Krystyna Jackowska, Paweł Krysiński

Applied Electrochemistry
Also of Interest
X-Ray Studies on Electrochemical Systems.
Synchrotron Methods for Energy Materials
Braun, 
ISBN ----, e-ISBN ----

Electrochemical Methods for the Micro- and Nanoscale.

Theoretical Essentials, Instrumentation and Methods for Applications in MEMS
and Nanotechnology
Kieninger, 
ISBN ----, e-ISBN ----

Electrochemical Energy Storage.

Physics and Chemistry of Batteries
Job, 
ISBN ----, e-ISBN ----

Maths in Chemistry.
Numerical Methods for Physical and Analytical Chemistry
Bansal, 
ISBN ----, e-ISBN ----

Solid-State Chemistry
Hofmann, 
ISBN ----, e-ISBN ----
Krystyna Jackowska, Paweł Krysiński

Applied
Electrochemistry

Aspects in Material and Environmental Science

2nd Edition, expanded and amended

Authors
Prof. Krystyna Jackowska
Professor Emeritus
University of Warsaw
Faculty of Chemistry
Laboratory of Electrochemistry
Pasteur 1
02-093 Warsaw
Poland

Prof. Paweł Krysiński

University of Warsaw
Faculty of Chemistry
Laboratory of Electrochemistry
Pasteur 1
02-093 Warsaw
Poland

ISBN 978-3-11-116034-4
e-ISBN (PDF) 978-3-11-116098-6
e-ISBN (EPUB) 978-3-11-116344-4

Library of Congress Control Number: 2024933394

Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie;
detailed bibliographic data are available on the Internet at http://dnb.dnb.de.

© 2024 Walter de Gruyter GmbH, Berlin/Boston

Cover image: The template-assisted electrodeposited Pd nanowires. Courtesy of Dr.
Magdalena Tagowska, Faculty of Chemistry, University of Warsaw
Typesetting: Integra Software Services Pvt. Ltd.
Printing and binding: CPI books GmbH, Leck

www.degruyter.com
Preface
A multiplicity of definitions exists for applied electrochemistry as a discipline. The
most general, attempting to embrace all aspects, can be formulated as science that
aims at improving our life, taking advantage of the phenomena occurring at the inter-
faces between metallic or semiconducting electrodes and electrolyte solutions (so-
called electrodics) as well as those occurring in the bulk of the electrolyte solutions
(so-called ionics). Such improvement can come through the understanding of these
phenomena, with subsequent construction and design of new devices or systems that
can be used not only by industry (e.g., catalysis) but also in more personalized appli-
cations, such as batteries for smartphones, pacemakers, solar panels, and so on.
To facilitate understanding of these phenomena, Basic concepts part in this book
present first a brief thermodynamic background allowing the reader to understand
the electrified interfaces and electrolyte solution under equilibrium and steady-state
conditions. Then it continues with a basic introduction to the structure of interfaces,
followed by the charge transfer processes occurring at metal/electrolyte interface
from the point of view of equilibrium and nonequilibrium phenomena, defining the
so-called reversible and irreversible electrodes. The next part begins with an intro-
duction to the selected electrochemical methods applied in the analytical and material
chemistry that benefit from the essential knowledge gained in the previous chapters.
The next part of this book guides through the Electrochemistry in material sci-
ence – selected topics. It aims at describing in more detail how the electrochemistry
can be used in corrosion science, catalysis, deposition of new material on a conduct-
ing support, allowing to get insight into the mechanism of deposition and its kinetics.
This chapter will lead to the nanostructured materials of different dimensions, organi-
zation, and topologies. Then, a very broad area, extremely important nowadays for
our population, namely – the energy storage and conversion, will be set forth in order
to show how the electrochemistry can be applied to understand the working of batter-
ies and fuel cells, to stimulate and push forward their development and design, thus
improving their reliability and durability. This part will conclude with the applica-
tions of electrochemistry in biology and medicine, thereby improving the quality of
life of patients and providing information on the energetics of living organisms.
Finally, the last part, Photoelectrochemistry in material science – selected topics,
will guide the reader through the selected topics of photoelectrochemistry, where light
acts as a power source for electrical energy generation, photocatalysis, and photoelectro-
catalysis, including the nanoscale processes at the semiconductor nanoparticles.
The goal of this book is to show all graduate and PhD students that electrochemis-
try not only has many applications for understanding of various phenomena in nowa-
days life but also has many applications in practical devices and can also stimulate
new science-enabled technologies, nourishing leaps from bench-top to large-scale indus-
tries, providing also means for protecting our environment. Our book is based on lec-
tures given at the University of Warsaw, Faculty of Chemistry. It addresses advanced

https://doi.org/10.1515/9783111160986-202
VI Preface

students and PhD students of chemistry, physics, engineering, and related subjects and
also those scientists who want to get a solid background knowledge of this area.
We hope that this background will be useful to the interested reader, encouