cm). The beam diameter (2ω 0) estimated at the focus in air was ~40 μm. The typical level
of liquid above the metal surface was ~7 mm. Typical fluence used was ~8 J/cm2. The scanning
speeds of X–Y stages were 0.1 mm/s in each direction. The motorized stages were scanned in such
a way as to draw periodic lines on the surface with the given spacing of (1) ~60 µm and (2) ~5 µm.
Based on the scanning configuration, two types of ablation were carried out: (1) single line ablation
wherein the average number of pulses incident per spot on the target was ~500 (0.1 μm/ms speed, 1
pulse/ms, assuming a spot diameter of ~50 μm on the target in liquid); and (2) double line ablation
in which the target surface was ablated twice (consequently the effective number of pulses per spot
was estimated to be ~1000).

11.6.2 Results from Ag Target Ablation in Water

11.6.2.1 Ag Targets Ablated with Different Fluences and Scanning Conditions
The prepared Ag colloids by multiple line, double line, single line ablation are designated as NP-1,
NP-2, NP-3, and NP-4 and the corresponding substrates as SS-1, SS-2, SS-3, and SS-4 for sepa-
rations of 5, 25, 50, and 75 µm, respectively (Podagatlapalli et al. 2012). As per the mechanism
shown in Figure 11.2b, SS-1 was influenced by multiple line ablations, whereas SS-2 was influ-
enced by double line ablation, since the estimated beam waist on the Ag substrate was >20 µm.
378 Semiconductor Nanocrystals and Metal Nanoparticles

Gaussian
2.0
α 407 (a) 400 μJ
(b) 600 μJ
1.5 407
(c) 800 μJ

Absorbance (a.u.)
(d) 1000 μJ
Zmax

1.0
407 d
b
0.5 c
407
a
0.0
400 600 800
SPR peak (nm)
(a) (b)

FIGURE 11.3 (a) Geometry of a focused Gaussian pulse using conical lens (Axicon) and formation of Bessel
beam pattern. (b) UV–Vis absorption spectra of Ag colloids prepared in double-distilled water. (Reproduced
from Podagatlapalli, G.K. et al., Laser Phys. Lett., 12, 036003, 2015. © Astro Ltd. With permission.)

Consequently, complete overwriting (ablation of preablated portion) was carried out for SS-1 and
partial overwriting for SS-2. Apart from these, SS-3 and SS-4 were obtained through single
ablation. Furthermore, Ag substrates (for a fixed scan separation of 20 µm) were also fabricated at
different fluences of ~4, ~8, ~12, ~16, and ~20 J/cm2. To achieve superior fabrication rate of NPs,
the laser beam should be focused exactly on the surface of the substrate. The theoretical beam
waist (ω 0) estimated at the focus (in air) was ~10 μm, while the dimensions of fabricated struc-
tures were measured to be ~20 μm in the single ablation case. Subsequent to the completion of the
scan, Ag substrates were removed from the liquid and preserved after proper cleaning. Colloidal
suspensions were preserved in air-tightened glass vials. The absorption spectra of obtained col-
loidal solutions and laser exposed substrates were recorded (immediately after ablation) using
a UV–Vis spectrometer (Jasco-V-670) equipped with an integrating sphere and working in the
spectral range of 250–2500 nm. The energy dispersive X-ray (EDX) spectra and high resolution
transmission electron microscope (HRTEM) images of the NP-1, NP-2, NP-3, and NP-4 were
recorded to identify the metallic fingerprints and NP size distribution along with their crystallo-
graphic information. Morphology of the laser-exposed portions of SS-1, SS-2, SS-3, and SS-4 was
characterized by a field emission electron microscope (FESEM) (Ultra 55 from Carl ZEISS) and
an AFM. Suspensions of Ag NPs were characterized by transmission electron microscope (TEM)
(SEI ­cecnai G2 S-Twin 200 kV instrument), providing an estimate of the size distribution and
morphologies of NPs. Ag colloids were centrifuged on carbon-coated copper grids and analyzed
using TEM operated at 200 keV. Details of the colloids and their characterization data have been
reported elsewhere (Podagatlapalli et al. 2013). HRTEM data of a single Ag NP exhibited paral-
lel Ag lattice planes having a separation of 2.2 Å corresponding to (111) lattice planes of silver.
Selected area electron diffraction (SAED) patterns revealed the polycrystalline nature of the Ag
nanospheres for all the four cases. In all of the cases, Ag NPs with polycrystallinity were evident
from a concentric ring pattern in SAED images. Measured lattice constants from the SAED pat-
tern agreed well with the reported values of silver and oxidized phases (minimal) of silver. The
recorded UV–Vis absorption spectra of NP-1, NP-2, NP-3, and NP-4 exhibited localized surface
plasmon resonance (LSPR) peaks in the neighborhood of 412, 417, 407, and 407 nm, respectively.
Depending on the geometry of ablation, the Ag NSs are labeled as SS-1 (line separation of
5 µm, multiple line ablation), SS-2 (line separation of 25 µm, double line ablation), and SS-3/SS-4
Applications of Metal Nanoparticles and Nanostructures 379

(line separation of 50 µm, 75 µm, single line ablation). The surface morphologies of the laser-
exposed portions of Ag substrates, characterized by FESEM, are presented in Figure 11.4a
through d, depicting NSs on SS-1 and SS-2 with dimensions of ~100 nm. There were no such
NSs on SS-3 and SS-4, which was confirmed from FESEM images. Because of multiple/double
line ablation, we could not observe distinguishable demarcation between two consecutive line
structures in SS-1 and SS-2, whereas for SS-3 and SS-4, we could observe separation between
the structures. The morphologies of ablated metallic surfaces suffer from incubation effects
(Podagatlapalli et al. 2013). The density and height of the nanosized domes were different for
SS-1 and SS-2. In addition to lateral NSs, there were small NP grains on both SS-1 and SS-2
substrates. These structures with Ag NP grains could enhance the Raman signals, and therefore,
SS-1 and SS-2 were used for Raman measurements of adsorbed rhodamine 6G (R6G) molecules.
The Raman spectra recorded for R6G on the laser-ablated surfaces SS-1 and SS-2 are shown in
Figure 11.5a for 532 nm excitation and in Figure 11.5b for 785 nm excitation. Excitation with
785 nm demonstrated higher Raman signature enhancement (compared to 532 nm excitation) and
lower fluorescence background from the absorbed analytes. Probably, the dome-like NSs provided

(a) (b)

2 μm EHT = 5.00 kV Signal A = InLens Date: 4 Apr 2012 1 μm EHT = 5.00 kV Signal A = InLens Date: 4 Apr 2012
2 μm Mag = 8.00 KX WD = 3.9 mm Time: 16:52:27
ZEISS
1 μm Mag = 20.00 KX WD = 3.9 mm Time: 16:54:53
ZEISS

(c) (d)

2 μm EHT = 5.00 kV Signal A = InLens Date: 4 Apr 2012 200 nm EHT = 5.00 kV Signal A = InLens Date: 4 Apr 2012
ZEISS ZEISS
2 μm Mag = 6.00 KX WD = 3.7 mm Time: 16:47:25 200 nm Mag = 50.00 KX WD = 3.9 mm Time: 16:57:47

FIGURE 11.4 FESEM images of the laser-exposed portions in Ag substrate in water (a) and (b) dome-like
structures formed on the substrate SS-1 because of the over wring, and its closer view (c) and (d) surface mor-
phology of SS-2 shows the dome-like structures. Closer view images show the fabricated Ag NPs grains on
the substrate. The scale bar in (a), (b), and (c) is 2 μm, while in (d) it is 200 nm. (Reprinted with permission
from Podagatlapalli, G.K., Hamad, S., Sreedhar, S., Prasad, M.D., and Venugopal Rao, S., Silver nano-entities
through ultrafast double ablation in aqueous media for surface enhanced Raman scattering and photonics
applications, J. Appl. Phys., 113, 073106. © 2013, American Institute of Physics.)
380 Semiconductor Nanocrystals and Metal Nanoparticles

772
610
7000

1184

Raman intensity (a.u.)

6000 1360
(II) 1309
1509
1644
5000 1572

4000 (I)

3000

500 1000 1500

(a) Raman shift (cm–1)

40,000 610 767 1269 1311

1178
1362
(III)
30,000 1508
Raman intensity (a.u.)

1647

20,000 (II)

10,000

(I)

0
500 1000 1500
(b) Raman shift (cm–1)

FIGURE 11.5 Raman signals recorded from R6G molecules (12 μΜ) in methanol (a) with micro-Raman
spectrometer (excitation at 532 nm). Red and blue spectra represent the enhanced Raman signatures of R6G
from SS-1 and SS-2 and (b) with bulk Raman spectrometer (excitation at 785 nm). Black spectrum was from
Ag plain surface. Red and blue spectra represent the enhanced Raman signatures of R6G from SS-1 and SS-2.
Signal collection time for both spectrometers was 5 s. (Reprinted with permission from Podagatlapalli, G.K.,
Hamad, S., Sreedhar, S., Prasad, M.D., and Venugopal Rao, S., Silver nano-entities through ultrafast double
ablation in aqueous media for surface enhanced Raman scattering and photonics applications, J. Appl. Phys.,
113, 073106. © 2013, American Institute of Physics.)

high local fields that contributed to significant Raman enhancements. Intensity enhancements of
~2439 were obtained for the prominent aromatic C–C stretch mode corresponding to 1362 cm−1
of R6G (12 μΜ) from SS-1 excited with 785 nm. This measurement was carried out by comparing
SERS spectra of R6G with reference spectra recorded at a higher concentration from the glass
side (nonplasmonic platform).
Figure 11.5 depicts the Raman spectra of R6G overlapped with the fluorescence background. This
could be explained on the basis of the distance between the analyte molecule and the nanomaterial.
Applications of Metal Nanoparticles and Nanostructures 381

Estimated intensity enhancements (ISERS/IRaman) for each mode of R6G from the multiply ablated
Ag NSs (SS-1) and doubly ablated Ag NSs (SS-2) with the excitation wavelengths of 532 and 785 nm
were ~103. The performance of SS-1 and SS-2 targets was evaluated by plotting the Raman signal
intensity enhancements. It was observed that superior Raman signal intensity enhancements were
observed from SS-2 (i.e., doubly ablated Ag NSs) for both the excitation wavelengths of 532 and
785 nm. Our observations demonstrated that the substrate with 25 µm line separations provided
optimal Raman enhancements due to double line ablation compared to the substrate obtained with
multiple line ablation. We believe that multiple line ablation, probably, caused a partial washout of
the nanoroughness compared to double line ablation. Measurements of R6G Raman enhancements
revealed that the performance of Ag NSs obtained from double line ablation was superior to NSs
fabricated by multiple line ablation. Therefore, we could conclude that scan separation of 25 µm
(double line ablation) provided better NSs and, hence, optimal Raman enhancements as a result of
double ablation.
Later, the combined effect of double line ablation on Ag targets with diverse input fluences
was investigated. For a fixed scan separation (25 µm), Ag substrates were ablated in DDW
at 4, 8, 12, 16, and 20 J/cm 2 input fluences. Figure 11.6 illustrates surface morphologies of
Ag substrates fabricated at different fluences (for a fixed line separation 25 µm). Along with
the formation of NSs, Ag NP grains were formed on these substrates. SERS activity of these
Ag substrates was investigated for RDX (C3H6N6O6; 1,3,5-trinitroperhydro-1,3,5-triazine) mol-
ecules dissolved in acetonitrile (ACN). Raman spectra recorded for RDX analyte adsorbed on
the Ag substrates fabricated with different laser fluences are illustrated in Figure 11.7a and b
for 532 and 785 nm excitations, respectively. The spectra were collected in the range of 250–
3500 cm−1, which covers most of the Raman bands of RDX. In the recorded SERS spectra of
RDX from the Ag nanostructured target (Figure 11.7a), only few peaks (Besner et al. 2007,
2008, Robitaille et al. 2013) were observed compared to the conventional bulk Raman spectra
(Figure 11.7b) of RDX. Raman signatures observed in the case of 785 nm excitation were higher
in number compared with 532 nm excitation. The modes corresponding to 383 and 924 cm−1
(Figure 11.7a) represent the signatures of ACN, and the remaining were from RDX. Intensity
enhancements for the observed Raman modes of RDX were estimated, and typical numbers are
provided as follows. In the case of excitation of 532 nm, modes corresponding to 2947, 2255, and
3005 cm−1 (C–H stretch) exhibited ~582, ~652, and ~86 times of the enhancement, respectively,
compared to the same modes observed from higher concentrations of RDX molecules adsorbed
on the glass slide. Some of the other modes observed were 1167 and 1381 cm−1 (CH 2 twisting).
However, fluorescence prevailed in the spectra obtained from the substrates prepared at fluences
of ~16 and ~20 J/cm 2 (Podagatlapalli et al. 2013). With 785 nm excitation (Figure 11.7b), the
dominant modes were 856 cm−1 (N–N stretch + NO2 axial scissoring), 924 cm−1 (CH 2 rocking
or combination), 1240 cm−1 (N–N and symmetric + NO2 stretch and may be with CH 2 twist),
1314 cm−1 (N–N stretching + CH 2 stretch), 1393 cm−1 (CH 2 twisting), 1457 cm−1 (CH 2 scissoring),
and 1560 cm−1 (ONO equatorial stretching). The observed Raman modes of RDX are in good
agreement with earlier reports (Sylvestre et al. 2005). The SERS activity of Ag targets fabricated
at different fluences for double line ablation was evaluated by plotting intensity enhancements
of RDX Raman modes with respect to the fluence at which NSs were prepared. In the case of
532 nm excitation, strong Raman signal enhancements were observed from the Ag substrates
fabricated at 12 J/cm 2. Our data analysis suggested that most of the RDX Raman modes were
enhanced strongly from the Ag NSs fabricated at 16 J/cm 2. From the analysis of Raman signal
enhancements of RDX with 532 nm and 785 nm excitations illustrated, optimal fluences were
~12 and ~16 J/cm 2, respectively (Podagatlapalli et al. 2013).

11.6.2.2 Ag Targets Ablated with Different Angles of Incidence

Morphologies of laser-exposed portions of Ag substrates fabricated with different angles of
incidence were investigated through FESEM and AFM imaging techniques. Figure 11.8a and c
382 Semiconductor Nanocrystals and Metal Nanoparticles

(a) (b)

100 200
100 nm EHT = 5.00 kV Signal A = InLens Date: 8 Jun 2012 200 nm EHT = 5.00 kV Signal A = InLens Date: 8 Jun 2012
ZEISS ZEISS
Mag = 200.00 KX WD = 4.4 mm Time: 15:32:23 Mag = 120.00 KX WD = 4.3 mm Time: 15:40:48

(c) (d)

200 100
100 nm EHT = 5.00 kV Signal A = InLens Date: 8 Jun 2012 100 nm EHT = 5.00 kV Signal A = InLens Date: 8 Jun 2012
ZEISS ZEISS
Mag = 100.00 KX WD = 3.9 mm Time: 15:47:37 Mag = 100.00 KX WD = 4.0 mm Time: 16:48:38

FIGURE 11.6 FESEM images (closer view of the structure) of the laser exposed portions in Ag substrate in
water prepared for different laser fluences: (a) 4 J/cm2, (b) 8 J/cm2, (c) 12 J/cm2, and (d) 16 J/cm2. The images
depict fabricated grains of Ag NPs on the substrate. The scale bar in (a) and (d) is 100 nm, and for (b) and (c),
it is 200 nm. (Reprinted with permission from Podagatlapalli, G.K., Hamad, S., Sreedhar, S., Prasad, M.D.,
and Venugopal Rao, S., Silver nano-entities through ultrafast double ablation in aqueous media for ­surface
enhanced Raman scattering and photonics applications, J. Appl. Phys., 113, 073106. © 2013, American
Institute of Physics.)

illustrates the FESEM images of laser-exposed portions of Ag substrates corresponding to NS-5

and NS-15. Figure 11.8b and d represents the AFM images of NS-5 and NS-15, respectively. AFM
images exhibited lateral NSs of dimensions <200 nm. Similarly, Figure 11.9 depicts the surface mor-
phologies of NS-30 and NS-45. Figure 11.9a and c illustrates the FESEM images of laser-exposed
portions of the Ag substrate corresponding to NS-30 and NS-45, respectively. Figure 11.9b and d
illustrates the AFM images of NS-30 and NS-45, respectively. AFM images confirmed the presence
of lateral NSs on the surface of NS-30. Ablation of targets was carried out in a controlled manner
without changing the ablation conditions except the angle of incidence. The surface topology of
laser-exposed portions of the substrates demonstrated different morphologies for NS-30, which is
evident from Figure 11.9b.
Nanostructured Ag targets fabricated at oblique laser incidence in DDW are labeled as NS-5 (Ag
target ablated at 5° of laser incidence), NS-15 (Ag target ablated at 15° of laser incidence), NS-30
(Ag target ablated at 30° of laser incidence), and NS-45 (Ag target ablated at 45° of laser incidence).
Applications of Metal Nanoparticles and Nanostructures 383

Plain
32,000 50 μJ
100 μJ
150 μJ
924 cm–1 2947 cm–1
200 μJ

Raman intensity (a.u.)

24,000 250 μJ
386 cm–1 1167 cm–1 2255 cm–1
1381 cm–1

16,000 3005 cm–1

8,000

00 800 1600 2400 3200

(a) Raman shift (cm–1)

16,000
1008 cm–1 1393 cm–1
Plain
379 cm–1 919 cm–1 50 μJ
–1
1240 cm 100 μJ
712 cm–1
12,000 150 μJ
601 cm–1
Raman intensity (a.u.)

–1
856 cm–1 1132 cm 200 μJ
250 μJ
1078 cm–1
1314 cm–1
8,000 1457 cm–1

1560 cm–1

4,000

400 800 1200 1600

(b) Raman shift (cm–1)

FIGURE 11.7 (a) Micro-Raman spectra of adsorbed RDX molecules (excitation at 532 nm) in acetonitrile.
The maximum enhancement of Raman signal was observed at pulse energy 150 µJ (fluence of 12 J/cm 2).
(b) Bulk Raman spectra of the adsorbed RDX (excitation at 785 nm) in acetonitrile. The maximum
enhancement of Raman signal was observed at pulse energy of 200 µJ (fluence of 16 J/cm 2). More signa-
tures were observed with 785 nm excitation compared to 532 nm excitation. For both the cases, the time of
integration was 5 s. (Reprinted with permission from Podagatlapalli, G.K., Hamad, S., Sreedhar, S., Prasad,
M.D., and Venugopal Rao, S., Silver nano-entities through ultrafast double ablation in aqueous media
for surface enhanced Raman scattering and photonics applications, J. Appl. Phys., 113, 073106. © 2013,
American Institute of Physics.)

The Raman spectra of R6G/ANTA adsorbed on NS-5, NS-15, NS-30, and NS-45 were recorded
with micro-Raman (WiTech ALPHA 300 instrument) using a continuous wave (cw) Nd:YAG laser
at 532 nm and bulk Raman spectrometers (Ocean Optics) with a cw Ar+ laser at 785 nm. In the
micro-Raman spectrometer, the laser beam was focused on the substrate with an objective lens
(100×) whose beam waist estimated was ~700 nm. Raman signals were collected in the backscat-
tering geometry. In bulk Raman spectrometer (excitation at 785 nm), the laser beam (diameter of
~1 mm) was directed toward the sample without any focusing lenses, and Raman signals were
384 Semiconductor Nanocrystals and Metal Nanoparticles

(a) (b)

[nm]
150

0
0
1
1
2
2
33
[μm]

200 nm EHT = 5.00 kV Signal A = InLens Date: 22 Mar 2013

ZEISS
Mag = 50.00 KX WD = 4.0 mm Time: 13:31:20

(c) (d)

[nm]

200 0

2
0 4
2 6
4
6 8
8
10 10
200 nm EHT = 5.00 kV Signal A = InLens Date: 22 Mar 2013
[μm]
ZEISS
Mag = 50.00 KX WD = 4.0 mm Time: 13:24:30

FIGURE 11.8 (a) FESEM imaging of laser-exposed Ag substrates NS-5. (b) AFM imaging of laser-exposed
Ag substrates NS-5. (c) FESEM imaging of laser-exposed Ag substrates NS-15. (d) AFM imaging of laser-
exposed Ag substrates NS-15. (Reprinted from Appl. Surf. Sci., 303, Podagatlapalli, G.K., Hamad, S.,
Mohiddon, M.A., and Venugopal Rao, S., Effect of oblique incidence on silver nanomaterials fabricated
in water via ultrafast laser ablation for photonics and explosives detection, 217–232. © 2014, with permission
from Elsevier.)

collected in the backscattering geometry. All the spectra were calibrated to Raman peak of silicon
wafer at 520 cm−1. The Raman spectra of R6G (10 µL) were recorded from the structured Ag sub-
strates. First, the performance of NS-5, NS-15, NS-30, and NS-45 was investigated by recording
the Raman spectra of R6G (25 µM) with excitation wavelengths of 785 and 532 nm. Figure 11.10a
through d depicts the Raman spectra of R6G (~10 µL drop) placed on the laser-exposed portion to
form a monolayer of analyte (excitation wavelength of 785 nm). Multilayered analyte adsorption on
NSs generally inhibits the activity of Raman signal enhancements since the multilayers screen the
effect of evanescent fields from NSs. Furthermore, the SERS activity of the substrate depends on the
distance between NSs and analyte molecule (typically should be few nm). To compare the Raman
spectra (excitation wavelength of 785 nm) of the analyte from the laser-ablated Ag surface, the
Raman spectra of R6G from the plain Ag surface (blue) and Si substrate (black) were also recorded.
The estimated Raman mode intensity enhancements of R6G (25 µM) at 1360 cm−1 were ~13, ~12,
~25, and ~14 from NS-5, NS-15, NS-30, and NS-45, respectively. Similarly, the Raman spectra of
R6G (25 nM) were also recorded from four substrates after proper cleaning, and comparison is
shown in Figure 11.11.
Applications of Metal Nanoparticles and Nanostructures 385

(a) (b)

650 [nm]
0

0
4

2 6
4
6 8
8
200 nm EHT = 5.00 kV Signal A = InLens Date: 22 Mar 2013
10 10
ZEISS
Mag = 50.00 KX WD = 4.0 mm Time: 11:57:39 [μm]

(c) (d)

300 [nm]

0
0

2
2

4 4

200 nm EHT = 5.00 kV Signal A = InLens Date: 22 Mar 2013 [μm]

ZEISS
Mag = 50.00 KX WD = 4.0 mm Time: 11:46:00

FIGURE 11.9 (a) FESEM imaging of laser-exposed Ag substrates NS-30. (b) AFM imaging of laser-
exposed Ag substrates NS-30. (c) FESEM imaging of laser-exposed Ag substrates NS-45. (d) AFM imaging
of laser-exposed Ag substrates NS-45. (Reprinted from Appl. Surf. Sci., 303, Podagatlapalli, G.K., Hamad, S.,
Mohiddon, M.A., and Venugopal Rao, S., Effect of oblique incidence on silver nanomaterials fabricated in
water via ultrafast laser ablation for photonics and explosives detection, 217–232. © 2014, with permission
from Elsevier.)

The performance of each substrate was evaluated using the enhancement factors (EFs) given by
(Podagatlapalli et al. 2014):

I SERS N Raman
EF =
I Raman N SERS

where
ISERS is the integrated intensity of R6G Raman band under consideration from the nanostructured
Ag substrate
IRaman is the integrated intensity of the same Raman band from higher molar R6G (0.25 M) on
the silicon substrate

Similarly, NSERS is the number of analyte molecules constituting the first monolayer on the nano-
structured substrate surface under the laser spot area and NRaman is the number of molecules covered
by the beam waist in the bulk R6G solution on Si target. NSERS and NRaman were estimated to be
386 Semiconductor Nanocrystals and Metal Nanoparticles

5000 4000
1362 1364
NS-5 + R6G NS-15 + R6G
4000 1510
1510 3000 1310
1310 610
767
1272
Intensity (a.u.)

Intensity (a.u.)
3000 767 1183 1274
1183
610 2000
2000

1000
1000

0 0
600 900 1200 1500 600 900 1200 1500
(a) Raman shift (cm–1) (b) Raman shift (cm–1)

6000
8000 NS-30 + R6G 1362
NS-45 + R6G 1360

4500
1510 772 1272
Intensity (a.u.)
6000
Intensity (a.u.)

1310
610 1308 1510
770 611 1187
3000
4000 1185 1269

1500
2000

0
0 600 900 1200 1500
500 1000 1500
(c) Raman shift (cm–1) (d) Raman shift (cm–1)

FIGURE 11.10 Raman spectra of R6G (25 µM) recorded from the (a) NS-5, (b) NS-15, (c) NS-30, and (d)
NS-45 with an excitation wavelength 785 nm. Time of integration for each spectrum was 10 s. (Reprinted
from Appl. Surf. Sci., 303, Podagatlapalli, G.K., Hamad, S., Mohiddon, M.A., and Venugopal Rao, S., Effect
of oblique incidence on silver nanomaterials fabricated in water via ultrafast laser ablation for photonics and
explosives detection, 217–232. © 2014, with permission from Elsevier.)

~120 and ~5 × 109, respectively. In this estimation, the adsorption factor was considered to be 0.3.
We arrived at this estimate by considering the beam waist at the focus to be ~0.65 µm for 532 nm
excitation and the size of the R6G molecule as 2 × 10 –18 m2. For aromatic C–C stretch mode corre-
sponding to 1362 cm−1, estimated EFs were ~2.5 × 108, ~2.6 × 108, ~5 × 109, and ~3.4 × 108 from the
substrates NS-5, NS-15, NS-30, and NS-45, respectively. Estimated intensity enhancements and EFs
revealed that a considerably superior performance of NS-30 (prepared at 30°) compared to others
for both µM and nM concentrations of R6G excited with 785 and 532 nm wavelengths.
After cleaning (several times with acetone) and sonication, four Ag NSs were utilized again
to record Raman spectra of an explosive molecule 5-amino, 3-nitro, -1H-1,2,4-nitrozole (ANTA)
dissolved in ACN (5 mM, 5 µM). The spectra are shown in Figure 11.12a and b for mM and μM
concentrations, respectively. As shown in Figure 11.12, dominant modes were 470, 488 cm−1 (both
are NO2 deformation), 724 cm−1 (ring deformation), 843 cm−1 (NO2 deformation + ring defor-
mation), 1026 cm−1 (N1–N2–C3 bend), 1125 cm−1 (NN-symmetric stretch), 1340 cm−1 (C–NO2
symmetric stretch), 1530 cm−1 (C–NH2 asymmetric stretch + NH2 bend), and 1588 cm−1 (C–NH2
symmetric stretch + NH2 bend). For C–NO2 symmetric stretch mode corresponding to 1340 cm−1,
estimated values of NSERS and NRaman were ~1.6 × 109 and ~4 × 1011, respectively. Estimated EFs
Applications of Metal Nanoparticles and Nanostructures 387

1540
1 NS-30 + R6G
2 NS-45 + R6G 1365
4500
3 NS-15 + R6G
4 NS-5 + R6G
1227 1506
1605
Intensity (a.u.) 1306
3000 746
610 1640
1

1500
2

3 4
0
500 1000 1500 2000

Raman shift (cm–1)

FIGURE 11.11 Raman spectra of R6G (25 nM) recorded from the NS-5, NS-15, NS-30, and NS-45 with an
excitation wavelength 532 nm and an integration time 5 s. (Reprinted from Appl. Surf. Sci., 303, Podagatlapalli,
G.K., Hamad, S., Mohiddon, M.A., and Venugopal Rao, S., Effect of oblique incidence on silver nanomaterials
fabricated in water via ultrafast laser ablation for photonics and explosives detection, 217–232. © 2014, with
permission from Elsevier.)

for the 1340 cm−1 mode were ~4.1 × 103, ~5.5 × 103, ~9.7 × 104, and ~4 × 103 from Ag substrates
NS-5, NS-15, NS-30, and NS-45, respectively (using adsorption factor [η]—0.6). Similarly, esti-
mated values of NSERS and NRaman were ~0.8 × 106 and ~1.6 × 1011, respectively, for 5 µM concen-
tration of ANTA (C–NO2 symmetric stretch mode corresponding to 1340 cm−1). Estimated EFs
were ~8.5 × 104, ~4.2 × 105, ~1.1 × 106, and ~4 × 105 for the Ag substrates NS-5, NS-15, NS-30,
and NS-45, respectively (adsorption factor [η]—0.4). As per our observations, estimated EFs for
R6G and ANTA (both mM and μM) from NS-30 demonstrated stronger enhancements (9.7 × 104,
1.1 × 106) compared to the other three substrates (Podagatlapalli et al. 2014).
The most important reason behind the obtained enhancement of the Raman signatures from the
corrugated surface achieved by laser ablation is that it probably acted like a random grating, thus
supporting the excitation of both the localized surface plasmons (presence of Ag NPs grains) and
surface plasmon polaritons (presence of random grating). Coupling of incident photons to surface
plasmons of nanomaterials determines the magnitude of evanescent field in the vicinity of nano-
materials. The essential condition for coupling in case of surface plasmon polaritons is the spac-
ing of the grating on a metallic target. Similarly, the size and shape of nanomaterials (NSs and
NPs) and wavelength of excitation are the key parameters to couple incident light to the localized
surface plasmons. Effective coupling of incident photons with surface plasmons of nanomaterials
enables nanotips to act as periodically oscillating dipole antennas with an oscillation frequency of
incident radiation. As a result, an analyte molecule in the vicinity of nanotip experiences collective
electric field of incident photons and field provided by a radiating dipole. Consequently, enhance-
ment of the Raman signals from both plasmons (localized and propagating) plays a critical role.
Further, we believe that the Raman signal enhancement of analytes from the laser-exposed portions
of Ag substrates depends on the degree of roughness and topology of the surfaces. Ablation of
Ag target at 30° incident angle enabled the surface to contain more NSs and random gratings com-
pared to NS-5, NS-15, and NS-45. In the case of NS-30, coupling of incident photons to the Ag surface at
30° incidence angle could have resulted in efficient ablation and formation of NSs. Thus, NS-30 was
388 Semiconductor Nanocrystals and Metal Nanoparticles

1340
30,000
1 NS-5 + ANTA
1126
2 NS-15 + ANTA
3 NS-30 + ANTA
4 NS-45 + ANTA
Intensity (a.u.)
20,000 1443

2458
3 2242 2671
2020
10,000 897

2
4

0 1

500 1000 1500 2000 2500 3000

(a) Raman shift (cm –1)

360 1330
1530
1 NS-45 + ANTA
2 NS-30 + ANTA
1400
1588 3 NS-15 + ANTA
320 4 NS-5 + ANTA
Intensity (a.u.)

724
1110
863 1
280 470

3
2
4
240

600 1200 1800 2400 3000

(b) Raman shift (cm–1)

FIGURE 11.12 Raman spectra of ANTA recorded from NS-5, NS-15, NS-30, and NS-45 with an excita-
tion wavelength 532 nm for (a) 5 mM concentration and (b) 5 μM concentration. Time of integration for each
spectrum was 5 s for data presented in (a) and 0.5 s for data presented in (b). (Reprinted from Appl. Surf. Sci.,
303, Podagatlapalli, G.K., Hamad, S., Mohiddon, M.A., and Venugopal Rao, S., Effect of oblique incidence
on silver nanomaterials fabricated in water via ultrafast laser ablation for photonics and explosives detection,
217–232. © 2014, with permission from Elsevier.)

consistent to provide large EFs and efficient trace detection. Furthermore, we succeeded in utilizing
the Ag NSs for the identification of two different analytes separately. For both analytes, R6G and
ANTA targets fabricated at 30° demonstrated a very significant Raman signal elevation.

11.6.2.3 Ag Targets Ablated with ps/fs Bessel Beams

Laser ablation dynamics of metals in liquids simultaneously depend on laser parameters such as wave-
length, repetition rate, pulse duration, energy per pulse, and focusing geometry. Most of the experiments
Applications of Metal Nanoparticles and Nanostructures 389

in this field were carried out by laser pulses with focused Gaussian pulses with conventional convex
lenses. In conventional Gaussian beam focusing experiments, efficient nanomaterial generation is pos-
sible when the spot size (beam waist) on the target is very small. However, the least beam waist mini-
mizes the depth of focus (DOF) (Zmax), which, in fact, introduces artifacts and creates a loss of control
on the position of beam waist for placing the sample exactly at focus. The focusing of Gaussian pulses
through an axicon (shown in Figure 11.3) eliminates the aforesaid problems by ­offering a considerable
DOF. Additionally, it maintains comparatively similar spot size without any translational spread due
to diffraction effects. Moreover, these nondiffracting Bessel beams exhibit nearly a constant inten-
sity profile along their propagation. Salient features of Bessel beams, such as self-reconstruction and
stability under nonlinear propagation, make them extensively useful in extremely localized and con-
trolled energy deposition in transparent materials. The interference of conical wave fronts produced by
an axicon is the primary reason to produce the Bessel beam that has the field and intensity distribution
given by E(ρ) = E0exp(−ρ2/ω2), where E0 is the on-axis field amplitude, ρ is the radial distance from
the propagation axis Z, and ω is the radius of Gaussian beam. We performed ablation experiments
to exploit the (1) uniqueness of nondiffracting Bessel beams over conventional Gaussian beam and
(2) compared the performance of Ag NPs fabricated by picosecond (ps)/femtosecond (fs) laser pulses.
The LSPR peak position for the Ag colloids prepared using Bessel beam and input energies of ~400,
~600, ~800, and ~1000 µJ was positioned at 407 nm, which is contrary to the reports on ablation till
date obtained using Gaussian laser pulses (Podagatlapalli et al. 2015). Although the absorbance of
colloids (measure of yield) changed in accordance with the input energy, LSPR peak position did not
change. LSPR peak position is the combined effect of all neighboring NPs covered under the diameter
of the light beam path. Other morphological characterizations such as TEM and UV–Vis absorption
provide global information on average sizes of NPs. LSPR peak possibly shifts due to the sizes of
neighboring NPs. But in the present case, it was different, demonstrating the possibility of fabricating
NPs with similar average sizes even at four different input energies.
Four Ag nanostructured targets were fabricated in DDW with ~2 ps pulses at ~400, ~600,
~800, and ~1000 μJ. Similarly, three more Ag nanostructured targets were fabricated in acetone
with ~40 fs pulses at ~600, ~800, and ~1000 μJ. These two sets of Ag NSs fabricated with ps
and fs Bessel beams were utilized to detect/identify explosive molecules, 2,4,6,8,10,12-hexanitro-
2,4,6,8,10,12-hexaazaisowurtzitane (CL-20). NSs were obtained for high-input laser energies such
as ~400, ~600, ~800, and ~1000 μJ, only. At lower energies (<400 μJ), we could not observe NSs
on laser-exposed portions. Morphologies were investigated by FESEM, and these surfaces were
utilized to record the Raman spectra of adsorbed CL-20 (5 µM). Raman spectra were recorded with
an excitation wavelength of 532 nm. Figures 11.13 and 11.14 demonstrate the FESEM images of
laser-exposed Ag targets ablated with ~2 ps and ~40 fs pulses, respectively. Insets depict the SERS
spectra (red) of CL-20 (5 µM) from the laser-exposed portions, and blue curves represent the 0.1 M
CL-20 Raman spectra from silicon wafer. Our data suggested that the highly elevated mode in CL-20
was 1330 cm−1 (CH bend + NO symmetric stretch) and EFs were estimated by considering ISERS and
IRaman of the mode mentioned from NSs as well as the silicon target. Estimated EFs were ~1.1 × 106,
~3.4 × 106, ~3.3 × 105, and ~3.8 × 105 for nanostructured targets prepared with ~2 ps pulses at ~400,
~600, ~800, and ~1000 µJ, respectively. Similarly, EFs of the Ag targets prepared with ~40 fs laser
pulses were estimated as 1.24 × 106, 1.9 × 106, and 8.07 × 105. From the comparison of EFs, we found
that in both regimes, the order of EF was similar, but a slightly lower enhancement was observed from
the Ag NSs prepared with ~2 ps pulses. To estimate these factors, beam waist of the excitation wave-
length (532 nm) at the focus was considered to be ~650 nm. On comparison of the EFs, we found that in
both regimes the order of magnitude was the same (~106) (Podagatlapalli et al. 2013).

11.6.3 Cu Targets Ablated with Different Energies and Scanning Conditions

Cu colloidal NPs have been produced by ps laser ablation in four different liquid media. To avoid the
ambiguity, Cu NPs prepared through multiple line (single line) ablation in acetone, dichloromethane
390 Semiconductor Nanocrystals and Metal Nanoparticles

(a) (b)

1585

1328
1267

1562
1331
1500
500

Intensity (a.u.)

Intensity (a.u.)

1204
1620

1640
1000

808
852
574
1259
775
400

941
806
500
300
300

225
200

10 μm EHT = 5.00 kV Signal A = 600 900 1200 1500 1800 10 μm 600 900 1200 1500 1800
EHT = 5.00 kV Signal A =
Mag = 3.00 KX WD = 3.8 Raman shift (cm–1) Mag = 1.00 KX WD = 3.7 Raman shift (cm–1)

(c) (d)

1548
1322
1384
1596

1255
1318
Intensity (a.u.)
Intensity (a.u.)

1243

900

1384
1124

360

1171
1030

1596
758
583
600
913
760
860

300
300
250
240 200
150
100
600 900 1200 1500 1800 600 900 1200 1500 1800
10 μm EHT = 5.00 kV Signal A = 10 μm EHT = 5.00 kV Signal A =
Mag = 1.00 KX WD = 3.7 Raman shift (cm–1) Mag = 1.00 KX WD = 3.7 Raman shift (cm–1)

FIGURE 11.13 FESEM images Ag targets with Bessel beams at energies (a) 400 µJ, (b) 600 µJ, (c) 800 µJ,
and (d) 1000 µJ. Insets show SERS spectra in red of CL-20 (5 µM) from the laser-exposed portions and
blue represent the 0.1 M CL-20 Raman spectra from silicon wafer. Time of integration was 0.5 s for all
the cases. (Reproduced from Podagatlapalli, G.K. et al., Laser Phys. Lett., 12, 036003, 2015. © Astro Ltd.
With permission.)

(DCM), ACN, and chloroform (CHCl3) were labeled as MCuNP1 (SCuNP1), MCuNP2 (SCuNP2),
MCuNP3 (SCuNP3), and MCuNP4 (SCuNP4), respectively. Figure 11.15 summarizes the Raman
spectra recorded for (1) MCuNP1 and SCuNP1, (2) MCuNP2 and SCuNP2, (3) MCuNP3 and
SCuNP3, (4) MCuNP4 and SCuNP4, respectively. In the case of MCuNP1, the Raman signal inten-
sity for the 531.04 cm−1 mode was ~2179, whereas the Raman intensity for pure acetone was ~9, and
therefore, the intensity enhancement was estimated to be ~242. Our observation was that Raman
signatures in MCuNP1 were elevated better compared to SCuNP1. In the case of MCuNP2 and
SCuNP2, the intensity enhancements were ~11 and ~5 for the 283.4 cm−1 mode and ~5.5 and ~5 for
the 701.6 cm−1 mode, respectively. For the case of MCuNP3 and SCuNP3, the intensity enhance-
ments were ~54.8 and ~30.7 for the 379.05 cm−1 mode, ~50 and ~29.6 for the 919.53 cm−1 mode,
and ~16.3 and ~11.6 for the 1374.92 cm−1 mode, respectively. Similarly, in the case of MCuNP4
and SCuNP4, the intensity enhancements were ~26.9 and ~5.6 for the 258.5 cm−1 mode, ~24.5 and
~5.1 for the 364.5 cm−1 mode, ~24.5 and ~5.8 for the 665.9 cm−1 mode, and 51.3 and 10.9 for the
754.6 cm−1 mode, respectively. Our studies clearly suggest that these NPs have potential for SERS
studies of other analyte molecules (Hamad et al. 2013).
Figure 11.16 illustrates the FESEM images of fabricated NSs on the surface of Cu ablated in
(1) acetone (MCuNS1), (2) DCM (MCuNS2), (3) ACN (MCuNS3), and (4) chloroform (MCuNS4),
Applications of Metal Nanoparticles and Nanostructures 391

(a) (b)

1344
1050

1605

1513
1583
1451
900 900

1410

Intensity (a.u.)

1402
1248
Intensity (a.u.)

1344

1104
750 750

936
1252

988
866

760
988
600 600

650
702
450 450

280 280
240 240
200 200
160 160
800 1200 1600 2000 800 1200 1600 2000
20 μm EHT = 5.00 kV Signal A 20 μm EHT = 5.00 kV Signal A
Mag = 643 X WD = 2 Raman shift (cm–1) Mag = 592 X WD = 2 Raman shift (cm–1)

(c)
6.0 6.0
EFs of Ag targets prepared at 2 ps
EFs of Ag targets prepared at 40 fs
EF (×106) 4.5 4.5

EF (×106)
3.0 3.0
750
1348
1512
Intensity (a.u.)

1385
1294
1176

600
1.5 1.5
790

1095
1018
979
892

450

300
0.0 0.0
20 μm
150 450 600 750 900 1050
EHT = 5.00 kV Signal A 800 1200 1600
Mag = 568 X WD = 2.
Raman shift (cm–1) (d) Energy/pulse (μJ)

FIGURE 11.14 FESEM images Ag targets with Bessel beams prepared with ~40 fs laser pulses at energies
(a) 600 µJ, (b) 600 µJ, and (c) 800 µJ. Insets show SERS spectra in red of CL-20 (5 µM) from the laser-exposed
portions and (d) comparison of enhancement factors for the Ag targets prepared with Bessel beams with ~2
ps and ~40 fs pulses. Time of integration was 0.5 s for all the cases. (Reproduced from Podagatlapalli, G.K.
et al., Laser Phys. Lett., 12, 036003, 2015. © Astro Ltd. With permission.)

and insets present their respective images with lower resolution (2 μm scale). The surface morphology
of multiply ablated Cu substrates MCuNS1 (Figure 11.16a) and MCuNS3 (Figure 11.16c) revealed the
formation of weakly distinguishable laser-induced periodic surface structures (LIPSS) with a period
of ~400 nm along with cylindrical Cu NP grains (length ~300 nm, diameter ~20 nm) and the forma-
tion of LIPSS with a periodicity 250–300 nm, respectively. In the case of MCuNS2 (Figure 11.16b),
the surface topography of the Cu substrate consisted of pillar-shaped structures of a few micrometers.
Moreover, most of the cubic NPs with the size of 200 nm were formed on top of the pillar. Similarly
in the case of MCuNS4 (Figure 11.16d), rough-ended pillars with sizes of a few micrometers were
formed on the Cu surface unlike in the case of MCuNS2. Figure 11.17 depicts the FESEM image
morphologies of fabricated NSs on the surface of Cu substrates when ablation was carried out
in (1) acetone (SCuNS1), (2) DCM (SCuNS2), (3) ACN (SCuNS3), and (4) chloroform (SCuNS4)
via single line ablation, and insets of Figure 11.17 depict their respective images with lower resolu-
tion (2 μm scale). We estimated that a lesser number of pulses (~500) per spot were incident on the
surface compared to the multiple ablation case (~4000). LIPSS with periods of ~300 and ~400 nm
for SCuNP1 (Figure 11.17a) and SCuNP3 (Figure 11.17c), respectively, were observed. In addition to
this, spherical NP grains (with NP size of ~50 nm) were formed on the periodic structure surfaces
in the case of SCuNP1, which was not observed in SCuNP3 case. Similarly, randomly corrugated
392 Semiconductor Nanocrystals and Metal Nanoparticles

20,000 24,000

787.38

701.6
20,000
16,000

283.4

739.3
16,000

1430.97
12,000

531.04

1066.97

1711.5
1223.6
391.14
12,000
8,000
8,000
4,000
4,000
Raman intensity (a.u.)

0 0
(a) 0 500 1000 1500 (b) 0 250 500 750 1000 1250 1500 1750

70,000
16,000

665.9
364.5
919.53

14,000 60,000
379.05

258.5
1374.92

12,000 50,000
10,000 40,000

754.6
8,000

1215.9
30,000
6,000
4,000 20,000

2,000 10,000
0 0
(c) 0 500 1000 1500 (d) 0 500 1000 1500

Raman shift (cm–1)

FIGURE 11.15 Raman spectra of (a) MCuNP1 (blue/top), SCuNP1 (wine/middle), and pure acetone (red/
bottom); (b) MCuNP2 (blue/top), SCuNP2 (wine/middle), and pure DCM (red/bottom); (c) MCuNP3 (blue/
top), SCuNP3 (wine/middle), and pure ACN (red/bottom); and (d) MCuNP4 (blue/top), SCuNP4 (wine/mid-
dle), and pure chloroform (red/bottom) recorded with an excitation wavelength of 785 nm and an integration
time of 5 s. (Reproduced with permission from Hamad, S., Podagatlapalli, G.K., Tewari, S.P., and Venugopal
Rao, S., Influence of picosecond multiple/single line ablation on copper nanoparticles fabricated for surface
enhanced Raman spectroscopy and photonics applications, J. Phys. D Appl. Phys., 46, 485501. © 2013 by the
IOP Publishing.)

structures along with NP grains and less number of triangular NPs (~250 nm size) were observed for
the cases SCuNP2 (Figure 11.17b) and SCuNP4 (Figure 11.17d), respectively (Hamad et al. 2014).
The observed ripples could be the resultant of the interference of the incident electromagnetic
wave and surface-scattered electromagnetic wave (SEW) as proposed by earlier groups working on
LIPSS. The periodicity of the ripples is related to the wavelength and angle of incidence of the laser
beam as in the following equation:
l
Ù^ =
l / l s ± sin q

where
∧⊥ is the period of ripples
λ is the incident wavelength of the laser
λ s is the wavelength of the surface wave

For a normal incidence, the periodicity of the ripples is nearly equal to the λ of SEW.
Applications of Metal Nanoparticles and Nanostructures 393

200 nm EHT = 5.00 kV Signal A = InLens Date: 6 Feb 2013 200 nm EHT = 5.00 kV Signal A = InLens Date: 6 Feb 2013
Mag = 50.00 KX WD = 4.4 mm Time: 12:20:59 Mag = 50.00 KX WD = 4.5 mm Time: 12:05:42

(a) (b)

200 nm EHT = 5.00 kV Signal A = InLens Date: 6 Feb 2013 200 nm EHT = 5.00 kV Signal A = InLens Date: 6 Feb 2013
Mag = 50.00 KX WD = 4.9 mm Time: 11:55:03 Mag = 50.00 KX WD = 4.6 mm Time: 12:29:41

(c) (d)

FIGURE 11.16 FESEM images of substrates (a) MCuNS1 in acetone, (b) MCuNS2 in DCM, (c) MCuNS3 in
ACN, and (d) MCuNS4 in CHCl3 fabricated through ultrafast multiple line ablation of Cu. Insets show their
corresponding views on a larger scale (2 μm scale). (Reprinted with permission from Hamad, S., Podagatlapalli,
G.K., Mohiddon, M.A., and Venugopal Rao, S., Cost effective nanostructured copper substrates prepared with
ultrafast laser pulses for explosives detection using surface enhanced Raman scattering, Appl. Phys. Lett., 104,
263104. © 2014, American Institute of Physics.)

Raman spectra of various analytes (ANTA, TNT, and R6G of concentrations 10 –4 and 10 –6 M)
were recorded using a micro-Raman spectrometer using an excitation source at 532 nm and a
100× microscope objective. Typically, a tiny drop (volume of ~10 μL) of ANTA was placed on all
of the 8 Cu NS substrates to achieve a monolayer of the analyte molecules, and they are Cu LIPSS
+ cylindrical NPs (MCuNS1), pillar-shaped structures + cubical NPs (MCuNS2), LIPSS with
250 nm periodicity (MCuNS3), rough-ended pillars (MCuNS4), LIPSS + spherical NP grains
(SCuNS1), randomly corrugated structures + spherical NP grains (SCuNS2), LIPSS with 400 nm
periodicity (SCuNS3), and small number of triangular NPs with ~250 nm size (SCuNS4). Raman
spectra were typically recorded for four times at various positions on the target, and their average
spectra were considered. Later on, utilized Cu substrates were cleaned and sonicated with ACN,
and again the same substrates were utilized for the second time to record the Raman spectra of
R6G. Subsequently, proper cleaning procedures were again followed, and the same substrates
394 Semiconductor Nanocrystals and Metal Nanoparticles

200 nm EHT = 5.00 kV Signal A = InLens Date: 6 Feb 2013 200 nm EHT = 5.00 kV Signal A = InLens Date: 6 Feb 2013
Mag = 50.00 KX WD = 4.0 mm Time: 13:58:53 Mag = 50.00 KX WD = 5.3 mm Time: 14:32:51

(a) (b)

200 nm EHT = 5.00 kV Signal A = InLens Date: 6 Feb 2013 200 nm EHT = 5.00 kV Signal A = InLens Date: 6 Feb 2013
Mag = 50.00 KX WD = 4.6 mm Time: 12:39:22 Mag = 50.00 KX WD = 4.5 mm Time: 14:17:05

(c) (d)

FIGURE 11.17 FESEM images of substrates (a) SCuNS1 in acetone, (b) SCuNS2 in DCM, (c) SCuNS3 in
ACN, and (d) SCuNS4 in chloroform fabricated through ultrafast single line ablation of Cu. Insets show cor-
responding views on a larger scale (2 μm scale). (Reprinted with permission from Hamad, S., Podagatlapalli,
G.K., Mohiddon, M.A., and Venugopal Rao, S., Cost effective nanostructured copper substrates prepared with
ultrafast laser pulses for explosives detection using surface enhanced Raman scattering, Appl. Phys. Lett., 104,
263104. © 2014, American Institute of Physics.)

were once more utilized, for the third time, to record the Raman spectra of TNT. The time sepa-
ration between these three trials was approximately few weeks. Raman spectra of ANTA from
Cu NSs fabricated via ps multiple and single line ablation are shown in Figure 11.18a and b.
Significantly observed modes from the data were C–NO2 symmetric stretching mode (1341 cm−1),
N–N symmetric stretching mode (1130 cm−1), NO2 deformation + ring deformation (840 cm−1),
ring torsion + NO2 deformation (589 cm−1), C–N symmetric stretch mode (1480 cm−1), N4–C5–N1
bending mode (960 cm−1), ring deformation mode (1070 cm−1), and ring deformation + N–H bend
(1303 cm−1). The Raman spectra of ANTA from Cu substrates were compared with the normal
Raman spectra of ANTA obtained using a Si target to estimate the EFs for 1341 cm−1 mode. The
intensity for 1341 cm−1 mode for MCuNP1 (LIPSS and cylindrical NPs) was ~5301 (arbitrary
units). However, the normal Raman intensity of ANTA from Si target for the same mode recorded
Applications of Metal Nanoparticles and Nanostructures 395

2000 14,000

1340 cm–1
MCuNS1
MCuNS2 12,000
1800 MCuNS3
10,000
Raman intensity (a.u.)
MCuNS4

1130 cm–1
1600

1067 cm–1
8,000

840 cm–1
960 cm–1
6,000
1400

589 cm–1
4,000
1200
2,000

1000 0
500 1000 1500 2000 2500

(a) Raman shift (cm–1)

2400 1340 cm–1 SCuNS1

SCuNS2
SCuNS3
1130 cm–1

2100
SCuNS4
Raman intensity (a.u.)

1800
1459 cm–1
1067 cm–1

1500
cm–1
960 cm–1
589 cm–1

844

1200

900

500 1000 1500 2000

(b) Raman shift (cm–1)

FIGURE 11.18 SERS spectra recorded from ANTA molecule adsorbed on different (a) multiple line and
(b) single line ablated Cu NSs. MCuNS1 (SCuNS1)—red/short dotted line, MCuNS2 (SCuNS2)—blue/short
dashed line, MCuNS3 (SCuNS3)—orange/short dashed line, MCuNS4 (SCuNS4)—olive/dashed dotted dot-
ted line. The integration time was 5 s (2 μm scale). (Reprinted with permission from Hamad, S., Podagatlapalli,
G.K., Mohiddon, M.A., and Venugopal Rao, S., Cost effective nanostructured copper substrates prepared with
ultrafast laser pulses for explosives detection using surface enhanced Raman scattering, Appl. Phys. Lett., 104,
263104. © 2014, American Institute of Physics.)

with the same experimental conditions was ~10 (arbitrary units). The intensity enhancement
(ISERS/IRaman), therefore, was estimated to be ~530. We estimated that ~50% of analyte molecules
were adsorbed on the Cu substrates (the adsorption coefficient (η) used was 0.5). EFs calculated
for the 1340 cm−1 mode of ANTA molecule were 1.2 × 106, 3.1 × 104, 7.3 × 104, 7.1 × 104, 1.4 × 105,
3.6 × 104, 7.1 × 104, and 7 × 10 4 for MCuNS1, MCuNS2, MCuNS3, MCuNS4, SCuNS1, SCuNS2,
SCuNS3, and SCuNS4, respectively. In the case of ANTA, large Raman signal enhancements
were observed from MCuNS1 and SCuNS1. In particular, in the case of MCuNS1 (SCuNS1),
nanocylindrical grains (NP grains) were observed on the top of LIPSS. Consequently, these sub-
strates could have provided high local fields, which resulted in the observed strong enhancements
(Hamad et al. 2014). Similarly, the Raman spectra of R6G (data presented in Figure 11.19) from
396 Semiconductor Nanocrystals and Metal Nanoparticles

1648 cm–1
1363 cm–1
MCuNS1
10,000

1507 cm–1
MCuNS2

1572 cm–1
MCuNS3

1180 cm–1
MCuNS4

1305 cm–1
611 cm–1
8,000 R6G on glass
Raman intensity (a.u.)

772 cm–1
6,000

4,000

2,000

0
500 1000 1500
(a) Raman shift (cm–1)

2500
SCuNS1
SCuNS2
SCuNS3
1363 cm–1

2000 SCuNS4
1305 cm–1

1572 cm–1
–1
1507 cm–1
611 cm–1

1180 cm–1
–1

1648 cm
R6G on glass
772 cm
Raman intensity (a.u.)

1500

1000

500

0
500 1000 1500
–1
(b) Raman shift (cm )

FIGURE 11.19 SERS spectra recorded from R6G molecule adsorbed on different (a) multiple line and
(b) single line ablated Cu NSs. MCuNS1 (SCuNS1)—red/short dotted line, MCuNS2 (SCuNS2)—blue/short
dashed line, MCuNS3 (SCuNS3)—orange/dashed dotted dotted line, MCuNS4 (SCuNS4)—olive/dashed dot-
ted line, and R6G on glass—black solid lines. The integration time was 5 s. (Reprinted with permission from
Hamad, S., Podagatlapalli, G.K., Mohiddon, M.A., and Venugopal Rao, S., Cost effective nanostructured
copper substrates prepared with ultrafast laser pulses for explosives detection using surface enhanced Raman
scattering, Appl. Phys. Lett., 104, 263104. © 2014, American Institute of Physics.)
Applications of Metal Nanoparticles and Nanostructures 397

the cleaned multiply and singly ablated Cu substrates recorded for the second time were analyzed.
The prominent enhanced modes of R6G were the aromatic C–C stretch, C–H out-plane bend,
and C–C–C ring in-plane bend at 1363, 772, and 610 cm−1, respectively. Evaluation of EFs was
carried out by comparing reference spectra of R6G (0.25 M concentration) recorded from a cover
slip. At low concentrations (~5 μM) of R6G, it is believed that 43% of molecules (η ~ 0.43) were
adsorbed on Cu NS substrates. Estimated EFs of R6G for the 1363 cm−1 mode compared with the
same mode in normal Raman spectra were 6.2 × 105, 9.1 × 106, 2.6 × 106, 1.8 × 107, 4.6 × 105, 1.1
× 106, 9.1 × 105, and 3.8 × 105 for MCuNS1, MCuNS2, MCuNS3, MCuNS4, SCuNS1, SCuNS2,
SCuNS3, and SCuNS4, respectively. Among the substrates investigated, MCuNS2, whose surface
topography comprised of μm sized pillars, and MCuNS4, containing roughened pillars along
with NP grains, demonstrated stronger enhancements (Hamad et al. 2014).
Figure 11.20 illustrates the Raman spectra of the third testing sample TNT, recorded from 8 Cu
targets subsequent to appropriate cleaning. The recorded Raman spectra revealed that NO2 sym-
metrical stretch mode corresponding to 1362 cm−1 was predominantly elevated. Additionally, two
more peaks were observed: one at 1616 cm−1, which corresponds to C–C aromatic stretching vibra-
tion, and another mode at 790 cm−1 (C–H out of plane bend). Corresponding EFs calculated for the
1362 cm−1 mode were ~4.2 × 104, ~7.7 × 104, ~2.2 × 105, ~1.6 × 104, ~2.5 × 105, ~5.8 × 104, ~1.9 × 105,
and ~5.2 × 104 for MCuNS1, MCuNS2, MCuNS3, MCuNS4, SCuNS1, SCuNS2, SCuNS3, and
SCuNS4, respectively. Raman spectra collected from three analytes using the same Cu targets
demonstrated a small reduction in enhancements from analyte 1 to analyte 3 (Hamad et al. 2014).
However, the enhancements were reasonably large even in the third trial. Our experimental data
demonstrated that nanostructured Cu targets possess potential for elevating the Raman signatures
of diverse analytes. We believe that three reasons could possibly contribute to the Raman signal
enhancement: (1) The coupling of incident photons to localized and propagating surface plasmons
leads to the elevation of evanescent fields (these evanescent fields influence the individual vibrating
dipoles to reradiate more number of photons, thus enhancing the Raman signal), (2) the aforemen-
tioned coupling of incident photons to LSPR might have fulfilled the condition that the dimensions
and shape of the nanomaterials allowed for perpendicular surface plasmon vibrations with respect
to the planar surface of the target, (3) whenever the separation between the nanodimensional mate-
rials (NPs or nanotips) is optimized, then combined electron resonances of individual structures
enhance the probability of coupling, and hence enormous amounts of evanescent fields are possible
and sometimes specifically compatible to certain modes of interest. Consequently, some substrates
demonstrated better Raman signature enhancement for a specific molecule. The estimated EFs in
our experiments were found to be superior compared to EFs obtained with Cu targets prepared by
other methods.

11.6.4 Cu Targets Ablated in Chloroform

In these experiments, the Cu metal target was placed at the bottom of a Pyrex cell filled with liquid
of ~5 mm thickness above the surface of metal target. The spot size on the sample surface was
estimated to be ~90 μm. Typical fluence of ~2.5 J/cm2 and an ablation time of 30 min were used.
The liquid-filled Pyrex cell with Cu target, placed on the X–Y stage (Newport), was translated to
draw periodic lines on Cu target at a separation of ~60 µm. Morphologies of the fabricated Cu NSs
in CHCl3 were characterized by FESEM, and the data are presented in Figure 11.21a. The FESEM
data demonstrated formation of micrometer-sized structures with sharp edges and magnified image
depicted random NSs with roughness of ~700 nm. The surface activity of Cu NS was investigated by
recording the Raman spectra of three adsorbed molecules (R6G [1 × 10 –9 and 5 × 10 –7 M], ANTA
[1 × 10 –6 and 2 × 10 –4 M], FOX-7 [1 × 10 –6 and 2 × 10 –4 M]) on the Cu NS. The Raman spectra
398 Semiconductor Nanocrystals and Metal Nanoparticles

1362 cm–1
1600
MCuNS1

1536 cm–1
1620 cm–1
MCuNS2
1400 MCuNS3
MCuNS4

1209 cm–1
1200

1086 cm–1
Raman intensity (a.u.)

825 cm–1
1000

800

600

400

200

500 1000 1500 2000

(a) Raman shift (cm–1)

2000
cm–1

SCuNS1
1209 cm–1

SCuNS2
1362

SCuNS3
1600
SCuNS4
1536 cm–1
1620 cm–1
Raman intensity (a.u.)

825 cm–1

1082 cm–1

1200

800

400

0
500 1000 1500 2000
(b) Raman shift (cm–1)

FIGURE 11.20 SERS spectra recorded from TNT molecule adsorbed on different (a) multiple line and
(b) single line ablated Cu NSs. MCuNS1 (SCuNS1)—red/short dotted line, MCuNS2 (SCuNS2)—blue/short
dashed line, MCuNS3 (SCuNS3)—orange/short dashed dotted line, MCuNS4 (SCuNS4)—olive/solid line.
The integration time was 5 s. (Reprinted with permission from Hamad, S., Podagatlapalli, G.K., Mohiddon,
M.A., and Venugopal Rao, S., Cost effective nanostructured copper substrates prepared with ultrafast laser
pulses for explosives detection using surface enhanced Raman scattering, Appl. Phys. Lett., 104, 263104.
© 2014, American Institute of Physics.)
Applications of Metal Nanoparticles and Nanostructures 399

(a)

2 μm EHT = 5.00 kV Signal A = InLens Date: 7 Feb 2012

ZEISS
Mag = 6.65 KX WD = 5.4 mm Time: 17:30:32

1500

1648
1507
1363

(i) R6G (5 × 10–7 M)

1572
1000
1305
611

1180
772

500
1337

0
Raman intensity (a.u.)

(b)
800
(ii) ANTA (2 × 10–4 M)
1126
844
955

400

6000 (iii) FOX-7 (2×10–4 M)

1209

1340

4000
860

1167
1026

1520

2000
1483
624

1065
483

1605
791

0
400 800 1200 1600
Raman shift (cm–1)

FIGURE 11.21 (a) FESEM image of NSs on the Cu substrate prepared in chloroform and (b) SERS spectra
recorded from (i) R6G (blue/top—5 × 10 –7 M), (ii) ANTA (wine/middle—2 × 10 –4 M), and (iii) FOX-7 (red/
bottom—2 × 10 –4 M) adsorbed on Cu NSs. (Reprinted from Chem. Phys. Lett., 621, Hamad, S., Podagatlapalli,
G.K., Mohiddon, M.A., and Venugopal Rao, S., Surface enhanced fluorescence from corroles and SERS stud-
ies of explosives using copper nanostructures, 171–176. © 2015, with permission from Elsevier.)
400 Semiconductor Nanocrystals and Metal Nanoparticles

were recorded at 532 nm with a 0.9 NA objective. SERS spectra of R6G (1 × 10 –9 and 5 × 10 –7 M)
were recorded with 10–20 µL deposited on Cu NS to form a monolayer. An average of 10 spectra
collected is illustrated in Figure 11.21b. EFs were evaluated by comparing the SERS spectra with
the normal Raman spectra of R6G that were deposited on a silicon surface for the most prominent
Raman peak of 1363 cm−1. In case of R6G, at 1 × 10 –9 and 5 × 10 –7 M concentrations, the absorbed
molecules were estimated to be ~35% and ~43%. The estimated EFs for 1363 cm−1 mode were ~6 × 108
and ~1.8 × 107 for 1 × 10 –9 and 5 × 10 –7 M concentrations, respectively. Large enhancement in the
case of R6G could be due to the rough edges arising from micrometer-sized structures on the Cu
surface. Later, the substrate was cleaned by sonication in acetone for half an hour and dried at
room temperature. After 2 weeks, the same Cu substrate was utilized again by recording the SERS
spectra of an explosive molecule (ANTA) at 1 × 10 –6 and 2 × 10 –4 M concentrations. SERS spectra
of the ANTA molecule adsorbed on the Cu NSs substrate are shown in Figure 11.21b. The main
characteristic mode was detected at 1337 cm−1 with lower intensity along with the 1577 cm−1 mode
assigned to C–NO2 symmetrical stretch and C–NH2 symmetrical stretch + NH2 bend, respectively.
For 2 × 10 –4 M concentration, the characteristic mode at 1337 cm−1 was elevated with good signal
to noise ratio, and some more modes were detected at 844, 955, and 1126 cm−1, which were not
observed in SERS spectra at 1 × 10 –6 M concentration. We estimated that ~47% and ~50% ANTA
molecules were adsorbed on the NS substrate for the concentrations of 1 × 10 –6 and 2 × 10 –4 M,
respectively. The evaluated EFs for the 1337 cm−1 mode were ~1.4 × 107 and ~2.8 × 105 at 1 × 10 –6
and 2 × 10 –4 M concentrations, respectively. One day later, the SERS spectra of another explosive
molecule (FOX-7) were recorded for the third time after appropriate cleaning of the substrate. SERS
spectra of FOX-7 (Figure 11.21b) demonstrated the characteristic Raman bands corresponding to
symmetric C–NO2 stretching and NH wagging at 1340 cm−1 with lower intensity at a concentra-
tion of 1 × 10 –6 M, whereas at 2 × 10 –4 M concentration, the mode intensity was clearly enhanced.
Moreover, the Raman mode at 1483 cm−1 was observed in both the cases. The calculated EFs for
the 1340 cm−1 mode were ~2.5 × 105 and ~4.2 × 104 for 1 × 10 –6 and 2 × 10 –4 M concentrations,
respectively (Hamad et al. 2015).

11.6.5 Cu Targets Ablated with Corroles and Surface-Enhanced

Fluorescence/SERS Studies
In this study, Cu targets were ablated in chloroform and in diluted triphenyl corrole (TPC,
2.5 × 10 –5 M) and tri tolyl corrole (TTC, 2.5 × 10 –5 M) solutions to fabricate chlorinated Cu NPs
and suspensions of chlorinated Cu NPs in corroles, respectively. Briefly, the Cu target was placed at
the bottom of the Pyrex cell filled with liquid. Typical height of the liquid above the surface of metal
target was ~5 mm. The spot size on the sample surface was estimated to be ~90 μm. Typical fluence
of ~2.5 J/cm2 and an ablation time of 30 min were used. The liquid-filled Pyrex cell with Cu target
was placed on a motorized X–Y stage, which was controlled by an ESP 300 motion controller. The
stages (Newport) were translated to draw periodic lines on Cu target at a separation of ~60 µm. To
avoid ambiguity, the fabricated Cu colloids in TTC, TPC, and pure chloroform are labeled as NP-1,
NP-2, and NP-3, respectively.
Figure 11.22a and c illustrates the TEM images of NP-1 and NP-2, and insets show the size
distribution. It was observed that the morphologies of fabricated NPs were spherical in nature.
Following a Gaussian fit of the size distribution histogram average sizes of NPs were found to be
~9.5 and ~38 nm in NP-1 and NP-2, respectively. Figure 11.22b and d illustrates the SAED spectrum
of NP-1 and NP-2, respectively. The shape of the SAED pattern revealed that the particles were
­nanocrystalline in structure and the measured interplanar separations (d) for NP-1 were 3.2 and 2.6 Å,
in agreement with those of copper chloride (CuCl) crystal planes [(200) and (211)] and 2.2, 1.82,
Applications of Metal Nanoparticles and Nanostructures 401

Avg. size = 9.5 nm

Std. div = 1.1 nm

Abundance
(220)
0 5 10 15 20 25 Cu (200)
Size (nm)
(211)

CuCl (200)

(111)

100 nm (a) (b)

5 1/nm
Avg. size = 38 nm
Std. div = 9 nm
Abundance

10 20 30 40 50 60 70 80
(200)
Size (nm) (123) (111)
(110)
(210)

(220)

200 nm (c) (d)

5 1/nm

Cu (220)
Cu (200)
Cu (110)
CuCl (110)

Avg. size = 11 nm
Std. div = 4.4 nm
Abundance

5 10 15 20 50 nm (e) 5 1/nm (f)

Size (nm)

FIGURE 11.22 TEM micrographs of corrole/Cu colloids prepared in liquid media: (a) NP1 in TTC,
(c) NP2 in TPC, and (e) NP3 in chloroform. Insets of (a), (c), and (e) illustrate the size distribution (b), (d),
and (f) SAED patterns of NP1 in TTC, NP2 in TPC, and NP3 in chloroform. (Reprinted from Chem. Phys.
Lett., 621, Hamad, S., Podagatlapalli, G.K., Mohiddon, M.A., and Venugopal Rao, S., Surface enhanced
­fluorescence from corroles and SERS studies of explosives using copper nanostructures, 171–176. © 2015,
with permission from Elsevier.)
402 Semiconductor Nanocrystals and Metal Nanoparticles

and 1.28 Å matching with Cu (111), (200), and (220) planes. In the case of NP-2, “d” values were 4.6,
2.8, and 1.7 Å in good agreement with CuCl planes of (110), (210), and (123) and the atomic spacing
values of 2.19, 1.81, and 1.36 Å in good agreement with Cu crystal planes of (111), (200), and (220).
SAED results revealed that Cu NPs and CuCl NPs were formed in the colloidal solutions (Hamad
et al. 2015). Similarly, from the TEM images of NP-3 presented in Figure 11.22e, it is evident
that a majority of produced NPs were smaller in size with a few NPs with large dimensions. The
mean diameter of small particles, evaluated by Gaussian fit, was ~11 nm (inset of Figure 11.22e).
Similarly, the estimated average size of NPs with larger dimensions was ~25 nm. Figure 11.22f
illustrates the SAED pattern of NP-3 and it is evident that NP-3 phase was nanocrystalline. The ring
pattern in SAED was indexed with cubic phase of CuCl crystal planes, which were confirmed from
the measured interplanar spacing (estimated from the diameter of rings in SAED image). The preva-
lence of CuCl (110) with “d” spacing of 4.5 Å and Cu (111), (200), and (220), with corresponding
“d” spacing values of 2.18, 1.81, and 1.29 Å, respectively, confirmed the formation of cubic phase
for both Cu and CuCl NPs.
Surface-enhanced fluorescence (SEF) spectra have been recorded using micro-Raman spec-
trometer with a 532 nm excitation source, and spectra were collected through a charge couple
device (CCD) camera. The laser beam was focused on analytes placed on Cu substrates using
microscopic objective (100×, NA = 0.9), and backscattered signal was collected through the
same objective. Two types of SEF measurements have been performed. A tiny drop (~10 μL)
of (1) manually mixed corroles solution (1 nM) and chlorinated Cu NPs and (2) colloidal cor-
role Cu NPs were deposited on the cover slips. Three SEF spectra were recorded at different
positions on the substrate, and an average of the three spectra obtained was considered for
analysis. Typical acquisition time used was 5 s. Figure 11.23 demonstrates the SEF spectra from
(1) NP-1 and (2) NP-2. The enhancements were estimated by comparing with characteristic
bands obtained in normal fluorescence spectra (NFS) of TTC (100 µM) and TPC (100 µM) at
higher concentrations. Comparison of SEF and NFS of both TTC and TPC revealed that the
intensity of fluorescence band in SEF spectra was enhanced by 13 and 333 times, respectively.
Evidently, similar bands appeared in SEF (622 nm for TTC and 646 nm for TPC) spectrum and
NFS as well. We did not observe any visible peak shift in SEF spectra compared to the peaks
in NFS. A tiny drop (~10 μL) of manually mixed solution of corroles (1 × 10 –9 M) and Cu/CuCl
NPs in chloroform was placed on the glass slide and dried. Recorded SEF spectra of these dried
samples are shown in Figure 11.23c and d. The spectra illustrated very high intensity at 626 and
645 nm for TTC and TPC, respectively. SEF spectra of lower concentration (10 –9 M) analytes
(TTC and TPC) were compared with the higher concentration (1 × 10 –4 M) spectra to estimate
the EFs and were found to be ~2.7 × 105 and ~7.1 × 105 for TTC and TPC, respectively (Hamad
et al. 2015). Fluorescence enhancements in the present case, possibly, occurred due to (1) elec-
tromagnetic mechanism that motivates the amplification of incident field through the excitation
of surface plasmons and (2) increase in quantum yield and radiative decay rates. In the case of
TTC–Cu–Cl and TPC–Cu–Cl colloidal solutions, fluorophores (TTC and TPC) were strongly
adsorbed on metal surfaces inciting quenching of fluorescence (Hamad et al. 2015). Moreover,
there could be reabsorption of the fluorescence emission band through the attached Cu NPs
whose SPR band falls in the range of fluorescence band of fluorophore. Due to the reabsorption,
fluorescence might have quenched significantly. In the second case, fluorophores (TTC and
TPC) might not completely sit on the Cu NPs surface since it was a passive mixing of Cu NPs
to TTC and TPC. In this case, due to the local field effect of Cu NPs that were in the proximity
of fluorophores, significant fluorescence enhancement was observed with very minimal quench-
ing of fluorescence. Furthermore, reabsorption of fluorescence emission band by the Cu NPs
in the neighborhood was possibly absent since their surface was passivated by the CuCl layer
(Hamad et al. 2015).
Applications of Metal Nanoparticles and Nanostructures 403

12,000
TTC Cu colloids (NP1) TPC Cu colloids (NP2)
TTC (100 μM) TPC (100 μM)
TTC (25 μM)
6000 TPC (25 μM)
9,000
622 nm
4000
6,000 646 nm

3,000 2000

0 0
Intensity (a.u.)

(a) 550 600 650 700 750 (b) 550 600 650 700 750

TTC (100 μM) TPC (100 μM)

8000 NP-3 + TTC (1 nM) 16,000 NP-3 + TPC (1 nM)

6000 12,000
645 nm

4000 626 nm 8,000

2000 4,000

0 0
(c) 550 600 650 700 750 800 (d) 550 600 650 700 750 800
Wavelength (nm)

FIGURE 11.23 SEF spectra recorded from (a) TTC Cu colloids (NP1) and (b) TPC Cu colloids (NP2). TTC
(100 µM) and TPC (25 µM) spectra are represented by orange/dotted line and black/dashed line, respectively.
The spectrum of NP1 (NP2) is represented by blue/solid line. SEF spectra recorded from (c) TTC and (d)
TPC corroles adsorbed on NP3 individually. TTC (TPC) spectra are represented by red/dotted lines and TTC/
TPC + NP3 by blue/solid lines. The integration time used was 5 s. (Reprinted from Chem. Phys. Lett., 621,
Hamad, S., Podagatlapalli, G.K., Mohiddon, M.A., and Venugopal Rao, S., Surface enhanced fluorescence
from ­corroles and SERS studies of explosives using copper nanostructures, 171–176. © 2015, with permis-
sion from Elsevier.)

11.7 NONLINEAR OPTICAL STUDIES OF Cu, Ag, AND Al COLLOIDS

11.7.1 NLO Properties of Cu Colloids
The third-order NLO properties of the Cu colloids (Cu NPs in acetone, CuCl NPs in DCM, CuO
NPs in ACN, and CuCl2 NPs in chloroform) were investigated at a wavelength of 800 nm with
~2 ps pulses using the standard Z-scan technique. These Cu colloids were generated by multiple
line (single line) ablation of Cu in acetone, DCM, ACN, and chloroform, which were designated as
MCuNP1 (SCuNP1), MCuNP2 (SCuNP2), MCuNP3 (SCuNP3), and MCuNP4 (SCuNP4), respec-
tively. Open-aperture Z-scan data provide information about the nonlinear absorption (related to
the imaginary part of third-order NLO susceptibility, χ(3)) properties, while the closed-aperture
(CA) Z-scan data provide information on sign and magnitude of the nonlinear refractive index n2
(related to the real part of χ(3)). The linear transmittance (LT) was typically >95% for all NPs except
MCuNP1, MCuNP3 (>80%).
404 Semiconductor Nanocrystals and Metal Nanoparticles

Figure 11.24 depicts the complete open-aperture data obtained for all investigated 8 NPs at both
lower (open squares—80 GW/cm2) and higher (open stars—125 GW/cm2) peak intensities. Three
distinct behaviors were observed: (1) reverse saturable absorption (RSA) type of behavior that could
be attributed to two-photon absorption (2PA; β), (2) an effective three-photon absorption (3PA; γeff )
type of behavior, and (3) a switching mechanism from saturable absorption (SA) to RSA. The open-
aperture data of MCuNP1 (Figure 11.24a) and MCuNP3 (Figure 11.24e) illustrate the behavior of
switching from SA to RSA at lower peak (open squares—80 GW/cm2) intensity and pure RSA at

1.01 1.00
(a)
1.00 (c)
0.96
0.99
Norm. transmittance

0.98 0.92

1.01 1.01

1.00 1.00

(b) 0.99
0.99 (d)
0.98
0.98
0.97
–40 –20 0 20 40 –40 –20 0 20 40

1.01
1.00
(e) 0.95
(g)
1.00 0.90
Norm. transmittance

0.85
0.80
0.99

1.01 1.00
(f)
0.95 (h)
1.00
0.90
0.99
0.85
–40 –20 0 20 40 –40 –20 0 20 40
Z (mm)

FIGURE 11.24 Open-aperture Z-scan curves obtained for (a) MCuNP1, (b) SCuNP1 with varying input
intensities I00 = 80 GW/cm2 (open squares), I00 = 125 GW/cm2 (open stars). Open-aperture Z-scan curves
obtained for (c) MCuNP2, (d) SCuNP2 with varying input intensities I00 = 80 GW/cm2 (open squares), I00 =
125 GW/cm2 (open stars). Open-aperture Z-scan curves obtained for (e) MCuNP3, (f) SCuNP3 with varying
input intensities I00 = 80 GW/cm2 (open squares), I00 = 125 GW/cm2 (open stars) and (g) MCuNP4, (h) SCuNP4
with varying input intensities I00 = 80 GW/cm2 (open squares), I00 = 125 GW/cm2 (open stars). Solid lines are
the theoretical fits. (Reproduced with permission from Hamad, S., Podagatlapalli, G.K., Tewari, S.P., and
Venugopal Rao, S., Influence of picosecond multiple/single line ablation on copper nanoparticles fabricated
for surface enhanced Raman spectroscopy and photonics applications, J. Phys. D: Appl. Phys., 46, 485501.
© 2013 by the IOP Publishing.)
Applications of Metal Nanoparticles and Nanostructures 405

higher peak intensity (open stars—125 GW/cm2). The data obtained at higher peak intensities were
fitted efficiently using only β. At lower peak intensity, SCuNP1 (Figure 11.24b) data were fitted to β,
while SCuNP3 (Figure 11.24f) did not show any NLO behavior. At higher peak intensity, both
SCuNP1 and SCuNP3 samples exhibited RSA type of behavior, and the data were fitted success-
fully using γeff (2PA + ESA) and β, respectively. The resonances of SCuNP1 (565 and 310 nm) and
SCuNP3 (588 nm) might have led to this observation. The open-aperture data of MCuNP2 (Figure
11.24c), SCuNP2 (Figure 11.24d), MCuNP4 (Figure 11.24g), and SCuNP4 (Figure 11.24h) revealed
that 2PA was dominant at both peak intensities with 800 nm excitation, which can be accredited to
the SPR band of nanocolloids formed at nearly 400 nm (Hamad et al. 2015).
Nonlinear absorption in plasmonic metal NPs is governed by the transitions of conduction/valence
band electrons and is explained as follows: (1) completely occupied “d” band to unoccupied conduc-
tion band and/or (2) within the conduction band (ground states to excited states) due to plasmon
resonances and (3) free carrier absorption from excited conduction band to high-lying states.
Figure 11.25 depicts a generic energy band diagram portraying various mechanisms of (1) SA: in the
cases of MCuNP1 and McuNP3, we observed residual absorption at 800 nm and, therefore, expected
saturation at lower peak intensities with probable transition from “d” band to the “p” (conduction)
band and is depicted by Ia in Figure 11.25. However, at higher peak intensities, one expects 2PA as
depicted by Ib with transitions from ground state conduction band (GSCB) to excited state conduc-
tion band (ESCB). Due to two-photon resonance (800 + 800 nm) between GSCB and ESCB, 2PA
dominates with negligible contribution from SA at higher peak intensities. (2) Instantaneous 2PA with
possible mechanisms is depicted in the second box marked as IIa (between “d” band and GSCB)
and IIb (between GSCB and ESCB) and (3) 2-step 3PA (2PA + ESA) at very high peak intensities
as depicted by processes in boxes IIIa and IIIb (excited state absorption induced from ESCB in this
case). The figure is generic in the sense that we have assumed a fixed gap between the states and dif-
ferent photon energies, whereas in reality, the band gaps could be different (including SPR positions)
for these eight different NPs with single energy photon excitation (1.55 eV for 800 nm ps photon).
The experimental data were fitted with appropriate equations to estimate nonlinear absorption

Higher-lying band σESA

IIIb
Ib IIb ESCB

β σESA

GSCB
IIIa

γe is due
σ0 (Is) Ia IIa to β + ESA
β

“d” state

FIGURE 11.25 A generalized energy band diagram of the observed CuNPs (pure Cu NPs, CuO NPs, CuCl
NPs, CuCl2 NPs) explaining various nonlinear absorption phenomena. ESCB refers to excited state conduc-
tion band and GSCB refers to ground state conduction band. (Reproduced with permission from Hamad, S.,
Podagatlapalli, G.K., Tewari, S.P., and Venugopal Rao, S., Influence of picosecond multiple/single line abla-
tion on copper nanoparticles fabricated for surface enhanced Raman spectroscopy and photonics applications,
J. Phys. D Appl. Phys., 46, 485501. © 2013 by the IOP Publishing.)
406 Semiconductor Nanocrystals and Metal Nanoparticles

coefficients. Among all, MCuNP4 exhibited largest β, n2, and χ(3) values (Hamad et al. 2013). The
values of nonlinear refractive index (n2) were obtained for all colloidal solutions at an intensity of 33
GW/cm2 using CA Z-scan method, and the data are shown in Figure 11.26. Figure 11.26a illustrates
the CA data of MCuNP1 (triangles) and SCuNP1 (circles), which exhibit positive nonlinearity (like sol-
vent) with n2 = 6 × 10–16 cm2/W, smaller than the solvent magnitude (n2 = 19 × 10 –16 cm2/W), and the
observed sign of n2 for MCuNP1 and SCuNP1 was negative. Figure 11.26b through d represent CA
data of MCuNP2 (SCuNP2), MCuNP3 (SCuNP3), and MCuNP4 (SCuNP4). The observed sign of
the nonlinearity was negative for MCuNP2 (SCuNP2) and positive for MCuNP3 (MCuNP3). The
data were fitted using standard equation to estimate the magnitudes of n2. The contribution from
solvents (DCM, ACN, and chloroform) (positive nonlinearity) was also identified. The order of mag-
nitude of n2 was found to be ~10 –15 cm2/W for all the colloids investigated, and the order of magnitude
for β was ~10 –11 cm/W (Hamad et al. 2013).

11.7.2 NLO Properties of Ag Colloids

Picosecond multiple/double line ablation of Ag targets in DDW was carried out to obtain Ag col-
loids. Ag colloids prepared in the multiple line ablation are labeled as Ag NP-1, whereas the colloids
prepared via double line ablation are labeled as Ag NP-2. Average sizes obtained for Ag NP-1 and Ag
NP-2 were ~13 and ~17 nm, respectively (Podagatlapalli et al. 2013). In both multiple line/double line
ablations, a broad plasmon band was observed, and even the absorbance of both colloids was similar.
When compared to single line ablation, multiple line/double line ablation provided large plasmon
band (which was evident from the recorded UV–Vis absorption spectra) widths and average sizes.
These Ag NPs were observed to have dark gray coloration compared to the other Ag NPs prepared

1.04 1.10

1.05
1.00
1.00

0.95
0.96
Norm. transmittance

–40 –20 0 20 40 –40 –20 0 20 40

(a) (b)

1.05 1.2

1.00 1.0

0.8
0.95

0.6
–40 –20 0 20 40 –40 –20 0 20 40
(c) (d)
Z (mm)

FIGURE 11.26 Closed-aperture Z-scan curves obtained for (a) MCuNP1 (open triangles), SCuNP1 (open
circles), (b) MCuNP2 (open triangles), SCuNP2 (open circles), (c) MCuNP3 (open triangles), SCuNP3
(open circles), and (d) MCuNP4 (open triangles), SCuNP4 (open circles). Closed-aperture studies were
performed at a peak intensity of 33 GW/cm2. Solid lines are the theoretical fits. (Reproduced with permission
from Hamad, S., Podagatlapalli, G.K., Tewari, S.P., and Venugopal Rao, S., Influence of picosecond multiple/
single line ablation on copper nanoparticles fabricated for surface enhanced Raman spectroscopy and photonics
applications, J. Phys. D: Appl. Phys., 46, 485501. © 2013 by the IOP Publishing.)
Applications of Metal Nanoparticles and Nanostructures 407

by double line/single line ablation. Open-aperture Z-scan was performed at peak intensities ~83 and
~138 GW/cm2, while the CA study was carried out at ~28 GW/cm2. Open-aperture data of NP-1
(LT ~90%) presented in Figure 11.27a demonstrated a switching behavior (from SA to RSA) with a
saturation intensity of ~6.5 × 107 W/cm2 at lower peak intensity (83 GW/cm2, β = 2 × 10 –12 cm/W)
represented by open circles in the plot. RSA was observed at higher peak intensities (138 GW/cm2)
with a strong 2PA coefficient (β = 4.5 × 10 –12 cm/W) whose data points are indicated by open stars.
The experimental data were fitted with standard equations. The CA data were fitted with theoretical
equations to extract n2. The CA study was carried out at low peak intensity (~28 GW/cm2) to get the
information on nonlinear refraction (NLR) of the sample. The CA data are shown in Figure 11.27b
and the measured n2 was ~3.4 × 10 –16 cm2/W. Figure 11.27c depicts the open-aperture data of Ag
NPs prepared by double line ablation (NP-2) (LT was ~92%), which signifies pure RSA both at lower
(β = 6.2 × 10 –12 cm/W) and higher (β = 1.78 × 10 –11 cm/W) peak intensities. The CA data for NP-2 are
shown in Figure 11.27d and n2 retrieved was ~4 × 10 –15 cm2/W from the fits of experimental data. The n2
value of the pure water was higher than the n2 of NP-2, indicating that the sign of n2 of NP-2 was nega-
tive. The measured χ(3) values for NP-1 and NP-2 were 2.3 × 10 –14 and 2.7 × 10 –13 e.s.u., respectively
(Podagatlapalli et al. 2013). The colloids obtained with different angles of incidence were studied

1.01

1.02
1.00
Norm. transmittance
Norm. transmittance

0.99 1.00

0.98
n2 = 3.4 × 10–16 cm2/W
0.98
I00 = 28 GW/cm2
0.97
–40 –20 0 20 40 –40 –20 0 20 40
(a) Z (mm) (b) Z (mm)

1.00 1.1 n2 = 4 × 10–15 cm2/W

Norm. transmittance

Norm. transmittance

I00 = 28 GW/cm2

0.98 1.0

0.96 0.9

0.8
0.94
–40 –20 0 20 40 –40 –20 0 20 40
(c) Z (mm) (d) Z (mm)

FIGURE 11.27 Open-aperture Z-scan curves obtained for (a) Ag colloids NP-1 with varying input intensi-
ties I00 = 83 GW/cm2 (open circles), I00 = 138 GW/cm2 (stars). Blue and red colors represent low intensities and
high intensities, respectively, (b) closed-aperture Z-scan curves obtained for Ag colloids NP-1 at peak inten-
sity 28 GW/cm2, (c) open-aperture Z-scan curves obtained for Ag colloids NP-2 with varying input intensities
I00 = 83 GW/cm2 (open circles), I00 = 138 GW/cm2 (stars), and (d) closed-aperture Z-scan curves obtained for
Ag colloids NP-2 at peak intensity 28 GW/cm2. Solid lines are the theoretical fits. (Reprinted with permission
from Podagatlapalli, G.K., Hamad, S., Sreedhar, S., Prasad, M.D., and Venugopal Rao, S., Silver nano-entities
through ultrafast double ablation in aqueous media for surface enhanced Raman scattering and photonics
applications, J. Appl. Phys., 113, 073106. © 2013, American Institute of Physics.)
408 Semiconductor Nanocrystals and Metal Nanoparticles

using the fs DFWM technique (at 800 nm), and the magnitude of third-order NLO susceptibility was
found to be 10 –14 e.s.u. (Podagatlapalli et al. 2014).

11.7.3 NLO Properties of Al Colloids

Al NPs were fabricated in oxygen-free polar and nonpolar liquids such as carbon tetrachloride
(CCl4) and chloroform (CHCl3) using ~40 fs laser pulses. Laser ablation of bulk Al target in aque-
ous media provided Al colloid NPs in CCl4 and CHCl3. Average sizes of Al NPs in CCl4 and CHCl3
(estimated using image-J software) were ~33 ± 4.5 and 15 ± 5 nm, respectively. The prepared Al
colloids were utilized to investigate NLO behavior using ~2 ps laser pulses with the Z-scan tech-
nique. The nonlinear absorption of the colloidal NPs was investigated using open-aperture Z-scan
method (Podagatlapalli et al. 2012). We could explain how the presence of Al NPs in solvents influ-
ences the NLO properties at different input energies. This investigation was performed by ~2 ps laser
at a repetition rate of 1 kHz and central wavelength at 800 nm as an excitation source. The colloidal
solution was placed in a quartz cuvette of ~1 mm path length in a sample holder located on a motor-
ized stage (Newport ILS 250 PP). Z-scan studies were performed by focusing the 4 mm diameter
input beam using an achromatic doublet (f = 20 cm). The beam waist at the focus was ~25 μm. Laser
beam was allowed to pass through the colloidal solution, and the transmitted light was observed
by a photodiode (SM1PD2A) along with lock-in amplifier (7265 DSP from Signal Recovery). The
translational stage and photodiode were interfaced to a personal computer.
Figure 11.28a through c depicts the open-aperture Z-scan data of Al colloids in CCl4 recorded
at three input peak intensities of 96, 130, and 190 GW/cm2, respectively. The optical properties of

1.4 1.6 1.4

β = 1.9 × 10–10 cm/W β = 3.3 × 10–10 cm/W
1.4 1.2 β = 2 × 10–9 cm/W
1.2 IS = 6.5 GW/cm2 IS = 5 GW/cm2
1.2 1.0
1.0 0.8
1.0
0.8 0.6
0.8 0.6 0.4
0.4 0.2
0.6
Norm. transmittance

–40–30–20–10 0 10 20 30 40 –40 –30–20 –10 0 10 20 30 40 –40–30–20–10 0 10 20 30 40

(a) (b) (c)

1.15
β = 4.9 × 10–11 cm/W β = 5 × 10–11 cm/W β = 6 . 3 × 10–11 cm/W
1.10 IS = 6 GW/cm2 1.0
1.0
1.05
L
J

1.00 0.9
0.9
0.95

0.90 0.8
–40 –30–20 –10 0 10 20 30 40 –40 –30–20 –10 0 10 20 30 40 –40–30–20–10 0 10 20 30 40
(d) (e) Z (mm) (f)

FIGURE 11.28 Open-aperture Z-scan curves obtained for Al colloidal-CCl4 NPs with varying input inten-
sities (a) I00 = 96 GW/cm 2, (b) I00 = 130 GW/cm 2, and (c) I00 = 190 GW/cm 2. Open-aperture Z-scan curves
obtained for Al colloidal-CCl4 NPs with varying input intensities (d) I00 = 69 GW/cm2, (e) I00 = 110 GW/cm2, and
(f) I00 = 140 GW/cm2. Solid lines are theoretical fits. (Reprinted from Chem. Phys. Lett., 530, Podagatlapalli,
G.K., Hamad, S., Sreedhar, S., Tewari, S.P., and Venugopal Rao, S., Fabrication and characterization of
aluminum nanostructures and nanoparticles obtained using femtosecond ablation technique, 93–97. © 2012,
with permission from Elsevier.)
Applications of Metal Nanoparticles and Nanostructures 409

metal NPs can be described using transitions of conduction electrons between the discrete energy
states in the quantum wells provided by the LSPRs of NPs. Large enhancements of the local field
over the surface of NPs can be obtained at the plasmon resonance frequency. At lower peak inten-
sities, the behavior was switching from SA to RSA (Figure 11.28a), and at higher peak intensi-
ties, the behavior switched completely to pure RSA (Figure 11.28c). Pure CCl4 did not show any
nonlinear absorption (NLA), but the presence of NPs (40%–50% LT) introduced NLA. The sum-
mary of obtained NLO coefficients for Al colloids in CCl4 is presented elsewhere (Podagatlapalli
et al. 2012). Similarly, Figure 11.28d through f demonstrates the open-aperture data of CHCl3 col-
loids (65% LT) recorded at the same experimental conditions as mentioned for Al colloids in CCl4,
and input peak intensities utilized were ~69, ~110, and ~140 GW/cm2, respectively. At lower peak
intensities, the behavior was switching type (SA to RSA) (Figure 11.28d), and at higher intensities
(Figure 11.28f), the behavior switched completely to RSA (pure RSA) with strong 2PA. The order
of magnitude of β was ~10 –9 cm/W for all the colloids investigated (Podagatlapalli et al. 2012). Our
future investigations will comprise evaluating the nonlinearity in these NPs using fs/ps pulses in
the entire visible and near-IR spectral regions. Furthermore, we will investigate the pure electronic
nonlinearities using low repetition rate ultrashort laser pulses.

11.8 ANTIBACTERIAL STUDIES OF Cu, Ag COLLOIDS

Three solvents, namely, acetone, DCM, and ACN were used for NPs fabrication. The effect of
CuNP1 (acetone), CuNP2 (DCM), and CuNP3 (ACN) on bacterial growth was assessed against
the gram-negative bacteria Escherichia coli using agar well diffusion and viability assay meth-
ods. Investigation of the antibacterial activity of different Cu NPs against E. coli, gram-nega-
tive bacteria, by agar well diffusion and viability assay methods was accomplished (Syed et al.
2012). Bacterial suspension was applied uniformly on the surface of Muller–Hinton agar plates.
Subsequently, agar wells were filled with 200 µL of three different Cu NPs and their solvents. The
plates were incubated at 37°C for 24 h, and later the diameter of inhibition zone was measured. In
viability assay method, E. coli culture was grown overnight in Luria–Bertani (LB) broth. Aliquots
of 100 µL of culture were subcultured in fresh LB media (2 mL) and treated with NPs and pure
solvents that were incubated for 3 h. Inoculums without NPs served as negative control in these
measurements. Hundred microliters from each test was taken out for plating it on LB agar plates.
The bacterial viability was observed visually, and bacterial colonies were scrutinized after 12 h of
incubation at 37°C.
Figure 11.29a and b illustrate the agar well diffusion method, exhibiting the zone of inhibition of
Cu NPs fabricated in different liquids and positive control, against E. coli (gram-negative bacteria).
Figure 11.29a demonstrates that CuNP1 and CuNP2 inhibit the growth of E. coli, gram-negative
bacteria, and the zones of inhibition were 25 and 28 mm, respectively. The effect of CuNP2 on
the inhibition zone was greater than CuNP1, which is consistent with the earlier published results
(Syed et al. 2012). Figure 11.29b shows that the zone of inhibition was maximum (35 mm) when
oxidized NPs (CuNP3) were used. Observed solvent effect on the bacterial inhibition was minimal.
Positive Cu ions released from NPs interact with negative charges of microbes during the interac-
tion of NPs with bacterial microorganisms leading to cellular distortions and causing bacteria to
lose its viability. The results of viability assay method displayed a drastic reduction in a number of
bacterial colonies after NPs treatment. The corresponding images of these agar plates are shown in
Figure 11.29c through f. Figure 11.29c represents the untreated agar plate, d through f illustrates the
treated E. coli (gram-negative bacteria) with CuNP3, CuNP2, and CuNP1, respectively. The assay
method was sensitive than the agar well diffusion technique, even detecting the background activity
of solvents. The background activity of three solvents (ACN, DCM, acetone) was less (by at least
75%) than the activity exhibited by those solvents that were potentiated with NPs evident from the
bacterial growth on LB plates. The Ag NPs dissolved in ACN also showed remarkable antibacterial
activity (26 mm). In the viability assay, E. coli culture was grown overnight in LB. Aliquots of
410 Semiconductor Nanocrystals and Metal Nanoparticles

Acetone (a) (b) (c)

ACN
ACN
Cu NPs
+
Acetone
28 mm
25 mm
35 mm
Cu + Ag NPs
DCM + ACN

DCM Cu+ACN 26 mm

(d) (e) (f)

Cu NPs
+ ACN
Cu NPs
+ DCM

(g) (h)

Cu NPs+ Ag NPs
Acetone + ACN

FIGURE 11.29 (a–c) Agar well diffusion method: zone of growth inhibition: (a) E. coli with Cu NPs in
acetone and DCM (dichloromethane), (b) E. coli with Cu NPs in acetonitrile (ACN), (c) E. coli with Ag NPs in
ACN (Hamad et al. 2012), (d–h) viability assay method: (d) agar growth plate showing E. coli after incubation
and agar plate of captured E. coli with (e) Cu NPs in ACN, (f) Cu NPs in DCM, (g) Cu NPs in acetone, and
(h) Ag NPs in ACN.

100 µL of the culture were subcultured in fresh LB media (2 mL) and treated with Ag NPs and pure
s­ olvents, and only inoculums without NPs served as negative control. Cultures were incubated for
3 h. Hundred microliters from each test was taken and plated on LB agar plates and then incubated
at 37°C overnight. Bacterial growth was investigated and colonies were counted. The data for Ag
NPs are presented in Figure 11.29h. Our studies clearly demonstrated that the antibacterial capabil-
ity of Cu and Ag NPs achieved ULAL.

11.9 CONCLUSIONS
We were successful in fabricating several metallic (Al, Ag, Cu, etc.) NPs and NSs utilizing the
technique of ULAL. Several parameters (solvent, angle of laser pulses incidence, Bessel beam, etc.)
were investigated for their effect on the ablation products. The achieved NPs and NSs were demon-
strated to possess potential for several applications including SERS studies of explosive molecules,
and their photonic and antibacterial properties were studied. Some of the highlights of our studies
are summarized as follows:

1. Multiple/double/single line ablation differs in the effective pulse number per spot and degree
of over writing. During the process of single/double/multiple line ablation, roughness offered
by the Ag target varied as a function of the degree of overwriting. The differences in roughness
and number of pulses per spot demonstrated different yields of NPs and morphologies of NSs.
Applications of Metal Nanoparticles and Nanostructures 411

2. Double line ablated Ag NSs fabricated at different fluences demonstrated higher Raman
intensity enhancements for the explosive molecules RDX (~mM). Fluence of 16 J/cm2
used in the fabrication of Ag NSs showed superior SERS performance (Podagatlapalli
et al. 2013).
3. Ablation with nondiffracting Bessel beams rooted out the ambiguities on short DOF in
conventional Gaussian beam focusing through spherical lenses. First-order non-diffracting
Bessel beams were generated by focusing Gaussian beam using an axicon (or conical lens).
Ablation of Ag in DDW/acetone with Bessel beams revealed the fabrication of Ag NPs with
the same dimensions even for different energies (both in ps/fs regimes). Ag colloids pre-
pared in DDW (ps ablation) at different input pulse energies exhibited the SPR peak at the
same wavelength. A similar phenomenon was observed in fs case (Ag in acetone). Outcome
of Bessel beam ablation might be from the unusual dynamics of cavitation ­bubbles produced
at the time of ablation. Identical average sizes were tentatively assigned to the simultane-
ous fabrication/fragmentation of Ag NPs by Bessel beam transverse profile. Identification/
detection of CL-20 (5 μM) was achieved through Ag NSs prepared by Bessel beam ablation
(both ps and fs cases) and the estimated EFs were >106.
4. Ablation at 30° angle demonstrated the possibility of higher coupling efficiency of inci-
dent photons to the surface plasmons on Ag target enhancing the effectiveness of ablation.
Consequently, higher yields of Ag NPs and well-structured nanosurfaces were achieved.
SERS studies of R6G/ANTA illustrated that Ag NSs fabricated at 30° angle of incidence
provided higher EFs compared to others (Podagatlapalli et al. 2014).
5. Our recent SERS studies on NPs and NSs of Au–Ag alloys prepared using fs ablation
clearly demonstrated the capability of detecting secondary explosives of CL-20 and FOX-7
(Podagatlapalli et al. 2015). Other combinations (Au–Cu, Ag–Cu) of alloy NPs and NSs
are being investigated.
6. Cu substrates fabricated using ps multiple/single line technique were utilized exultantly
to investigate the surface plasmon (localized and propagating)–mediated enhancements
of different analytes using SERS studies. The reproducibility of the Cu targets for the
trace-level detection of explosive molecules was dealt with a simple cleaning procedure.
Trace-level detection of explosive compounds such as ANTA and TNT was performed
along with R6G individually after subjecting surfaces to appropriate cleaning procedures
and the obtained EFs were >105.
7. TTC–Cu–Cl and TPC–Cu–Cl colloidal solutions in TTC, TPC, and Cu/CuCl nano-
composites in chloroform were prepared by means of ps laser ablation of bulk Cu target.
Significant enhancements (~105) in the fluorescence were obtained for TTC/TPC placed
on Cu/CuCl nanocomposites compared to the case wherein fluorophores were directly
adsorbed (EF ~333) on the substrates (TTC–Cu–Cl, TPC–Cu–Cl colloids). The possible
reasons behind the enhancement were deliberated in detail. The enhancement in fluo-
rescence was attributed to the generated localized electromagnetic field of the Cu NPs
(Hamad et al. 2015).
8. Photonic and antibacterial properties of Cu and Ag NPs were successfully demonstrated,
and the NLO coefficients obtained for these NPs are on par or better than some of the
recently reported successful NLO moieties.

11.10 FUTURE SCOPE

Metallic NPs and NSs achieved through laser ablation (with ns, ps, or fs pulses) find further applica-
tions in a variety of fields. This chapter could not cover most of these applications. For example, Guo
and coworkers extensively worked on laser structuring of metallic targets (Vorobyev and Guo 2005,
2008, 2013) and tailored their reflectivity properties. Vorobyev and Guo (2005) demonstrated a
significant absorption enhancement due to nanostructuring of gold. They believed that the physical
412 Semiconductor Nanocrystals and Metal Nanoparticles

mechanism of the enhanced absorption was due to collective effects of nanostructural, microstruc-
tural, and macrostructural surface modifications induced by ablation incurred by fs laser pulses.
Vorobyev and coworkers (2009) applied the fs laser blackening technique directly to a tungsten
incandescent lamp filament and radically brightened the tungsten lamp and enhanced its emission
efficiency to approach 100%. Vorobyev and Guo (2013) again studied the origin and formation of
random surface NSs produced on metals using the fs laser ablation technique. Tan and coworkers
(2013) in their recent review article clearly demonstrated the possibility of preparing functional
nanomaterials using ULAL. They also provided the avenues for various applications of these func-
tional nanomaterials. Several new studies (De Bonis et al. 2014, Jiang and Pinchuk 2015, Li et al.
2014, Linic et al. 2015, Maximova et al. 2015, Nguyen et al. 2015, Sportelli et al., 2015, Tan et al.
2013a,b, 2015, Weng et al. 2014, Zhang et al. 2015) reported in the last couple of years as well pro-
vide a glimpse of potential of the fs laser pulses for creating metallic and nonmetallic NPs and NSs
for various applications. For example, Linic and coworkers (2015) reported the significance of metal
NPs in photochemical transformations. Sportelli and group (2015) used laser ablation–synthesized
NPs for food packaging applications. Finally, Zhang and coworkers (2015) reported intense THz
emission from random metallic NSs with an fs pulse irradiation.

ACKNOWLEDGMENTS
All the authors acknowledge continuous financial support from DRDO, India, through a grants-in-
aid scheme and fellowships. The authors also acknowledge the support of Dr. Md. A. Mohiddon
during the Raman measurements presented in this work. The authors also wish to acknowledge the
partial financial support from UPE-I and UPE-II of University of Hyderabad, India.

REFERENCES
Alnassar, S.I., Akman, E., Oztoprak, B.G., Kacar, E., Gundogdu, O., Khaleel, A., Demir, A. 2013. Study of
the fragmentation phenomena of TiO2 nanoparticles produced by femtosecond laser ablation in aqueous
media. Opt. Laser Technol. 51: 17–23.
Amendola, V., Meneghetti, M. 2007. Controlled size manipulation of free gold nanoparticles by laser irradia-
tion and their facile bioconjugation. J. Mater. Chem. 17: 4705–4710.
Amendola, V., Dini, D., Polizzi, S., Shen, J., Kadish, K.M., Calvete, M.J.F., Hanack, M., Meneghetti, M.
2009a. Self-healing of gold nanoparticles in the presence of zinc phthalocyanines and their very
efficient nonlinear absorption performances. J. Phys. Chem. C 113: 8688–8695.
Amendola, V., Meneghetti, M. 2009b. Laser ablation synthesis in solution and size manipulation of noble
metal nanoparticles. Phys. Chem. Chem. Phys. 11: 3805–3821.
Amendola, V., Meneghetti, M. 2013. What controls the composition and the structure of nanomaterials gener-
ated by laser ablation in liquid solution? Phys. Chem. Chem. Phys. 15: 3027–3046.
Amendola, V., Polizzi, S., Meneghetti, M. 2006. Laser ablation synthesis of gold nanoparticles in organic
solvents. J. Phys. Chem. B 110: 7232–7237.
Amendola, V., Polizzi, S., Meneghetti, M. 2012. Laser ablation synthesis of silver nanoparticles embedded in
graphitic carbon matrix. Sci. Adv. Mater. 4: 497–500.
Amendola, V., Riello, P., Meneghetti, M. 2011a. Magnetic nanoparticles of iron carbide, iron oxide, iron@
iron oxide, and metal iron synthesized by laser ablation in organic solvents. J. Phys. Chem. C 115:
5140–5146.
Amendola, V., Riello, P., Polizzi, S., Fiameni, S., Innocenti, C., Sangregorio, C., Meneghetti, M. 2011b.
Magnetic iron oxide nanoparticles with tunable size and free surface obtained via a “green” approach
based on laser irradiation in water. J. Mater. Chem. 21: 18665–18673.
Amendola, V., Scaramuzza, S., Litti, L., Meneghetti, M., Zuccolotto, G., Rosato, A., Nicolato, E. et al. 2014.
Magneto-plasmonic Au–Fe alloy nanoparticles designed for multimodal SERS-MRI-CT imaging.
Small 10: 2476–2486.
Amoruso, S., Ausanio, G., Barone, A.C., Bruzzese, R., Gragnaniello, L., Vitiello, M., Wang, X. 2005a.
Ultrashort laser ablation of solid matter in vacuum: A comparison between the picosecond and femto-
second regimes. J. Phys. B: Atom. Mol. Opt. Phys. 38: L329–L338.
Applications of Metal Nanoparticles and Nanostructures 413

Amoruso, S., Ausanio, G., Bruzzese, R., Vitello, M., Wang, X. 2005b. Femtosecond laser pulse irradiation of
solid targets as a general route to nanoparticle formation in a vacuum. Phys. Rev. B 71: 033406.
Ancona, A., Sportelli, M.C., Trapani, A., Picca, R.A., Palazzo, C., Bonerba, E., Mezzapesa, F.P., Tantillo, G.,
Trapani, G., Cioffi, N. 2014. Synthesis and characterization of hybrid copper-chitosan nano-antimicro-
bials by femtosecond laser-ablation in liquids. Mater. Lett. 136: 397–400.
Arboleda, D.M., Santillán, J.M.J., Herrera, L.J.M., Van Raap, M.B.F., Zélis, P.M., Muraca, D., Schinca, D.C.,
Scaffardi, L.B. 2015. Synthesis of Ni nanoparticles by femtosecond laser ablation in liquids: Structure
and sizing. J. Phys. Chem. C 119: 13184–13193.
Arruebo, M., Pacheco, R.F., Ibarra, M.R., Santamaria, J. 2007. Magnetic nanoparticles for drug delivery.
NanoToday 2: 22–32.
Asahi, T., Mafuné, F., Rehbock, C., Barcikowski, S. 2015. Strategies to harvest the unique properties of laser-
generated nanomaterials in biomedical and energy applications. Appl. Surf. Sci. 348: 1–3.
Bagga, K., Barchanski, A., Intartaglia, R., Dante, S., Marotta, R., Diaspro, A., Sajti, C.L., Brandi, F. 2013.
Laser-assisted synthesis of Staphylococcus aureus protein-capped silicon quantum dots as bio-func-
tional nanoprobes. Laser Phys. Lett. 10: 065603.
Balling, P., Schou, J. 2006. Femtosecond-laser ablation dynamics of dielectrics: Basics and applications for
thin films. Rep. Prog. Phys. 76: 036502.
Barchanski, A., Funk, D., Wittich, O., Tegenkamp, C., Chichkov, B.N., Sajti, C.L. 2015. Picosecond laser
fabrication of functional gold-antibody nanoconjugates for biomedical applications. J. Phys. Chem. C
119: 9524–9533.
Barcikowski, S., Hahn, A., Guggenheim, M., Reimers, K., Ostendorf, A. 2010. Biocompatibility of nano-
actuators: Stem cell growth on laser-generated nickel–titanium shape memory alloy nanoparticles.
J. Nanopart. Res. 12: 1733–1742.
Barcikowski, S., Mafune, F. 2011. Trends and current topics in the field of laser ablation and nanoparticle
generation in liquids. J. Phys. Chem. C 115: 4985.
Barcikowski, S., Meńndez-Manjón, A., Chichkov, B., Brikas, M., Račiukaitis, G. 2007. Generation of nanopar-
ticle colloids by picosecond and femtosecond laser ablations in liquid flow. Appl. Phys. Lett. 91: 083113.
Barmina, E.B., Stratakis, E., Fotakis, C., Shafeev, G.A. 2010. Generation of nanostructures on metals by laser
ablation in liquids: New results. Quant. Electron. 40: 1012–1020.
Barmina, E.V., Barberoglu, M., Zorba, V., Simakin, A.V., Stratakis, E., Fotakis, C., Shafeev, G.A. 2009.
Surface nanotexturing of tantalum by laser ablation in water. Quant. Electron. 39: 89–93.
Bärsch, N., Jakobi, J., Weiler, S., Barcikowski, S. 2009. Pure colloidal metal and ceramic nanoparticles from
high-power picosecond laser ablation in water and acetone. Nanotechnology 20: 445603.
Batani, D. 2010. Short-pulse laser ablation of materials at high intensities: Influence of plasma effects. Laser
Part. Beams 28: 235–244.
Batani, D., Vinci, T., Bleiner, D. 2014. Laser-ablation and induced nanoparticle synthesis. Part. Beams 32: 1–7.
Bertorelle, F., Ceccarello, M., Pinto, M., Fracasso, G., Badocco, D., Amendola, V., Pastore, P., Colombatti, M.,
Meneghetti, M. 2014. Efficient AuFeOx nanoclusters of laser-ablated nanoparticles in water for
cells guiding and surface-enhanced resonance Raman scattering imaging. J. Phys. Chem. C 118:
14534–14541.
Besner, S., Kabashin, A.V., Meunier, M. 2007. Two-step femtosecond laser ablation-based method for the syn-
thesis of stable and ultra-pure gold nanoparticles in water. Appl. Phys. A Mater. Sci. Process. 88: 269–272.
Besner, S., Kabashin, A.V., Meunier, M., Winnik, F.M. 2005. Fabrication of functionalized gold manoparticles
by femtosecond laser ablation in aqueous solutions of biopolymers. Proc. SPIE 5969: 59690B.
Besner, S., Kabashin, A.V., Winnik, F.M., Meunier, M. 2008. Ultrafast laser based “green” synthesis of non-toxic
nanoparticles in aqueous solutions. Appl. Phys. A Mater. Sci. Process. 93: 955–959.
Besner, S., Kabashin, A.V., Winnik, F.M., Meunier, M. 2009. Synthesis of size-tunable polymer-protected gold
nanoparticles by femtosecond laser-based ablation and seed growth. J. Phys. Chem. C 113: 9526–9531.
Besner, S., Meunier, M. 2010. Laser synthesis of nanomaterials. In Laser Precision Microfabrication, Sugioka,
K., Meunier, M., and Piqué, A. (Eds.). Springer Series in Materials Science, Springer-Verlag, Berlin,
Germany, Vol. 135, pp. 163–187.
Boulais, E., Lachaine, R., Hatef, A., Meunier, M. 2013. Plasmonics for pulsed-laser cell nanosurgery:
Fundamentals and applications. J. Photochem. Photobiol. C Photochem. Rev. 17: 26–49.
Chelnokov, E., Rivoal, M., Colignon, Y., Gachet, D., Bekere, L., Thibaudau, F., Giorgio, S., Khodorkovsky, V.,
Marine, W. 2012. Band gap tuning of ZnO nanoparticles via Mg doping by femtosecond laser ablation
in liquid environment. Appl. Surf. Sci. 258: 9408–9411.
Chen, J.K., Beraun, J.E. 2001. Numerical study of ultrashort laser pulses interaction with metal films. Numer.
Heat Trans. A Appl. 40: 1–20.
414 Semiconductor Nanocrystals and Metal Nanoparticles

Chen, Y.-H., Yeh, C.-S. 2001. A new approach for the formation of alloy nanoparticles: Laser synthesis of
gold–silver alloy from gold–silver colloidal mixtures. Chem. Commun. 4: 371–372.
Chichkov, B.N., Momma, C., Nolte, S., von Alvensleben, F., Tunnermann, A. 1996. Femtosecond, picosecond
and nanosecond laser ablation of solids. Appl. Phys. A Mater. Sci. Process. 63: 109–115.
Compagnini, G., Messina, E., Puglisi, O., Cataliotti, R.S., Nicolosi, V. 2008. Spectroscopic evidence of a
core–shell structure in the earlier formation stages of Au–Ag nanoparticles by pulsed laser ablation in
water. Chem. Phys. Lett. 457: 386–390.
Compagnini, G., Scalisi, A.A., Puglisi, O. 2003. Production of gold nanoparticles by laser ablation in liquid
alkanes. J. Appl. Phys. 94: 7874–7877.
Compagnini, G., Sinatra, M.G., Messina, G.C., Patan, G., Scalese, S., Puglisi, O. 2012. Monitoring the forma-
tion of inorganic fullerene-like MoS2 nanostructures by laser ablation in liquid environments. Appl.
Surf. Sci. 258: 5672–5676.
Dallaire, A.-M., Rioux, D., Rachkov, A., Patskovsky, S., Meunier, M. 2012. Laser-generated Au–Ag nanopar-
ticles for plasmonic nucleic acid sensing. J. Phys. Chem. C 116: 11370–11377.
De Bonis, A., Sansone, M., Alessio, L.D., Galasso, A., Santagata, A., Teghil, R. 2013. Dynamics of laser-
induced bubble and nanoparticles generation during ultra-short laser ablation of Pd in liquid. J. Phys.
D Appl. Phys. 46: 445301.
De Bonis, A., Sansone, M., Galasso, A., Santagata, A., Teghil, R. 2014. The role of the solvent in the ultrashort
laser ablation of palladium target in liquid. Appl. Phys. A Mater. Sci. Process. 117: 211–216.
De Giacomo, A., Dell’Aglio, M., Santagata, A., Gaudiuso, R., De Pascale, O., Wagener, P., Messina, G.C.,
Compagnini, G., Barcikowski, S. 2013. Cavitation dynamics of laser ablation of bulk and wire-shaped
metals in water during nanoparticles production. Phys. Chem. Chem. Phys. 15: 3083–3092.
Del Fatti, N., Vallée, F. 2001. Ultrafast optical nonlinear properties of metal nanoparticles. Appl. Phys. B
Lasers Opt. 73: 383–390.
Delfour, L., Itina, T.E. 2015. Mechanisms of ultrashort laser-induced fragmentation of metal nanoparticles in
liquids: Numerical insights. J. Phys. Chem. C 119: 13893–13900.
Dell’Aglio, M., Gaudiuso, R., De Pascale, O., De Giacomo, A. 2015. Mechanisms and processes of pulsed
laser ablation in liquids during nanoparticle production. Appl. Surf. Sci. 348: 4–9.
Dolgaev, S.I., Simakin, A.V., Voronov, V.V., Shafeev, G.A., Bozon-Verduraz, F. 2002. Nanoparticles produced
by laser ablation of solids in liquid environment. Appl. Surf. Sci. 186: 546–551.
Elsayed, K.A., Imam, H., Ahmed, M.A., Ramadan, R. 2013. Effect of focusing conditions and laser parameters
on the fabrication of gold nanoparticles via laser ablation in liquid. Opt. Laser Technol. 45: 495–502.
Eroshova, O.I., Perminov, P.A., Zabotnov, S.V., Gongal’skii, M.B., Ezhov, A.A., Golovan’, L.A., Kashkarov,
P.K. 2012. Structural properties of silicon nanoparticles formed by pulsed laser ablation in liquid media.
Cryst. Rep. 57: 831–835.
Fan, G., Ren, S., Qu, S., Wang, Q., Gao, R. 2014. Stability and nonlinear optical properties of Cu nanoparticles
prepared by femtosecond laser ablation of Cu target in alcohol and water. Opt. Commun. 330: 122–130.
Fan, M., Lai, F., Chou, H., Lu, W., Hwang, B., Brolo, A.G. 2013. Surface-enhanced Raman scattering (SERS)
from Au:Ag bimetallic nanoparticles: The effect of the molecular probe. Chem. Sci. 4: 509–515.
Fazio, E., Neri, F. 2013. Nonlinear optical effects from Au nanoparticles prepared by laser plasmas in water.
Appl. Surf. Sci. 272: 88–93.
Gamaly, E.G., Rode, A.V., Luther-Davies, B., Tichonchuk, V.T. 2002. Ablation of solids by femtosecond lasers:
Ablation mechanism and ablation thresholds for metals and dielectrics. Phys. Plasmas 9: 949–957.
Ganeev, R.A., Baba, M., Ryasnyansky, A.I., Suzuki, M., Kuroda, H. 2004. Characterization of optical and
nonlinear optical properties of silver nanoparticles prepared by laser ablation in various liquids. Opt.
Commun. 240: 437–448.
Ganeev, R.A., Baba, M., Ryasnyansky, A.I., Suzuki, M., Kuroda, H. 2005. Laser ablation of GaAs in liquids:
Structural, optical, and nonlinear optical characteristics of colloidal solutions. Appl. Phys. B Lasers
Opt. 80: 595–601.
Gökce, B., Van’t Zand, D.D., Menéndez-Manjón, A., Barcikowski, S. 2015. Ripening kinetics of laser-gener-
ated plasmonic nanoparticles in different solvents. Chem. Phys. Lett. 626: 96–101.
Guisbiers, G., Wang, Q., Khachatryan, E., Arellano-Jimenez, M.J., Webster, T.J., Larese-Casanova, P., Nash, K.L.
2015. Anti-bacterial selenium nanoparticles produced by UV/VIS/NIR pulsed nanosecond laser abla-
tion in liquids. Laser Phys. Lett. 12: 016003.
Guo, S., Wang, E. 2011. Noble metal nanomaterials: Controllable synthesis and application in fuel cells and
analytical sensors. NanoToday 6: 240–264.
Hajiesmaeilbaigi, F., Motamedi, A. 2007. Synthesis of Au/Ag alloy nanoparticles by Nd:YAG laser irradiation.
Laser Phys. Lett. 4: 133–137.
Applications of Metal Nanoparticles and Nanostructures 415

Hamad, S. 2014. Ultrafast laser fabricated nanoparticles and nanostructures: Characterization, spectroscopy
and applications. PhD thesis, University of Hyderabad, Hyderabad, India.
Hamad, S., Podagatlapalli, G.K., Mohiddon, M.A., Venugopal Rao, S. 2014b. Cost effective nanostructured
copper substrates prepared with ultrafast laser pulses for explosives detection using surface enhanced
Raman scattering. Appl. Phys. Lett. 104: 263104.
Hamad, S., Podagatlapalli, G.K., Mohiddon, M.A., Venugopal Rao, S. 2015. Surface enhanced fluorescence
from corroles and SERS studies of explosives using copper nanostructures. Chem. Phys. Lett. 621:
171–176.
Hamad, S., Podagatlapalli, G.K., Nageswara Rao, S.V.S., Pathak, A.P., Venugopal Rao, S. 2014a. Excited state
dynamics of silicon nanocrystals fabricated using ultrafast laser ablation in liquids. In 12th International
Conference on Fiber Optics and Photonics, OSA Technical Digest (online) (Optical Society of America,
2014), paper T3A.48.
Hamad, S., Podagatlapalli, G.K., Sreedhar, S., Tewari, S.P., Venugopal Rao, S. 2012. Femtosecond and pico-
second ablation of aluminum for synthesis of nanoparticles and nanostructures and their optical charac-
terization. Proc. SPIE 8245: 82450L.
Hamad, S., Podagatlapalli, G.K., Tewari, S.P., Venugopal Rao, S. 2013. Influence of picosecond multiple/
single line ablation on copper nanoparticles fabricated for surface enhanced Raman spectroscopy and
photonics applications. J. Phys. D Appl. Phys. 46: 485501.
Hamad, S., Podagatlapalli, G.K., Vendamani, V.S., Nageswara Rao, S.V.S., Pathak, A.P., Tewari, S.P.,
Venugopal Rao, S. 2014c. Femtosecond ablation of silicon in acetone: Tunable photoluminescence
from generated nanoparticles and fabrication of surface nanostructures. J. Phys. Chem. C 118:
7139–7151.
Harilal, S.S., Freeman, J.R., Diwakar, P.K., Hassanein, A. 2014. Laser-Induced Breakdown Spectroscopy.
Springer Series in Optical Sciences, Springer, Berlin, Germany, Vol. 182, pp. 143–166, Chapter 6.
Hatanaka, K., Itoh, T., Asahi, T., Ichinose, N., Kawanishi, S., Sasuga, T., Fukumura, H., Masuhara, H. 1999.
Femtosecond laser ablation of liquid toluene: Molecular mechanism studied by time-resolved absorp-
tion spectroscopy. J. Phys. Chem. A 103: 11257–11263.
Hatef, A., Meunier, M. 2015. Plasma-mediated photothermal effects in ultrafast laser irradiation of gold
nanoparticle dimers in water. Opt. Exp. 23: 1967–1980.
Hu, H., Liu, T., Zhai, H. 2015. Comparison of femtosecond laser ablation of aluminum in water and in air by
time-resolved optical diagnosis. Opt. Exp. 23: 628–635.
Hu, X., Gong, H., Xu, H., Wei, H., Cao, B., Liu, G., Zeng, H., Cai, W. 2011. Influences of target and liquid
media on morphologies and optical properties of ZnO nanoparticles prepared by laser ablation in solution.
J. Am. Ceramic Soc. 94: 4305–4309.
Ibrahimkutty, S., Wagener, P., Menzel, A., Plech, A., Barcikowski, S. 2012. Nanoparticle formation in a cavi-
tation bubble after pulsed laser ablation in liquid studied with high time resolution small angle X-ray
scattering. Appl. Phys. Lett. 101: 103104.
Intartaglia, R., Bagga, K., Brandi, F. 2014. Study on the productivity of silicon nanoparticles by picosecond
laser ablation in water: Towards gram per hour yield. Opt. Exp. 22: 3117–3127.
Intartaglia, R., Bagga, K., Brandi, F., Das, G., Genovese, A., Di Fabrizio, E., Diaspro, A. 2011. Optical proper-
ties of femtosecond laser-synthesized silicon nanoparticles in deionized water. J. Phys. Chem. C 115:
5102–5107.
Intartaglia, R., Bagga, K., Genovese, A., Athanassiou, A., Cingolani, R., Diaspro, A., Brandi, F. 2012a.
Influence of organic solvent on optical and structural properties of ultra-small silicon dots synthesized
by UV laser ablation in liquid. Phys. Chem. Chem. Phys. 14: 15406–15411.
Intartaglia, R., Bagga, K., Scotto, M., Diaspro, A., Brandi, F. 2012b. Luminescent silicon nanoparticles pre-
pared by ultrashort pulsed laser ablation in liquid for imaging applications. Opt. Mater. Exp. 2: 510–518.
Intartaglia, R., Barchanski, A., Bagga, K., Genovese, A., Das, G., Wagener, P., Di Fabrizio, E., Diaspro, A.,
Brandi, F., Barcikowski, S. 2012c. Bioconjugated silicon quantum dots from one-step green synthesis.
Nanoscale 4: 1271–1274.
Intartaglia, R., Das, G., Bagga, K., Gopalakrishnan, A., Genovese, A., Povia, M., Di Fabrizio, E., Cingolani,
R., Diaspro, A., Brandi, F. 2013. Laser synthesis of ligand free bimetallic nanoparticles for plasmonic
applications. Phys. Chem. Chem. Phys. 15: 3075–3082.
Ishikawa, Y., Kawaguchi, K., Shimizu, Y., Sasaki, T., Koshizaki, N. 2006. Preparation of Fe–Pt alloy particles
by pulsed laser ablation in liquid medium. Chem. Phys. Lett. 428: 426–429.
Ismail, R.A., Sulaiman, G.M., Abdulrahman, S.A., Marzoog, T.R. 2015. Antibacterial activity of magnetic
iron oxide nanoparticles synthesized by laser ablation in liquid. Mater. Sci. Eng. C 53: 286–297.
Itina, T.E. 2011. On nanoparticles formation by laser ablation in liquids. J. Phys. Chem. C 115: 5044–5048.
416 Semiconductor Nanocrystals and Metal Nanoparticles

Ivanov, D.S., Zhigilei, L.V. 2003. Combined atomistic-continuum modeling of short-pulse laser melting and
disintegration of metal films. Phys. Rev. B 68: 064114.
Jain, P.K., Huang, X., El-Sayed, I.H., El-Sayed, M. 2008. Noble metals on the nanoscale: Optical and photo-
thermal properties and some applications in imaging, sensing, biology, and medicine. Acc. Chem. Res.
41: 1578–1586.
Jiang, K., Pinchuk, A.O. 2015. Noble metal nanomaterials: Synthetic routes, fundamental properties, and
promising applications. Solid State Phys., 66: 131–211.
Jiang, Y., Liu, P., Liang, Y., Li, H.B., Yang, G.W. 2011. Promoting the yield of nanoparticles from laser ablation
in liquid. Appl. Phys. A: Mater. Sci. Process. 105: 903–907.
Jin, Z., Gu, W., Shi, X., Wang, Z., Jiang, Z., Liao, L. 2014. A novel route to surface-enhanced Raman scat-
tering: Ag nanoparticles embedded in the nano-gaps of a Ag substrate. Adv. Opt. Mater. 2: 588–596.
Kabashin, A.V., Meunier, M. 2003. Synthesis of colloidal nanoparticles during femtosecond laser ablation of
gold in water. J. Appl. Phys. 94: 7941–7943.
Kabashin, A.V., Meunier, M. 2006. Laser ablation-based synthesis of functionalized colloidal nanomaterials
in biocompatible solutions. J. Photochem. Photobiol. A Chem. 182: 330–334.
Kabashin, A.V., Meunier, M., Kingston, C., Luong, J.H.T. 2003. Fabrication and characterization of gold
nanoparticles by femtosecond laser ablation in an aqueous solution of cyclodextrins. J. Phys. Chem. B
107: 4527–4531.
Kamat, P.V. 2002. Photophysical, photochemical and photocatalytic aspects of metal nanoparticles. J. Phys.
Chem. B 106: 7729–7744.
Karavanskii, V.A., Simakin, A.V., Krasovskii, V.I., Ivanchenko, P.V. 2004. Nonlinear optical properties
of colloidal silver nanoparticles produced by laser ablation in liquids. Quant. Electron. 34: 644–648.
Kazakevich, P.V., Simakin, A.V., Voronov, V.V., Shafeev, G.A. 2006. Laser induced synthesis of nanoparticles
in liquids. Appl. Surf. Sci. 252: 4373–4380.
Kazakevich, P.V., Voronov, V.V., Simakin, A.V., Shafeev, G.A. 2004. Production of copper and brass nanopar-
ticles upon laser ablation in liquids. Quant. Electron. 34: 951–956.
Kelly, K.L., Coronado, E., Zhao, L.L., Schatz, G.C. 2003. The optical properties of metal nanoparticles:
The influence of size, shape, and dielectric environment. J. Phys. Chem. B 107: 668–677.
Kenth, S., Sylvestre, J.-P., Fuhrmann, K., Meunier, M., Leroux, J.-C. 2011. Fabrication of paclitaxel nanocrys-
tals by femtosecond laser ablation and fragmentation. J. Pharma. Sci. 100: 1022–1030.
König, J., Nolte, S., Tünnermann, A. 2005. Plasma evolution during metal ablation with ultrashort laser pulses.
Opt. Exp. 13: 10597–10607.
Kuladeep, R., Jyothi, L., Ali, S.S., Deepak, K.L.N., Narayana Rao, D. 2012. Laser-assisted synthesis of Au–Ag
alloy nanoparticles with tunable surface plasmon resonance frequency. Opt. Mater. Exp. 2: 161–172.
Kuzmin, P.G., Shafeev, G.A., Bukin, V.V., Garnov, S.V., Farcau, C., Carles, R., Warot-Fontrose, B., Guieu, V.,
Viau, G. 2010. Silicon nanoparticles produced by femtosecond laser ablation in ethanol: Size control,
structural characterization, and optical properties. J. Phys. Chem. C 114: 15266–15273.
Lachaine, R., Boulais, E., Meunier, M. 2014. From thermo- to plasma-mediated ultrafast laser induced
plasmonic nanobubbles. ACS Photon. 1: 331–336.
Lalayan, A.A. 2005. Formation of colloidal GaAs and CdS quantum dots by laser ablation in liquid media.
Appl. Surf. Sci. 248: 209–212.
Lee, I., Han, S.W., Kim, K. 2001. Production of Au–Ag alloy nanoparticles by laser ablation of bulk alloys.
Chem. Commun. 18: 1782–1783.
Leitz, K.-H., Redlingshöer, B., Reg, Y., Otto, A., Schmidt, M. 2011. Metal ablation with short and ultrashort
laser pulses. Phys. Proc. 12(Part 2): 230–238.
Li, Y.-J., Chiu, W.-J., Unnikrishnan, B., Huang, C.-C. 2014. Monitoring thrombin generation and screening
anticoagulants through pulse laser-induced fragmentation of biofunctional nanogold on cellulose mem-
branes. ACS Appl. Mater. Int. 6: 15253–15261.
Liang, Y., Liu, P., Xiao, J., Li, H.B., Wang, C.X., Yang, G.W. 2014. A general strategy for one-step fabrication
of one-dimensional magnetic nanoparticle chains based on laser ablation in liquid. Laser Phys. Lett.
11: 056001.
Linic, S., Aslam, U., Boerigter, C., Morabito, M. 2015. Photochemical transformations on plasmonic metal
nanoparticles. Nat. Mater. 14: 567–576.
Link, S., Mostafa El-Syed, A. 2003. Optical properties and ultrafast dynamics of metallic nanocrystals. Annu.
Rev. Phys. Chem. 54: 331–366.
Liu, B., Hu, Z., Che, Y., Chen, Y., Pan, X. 2007. Nanoparticle generation in ultrafast pulsed laser ablation of
nickel. Appl. Phys. Lett. 90: 044103.
Applications of Metal Nanoparticles and Nanostructures 417

Liu, P., Cai, W., Fang, M., Li, Z., Zeng, H., Hu, J., Luo, X., Jing, W. 2009. Room temperature synthesized rutile
TiO2 nanoparticles induced by laser ablation in liquid and their photocatalytic activity. Nanotechnology
20: 285707.
Liu, P., Cui, H., Wang, C.X., Yang, G.W. 2010. From nanocrystal synthesis to functional nanostructure fabrica-
tion: Laser ablation in liquid. Phys. Chem. Chem. Phys. 12: 3942–3952.
Lorazo, P., Lewis, L.J., Meunier, M. 2003. Short-pulse laser ablation of solids: From phase explosion to frag-
mentation. Phys. Rev. Lett. 91: 225502/1–225502/4.
Lorazo, P., Lewis, L.J., Meunier, M. 2006. Thermodynamic pathways to melting, ablation, and solidification
in absorbing solids under pulsed laser irradiation. Phys. Rev. B 73: 134108.
Makarov, G.N. 2013. Laser applications in nanotechnology: Nanofabrication using laser ablation and laser
nanolithography. Phys. Usp. 56: 643–682.
Margetic, V. 2002. Femtosecond laser ablation. PhD thesis, Institute of Spectrochemistry and Applied
Spectroscopy, Dortmund, Germany.
Maximova, K., Aristov, A., Sentis, M., Kabashin, A.V. 2015. Size-controllable synthesis of bare gold nanopar-
ticles by femtosecond laser fragmentation in water. Nanotechnology 26: 065601.
Mendivil, M.I., Krishnan, B., Sanchez, F.A., Martinez, S., Aguilar-Martinez, J.A., Castillo, G.A., Garcia-
Gutierrez, D.I., Shaji, S. 2013. Synthesis of silver nanoparticles and antimony oxide nanocrystals by
pulsed laser ablation in liquid media. Appl. Phys. A Mater. Sci. Process. 110: 809–816.
Menéndez-Manjon, A., Barcikowski, S., Shafeev, G.A., Mazhukin, V.I., Chichkov, B.N. 2010. Influence of
beam intensity profile on the aerodynamic particle size distributions generated by femtosecond laser
ablation. Laser Part. Beams 28: 45–52.
Menéndez-Manjón, A., Schwenke, A., Steinke, T., Meyer, M., Giese, U., Wagener, P., Barcikowski, S. 2013.
Ligand-free gold-silver nanoparticle alloy polymer composites generated by picosecond laser ablation
in liquid monomer. Appl. Phys. A Mater. Sci. Process. 110: 343–350.
Menéndez-Manjón, A., Wagener, P., Barcikowski, S. 2011. Transfer-matrix method for efficient ablation by
pulsed laser ablation and nanoparticle generation in liquids. J. Phys. Chem. C 115: 5108–5114.
Mortazavi, S.Z., Parvin, P., Reyhani, A., Golikand, A.N., Mirershadi, S. 2011. Effect of laser wavelength at IR
(1064 nm) and UV (193 nm) on the structural formation of palladium nanoparticles in deionized water.
J. Phys. Chem. C 115: 5049–5057.
Murphy, C.J., Sau, T.K., Gole, A.M., Orendorff, C.J., Gao, J., Gou, L., Hunyadi, S.E., Li, T. 2005.
Anisotropic metal nanoparticles: Synthesis, assembly, and optical applications. J. Phys. Chem. B
109: 13857–13870.
Narayanan, K.B., Sakthivel, N. 2010. Biological synthesis of metal nanoparticles by microbes. Adv. Colloid
Int. Sci. 156: 1–13.
Nath, A., Das, A., Rangan, L., Khare, A. 2012. Bacterial inhibition by Cu/Cu2O nanocomposites prepared via
laser ablation in liquids. Sci. Adv. Mater. 4: 106–109.
Nedderson, J., Chumanov, G., Cotton, T.M. 1993. Laser ablation of metals: A new method for preparing SERS
active colloids. Appl. Spectrosc. 47: 1959–1964.
Nguyen, V., Yan, L., Si, J., Hou, X. 2015. Femtosecond laser-induced size reduction of carbon nanodots in
solution: Effect of laser fluence, spot size, and irradiation time. J. Appl. Phys. 117: 084304.
Nichols, W.T., Sasaki, T., Koshizaki, N. 2006a. Laser ablation of a platinum target in water. I. Ablation mecha-
nisms. J. Appl. Phys. 100: 114911.
Nichols, W.T., Sasaki, T., Koshizaki, N. 2006b. Laser ablation of a platinum target in water. II. Ablation rate
and nanoparticle size distributions. J. Appl. Phys. 100: 114912.
Nichols, W.T., Sasaki, T., Koshizaki, N. 2006c. Laser ablation of a platinum target in water. III. Laser-induced
reactions. J. Appl. Phys. 100: 114913.
Nie, S., Emory, S.R. 1997. Probing single molecules and single nanoparticles by surface-enhanced Raman
scattering. Science 275: 1102–1106.
Niu, K.Y., Yang, J., Kulinich, S.A., Sun, J., Li, H., Du, X.W. 2010. Morphology control of nanostructures via
surface reaction of metal nanodroplets. J. Am. Chem. Soc. 132: 9814–9819.
Noël, S., Hermann, J., Itina, T. 2007. Investigation of nanoparticle generation during femtosecond laser
ablation of metals. Appl. Surf. Sci. 253: 6310–6315.
Nolte, S., Chichkov, B.N., Welling, H., Shani, Y., Liebermann, K., Terkel, H. 1999. Nanostructuring with
spatially localized femtosecond laser pulses. Opt. Lett. 24: 914–916.
Olea-Mejía, O., Fernández-Mondragón, M., Rodríguez-de la Concha, G., Camacho-López, M. 2015. SERS-
active Ag, Au and Ag–Au alloy nanoparticles obtained by laser ablation in liquids for sensing methylene
blue. Appl. Surf. Sci. 348: 66–70.
418 Semiconductor Nanocrystals and Metal Nanoparticles

O’Malley, S.M., Amin, M., Borchert, J., Jimenez, R., Steiner, M., Fitz-Gerald, J.M., Bubb, D.M. 2014.
Formation of rubrene nanocrystals by laser ablation in liquids utilizing MAPLE deposited thin films.
Chem. Phys. Lett. 595–596: 171–174.
Onodera, Y., Nunokawa, T., Odawara, O., Wada, H. 2013. Upconversion properties of Y2O3:Er,Yb nanopar-
ticles prepared by laser ablation in water. J. Lumin. 137: 220–224.
Ou, Y., Wang, L., Zhu, L., Wan, L., Xu, Z. 2014. In-situ immobilization of silver nanoparticles on self-
assembled honeycomb-patterned films enables surface-enhanced Raman scattering (SERS) substrates.
J. Phys. Chem. C 118: 11478−11484.
Papagiannouli, I., Aloukos, P., Rioux, D., Meunier, M., Couris, S. 2015. Effect of the composition on the
nonlinear optical response of Au xAg1–x nano-alloys. J. Phys. Chem. C 119: 6861−6872.
Pelton, M., Aizpuru, J., Bryant, G. 2008. Metal-nanoparticle plasmonics. Laser Photon. Rev. 2: 136–159.
Pereira, A., Cros, A., Delaporte, P., Georgiou, S., Manousaki, A., Marine, W., Sentis, M. 2004. Surface nano-
structuring of metals by laser irradiation: Effects of pulse duration, wavelength and gas atmosphere.
Appl. Phys. A Mater. Sci. Process. 79: 1433–1437.
Perez, D., Beland, L.K., Deryng, D., Lewis, L.J., Meunier, M. 2008. Numerical study of the thermal ablation
of wet solids by ultrashort laser pulses. Phys. Rev. B 77: 014108.
Perez, D., Lewis, L.J. 2003. Molecular-dynamics study of ablation of solids under femtosecond laser pulses.
Phys. Rev. B 67: 184102.
Petersen, S., Jakobi, J., Barcikowski, S. 2009. In situ bioconjugation—Novel laser based approach to pure
nanoparticle-conjugates. Appl. Surf. Sci. 255: 5435–5438.
Podagatlapalli, G.K. 2014. Ultrafast laser ablation of metals in liquids for explosives detection using SERS and
development of CARS experiments. PhD thesis, University of Hyderabad, Hyderabad, India.
Podagatlapalli, G.K., Hamad, S., Mohiddon, M.A., Venugopal Rao, S. 2014a. Effect of oblique incidence on
silver nanomaterials fabricated in water via ultrafast laser ablation for photonics and explosives detec-
tion. Appl. Surf. Sci. 303: 217–232.
Podagatlapalli, G.K., Hamad, S., Mohiddon, M.A., Venugopal Rao, S. 2015a. Fabrication of nanoparticles and
nanostructures using ultrafast laser ablation of silver with Bessel beams. Laser Phys. Lett. 12: 036003.
Podagatlapalli, G.K., Hamad, S., Sreedhar, S., Prasad, M.D., Venugopal Rao, S. 2013. Silver nano-entities
through ultrafast double ablation in aqueous media for surface enhanced Raman scattering and photon-
ics applications. J. Appl. Phys. 113: 073106.
Podagatlapalli, G.K., Hamad, S., Sreedhar, S., Tewari, S.P., Venugopal Rao, S. 2012. Fabrication and
characterization of aluminum nanostructures and nanoparticles obtained using femtosecond ablation
technique. Chem. Phys. Lett. 530: 93–97.
Podagatlapalli, G.K., Hamad, S., Venugopal Rao, S. 2014b. Fabrication of hybrid Ag–Au nanomaterials for
explosives detection. In 12th International Conference on Fiber Optics and Photonics, OSA Technical
Digest (online) (Optical Society of America, 2014), paper S5A.51.
Podagatlapalli, G.K., Hamad, S., Venugopal Rao, S. 2014c. Silver nanomaterials in aqueous media fabricated
with non-diffracting picosecond Bessel beam and applications. In Light, Energy and the Environment,
OSA Technical Digest (online) (Optical Society of America, 2014), paper JW6A.13.
Podagatlapalli, G.K., Hamad, S., Venugopal Rao, S. 2015b. Trace-level detection of secondary explosives
using hybrid silver-gold nanoparticles and nanostructures achieved with femtosecond laser ablation.
J. Phys. Chem. C 119: 16972–16983.
Povarnitsyn, M.E., Itina, T.E., Levashov, P.R., Khishchenko, K.V. 2013. Mechanisms of nanoparticle formation
by ultra-short laser ablation of metals in liquid environment. Phys. Chem. Chem. Phys. 15: 3108–3114.
Povarnitsyn, M.E., Itina, T.E., Sentis, M., Khishchenko, K.V., Levashov, P.R. 2007. Material decomposition
mechanisms in femtosecond laser interactions with metals. Phys. Rev. B 75: 235414.
Pradeep, T., Anshup. 2009. Noble metal nanoparticles for water purification: A critical review. Thin Solid
Films 517: 6441–6478.
Qua, L., Song, Q., Li, Y., Peng, M., Li, D., Chen, L., Fossey, J.S., Long, Y. 2013. Fabrication of bimetallic
microfluidic surface-enhanced Raman scattering sensors on paper by screen printing. Anal. Chim. Acta
792: 86–92.
Radziemski, J., Creamers, D.A. 1989. Laser Induced Plasma and Applications. Marcel Dekker, New York.
Rehbock, C., Jakobi, J., Gamrad, L., van der Meer, S., Tiedemann, D., Taylor, U., Kues, W., Rath, D.,
Barcikowski, S. 2014. Current state of laser synthesis of metal and alloy nanoparticles as ligand-free
reference materials for nano-toxicological assays. Beilstein J. Nanotechnol. 5: 1523–1541.
Riabinina, D., Chaker, M., Margot, J. 2012. Dependence of gold nanoparticle production on pulse duration by
laser ablation in liquid media. Nanotechnology 23: 135603.
Applications of Metal Nanoparticles and Nanostructures 419

Rioux, D., Laferrière, M., Douplik, A., Shah, D., Lilge, L., Kabashin, A.V., Meunier, M.M. 2009. Silicon
nanoparticles produced by femtosecond laser ablation in water as novel contamination-free photosensi-
tizers. J. Biomed. Opt. 14: 021010.
Robitaille, A., Boulais, E., Meunier, M. 2013. Mechanisms of plasmon-enhanced femtosecond laser nanoabla-
tion of silicon. Opt. Exp. 21: 9703–9710.
Saikiran, V., Vendamani, V.S., Hamad, S., Nageswara Rao, S.V.S., Venugopal Rao, S., Pathak, A.P. 2014.
150 MeV Au ion induced modification of Si nanoparticles prepared by laser ablation. Nucl. Instrum.
Methods Phys. Res. B 333: 99–105.
Sajti, C.L., Sattari, R., Chichkov, B.N., Barcikowski, S. 2010. Gram scale synthesis of pure ceramic nanopar-
ticles by laser ablation in liquid. J. Phys. Chem. C 114: 2421–2427.
Sakamoto, M., Fujistuka, M., Majima, T. 2009. Light as a construction tool of metal nanoparticles: Synthesis
and mechanism. J. Photochem. Photobiol. C Photochem. Rev. 10: 33–56.
Salminen, T., Dahl, J., Tuominen, M., Laukkanen, P., Arola, E., Niemi, T. 2012. Single-step fabrication of
luminescent GaAs nanocrystals by pulsed laser ablation in liquids. Opt. Mater. Exp. 2: 799–813.
Santagata, A., Guarnaccio, A., Pietrangeli, D., Szegedi, Á., Valyon, J., De Stefanis, A., De Bonis, A. et al.
2015. Production of silver–silica core–shell nanocomposites using ultra-short pulsed laser ablation in
nanoporous aqueous silica colloidal solutions. J. Phys. D Appl. Phys. 48: 205304.
Sasaki, T., Shimizu, Y., Koshizaki, N. 2006. Preparation of metal oxide-based nanomaterials using nanosec-
ond pulsed laser ablation in liquids. Photochem. Photobiol. A Chem. 182: 335–341.
Sau, T.K., Rogach, A.L. 2010. Nonspherical noble metal nanoparticles: Colloid-chemical synthesis and mor-
phology control. Adv. Mater. 22: 1781–1804.
Schwenke, A., Dalüge, H., Kiyan, R., Sajti, C.L., Chichkov, B.N. 2013. Non-agglomerated gold-PMMA nano-
composites by in situ-stabilized laser ablation in liquid monomer for optical applications. Appl. Phys. A:
Mater. Sci. Process. 111: 451–457.
Schwenke, A., Wagener, P., Nolte, S., Barcikowski, S. 2011. Influence of processing time on nanoparticle
generation during picosecond-pulsed fundamental and second harmonic laser ablation of metals in tet-
rahydrofuran. Appl. Phys. A Mater. Sci. Process. 104: 77–82.
Semaltianos, N.G. 2010. Nanoparticles by laser ablation. Crit. Rev. Solid State Mater. Sci. 35: 105–124.
Semaltianos, N.G., Logothetidis, S., Perrie, W., Romani, S., Potter, R.J., Edwardson, S.P., French, P., Sharp,
M., Dearden, G., Watkins, K.G. 2010. Silicon nanoparticles generated by femtosecond laser ablation in
a liquid environment. J. Nanopart. Res. 12: 573–580.
Serkov, A.A., Barmina, E.V., Simakin, A.V., Kuzmin, P.G., Voronov, V.V., Shafeev, G.A. 2015. Generation of
core–shell nanoparticles Al@Ti by laser ablation in liquid for hydrogen storage. Appl. Surf. Sci. 348: 71–74.
Shafeev, G.A. 2008. Formation of nanoparticles under laser ablation of solids in liquids. In Nanoparticles:
New Research, Lombardi, S.L. (Ed.), Nova Science Publishers, Inc. pp. 1–37.
Shah, A., Rahman, L.-U., Qureshi, R., Rehman, Z.-U. 2012. Synthesis, characterization and applications of
bimetallic (Au–Ag, Au–Pt, Au–Ru) alloy nanoparticles. Rev. Adv. Mater. Sci. 30: 133–149.
Shaheen, M.E., Gagnon, J.E., Fryer, B.J. 2013. Femtosecond laser ablation of brass in air and liquid media.
J. Appl. Phys. 113: 213106.
Simakin, A.V., Voronov, V.V., Kricihenko, N.A., Shafeev, G.A. 2004. Nanoparticles produced by laser ablation
of solids in liquid environment. Appl. Phys. A Mater. Sci. Process. 79: 1127–1132.
Simakin, A.V., Voronov, V.V., Shafeev, G.A., Brayner, R., Bozon-Verduraz, F. 2001. Nanodisks of Au and Ag
produced by laser ablation in liquid environment. Chem. Phys. Lett. 348: 182–186.
Siuzdak, K., Sawczak, M., Klein, M., Nowaczyk, G., Jurga, S., Cenian, A. 2014. Preparation of platinum modi-
fied titanium dioxide nanoparticles with the use of laser ablation in water. Phys. Chem. Chem. Phys. 16:
15199–15206.
Sportelli, M.C., Ancona, A., Picca, R.A., Trapani, A., Volpe, A., Trapani, G., Cioffi, N. 2015. Laser ablation
synthesis in solution of nanoantimicrobials for food packaging applications. Mater. Res. Soc. Symp.
Proc. 1804: 37–42.
Spyropoulos, G.D., Stylianakis, M.M., Stratakis, E., Kymakis, E. 2012. Organic bulk heterojunction photo-
voltaic devices with surfactant-free Au nanoparticles embedded in the active layer. Appl. Phys. Lett.
100: 213904.
Stratakis, E. 2012. Nanomaterials by ultrafast laser processing of surfaces. Sci. Adv. Mater. 4: 407–431.
Stratakis, E., Barberoglou, M., Fotakis, C., Viau, G., Garcia, C., Shafeev, G.A. 2009a. Generation of Al
nanoparticles via ablation of bulk al in liquids with short laser pulses. Opt. Exp. 17(15): 12650–12659.
Stratakis, E., Zorba, V., Barberoglou, M., Fotakis, C., Shafeev, G.A. 2009b. Femtosecond laser writing of
nanostructures on bulk Al via its ablation in air and liquids. Appl. Surf. Sci. 255(10): 5346–5350.
420 Semiconductor Nanocrystals and Metal Nanoparticles

Stratakis, E., Zorba, V., Barberoglou, M., Fotakis, C., Shafeev, G.A. 2009c. Laser writing of nanostructures
on bulk Al via its ablation in liquids. Nanotechnology 20(10): 105303.
Syed, H., Podagatlapalli, G.K., Hussian, A., Ahmed, N., Sreedhar, S., Tewari, S.P., Rao, S.V. 2012. Fabrication
of metal nano-entities using ultrafast ablation for SERS, photonics, and biomedical applications. In
International Conference on Fibre Optics and Photonics, OSA Technical Digest (online) (Optical
Society of America, 2012), paper MPo.31.
Sylvestre, J.-P., Kabashin, A.V., Sacher, E., Meunier, M. 2005. Femtosecond laser ablation of gold in water:
Influence of the laser-produced plasma on the nanoparticle size distribution. Appl. Phys. A Mater. Sci.
Process. 80: 753–758.
Sylvestre, J.-P., Kabashin, A.V., Sacher, E., Meunier, M., Luong, J.H.T. 2004a. Stabilization and size control
of gold nanoparticles during laser ablation in aqueous cyclodextrins. J. Am. Chem. Soc. 126: 7176–7177.
Sylvestre, J.-P., Poulin, S., Kabashin, A.V., Sacher, E., Meunier, M., Luong, J.H.T. 2004b. Surface chemistry
of gold nanoparticles produced by laser ablation in aqueous media. J. Phys. Chem. B 108: 16864–16869.
Tagami, T., Imao, Y., Ito, S., Nakada, A., Ozeki, T. 2014. Simple and effective preparation of nano-pulverized
curcumin by femtosecond laser ablation and the cytotoxic effect on C6 rat glioma cells in vitro.
Int. J. Pharma. 468: 91–96.
Tan, D., Lin, G., Liu, Y., Teng, Y., Zhuang, Y., Zhu, B., Zhao, Q., Qiu, J. 2011. Synthesis of nanocrystalline
cubic zirconia using femtosecond laser ablation. J. Nanopart. Res. 13: 1183–1190.
Tan, D., Liu, X., Dai, Y., Ma, G., Meunier, M., Qiu, J.A. 2015. Universal photochemical approach to ultra-
small, well-dispersed nanoparticle/reduced graphene oxide hybrids with enhanced nonlinear optical
properties. Adv. Opt. Mater. 3: 836–841.
Tan, D., Yamada, Y., Zhou, S., Shimotsuma, Y., Miura, K., Qiu, J. 2013a. Photoinduced luminescent carbon
nanostructures with ultra-broadly tailored size ranges. Nanoscale 5: 12092–12097.
Tan, D., Zhou, S., Qiu, J., Khusro, N. 2013b. Preparation of functional nanomaterials with femtosecond laser
ablation in solution. J. Photochem. Photobiol. C Photochem. Rev. 17: 50–68.
Taylor, U., Barchanski, A., Garrels, W., Klein, S., Kues, W., Barcikowski, S., Rath, D. 2011. Toxicity of gold
nanoparticles on somatic and reproductive cells. Adv. Exp. Med. Biol. 733: 125–133.
Tian, X., Guan, X., Luo, N., Yang, F., Chen, D., Peng, Y., Zhu, J., He, F., Li, L., Chen, X. 2014. In vivo immu-
notoxicity evaluation of Gd2O3 nanoprobes prepared by laser ablation in liquid for MRI preclinical
applications. J. Nanopart. Res. 16: 2594.
Tiedemann, D., Taylor, U., Rehbock, C., Jakobi, J., Klein, S., Kues, W.A., Barcikowski, S., Rath, D. 2014.
Reprotoxicity of gold, silver, and gold-silver alloy nanoparticles on mammalian gametes. Analyst 139:
931–942.
Tilaki, R.M., Zad, A.I., Mahdavi, S.M. 2007. Size, composition and optical properties of copper nanoparticles
prepared by laser ablation in liquids. Appl. Phys. A Mater. Sci. Process. 88: 415–419.
Truong, S.L., Levi, G., Bozon-Verduraz, F., Petrovskaya, A.V., Simakin, A.V., Shafeev, G.A. 2007a. Generation
of Ag nanospikes via laser ablation in liquid environment and their activity in SERS of organic mol-
ecules. Appl. Phys. A Mater. Sci. Process. 89: 373–376.
Truong, S.L., Levi, G., Bozon-Verduraz, F., Petrovskaya, A.V., Simakin, A.V., Shafeev, G.A. 2007b. Generation
of nanospikes via laser ablation of metals in liquid environment and their activity in surface-enhanced
Raman scattering of organic molecules. Appl. Surf. Sci. 254: 1236–1239.
Tsuji, T., Kakita, T., Tsuji, M. 2003. Preparation of nano-size particles of silver with femtosecond laser abla-
tion in water. Appl. Surf. Sci. 206: 314–320.
Tsuji, T., Mizuki, T., Yasutomo, M., Tsuji, M., Kawasaki, H., Yonezawa, T., Mafuné, F. 2010. Efficient fabrica-
tion of substrates for surface-assisted laser desorption/ionization mass spectrometry using laser ablation
in liquids. Appl. Surf. Sci. 257: 2046–2050.
Upadhyay, A.K., Inogamov, N.A., Rethfeld, B., Urbassek, H.M. 2008. Ablation by ultrashort laser pulses:
Atomistic and thermodynamic analysis of the processes at the ablation threshold. Phys. Rev. B 78:
045437.
Usui, H., Shimizu, Y., Sasaki, T., Koshizaki, N. 2005. Photoluminescence of ZnO nanoparticles prepared by
laser ablation in different surfactant solutions. J. Phys. Chem. B 109: 120–124.
Vadavalli, S., Valligatla, S., Neelamraju, B., Dar, M.H., Chiasera, A., Ferrari, M., Desai, N.R. 2014. Optical
properties of germanium nanoparticles synthesized by pulsed laser ablation in acetone. Front. Phys. 2: 57.
Vendamani, V.S., Hamad, S., Saikiran, V., Pathak, A.P., Venugopal Rao, S., Ravi Kanth Kumar, V.V.,
Nageswara Rao, S.V.S. 2015. Synthesis of ultra-small silicon nanoparticles by femtosecond laser ablation
of porous silicon. J. Mater. Sci. 50: 1666–1672.
Applications of Metal Nanoparticles and Nanostructures 421

Venugopal Rao, S., Podagatlapalli, G.K., Hamad, S. 2014. Ultrafast laser ablation in liquids for nanomaterials
and applications. J. Nanosci. Nanotechnol. 14: 1364–1388.
Von Der Linde, D., Sokolowski-Tinten, K. 2000. Physical mechanisms of short-pulse laser ablation. Appl.
Surf. Sci. 154: 1–10.
Vorobyev, A.Y., Guo, C. 2005. Enhanced absorptance of gold following multi-pulse femtosecond laser ablation.
Phys. Rev. B 72: 195422.
Vorobyev, A.Y., Guo, C. 2008. Colorizing metals with femtosecond laser pulses. Appl. Phys. Lett. 92: 041914.
Vorobyev, A.Y., Guo, C. 2013. Direct femtosecond laser surface nano/microstructuring and its applications.
Laser Photon. Rev. 7: 385–407.
Vorobyev, A.Y., Makin, V.S., Guo, C. 2009. Brighter light sources from black metal: Significant increase in
emission efficiency of incandescent light sources. Phys. Rev. Lett. 102: 234301.
Wang, C.X., Liu, P., Cui, H., Yang, G.W. 2005. Nucleation and growth kinetics of nanocrystals formed upon
pulsed-laser ablation in liquid. Appl. Phys. Lett. 87: 201913.
Wang, T., Hu, F., Ikhile, E., Liao, F., Li, Y., Shao, M. 2015. Two-step-route to Ag–Au nanoparticles grafted on
Ge wafer for extra-uniform SERS substrates. J. Mater. Chem. 3: 559–563.
Weng, C.-I., Cang, J.-S., Chang, J.-Y., Hsiung, T.-M., Unnikrishnan, B., Hung, Y.-L., Tseng, Y.-T., Li, Y.-J.,
Shen, Y.-W., Huang, C.-C. 2014. Detection of arsenic(III) through pulsed laser-induced desorption/
ionization of gold nanoparticles on cellulose membranes. Anal. Chem. 86: 3167–3173.
Xia, Y., Rogers, J.A., Paul, K.E., Whitesides, G.M. 1999. Unconventional methods for fabricating and pattern-
ing nanostructures. Chem. Rev. 99: 1823–1848.
Yan, Z., Chrisey, D.B. 2012. Pulsed laser ablation in liquid for micro-/nanostructure generation. J. Photochem.
Photobiol. C Photochem. Rev. 13: 204–223.
Yang, G.W. 2007. Laser ablation in liquids: Applications in the synthesis of nanocrystals. Prog. Mater. Sci.
52: 648–698.
Yang, G.W. 2011. Laser Ablation in Liquids: Principles and Applications in the Preparation of Nanomaterials.
Pan Stanford Publishing, Singapore.
Zakharko, Y., Rioux, D., Patskovsky, S., Lysenko, V., Marty, O., Bluet, J.-M., Meunier, M. 2011. Direct syn-
thesis of luminescent SiC quantum dots in water by laser ablation. Phys. Stat. Solidi Rapid Res. Lett.
5: 292–294.
Zavedeev, E.V., Petrovskaya, A.V., Simakin, A.V., Shafeev, G.A. 2006. Formation of nanostructures upon laser
ablation of silver in liquids. Quant. Electron. 36: 978–980.
Zeng, H., Du, X.-W., Singh, S.C., Kulinich, S.A., Yang, S., He, J., Cai, W. 2012. Nanomaterials via laser abla-
tion/irradiation in liquid: A review. Adv. Funct. Mater. 22: 1333–1353.
Zhakhovskii, V.V., Inogamo, N.A., Petrov, Y.V., Ashitkov, S.I., Nishihara, K. 2009. Molecular dynamics simu-
lation of femtosecond ablation and spallation with different interatomic potentials. Appl. Surf. Sci. 255:
9592–9596.
Zhang, L., Mu, K., Zhao, J., Wu, T., Wang, H., Zhang, C., Zhang, X.-C. 2015. Intense thermal terahertz-to-
infrared emission from random metallic nanostructures under femtosecond laser irradiation. Opt. Exp.
23: 14211–14218.
Zheng, Z., Shan, G., Li, J., Chen, Y., Liu, Y. 2014. Au/Ag nano alloy shells as near infrared SERS nano-probe
for the detection of protein. Mater. Res. Exp. 1: 045408.
Zhigilei, L.V., Lin, Z., Ivanov, D.S. 2009. Atomistic modeling of short pulse laser ablation of metals:
Connections between melting, spallation, and phase explosion. J. Phys. Chem. C 113: 11892–11906.
Zimbone, M., Buccheri, M.A., Cacciato, G., Sanz, R., Rappazzo, G., Boninelli, S., Reitano, R., Romano, L.,
Privitera, V., Grimaldi, M.G. 2015. Photocatalytical and antibacterial activity of TiO2 nanoparticles
obtained by laser ablation in water. Appl. Catal. B Environ. 165: 487–494.
 12 Exploring the LaAlO3/SrTiO3
Two-Dimensional Electron Gas
From Fundamental to
Technical Applications
Ngai Yui Chan, Fan Zhang, Kit Au, Wing Chong Lo,
Helen La Wa Chan, and Jiyan Dai

CONTENTS
12.1 Introduction........................................................................................................................... 423
12.2 LAO/STO Preparation and Properties................................................................................... 424
12.3 LAO/STO Device Applications............................................................................................. 428
12.3.1 Polar Liquid Molecule–Induced Transport Property Modulation
at the LAO/STO Heterointerface............................................................................... 428
12.3.1.1 Polar Liquid Sensing................................................................................... 428
12.3.1.2 Schottky-Diode-Like Junction.................................................................... 430
12.3.1.3 Mechanism Discussion............................................................................... 430
12.3.2 Palladium Nanoparticle Enhanced Giant Photoconductivity at LaAlO3/SrTiO3
2DEG Heterostructure............................................................................................... 433
12.3.2.1 Pd/LAO Film Growth................................................................................. 433
12.3.2.2 Surface Morphology................................................................................... 434
12.3.2.3 Photoresponse under UV Light and Room Light....................................... 436
12.3.2.4 Wavelength-Dependent Photoconductivity................................................. 437
12.3.2.5 Time-Dependent Resistance of the LAO/STO Interface............................ 439
12.3.2.6 Current Voltage Characteristics under Different Ambient.........................440
12.3.2.7 Mechanism Discussion............................................................................... 441
12.3.3 Highly Sensitive Gas Sensor by the LaAlO3/SrTiO3 Heterostructure
with Pd NPs Surface Modulation..............................................................................444
12.3.3.1 Gas Sensing Characteristics of Pd NPs–Coated LAO/STO.......................444
12.3.3.2 Response to Other Gases............................................................................ 450
12.3.3.3 Mechanism Discussion............................................................................... 453
12.4 Conclusion and Perspectives.................................................................................................. 456
Acknowledgments........................................................................................................................... 457
References....................................................................................................................................... 457

12.1 INTRODUCTION
“Often, it may be said that the interface is the device,” as stated by Prof. Herbert Kroemer in his
Nobel lecture in 2000 [1]. A rich set of fundamental physical phenomena can be explored in the
complex oxide interface. At the interface, surprising novel properties were discovered, which are
not present in their bulk counterparts. The interface between lanthanum aluminate (LaAlO3) (LAO)
and strontium titanate (SrTiO3) (STO) is a remarkable system that was discovered in 2004 [2];

423
424 Semiconductor Nanocrystals and Metal Nanoparticles

despite both materials being insulators, a quasi-two dimensional electron gas (2DEG) can be found
at their interface. The discovery of 2DEG at the interface between the LAO and STO heterostructure
with intriguing properties such as the ability of nanopatterning [3], magnetism [4], superconduc-
tivity [5], resistance switching behavior [6], and photoconductivity [7] has been studied extensively
in recent years. The study of such properties of the LAO/STO interface is one of emerging areas in
the field of condensed matter physics.
Intense research has been conducted to explain the observed conducting mechanism, including
the electronic reconstruction (polar catastrophe model) [8], the presence of oxygen vacancies [9–11],
and the cation intermixing [12,13] at the interface. However, the exact mechanism to explain these
phenomena is still under debate, and researchers continue to investigate this system.
For natural analogue of conventional semiconductor heterostructure, from the view of electron
mobility, the LAO/STO is not a feasible candidate for high-speed electronic device applications,
with mobility around 5–10 cm2 V−2 s−1 at room temperature and up to 103 cm2 V−2 s−1 at lower tem-
peratures [2]. For comparison, graphene, a well-known two-dimensional material, has an electron
mobility higher than 15,000 cm2 V−2 s−1 at room temperature [14], high-mobility semiconductor
2DEG systems, such as the Ga xAl1−xAs/GaAs heterostructure, have electron mobility 3 orders of
magnitude higher than that of LAO/STO at low temperatures [15], and the molecular beam epitaxy–
grown MgxZn1−xO/ZnO oxide 2DEG system has a mobility exceeding 700,000 cm2 V−2 s−1 [16].
However, the LAO/STO heterostructure, which exhibits novel properties like sharp switching from
insulating/conducting state [17] and optical transparency [18], has the ability for combination with
other oxides that have outstanding properties.
The significant results observed in this system pave way for potential applications in multifunc-
tional oxide-based electronic devices. Even though the discovery of LAO/STO 2DEG was included
in the Science Magazine Top 10 Breakthroughs of the Year 2007 [19], a relatively weak area in this
field of research remains—the device application for such an oxide interface 2DEG. The main dif-
ficulty in applying this 2DEG system in electronic devices is its relatively low response to external
stimulus such as light and gas (since gas cannot react and be absorbed onto the heterostructure).
In this chapter, a method to modify the surface of the LAO/STO heterostructure to significantly
enhance its response to ultraviolet light and gases is introduced. This method has brought this
fascinating 2DEG system closer to real-world applications in sensing devices.
The works presented in this chapter are separated into four main sections, concerning the fab-
rication of the LAO/STO heterostructure and the fabrication of the polar liquid sensor, ultraviolet
sensor, and gas sensor. The study of the LAO/STO heterostructure provides an unexplored regime
to understand its potential for device application.

12.2 LAO/STO PREPARATION AND PROPERTIES

To fabricate the atomically aligned heterostructure or interface, an atomically flat substrate is
required. TiO2-site terminated SrTiO3 substrate was prepared with conventional SrTiO3 (001) sub-
strate, with miscut angle of the substrate smaller than 0.5°. The preparation procedure of the LaAlO3/
SrTiO3 heterostructure is summarized in Figure 12.1. First, the as-received SrTiO3 substrates were
treated with buffered hydrofluoric acid (BHF) by NH4F:HF = 7:1, followed by a heat treatment
process to obtain an atomically flat substrate [20–22]. Then, a LaAlO3 thin film was grown on the
treated SrTiO3 substrate using pulsed laser deposition (the deposition parameters are summarized
in Table 12.1); during the deposition, the growth of the thin film was monitored by reflection high
energy electron diffraction (RHEED). Finally, electrical wires were bonded at the corners (by the
van der Pauw geometry) for subsequent electrical properties measurement.
The roughness of the top layer increases with thin film deposition. A decrease in the intensity
of the RHEED beam is observed. The intensity of the beam reaches a minimum when half of the
monolayer of the LaAlO3 thin film is deposited on top of the SrTiO3 substrate. Afterward, the inten-
sity of the RHEED beam increases until a full monolayer of the LaAlO3 is deposited on the SrTiO3
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 425

LAO
2DEG
Surface Deposition Bonding STO
treatment

FIGURE 12.1 Fabrication process of the LaAlO3/SrTiO3 heterostructure. Surface treatment process was
performed on the as-received SrTiO3 (001) substrate. LaAlO3 thin film was deposited on top of the TiO2-site-
terminated substrate. After the deposition, electrical wires were bonded with the van der Pauw geometry of
the LaAlO3/SrTiO3 heterostructure for subsequent electrical measurement.

TABLE 12.1
Experimental Parameters Used for Thin Film
Deposition
Target to substrate distance 60 mm
Laser energy 250–280 mJ
Repetition rate of the laser 1 Hz
Deposition temperature 750°C
Base vacuum <10–5 Pa

surface. From the RHEED measurement shown in Figure 12.2a, it is observed that each oscillation
corresponds to one monolayer deposition (with thickness ~0.4 nm), so the surface crystallinity of
the sample can be determined. In addition, as shown in Figure 12.2b and c, the electron diffraction
pattern of the bare SrTiO3 substrate and after deposition (LaAlO3/SrTiO3 heterostructure) can be
obtained.
X-ray diffraction (XRD) and transmission electron microscopy (TEM) were used to investigate
the quality of the LaAlO3 thin film and its epitaxial relationship with the SrTiO3 substrate after
deposition. XRD analysis is one of the nondestructive techniques for analyzing a wide range of
materials. Figure 12.3 shows a θ–2θ scan of the 20 unit cells (uc) (~8 nm)-thick LaAlO3 thin film
on the TiO2-site-terminated SrTiO3 substrate in the range of 2θ = 18°–28°, where only the (001)
reflection of the LaAlO3 can be observed with the (001) reflection of SrTiO3 substrate, indicating a
c-axis-oriented growth. In the XRD diffractogram, Kiessig fringes were obtained, resulting from
the constructive and destructive interference between the film and the film–substrate interface,
which indicates an abrupt and highly ordered crystalline structure between the LaAlO3/SrTiO3
interfaces.
The LaAlO3/SrTiO3 heterostructure was characterized by cross-sectional TEM. The cross-
sectional TEM sample was prepared by mounting the LAO/STO sample face to face and mechani-
cally grinding the cross section to 20 μm and then finally ion-milling to a thickness below 100 nm.
As shown in Figure 12.4, the LaAlO3–SrTiO3 interface is atomically sharp and abrupt, implying
that the surface of the STO is atomically flat. The bright dots representing LaAlO3 unit cells can be
clearly identified, from which the number of layers can be counted accurately with an uncertainty of
one unit cell. The LAO thin film, as shown in the figure, has a thickness of around 20 uc.
426 Semiconductor Nanocrystals and Metal Nanoparticles

RHEED Osc. of sample 24 (10 uc)

1.1
1.0 (b)
0.9
RHEED I/I0 (a.u.)

0.8
0.7
0.6
0.5
0.4
(c)
–2 0 2 4 6 8 10 12
(a) Number of units cells

FIGURE 12.2 (a) RHEED oscillations during the growth of LaAlO3 thin film on top of the SrTiO3 substrate
with 10 unit cell (uc) thickness in LaAlO3. (b) RHEED diffraction along the (001) plane of the SrTiO3 before
the deposition. (c) RHEED diffraction of the LaAlO3 plane after the deposition, which confirms the crystal-
linity of the film.

STO (001)
Intensity (a.u.)

18 20 22 24 26 28
2θ

FIGURE 12.3 XRD pattern of the LaAlO3/SrTiO3 heterostructure.

A variety of electrical property characterizations can be conducted by employing the Physical

Properties Measurement System. Sheet resistance versus temperature can be measured while the
temperature of the system varies during measurement. The resistance measurement of the sample
was constructed in the van der Pauw geometry [23] for a sample with size of 5 × 5 mm2, where
four aluminum wires were ultrasonically bonded on the sample corners and ohmic behavior was
observed from the linear current–voltage characteristics.
To clarify that the as-grown LAO/STO sample at the interface is a standard metallic conduc-
tor, its sheet resistance and mobility as a function of temperature was measured as shown in
Figure 12.5. To reduce the effect of photoexcitation of excess charge carrier, the experiment was
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 427

20
STO 19
18
17
16
15
14
13
12
11
10
LAO 9
8
7
6
5
4
3
2
1

5 nm

FIGURE 12.4 High-resolution TEM image of the LAO/STO interface. The number of unit cells (around
20 uc) can be counted.

104
Rxx (Ω/sq.)

103

102
1 10 100
(a) Temperature (K)

103
Mobility (cm2 V−2 s−1)

102

101

1 10 100
(b) Temperature (K)

FIGURE 12.5 (a) Sheet resistance and (b) the mobility of the LAO/STO sample as a function of temperature.
428 Semiconductor Nanocrystals and Metal Nanoparticles

performed by shielding the sample from any light for 1 day before the experiment. It should be noted
that, although the structures of LaAlO3/SrO-site-terminated SrTiO3 and of LaAlO3/TiO2-site-
terminated SrTiO3 heterostructure are very similar, different electronic properties can be measured.
The former gives insulating properties while the latter shows metallic properties. In this thesis,
the LaAlO3/TiO2-site-terminated SrTiO3 heterostructure is investigated; the n-type sheet carrier
charge density is measured to be ns = 2 × 1013 cm−2 by a Hall effect measurement. As shown in the
figure, sheet resistance decreases with temperature, and the mobility of the interface increases up
to μ = 103 cm2 V−1 s−1 at 2.5 K.

12.3 LAO/STO DEVICE APPLICATIONS

12.3.1 Polar Liquid Molecule–Induced Transport Property
Modulation at the LAO/STO Heterointerface
The conducting mechanism behind the LAO/STO interface was reconciled by the intrinsic polar
catastrophe model [2,17]. In this model, a proposed charge transfer from the LAO layer (includ-
ing surface) to the interface is associated with a buildup of the internal electric field in the LAO
layer, which originates from the polar discontinuity between LAO (alternating charged layers) and
STO (charge-neutral layers). Such an internal electric field, also predicted from first principles
calculations for STO/LAO/vacuum stacks [24,25], was evidenced by various experiments [26–28].
For example, an experimental observation of an expansion of the LAO c-axis in ultrathin layers
indicated strong electrostrictive effect produced by the dielectric LAO [29]. Recent experimental
reports also clearly revealed the importance of the correlation between the interface and the sur-
face charge states, making it possible to manipulate the interfacial conduction using a conducting
atomic force microscopy technique [6,30] by handling with the surface charges and adsorbates
[31–33], or a capping layer of STO [34]. The fact that the interface conducting states can be highly
sensitive to surface adsorbates leads to a strong possibility of LAO/STO interfaces functioning as
polar molecule sensors.
Recently, Xie et al. [35] reported that surface adsorption of polar liquids, such as water, induced
a huge change (by factor of 3) in the conductivity at LAO/STO interfaces. This conductivity change
has been attributed to an increase in sheet carrier density of the 2DEG by more than 2 × 1013 cm−2,
suggesting that the adsorbent has a great influence on charge transfer from the film surface to
the interface. These results also suggested that sensor application is possible utilizing the surface
interface coupling-induced conductivity modulation. In this chapter, the room temperature cur-
rent–voltage (I–V) relationships of the LAO/STO interface using an in-plane field effect transistor
device structure have been characterized, where the polar molecules act as a gate voltage to affect
the source/drain current across the interface channel. The local conductivity of the buried inter-
face changes dramatically from metallic to semiconducting when a droplet of the polar liquid such
as water is put onto the film surface. These observations provide a direct evidence of the built-in
electric field in LAO layer and demonstrate that such a fascinating interface system can be used for
high-performance polar molecule sensors.

12.3.1.1 Polar Liquid Sensing

I–V curves of the interface were measured before and after placing a droplet of liquid at the center
of the sensor. A variety of liquids, including water, acetone, gasoline, hexane, etc., have been tested.
Figure 12.6a shows the I–V curves with and without a water droplet on top of the LAO surface. It
can be seen that for the sample tested in air, the I–V characteristic is typically linear with perfect
Ohmic behavior, illustrating a metallic conduction at the LAO/STO interface. When a drop of
deionized (DI) water is added to the exposed LAO surface, the I–V curves change to a typical
field-effect transistor (FET) source–drain type current with a strong saturation at high voltages.
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 429

20 Original
With DI water

10

I (μA) 0
Electrode
Water
–10
PC
Bb
oar
–20 d
LAO/STO

–10 –5 0 5 10
(a) V (V)

H H
Water droplet VD
O

LAO

2DEG
STO

(b)

FIGURE 12.6 (a) I–V curves of the LAO/STO interface with/without deionized (DI) water droplet on top of
LAO surface. The inset is a schematic diagram showing the sensor structure. (b) Schematic diagram illustrat-
ing water molecule alignment along the electrostatic field direction on the sample surface. The larger arrow
indicates relatively larger polarization field, and the smaller arrow represents smaller polarization field in
the LAO layer.

This metal/semiconductor/metal junction-like behavior indicates that there is a strong field effect
induced by water molecules that changes the local interfacial conduction from metallic to semicon-
ducting. Since LAO is a polarized ionic oxide with the polarization pointing from the substrate to the
­surface [35], water molecules, which are polar, have a tendency to align their dipole moments with
the electrostatic field to reach the most stable state (Figure 12.1b). As also illustrated in Reference
35, the water molecules’ alignment weakens the local electric potential in LAO and thus modifies
the band structure of LAO. For samples with water droplet on top of the LAO surface, the weaken-
ing of the local electric field in the LAO layer underlying the water droplet results in a transition
of the interfacial transport property from metallic 2DEG to n-type semiconducting dominated, as
shown in Figure 12.6b. Therefore, a metal/n-type semiconductor/metal junction, analogous to a nar-
row channel field-effect transistor, can be formed, whose I–V characteristic far deviates from linear
relation. The liquid was kept on top of the sample surface while the measurements were performed.
The blowing off of all visible liquid immediately after placing the droplet recovers the I–V curve to
its perfect Ohmic behavior with a smaller resistance. After the sample was heated to above 373 K,
the linear I–V characteristics resumed. It was also noted that nonpolar liquids such as gasoline and
hexane have no obvious influence on the I–V curve (not shown since they are exactly the same as
that without water droplet).
430 Semiconductor Nanocrystals and Metal Nanoparticles

12.3.1.2 Schottky-Diode-Like Junction

It is interesting to see that the 2DEG is sensitive to the position of water droplet on the LAO sur-
face. When the water droplet is placed off from the device center (for this experiment, there is no
insulating parylene coating), the I–V curve becomes asymmetric. Figure 12.7a shows a rectified I–V
characteristic when the water droplet is close to one side of the electrodes. A Schottky-diode-like
junction is realized (Figure 12.7a) in this polar molecule–modulated oxide interface. A schematic
diagram is shown in Figure 12.7b to illustrate the mechanism of forming the Schottky junction.

12.3.1.3 Mechanism Discussion

Figure 12.8 illustrates the band diagram qualitatively demonstrating the proposed conducting
mechanism across the junctions of 2DEG and n-type semiconducting LAO/STO interface under the
water droplet. Figure 12.8a shows the state when water droplet is placed at the center of the sensor
without applying a bias voltage VD, where two Schottky junctions are formed; while under a VD, the
band structure changes, and the current is restricted by the thickness of the depletion layer through a
tunneling mechanism (Figure 12.8b). In the case when water droplet is placed at one side closing to
the electrode, where one Schottky junction is formed between the metallic and n-type semiconduct-
ing LAO/STO interface, the current is rectified. Figure 12.8c and d demonstrates the band structure
of this junction. No saturation was observed when positive VD was applied. This indicates that, for

30 Original
Water drop touch to positive wire
20

10
I (μA)

–10

–20

–30 T = 298 K

–10 –5 0 5 10
(a) V (V)

H H Water droplet VD
O

LAO

2DEG STO

(b)

FIGURE 12.7 (a) I–V curves of the LAO/STO interface with water on top of LAO surface but close to one
end of electrode. (b) Schematic diagram illustrating water molecule alignment along the electrostatic field
direction on the sample surface.
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 431

(a) (b)

εF
εF
2DEG 2DEG 2DEG
n-type semi- n-type semi-
conductor 2DEG
conductor

(c) (d)
εF
εF
n-type semi-
2DEG n-type semi- 2DEG conductor
conductor

VD > 0 VD < 0

FIGURE 12.8 Schematic band diagrams of Schottky junctions formed along the LAO/STO interface of the
sample with water droplet at the center of the sample for (a) zero bias voltage and (b) nonzero bias voltage.
(c, d) Schematic band diagrams when water droplet is close to one side of the electrode: (c) when V D > 0 and
(d) when V D < 0.

the 2DEG, the Fermi level of the metallic region is higher than that of the n-type semiconducting
region, which is different from a typical Schottky diode. The typical charge carrier density of 2DEG
at the LAO/STO interface, usually with a value of ~1013 cm−2, is much lower than that of normal
metals. As a result, a nonnegligible local depletion layer in metallic 2DEG region can be formed
near the junction. In this scenario, when VD > 0, as shown in Figure 12.8c (2DEG is negatively
biased), the depletion layer in the metallic region is suppressed. Therefore, a quasi-Ohmic behavior
can be observed. While for VD < 0, as shown in Figure 12.8d (2DEG is positively biased), the cur-
rent is likely to be limited due to the locally formed barrier in metallic region. As a consequence, a
“rectifier” is realized.
In order to study the tuning effect to the transport property of the 2DEG by water molecules
when they change from liquid to solid (ice), the water temperature was cooled below its ice point
by adding liquid nitrogen (LN2) around the sample. As can be seen in Figure 12.9, after the water
droplet is frozen to LN2 temperature, the linear I–V characteristic of the interface resumed, but with
a larger slope (lower resistance) due to the decreased temperature. By comparing the result to a
control sample, where there is no water droplet on the surface, one can see that, at LN2 temperature,
ice has no effect to the I–V characteristics. Figure 12.10 shows the curve for the measured current
when we freeze the water droplet to ice at LN2 temperature and then melt ice back to water, where
the ice and melting points are indicated. During the current measurement, a fixed bias voltage of
10 V at VD is applied.
In order to more clearly understand the mechanism of the 2DEG tuning by water molecules at
difference temperatures, water temperature is adjusted below its ice point with a cryostat placed
around the sample. Figure 12.11 shows temperature-dependent I–V curves of the LAO/STO inter-
face with water droplet/ice on top. One can see that when water transforms to ice at a temperature
of 260 K, which is lower than ice point, the I–V curve shows a hysteresis characteristic, when it is
linear when the voltage is less than 20 V, then decreasing when voltage increases. This phenomenon
may be due to the melting of the most interfacial layer of water molecules due to Joule heating
induced by current; while at a temperature much lower than the ice point, there is negligible hyster-
esis. Other unknown reasons may also be responsible for this interesting phenomenon, and further
432 Semiconductor Nanocrystals and Metal Nanoparticles

80
2.0
1.5
60 1.0
0.5

I (mA)
0.0
40 –0.5
–1.0 Sample A
–1.5
Sample B
20 –2.0
–20 –10 0 10 20
V (V)
I (μA)

–20

–40
Sample A
–60 Sample B
Sample with water drop
–80
–20 –10 0 10 20
V (V)

FIGURE 12.9 I–V curves of the LAO/STO interface for samples A (control) and B with water on top of
the LAO surface. The inset is I–V curves of samples A and B at LN2 temperature, where the water droplet in
sample B changes to ice. The two curves are overlapped and indistinguishable.

16 4
1. Without water
14 2. Water on LAO surface
3. Water changed to ice
1
12 4. Ice on LAO surface
5. Ice melted
10 6. Water on LAO surface
I (μA)

5
8
6
6

4
3
2
VD = 10 V
2

0 50 100 150 250 250 300

Times (s)

FIGURE 12.10 Current–time curve of the LAO/STO interface with water/ice on top of the sample surface
when a fixed voltage is added on VD.

study is needed. Theoretically, from the first principle calculations, the adsorption of a thin layer
of polar water molecules on the surface of LaAlO3 can remarkably enhance the carrier density of
the interfacial 2DEG by at least 50%, which is qualitatively consistent with reported experimental
results [36].
These results suggest that this type of sensor may be extended to applications in polar gas molecule
sensors if a catalytic layer can be coated on the sensor surface. Another application as a biosensor,
such as a DNA sensor, can be expected due the fact that many types of DNA carry negative charges.
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 433

200

150

100

50

0
I (μA)

–50

–100 Bare sample at 290 K

A drop of water at center at 290 K
–150 A drop of ice at center at 270 K
A drop of ice at center at 260 K
–200 A drop of ice at center at 240 K

–60 –40 –20 0 20 40 60

V (V)

FIGURE 12.11 Temperature-dependent I–V curves of the LAO/STO interface with water or ice on top of
the LAO surface.

12.3.2 Palladium Nanoparticle Enhanced Giant Photoconductivity

at LaAlO3/SrTiO3 2DEG Heterostructure

12.3.2.1 Pd/LAO Film Growth

A LAO film was deposited on a TiO2-terminated SrTiO3 (001) substrate with a size of 5 mm × 5 mm
by laser molecular beam epitaxy (laser-MBE) using a krypton flouride (KrF) Excimer laser with
wavelength 248 nm, at a repetition rate of 1 Hz, and a single crystalline LaAlO3 target [37]. During
deposition, the substrate temperature was maintained at 750°C with base vacuum lower than 2 × 10–5
Pa; the growth was monitored by reflection high energy electron diffraction (RHEED). In order to
reduce oxygen vacancies in the film, after deposition, the samples were annealed in situ at a reduced
temperature of 550°C at 1000 Pa O2 for 1 h. The sample was then cooled down to room tempera-
ture in the same ambient. After deposition, Hall and sheet resistance measurements were made to
characterize the 2DEG nature of the interface. Palladium nanoparticles (NPs) were deposited on
LAO/STO surface using DC magnetron sputtering under a power of 15 W at room temperature with
Ar gas (99.995%) of 10 sccm flow rate and a base pressure of 100 mTorr. The deposition time was
optimized to make sure that the Pd NPs did not form a complete conducting path on the LAO/STO
surface. A TEM grid with carbon film was placed close to the sample during the deposition of pal-
ladium for subsequent TEM structural analysis (since it is room temperature growth, the Pd NPs on
the carbon film can roughly represent the structure of the Pd NPs on LAO surface). High-resolution
TEM was used for the nanoparticles analysis. The size of the Pd NPs may affect the experimental
result. Figure 12.12 shows the TEM images for the Pd NPs deposited by our magnetron sputtering
system for 6 and 15 s. As can be seen in Figure 12.12a and b, the Pd NPs deposited for 6 s gave
the best uniformity with sizes close to ~2 nm. Increasing the deposition time of Pd NPs up to 15 s,
as shown in Figure 12.12c and d, increases the size of the Pd NPs but is not uniformly distributed
(ranging from 2 to 10 nm). Therefore, to minimize the effect of Pd size distribution to the electrical
performance, we choose to use Pd NPs with a mean size of ~2 nm. Increasing the time of Pd deposi-
tion results in a Pd thin film, which can conduct electricity.
The samples were ultrasonically wire-bounded with Al wires and the two-probe electrical char-
acteristics were measured by the electrometer under a DC bias of 10 V in a stainless steel vacuum
434 Semiconductor Nanocrystals and Metal Nanoparticles

50 nm 10 nm

(a) (b)

50 nm 10 nm

(c) (d)

FIGURE 12.12 (a) Low-magnification TEM images of the Pd NPs deposited by magnetron sputtering for
6 s, (b) high-resolution images showing the uniformly distributed Pd NPs with size around ~2 nm, (c) low
magnification TEM images of the Pd NPs deposited for 15 s, and (d) the corresponding high-resolution images
showing Pd particles with various sizes.

chamber with oxygen, dry air, and argon, respectively. Oxygen and argon gases are of 99.9995%
purity. Ultraviolet light source was induced into the chamber for photoresistance measurement.
Lights with different wavelengths were generated with a standard system equipped with a mono-
chromator and a dual-channel power meter.
LaAlO3 thin films with 5, 10, 20, and 40 uc thicknesses were deposited on TiO2-terminated
STO (001) substrates, where the growth conditions were the same as those reported elsewhere
[38]. After the deposition, Pd NPs were deposited on top of the surface of the LAO/STO hetero-
structure by means of DC magnetron sputtering at room temperature. This room-temperature-
grown Pd NPs layer has very weak adhesion and can be removed easily by surface cleaning using
cotton tips with methanol, and thus should not lead to an epitaxial growth on the LAO layer.
Control samples (LAO/STO) without Pd NPs deposition were also fabricated as a reference. Pd
NPs were also deposited on TiO2-site-terminated STO substrate (without LAO) showing that Pd
NPs were well below the conduction percolation threshold and no photoconductivity effect was
observed.

12.3.2.2 Surface Morphology

Figure 12.13a is an AFM image showing the surface morphology of the Pd NPs–coated LAO/STO
sample. The disappearance of surface terrace suggests the coverage of LAO surface by Pd NPs.
Figure 12.13b is a HRTEM image of Pd NPs on carbon film, where it can be seen that the mean
particle size of the Pd NPs is ~2 nm and particles are uniformly distributed. In order to confirm the
existence of Pd NPs on LAO surface and to know their structural characteristics, a TEM carbon
film is used together with LAO/STO when coating the Pd NPs. The high-resolution images of the
Pd NPs also show clear lattice fringes of Pd, and the inset selected area electron diffraction (SAED)
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 435

2.0 nm

0 nm
2 μm

(a)

(331)
(311)
(220)
(200)

(111)

0.22 nm

10 nm

(b)

FIGURE 12.13 (a) AFM image of Pd NPs–modulated LAO surface on STO substrate. (b) HRTEM images
showing uniformly distributed Pd NPs with an average size of around 2 nm. The lattice spacing of the particles
is identified to be around 0.22 nm corresponding to Pd (111) atomic plane. The inset is a SAED pattern show-
ing the polycrystalline structure of the Pd NPs.

pattern can be indexed as the nanocrystalline fcc structure of Pd. (Figure 12.14 shows the surface
morphology of the bare TiO2-terminated STO substrate and LAO/STO surface by AFM where clear
terrace can be observed.) Figure 12.15a and c shows the morphology of the Pd NPs on the carbon
film, and the energy dispersive x-ray analysis result from Pd NPs is shown in Figure 12.15b and d.
It is noted that, the Pd NPs on top of the LAO/STO surface can be removed easily using cotton
tips. Figure 12.15c and d shows the TEM analysis and EDS results for the Pd NPs after removing
the Pd NPs by cotton tips. The Pd NPs show the same morphology as what we observed in Figure
12.15a, and the EDX result is same as shown in Figure 12.15b. The actual surface morphology has
been studied by preparing cross-sectional TEM samples for the Pd/LAO/STO structure. However,
the picture obtained is not very meaningful since during sample preparation, where rinsing and
heating the substrates were inevitable, Pd NPs were rinsed away or aggregate during the treatment.
436 Semiconductor Nanocrystals and Metal Nanoparticles

0.5 nm 0.6 nm

0 nm 0 nm
2 μm 2 μm

(a) (b)

FIGURE 12.14 (a) Bare TiO2-site-terminated STO (001) substrate. (b) Surface morphology of the surface
after 5 uc LAO thin film deposition; clear terraces can still be observed.

(a) (c)

50 nm 50 nm

Pd
(b) Spectrum 4 (d) Spectrum 2

Cu

Pd
Cu

Cu
Cu Cu Pd
Cu Pd

0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20
Full Scale 253 cts Cursor. 20.232 (1 cts) keV Full Scale 480 cts Cursor. 20.064 (3 cts) keV

FIGURE 12.15 (a) TEM analysis of the Pd NPs on the carbon grid showing uniform distribution. (b)
Elemental analysis was performed with energy dispersive system (EDX); the two significant peaks at low
energy correspond to C-Kalpha and O-Kalpha. The existence of palladium atoms in the nanoparticles can be
observed in the peak listed in 2.4 eV. (c) TEM analysis of Pd NPs removed with cotton tips from the sample’s
surface. (d) The corresponding EDX analysis.

12.3.2.3 Photoresponse under UV Light and Room Light

With LaAlO3 surface modification by Pd NPs, LaAlO3/SrTiO3 (LAO/STO) interfacial 2DEG pres-
ents giant optical switching effect with photoconductivity on/off ratio as high as 750% under UV
light irradiation for wavelength shorter than 400 nm. This giant optical switching behavior can be
explained by the Pd nanoparticle’s catalytic effect and surface/interface charge coupling, which will
be discussed in the following section.
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 437

Photoresponses of bare (i.e., without Pd NPs coating) and Pd NPs–coated LAO/STO samples
under a repeatable on/off UV light (365 nm) are shown in Figure 12.16. It can be seen that when
the UV light is turned on, both samples change from a high-resistance state to low-resistance state;
when the light is off, the resistance of the interface increases as a function of time. The resistance of
the interface decreases within 0.1 s after the UV light is irradiated on both samples, which is then
followed by a slow photorecovery process. The increase in resistance of the Pd NP–coated sample at
each time interval is much higher than the bare sample. Within 2 min after the light is switched off
in each cycle, the resistance is yet to be saturated; this can be attributed to the persistent photocon-
ductivity effect [39–41]. This result reveals that after the surface modification by Pd NPs, the LAO/
STO interfacial 2DEG becomes more sensitive to UV light, making it possible for its application as
UV light sensor.
The interface of the Pd NPs–coated LAO/STO sample is also very sensitive to UV radiation from
tube lamps, and Figure 12.17 shows the typical photoemission of the conventional tube lamp, where
UV radiation with wavelength lower than 365 nm can be observed. Figure 12.18 shows the response
of the Pd NPs–coated LAO/STO sample under room light, and around 30% change in resistance can
be observed under on/off room light conditions. Overexposure to fluorescent tube lamp (FT) was
found to be harmful to humans [42,43]. The phosphor coating inside a typical fluorescent tube lamb
absorbs UV radiation and emits a broad range of visible light; however, not all UV radiation can be
absorbed by this phosphor coating.

12.3.2.4 Wavelength-Dependent Photoconductivity

To understand the mechanism of such Pd NPs–induced enhancement effect, wavelength and time
dependences of the photoconductivity for the two samples were characterized. Figure 12.19 illus-
trates the wavelength-dependent photoconductivity of the LAO/STO samples (with and without Pd
NPs on surface) at room temperature. The photoconductivity of the heterostructure was measured
at 10 V bias, and the samples were irradiated by light with wavelength ranging from 700 to 300 nm
(from low photon energy to high photon energy). The samples were put in the dark environment
before measurement and the measurement started after the resistance reached a stable state. It is
noted that, in the dark environment, the resistance of the Pd NPs surface–modulated LAO/STO
heterostructure (~3 MΩ) is much higher than that of the bare LAO/STO 2DEG, which is ~0.25 MΩ.
The change in the photoconductivity is defined by (Iphoto−Idark/Idark) × 100%, where Iphoto corresponds

6
UV on Pd NPs
Bare
5
Resistance (×105 Ω)

2
UV off

1
0 200 400 600 800 1000 1200
Time (s)

FIGURE 12.16 Photoresponse characteristics of the Pd NPs–coated LAO/STO (red) and LAO/STO (black)
heterostructure showing the reversible switching behavior under periodic illumination of a 365 nm UV light
with incident power densities 10 mW/cm2.
438 Semiconductor Nanocrystals and Metal Nanoparticles

Fluorescent lamp
365 nm

Intensity (a.u.)

313 nm

334 nm

250 300 350 400

Wavelength (nm)

FIGURE 12.17 Photoemission spectra of the conventional tube lamp, where UV emission with wavelength
lower than 365 nm can be observed.

30

28 FT light on

26
Resistance (×105 Ω)

24

22

20

18
FT light off
16

500 1000 1500 2000

Time (s)

FIGURE 12.18 Photoresponse characteristics of the Pd NPs–coated LAO/STO heterostructure showing the
reversible switching behavior under periodic illumination of a fluorescent (FT) light.

to the photocurrent under light illumination and Idark corresponds to the dark current. Strong pho-
toconductivity is observed at ~380 nm, while the increase in photoconductivity for both samples
in the range of wavelength 420–700 nm is not obvious. The change in photocurrent for the bare
sample at 380 nm is only 18% (a small increase in 420–700 nm is due to the surface trap state), but
for the Pd NPs–coated LAO/STO sample, the change is 750%, which is a giant photoconductivity. It
is worth noting that the 380 nm UV light is around the band-gap energy of STO substrate (3.3 eV),
suggesting that the increase of photocurrent is due to the absorption of photon energy by STO and
the generation of electron–hole pairs; while the electrons diffuse to the interface as a contribution to
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 439

800
Pd NPs
700

600
20 Bare
500
15

ΔI/I0 (%)
ΔI/I0 (%)
400 10

300 5

0
200
300 400 500 600 700
100 Wavelength (nm)

300 400 500 600 700

Wavelength (nm)

FIGURE 12.19 Wavelength-dependent photoresponse, that is, change of photocurrent for the Pd NPs–coated
and bare (inset) LAO/STO samples under 10 V bias.

the carriers. By contrast, for the photons with energy below the bandgap energy, only very limited
number of charge carriers can be generated corresponding to the mid-gap defect states of the STO
substrate from oxygen vacancies during the deposition of the thin film.

12.3.2.5 Time-Dependent Resistance of the LAO/STO Interface

In order to investigate the effect to photoconductivity in the UV range, time-dependent resistance of
the interface for the bare and Pd NPs–coated LAO/STO samples were measured at different ambi-
ent conditions (oxygen, dry air, and argon). To circumvent the scattering of the data from sample to
sample and rule out the effects of the unwanted experimental factors, the Pd NPs–coated LAO/STO
heterostructure was prepared from the same piece of sample and the Pd NPs were grown under the
same condition for all measurements. As shown in Figure 12.20a, for the recovery process of the
LAO/STO sample, the resistance of the 2DEG increases less than 30% in an hour after switching
off the UV light in the oxygen and argon environments, suggesting that the change of gas does not
affect the transport properties of the interface. However, very different electrical responses for the
Pd NPs–coated LAO/STO sample under oxygen, dry air, and argon were observed, as shown in
Figure 12.20b. With argon gas in the chamber, resistance of the Pd NP/LAO/STO sample recovers
very slowly, with a recovery speed similar to that in the bare sample, as shown in Figure 12.20a.
When the gas ambient is changed from argon to dry air, the speed of the recovery process increases
as the result of oxygen adsorption where Pd NPs are treated as the available surface adsorption
sites for oxygen molecules. Further studies show that recovery of resistance is even faster under
pure oxygen gas because more oxygen molecules are available for chemisorption on the Pd NPs
surface. Figure 12.20a and b shows the significant difference of electrical response between the Pd
NPs–modulated and the bare LAO/STO samples, suggesting enhanced photoconductivity upon UV
light illumination, that is, the Pd NPs on the LAO/STO heterostructure significantly enhance the
photoconductivity of the interface.
The curve under a different environment has been fitted using the stretched exponential relax-
ation law as listed in Equation 12.1.
b
ætö
I photo (t ) = I photo (0)exp - ç ÷ (12.1)
ètø
440 Semiconductor Nanocrystals and Metal Nanoparticles

0.26

0.25
O2
0.24

Resistance (MΩ)
Ar

0.23
UV off
0.22

0.21

0.20
20 40 60 80 100 120
(a) Time (min)

5.0
4.5
4.0
3.5 UV off O2
Resistance (MΩ)

Air
3.0
Ar
2.5
2.0
1.5
1.0
0.5
0.0
200 400 600 800 1000
(b) Time (min)

FIGURE 12.20 Photorecovery process in oxygen, dry air, and argon in bare (a) and Pd NPs-coated (b) LAO/
STO samples at room temperature.

The results are shown in Figure 12.21, and the fitting parameters are listed in Table 12.2,
where Iphoto(t) is the photocurrent of the interface at specific time t, τ is the characteristic relaxation
time, and β is the stretching parameter with value between 0 to 1. For β close to a value of 1, it shows
a Debye-like behavior, a model used to describe relaxation behavior. The rate of change of the resis-
tance is initially rapid, but it becomes continually slower as time progresses.

12.3.2.6 Current Voltage Characteristics under Different Ambient

The I–V characteristics for the bare and Pd NPs–coated LAO/STO samples were measured under
the same condition. Before the measurement, both samples were put in the dark for 24 h. The I–V
curves for bare LAO/STO sample in the dark and under UV light are shown in Figure 12.22a, where
ohmic conducting behavior of the interface can be seen, and the resistances were found to be 26 kΩ
in the dark and 20 kΩ under UV light. The I–V curve of the bare sample measured under normal
room light illumination is almost the same as shown in Figure 12.22a, showing less change in resis-
tance. Figure 12.22b shows the I–V curves for the Pd NPs–coated sample measured in the dark,
room light, and UV irradiation. It is apparent that the I–V curves will exhibit very large changes
in slope under different illuminations, indicating a significant change in resistance at the interface.
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 441

1.0

0.8

Ippc(t)/Ippc(0)
0.6
Ar (c)

0.4
Air (b)
0.2
O2 (a)
0.0
0 20 40 60 80 100
Time (min)

FIGURE 12.21 Photorecovery process in (a) oxygen, (b) dry air, and (c) argon. Ippc refers to the persistent
photocurrent.

TABLE 12.2
Characteristic Relaxation Time τ
and Stretching Parameter β Fitted
by Iphoto(t) = Iphoto(0)exp−(t/τ)β in
Different Ambients
τ (min) β
In oxygen 17.47 0.498
In dry air 39.57 0.520
In argon 201.94 0.371

12.3.2.7 Mechanism Discussion

To explain this enhanced photoconductivity effect of the Pd NPs–coated sample, we propose the
following model (as shown in Figure 12.23). It is believed that the presence of Pd NPs and the
resultant charge coupling and exchange with the oxygen-deficient LAO film facilitate the interfacial
redox reaction at room temperature [10,44–46]. From the view point of electronic band structure,
the work function of the Pd NPs relative to the electron affinity of STO determines the transfer of
electrons from the interface to the Pd NPs. Metals such as Pd with work function (5.6 eV) larger than
the electron affinity of STO (4.0 eV) result in a strong suppression of 2DEG carrier densities at the
interface due to the reduction of an internal built-in electric field in the LAO layer; a similar effect
has been observed in a recent first principle calculation [47–49]. In addition to this Pd NPs electron
affinity–induced degradation of the 2DEG, the Pd NPs catalytic electrochemical reaction with oxy-
gen molecules is believed to play a more important role in reducing the sheet carrier density (ns) of
the LAO/STO 2DEG. The measured ns for bare sample is ~6.5 × 1013 cm−2 under UV irradiation
and ~1.2 × 1013 cm−2 in the dark; while for the Pd NPs–coated LAO/STO, ns under UV irradiation
is measured to be ~3.9 × 1013 cm−2 and in the dark it is ~3.4 × 1012 cm−2.
To compare the conducting behavior between the bare LAO/STO and Pd NPs–coated LAO/STO
samples, the resistance–temperature ratios of the samples were measured from room temperature
down to 15 K in the “dark” and “under UV irradiation” environments; the UV light with wave-
length 365 nm was irradiated on the sample during cooling, and the results are shown in Figure 12.24.
442 Semiconductor Nanocrystals and Metal Nanoparticles

90

60

30

I (μA)
0

–30

–60
UV
–90 Dark

(a) –20 –10 0 10 20

60

40

20
I (μA)

–20
UV
–40 FT
Dark

–60
–20 –10 0 10 20
(b) Voltage (V)

FIGURE 12.22 I–V curves of (a) bare LAO/STO interface in the dark and under UV light and (b) Pd NPs–
coated LAO/STO in the dark, under fluorescent tube lamp (FT), and UV light.

One can see from Figure 12.24a and b that the bare LAO/STO sample shows metallic behavior
under UV irradiation and the dark environment. For the Pd NPs–coated sample, the interface shows
metallic behavior under UV irradiation as shown in Figure 12.24c, while in the dark (as shown in
Figure 12.24d), the conducting behavior of the interface changes to insulating. This result suggests
that the presence of metallic elements on the LAO layer affects the electrical properties of the inter-
face. It is well known that Pd NPs are very sensitive to the surrounding environment due to the large
surface-to-volume ratio and the Pd NPs surface usually absorbs oxygen molecules when exposed
to air [50,51]. The Pd NPs catalytically activate the dissociation of molecular oxygen, and due to
the catalytic effect, the interaction between oxygen molecules from the ambient and Pd NPs results
in the dissociation of O2, that is, O2 + 2e− → 2O−. A similar phenomenon has been observed in Pd
NPs–coated ZnO systems [51,52].
It was proposed that when oxygen molecules from the ambient were absorbed on the exposed
surface of Pd NPs, electrons were extracted from the ZnO conduction band, and depletion layers
were formed at the surface of ZnO, causing a decrease in carrier concentration. In the case of the
LAO/STO system, we believe that oxygen molecules dissociate and generate chemisorbed oxygen
species due to the enhanced interaction by Pd NPs on the surface, and the oxygen molecules tend to
capture free electrons from the Pd NPs and cause interfacial 2DEG electrons to diffuse out to the
LAO film or the Pd NPs. Therefore, with the absorption of oxygen on the Pd NPs surface, O− tends
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 443

UV
Pd NPs

O–

O
LA

G
2DE

STO
(a)

1 2 UV
O2 O2 O2
O2

Pd NPs
– – – –
O– O– O– O– O– O– O– O O O O– O– O– O O– O–
LAO
2DEG e– e– e– e– e–
h+ h+
STO
(b)

FIGURE 12.23 (a) Schematic diagram of the Pd NPs–surface modulated LAO/STO with UV light irra-
diation. (b) Schematic depiction of the processes: Process 1—oxygen molecules absorbed on Pd NPs attract
electrons from 2DEG, making it negatively charged; Process 2—UV light irradiates on the surface and excites
electron–hole pairs inside STO, the photoexcited electrons become the excess charge carriers to the 2DEG
interface, and photoexcited holes diffuse through LAO and combine with the negatively charged oxygen on
the surface of LAO.

to “spillover” the surface of the LAO [53], resulting in more negative charge carriers being trapped
on the LAO surface, and, therefore, the carrier density of the interfacial 2DEG decreases and the
resistance of the interface increases.
On the other hand, when UV light irradiates on the surface, more electrons are induced in the
2DEG interface due to the generation of photon-excited electrons and holes in the STO substrate
(suggested by the fact that the photocurrent starts to increase when the photon energy exceeds the
band-gap energy of STO). The persistent increase of the photoresistivity when light is off is due to
the slow recombination of electrons and holes [39]; the slow recombination process is believed to
be due to the internal electric field buildup from the polar discontinuity in the LAO layer. This field
tends to separate electrons and holes (electrons move toward the interface and holes move toward
the surface), and, therefore, the recombination is generally prohibited; but the interdiffusion of elec-
trons and holes still results in slow recombination.
Electrical measurement were conducted on samples with Pd NPs coated on the 10 and 40 uc
LAO/STO, and the electrical performance of the sample varies with testing condition as shown in
Figure 12.25. Compared to the 5 uc Pd NPs–coated LAO/STO sample. It is apparent that the 10 and
40 uc thick samples show a slower and smaller photoresponse, as per the data given in Table 12.3. It is
believed that the thicker LAO film hinders the diffusion of electrons to the Pd NPs due to the increased
diffusion length and so the ratio of resistance recovery is not as high as in the case of the 5 uc.
In summary, the Pd NPs modification of the LAO/STO heterostructure surface exhibits a large
oxygen-sensitive photoconductivity at room temperature. This interesting phenomenon is attributed
444 Semiconductor Nanocrystals and Metal Nanoparticles

0.12
0.12
0.10 0.10
0.08 0.08
Rxx (MΩ/sq.)

Rxx (MΩ/sq.)
0.06 0.06

0.04 0.04

0.02 0.02

0.00 0.00
0 50 100 150 200 250 300 0 50 100 150 200 250 300
(a) (b)

0.25 12

0.20 10
Rxx (MΩ/sq.)

Rxx (MΩ/sq.)
0.15 8

6
0.10
4
0.05
2

0 50 100 150 200 250 300 0 50 100 150 200 250 300
(c) Temperature (K) (d) Temperature (K)

FIGURE 12.24 Resistance-temperature curve of the bare LAO/STO sample (a) under UV and (b) in the
dark, for Pd NPs–coated LAO/STO samples (c) under UV and (d) in the dark.

to Pd NPs’ catalytic effect and surface/interface charge coupling. These results are interesting in
physics and can probably be used for sensor applications such as UV light sensing and gas sensing.
The persistent photocurrent can be attributed to the existence of internal field in the LAO layer.
These also provide evidence for the polar catastrophe model for the explanation of conducting inter-
face, as discussed widely in the literature.

12.3.3 Highly Sensitive Gas Sensor by the LaAlO3/SrTiO3

Heterostructure with Pd NPs Surface Modulation
12.3.3.1 Gas Sensing Characteristics of Pd NPs–Coated LAO/STO
Pd NPs with size around 2 nm were deposited by DC magnetron sputtering in pure argon ambient
on the surface of LAO/STO heterostructure at room temperature [54]. Figure 12.26a schematically
shows the device structure, and the actual picture of the sample is shown in Figure 12.26b, where
the aluminum wires are ultrasonically bonded at the four corners for electrical characterization. It
is apparent that the Pd NPs, which appear as dark spots, are distributed uniformly and densely on
the surface of the LAO layer. The Pd NPs are crystallized and the density does not lead to a com-
plete conduction percolation path. The high-resolution image shown in Figure 12.26c and the ring-
shaped selected area electron pattern (SAED) shown in Figure 12.26d indicate the nanocrystalline
fcc structure of the Pd NP. The small size and high density distribution of the nanoparticles should
enhance the responsivity to the targeted gases [55,56].
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 445

0.26

0.24

0.22

Resistance (MΩ) 0.20

0.18
UV off O2
0.16 Air
Ar
0.14
200 400 600 800 1000
(a) Time (min)

0.40

0.36
Resistance (MΩ)

0.32

UV off O2
0.28
Air
Ar
0.24

200 400 600 800 1000

(b) Time (min)

FIGURE 12.25 Thickness-dependent response to gases (a) 10 uc and (b) 40 uc.

TABLE 12.3
Change of Resistance in Different Gas Environments
5 uc (%) 10 uc (%) 40 uc (%)
Bare LAO/STO 42 13 27
In argon 130 18 45
In dry air 633 29 48
In oxygen 2071 41 57

12.3.3.1.1 Effect of Gas Concentration

Figures 12.27 through 12.29 show the hydrogen gas sensing characteristics for the Pd NPs–coated
LAO/STO heterostructure with thickness of 5 uc at room temperature and elevated (80°C) tempera-
ture, where it is of particular interest that the conductance of the sample is extremely sensitive to
H2 gas. It is noticed that our pristine LAO/STO heterostructure does not show any response to H2
gas as reflected from its current–voltage characteristics, and there is no experimental report either
to show the gas sensing ability of such heterostructure. The current–voltage characteristics for the
446 Semiconductor Nanocrystals and Metal Nanoparticles

H2O C2H6O
(CH3)2OH
H2

5 mm

5 mm

(b)

(111)

A Pd NP
LAO
2DEG
STO
(a) (c) (d)

FIGURE 12.26 (a) Schematic diagram of the Pd NPs–coated LAO/STO heterostructure (the surface is a
low-magnification TEM image representing the Pd NPs), (b) actual picture of the device, (c) high-resolution
TEM image of a single Pd NP, and (d) SAED pattern of the Pd NP.

bare LAO/STO and Pd NPs–coated LAO/STO sample at 20 ppm H2 and pure oxygen are shown in
Figure 12.30. Figure 12.30a shows the I–V response of the interface, where typical ohmic behavior
with same resistance can be seen in the O2 and H2 ambient, suggesting that the change of gas ambi-
ent does not affect the electrical properties of the bare LAO/STO interface. However, as shown in
Figure 12.30b, the electrical response of the Pd NPs–coated LAO/STO interface changes after the
exposure to H2 and O2, and the I–V response shows a decrease in resistance for the 5 uc sample after
exposure to 20 ppm H2 gas. This suggests that the response to H2 gas is mainly attributed to the
presence of Pd NPs on the surface of LAO/STO.
It is believed that the reason H2 gas sensing in LAO/STO 2DEG has not been observed before is
due to the low reactivity of the LAO surface to any gas. The Pd NPs on the LAO surface, therefore,
play a crucial role in the functionalization of H2 gas sensing. It is also worth mentioning that the
response to a gas is an interfacial effect, since the experiment with surface electrode contact without
connecting to the interface does not show any measurable conductance and response upon exposure
to hydrogen.
For a typical gas sensor, three physical parameters are usually studied for the sensing perfor-
mance, including sensitivity S(%), response time τR, and recovery time τs. Among them, S (%) is
defined as the relative variation of the current at the interface and can be calculated as S = ΔI/I0 (%),
where ΔI = Ig−I0 (I0 is the current of the interface before exposing to hydrogen and Ig is the peak
current in presence of hydrogen). The parameters of τR and τs are defined respectively as the time
reach 90% of the saturated current of full response and recovery.
The transient response of the gas sensor to hydrogen at room temperature is shown in
Figure 12.27a, illustrating the variations in current in the interface corresponding to On (for 5 min)
and Off of H2 gas flow for two cycles with 20 ppm H2 gas concentration. The results show that the
interfacial current quickly increases when the device is exposed to H2 gas, with τR ~ 7.3 min, fol-
lowed by a slow approach to saturation. After the current in the interface saturated, the hydrogen
gas inside the chamber was pumped out and synthetic air was flushed to the sample for the recov-
ery process; the current of the interface decreased and essentially resumed the original state with
τs ~ 36.7 min. The responses of the Pd NPa–coated LAO/STO heterostructure measured at room
temperature was also recorded in the environment with hydrogen gas at different concentrations
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 447

τR ~ 7.3 min
H2
2 × 10–5

Current (A)
τs ~ 36.73 min

10–5

0 50 100 150 200 250

(a) Time (min)

2 ppm 4 ppm 6 ppm 8 ppm 10 ppm 12 ppm 14 ppm

800

700

600

500
S (%)

400

300

200

100

0
0 50 100 150
(b) Time (min)

FIGURE 12.27 (a) Electrical response for the Pd NPs–coated LAO/STO heterostructure exposed to 20 ppm
H2 gas loading and de-loading cycle at room temperature. (b) Real-time response of sensitivity under exposure
to H2 gas with various concentrations (2–14 ppm) at room temperature.

(from 2 to 14 ppm). The results shown in Figure 12.27b are plotted as the relationship between
sensitivity and time. A detectable change in current of ~80% was achieved for the sample with
hydrogen concentration as low as ~2 ppm. It is apparent that exposing the sensor to a higher con-
centration of hydrogen gas induces faster increase of the current to a higher peak value. These
sawtooth responses are superimposed on a monotonically increasing background; this fact can
be attributed to the charge accumulation and slow recovery rate of the sensor operated at room
temperature. Nevertheless, it shows that the Pd NPs–modulated LAO/STO heterostructure is very
sensitive to H2 gas.
Temperature is an important factor that greatly influences the hydrogen sensing response based
on the catalytic effect. Usually, higher temperature would lead to higher sensing performance due to
the lowering of activation energy for gas adsorption and desorption of devices. Figure 12.28a shows
the cyclic performance of the device with 20 ppm H2 operated at 80°C, where the sensor shows both
shorter, τR ~ 0.8 min, and longer, τs ~ 0.95 min, response times, which are much faster compared to
448 Semiconductor Nanocrystals and Metal Nanoparticles

τR ~ 0.8 min H2

10–5

Current (A)
τs ~ 0.95 min

10–6

0 5 10 15 20 25 30 35
(a) Time (min)

2 ppm 4 ppm 6 ppm 8 ppm 10 ppm 12 ppm 14 ppm

4000

3000
S (%)

2000

1000

0
0 50 100 150
(b) Time (min)

FIGURE 12.28 (a) Electrical response to H2 gas for the Pd NPs–coated LAO/STO heterostructure.
(b) Sensitivity response at 80°C.

the sensor operated at room temperature. The shorter response time can be explained by faster dif-
fusion and dissociation of hydrogen molecules at higher temperature, where the surface has higher
reactivity and more active sites for H2 molecules adsorption to Pd NP. Figure 12.28b shows the
sensor’s response to H2 gas with different concentrations, where it can be seen that at higher tem-
perature, the sensitivity is much higher. A response with sensitivity up to 2400% has been recorded
for a device exposed to 2 ppm H2 gas.
As shown in Figure 12.29a, the response time τR of the Pd NPs–coated LAO/STO heterostructure
at 80°C generally tends to be shortened (for [H2] = 2 ppm, τ = 3.2 min and for [H2] = 20 ppm, τ =
0.8 min). Upon exposure to dry air for the recovery process, dissolved hydrogen on Pd NPs reacts
with the oxygen in air and forms H2O by the reaction of 2H2 + O2 = 2H2O. At relatively low working
temperatures, the response time is longer due to slow desorption of the water molecules formed on
the surface. The sensor was able to detect H2 gas with low concentration at high temperature, while
it was relatively less sensitive to low H2 concentrations at room temperature. Figure 12.29b plots the
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 449

3.5
Fitting
3.0

Response time (min)

2.5

2.0

1.5

1.0

0.5
4 8 12 16 20
(a) H2 concentration (ppm)

1000 4900
900
4200
800
700 3500
600
2800
S (%)

500
400 2100

300 1400
200 At 80°C
700
100 At room temperature
0 0
2 4 6 8 10 12 14
(b) H2 concentration (ppm)

FIGURE 12.29 (a) The relationship between response time and hydrogen concentration for a sensor oper-
ated at 80°C. (b) Combined sensor sensitivity at room temperature and 80°C.

sensitivities for the sensor to various concentrations of H2 gas at room temperature and at 80°C. It
is apparent that at low H2 concentration, the sensitivity increases with the increase of H2 concentra-
tion; while at high concentration, it begins to saturate probably due to the lack of adsorption sites
on the Pd NPs surface.

12.3.3.1.2 Effect of Oxygen Concentration in Hydrogen Gas

Instead of testing the sensor using H2 gas balanced in argon, the sensor was found to be able to detect
H2 gas balanced with air as well, where coexistence of oxygen gas (oxidizing gas) during hydrogen
gas (reducing gas) detection should play a very important role in the sensing characteristics of the
Pd NPs–coated LAO/STO heterostructures. Figure 12.31 shows the time-dependent interfacial cur-
rent of the Pd NPs–coated LAO/STO sensor exposed to various concentrations of H2 in synthetic
air (10, 40, 380, and 3500 ppm) at 80°C, with oxygen gas recovery after each H2/air measurement.
With the introduction of the gas mixture to the sensor, the current of the interface increases, while
it also increases with the increase in hydrogen concentration, showing a faster increase in current
to a higher peak value.
450 Semiconductor Nanocrystals and Metal Nanoparticles

100

50

Current (µA)
0

–50
In H2
In O2
–100

–10 –5 0 5 10
(a) Voltage (V)

200
150
100
Current (µA)

50
0
–50
–100
In H2
–150 In O2
–200

–10 –5 0 5 10
(b) Voltage (V)

FIGURE 12.30 Current–voltage characteristics in H2 and O2 ambient of the (a) bare LAO/STO and (b) Pd
NPs–coated LAO/STO samples.

12.3.3.1.3 Effect of LAO Thickness

It is of particular interest to show the sensors’ response to H2 gas for the samples with different
LAO thicknesses, so a series of samples with different LAO thicknesses (10, 20, and 40 uc) and the
same configuration were tested. Figure 12.32 shows the response curves of the sensors to different
hydrogen concentrations at room temperature and at 80°C. All the samples respond to hydrogen gas
and the responses tend to saturate at a high hydrogen concentration. Compared to the sample of 5 uc
LAO thickness, smaller and slower gas responses for higher LAO thicknesses have been observed.
It is believed that a thicker LAO film reduces the field effect and hinders the diffusion of charge
carrier to the Pd NPs with the increase in diffusion length, resulting in the heterostructure’s lower
sensitivity to H2 gas.

12.3.3.2 Response to Other Gases

The working mechanism of the gas sensor lies in the conversion of electrical conductivity due to
surface reactions such as oxidation or reduction that is caused by different gas exposure. Oxidizing
(reducing) gases can serve as electron-withdrawing (donating) groups and change the channel car-
rier concentration through charge coupling with the interface; while the adsorbents from the ambi-
ent gas molecules interact strongly with the Pd NPs–coated LAO/STO heterostructure. In addition
 Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 451

10 ppm 40 ppm 380 ppm 3500 ppm

20 O2 H2 + Air O2 H2 + Air O2 H2 + Air O2 H2 + Air O2

0.005% H2 0.02% H 2 0.18% H 1.78% H 2 2
99.995% Air 99.98% Air 99.82% Air 98.22% Air

16

Current (µA) 12

0
0 20 40 60 80 100
Time (min)

FIGURE 12.31 Time-dependent interfacial current changes of the Pd NPs–coated LAO/STO sensor exposed
to various concentrations of H2 gas in synthetic air (10, 40, 380, and 3500 ppm) at 80°C and then back to O2 gas
for recovery process. The concentration of the H2/air gas was monitored by using the precalibrated mass flow
controllers. Hydrogen gas and synthetic air were introduced to the gas mixer by a two-way valve using two sepa-
rate mass flow controllers, with the test gases allowed to flow through the test chamber with the sensor installed.

2 ppm 4 ppm 6 ppm 8 ppm 10 ppm 12 ppm 14 ppm

600
700 5 uc 10 uc
10 uc 20 uc
20 uc 500
600 40 uc
40 uc
Sensitivity (%)

500 400
Sensitivity (%)

400 300
300
200
200
100 100

0 0
0 20 40 60 80 100 120 140 160 180 2 4 6 8 10 12 14
(a) Time (min) (b) H2 concentration (ppm)

2 ppm 4 ppm 6 ppm 8 ppm 10 ppm 12 ppm 14 ppm

5000 1000
5 uc
10 uc 900 10 uc
20 uc 20 uc
4000 40 uc
800 40 uc
700
Sensitivity (%)

Sensitivity (%)

3000 600
500
2000 400
300
1000 200
100
0 0
0 20 40 60 80 100 120 140 160 180 2 4 6 8 10 12 14
(c) Time (min) (d) H2 concentration (ppm)

FIGURE 12.32 Real-time sensitivity response of Pd NPs–coated LAO/STO heterostructure with different
LAO layer thicknesses at (a, b) room temperature and (c, d) 80°C.
452 Semiconductor Nanocrystals and Metal Nanoparticles

to H2 gas, other reducing gases such as ethanol, acetone, and water vapor were tested. The sensor
was put in the argon ambient to obtain the base line and a fixed amount of the chemical vapor was
flowed to the sensor for sensitivity test. As shown in Figure 12.33a and b, sensitivities to acetone and
ethanol gases reached ~64% and 48%, respectively, and the response times were τR,acetone ~ 25 min
and τR,ethanol ~ 8 min, respectively, for acetone and ethanol gases. We believe that this gas-sensing
characteristic is due to the exchange of electrons between ionosorbed species and nanoparticles,
as well as the LAO film.
The reaction of ethanol and acetone with ionic oxygen species can be described by CH3CH2OH +
6O− → 2CO2 + 3H2O + 6e− and CH3COCH3 + O− → CH3CO + CH2 + OH− + e−. The reductive gas
reduces the molecular oxygen on top of the Pd NPs surface, and, thus, the electron concentra-
tion at the interface increases and resumes to a stable state. The behavior of the device to water
vapor is similar to acetone and ethanol, however, much higher sensitivity and shorter response
time are observed. As shown in Figure 12.33c, the conductance of the Pd NPs–coated LAO/STO
heterostructure increases with sensitivity up to ~8000% after exposure to H2O vapor at room
temperature.
The H2O molecules should remove adsorbed oxygen on the LAO surface. As mentioned by Xie
et al. [35], H2O are known to alter the charge carrier density of the LAO/STO interface. Sensing of
the H2O molecules with reasonable value of sensitivity and slow response and recovery have been
reported, and H2O molecules were proposed as the major polar gases in the air that can influence

70 Acetone
60
Ethanol

60 τR 50
τR
50
Acetone out 40
Sensitivity (%)

Sensitivity (%)

40
Ethanol out
30
30
20
20
Acetone in 10
10 Air Ethanol in Air
0 0
10 20 30 40 50 5 10 15 20 25
(a) Time (min) (b) Time (min)

9000
Water vapor
8000 τR
7000
Water out
6000
Sensitivity (%)

5000
4000
3000
2000
Water in
1000
Air
0
5 10 15 20
(c) Time (min)

FIGURE 12.33 Real-time sensitivity response of the Pd NPs–coated LAO/STO heterostructure to (a) acetone,
(b) ethanol, and (c) water vapor in argon ambient at room temperature.
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 453

interface conductance. One probable reaction is that water vapor can be dissociated on the Pd sur-
face giving rise to H + and OH− ions, that is, H2O↔H + + (OH)−. A first principle calculation shows
that H2O binds strongly to the AlO2 outer surface [57,58] and modulates the conductivity at the
interface. However, as dry air is fed in for recovery process, the recovery rate is much slower than
the case in acetone and ethanol. Water molecules adsorbed on the Pd NPs surface with slower
evaporation rate lead to lower surface adsorption for chemisorptions of oxygen species, resulting in
a slower recovery rate.

12.3.3.3 Mechanism Discussion

A model (as shown in Figure 12.34) has been proposed to explain the enhanced sensitivity to gases
for the Pd NPs–coated LAO/STO heterostructure. Palladium has been used in sensing applica-
tions due to its high diffusion coefficient, solubility, and selectivity with respect to hydrogen [59].
As shown in the diagram, direct adsorption of the gaseous molecules or the following by-products
formed by the reaction on the Pd NPs surface occur on top of the LAO layer. The presence of Pd
NPs results in charge coupling and exchange with the oxygen-deficient LAO film, which facilitates
the interfacial redox reaction of oxygen [44,46,60], where the sensing mechanism can be explained
in terms of oxidizing/reducing gas effect. Therefore, the surface is active and promotes further
adsorption of oxygen from the atmosphere due to the presence of Pd.
The surface of the Pd contains a number of oxygen-related physisorption species such as O2− and
−
O , which “spillover” on the surface. On the other hand, when the heterostructure is exposed to
hydrogen gas, hydrogen acts as the reducing gas, which decreases the concentration of oxygen spe-
cies on the surface and eventually increases the concentration of the charge carrier at the interface.
Electrically, the work function of the Pd metal relative to the electron affinity of STO determines
the transfer direction of electrons between the 2DEG to the Pd NPs. The presence of Pd metal on
the surface with work function 5.6 eV is larger than the electron affinity of STO (4.0 eV), leading to
a strong suppression of 2DEG carrier densities at the LAO/STO interface due to a reduction in the
builtin electric field in the LAO layer; a similar phenomenon has been observed based on the first
principle calculation in the LAO/STO system [47,48].

Oxygen adsorption Oxygen desorption

O2 (g) O2 (ad)

1 O2 (ad) + e– O2− (ad) 2 (i) (ii) 2H2 + O2– 2H2 O + e–

O2 O2 O2 H2 H2 H2 H2
O2 H2O H2O

O– O– O– O– O– O–
PdHx

LAO
2DEG e– e– e–

STO

FIGURE 12.34 Proposed mechanism for H2 gas sensing process: Process 1: the oxygen molecules are
adsorbed on the Pd NPs and attract electrons from the 2DEG, making the Pd NPs negatively charged, which
decrease the charge carrier concentration of the interface; Process 2: (i) Pd NPs changes to palladium hydride
(PdH x) after exposure to hydrogen gas and apparently decrease the work function of Pd, which lowers the bar-
rier height of the interface. (ii) The atomic hydrogen reacts with the oxygen molecules adsorbed on the Pd NPs
surface, the hydrogen molecules react with the adsorbed oxygen species, 2H2 + O2− (ad) → 2H2O + e−, and
the as-released electrons enhance the conductivity of the interfaces.
454 Semiconductor Nanocrystals and Metal Nanoparticles

As reported, the hydrogen-sensing response of semiconductor-based gas sensors with noble

metal electrodes is related to Schottky contact between the noble metal electrode and the hetero-
structure [53,61,62]. The hydrogen molecules may be adsorbed and dissociated into hydrogen atoms
on Pd surface and dissolve into the Pd bulk; and consequently the Pd NPs change to palladium
hydride (PdH x) that apparently decrease the work function of Pd and lower the barrier height, result-
ing in less suppression of the carrier density at the interface, and, therefore, fewer electrons from the
interface may transfer to (PdH x) [63,64].
In addition to the Pd NPs electron affinity–induced modulation of the interfacial conductivity,
the Pd NPs catalytic electrochemical reaction with oxygen molecules and the consequence of charge
exchange with the LAO film is believed to play a more important role in reducing the charge carrier
density of the LAO/STO 2DEG. The Pd NPs surface usually absorbs dissociated oxygen molecules
when exposed to air [50]. It is known that the transport properties of the Pd NPs–coated LAO/STO
heterostructure are strongly affected by the gas environment, especially oxygen, and the conductance
of the interface decreases when the device is exposed to oxygen ambient [37]. The Pd NPs act as cata-
lysts that activate the dissociation of molecular oxygen [50], and, therefore, the presence of ambient
oxygen has a considerable influence on the performance of Pd NPs–coated LAO/STO 2DEG.
When the sensor is exposed to H2 gas, the hydrogen molecules react with the adsorbed oxygen
species, 2H2 + O2− (ad) → 2H2O + e−, and the as-released electrons will enhance the conductivity
of the interface. Afterward, when air is introduced into the device, oxygen molecules in air react
with the surface adsorbed-hydrogen atoms to form H2O and evaporate due to exothermic reaction.
This reaction process is exothermic and the produced H2O molecules are desorbed quickly from
the surface. A depletion region at the interface will be rebuilt by the adsorbed oxygen species on
the Pd surface and results in O2− or O− formation.
To study the importance of the presence of oxygen species for the sensing response, three
­carrier gases, synthetic air (~20% oxygen balanced in 80% argon gas), pure oxygen, and argon,
were selected to study the behavior of recovery process after the sensor was exposed to hydrogen.
As shown in Figure 12.35, the resistance of the LAO/STO 2DEG has been measured for the sensor
exposed to different carrier gases.
First, measurement began with the sensor that was exposed to pure oxygen ambient environment
to obtain a baseline (0.83 MΩ). Second, the resistance of the sample decreased to ~0.17 MΩ when
20 ppm hydrogen gas was introduced for 5 min. Third, synthetic air was introduced to the sample

O2 H2 Air H2 O2 H2 Ar H2 O2
0.8
Resistance (MΩ)

0.6 ~21%

0.4

0.2

0 20 40 60 80
Time (min)

FIGURE 12.35 Variation of the resistance of the Pd NPs–coated LAO/STO sensor exposed to 20 ppm
hydrogen with recovery process conducted under synthetic air, pure oxygen, and argon gas.
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 455

for 10 min, which enabled the recovery process, and the resistance of the sample reached ~0.53 MΩ.
Fourth, hydrogen gas was introduced again to the sensor for 5 min and the resistance of the sample
reached ~0.17 MΩ. Then, pure oxygen was introduced to the sample for 10 min, and the resistance
of the sample reached ~0.64 MΩ after the recovery process. Compared with the case where the
sample recovers in synthetic air, the resistance of the sample differs for ~21%, which implies that
the recovery process is extremely sensitive to the presence of oxygen. There was little or no recov-
ery when this experiment was conducted in argon gas. The result demonstrates the key role of the
oxygen adsorbed on the surface of the LAO/STO heterostructure.
It should be noted that the type of chemisorbed oxygen species depends strongly on temperature,
where a relatively higher temperature favors the redox reaction, and the interface 2DEG will become
more sensitive for H2 sensing. It may be argued that the hydrogen sensitivity for the Pd NPs–coated
LAO/STO may be due to the volume expansion of the Pd NPs lattice. For bulk palladium, a lattice
expansion as high as 3%–4% has been reported [65], and the hydrogen sensing mechanism has been
explained by the tunneling current between the Pd NPs. To rule out this effect, Pd NPs were also
deposited on the TiO2-site-terminated STO substrate (without LAO), and the results show that the
Pd NPs are well below the conduction percolation threshold in hydrogen ambient.
Indeed, the proposed mechanism should be supported by experimental data [66]. The use of
Kelvin probe force microscopy (KPFM) can determine the work function of the sample surface, and
the ambient controlled KPFM is needed to measure the work function of the sample under different
gas ambient. With KPFM, the work function of the surface can be determined at atomic scales. In
order to have a deeper understanding to the observed phenomenon of the Pd NPs–coated LAO/STO
sample, the transport and the KPFM measurement were performed in a glove box, the measurement
was made under nitrogen, nitrogen (98%)/hydrogen (2%) and oxygen ambient. In order to suppress
the influence of the persistent photocurrent, the samples, including the bare LAO/STO and the Pd
NPs–coated LAO/STO, were stored in the dark for at least 1 day.
The conductance (G) and the work function (W) of the Pd NPs–coated LAO/STO heterostructure
show significant response upon exposure to hydrogen gas. The conductance of the Pd NPs–coated
LAO/STO heterostructure (GPd/LAO/STO) is smaller than that of the bare LAO/STO heterostructure
(GLAO/STO), suggesting that the presence of Pd NPs influences the sheet carrier density of the LAO/
STO interface significantly. According to Xie et al. [35], the sheet carrier density σ is directly related
to the uncompensated potential, Vuncom. A potential drop across the polar LAO film is given by

e LAO
Ds = - DVuncom (12.2)
edLAO

where
Vuncom is the uncompensated potential
e is the unit charge
εLAO is the dielectric constant of LaAlO3

Due to the large work function of the Pd element, a Schottky contact is formed at the Pd–LAO
interface and hence the potential profile of the LAO thin film changes. Such band bending alters the
uncompensated potential, Vuncom, and the work function. Assuming that the electron mobility of the
LAO/STO heterostructure and the Pd NPs–coated LAO/STO heterostructure is the same, the rela-
tion GLAO/STO − GPd/LAO/STO = (σLAO/STO − σPd/LAO/STO) × e × μ can be obtained. A large difference of
GPd/LAO/STO and the work function of Pd NPs–coated LAO/STO has been observed in the nitrogen
and hydrogen/nitrogen ambient. Two possible processes, that is, the PdH x formation and the dis-
sociate or adsorption of hydrogen gas on the LAO/STO surface, can be considered. PdHx has a smaller
work function compared with Pd, which subsequently lowers the Vuncom and raises the charge carrier
concentration σ, according to Equation 12.2. Thus, hydrogen adsorption on the Pd NPs–coated
456 Semiconductor Nanocrystals and Metal Nanoparticles

LAO STO surface can readily explain the notable decrease of the work function of the Pd NPs–coated
LAO/STO heterostructure.
The gas-sensing ability has been attributed to charge coupling between the desorbed gas mol-
ecules/Pd NPs and the LAO/STO interface through the LAO layer. These results not only promise
the potential interest for understanding the oxide interfacial 2DEG, but also its application in all-
oxide devices, and, thus, opening a new route to complex oxide physics and ultimately to the design
of devices in oxide electronics.

12.4 CONCLUSION AND PERSPECTIVES

In this chapter, the interfacial properties of the oxide heterostructure and its application in sens-
ing device were reviewed. In particular, through surface modification, the LAO/STO 2DEG was
demonstrated to have great potential in polar liquid sensor, UV light sensor, and gas sensor. The
following conclusions can be made.
The potential application of LAO/STO interfacial 2DEG as polar molecule sensor was demon-
strated with an explanation of the sensing mechanism, where a Schottky junction model was pro-
posed. These results present another evidence to the existence of a polarization field inside the LAO
layer due to polar catastrophe and charge exchange between the LAO layer and the STO surface.
With LaAlO3 surface modification by Pd NPs, the LaAlO3/SrTiO3 interfacial 2DEG presents
a giant optical switching effect with a photoconductivity on/off ratio as high as 750% under UV
light irradiation. Strong photoconductivity has been observed at wavelengths around 380 nm, while
the increase of photoconductivity in the wavelength range of 420–700 nm is not obvious, sug-
gesting that the increase in photocurrent is due to the absorption of photon energy by the STO
substrate from the generation of electron–hole pairs. For the Pd nanoparticle decorated LAO/STO
heterostructure, the recovery process is found to be dependent on oxygen concentration, which can
be fitted by the stretched exponential relaxation law. The palladium electron affinity induced the
degradation of the interface and the palladium nanoparticle catalytic electrochemical reaction with
oxygen molecules is believed to play an important role in the UV sensing mechanism. This result
shows that the LAO/STO heterostructure can probably be used for sensor applications in the field of
UV light sensing and gas sensing.
A highly sensitive hydrogen (H2) gas sensing characteristic has been demonstrated on the same
system where the Pd NPs are accommodated to the surface of LAO/STO 2DEG heterostructure
to trigger its sensitivity and selectivity to gases. The surface-modified LAO/STO heterostructure
is demonstrated to be highly sensitive to different gases. Hydrogen, oxygen, ethanol, acetone, and
water vapor can be detected via changes in the transport properties of the interface. The sensor is
found to be able to detect hydrogen gas with concentration as low as ~2 ppm at room temperature.
Exposing the sensor to a higher concentration of hydrogen induces a faster increase in sensitivity.
Increasing temperature would lead to higher sensing performance due to the lowering of activa-
tion energy for gas adsorption and desorption on the devices, and shorter response/recovery time
is observed for sensors operated at 80°C. The gas-sensing ability has been attributed to charge
coupling between the desorbed gas molecules/Pd NPs and the LAO/STO interface through the
LAO layer.
These results provide further evidence for understanding the mechanism of the oxide interfa-
cial 2DEG. It suggests that there is an energy well at the interface and an internal electric field
in the LAO film. Movable electrons induced by oxygen vacancy can flow in and out of the inter-
face driven by the internal field. The persistent photocurrent may be due to the internal field that
hinders the recombination of electron and holes generated by UV light irradiation. The results
reported in this chapter demonstrate the LAO/STO 2DEG application in all-oxide devices, and
thus opens a new route to complex oxide physics and ultimately to the design of devices in oxide
electronics.
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 457

ACKNOWLEDGMENTS
Ngai Yui Chan thanks the support of the Hong Kong PhD Fellow Scheme from the Research Grants
Council of Hong Kong No. RUY3. Jiyan Dai acknowledges the support from the Hong Kong NSFC/
RGC under Grant No. N-PolyU517/14 and the Strategic Importance Projects of the Hong Kong
Polytechnic University under Grants No. 1-ZE25 and 1-ZVCG.

REFERENCES
1. H. Kroemer. Quasi-electric fields and band offsets: Teaching electrons new tricks. Nobel Lecture,
December 8, 2000, http://go.nature.com/5SFA6C. Accessed on April 12, 2016.
2. A. Ohtomo and H. Y. Hwang. A high-mobility electron gas at the LaAlO3/SrTiO3 heterointerface.
Nature, 427(6973), 423–426, 2004.
3. C. W. Schneider, S. Thiel, G. Hammerl, C. Richter, and J. Mannhart. Microlithography of electron gases
formed at interfaces in oxide heterostructures. Appl. Phys. Lett., 89(12), 122101, 2006.
4. A. Brinkman, M. Huijben, M. van Zalk, J. Huijben, U. Zeitler, J. C. Maan, W. G. van der Wiel, G.
Rijnders, D. H. A. Blank, and H. Hilgenkamp. Magnetic effects at the interface between non-magnetic
oxides. Nat. Mater., 6(7), 493–496, 2007.
5. N. Reyren, S. Thiel, A. D. Caviglia, L. F. Kourkoutis, G. Hammerl, C. Richter, C. W. Schneider et al.
Superconducting interfaces between insulating oxides. Science, 317(5842), 1196–1199, 2007.
6. C. Cen, S. Thiel, G. Hammerl, C. W. Schneider, K. E. Andersen, C. S. Hellberg, J. Levy, and J. Mannhart.
Nanoscale control of an interfacial metal-insulator transition at room temperature. Nat. Mater., 7(4),
298–302, 2008.
7. P. Irvin, Y. Ma, D. F. Bogorin, C. Cen, C. W. Bark, C. M. Folkman, C. B. Eom, and J. Levy. Rewritable
nanoscale oxide photodetector. Nat. Photon., 4(12), 849–852, 2010.
8. N. Nakagawa, H. Y. Hwang, and D. A. Muller. Why some interfaces cannot be sharp. Nat. Mater., 5(3),
204–209, 2006.
9. A. Kalabukhov, R. Gunnarsson, J. Börjesson, E. Olsson, T. Claeson, and D. Winkler. Effect of oxygen
vacancies in the SrTiO3 substrate on the electrical properties of the LaAlO3 ∕SrTiO3 interface. Phys. Rev.
B, 75, 121404R, 2007.
10. Z. Q. Liu, C. J. Li, W. M. Lü, X. H. Huang, Z. Huang, S. W. Zeng, X. P. Qiu et al. Origin of the two-
dimensional electron gas at LaAlO3/SrTiO3 interfaces: The role of oxygen vacancies and electronic
reconstruction. Phys. Rev. X, 3(2), 021010, 2013.
11. M. Basletic, J. L. Maurice, C. Carrétéro, G. Herranz, O. Copie, M. Bibes, E. Jacquet, K. Bouzehouane,
S. Fusil, and A. Barthélémy. Mapping the spatial distribution of charge carriers in LaAlO3/SrTiO3
heterostructures. Nat. Mater., 7(8), 621–625, 2008.
12. L. Qiao, T. C. Droubay, T. C. Kaspar, P. V. Sushko, and S. A. Chambers. Cation mixing, band offsets
and electric fields at LaAlO3/SrTiO3(001) heterojunctions with variable La:Al atom ratio. Surf. Sci.,
605(15–16), 1381–1387, 2011.
13. S. A. Chambers, M. H. Engelhard, V. Shutthanandan, Z. Zhu, T. C. Droubay, L. Qiao, P. V. Sushko
et al. Instability, intermixing and electronic structure at the epitaxial LaAlO3/SrTiO3(001) heterojunc-
tion. Surf. Sci. Rep., 65(10–12), 317–352, 2010.
14. J. E. Drut and T. A. Lähde. Is graphene in vacuum an insulator? Phys. Rev. Lett., 102, 023802, 2009.
15. K. Hirakawa, H. Sakaki, and J. Yoshino. Mobility modulation of the two-dimensional electron gas via
controlled deformation of the electron wave function in selectively doped AlGaAs–GaAs heterojunc-
tions. Phys. Rev. Lett., 54, 1279, 1985.
16. J. Falson, D. Maryenko, Y. Kozuka, A. Tsukazaki, and M. Kawasaki. Magnesium doping controlled
density and mobility of two-dimensional electron gas in MgxZn1−xO/ZnO heterostructures. Appl. Phys.
Exp., 4, 091101, 2011.
17. S. Thiel, G. Hammerl, A. Schmehl, C. W. Schneider, and J. Mannhart. Tunable quasi-two-dimensional
electron gases in oxide heterostructures. Science, 313(5795), 1942–1945, 2006.
18. J. Mannhart and D. G. Schlom. Semiconductor physics: The value of seeing nothing. Science, 430,
620–621, 2004.
19. Breakthrough of the Year, The runners-up. Science, 318(5858), 1844–1849, December 21, 2007.
20. M. Kawasaki, K. Takahashi, and T. Maeda. Atomic control of the SrTiO3 crystal surface. Science,
266(5190), 1540–1542, 1994.
458 Semiconductor Nanocrystals and Metal Nanoparticles

21. M. Kawasaki, A. Ohtomo, and T. Arakane. Atomic control of SrTiO3 surface for perfect epitaxy of
perovskite oxides. Appl. Surf. Sci., 107, 102–106, 1996.
22. G. Koster, B. L. Kropman, G. J. H. M. Rijnders, D. H. A. Blank, and H. Rogalla. Quasi-ideal strontium
titanate crystal surfaces through formation of strontium hydroxide. Appl. Phys. Lett., 73(20), 2920,
1998.
23. L. J. van der Pauw. A method of measuring specific resistivity and Hall effect of discs of aribtrary shape.
Philips Res. Reports, 13, 1–9, 1958.
24. Z. Popović, S. Satpathy, and R. Martin. Origin of the two-dimensional electron gas carrier density at the
LaAlO3 on SrTiO3 interface. Phys. Rev. Lett., 101, 256801, 2008.
25. J. Lee and A. A. Demkov. Charge origin and localization at the n-type SrTiO3/LaAlO3 interface. Phys.
Rev. B, 78, 193104, 2008.
26. G. Singh-Bhalla, C. Bell, J. Ravichandran, W. Siemons, Y. Hikita, S. Salahuddin, A. F. Hebard, H. Y. Hwang,
and R. Ramesh. Built-in and induced polarization across LaAlO3/SrTiO3 heterojunctions. Nat. Phys.,
7, 80–86, 2011.
27. C. W. Bark, D. A. Felker, Y. Wang, Y. Zhang, H. W. Jang, C. M. Folkman, J. W. Park et al. Tailoring a
two-dimensional electron gas at the LaAlO3/SrTiO3 (001) interface by epitaxial strain. Proc. Natl. Acad.
Sci., 108(12), 4720–4724, 2011.
28. S. Pauli, S. Leake, B. Delley, M. Björck, C. Schneider, C. Schlepütz, D. Martoccia, S. Paetel, J. Mannhart,
and P. Willmott. Evolution of the interfacial structure of LaAlO3 on SrTiO3. Phys. Rev. Lett., 106(3),
036101, 2011.
29. C. Cancellieri, D. Fontaine, S. Gariglio, N. Reyren, A. Caviglia, A. Fête, S. Leake et al. Electrostriction
at the LaAlO3/SrTiO3 interface. Phys. Rev. Lett., 107, 056102, 2011.
30. C. Cen, S. Thiel, J. Mannhart, and J. Levy. Oxide nanoelectronics on demand. Science, 323(5917),
1026–1030, 2009.
31. F. Bi, D. F. Bogorin, C. Cen, C. W. Bark, J. W. Park, C. B. Eom, and J. Levy. Water-cycle’ mechanism
for writing and erasing nanostructures at the LaAlO3/SrTiO3 interface. Appl. Phys. Lett., 97(17), 173110,
2010.
32. Y. Xie, C. Bell, T. Yajima, Y. Hikita, and H. Y. Hwang. Charge writing at the LaAlO3/SrTiO3 surface.
Nano Lett., 10(7), 2588–2591, 2010.
33. Y. Xie, C. Bell, Y. Hikita, and H. Y. Hwang. Tuning the electron gas at an oxide heterointerface via free
surface charges. Adv. Mater., 23(15), 1744–1747, 2011.
34. R. Pentcheva, M. Huijben, K. Otte, W. E. Pickett, J. E. Kleibeuker, J. Huijben, H. Boschker et al.
Parallel electron–hole bilayer conductivity from electronic interface reconstruction. Phys. Rev. Lett.,
104, 166804, 2010.
35. Y. Xie, Y. Hikita, C. Bell, and H. Y. Hwang. Control of electronic conduction at an oxide heterointerface
using surface polar adsorbates. Nat. Commun., 2, 494, 2011.
36. W. C. Lo, K. Au, N. Y. Chan, H. Huang, C.-H. Lam, and J. Y. Dai. First principles study of transport prop-
erties of LaAlO3/SrTiO3 heterostructure with water adsorbates. Solid State Commun., 169, 46–49, 2013.
37. N. Y. Chan, M. Zhao, N. Wang, K. Au, J. Wang, L. W. H. Chan, and J. Dai. Palladium nanoparticle
enhanced giant photoresponse at LaAlO3/SrTiO3 two-dimensional electron gas heterostructures. ACS
Nano, 7, 8673–8679, 2013.
38. K. Au, D. F. Li, N. Y. Chan, and J. Y. Dai. Polar liquid molecule induced transport property modulation
at LaAlO3/SrTiO3 heterointerface. Adv. Mater., 24(19), 2598–2602, 2012.
39. A. Rastogi and R. C. Budhani. Solar blind photoconductivity in three-terminal devices of LaAlO3/
SrTiO3 heterostructures. Opt. Lett., 37(3), 317–319, 2012.
40. A. Tebano, E. Fabbri, D. Pergolesi, G. Balestrino, and E. Traversa. Room-temperature giant persistent
photoconductivity in SrTiO3/LaAlO3 heterostructures. ACS Nano, 6, 1278–1283, 2012.
41. E. Di Gennaro, U. Scotti, C. Aruta, C. Cantoni, A. Gadaleta, A. R. Lupini, D. Maccariello, D. Marré
et al. Persistent Photoconductivity in 2D Electron Gases at Different Oxide Interfaces, Adv. Optical
Mater., 1, 834–843, 2013.
42. A NEMA Lighting Systems Division Document, Ultraviolet radiation from fluorescent lamps, Lamp
Section, National Electrical Manufacturers Association, Rosslyn, VA, May 4, 1999.
43. Health and Consumers Scientific Committees, Health effects of artificial light, http://ec.europa.eu/health/
scientific_committees/opinions_layman/artificial-light/en/about-artificial-light.htm. Accessed on April
12, 2016.
44. S. W. Lee, Y. Liu, J. Heo, and R. G. Gordon. Creation and control of two-dimensional electron gas using
Al-based amorphous oxides/SrTiO3 heterostructures grown by atomic layer deposition. Nano Lett.,
12(9), 4775–4783, 2012.
Exploring the LaAlO3/SrTiO3 Two-Dimensional Electron Gas 459

45. F. Trier, D. V. Christensen, Y. Z. Chen, A. Smith, M. I. Andersen, and N. Pryds. Degradation of the
interfacial conductivity in LaAlO3/SrTiO3 heterostructures during storage at controlled environments.
Solid State Ion., 230, 12–15, 2012.
46. Y. Chen, N. Pryds, J. E. Kleibeuker, G. Koster, J. Sun, E. Stamate, B. Shen, G. Rijnders, and S. Linderoth.
Metallic and insulating interfaces of amorphous SrTiO3-based oxide heterostructures. Nano Lett., 11(9),
3774–3778, 2011.
47. R. Arras, V. G. Ruiz, W. E. Pickett, and R. Pentcheva. Tuning the two-dimensional electron gas at the
LaAlO3/SrTiO3(001) interface by metallic contacts. Phys. Rev. B, 85, 125404, 2012.
48. A. Janotti, L. Bjaalie, L. Gordon, and C. G. Van de Walle. Controlling the density of the two-dimensional
electron gas at the SrTiO3/LaAlO3 interface. Phys. Rev. B, 86, 241108, 2012.
49. J. Lee, C. Lin, and A. Demkov. Metal-induced charge transfer, structural distortion, and orbital order in
SrTiO3 thin films. Phys. Rev. B, 87, 165103, 2013.
50. I. Meusel, J. Hoffmann, J. Hartmann, M. Heemeier, M. Bäumer, J. Libuda, and H. Freund. The interac-
tion of oxygen with alumina-supported palladium particles. Catal. Lett., 71(1), 5–13, 2001.
51. A. Bera and D. Basak. Pd-nanoparticle-decorated ZnO nanowires: Ultraviolet photosensitivity and
photoluminescence properties. Nanotechnology, 22, 265501, 2011.
52. Y. Chang, J. Xu, Y. Zhang, S. Ma, L. Xin, L. Zhu, and C. Xu. Optical properties and photocatalytic
performances of Pd modified ZnO samples. J. Phys. Chem. C, 113(43), 18761–18767, 2009.
53. A. Kolmakov, D. O. Klenov, Y. Lilach, S. Stemmer, and M. Moskovits. Enhanced gas sensing by
individual SnO2 nanowires and nanobelts functionalized with Pd catalyst particles. Nano Lett., 5(4),
667–673, 2005.
54. N. Y. Chan, M. Zhao, J. Huang, K. Au, M. H. Wong, H. M. Yao, W. Lu et al. Highly sensitive gas sensor
by the LaAlO3/SrTiO3 heterostructure with Pd nanoparticle surface modulation. Adv. Mater., 26(34),
5962–5968, 2014.
55. F. Yang, S. Kung, M. Cheng, J. C. Hemminger, and R. M. Penner. Smaller is faster and more sensitive:
The effect of wire size on the detection of nanowires. ACS Nano, 4(9), 5233–5244, 2010.
56. M. A. Lim, H. Kim, C. Park, Y. W. Lee, S. W. Han, and Z. Li. A new route toward ultrasensitive, flexible
chemical sensors: Metal nanotubes by wet-chemical synthesis along sacrificial nanowire templates. ACS
Nano, 6(1), 598–608, 2012.
57. F. Li, M. Liang, W. Du, M. Wang, Y. Feng, Z. Hu, L. Zhang, and E. G. Wang. Writing charge into the
n-type LaAlO3/SrTiO3 interface: A theoretical study of the H2O kinetics on the top AlO2 surface. Appl.
Phys. Lett., 101(25), 251605, 2012.
58. Y. Li and J. Yu. Modulation of electron carrier density at the n-type LaAlO3/SrTiO3 interface by water
adsorption. J. Phys. Condens. Matter, 25(26), 265004, 2013.
59. S. Semancik and T. B. Fryberger. Model studies of SnO2 based gas sensors: Vacancy defects and Pd
additive effects. Sens. Actuat. B Chem., 1, 97–102, 1990.
60. F. Trier, D. V. Christensen, Y. Z. Chen, A. Smith, M. I. Andersen, and N. Pryds. Degradation of the
interfacial conductivity in LaAlO3/SrTiO3 heterostructures during storage at controlled environments.
Solid State Ion., 230, 12–15, 2013.
61. H. Hasegawa and M. Akazawa. Hydrogen sensing characteristics and mechanism of Pd ∕AlGaN∕GaN
Schottky diodes subjected to oxygen gettering. J. Vac. Sci. Technol. B Microelectron. Nanom. Struct.,
25(4), 1495, 2007.
62. K. Skucha, Z. Fan, K. Jeon, A. Javey, and B. Boser. Palladium/silicon nanowire Schottky barrier-based
hydrogen sensors. Sens. Actuat. B Chem., 145(1), 232–238, 2010.
63. T. C. Lin and B. R. Huang. Palladium nanoparticles modified carbon nanotube/nickel composite rods
(Pd/CNT/Ni) for hydrogen sensing. Sens. Actuat. B Chem., 162(1), 108–113, 2012.
64. M. G. Chung, D. H. Kim, D. K. Seo, T. Kim, H. U. Im, H. M. Lee, J. B. Yoo, S. H. Hong, T. J. Kang,
and Y. H. Kim. Flexible hydrogen sensors using graphene with palladium nanoparticle decoration.
Sens. Actuat. B Chem., 169(13382), 387–392, 2012.
65. F. Favier, E. C. Walter, M. P. Zach, T. Benter, and R. M. Penner. Hydrogen sensors and switches from
electrodeposited palladium mesowire arrays. Science, 293(5538), 2227–2231, 2001.
66. H. Kim, N. Y. Chan, J. Dai, and D. W. Kim. Enhanced surface-and-interface coupling in Pd-nanoparticle-
coated LaAlO3/SrTiO3 heterostructures: Strong gas- and photo-induced conductance modulation. Sci.
Rep., 5, 8531, 2014.
 13 Novel Nanoelectronic Device
Applications of Nanocrystals
and Nanoparticles

Z. Liu

CONTENTS
13.1 Introduction........................................................................................................................... 461
13.2 Single-Electron Devices........................................................................................................ 462
13.2.1 Single-Electron Box................................................................................................... 462
13.2.2 Coulomb Blockade Effects........................................................................................ 463
13.2.3 Working Principles of SETs.......................................................................................465
13.2.4 Fabrication of Si-Based SETs.................................................................................... 467
13.2.5 Device Applications of SETs..................................................................................... 467
13.3 Memristors.............................................................................................................................469
13.3.1 Memristors Based on Fe3O4 Nanoparticle Assemblies............................................. 471
13.3.2 Memristors Based on Amorphous Si Embedded with Ag Nanoparticles................. 472
13.3.3 Electronic Synapses Realized with Si-Based Memristors......................................... 475
13.3.4 Electrochemical Dynamics of Metallic Nanoparticles in Dielectrics....................... 478
13.4 Self-Learning Devices...........................................................................................................484
13.4.1 Self-Learning Behaviors in Ni-Rich NiO Thin Films...............................................484
13.4.2 Transient Memory Loss Behavior............................................................................. 490
13.5 Printing and Flexible Thin-Film Transistors......................................................................... 493
13.5.1 Printable CdSe Nanocrystal for Inorganic TFTs....................................................... 495
13.5.2 Spin-Coated CdSe Nanocrystals for Flexible TFTs.................................................. 497
References....................................................................................................................................... 498

13.1 INTRODUCTION
Nanotechnology builds on advances in microelectronics for almost half a century. The miniatur-
ization of electrical components greatly increased the utility and portability of computers, imag-
ing equipment, microphones, and other electronics. Scaling down of electronic device sizes is the
fundamental strategy for improving the performance of integrated circuits. The downscaling of
sizes of metal-oxide-semiconductor field-effect transistors (MOSFETs), yielding higher speeds and
larger packing densities at a lower cost for each generation of manufacturing technology, has been
the basis of the development of semiconductor industry for the past several decades [1,2]. However,
in the early years of the twenty-first century, the scaling of MOSFETs entered the deep sub-50 nm
regime. In this deep nanoscaled regime, fundamental limits of MOSFETs and technological chal-
lenges with regard to the scaling of MOSFETs are encountered. In order to extend the prodigious
progress of IC performance, it is essential to explore new design, architectures, and physical mecha-
nisms in the search for the next device breakthrough. Next-generation nanoelectronic devices may
have an operation principle effective to utilize quantum-mechanical effects for smaller dimensions
and thus provide a new functionality beyond that is attainable with MOSFETs.

461
462 Semiconductor Nanocrystals and Metal Nanoparticles

Semiconductor and metal nanoparticles are of a unique electronic nature due to the quantum-
mechanical rules governing their tiny nanoscale sizes. Based on nanocrystals and nanoparticles,
which are synthesized as standalone films or dispersed in certain dielectric matrices, many new
electronic devices have been developed to further exploit the performance and efficiency of nano-
electronics. In this chapter, several kinds of novel nanoelectronic devices based on nanocrystals
and nanoparticles will be introduced, including single-electron devices, memristors, self-learning
devices, and flexible or printing thin-film transistors (TFTs).

13.2 SINGLE-ELECTRON DEVICES

Single-electron devices are promising as new nanoscale devices because they retain their scalabil-
ity even at the atomic scale; moreover, they can control the motion of even a single electron [3,4].
Therefore, if single-electron devices are used as ultra large-scale integration (ULSI) elements, then
such ULSI circuits will have the attributes of extremely high integration and extremely low power
consumption. In this respect, scalability means that the performance of electronic devices increases
with the reduction of device dimensions. Power consumption is roughly proportional to the number
of the electrons transferred from the voltage source to the ground in logic operations. Therefore, the
utilization of single-electron devices in ULSI circuits is expected to reduce the power consumption of
the ULSI circuits. The operation and operation principle of single-electron transistors (SETs) is briefly
explained in this section. Next, the analytical device model of a SET, which is a typical functional
single-electron device, for circuit simulation is derived and the methodology of designing logic circuits
with SETs is discussed. Finally, some applications of single-electron transistors are introduced.

13.2.1 Single-Electron Box

Let a small conductor (traditionally called an island) be electrically neutral; that is, the material
has exactly as many electrons as it has protons in its crystal lattice. In this state, the island does
not generate any appreciable electric field beyond its borders, and a weak external force may bring
in an additional electron from outside. In most single-electron devices, this injection is carried
out by electron tunneling through an energy barrier created by a thin insulating layer. Now the
net charge of the island becomes −e instead of being neutral, and an electric field E is produced,
which repulses the following electrons to be added into the island. Though the fundamental charge
e = −1.6 × 10 –19 C is very small, the field E, which is inversely proportional to the square of the
island size, may become rather strong for nanoscale-sized structures. For example, the field is as
large as about 140 kV/cm on the surface of a 10 nm sphere in vacuum [5].
Figure 13.1 shows the conceptually simplest device, the single-electron box [6]. The device con-
sists of just one small island separated from a larger electrode (as the electron source) by a tunnel
barrier. An external electric field may be applied to the island using another electrode (i.e., the
gate electrode) separated from the island by a thicker insulator that prohibits electron tunneling.

Quantum Electron
dot tunneling

Source
Gate electrode
electrode
Insulating layer

FIGURE 13.1 Schematic of a single-electron box. The single-electron box consists of a quantum dot, an
electrode close to the dot allowing electron tunneling, and a gate electrode coupled to the dot, which is able to
modulate the energy levels in the dot.
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 463

The applied electric field on the gate electrode changes the electrochemical potential of the island
and thus determines the conditions of electron tunneling. Elementary electrostatics shows that the
free (Gibbs) energy of the system may be described as [5]

Q 2 æ C0 ö
W= + ÷ QU + const. (13.1)
2CS çè CS ø

where
Q = −ne is the island charge (n is the number of uncompensated electrons)
C0 is the island gate capacitance
CΣ is the total capacitance of the island (including C0)

Usually, this expression can be rewritten as [5]

(ne - Qe )2
W= + const. (13.2)
2CS

where parameter Qe, defined as [5]

Qe º UC0, (13.3)

which is usually called the “external charge.” An elementary calculation using these equations
shows that Q is a step-like function of Qe, that is, of the gate voltage, with a fixed distance between
the neighboring steps [5]:

DQe = e, (13.4)

e
DU = = const. (13.5)
C0

13.2.2 Coulomb Blockade Effects

The free energy F(n) of a system having n electrons in the island is expressed as

F (n) = WC (n) - A(n), (13.6)

where
WC (n) is the charging energy
A(n) is the work done by the gate voltage in order to make the electron number increase from
0 to n [7]

When tunneling events do not occur, the tunneling junction behaves like a normal capacitor, and
the polarization charge on the capacitors does not have to be associated with a discrete number of
electrons, n. This polarization charge is essentially due to a rearrangement of the electron gas with
respect to the positive background of ions. Therefore, the polarization charge takes a continuous
range of values, although the number of electrons in the quantum dot takes a discrete number of
464 Semiconductor Nanocrystals and Metal Nanoparticles

electrons, n. The polarization charges on the tunneling junction and gate capacitor are obtained
from the following relationship [7]:

Qt - Qg = -ne, (13.7)

Qt Qg
+ = Vg , (13.8)
Ct C g

where Qt and Qg are the polarization charges on the tunneling junction and the gate capacitor,
respectively. By using Qt and Qg, the charging energy WC (n) of the quantum dot is expressed as [7]

e2 n2 1 CtCgVg2
WC (n) = + , (13.9)
2CS 2 CS

where CΣ = Ct + Cg. In addition, the work A(n) done by the gate voltage source in order to make the
electron number of the quantum dot change from 0 to n is expressed as [7]

Cg C C V2
ò
A(n) = I (t ) × Vg dt = QgVg = en
CS
Vg + t g g .
CS
(13.10)

In order to maintain the electron number in the quantum dot, the following condition is required:
F(n) < F(n ± 1).
Free energy change ∆F(n, n + 1) that accompanies a transition of the electron number from n to
n + 1 is also simply expressed with critical charges Qc as [7]

e
DF (n, n + 1) = F (n + 1) - F (n) = (Qt - Qc ). (13.11)
Ct

Experimentally, a typical single-electron tunneling and Coulomb blockade effect was reported
by Chan et al. [8] in a nanostructure incorporated with gold nanocrystals (Au NCs). Figure 13.2
shows the HfAlO/Au NCs/HfAlO trilayer structure on a p-Si substrate. The Au NCs are uni-
formly distributed between the HfAlO control layer and the tunnel layer. The thicknesses of the
tunnel layer and the control layer are about 15 and 20 nm, respectively. The interface between the
HfAlO tunnel layer and the silicon substrate is mainly due to the oxygen diffusion into the sam-
ple during postdeposition annealing. The inset shows the plane-view TEM image of the trilayer
structure, where we can see that the self-organized and uniformly distributed Au NCs are spheri-
cal in shape, where the range of diameter of Au NCs is from 3 to 4 nm, with a density of about
1.6 × 1012 cm−2.
Figure 13.3 shows the I–V curves of the capacitor structure (with and without Au NCs) measured
with positive bias at different temperatures [8]. Coulomb blockade effect can be observed in the
trilayer structure in the figure, indicated by the periodically modulated current as the bias voltage
increases. As apparent from Figure 13.3, the curve of the control layer (without Au NCs) does not
show any steps in the I–V curve, suggesting that the Au NCs are responsible for the Coulomb block-
ade effect. From the I–V curve measured at 20 K, an average Coulomb voltage can be determined
as ∆V = 31 mV, which corresponds to the Coulomb capacitance C = 5.1 aF based on the relationship
∆V = e/C. It is interesting to notice that peaks instead of flat steps are observed in the I–V curves as
shown in Figure 13.3. It is believed that the current peaks are due to the interplay of resonant tun-
neling and Coulomb blockade effect [8].
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 465

5 nm

HfAlO

Au NCs

HfAlO

p-Si substrate
Pt
Au NC HfAlO

p-Si
20 nm

FIGURE 13.2 Cross-sectional HRTEM image of the trilayer floating gate memory structure, where Au NCs
are evenly distributed in HfAlO matrix. The inset shows the plane-view HRTEM image of the Au NCs. (From
Chan, K.C. et al., Appl. Phys. Lett., 92, 143117, 2008.)

350

300 Three layer

20 K
250 50 K
HfAlO in 20 K
Current (µA)

200

150

100

50

0
2.4 2.5 2.6 2.7 2.8 2.9
Voltage (V)

FIGURE 13.3 I–V curves at 20 and 50 K, showing the steps and the peaks in current. For the control layer
with only HfAlO, there are no steps or peaks that can be observed. (From Chan, K.C. et al., Appl. Phys. Lett.,
92, 143117, 2008.)

13.2.3 Working Principles of SETs

SETs are three-terminal switching devices, which can transfer electrons one by one from source
to drain. The schematic structure of SETs is shown in Figure 13.4. As shown in the figure, the
structure of SETs is almost the same as that of MOSFETs. However, SETs have tunneling junc-
tions in place of pn-junctions of the MOSFETs and a quantum dot in place of the channel region
of the MOSFETs.
466 Semiconductor Nanocrystals and Metal Nanoparticles

Gate Electron
tunneling

Drain Source

Quantum dot

FIGURE 13.4 Schematic structure of a single-electron transistor.

The electrical potential of the quantum dot (island) can be tuned by the gate, capacitively coupled
to the island. As shown in Figure 13.5a, in the blocking state, no accessible energy levels are within
tunneling range of the electron on the source contact. All energy levels on the island electrode with
lower energies are occupied; therefore, the electron on the source side is not able to tunnel through
the isolating junction (dashed arrow, meaning tunneling is prohibited). However, when a positive
voltage is applied to the gate electrode, the energy levels of the island electrode are lowered. The
electron can now tunnel onto the island (solid arrow), occupying a previously vacant energy level,
as illustrated in Figure 13.5b. From there it can tunnel onto the drain electrode where it inelastically
scatters and reaches the drain electrode Fermi level.
The energy levels of the island electrode are evenly spaced with a separation of ΔE. ΔE is the
energy needed to transfer each subsequent electron to the island, which acts as a self-capacitance C.
The lower the C, the bigger the ΔE. It is crucial for ΔE to be larger than the energy of thermal fluc-
tuations kBT; otherwise, an electron from the source electrode can always be thermally excited onto
an unoccupied level of the island electrode, and no blocking can be observed.
The source-to-drain current of SETs can be calculated by using the tunneling rate of an electron
through the tunneling junction. The rate Γ(n, n + 1) of an electron tunneling through a tunneling junc-
tion, which accompanies a transition of the electron number in the dot from n to n + 1, is given by [4]

1 DF (n, n + 1)
G(n, n + 1) = , (13.12)
e Rt 1 - exp éë -DF (n, n + 1)/kBT ùû
2

where
ΔF(n, n + 1) is the free energy change that accompanies the tunneling
Rt is the tunneling resistance of the junction

Quantum Quantum
Drain dot Source Drain dot Source

ΔE
ΔE
(a) (b)

FIGURE 13.5 Energy levels of source, island (i.e., the quantum dot) and drain in a SET for both the blocking
state (a) and the transmitting state (b).
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 467

In the following, the tunneling rate through the junction of the source is denoted by Γs(n, n + 1) and
the tunneling rate through the junction of the drain is denoted by Γd(n, n + 1). The probability pn of
finding n electrons in the dot may change by leaving this state or by coming into this state from the
states n – 1 or n + 1 [4]

dpn
= G tot (n + 1, n) pn +1 + G tot (n - 1, n) pn -1 - [G tot (n, n+1) + G tot (n, n -1)] pn, (13.13)
dt
where

G tot (n, n+1) = G s (n, n + 1) + G d (n, n + 1).

A normalization condition exists:

+¥

åp
n =-¥
n = 1. (13.14)

The current (I) of a SET is obtained by [4]

I = eSpn éëG s (n, n + 1) - G d (n, n + 1) ùû. (13.15)

The characteristics of single-electron devices are usually calculated with numerical simulators [9,10].

13.2.4 Fabrication of Si-Based SETs

Since single-electron phenomena can be observed in any conductive substances, single-electron
devices are fabricated using a variety of materials such as aluminum, GaAs heterostructures, and
silicon [11,12]. However, in order to utilize single-electron devices as elemental devices of ULSI
­circuits, silicon SETs are essential. This can be achieved if fabrication techniques of nanometer-
scaled silicon quantum dots are established. Besides, fabricating silicon SETs by using MOS
processes is advantageous because we can use highly advanced fabrication tools developed for
complementary metal oxide semiconductor (CMOS) ULSI circuits.
Generally, silicon quantum dots formation is possible through two approaches: (1) patterned by
fine-lithography techniques and (2) grown by deposition processes. Using the former approach, it
is possible to accurately define the structures and positions of quantum dots. Takahashi et al. pro-
posed a novel silicon quantum dot fabrication called pattern-dependent oxidation (PADOX) [13]. In
2000, Ono et al. developed the vertical version of PADOX (V-PADOX) [14]. In PADOX, laterally
broad 2-D regions are essential for tunneling barrier formation. The process is also known as Si
width modulation. On the other hand, in V-PADOX, vertically broad, namely, thick, 2-D regions
are utilized for tunneling barrier formation, known as thickness modulation. The advantage of the
V-PADOX is that it makes it possible to form two tiny islands in a small area by utilizing not a litho-
graphic process but the oxidation process, which includes the accumulation of stress in small struc-
tures. Thus, two SETs can be fabricated in an extremely small area. Both PADOX and V-PADOX
are able to be combined with the conventional CMOS. However, the latter approach is favorable
from the viewpoints of throughput and fabricated quantum-dot sizes.

13.2.5 Device Applications of SETs

The implementation of SETs is able to increase the packaging density and to decrease the power
consumption of integrated circuits. Therefore, many logic circuits consisting of SETs have been
developed. There are generally two methods for implementing logic operations in the circuits of
SETs: (1) by representing a bit by a single electron and using field effect to transfer electrons one by
468 Semiconductor Nanocrystals and Metal Nanoparticles

one and (2) by representing a bit by more than one electron and using field effect to switch the current
on or off. The former method is more attractive from the power consumption standpoint. However,
in that case even one erroneous electron caused by noise or thermal agitation will completely alter
the logic operation results. Therefore, from the viewpoint of operation reliability, the latter method
is preferable. Logic circuits realized by the former method resemble those consisting of charge-
coupled devices, whereas logic circuits realized by the latter method resemble those consisting of
MOSFETs. Since present logic circuits consist of MOSFETs, the latter method is preferable.
Figure 13.6a shows the realization of a SET based on nanocrystals reported by Klein et al. [15].
CdSe nanocrystals with a diameter of 5.5 nm were bound to two closely spaced gold leads by
bifunctional linker molecules. The leads were fabricated on a degenerately doped silicon wafer,
which was used as the gate electrode to tune the charge state of the CdSe nanocrystal. Figure 13.6b
shows a field-emission scanning electron microscope (SEM) image of a completed device [15]. It
can be seen that a number of nanocrystals are distributed in the ~10 nm gap between the two elec-
trodes. Although the gap is so small and even filled with nanocrystals, most devices have immeasur-
ably high impedance (R > 100 GΩ). Only about 1 in 20 have a measurable resistance, typically in
the range 10 MΩ to 1 GΩ. Devices with relatively small resistance typically behave like transport
occurs through a single nanocrystal, even though the number of nanocrystals in the junction region
is quite large as can be seen in Figure 13.6b. Since electron tunneling through the linker molecules
has an exponential decay length of less than 1 Å [15], only a well-placed nanocrystal, within 2 nm
of each lead, can contribute to the current conduction.
Figure 13.7 shows the linear device conductance (G) measured at T = 4.2 K as a function of the
gate voltage (Vg) applied to the heavily doped Si substrate. As the gate bias was swept from −9 to
−5 V, the conductance increased to a peak and then declined back to zero. The insets in Figure 13.7
show the plots of the current (I) flowing through the nanocrystal as a function of the voltage (V)
applied between the two leads at two fixed values of Vg. The I–V measured at Vg, which is away
from the linear response peak corresponding to Vg of ~0.7 V, shows a suppressed conductance
at small V whereas the I–V characteristic taken at the center of the peak has a finite linear conduc-
tance for small V. The conductance peak observed in Figure 13.7 is a Coulomb oscillation. The
suppressed conductance on each side of the peak is a consequence of the finite energy required to
add (remove) an electron to (from) the nanocrystal in its ground state. This energy is analogous
to the electron affinity (ionization energy) of a molecule. The peak occurs when the two charge

Nanocrystals

Au leads
V +
Linker molecules
SiO2 insulating layer I

Doped Si substrate
Vg 50 nm

(a) (b)

FIGURE 13.6 (a) Schematic of the single-electron device based on CdSe nanocrystals sit on top of two
closely placed Au leads. The setup of measurement probes applied on the devices is also demonstrated.
(b) Field-emission scanning electron microscope (SEM) image of CdSe nanocrystals with 5.5 nm in diameter
distributed over and in the 10 nm gap of two 13 nm thick leads. (From Klein, D.L. et al., Nature, 389, 699, 1997.)
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 469

5 40
20

I (pA)
40
4 0
20
–20

I (pA)
G (nS)
3 0
–40
–20 –20 –10 0 10 20
V (mV)
2 –40
–20 –10 0 10 20
V (mV)
1

0
–9 –8 –7 –6 –5
Vg (V)

FIGURE 13.7 Conductance (G) versus gate voltage (Vg) for a single nanocrystal transistor measured at
T = 4.2 K. The dots are the measured values and the solid curve is a fit to the data by using the standard
Coulomb blockade model [16] with a temperature T = 5 K. The insets show the I–V characteristics measured
at the gate voltages, indicated by arrows. (From Klein, D.L. et al., Nature, 389, 699, 1997.)

states of the nanocrystal have the same total (including electrostatic) energy and an extra electron
can therefore hop on and off the nanocrystal at no energy cost. The maximum size of the gap
in the I–V curves away from the peak provides a measure of the addition (removal) energy for
electrons.
A map of the differential conductance (dI/dV) of the device as a function of both V and Vg is plotted
as a gray scale in Figure 13.8. The evolution of the Coulomb gap with Vg can be seen in Figure 13.8a
and b. The Coulomb gap is indicated with light-colored diamond-shaped regions in the figure. It can
be observed that the Coulomb gap is zero at Coulomb oscillation; it increases to a maximum value
approximately halfway between two oscillations, and then decreases to zero at the next Coulomb
oscillation. Figure 13.8c shows a schematic evolution of the Coulomb gap for successive electrons
along with the inferred addition energies, which range from ~15 to 30 meV. Additional I–V measure-
ments on eight other samples without a gate have been performed and the typical addition energies
are determined to be in the range of 30–60 meV.
SETs have the properties of low power and good scalability; therefore, they are very promising to
be used in future ULSI circuits. The working physics and mechanisms of SETs are discussed in this
section. Researchers have already developed a number of processes to form the essential quantum dots
mainly based on semiconductor nanocrystals. But in order to operate SET circuits at room tempera-
ture, the size of the quantum dot must be much smaller than 10 nm. Fabricating a structure smaller
than 10 nm is still quite difficult with the present technology. The development of ULSI nanofabrica-
tion techniques is desirable to realize such nanoscale devices. Another challenge is operation stability,
which involves removing random background charge or, alternatively, providing continuous charge
transfer in nanoscale resistors. In future, it is important to develop an operation scheme under which
the functionality of SET circuits can surpass that of conventional CMOS circuits.

13.3 MEMRISTORS
A memristor, a contraction of “memory resistor,” is a two-terminal electronic device whose resis-
tance can be precisely modulated by the current or flux flowing through it. It was theoretically pro-
posed by Leon Chua in 1971 as the fourth basic circuit element, other than the resistor, capacitor,
470 Semiconductor Nanocrystals and Metal Nanoparticles

40
(b)

(a) 20
20

V (mV)
V (mV)

0
0

–20
–20

–40
–15 –10 –5
Vg (V)
–10 –5
Vg (V)

0 10 20 30
dI/dV (nA/V)

Δ3h = 29 meV
Δ4h = 22 meV
20 Δ2h = 14 meV
V (mV)

3 holes 2 holes 1 holes 0 holes

0

–20

0 5 10
ΔVg (V)
(c)

FIGURE 13.8 (a, b) Composite grayscale plots of the differential conductance (dI/dV) of a CdSe nanocrystal
plotted as a function of both Vg and V. The white diamond shaped regions correspond to the Coulomb gap. (c) A
schematic illustration of the data from (a) and (b), indicating the inferred number of holes on the nanocrystal as
a function of Vg and the addition energy for successive holes. (From Klein, D.L. et al., Nature, 389, 699, 1997.)

and inductor [17]. Figure 13.9 shows the four fundamental two-terminal circuit elements and illus-
trates the relationship between different circuit variables [18]. It can be seen that the memristor is
defined by memristance, which can be described by a relationship between charge (q) and flux (φ).
Mathematically, memristive systems can be described by the following set of equations [18]:

v = R(w, i )i, (13.16)

dw
= f (w, i ), (13.17)
dt

where
R is the resistance
w is the internal state variable
f is a function describing the change of w as a function of inputs and the current state
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 471

d = vdt
Resistor Capacitor
dv = Rdi dq = Cdv

i dq = idt q

Inductor Memristor
d = Ldi d = Mdq

Memristive systems

FIGURE 13.9 The four fundamental circuit elements—resistor, inductor, capacitor, and memristor—and
the functional relationships of a pair of the four basic circuit variables (i.e., voltage [V], current [I], charge [q],
and flux [φ]) defined by the circuit elements (i.e., resistance [R], inductance [L], capacitance [C], and
memristance [M]). (From Strukov, D.B. et al., Nature, 453, 80, 2008.)

Although the memristor is physically a simple two-terminal device, it can exhibit highly com-
plex behaviors when different terms are incorporated into these equations. As a result, memristor
devices existed only in theory and papers for almost 40 years.
In 2008, a research group lead by R.S. Williams from Hewlett Packard demonstrated the first
workable memristor in a laboratory, which was composed of nanoscale titanium oxide thin films
sandwiched by two platinum electrodes [18]. The TiOx -based device exhibited a fascinating prop-
erty, which was the dependence of device resistance on time-varying current by means of a chemical
reaction in the interface region between TiO2 and TiOx layers. Specifically, the boundary between
the high-resistance, pure TiO2 layer and the low-resistance TiOx layer can be brought to a nanometer
scale upon the application of electric fields between the top and bottom electrodes. Since the find-
ing of the missing piece in the puzzle of circuit elements shown in Figure 13.9, intensive studies on
memristors have been carried out in multiple aspects, including materials, device structures, and
neural applications.

13.3.1 Memristors Based on Fe3O4 Nanoparticle Assemblies

Regarding nanocrystals and nanoparticles, Kim et al. reported a new addition to the memristor
family with nanoparticle assemblies consisting of an infinite number of monodispersed single
crystalline Fe3O 4 nanoparticles [19]. A nonhydrolytic chemical method was used to prepare
single crystalline Fe3O 4 nanoparticles, which have diameters of 7, 9, 12, and 15 nm with a
well-controlled size monodispersity (σ ≈ 5%). The surface ligands of synthesized nanopar-
ticles were carefully removed by washing with tetramethylammonium hydroxide solution for
a pristine, inorganic surface. The nanoparticle assemblies in the form of compact pellets with
dimensions of 0.5 × 1 × 4 mm were then fabricated by coldpressing in a die under 160 Pa for
472 Semiconductor Nanocrystals and Metal Nanoparticles

15 min. There was no heat treatment during the preparation of the pellets, and thus the surface
properties of the nanoparticles were not altered.
The voltage–current (V–I) characteristics of the nanoparticle assembly pellets were measured
using a conventional four-point probing method, as shown in Figure 13.10a. In the case of nanopar-
ticles with a diameter (D) of 7 nm (see Figure 13.10b), a type of hysteresis is observed at room
­temperature, as shown in Figure 13.10c. The bistable V–I characteristics illustrated that the switch-
ing in measured voltages are directly related to the existence of hysteretic behavior as the current
sweeping was applied in sequences of 1–6, corresponding to 0 → +20 nA → 0 → −20 nA → 0. As the
current was increased from 0 to +20 nA, the switching from a low-resistance state (RON ≈ 2 × 107 Ω)
to a high-resistance state (ROFF ≈ 4 × 108 Ω) occurred at I ≈ +16 nA, as shown in Figure 13.10c. In
contrast, ohmic behavior was seen when the current decreases from +20 nA to 0. The same behavior
was observed when the current sweeping direction was reversed. This type of instantaneous switch-
ing from the low-resistance (i.e., on) state to the high-resistance (i.e., off) state occurred in a typi-
cal ROFF/RON ratio of 20:1. Stable V–I characteristics were observed when sweeps were carried out
repeatedly, suggesting good endurance in the memristors based on Fe3O4 nanoparticle assemblies.

13.3.2 Memristors Based on Amorphous Si Embedded with Ag Nanoparticles

Similar to the TiO2/TiOx bilayer structure invented by the HP research group, a memristor based on
­ g nanoparticle–embedded Si thin films was reported by Jo et al. [20]. The memristor device consists
A
of a bottom tungsten nanowire electrode, a sputtered silicon layer (2–4 nm), an amorphous silicon

(a) Fe3O4 nanoparticle assemblies

10 +Imax

D = 7 nm ROFF
3
T = 295 K
5
Voltage (V)

4 RON
0 2
(400) 5 1
2.06 Å
(220)
–5
2.99 Å 6
–Imax
20 nm 1 nm –10
–20 –10 0 10 20
(b) (c) Current (×10–9 A)

FIGURE 13.10 (a) Schematic of the measurement setup for V–I analysis of the nanoparticle assemblies.
(b) Transmission electron microscopy (TEM) and high-resolution TEM (HRTEM) images of 7 nm sized
Fe3O4 nanoparticles. (c) V–I characteristics measured at room temperature for 7 nm sized Fe3O4 nanoparticle
assemblies. (From Kim, T.H. et al., Nano Lett., 9, 2229, 2009.)
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 473

(a-Si) layer (2.5–4.5 nm) deposited by plasma-enhanced chemical vapor deposition (PECVD), a
co-sputtered silver and silicon layer (20–30 nm thick), and a top chrome/platinum nanowire electrode,
as schematically illustrated in Figure 13.11. The bottom tungsten nanowire electrode was fabricated
by e-beam lithography, mask layer lift-off, and reactive ion etching. After that, a-Si deposition,
co-sputtering, and top metal lift-off processes were performed, followed by another reactive ion
etching step to remove the co-sputtered layer outside the cross-point regions defined by the perpen-
dicularly crossed top and bottom electrodes.
As schematically shown in Figure 13.11, the device consists of multiple film structure with
active layer of sputtered Ag and Si, which has a properly designed Ag/Si mixture ratio gradient.
More specifically, a Ag-rich (high-conductivity) region and a Ag-poor (low-conductivity) region
were formed, respectively, in the fabrication process, as shown in Figure 13.11. Typically, resis-
tance switching devices require an electroforming process to activate the switching property
regardless of switching material. During electroforming, metal ions or particles are injected
into the host medium and cause semipermanent structural modifications inside the insulating
storage layer. The forming process creates localized conducting paths (filaments), resulting
in discrete, abrupt resistance switching characteristics. Through co-sputtering of Ag and Si,
nanoscale Ag particles and clusters were incorporated into the Si host medium during device
fabrication, and a uniform conduction front between the Ag-rich and Ag-poor regions was
formed. As a result, the forming process was not required in the Ag/Si-based devices. In addi-
tion, instead of discrete, localized conducting filament formation, the continuous motion of
the conduction front under applied bias in the co-sputtered memristor device results in reliable
“analog” switching behaviors (i.e., a gradual change in resistance instead of abrupt changes),
as shown in Figure 13.12a.
Figure 13.12a shows the measured current (blue lines) through the memristor as a function
of the applied voltage across the device for five consecutive positive voltage sweepings followed
by five consecutive negative voltage sweepings. Being different from resistive switching devices
that show abrupt conductance switching, the conductance in devices based on Ag-embedded Si
thin films continuously increased (or decreased) during the positive (or negative) voltage sweeps,
and the slope of I–V curve in each subsequent sweep picked up where the last sweep left off.

Top
Cr/Pt
electrode

Ag + Si

a-Si

Bottom
W
electrode

FIGURE 13.11 The device structure of a two-terminal memristor. The active layer is a two-layered structure
based on amorphous Si thin films with and without Ag nanoparticles, respectively. (From Jo, S.H. et al., Nano
Lett., 10, 1297, 2010.)
474 Semiconductor Nanocrystals and Metal Nanoparticles

1.0 0.0
1.0

Front position w
Current (µA)

Calculated
0.5 Extracted –0.2
0.5
2 3 4 5
No. of sweeps
Calculated
Data

0.0 –0.4
0 1 2 3 –2 –1 0
(a) Voltage (V)

0.8

2
0.4

Current (µA)
Voltage (V)

0 0.0

Current –0.4
–2 Voltage

0.0 4.8 9.6 14.4 19.2

(b) Time (s)

FIGURE 13.12 Memristor characteristics in Ag-embedded a-Si thin-film structures. (a) Measured (blue
lines) and calculated (orange lines) I–V characteristics of the memristor (device size: 500 nm × 500 nm).
The inset shows comparison of calculated and extracted values of the normalized Ag front position w during
positive DC sweepings. (b) The replotted current and voltage data from (a) as a function of time, highlighting
the current change in sequential voltage sweeps. (From Jo, S.H. et al., Nano Lett., 10, 1297, 2010.)

The I–V relationship in this device can be well fitted by a memristor circuit model (orange lines
in Figure 13.12a) described as [18]

1
i (t ) = v(t ), (13.18)
RON w(t ) + ROFF (1 - w(t ))

where w(t) represents the normalized position of the conduction front between the Ag-rich and
Ag-poor regions within the active device layer and its value varies between 0 and 1. When a posi-
tive voltage bias is applied on the device, Ag ions driven by the electric field move from the Ag-rich
region to the Ag-poor region and increase w, and vice versa. As w(t) approaches 1 (or 0), the device
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 475

reaches the smaller (highest) resistance state with a resistance of RON (or ROFF). In this model, the
position w(t) can be assumed to be a linear function of the flux-linkage φ(t) = ∫v(t)dt through
the device. Equation 13.18 can then be rewritten as [20]

i(t ) = G(j(t ))v(t ), (13.19)

which is the equation for a flux-controlled memristor and G(φ(t)) is the memductance. However, bias
voltages with amplitude greater than a threshold (VT) of 2.2 V were required to drive the Ag ions inside
the a-Si matrix and voltages smaller than V T have negligible effect on the memristor resistance.
Based on the memristor model discussed, the current values (orange lines) were calculated as a
function of the voltage sweepings and plotted in Figure 13.12a as orange lines [20]. The measured
current (blue lines) is also shown in Figure 13.12a and a good consistency can be observed between
the measured and calculated curves. In addition, the values of w(t) (orange lines) used to calcu-
late the current during five consecutive positive voltage sweepings is also included as an inset of
Figure 13.12a, by assuming w(t) to be a linear function of the flux-linkage φ(t). In comparison, the
values of w(t) can be also extracted from the data using Equation 13.18, which are plotted in the inset
of Figure 13.12a, indicated with blue lines. The good agreements between the calculated and the
measured values in i(t) and w(t) suggest that the device characteristics above the V T can be explained
by the flux-controlled memristor model using Equations 13.18 and 13.19, where the front position is
roughly a linear function of the flux-linkage. Due to the existence of a threshold effect, the device is
actually not a true memristor but falls in the more broadly defined memristive device category [21].
The flux-controlled memristor model introduced suggests that the conductance in the device can
be incrementally modulated by tuning the duration and sequence of the applied programming volt-
age. Figure 13.13a shows the memristor response when the device was programmed by 100 consecutive
identical positive pulses of 3.2 V, 300 μs followed by 100 identical consecutive negative voltage pulses of
−2.8 V, 300 μs [20]. The device current was measured at a small read voltage of 1 V after each program-
ming pulse. As expected from the previous sweeping I–V characteristics of the device shown in Figure
13.12a, the application of positive voltage pulses (i.e., the potentiating signal, P) incrementally increases
the memristor current (equivalently the conductance), and the application of negative voltage pulses (i.e.,
depressing signal, D) incrementally decreases the memristor current or conductance, as can be seen from
Figure 13.13a. It is thus noticed that the memristor conductance can be controlled and tuned by the flux-
linkage φ(t) as a result of the application of programming pulses. This effect was further demonstrated
in Figure 13.13b involving the application of mixed bipolar pulses. Mixed positive and negative voltage
pulses with fixed pulse height but different duration were applied to the device and the change in mem-
ristor conductance (ΔG) was measured after each pulse. A proportional correlation between ΔG and
the duration of the applied pulses was observed. More specifically, the application of a longer positive
(or negative) pulse caused a larger increase (or decrease) in the memristor conductance, and vice versa.

13.3.3 Electronic Synapses Realized with Si-Based Memristors

A synapse bears striking resemblance to the memristor, introduced in the previous section. Similar
to a biological synapse, the conductance of a memristor can be incrementally tuned by carefully
controlling the flux-linkage through it. Realization of an electronic device that can emulate the func-
tions of biological synapse is essential to the development of next-generation computing technology.
Currently, the computers based on the von Neumann architecture, which fetch, decode, and execute
instructions in a sequential process, are ideal to solve structured problems, such as mathematical
problems. For solving problems where interactions with the real world are involved and no specific
instructions are defined, modern digital computers become very inefficient, and such cases usually
require ­systems with great complexity and huge power consumption. Though current digital comput-
ers possess the computing speed and capability to emulate brain functionality of certain animals, the
associated complexity and energy dissipation in the system grows exponentially along the hierarchy
476 Semiconductor Nanocrystals and Metal Nanoparticles

0.4

0.3
Current (100 nA)
P D

0.2

0.1

0.0
0 40 80 120 160 200
(a) Pulse #

4
Applied signal (V)

–2
Time (ms)

15 0 2 4 6 8 10 12 14
Conductance change (%)

–15

(b)

FIGURE 13.13 The evolution of the memristor current as a result of the programming pulses. (a) The incre-
mental increase and decrease of device current read at 1 V as a result of consecutive positive potentiating (P)
pulses (3.2 V, 300 μs) followed by negative depressing (D) pulses (−2.8 V, 300 μs). (b) Mixed potentiating and
depressing pulses with different pulse widths used for programming (top); and change of the measured device
conductance after the application of each pulse. The conductance change was normalized to the maximum
memristor conductance. (From Jo, S.H. et al., Nano Lett., 10, 1297, 2010.)

of animal intelligence. For example, to perform certain cortical simulations at the cat scale even at
83 times slower firing rate, the IBM team had to employ Blue Gene/P (BG/P), which is a supercom-
puter equipped with 147,456 CPUs and 144 TB of main memory [20,22]. On the other hand, being
dramatically different from the von Neumann digital architecture, which stores and processes data
separately in modern computers, brains of biological creatures are configured in large connectivity
between neurons (~104 in a mammalian cortex), offering highly parallel processing power and high
efficiency. The connectivity strength between two neurons (i.e., the so-called synaptic weight) can
be precisely adjusted by the ionic flow through them and it is widely believed that the adaptation of
synaptic weights enables the memorization and learning ability of biological systems.
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 477

Memristor
Memristor synapse

Ag + Si s
on
eur
Si -n
re
Sp
O
CM
Synapse
CM
OS
pos
t-ne
To pre-neuron u ron
s
To post-neuron
(a) (b)

FIGURE 13.14 (a) Schematic illustration of the concept of using the memristor introduced in Figure 13.11
to emulate a synapse between two neurons. (b) Schematic of a neuromorphic system with CMOS neurons and
memristor synapses in a cross-point configuration. (From Jo, S.H. et al., Nano Lett., 10, 1297, 2010.)

Jo et al. demonstrated the experimental implementation of biological synapses with nanoscale

silicon-based memristors, which can emulate an advanced synaptic function, namely, spiking-time-
dependent plasticity (STDP) [20]. STDP is an important synaptic adaption rule for competitive
Hebbian learning and it can be achieved in a hybrid synapse/neuron circuit composed of CMOS neu-
rons and nanoscale memristor synapses, as illustrated in Figure 13.14a. These demonstrations provide
a direct experimental support for memristor-based neuromorphic systems. Figure 13.14b shows the
implementation of a crossbar hardware structure, where a two-terminal memristor synapse is formed
at each cross-point and connects CMOS-based pre- and postsynaptic neurons [20]. The crossbar-
configured network based on memristors can potentially offer connectivity and function density com-
parable to those of biological systems. In such a network, data will be processed and stored in a
parallel approach, which is analogous to biological systems rather than digital computers. As shown
in Figure 13.14b, every CMOS neuron in the “pre-neuron” layer is directly connected to every neuron
in the “post-neuron” layer at each cross-point with unique synaptic weights. A high synaptic density of
1010 cm−2 can be potentially obtained for crossbar networks with 100 nm pitch [20].
Memristor devices based on Ag nanoparticles embedded with a-Si thin films are capable of
emulating biological synapses with properly designed CMOS neuron components to provide pro-
gramming voltage pulses with controlled pulse width and height [20]. The neuron circuit consists
of two CMOS-based integrate-and-fire neurons connected by a nanoscale memristor with an active
device area of 100 nm × 100 nm. Specifically, being analogous to the synaptic behavior of biologi-
cal systems, the neuron circuit generates a potentiating or a depressing pulse across the memristor
synapse when the presynaptic neuron spikes before or after the postsynaptic neuron, with the pulse
width being an exponentially decaying function of the relative neuron spike timing (Δt = tpre-neuron −
tpost-neuron, where tpre-neuron and tpost-neuron are the time when the presynaptic neuron and postsynaptic
neuron fire spikes, respectively).
Figure 13.15 shows the measured change of the memristor synaptic weight after each neuron spik-
ing event obtained in the hybrid CMOS-neuron/memristor-synapse circuit [20]. When the pre-neuron
spikes before (or after) the post-neuron, the synaptic weight of the memristor increases (or decreases).
Moreover, as can be seen from Figure 13.15, the change in the synaptic weight versus the spike tim-
ing difference Δt can be well fitted with exponential decay functions. It is thus verified that an STDP
function analogous to that of biological synaptic systems can be obtained in such memristor synapses.
Figure 13.16 presents the endurance of the device after continuous applications of the potentia-
tion and depression pulses [20]. As can be seen in Figure 13.16b, after 1.5 × 108 times of operations
with P/D pulses, the device shows certain degradation but the evolution of device conductance
478 Semiconductor Nanocrystals and Metal Nanoparticles

20

10

Δ Synaptic weight (%)

–10

–20
–60 –40 –20 0 20 40 60
Δ Spike timing (ms)

FIGURE 13.15 The measured change of synaptic weight in the memristor as a function of the relative ­neuron
spiking timing (Δt), demonstrating the STDP function. The inset shows an SEM image of the fabricated
­memristor crossbar array with a scale bar of 300 nm. (From Jo, S.H. et al., Nano Lett., 10, 1297, 2010.)

still follows the P/D modulation. Assuming that biological synapses are communicated at a rate
of 1 Hz [23], this endurance provides the capability of continuous synaptic operation for around
5 years. The endurance characteristic together with the large density offered by the simple two-
terminal memristor synapses in the crossbar configuration, as shown in Figure 13.14b, make the
hybrid CMOS–neuron/memristor–synapse approach very promising for the hardware implementa-
tion of biology-inspired neuromorphic systems.

13.3.4 Electrochemical Dynamics of Metallic Nanoparticles in Dielectrics

In device applications, nanocrystals and nanoparticles are usually embedded in certain matrices with
insulating or semiconducting properties. The ability to control and manipulate nanoscale inclusions
in the solid-state medium will greatly expand the functions of materials and create new devices
for electronic and optoelectronic applications. Therefore, it is vital to understand how nanopar-
ticles behave in electrical and chemical aspects. As mentioned above, novel devices like memristors
and electronic synapses are mostly based on metallic nanoclusters dispersed in a dielectric matrix.
Thus, in this section, the metallic nanoparticles in dielectrics are taken as an example to elaborate
the electrochemical dynamics in the nanocomposite materials.
Yang et al. studied the origin of the dynamic growth and migration processes of metal nano-
clusters in dielectrics with microscopic techniques [24]. In situ transmission electron microscopy
(TEM) studies were carried out on samples consisting of individual metal nanoclusters embedded
in a dielectric host film. Detailed analysis and tracking of individual clusters were conducted and
the migration inside the bulk of the film was investigated. By tracing the dynamic motion of nano-
clusters under electric field in an in situ TEM, fundamental atomic/ionic dynamics were revealed.
The experimental strategy is illustrated in Figure 13.17a through d, where Ag nanoclusters were
used. As illustrated in Figure 13.17a, Ag nanoclusters were included in the SiO2 film by annealing a
stacked SiO2/Ag/SiO2 structure with an Ag layer of ~1 nm thickness. Au electrodes were then depos-
ited onto the SiO2/Ag/SiO2 stack. Subsequently, successive focused ion beam lift-out, patterning,
and thinning processes were performed to prepare an array of Au/SiO2/Ag nanoclusters/SiO2/Au
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 479

0.8

0.7

P D
Current (100 nA)
0.6

0.5

0.4

0.3
0 20 40 60 80
(a) Pulse # (after 1.0 × 107 cycles)

1.2

1.0
P D
Current (100 nA)

0.8

0.6

0.4

0.2
0 20 40 60 80
(b) Pulse # (after 1.5 × 108 cycles)

FIGURE 13.16 Endurance of the memristor device after (a) 1.0 × 107 and (b) 1.5 × 108 potentiating/
depressing (P/D) cycles. In each test, 3.1 V, 800 μs potentiating pulses, −2.9 V, 800 μs depressing pulses, and
1 V, 2 ms read pulses were used. After each P or D pulse, the device conductance was measured by a read pulse
and recorded. (From Jo, S.H. et al., Nano Lett., 10, 1297, 2010.)

devices for cross-sectional TEM studies. As shown in Figure 13.17d, the top Au electrode of a device
is contacted by a movable W probe that applies voltages during in situ TEM analysis. Figure 13.17e
shows a TEM image for an embedded Ag nanoparticle along its [011] zone axis, and Figure 13.17f
displays the corresponding fast Fourier transformation results, which clearly reveal the crystalline
face-centered cubic (FCC) structure of the Ag nanocluster.
480 Semiconductor Nanocrystals and Metal Nanoparticles

Au
SiO2
Au
NiCr BE
SiO2 W probe
Si

(a) (b) (c) (d)

Ag –– –
(111) (111) 1 1
–
(200)
+ – + –
(200) 2 2
3 3
–– –
(111) (111) 0s 0.3 s
(e) (f ) (g) (h)

+ – + – + – + –
2 2
3 3 3
0.6 s 1.5 s 22 s 4 40 s 4
(i) (j) (k) (l)

+ – + – + – + –
0s 27 s 34 s 50 s
(m) (n) (o) (p)

E E E

+ – + – + –
δ– δ+ δ– δ+δ–δ+ δ–δ+

(q) (r) (s)

FIGURE 13.17 Dynamic growth and migration of Ag nanoclusters in a SiO2 film. (a) Schematic of the
device structure consisting of Ag nanoclusters embedded SiO2 film. (b) SEM image showing the FIB lift-out
process. Scale bar: 20 mm. (c) SEM image showing an array of samples prepared by FIB. Scale bar: 1 mm.
(d) TEM image showing the in situ experimental setup. Scale bar: 200 nm. (e) High-resolution TEM image
and (f) the corresponding fast Fourier transformation results. Scale bar: 1 nm. (g–l) Real-time TEM images
revealing the electrical field–induced growth and migration of Ag nanoclusters. Scale bar: 10 nm. (m–p)
In situ TEM images together with simulation results showing the dissolution and growth processes between
two Ag nanoclusters. Scale bar: 10 nm. (q) Schematic of a polarized Ag nanocluster embedded in SiO2 under
electric field, overlaid with the electric potential. (r) New cluster nucleation and formation of another bipolar
electrode nearby. (s) The dissolution of the original cluster and the growth of the new cluster lead to an effec-
tive movement of a cluster along the electric field. (From Yang, Y. et al., Nat. Commun., 5, 4232, 2014.)
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 481

The evolution and migration of embedded Ag nanoclusters can be regularly observed when an
electric field (~0.8 MV/cm corresponding to an applied voltage of 3 V) is applied across the Au
electrode, as shown in Figure 13.17g through l. The Ag nanoclusters formed in SiO2 have diameters
of ~5 nm. At location 1, dynamic interactions between two Ag nanoclusters were observed. As
shown in Figure 13.17g through i, the nanocluster on the left was gradually dissolved and included
into another cluster on the right side, resulting in an overall movement of Ag nanoclusters along
the direction of the applied electric field. Similarly, clusters 2 and 3 were also gradually dissolved.
Cluster 2 disappeared with a void left in the film, as shown in Figure 13.17j and k. As for cluster 3, its
dissolution resulted in the creation of a new cluster (namely, cluster 4) nearby, indicated by the lower
circle in Figure 13.17k and l. Figure 13.17m through p shows the interactions between another pair
of Ag nanoclusters, illustrating the gradual dissolution of the original nanocluster in the upstream
and the nucleation and growth of a new nanocluster in the downstream. Effective growth along the
field direction was achieved as the original Ag nanocluster was completely dissolved in a location
where a void was left behind (indicated by a circle in Figure 13.17p).
The dynamic process of Ag nanoclusters revealed by in situ TEM can be explained by electro-
chemical redox processes in bipolar electrodes with strong influence from kinetic effects. Under an
electric field inside the dielectric, the metal nanoclusters can be treated as bipolar electrodes, with an
effective cathode (δ−) facing the field direction and an effective anode (δ+) on the opposite side, as
shown in Figure 13.17q. In this model, electrochemical oxidation processes occur on the anode side
of the clusters and Ag+ ions are generated. The Ag+ ions drift under the applied electric field and then
reduced upon reaching a cathode (e.g., the cathode side of another Ag cluster in the downstream). In
the Au/SiO2/Ag nanoclusters/SiO2/Au devices, there is no Ag supply to the first Ag cluster upstream,
so it experiences only net oxidation and thus gradually dissolves. The created Ag+ ions can be reduced
and deposited onto the second nanocluster downstream. If oxidation on the anode side of the sec-
ond nanocluster is suppressed due to kinetic limitations, the second nanocluster will experience a net
growth. As a result, an effective cluster displacement along the electric field is realized by the merging
of the first and second nanoclusters downstream, as shown in Figure 13.17q through s. On the other
hand, the redox and migration processes of a specific nanocluster are determined by the local struc-
tural and or electrochemical environment sensed by the cluster and are affected by film inhomogeneity
and electric field distribution, so the dynamic growth and migration are stochastic in nature.
It is found that the formation of the second nanocluster inside the SiO2 film requires a nucle-
ation process, which is the initial formation of a critical nucleus. This process can be regarded as a
chemical process of formation of stable atomic configuration to minimize the excess surface energy
and allow further growth. As a result, the nucleation process is spatially selective instead of being
homogeneous, as shown in Figure 13.17. If nucleation could not proceed, the Ag+ ions can reach
the Au counter electrode and become reduced there. This appears to be the case for cluster 2 in
Figure 13.17g through l. However, a significant geometric change may not be observed in this case
due to the much larger volume of the electrode.
These analyses show that kinetic factors of the electrochemical redox reactions and ionic drift
strongly influence the dynamics of metal nanoclusters. The evolution of the shape and size of a
nanocluster depends on the competition between the oxidation and reduction rates on the cathode
and anode sides of the bipolar electrodes. The relationship can be described as [24]

dV i
µ Gired - Giox, (13.20)
dt

where
V i is the volume of cluster i
Giox and Gired is the oxidation and reduction rate, respectively, on the anode and cathode sides of
cluster i, respectively
482 Semiconductor Nanocrystals and Metal Nanoparticles

For two neighboring nanoclusters (i and i − 1) with small distance, the exchange of Ag ions can be
approximated as [24]

Gired = Giox-1, (13.21)

accounting for the conservation of the Ag material.

The experimental results show that the clusters in the upstream merge into their counter-
parts in the downstream, for example, clusters 1 and 3 in Figure 13.17g through l and the clus-
ter in Figure 13.17m. Figure 13.17m through p shows the consistency between the experimental
observations and simulation results using Equations 13.20 and 13.21 by assuming Giox = 0 for the
second nanocluster. Other behaviors such as displacement (movement) of nanoclusters along the
electric field toward the counter electrode can also be expected from Equation 13.20, as shown
in Figure 13.17s. Since SiO2 is an amorphous insulator, the conditions for ion migration and
nucleation are inhomogeneous and depend strongly on the local microstructure, electric field
distribution, electrochemical environment, and the presence of defects, as observed in Figure 13.17g
through p. Different metal systems were extensively studied by Yang et al. as well, and it is
concluded that the apparent movements of metal nanoclusters are always along the electric field
direction. However, different electric fields are required for the migration of metals within the
measurement duration.
In situ TEM studies were also conducted on filament effects in actual resistive memory struc-
tures [24]. As shown in Figure 13.18a, devices with a Ag/SiO2/W structure were formed by evapo-
rating SiO2 film onto a W probe, which then connects with a high-purity Ag wire. Measurement
results for a device with a ~40 nm SiO2 film are shown in Figure 13.18b through e.
The filament growth in the SiO2 film started with the appearance of several Ag clusters near the
Ag electrode after ~3 min (Figure 13.18b). These clusters then behave as bipolar electrodes during
subsequent growth. As expected, over time these clusters move closer to the W electrode, follow-
ing the splitting–merging processes (Figure 13.17m through p). Due to the higher concentration
of Ag+ ions near the Ag electrode and therefore the higher probability to overcome the nucleation
barrier, more Ag clusters will be nucleated near the Ag electrode inside SiO2, and this repeated
nucleation and growth leads to a filament shape, as observed in Figure 13.18d. The growth mode
results in a conically shaped filament (Figure 13.18d and e) with its base at the Ag/SiO2 interface
and expanding toward the inert W electrode. Such a growth direction can be explained by the
generalized framework discussed earlier. The overall growth of the filament can then be charac-
terized by the step-by-step movements of Ag cluster components driven by the electrochemical
kinetics, as illustrated by Figure 13.18f through j. The details include (f) Ag ionization at the
active Ag electrode; (g) nucleation of Ag nanoclusters near the active electrode, where the nano-
clusters act as bipolar electrodes in subsequent filament growth; (h) growth of Ag nanoclusters
at new positions that are closer to the W electrode; (i) migration of Ag nanoclusters leads to void
formation marked by the dashed line; and (j) continued filament growth by void refilling from the
active electrode.
In addition, the formation of a void was observed in the SiO2 film at the interface region in
Figure 13.18d and e after large amounts of Ag atoms passed through. The void formation indicates
the plastic deformation of the dielectric film due to increased mechanical stress. Similar voids have
also been observed in the cases of discrete nanoclusters (e.g., Figure 13.17j through l and p). Apart
from facilitating fast ion diffusion, the voids provide space for continuous movement of Ag clusters
in a bootstrapping fashion, as shown in Figure 13.18d and e. As the existing Ag nanoclusters move
away toward the W electrode, new nanoclusters can move in and fill the void. The continuous move-
ment of Ag clusters facilitates filament growth toward the inert W electrode. The observation of
dynamic void formation and filament growth processes is fundamentally important for the under-
standing and optimization of memristive device operations.
 Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles
Source

SiO2 e-Beam
–W Ag + Ag Ag Ag Ag
W W W W

Sensor 180 s 430 s 437 s 437 s

(a) (b) (c) (d) (e)

+ + + + +
– – – – –
δ+ δ–
(f ) (g) (h) (i) (j)

FIGURE 13.18 In situ TEM observation of bootstrapping filament growth and void formation. (a) Schematic of the experimental setup. (b–e) Real-time TEM images
showing continuous filament growth in a 40 nm SiO2 film under the application of 8 V. Scale bar: 20 nm. (f–j) Schematic illustration of the bootstrapping filament growth
mode. (From Yang, Y. et al., Nat. Commun., 5, 4232, 2014.)

483
484 Semiconductor Nanocrystals and Metal Nanoparticles

13.4 SELF-LEARNING DEVICES

As one kind of smart materials, nanoparticles/nanoclusters-related materials can be used in smart
devices for neural applications. Neural networks have an important characteristic, which is the
learning ability, and thus human/living organisms can react according to input signals based on
their experiences. Such neural networks have been realized with conventional MOS devices, but
complicated software programming with additional devices and circuits are required due to lack of
inherent self-learning ability in the conventional devices [25]. As a result, due to hardware complex-
ity and huge energy dissipation, biomimetic computing architectures with self-learning abilities are
difficult to be employed in current computational systems, which is similar to the case discussed [22].
Realization of devices with inherent self-learning abilities could be one of the fundamental solu-
tions to overcome this obstacle.

13.4.1 Self-Learning Behaviors in Ni-Rich NiO Thin Films

Liu et al. demonstrate a resistive switching device based on a Ni-rich nickel oxide thin film that
has an inherent self-learning ability similar to that of the neural networks in the human brain [26].
Depending on the experience of voltage pulses or sweeping in a way similar to that of the human
brain, the device exhibited multiple resistance states corresponding to short-term memory (STM)
or long-term memory (LTM). In addition, through the modulation of its resistance according to the
number of external electrical pulses, the device was able to mimic neuroplasticity, which is another
important function of neural networks.
As shown in Figure 13.19a, device fabrication started with thermal growth of 200 nm SiO2 thin
film on a p-type Si wafer. A 10 nm Ni layer followed by 200 nm Au layer were deposited onto the
SiO2 film with electron-beam evaporation to form the bottom electrode. A Ni-rich nickel oxide thin
film with a thickness of ~60 nm was then deposited onto the bottom electrode. The deposition was
carried out by radio-frequency (rf) (13.6 MHz) magnetron sputtering of a NiO target (99.99% in
purity) with an Ar flow rate of 75 sccm and the rf power of 150 W. A 120 nm Au/10 nm Ni layer was
deposited onto the nickel oxide thin film by electron-beam evaporation to form top electrodes. The
devices actually had a simple metal-insulator-metal (MIM) structure. An analysis on the chemi-
cal states of the synthesized nickel oxide thin film with x-ray photoelectron spectroscopy (XPS)
revealed that the Ni:O atomic ratio was 1.5:1, indicating that the as-deposited nickel oxide thin film
was Ni rich. Naturally, the elemental Ni component was believed to be dispersed in the NiO matrix
as Ni nanoparticles or nanoclusters.
The device shown in Figure 13.19a is basically a typical MIM structure used in the resis-
tive random access memory (RRAM); therefore, it exhibits a resistive switching behavior [26].
Figure 13.19b shows the typical I–V characteristics of the device in voltage-sweeping measurement
with a current compliance of 0.1 A. When the voltage stated with 0 V was increased to ~1.8 V,
there was an abrupt increase in the current, showing that the device resistance was switched from
a high-resistance state to a low-resistance state. In the second voltage sweeping, a sudden drop of
current appeared when the voltage reached ~0.6 V, showing that the resistance was switched back
to the high-resistance state. This is a typical resistive switching behavior, which can be controlled
by voltage sweeping applied to the electrodes of the device. The formation and rupture of conduc-
tive filaments in the material system are believed to be responsible for the resistive switching [26].
The self-learning ability was realized based on the resistive switching characteristics of the NiO
thin film–based device, and achieving STM, LTM, and a transition from STM to LTM was possible
with this device.
The memorization was realized in the device as a function of the number of the electrical pulses
applied across the top and bottom electrodes [26]. Pulses of 0.5 V, 0.1 s apart were applied to the
device in a low-resistance state, and the evolution of device resistance was recorded. As shown in
Figure 13.20a, three memory stages can be observed, which are the unmemorized stage, the STM
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 485

100 µm

Au (120 nm)/Ni (10 nm)

Ni-rich NiO (60 nm)

Au (200 nm)/Ti (10 nm)

SiO2 (200 nm)

p-Type Si substrate

(a)

Low-resistance state

10–1

10–2
Current (A)

High-resistance state
10–3

10–4
First voltage sweeping
Second voltage sweeping
10–5
0.0 0.5 1.0 1.5 2.0 2.5 3.0
(b) Voltage (V)

FIGURE 13.19 (a) Schematic of the self-learning device structure and (b) I–V characteristics of the device in
the voltage sweeping measurement. (From Liu, Y. et al., Appl. Phys. A, 105, 855, 2011.)

stage, and the LTM stage. In the unmemorized stage of the first 16 pulses, the device resistance
was ~5 Ω and there was no obvious change. This stage imitates an unmemorized state in human
neural networks, meaning the information is not stored in the brain (i.e., no neuronal connections
are established) although we experience some external impression. In the next 17th–36th pulses
corresponding to the STM stage, the device resistance increased gradually with the number of
applied pulses. This is an intermediate state between the unmemorized stage and the LTM stage
since the resistance is not stable. These slight increases in device resistance in the intermediate
states originated from the rupture of localized filaments in the material system [27]. In this stage,
the memorization is formed and gradually enhanced with external impression. However, it is still
not stable, and the memorization can be easily lost. Experimental results in Figure 13.20b show
that the STM state can switch back to the initial state after a waiting time of ~1500 s, exhibiting
a memory loss behavior. In the LTM stage, the resistance was stabilized in a saturation state. The
saturated resistance can be maintained for a considerably long time if the conductive filaments in
the material system are fully ruptured due to Joule heating process [28,29]. The stable memorization
is analogous to the LTM in human brain.
486 Semiconductor Nanocrystals and Metal Nanoparticles

Unmemorized STM LTM

350

300

250
Resistance (Ω)
200

150

100

50

0 20 40 60 80
(a) Number of pulses

100

STM
10
R/R0

Initial state
1

0.1
100 101 102 103 104 105
(b) Waiting time (s)

FIGURE 13.20 (a) The evolution of device resistance as a function of the number of pulses and (b) the
change of normalized resistance of the STM state with waiting time. (From Liu, Y. et al., Appl. Phys. A, 105,
855, 2011.)

As can be seen from Figure 13.20a, for the STM to occur, 16 pulses of 0.5 V, 0.1 s are required.
If the voltage magnitude or the duration of applied pulses changes, a different number of pulses will
be needed to trigger the STM formation. Figure 13.21 shows the dependence of minimum number
of pulses on the voltage magnitude and pulse width for the STM to occur in the device. The STM
was considered to be formed when the resistance was increased by an order. For each pulse voltage/
width, 10 devices located at different places on the wafer were measured to obtain the minimum
number of pulses required statistically. Small pulse voltages (i.e., below 0.4 V) were not able to
form STM regardless of the pulse width. This is due to the fact that localized Joule heating and or
electric fields at small voltages are not sufficient to cause the rupture of localized filaments, which
are formed in the low-resistance states [27]. For the pulse with a width of 0.005 s, an average of
43 pulses with a voltage magnitude of 0.5 V were required to form the STM; however, the average
number was reduced to 20 if 0.6 V pulses were used. On the other hand, with the voltage magnitude
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 487

100
Max. 0.5 V for 0.005 s
0.5 V for 0.01 s
80 75% 0.5 V for 0.1 s
50%
0.6 V for 0.005 s
25%
0.6 V for 0.01 s
Number of pulses
60 Min. 0.6 V for 0.1 s

40

20

FIGURE 13.21 Minimum numbers of pulses required for the STM to occur for different combinations of
voltage magnitude and pulse width. (From Liu, Y. et al., Appl. Phys. A, 105, 855, 2011.)

fixed at 0.5 or 0.6 V, the average number of pulses decreased as well when the pulse width was
increased, as can be seen from Figure 13.21. In addition, the resistance change is also a function
of the interval between applied electrical pulses. For a longer interval between pulses, more pulses
would be needed to switch the device. The dependence of minimum number of pulses for the STM
to occur on the voltage magnitude and pulse width is related to the heat accumulation process dur-
ing the rupture of the conductive filaments. Pulses with a larger voltage magnitude, a larger pulse
width, a larger duty cycle (i.e., a short interval in between pulses), or a combination of these are able
to accumulate a high density of heat energy in a localized region and, thus, easily switch the device
from the unmemorized state to the STM state through the melting of conductive filaments.
Instead of voltage pulses, repeated forward and backward voltage sweepings can also be used
to realize memorization transition from STM to LTM [26]. Figure 13.22a shows the dual-sweeping
I–V characteristics for 10 cycles. A quasi-linear I–V characteristic was observed for each sweep-
ing, suggesting that ohmic conduction might be the dominant current conduction process in the
device. Device resistance was obtained at +0.05 V in backward sweepings, and the evolution of the
resistance with measurement cycles is shown in Figure 13.22b. As shown in Figure 13.22a, the I–V
curves show a hysteresis in the forward and backward measurement, and the device current continu-
ously decreases with the sweeping cycles. In the first five cycles, very minimal hysteresis can be
observed, indicating that the device is maintained in the unmemorized state. The resistance in this
state is almost a constant of ~4 Ω as shown in Figure 13.22b. For the sixth cycle, a large hysteresis
occurs, and the resistance is suddenly increased to 25 Ω. The resistance increased again to ~52 Ω
for the seventh cycle, indicating that the device was entering the STM state. For the 9th and 10th
cycles, the two curves were almost consistent. No significant hysteresis can be observed and the
resistance was saturated at ~78 Ω. It is believed that the device was in the LTM state after the ninth
dual-sweeping measurement.
Figure 13.23 shows the statistical minimum number of cycles required for the STM to occur as
a function of the ending voltage of a sweeping cycle. The results were similar to the dependence of
the minimum number of the pulses required for the STM to occur on the pulse voltage described
earlier. As can be seen in Figure 13.23, the average minimum number of cycles decreases when the
ending voltage is increased. For example, six cycles of voltage sweeping were required to switch
the device into STM for an ending voltage of 0.4 V; however, only four cycles were needed for an
ending voltage of 0.5 V. This indicated that a larger portion of conductive filaments were ruptured
in a cycle with a higher ending voltage.
488 Semiconductor Nanocrystals and Metal Nanoparticles

0.10
1st

0.08

Current (A) 0.06

6th

0.04

0.02

10th
0.00
0.0 0.1 0.2 0.3 0.4 0.5
(a) Voltage (V)

Unmemorized STM LTM

100
Resistance (Ω)

10

1
0 2 4 6 8 10
(b) Number of cycles

FIGURE 13.22 (a) I–V characteristics of the forward and backward voltage sweepings and (b) evolution of
the resistance measured at +0.05 V in the backward sweeping with sweeping cycles. (From Liu, Y. et al., Appl.
Phys. A, 105, 855, 2011.)

Figure 13.24 shows a possible circuit application of the self-learning device. The input voltage
pulses applied on the Vin have a minimum value of 0 V and a maximum value of Vim. The out-
put pulses at the Vout terminal should have a voltage magnitude changing with time, as shown in
Figure 13.24. It should have a minimum voltage of 0 V and a maximum value of Vom, which can be
described by
-1
é R1 ù
Vom = Vim ê1 + ú , (13.22)
ë ( R2 + R D ) û

where R D is the resistance of the self-learning device, which changes with the number of pulses.
R1 and R2 can be carefully chosen according to the range of device resistance to make the circuit
sensitive. Since R D will gradually increase to a saturated value with the number of applied pulses,
the output voltage will also increase to a saturated value with the number of pulses. The illustration
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 489

Max.
12
75%
50%
10 25%

Number of cycles
Min.
8

0.4 V 0.45 V 0.5 V 0.6 V

Ending voltage

FIGURE 13.23 Minimum number of voltage sweeping cycles required for the STM to occur as a function of
the ending voltage of the sweeping cycle. (From Liu, Y. et al., Appl. Phys. A, 105, 855, 2011.)

Vin Vin
Vin

0V
R1
Vom
Vout
Vout
0V

Self-learning device STM LTM

R2
Time

FIGURE 13.24 A possible circuit implementation of the self-learning device. (From Liu, Y. et al., Appl.
Phys. A, 105, 855, 2011.)

of input and output signals as a function of time is illustrated in Figure 13.24, indicating that the
memorization from STM to LTM can be realized with such a circuit implementation based on the
self-learning device discussed earlier.
Neuroplasticity is used to describe brain changes in response to experience [30]. It is mainly
based on the creation of interconnections between neurons. Similarly, in the resistive switching
device, the internal structure changes through formation and rupture of conductive filaments as
a result of the application of a series of electrical pulses. The change in internal structure is dem-
onstrated by a sharp change or switching in device resistance. The resistance can be increased by
electrical pulses, as shown in Figure 13.20a, and a sharp switching from a low-resistance state to
490 Semiconductor Nanocrystals and Metal Nanoparticles

a high-resistance state can be achieved with pulses with suitable voltage and width. On the other
hand, sharp switching from a high-resistance state to a low-resistance state can be also achieved
with pulses with sufficiently large voltage and pulse width if a high-resistance state already exists.
Figure 13.25 shows the response of the resistance of a device in a high-resistance state to the appli-
cation of electrical pulses of 1.9 V, 0.1 s. The high resistance of ~104 Ω was maintained for five
pulses. However, the application of the sixth pulse switched the device to a low-resistance state of
19 Ω. The resistance remained unchanged for the application of subsequent pulses. This indicates
that the device could be switched from a high-resistance state to a low-resistance state by electrical
pulses without experiencing any intermediate states. It is believed that stable filaments were fully
formed in the device at the sixth pulse, which is similar to a neuroplasticity behavior, where neuro-
nal c­ onnections are created after certain times of experiences.
Figure 13.26 shows the statistical minimum numbers of pulses required to switch a device from
a high-resistant state to a low-resistance state at various pulse voltages with the pulse width fixed
at 0.1 s [26]. It was found that voltages below 1.6 V were not able to switch a device from the high-
resistance state to the low-resistance state. Twenty-five pulses at 1.7 V on average over five measured
devices were needed for the switching to occur, and the average number became smaller for pulses
with higher voltages.

13.4.2 Transient Memory Loss Behavior

In biological systems, “memory” is an essential building block for learning. Unlike modern semi-
conductor memories, human memory is by no means eternal. Yet, forgetfulness is not always a
disadvantage since it helps to remove redundant information and release memory storage for more
important or more frequently accessed information. Moreover, it is thought to be necessary for
individuals to adapt to new environments. Eventually, only memories that are of great significance
are transformed from STM into LTM through stimulation based on repeated experiencing. No mat-
ter what the exact mechanisms are, forgetting is an indisputable fact. Simply speaking, short-term
memory retention can be attributed to transient activation or transmission of certain chemicals,
which could be ions, transmitters, receptors, etc. In the human brain, memory immediately starts to
decay after an impression. For stimulations with a longer interval in between, memory decay will be
more significant and more impressions will be needed to achieve the STM and the LTM.

105

104
Resistance (Ω)

103

102

101
0 5 10 15 20
Number of pulses

FIGURE 13.25 Device resistance in response to electrical pulses resembling neuroplasticity in human brain.
The resistance was measured at +1.9 V. (From Liu, Y. et al., Appl. Phys. A, 105, 855, 2011.)
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 491

60

Max.

75%
40 50%

Number of pulses
25%
Min.

20

1.7 1.8 1.9 2.0

Pulse voltage (V)

FIGURE 13.26 Minimum numbers of pulses needed to switch a device from the high-resistance state to the
low-resistance state at various pulse voltages with the pulse width fixed at 0.1 s. (From Liu, Y. et al., Appl.
Phys. A, 105, 855, 2011.)

Hu et al. reported transient memory loss effect in resistive switching devices based on Ni-rich
NiO thin films [31]. Similar to the learning-behavior introduced in Section 13.4.1, the device based
on Ni-rich NiO thin film shows a gradual change in conductance under the application of consecu-
tive voltage pulses [31], resembling the STM to LTM transition in biological systems, as shown
in Figure 13.27. The difference is that the device exhibited a conductance increment rather than a
resistance increment (conductance reduction), shown in Figure 13.20a. As shown in Figure 13.27,

Unmemorized STM LTM

101

Pulse width = 0.1 s

G/G0

Pulse height = 2.8 V

Interval = 0.1 s
Read at 0.03 V

100

0 5 10 15 20
Number of pulses

FIGURE 13.27 Device conductance (G) measured at 0.03 V as a function of the number of the applied
electrical pulses (2.8 V in voltage magnitude, 0.1 s in width, and 0.1 s in interval between pulses). The initial
conductance (G 0) is 387.5 μS. (From Hu, S.G. et al., Appl. Phys. A, 109, 349, 2012.)
492 Semiconductor Nanocrystals and Metal Nanoparticles

STM has been achieved after four electrical pulses, which is similar to the situation that our brain
can quickly memorize a scene or a matter after several impressions. The memorization is con-
tinuously enhanced with the number of voltage pulses until a saturation state that is reached after
13 electrical pulses, and the memory state can be maintained for a long time. The conductance state
corresponding to the STM is an unstable intermediate state with a typical retention time of tens of
minutes; in contrast, the saturated conductance state corresponding to the LTM has a retention time
longer than a few hours. The formation of STM conductance state here is due to the formation of the
localized conductive filaments in the material system [27–29]. With more electrical pulses applied,
more and or stronger localized filaments are created, and thus the device conductance increases
[32]. Once stable and strong filaments are formed, the LTM state with a saturated conductance is
formed. On the other hand, trapping of electrons in the oxide film could also affect the conduction.
The defects in the oxide, including O vacancies [33] and Ni vacancies [34], can serve as the electron
traps that modulate current conduction during electrical stimulations.
Besides the learning ability that yields memorization, memory loss behavior (i.e., the forgetting
effect) is another important function of the human brain. The resistive switching device can also be
used to mimic memory loss function. Successive voltage pulses with a fixed height of 1.8 V and a
width of 0.005 s, but with different intervals (i.e., the off time between two consecutive pulses) of
0.005, 0.05, and 0.5 s, respectively, were applied to the device. The device conductance was mea-
sured at a reading voltage of 0.03 V after each pulse. Figure 13.28 shows the evolution of device
conductance, which is normalized to the initial conductance of G 0 as a function of the number of
voltage pulses for different intervals [31]. As can be seen in this figure, to achieve STM with the
conductance increased by one order, 13, 22, and 43 pulses are required for the intervals of 0.005,
0.05, and 0.5 s, respectively. It was believed that voltage-induced Joule heating plays an important
role in the process of CF formation in the resistive switching device [28,29,35,36]. Besides, heat dif-
fusion during the interval suppresses the formation of CFs. Therefore, the number of pulses required
to form the STM conductance state increases with the interval. This phenomenon is identical to the
memory-loss function in the human brain, where a piece of memory could be lost if impressions
were not continuously received.
Figure 13.29 shows the dependence of the minimum number of pulses that is required to
achieve the STM state on the pulse interval [31]. The occurrence of the STM is defined as when

12

10

Interval
G/G0

6 0.005 s
Pulse height = 1.8 V
Pulse width = 0.005 s 0.05 s
4 0.5 s

0
0 10 20 30 40 50
Number of pulses

FIGURE 13.28 Change in device conductance with number of pulses for different intervals. The pulse volt-
age and width are fixed at 1.8 V and 0.005 s, respectively. (From Hu, S.G. et al., Appl. Phys. A, 109, 349, 2012.)
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 493

Pulse height = 1.8 V

Pulse width = 0.005 s

Minimum number of pulses required

Max. Read at 0.03 V
Mean
102
75%
50%
25%
Min.

101

0.005 0.05 0.5 1

Interval (s)

FIGURE 13.29 Statistical minimum number of pulses required to achieve the STM conductance state for
different intervals between pulses. (From Hu, S.G. et al., Appl. Phys. A, 109, 349, 2012.)

the device conductance is increased by one order. The pulse height and width were fixed at
1.8 V and 0.005 s, respectively. More than 20 devices selected from different locations on the
wafer were measured to obtain the statistical average value. The longer the interval between
pulses, the larger the number of pulses required on average. For example, 12.2 pulses on aver-
age are needed to achieve the STM for an interval of 0.005 s; while 81.9 pulses on average are
required for an interval of 1 s. This characteristic in the resistive switching device can simulate
the human brain’s behavior—that it is difficult to form the STM with longer intervals between
events due to memory loss during pulse intervals, and thus more impressions are required to
form a solid memorization.
The self-learning ability of a device has been demonstrated with repeated forward and back-
ward voltage sweepings [25,26]. Similarly, the “forgetting” ability can be also realized with volt-
age sweepings. Figure 13.30 shows typical I–V characteristics of the device measured during the
repeated forward and backward voltage sweepings between 0 and 2.2 V. The interval between the
forward and backward sweeping cycles was set to 10 and 20 s, with results shown in Figure 13.30a
and b, respectively. The device current increases gradually after each sweeping cycle, showing
that the conductance increases with voltage sweeping. To achieve the STM state with conductance
increased by an order, 11 and 17 sweeping cycles are required for the sweeping intervals of 10 and 20 s,
respectively, as shown in Figure 13.30.
Figure 13.31 shows the statistical minimum number of sweeping cycles that is required to form
the STM state as a function of the interval. As can be seen clearly in this figure, more sweeping
cycles are needed for a longer interval. The results shown in Figure 13.30 and Figure 13.31 ­suggests
that the formation of conductive filaments are depressed during the interval period and thus more
sweeping cycles are needed to form sufficient conductive filaments for a longer interval. The depres-
sion of the conductive filament formation during intervals is analogous to the memory-loss function
in the human brain.

13.5 PRINTING AND FLEXIBLE THIN-FILM TRANSISTORS

Moore’s law has been the primary driving force for the advancement of ICs, which leads to the
number of transistors per microelectronic chip being doubled every 1½ years. However, the cost of
a chip per unit of area has remained relatively static for decades. The prevalent Si-based technology
494 Semiconductor Nanocrystals and Metal Nanoparticles

1.0 × 10–4
11th

Current (A) Interval = 10 s

5.0 × 10–5

1st

0.0

0.0 0.5 1.0 1.5 2.0 2.5

(a) Voltage (V)

1.5 × 10–4

17th
Interval = 20 s

1.0 × 10–4
Current (A)

5.0 × 10–5

1st
0.0

0.0 0.5 1.0 1.5 2.0 2.5

(b) Voltage (V)

FIGURE 13.30 Measured I–V characteristics of the device in the repeated forward and backward voltage
sweepings, with the interval between two consecutive voltage sweeping being (a) 10 s and (b) 20 s, respec-
tively. (From Hu, S.G. et al., Appl. Phys. A, 109, 349, 2012.)

is not always a cost-effective solution for certain IC products, such as the popular wearable devices
nowadays. Hence, there is interest in developing solution-based techniques (e.g., printing and spin-­
coating) for microelectronics fabrication that are inexpensive, allow fabrication on flexible s­ ubstrates
(e.g., plastics and fiber papers), and have large-area synthesizing ability.
The primary focus has been on organic materials for solution-based printing [37,38]. Organic
semiconductors such as poly(3-hexylthiophene) based on solution-based synthesis approaches have
demonstrated a field effect mobility of ~0.1 cm2 V−1 s−1 [38]. The low mobilities of the solution-
processed organic thin-film transistors (TFTs) make them inappropriate for applications in micro-
processors, display drivers, and active matrix backplanes for high-resolution displays. Inorganic
semiconductors are the basis for almost all high-performance microelectronic devices. They can
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 495

102
Max.
Mean

Minimum number of cycles required

75%
50%
25%
Min.

101

0 10 20
Interval (s)

FIGURE 13.31 Statistical minimum number of the sweeping cycles required to achieve the STM conduc-
tance state for different intervals. (From Hu, S.G. et al., Appl. Phys. A, 109, 349, 2012.)

have an intrinsic mobility as high as 1000 cm2 V−1 s−1 and a lifetime more than 50 years. Inorganic
microelectronic devices and circuits are generally fabricated with conventional Si-related tech-
niques such as thermal oxidation, evaporation, sputtering, chemical vapor deposition, ion implanta-
tion, photolithography, and so on. Unfortunately, the solution-based approach is not a convenient
way to synthesize inorganic semiconductors since they are not intrinsically soluble in any conve-
nient solvents and do not offer structural variability as a mechanism to alter their solubility [39].

13.5.1 Printable CdSe Nanocrystal for Inorganic TFTs

Ridley et al. reported a pioneering work on the method of printing inorganic thin films by using
stable CdSe nanocrystal solutions as precursors to form polycrystalline thin films [39]. Unlike its
bulk crystalline counterpart, the population of surface atoms in a nanocrystal is of nonnegligible
percentage. Therefore, nanocrystals have properties somewhere between those of a bulk solid and
an atomic species, exhibiting size-dependent physical, electrical, and optical properties [40].
A typical synthesis was carried out in a nitrogen-filled glove box under room temperature.
Methanolic solutions of the reagents CdI2 and Na2Se (at 7.4 mM concentrations) were prepared.
Equimolar amounts of the reagents were then added to the reaction vessel, which was charged
with pyridine. Pyridine generally constituted one-third of the total reaction volume. The reaction
proceeds instantaneously, forming a yellow solution that quickly precipitates the nanocrystalline
­product. The nanocrystals are then isolated by centrifugation and decanting of the supernatant, and
the soluble by-product, NaI, is removed by repeatedly dispersing the product in methanol and isolat-
ing the particles after centrifugation and decanting. After dissolution of the nanocrystals in pyri-
dine, they are precipitated by addition of hexane and again isolated by centrifugation and decanting.
Finally, the nanocrystals are redissolved in pyridine and filtered through a 0.2 mm filter.
After synthesis of CdSe nanocrystals, the fabrication of TFT devices were carried out in the
glove box as well [39]. A single drop of the pyridine CdSe nanocluster solution was deposited onto
the area between the source and drain electrodes (Cr/Au) of a TFT structure. The CdSe nanocrys-
tals were heated by placing the wafer on a hot plate. Various heat treatments ranging from 150°C
to 350°C were used to sinter the nanocrystals in the channel region. Upon cooling, Norland Optical
496 Semiconductor Nanocrystals and Metal Nanoparticles

Adhesive 73, a photocurable polymer adhesive, was deposited and cured to encapsulate the CdSe
channel region of the TFT. The fabricated device was a coplanar inverted TFT consisting of n+-Si
gate, 100 nm thermal SiO2, Cr/Au (10 nm/100 nm) source and drain contacts, and a channel layer
with a length of 8 μm and width of 293 μm [39]. The TFTs were measured in air in a dark box at
room temperature. It was found that a field effect was not observed from unencapsulated devices
and those devices heated to 150°C. However, TFTs processed at 250°C were found to exhibit a field
effect, with the mobilities proportional to the treatment temperature.
Figure 13.32a shows the electrical characteristic of the TFT ramped from room temperature at
80°C/min and held at 350°C for an hour [39]. With a gate sweep of VGS = −40 to +40 V at VDS = 2.5 V,
the ID –VGS curve shows an ON/OFF ratio of 3.1 × 104, a linear regime mobility of 1 cm2 V−1 s−1

10–3

10–4

10–5
ID (A)

10–6

10–7

10–8
–40 –20 0 20 40
(a) VGS (V)

0.5

0.4
VGS = 40 V

0.3
ID (mA)

VGS = 20 V
0.2

0.1 VGS = 0 V

VGS = –20 V
0
0 2 4 6 8 10
(b) VDS (V)

FIGURE 13.32 Electrical characteristics of a TFT based on printed CdSe nanocrystals. (a) Logarithmic
ID –VGS curves for a gate sweep of VGS = –40 to +40 V at VDS = 2.5 V, and (b) ID –VDS curves with a gate sweep-
ing from VGS = −40 to +40 V in steps of 20 V. (From Ridley, B.A. et al., Science, 286, 746, 1999.)
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 497

(extracted by equating the slope of an ID –VGS plot to (W/L)μCoxV DS at V DS = 2.5 V, where Cox is the
capacitance of the gate oxide), a threshold voltage of 6.7 V (extracted from the VGS intercept of an
ID –VGS plot), and a subthreshold slope of ~7 to 10 V/dec. The hysteresis observed in Figure 13.32a
may be due to interface states at the semiconductor–insulator interface or simply due to poor
encapsulation and oxidation of the TFT. The negative resistance seen in the saturation region of
the ID −V DS characteristics shown in Figure 13.32b is also likely to be caused by hysteresis in the
semiconductor film. The mobility of these TFTs is approximately an order of magnitude larger
than the mobility of printed organic TFTs. The field-effect mobility can be increased further by
improvements in nanocrystal solution purity and processing techniques, which is very promising
for practical circuit applications. It is also worth mentioning that the processing temperatures
for this device are compatible with high glass transition temperature plastic substrates, such as
polyimide.

13.5.2 Spin-Coated CdSe Nanocrystals for Flexible TFTs

Using the same CdSe nanocrystals (NCs) but with different synthesizing approach, Kim et al.
reported flexible, low-voltage, and high-performance TFT devices, and also demonstrated fully
functional integrated circuits based on such devices [41]. In order to prepare the solution for spin
coating, highly monodispersed CdSe NCs with compact thiocyanate ligands were synthesized in
a nitrogen glove box [42,43]. NCs with thiocyanate ligands were then redispersed in dimethylfor-
mamide and spincast on the top of flexible substrates to form uniform, crack-free, and randomly
close-packed NC thin films as semiconducting channels in TFTs. Unlike organic semiconductors,
the NC thin films have a morphology and mobility largely insensitive to surface roughness except
with a root-mean-squared value as large as a few nanometers [44]. To fabricate flexible devices,
either a 25 or 50 mm thick polyimide, covered with 30 nm of atomic layer deposited (ALD) Al2O3 at
250°C, was used as the substrate. The Al2O3 encapsulated the polyimide substrate as a buffer layer,
which can eliminate severe delamination and cracking of the deposited NC thin films and metal
electrodes during subsequent thermal processing. A 20 nm thick Al back gate was deposited onto
the buffer layer by thermal evaporation through a shadow mask. The device was exposed to oxygen
plasma to form some native Al2O3 on the Al gate and to create additional hydroxyl groups. A 30 nm
ALD Al2O3 was grown as the gate dielectric layer. The measured capacitance of the Al2O3 dielectric
layer was 0.253 ± 0.019 mF/cm2, allowing for low-voltage operation. In addition, the gate leakage
was characterized at pA levels, indicating the combination of CdSe NC thin film channel and the
ALD Al2O3 dielectric is suitable for flexible electronic applications [41]. Inside a nitrogen glove box,
In/Au (50 nm/40 nm) electrodes were thermally deposited through a shadow mask onto the NC
thin film to complete back-gate/top-contact TFTs. Figure 13.33a shows the schematic of the whole
device structure, and Figure 13.33b demonstrates the real circuits fabricated on a Kapton substrate
based on the CdSe NC TFTs.
Figure 13.33c shows typical output characteristics (ID −VDS) [41]. An n-type device behavior that
is modulated by a small positive voltage as low as 2 V can be observed from this figure. The
extracted electron field-effect mobilities from the transfer curves (ID –VG) shown in Figure 13.33d
are 21.9 ± 4.3 cm2 V−1 s−1 in the linear regime (VDS = 0.1 V) and are 18.4 ± 3.6 cm2 V−1 s−1 in the
saturation regime (VDS = 2 V). These CdSe NC TFTs built on plastic substrates operate well at low
voltages. Good device performances, including high ION/IOFF ratio (>106), low subthreshold swing
(S = 0.28 ± 0.09 V/dec), low threshold voltage (V T = 0.38 ± 0.15 V), and low hysteresis (ΔV T = 0.25 ±
0.07 V) at VDS = 2 V were achieved in these devices. The low hysteresis was attributed to passivation
of the NC surface by indium and the selection of ALD Al2O3 as the gate dielectric, which helped to
reduce the density of trap states at the NC surface and at the semiconductor–gate dielectric inter-
face [43]. In addition, the small variation in device parameters and large-area uniformity of these
NC TFTs enables their integration in flexible TFT–based integrated circuits.
498 Semiconductor Nanocrystals and Metal Nanoparticles

Drain

Source
nm
100

e
CdS
Al 2O te
3

l G a
A
Al 2O n
3

a p to
K

(a) (b)

50 10–2
VG = 2 V 10–3
40 10–4
VDS = 2.0 V
10–5
1.6 V
30 10–6
ID (µA)

ID (A)

10–7
1.2 V VDS = 0.1 V
20 10–8
10–9
10 0.8 V 10–10
0–0.4 V 10–11
0 10–12
0.0 0.5 1.0 1.5 2.0 –2 –1 0 1 2
(c) VDS (V) (d) VG (V)

FIGURE 13.33 Flexible TFTs based on CdSe nanocrystals. (a) Schematic and (b) photograph of a flexible
TFT based on CdSe NCs atop a Kapton substrate. (c) Output ID –VDS and (d) transfer ID –VG characteristics.
(From Kim, D.K. et al., Nat. Commun., 3, 1216, 2012.)

REFERENCES
1. G. Baccarani, M. R. Wordeman, and R. H. Dennard, Generalized scaling theory and its application to a
1/4 micrometer MOSFET design, IEEE Transactions on Electron Devices, 31, 452–462, 1984.
2. R. H. Dennard, V. L. Rideout, E. Bassous, and A. R. LeBlanc, Design of ion-implanted MOSFET’s with
very small physical dimensions, IEEE Journal of Solid-State Circuits, 9, 256–268, 1974.
3. K. K. Likharev, Single-electron transistors: Electrostatic analogs of the DC SQUIDS, IEEE Transactions
on Magnetics, 23, 1142–1145, 1987.
4. H. Grabert and M. H. Devoret, Single Charge Tunneling: Coulomb Blockade Phenomena in
Nanostructures. New York: Plenum Press, 1992.
5. K. K. Likharev, Single-electron devices and their applications, Proceedings of the IEEE, 87, 606–632,
1999.
6. D. V. Averin and K. K. Likharev, Single electronics: A correlated transfer of single electrons and cooper
pairs in systems of small tunnel junctions, in: Mesoscopic Phenomena in Solids, B. L. Altshuler, P. A.
Lee, and W. Richard Webb, eds. Amsterdam, the Netherlands: Elsevier, 1991.
7. K. Uchida, Single-electron devices for logic applications, in: Nanoelectronics and Information
Technology—Advanced Electronic Materials and Novel Devices, R. Waser, ed., 3rd ed. John Wiley &
Sons, Weinheim, Germany, 2012.
8. K. C. Chan, P. F. Lee, and J. Y. Dai, Single electron tunneling and Coulomb blockade effect in HfAlO∕Au
nanocrystals/HfAlO trilayer nonvolatile memory structure, Applied Physics Letters, 92, 143117, 2008.
Novel Nanoelectronic Device Applications of Nanocrystals and Nanoparticles 499

9. C. Wasshuber, H. Kosina, and S. Selberherr, SIMON-A simulator for single-electron tunnel devices
and circuits, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 16,
937–944, 1997.
10. M. Kirihara, K. Nakazato, and M. Wagner, Hybrid circuit simulator including a model for single electron
tunneling devices, Japanese Journal of Applied Physics, 38, 2028, 1999.
11. T. A. Fulton and G. J. Dolan, Observation of single-electron charging effects in small tunnel junctions,
Physical Review Letters, 59, 109–112, 1987.
12. U. Meirav, M. A. Kastner, and S. J. Wind, Single-electron charging and periodic conductance resonances
in GaAs nanostructures, Physical Review Letters, 65, 771–774, 1990.
13. Y. Takahashi, Silicon single-electron devices for logic applications, in: Proceeding of the 32nd European
Solid-State Device Research Conference, Firenze, Italy, 2002, pp. 61–70.
14. Y. Ono, Y. Takahashi, K. Yamazaki, M. Nagase, H. Namatsu, K. Kurihara et al., Fabrication method for
IC-oriented Si single-electron transistors, IEEE Transactions on Electron Devices, 47, 147–153, 2000.
15. D. L. Klein, R. Roth, A. K. L. Lim, A. P. Alivisatos, and P. L. McEuen, A single-electron transistor made
from a cadmium selenide nanocrystal, Nature, 389, 699–701, 1997.
16. L. P. Kouwenhoven and P. L. McEuen, Single electron tunneling through a quantum dot, in: Nanoscience and
Technology, G. Timp, ed. New York: AIP Press.
17. L. O. Chua, Memristor—The missing circuit element, IEEE Transactions on Circuit Theory, 18,
507–519, 1971.
18. D. B. Strukov, G. S. Snider, D. R. Stewart, and R. S. Williams, The missing memristor found, Nature,
453, 80–83, 2008.
19. T. H. Kim, E. Y. Jang, N. J. Lee, D. J. Choi, K.-J. Lee, J.-T. Jang et al., Nanoparticle assemblies as mem-
ristors, Nano Letters, 9, 2229–2233, 2009.
20. S. H. Jo, T. Chang, I. Ebong, B. B. Bhadviya, P. Mazumder, and W. Lu, Nanoscale memristor device as
synapse in neuromorphic systems, Nano Letters, 10, 1297–1301, 2010.
21. L. O. Chua and K. Sung Mo, Memristive devices and systems, Proceedings of the IEEE, 64, 209–223, 1976.
22. R. Ananthanarayanan, S. K. Esser, H. D. Simon, and D. S. Modha, The cat is out of the bag: Cortical
simulations with 109 neurons, 1013 synapses, in: Proceedings of the Conference on High Performance
Computing Networking, Storage and Analysis, Portland, OR, 2009, pp. 1–12.
23. V. Lev-Ram, S. T. Wong, D. R. Storm, and R. Y. Tsien, A new from of cerebellar long-term potentiation
is postsynaptic and depends on nitric oxide but not cAMP, Proceedings of the National Academy of
Sciences of the United States of America, 99, 8389, 2002.
24. Y. Yang, P. Gao, L. Li, X. Pan, S. Tappertzhofen, S. Choi et al., Electrochemical dynamics of nanoscale
metallic inclusions in dielectrics, Nature Communications, 5, 4232, 2014.
25. T. Hasegawa, T. Ohno, K. Terabe, T. Tsuruoka, T. Nakayama, J. K. Gimzewski et al., Learning abilities
achieved by a single solid-state atomic switch, Advanced Materials, 22, 1831–1834, 2010.
26. Y. Liu, T. P. Chen, Z. Liu, Y. F. Yu, Q. Yu, P. Li et al., Self-learning ability realized with a resistive
switching device based on a Ni-rich nickel oxide thin film, Applied Physics A, 105, 855–860, 2011.
27. D. C. Kim, S. Seo, S. E. Ahn, D.-S. Suh, M. J. Lee, B.-H. Park et al., Electrical observations of filamen-
tary conductions for the resistive memory switching in NiO films, Applied Physics Letters, 88, 202102,
2006.
28. A. Sawa, Resistive switching in transition metal oxides, Materials Today, 11, 28–36, 2008.
29. R. Waser and M. Aono, Nanoionics-based resistive switching memories, Nature Materials, 6, 833–840,
2007.
30. R. J. Davidson and A. Lutz, Buddha’s brain: Neuroplasticity and meditation [in the spotlight], IEEE
Signal Processing Magazine, 25, 176–174, 2008.
31. S. G. Hu, Y. Liu, T. P. Chen, Z. Liu, Q. Yu, L. J. Deng et al., Realization of transient memory-loss with
NiO-based resistive switching device, Applied Physics A, 109, 349–352, 2012.
32. M.-J. Lee, S. Han, S. H. Jeon, B. H. Park, B. S. Kang, S.-E. Ahn et al., Electrical manipulation of nano-
filaments in transition-metal oxides for resistance-based memory, Nano Letters, 9, 1476–1481, 2009.
33. M. C. Ni, S. M. Guo, H. F. Tian, Y. G. Zhao, and J. Q. Li, Resistive switching effect in SrTiO3−​
δ/Nb-doped SrTiO3 heterojunction, Applied Physics Letters, 91, 183502, 2007.
34. W.-L. Jang, Y.-M. Lu, W.-S. Hwang, T.-L. Hsiung, and H. P. Wang, Point defects in sputtered NiO films,
Applied Physics Letters, 94, 062103, 2009.
35. S. Seo, M. J. Lee, D. H. Seo, E. J. Jeoung, D.-S. Suh, Y. S. Joung et al., Reproducible resistance switching
in polycrystalline NiO films, Applied Physics Letters, 85, 5655–5657, 2004.
500 Semiconductor Nanocrystals and Metal Nanoparticles

36. K. Tsunoda, K. Kinoshita, H. Noshiro, Y. Yamazaki, T. Iizuka, Y. Ito et al., Low power and high speed
switching of Ti-doped NiO ReRAM under the unipolar voltage source of less than 3 V, in: 2007 IEEE
International Electron Devices Meeting—IEDM’07, Piscataway, NJ, December 10–12, 2007, pp. 767–770.
37. Z. Bao, A. Dodabalapur, and A. J. Lovinger, Soluble and processable regioregular poly(3‐­hexylthiophene)
for thin film field-effect transistor applications with high mobility, Applied Physics Letters, 69, 4108–
4110, 1996.
38. H. Sirringhaus, N. Tessler, and R. H. Friend, Integrated optoelectronic devices based on conjugated
polymers, Science, 280, 1741–1744, 1998.
39. B. A. Ridley, B. Nivi, and J. M. Jacobson, All-inorganic field effect transistors fabricated by printing,
Science, 286, 746–749, 1999.
40. A. P. Alivisatos, Electrical studies of semiconductor-nanocrystal colloids, MRS Bulletin, 23, 18–23,
1998.
41. D. K. Kim, Y. Lai, B. T. Diroll, C. B. Murray, and C. R. Kagan, Flexible and low-voltage integrated
circuits constructed from high-performance nanocrystal transistors, Nature Communications, 3, 1216,
2012.
42. A. T. Fafarman, W.-K. Koh, B. T. Diroll, D. K. Kim, D.-K. Ko, S. J. Oh et al., Thiocyanate-capped
nanocrystal colloids: Vibrational reporter of surface chemistry and solution-based route to enhanced
coupling in nanocrystal solids, Journal of the American Chemical Society, 133, 15753–15761, 2011.
43. J.-H. Choi, A. T. Fafarman, S. J. Oh, D.-K. Ko, D. K. Kim, B. T. Diroll et al., Bandlike transport in
strongly coupled and doped quantum dot solids: A route to high-performance thin-film electronics,
Nano Letters, 12, 2631–2638, 2012.
44. T. Sekitani, U. Zschieschang, H. Klauk, and T. Someya, Flexible organic transistors and circuits with
extreme bending stability, Nature Materials, 9, 1015–1022, 2010.
Index
A electronic structure, 101–102
gain, 104–106
Ag target ablation in water line-width enhancement factor, 106–107
with different angles of incidence macroscopic and microscopic polarization, 105
coupling of incident photons, 387 phenomenological gain compression factor, 106
EFs, 385–387
FESEM and AFM imaging techniques, 381–382,
384–385 C
Raman spectra of ANTA, 386, 388
Raman spectra of R6G, 384, 386–387 CdS-sensitized ZnO nanorods
Raman spectra of R6G/ANTA, 383–384 CdS nanoparticle-decorated ZnO nanorods and P3HT
with different fluences and scanning conditions polymer blending, 63
ablation mechanism, 376–377 current density vs. voltage characteristics, 64
EDX spectra, 378 energy band diagram, 66
FESEM images, 378–379 EQE, 65–66
HRTEM images, 378 FESEM image, 64
micro-Raman spectra of adsorbed RDX molecules, PEDOT:PSS layer, 64
381, 383 Compound semiconductor nanocrystals
Raman spectra of R6G molecules, 379–381 CdSe nanocrystals
SAED patterns, 378 hydrothermal methods, 3
surface morphologies, 381–382 MPA-capped CdSe nanocrystals synthesis, 3
with ps/fs Bessel beams TEM, 3–4
FESEM images, 389–391 XRD, 3–5
Gaussian beam focusing experiments, 389 CdS nanocrystals
laser parameters, 388 aqueous-phase synthesis, 4–5
LSPR peak position, 389 colloid synthesis technique, 5
nonlinear propagation, 389 microemulsion by ultrasonic irradiation, 5
Alloyed semiconducting QDs, 296–297 TEM, 5–6
Amplified spontaneous emission (ASE), 115 wet chemical precipitation method, 5
Anodized alumina films (AAFs), 347 XRD, 6–7
Antibacterial activity, Cu, Ag colloids, 409–410 chemical methods, 2
Au-functionalized ZnO sensor CuCl nanocrystals
chemical sensor, 79 Bridgman method, 12
depletion layer, 81 chemical methods, 12
dynamic response and recovery, 79–81 Czochralski method, 12
sensing mechanism, 81 hydrothermal methods, 12
sensitivity, 78 SEM characterization, 12–13
sensor measurements, 78 XRD, 14
Auger spectroscopy GaAs nanocrystals
advantages, 158 AFM, 8, 10
Ag M4N45N45 Auger spectra and photoelectron spectra, electrochemical techniques, 7
159–160 laser ablation, 7–8
Auger chemical shift, 158 TEM, 8–9
binding energies, 159–160 wet process, 8
deconvoluted photoelectron spectra, 159, XRD, 8–9
161–162 III–V compound, 7
modified Auger parameter, 159 InAs nanocrystals
photoionization process, 158 AFM, 11–12
thermal annealing, 162 chemical synthesis, 9
XPS measurements, 158–159 deposition, 9–10
epitaxial method, 10
B SEM, 10–11
surface passivation, 9
Bulk-structure lasers, 96 TEM, 10–11
Buried InAs/InP quantum dots XRD, 11, 13
carrier-induced refractive index, 104–105 potential applications, 2
carrier scattering processes, 103–104 TOP/TOPO/HPA method, 2–3
complex optical susceptibility, 105 TOP/TOPO method, 2

501
502 Index

Core–shell nanoparticles, 359–360 Electroluminescence (EL)

CuO/ZnO nanorod heterostructures optical transitions, 255
depletion layer, 78 from Si+-implanted Si3N4 thin films
energy band diagram, 77–78 annealing effect, 273–276
operating temperature, 77 electrically tunable EL, 277–279
photochemical process, 77 properties and mechanisms, 270–273
response and recovery characteristics, 77 structure fabrication, 269–270
sensing performance, 78–79 from Si+-implanted SiO2 thin films, 264–266
Cu target ablation silicon nanocrystals, 203–204
in chloroform, 397, 399–400 Electron–electron scattering, 368
with corroles and surface-enhanced fluorescence/ Electron–lattice coupling, 368
SERS studies Electrooptic (EO) modulators, 99
fluorescence enhancements, 402 Electrophoretic deposition (EPD), 301
liquid-filled Pyrex cell, 400 Elemental semiconductor nanocrystals
surface-enhanced fluorescence (SEF) spectra, Ge nanocrystals
402–403 GeCl4 decomposition, 16–17
TEM images, 400–402 MBE, 16
with different energies and scanning conditions supercritical fluid method, 16
EFs, 397 TEM, 16–18
FESEM images, 390–394 XRD, 18
periodicity of ripples, 392 Si nanocrystals
ps laser ablation, 389 AFM, 15–16
Raman signal enhancement, 397 chemical etching, 14
Raman spectra, 390, 392 dry etching, 14–15
Raman spectra of ANTA, 393–395 electrochemical etching, 14
Raman spectra of R6G, 395–397 SEM, 15
Raman spectra of TNT, 394, 397–398 TEM, 15
XRD, 16–17
D Enhancement factors (EFs)
Ag target ablation with different angles of incidence,
Difference-frequency generation (DFG), 328, 385–387
334–337 Cu target ablation with different energies and scanning
complex amplitudes, 329 conditions, 397
electric susceptibility, 330 Excitation transition, 255
energy-level description, 313 External quantum efficiency (EQE), 65–66
geometry, 313
nonzero frequencies, 329 F
optical parametric amplification, 330
SONS, 330 Flow-cell synthesis, 346
Direct bandgap semiconductors, 255–256 Forouhi–Bloomer (FB) model, 239–240
Directly modulated (DM) laser, 99 Free-running quantum dot lasers
Drude–Lorentz–Sommerfeld (DLS) model, amplitude modulation response
132–133 carrier dynamics model, 107–108
Drude’s free-electron model, 155 carrier occupation probabilities, 108
Dye-sensitized solar cells (DSSCs), 63 damping factor, 110
cadmium selenium telluride, 296–297 discrete states, 107–108
MO semiconductor, 283–284 linearized differential rate equation, 109
Nb2O5, 291–292 modulation transfer function, 109
SnO2, 290–291 resonance frequency, 110
SrTiO3, 293 semiclassical rate equation model, 107
TiO2, 288–290 sinusoidal current modulation, 109
ZnO, 291 carrier capture and relaxation process impacts
structure and mechanism, 284–286 carrier capture time, 116, 118–120
carrier relaxation time, 116–117
E carrier scattering times, 116, 118
FM/AM index ratio, 119
Effective medium approximation (EMA), 246–247 modulation bandwidth, 117
Electrically tunable electroluminescence, 277–279 line-width enhancement factor
Electric vehicles (EVs), 356–357 below-threshold α-factor, 115
Electroabsorption (EA) modulators, see Electrooptic carrier-induced frequency shift, 111
modulators complex gain, 112
Electrodeposition, metallic nanopastes, 347–348 differential gain, 112
Index 503

differential rate equations, 114 dielectric function, 248–249

electric field, 111, 113 spectral fittings, 246–248
FM/AM index ratio, 115 synthesis procedure, 245
free carrier plasma effect, 111 Tauc plot, 248, 250
frequency shifts, 113 TEM image, 245, 247
material gain, 112 volume fraction, 247
nonzero α-factor, 111 TDLDA, 234
normalized bias current, 116, 118 wavelength, 231
Qdot material and laser parameters, 115–116 Green synthesis (GS), 359
small-signal carrier density vs. modulation
frequency, 116–117 H
Hakki-Paoli method, 115
G High-indexed (311)B InP substrates, 97–98
High-power light-emitting diodes (HPLEDs), 359
GaAs-based Qdot laser, 96 High-power semiconductor laser (HPSL), 359
Gas sensor with Pd NPs surface modulation Hybrid energy systems (HESs), 357
ethanol, acetone, and water vapor, 452 Hydrothermal/solvothermal (HT/ST) synthesis,
gas concentration 346–347
current–voltage characteristics,
445–446, 450 I
physical parameters, 446
response time and hydrogen concentration, Indirect bandgap semiconductors, 255–256
448–449 InGaAs/GaAs quantum dots, 96
sensitivity and time response, 447 InP-based quantum dot lasers
sensitivity response, 447–448 (311)B InP substrate, 97–98
transient response, 446–447 dynamic performance, 99–101
H2O molecules, 452–453 (100)InP substrate, 98–99
LAO thicknesses, 450–451 Ion-beam sputtering
mechanism, 453–456 electron microscopy results, 179–181
oxidizing (reducing) gases, 450, 452 on gold and silver metal films, 177–178
oxygen concentration in hydrogen gas, 449, 451 linear collision cascade, 176–177
Pd NPs-coated LAO/STO heterostructure, momentum transfer, 176
444, 446 morphological differences, 176
SAED pattern, 444, 446 nanostructuring, 176
Generalized Penn model (GPM), 228 nonequilibrium processing, 176
Germanium nanocrystals (nc-Ge) optical absorption results, 181–183
ab initio absorption spectra, 236–237 radiation-induced release of interstitials, 177
ab initio approaches, 216 RBS, 178
average oscillator strengths, 234–235 surface diffusion, 176
density of states (DOS), 232–233 Ion-exchange processing, Ag nanoparticles
DFT, 231 Drude’s free-electron model, 155
electron–hole pair, 237 exciton and surface-plasmon interactions, 173–175
electronic structures, 232 glass composition, 154
elemental semiconductor nanocrystals linear and nonlinear optical properties of glasses, 152
GeCl4 decomposition, 16–17 low-frequency Raman scattering spectra, 166–169
MBE, 16 modified Auger parameters, 158–162
supercritical fluid method, 16 optical absorption spectra, 154–155
TEM, 16–18 optical absorption spectroscopy, 162–167
XRD, 18 PL spectra, 169–173
embedded in SiO2 matrices, 249–251 processing conditions, 154
energy levels and optical transitions, 233–234 RBS measurements, 155–157
exciton binding energy, 236 XPS, 157–158
frequency-dependent dielectric functions, 236, 238 Ion implantation
HOMO–LUMO transition, 233–234 Ag nanoparticles
lowest electron–hole pair excitation energies, activation energy, 151
234–236 growth behavior, 150
optical gap, 237–238 optical absorption spectra, 149
quantum confinement effects, 231 in plain soda glass matrix, 151
self-assembled nc-Ge polarized low-frequency Raman spectra, 149–150
AFM images, 245, 247 recoil-implantation, 149
Bruggeman EMA, 246–247 SEM image, 151–152
504 Index

Au nanoparticles, 152–153 Heisenberg uncertainty principle, 197

coarsening stages, 148 MEG, 202
particle nucleation and growth, 147–148 optical gain, 202
supersaturation of metal atoms, 147–148 and photoconductivity, 192
porous silicon, 196
J quantum confinement of excitons, 196–197
room-temperature luminescence, 196–197
Joule heating, 353–354 sensitizers, 203
surface termination and passivation, 197
L Liquid-phase sintering (LPS), 349–351
Localized surface plasmon resonances (LSPRs), 58
LaAlO3/SrTiO3, see Lanthanum aluminate and strontium Lorentz oscillator model, 239–240
titanate heterostructure Low-frequency Raman scattering (LFRS)
Lanthanum aluminate and strontium titanate (LAO/STO) measurements, 140
heterostructure cobalt nanoparticles
electrical property characterizations, 426 cross-polarized Raman measurements, 143
fabrication process, 424–425 low-frequency vibrational mode intensity, 143–144
Hall effect measurement, 428 optical absorption spectra, 143, 145
highly sensitive gas sensor with Pd NPs surface polarization dependence, 143, 145
modulation quadrupolar modes, 144, 146
ethanol, acetone, and water vapor, 452 Raman peak frequency, 144
gas concentration, 445–450 Raman scattering intensity enhancement, 143, 146
H2O molecules, 452–453 Raman spectra, 146
LAO thicknesses, 450–451 silver nanoparticles, 142–143
mechanism, 453–456 Low-pressure chemical vapor deposition (LPCVD),
oxidizing (reducing) gases, 450, 452 195, 243
oxygen concentration in hydrogen gas, 449, 451
Pd NPs-coated LAO/STO heterostructure, 444, 446 M
SAED pattern, 444, 446
high-resolution TEM image, 425, 427 Memory resistor, see Memristors
insulating properties, 428 Memristors, 469
metallic properties, 428 amorphous Si embedded with Ag nanoparticles
palladium nanoparticle enhanced giant current evolution, 475–476
photoconductivity device structure, 472–473
current voltage characteristics under different electroforming, 473
ambient, 440, 442 flux-controlled memristor model, 475
mechanism, 441–445 I–V characteristics, 473–475
Pd/LAO film growth, 433–434 electrochemical dynamics, metallic nanoparticles in
photoresponse under UV light and room light, dielectrics, 478–483
436–438 electronic synapses realized with Si-based memristors
surface morphology, 434–436 biological synapse, 475, 477
time-dependent resistance, 439–441 Blue Gene/P, 476
wavelength-dependent photoconductivity, 437–439 device endurance, 477–479
polar liquid molecule-induced transport property STDP, 477–478
modulation von Neumann architecture, 475
intrinsic polar catastrophe model, 428 Fe3O4 nanoparticle assemblies, 471–472
mechanism, 430–433 four fundamental circuit elements, 470–471
polar liquid sensing, 428–429 memristive systems, 470
Schottky-diode-like junction, 430 TiOx -based device, 471
RHEED measurement, 424–426 Metal chalcogenide semiconductors, 284
sheet resistance cadmium chalcogenide quantum dots, 295–296
vs. mobility, 426–427 cadmium selenium telluride, 296–297
vs. temperature, 426 lead chalcogenide quantum dots, 293–295
2DEG interface, 424 zinc chalcogenide quantum dots, 296
XRD pattern, 425–426 Metallic nanopastes
Laser ablation in liquid (LAL) advancements
with ns pulses, 368 core–shell NPs, 359–360
pulsed LAL, 372 green synthesis, 359
review articles, 368–370 nanoalloys and nanoscale metallic compounds,
Laser sintering, 354–355 360–361
Layer-by-layer (LbL) technique, 300 metal nanomaterial synthesis
Light emission, silicon nanocrystals bottom-up approach, 345
embedded in silicon dioxide and silicon nitride, 198–200 chemical reduction methods, 345–347
energy state quantization, 197 comparison of synthesis methods, 348–349
Index 505

electrodeposition, 347–348 optical response

laser ablation, 348 dipolar surface plasmon excitation, 131
top-down approach, 345 DLS model, 132–133
power electronic applications, 356–357 interband transitions and core effects, 136
electric vehicles (EVs), 356–357 Mie theory, 131–132
military, aerospace, and high-power semiconductor quantum surface plasmon resonances, 137–139
device packaging, 359 quasi-static responses, 133–136
renewable energy (RE), 357–358 size-dependent dielectric functions, 137
smart grid and microgrids, 357–358 PL properties, 129
sintering propagation length, 128
brazing, 349 quantum-size behavior, 128
comparison of sintering methods, 355–356 SPP modes, 128–129
effect of pressure, 351–352 surface-plasmon-enhanced Raman scattering, 129
joule heating, 353–354 surface-plasmon resonance absorption, 18
laser sintering, 354–355 synthesis methods, 146–147
liquid-phase sintering, 349–350 top-down methods, 146
mass transport mechanisms, 350–351 alternative fabrication approaches, 174
particle size, 352–353 ion-beam sputtering, 176–183
photonic sintering, 354 submicrometric and nanometric features, 176
shape effect, 353 vibrational spectroscopy
soldering, 349 LFRS measurements, 140, 142–146
solid-state bonding, 349 spheroidal modes, 140–141
solid-state sintering, 349–350 torsional modes, 140
thermal effects, 351–352 vibrational frequency calculation, 141–142
thermal sintering, 353 Metal-oxide-semiconductor field-effect transistors
thermodynamic driving force, 350 (MOSFETs), 461
ultrasonic sealing, 354 Metal oxide (MO) semiconductors, 283–284
Metal nanomaterial synthesis Nb2O5, 291–292
bottom-up approach, 345 SnO2, 290–291
chemical reduction methods SrTiO3, 293
at ambient pressure, 346, 348–349 TiO2, 288–290
capping agents, 345 ZnO, 291
hydrothermal/solvothermal synthesis, 346–347 Microwave (MW) heating, 346–347
ionic precursor and reducing agent, 345 Mie scattering theory, 131–132, 155
polymers, 345 Modified Penn model, 227
polyol methods, 345–346 Molecular beam epitaxy (MBE), 7–8, 10, 16
sonochemical and sonoelectrochemical Multiple exciton generation (MEG), 202
methods, 347
electrodeposition, 347–349 N
laser ablation, 348–349
top-down approach, 345 Nanocrystals/nanoparticles
Metal nanoparticles compound semiconductor nanocrystals, 3–14
Ag nanoparticles definition, 1
chemical reduction method, 22 elemental semiconductor nanocrystals, 14–18
ion exchange/irradiation method, 21–22 embedded in dielectric film
STM, 23–24 Al nanoparticles, 28–31
TEM, 22–23 Ge nanocrystals, 26–29
XRD, 23 Si nanocrystals, 23–27
Au nanoparticles memristors, 469
biological methods, 19–20 amorphous Si embedded with Ag nanoparticles,
chemical methods, 18–19 473–476
physical methods, 19 electrochemical dynamics, metallic nanoparticles
TEM and HRTEM, 19–20 in dielectrics, 478–483
XPS technique, 21 electronic synapses realized with Si-based
XRD, 20–21 memristors, 475–479
bottom-up methods, 146 Fe3O4 nanoparticle assemblies, 471–472
ion-exchange processing, 152, 154–174 four fundamental circuit elements, 470–471
ion implantation, 147–153 memristive systems, 470
nucleation and growth control, 147 TiOx -based device, 471
optoelectronic properties, 147 MOSFETs, 461
thermochemical stability, 147 printing and flexible thin-film transistors
nanoscale dimensions, 128 inorganic microelectronic devices and circuits, 495
nonlinear optical susceptibility, 129 inorganic semiconductors, 494–495
optical properties, 129 Moore’s law, 493
506 Index

organic semiconductors, 494 O

printable CdSe nanocrystal, inorganic TFTs,
495–497 Optical absorption spectroscopy
solution-based printing, 494 Arrhenius graph, 166–167
spin-coated CdSe nanocrystals for flexible TFTs, calculated and experimental optical absorption
497–498 spectra, 165–166
self-learning devices chemical identification, 162–163
neural networks, 484 damping, 166
Ni-rich NiO thin films, 484–491 degree of supersaturation, 165
transient memory loss behavior, 490–495 excitons, 163–164
single-electron devices ion-exchanged glass sample before and after thermal
Coulomb blockade effects, 463–465 annealing, 162–163
SETs, 465–470 optical absorbance, 166
single-electron box, 462–463 photoexcitation of electrons, 163
ULSI elements, 462 quasi-continuous exciton energy bands, 164
synthesize techniques, 1 recorded optical absorption spectra, 164
Nanomaterials room-temperature optical absorption spectrum,
vs. bulk materials, 130 162–163
definition, 128 silver atom nucleation, 165
particle sizes vs. physical properties, 130 silver monoxide, 163
Nanoparticles/nanostructures Optical properties, semiconductor
antibacterial activity, Cu, Ag colloids, dielectric function, 215
409–410 electronic band structures, 215–216
LAL Maxwell’s equations, 215
with ns pulses, 368 Optical rectification (OR), 328, 335–336, 338–339
review articles, 368–370 complex amplitudes, 329
NLO properties cubic QDs, 309
Ag colloids, 406–408 semiparabolic QDs, 309
Al colloids, 408–409 SONS, 331
Cu colloids, 403–406 Ostwald ripening, 148
ULA, 368
Ag target ablation in water, 377–391 P
Cu target ablation, 389–403
experimental details, 376–378 Palladium nanoparticle enhanced giant photoconductivity
metals in ambient air, 370–371 current voltage characteristics under different ambient,
metals in liquids, 371–372 440, 442
parameters, 373–375 mechanism
Nanoscale plasmonic materials, 368 catalytic effect, 444
Nanostructures electrical measurement, 443
1D, 297–298 electrons and holes recombination, 443
1D–2D–3D, 299–300 I–V curves, 441–442
3D, 298 Pd NPs electron affinity-induced degradation, 441
2D, 298 Pd NPs-surface modulated LAO/STO with UV
Near-band edge (NBE) emission light irradiation, 441, 443
Mg- and Mn-doped ZnO tetrapods, 49 persistent photocurrent, 444
nanopencils, 56 resistance–temperature curve, 441–442, 444
ZnO nanoneedles and nanonails, 56 surface/interface charge coupling, 444
NiO-sensitized ZnO nanosheets thickness-dependent response to gases,
adsorption density, 75 443, 445
depletion layer, 76–77 Pd/LAO film growth, 433–434
energy band diagram, 75–76 photoresponse under UV light and room light
operating temperature, triethylamine, 75 giant optical switching effect, 436
response and recovery characteristics, 75 persistent photoconductivity effect, 437
sensing mechanism, 75–76 photoemission spectra, conventional tube lamp,
Nonlinear optical (NLO) properties 437–438
Ag colloids, 406–408 photoresponse characteristics, 437–438
Al colloids, 408–409 surface morphology
Cu colloids AFM image, 434–435
closed-aperture Z-scan curves, 406 bare TiO2-site-terminated STO (001) substrate,
energy band diagram, 405 435–436
nonlinear absorption, plasmonic metal HRTEM image, 434–435
NPs, 405 TEM analysis, 435–436
open-aperture Z-scan data, 403–405 time-dependent resistance, 439–441
Nonlinear optical susceptibility, 310–311 wavelength-dependent photoconductivity, 437–439
Index 507

PbS-sensitized ZnO nanorods MSM device diagram, 61–62

current density vs. voltage characteristics, 67 photoconductive gain, 60–61
energy band diagram, 67–68 spectral photoresponse, 60–63
MEH-PPV/ZnO device, 67–68 TEM, 58–59
PbS QDs, 66 UV–visible absorption spectra, 58–59
Photoelectrochemical cells Plasmons, 127
components, 283 Polar liquid molecule-induced transport property
DSSC modulation
MO semiconductor, 283–284, 288–293 intrinsic polar catastrophe model, 428
structure and mechanism, 284–286 mechanism
nanostructures current–time curve, 431–432
1D, 297–298 linear I–V characteristic, 431–432
1D–2D–3D, 299–300 Schottky junction band diagram, 430–431
3D, 298 temperature-dependent I–V curves, 431, 433
2D, 298 tuning effect, 431
QDSSC polar liquid sensing, 428–429
metal chalcogenide semiconductors, 284, 293–297 Schottky-diode-like junction, 430
MO semiconductor, 283–284, 288–293 Polymer–metal oxide hybrid bulk heterojunction solar
structure and mechanism, 286–288 cells, 63
semiconductor assembly process Polymer–metal oxide hybrid solar cell, 63
EPD, 301 Power electronic-based energy storage systems
LbL technique, 300 (PEBESSs), 357
self-assembly technique, 300 Power electronics
Photoinduced Au–ZnO plasmonic gas sensors applications
light-induced gas sensing, 82 electric vehicles (EVs), 356–357
LSPR characteristics, 82 military, aerospace, and high-power semiconductor
photoresponsivity, 82 device packaging, 359
room-temperature sensors, 82 renewable energy (RE), 357–358
sensing mechanism, 84–86 smart grid and microgrids, 357–358
sensor response, 82–83 bonding materials, 344
time pulse sequence, 83 packaging components, 344
wavelength sensitivity, 84 power converters, 344
wavelength-tunable gas sensors, 82 power semiconductor modules, 344
Photoluminescence (PL) Power semiconductor modules, 344
dynamics Printing and flexible thin-film transistors
exciton migration, 202 inorganic microelectronic devices and circuits, 495
PL decay times, 199–201 inorganic semiconductors, 494–495
PL intensity, 199, 201 Moore’s law, 493
PL recombination rate, 200, 202 organic semiconductors, 494
quantum confinement, 201 printable CdSe nanocrystal, inorganic TFTs, 495–497
optical transitions, 255 solution-based printing, 494
porous silicon, 256 spin-coated CdSe nanocrystals for flexible TFTs, 497–498
from Si+-implanted Si3N4 thin films
excitation transition process, 268 Q
PL excitation (PLE) spectra, 268–269
PL intensity with annealing temperature, 267 Quantum dash (Qdash) lasers
vs. pure silicon nitride thin film PL spectra, 266–267 dynamic characteristics, 99–101
from Si+-implanted SiO2 thin films, 263–264 nanostructures, 96–97
spectra Quantum dot (Qdot) lasers
band structure, 171–172 buried InAs/InP quantum dots
high-temperature-annealed samples, 173 carrier-induced refractive index, 104–105
intensity shift and quenching, 170 carrier scattering processes, 103–104
photoexcited electrons, 169–170 complex optical susceptibility, 105
room-temperature PL spectra before and after electronic structure, 101–102
annealing, 169 gain, 104–106
thermal annealing, 172 line-width enhancement factor, 106–107
thermal decomposition, 172 macroscopic and microscopic polarization, 105
time-resolved PL decay, 171 phenomenological gain compression factor, 106
Photonic sintering, 354 dynamic characteristics, 99–101
Plasma-enhanced chemical vapor deposition fabrication, 96
(PECVD), 195 free-running quantum dot lasers
Plasmonic ZnO photodetector amplitude modulation response, 107–111
combined SAED pattern, 58–59 carrier capture and relaxation processes, 116–120
LSPR effect, 61 line-width enhancement factor, 111–118
508 Index

GaAs-based Qdot laser devices, 96–97 R

InAs/GaAs Qdot systems, 97
InAs/InP Qdots, 97 Recombination transition, 255
InGaAs/GaAs quantum dots, 96 Resonant absorption of light, 127
InP-based quantum dot lasers Rutherford backscattering (RBS)
(311)B InP substrate, 97–98 ion-beam sputtering, 178
dynamic performance, 99–101 ion-exchange processing
(100)InP substrate, 98–99 accumulated mass of silver atoms per unit area, 156
Quantum dots diffusion theory, 156
vs. bulk semiconductor, 308–309 He+ ions, 155
calculated subbands, 319–321 near-surface accumulation, 155
density matrix formulation, optical susceptibility silver ion-exchanged soda glass, 155–156
electric susceptibility, 323 temperature dependence, 157
equation of motion, 320–321 thermal diffusion, 157
Hamiltonian operator, 321
IB dipole moment, 324 S
IB SONS, 323
ISB dipole moment, 324 Second-harmonic generation (SHG), 309, 328, 332–333
ISB SONS, 324 complex amplitudes, 329
perturbation series, 322 energy-level diagram, 312
polarization density, 323 geometry, 312
polarization per unit volume, 322 nonlinear polarization, 312–313
potential profiles and relevant parameters, 323 nonzero frequencies, 329
inhomogeneity, 331–332 piezoelectric crystals, 311
quantum disk model Second-order nonlinearity
Hamiltonian, 315–316 DFG, 328
longitudinal eigenenergy, 319 complex amplitudes, 329
potential energy profile, 317 electric susceptibility, 330
transverse eigenenergy, 316, 319 energy-level description, 313
wave function, 316–318 geometry, 313
second-order nonlinearity, 313–314 nonzero frequencies, 329
semiconductor nanostructures, 308 optical parametric amplification, 330
semiconductor nonlinearities, 309–310 SONS, 330
SONS electric polarization vector, 311
conduction ISB transitions, 324, 326, 328 noncentrosymmetrical crystals, 311
DFG, 334–337 nonlinear optics, 313
IB transition, 324–325 OR, 328
lens-shaped QD, 314 complex amplitudes, 329
millimeter wavelengths, 338 cubic QDs, 309
momentum matrix element, 328, 336–337 semiparabolic QDs, 309
nonmonotonic behavior, 309 SONS, 331
OR, 335–336, 338–339 QDs, 313–314
quantum confinement, 309 SFG, 328
self-assembled QDs, 314 complex amplitudes, 329
SFG, 334–335 energy-level description, 313
SHG, 332–333 geometry, 313
Stark shift, 336 nonzero frequencies, 329
symmetry breaking, 314 SONS, 330
valence ISB transitions, 324, 327–328 SHG, 328
structure, 314–315 complex amplitudes, 329
surface states, 309 energy-level diagram, 312
zero-dimensional semiconductor systems, 308 geometry, 312
Quantum dot-sensitized solar cell (QDSSC) nonlinear polarization, 312–313
metal chalcogenide semiconductors, 284 nonzero frequencies, 329
cadmium chalcogenide quantum dots, 295–296 piezoelectric crystals, 311
cadmium selenium telluride, 296–297 Second-order nonlinear susceptibility (SONS)
lead chalcogenide quantum dots, 293–295 conduction ISB transitions, 324, 326, 328
zinc chalcogenide quantum dots, 296 DFG, 330, 334–337
MO semiconductor, 283–284, 288–290 IB transition, 324–325
structure and mechanism, 286–288 lens-shaped QD, 314
Quantum-well (Qwell) lasers, 96–97 millimeter wavelengths, 338
Quantum wire (Qwire) lasers, 96 momentum matrix element, 328, 336–337
Quasi-two dimensional electron gas (2DEG) nonmonotonic behavior, 309
interface, 424 OR, 331, 335–336, 338–339
Index 509

quantum confinement, 309 integrated EL intensity, 273–274

self-assembled QDs, 314 I–V characteristics, 273
SFG, 330, 334–335 electrically tunable EL, 277–279
SHG, 332–333 properties and mechanisms
Stark shift, 336 band diagram, 271–272
symmetry breaking, 314 band tail radiative recombination, 272–273
valence ISB transitions, 324, 327–328 EL spectra at different gate voltages, 270
Self-learning devices gate current and integrated EL intensity, 270–271
neural networks, 484 UV and NIR band, 271
Ni-rich NiO thin films violet emission, 272
circuit implementation, 488–489 structure fabrication, 269–270
forward and backward voltage sweepings, 487–488 Silicon nanocrystals (nc-Si)
I–V characteristics, 484–485, 487–488 ab initio approaches, 216
metal-insulator-metal (MIM) structure, 484–485 applications, 191, 205–206
minimum number of voltage sweeping cycles, bulk Si crystal, 215
487, 489 colloidal chemistry, 196
minimum numbers of pulses, 486–487, 490–491 electrochemistry, 193–194
neuroplasticity, 489–490 electrons and holes, carriers, 192
pulse voltages, 486–487 elemental semiconductor nanocrystals
resistance evolution, 486 AFM, 15–16
resistive switching device, 484 chemical etching, 14
STM and LTM stages, 484–485 dry etching, 14–15
transient memory loss behavior electrochemical etching, 14
device conductance, 491–492 SEM, 15
forward and backward voltage sweepings, 493–494 TEM, 15
I–V characteristics, 493–494 XRD, 16–17
memory decay, 490 embedded in dielectric matrices
mimic memory loss function, 492 silicon dioxide, 194–196, 239–243
short-term memory retention, 490 silicon nitride–silicon oxynitride, 195–196,
statistical minimum number of pulses, 492–493 243–246
statistical minimum number of sweeping cycles, fundamental luminescence properties
493, 495 amorphous silicon quantum dot bandgap, 258
STM and LTM, 491–492 exciton Bohr radius, 257–258
Semiconductor lasers low-energy peak, oxidized nc-Si, 261
bulk-structure lasers, 96 luminescence wavelength, 257
optical data communication and telecommunication NIR band, 261–262
applications, 95–96 peak energy vs. average diameter, 259
Qdash lasers, 96–97 photon emission, 260
Qdot lasers, 96–97 PL lifetime, temperature dependence, 260
Qwell lasers, 96–97 PL spectra from Si-implanted SiO2 layers, 261–263
Qwire lasers, 96 PL spectra of nc-Si embedded in SiO2 thin films,
Semiconductor nanocrystals 258–259
compound PL spectrum at room temperature, 261
CdSe nanocrystals, 3–5 quantum confinement effect, 257–258
CdS nanocrystals, 4–7 radiative electron–hole recombination processes,
chemical methods, 2 257–258
CuCl nanocrystals, 11–14 spatial localization, 257
GaAs nanocrystals, 7–10 SRN materials, 262–263
III–V compound, 7 growth by laser ablation, 196
InAs nanocrystals, 9–13 history, 193
MPA-capped CdSe nanocrystals synthesis, 3 hydrogenated nc-Si
potential applications, 2 absorption coefficient, 223, 225
TOP/TOPO/HPA method, 2–3 average Si–Si bond lengths, 217
TOP/TOPO method, 2 cluster relaxation, 217–218
elemental cubic supercells, 216
Ge nanocrystals, 16–18 dielectric functions, 223–224, 226
Si nanocrystals, 14–17 electron energy vs. k, 227
Semiconductor nanostructures, 308 electron–hole pair, 217
Semiconductor nonlinearities, 309–310 energy separation, 227
Si-based LEDs, 256 exciton Coulomb energy, 221–222
Si+-implanted Si3N4 thin films, electroluminescence GPM, 228
annealing effect ground-state total energies, 220
EL band evolution, 274–276 HOMO–LUMO transition, 218, 220
EL spectrum deconvolution, 274–275 hydrogen-passivated spherical Si clusters, 220
510 Index

hydrogen passivation, 216–217 Coulomb gap, 469–470

Kohn–Sham energy levels, 217–219 fabrication, 467
optical absorption spectra, 223 logic operation implementation, 467–468
quantum size effect, 220 working principles, 465–467
quasiparticle and LDA band gaps, 220–221 single-electron box, 462–463
quasiparticle gaps, optical gaps and experimental ULSI elements, 462
absorption data, 221–222 Single-electron transistors (SETs)
size-dependent dielectric constant, 227–228 CdSe nanocrystals, 468
static dielectric constant, 226–229 composite grayscale plots, 469–470
structural properties, 217 conductance vs. gate voltage, 468–469
TDLDA absorption spectra, 223–224 Coulomb gap, 469–470
total polarization dielectric constant, 229 fabrication, 467
light emission and photoconductivity, 192 logic operation implementation, 467–468
miniaturization and integration, 192 working principles
nonvolatile memories, 205 electron probability, 467
optical properties energy levels, 466
electroluminescence, 203–204 SETs structure, 465–466
light emission, 196–199, 202–203 source-to-drain current, 466
PL dynamics, 199–202 Smart grid and microgrids, 357–358
oxidation of porous silicon, 194 SnO2 nanowire/ZnO nanorod heterostructures
oxidized nc-Si operating temperature, 71–72
bonding structures, 231 pulsed laser deposition and hydrothermal
hydroxyl-passivated Si35(OH)36 nanocrystal, 231 methods, 71
interface radiative states, 229 sensing mechanism, 72–75
optical gaps, 229–230 sensing responses, 72
oxygen contamination, 229 sensing transients, 71–72
oxygen-free and oxygen-rich Si nanocrystals Solid-state sintering, 349–350
absorption spectrum, 231–232 Sonochemical synthesis, 347
single-bonded surface oxygen atoms, 231 Sonoelectrochemical methods, 347
Si–O vibrations, 229 Spiking-time-dependent plasticity (STDP), 477–478
surface and interface properties, 229 Stain etching, 194
TDDFT, 229 Sum-frequency generation (SFG), 328, 334–335
Silicon nanocrystals embedded in dielectric matrices complex amplitudes, 329
Si3N4 thin films, 195–196 energy-level description, 313
annealing effect, 245–246 geometry, 313
dielectric function, 244–246 nonzero frequencies, 329
Kramers–Kronig-consistent Tauc–Lorentz SONS, 330
model, 244 Surface-modified ZnO sensors
LPCVD, 243 metal-sensitized ZnO gas sensors
multilayer optical model, 243–244 Au-functionalized ZnO sensor, 79–81
spectral fittings, 245 photoinduced Au–ZnO plasmonic gas sensors, 82–86
TEM image, 243–244 semiconductor-sensitized gas sensors
SiO2 thin films, 194–196 CuO/ZnO nanorod heterostructures, 77–79
bandgap expansion, 241, 243 NiO-sensitized ZnO nanosheets, 75–77
dielectric function, 239–241 SnO2 nanowire/ZnO nanorod heterostructures, 71–75
FB model, 239–241 Surface-plasmon-enhanced Raman scattering, 129
refractive index and extinction coefficient, 239, 242 Surface-plasmon polaritons (SPPs), 128–129
screening effect, 240 Surface-plasmon resonance (SPR), 127
static dielectric constant, 239–240, 242
Tauc plots, 241 T
Silicon-rich silicon nitride (SRN) materials, 262–263
Single-electron devices Thermal sintering, 353
Coulomb blockade effects Time-dependent density functional theory (TDDFT), 229
charging energy of quantum dot, 464 Time-dependent local density approximation
free energy, 463 (TDLDA), 234
free energy change, 464
HfAlO/Au NCs/HfAlO trilayer structure, 464–465 U
I–V curves, capacitor structure, 464–465
polarization charge, 463–464 Ultrafast laser ablation (ULA), 368
SETs Ag target ablation in water
applications, 467–470 with different angles of incidence, 381–388
CdSe nanocrystals, 468 with different fluences and scanning conditions,
composite grayscale plots, 469–470 377–383
conductance vs. gate voltage, 468–469 with ps/fs Bessel beams, 388–391
Index 511

Cu target ablation growing methods, 41

in chloroform, 397, 399–400 morphologies, 41–42
with corroles and surface-enhanced fluorescence/ photovoltaic devices
SERS studies, 400–403 CdS-sensitized ZnO nanorods, 63–66
with different energies and scanning conditions, cosensitized (CdS and CdSe) ZnO nanowires,
389–398 68–69
experimental details, 376–378 DSSCs, 63
metals in ambient air, 370–371 dye- and quantum-dot-sensitized solar cells, 63
metals in liquids, 371–372 hybrid solar cells, 63
parameters PbS-sensitized ZnO nanorods, 66–68
energy per pulse, spot size, and fluence, 374–375 shape- and size-dependent properties, 56
input wavelength, 373 tetrapods
number of pulses, 375 electrical properties, 49–53
pulse duration, 373–374 FE properties, 53–55
Ultra large-scale integration (ULSI) elements, 462 optical properties, 42–49
Ultrasonic sealing, 354 tripods, optical properties, 42–44
Ultraviolet (UV) emitters, 56–57 UV emitters, 56–57
UV photodetectors
V charge carrier density, 58
LSPR-induced absorption and scattering
Vibrational spectroscopy phenomena, 58
LFRS measurements, 140 MSM photodetectors/photoconductors, 58
cobalt nanoparticles, 143–146 plasmonic ZnO photodetector, 58–63
silver nanoparticles, 142–143 ZnO nanowire, 58
spheroidal modes, 140–141 ZnO tetrapods
torsional modes, 140 electrical properties
vibrational frequency calculation, 141–142 complex-plane impedance spectroscopy, 49
relaxation time, 49
X Sn-doped tetrapods, 50–53
undoped tetrapods, 49–50
X-ray photoelectron spectroscopy (XPS) field emission (FE) properties
Auger spectroscopy, 158–159 current density, 53
Au nanoparticles, 21 Mg-doped tetrapods, 55
ion-exchange processing, Ag nanoparticles, 157–158 Sn-doped tetrapods, 54–55
optical properties
Z advantage, 42
Mg and Mn doped tetrapods, 49
Zinc oxide (ZnO) room-temperature PL characteristics, 42–43
noncentrosymmetric structure, 40 Sn- and P-doped tetrapods, 45–48
n-type semiconductor, 40 three and four rod-shaped arms, 41
oxygen adsorption and desorption, 40–41 undoped tetrapods, 44–45
piezoelectric and pyroelectric properties, 40 ZnO tripods, optical properties
wurtzite structure, 39–40 three and four rod-shaped arms, 41
zinc blende (ZB) structure, 40 undoped tripods
ZnO nanostructures Bose–Einstein-type expression, 43
biosensors, 86–87 PL emission characteristics, 43–44
gas sensors, 69–70 room-temperature PL characteristics,
parameters, 70–71 42–43
surface-modified ZnO sensors, 71–86 temperature-dependent PL spectra, 42, 44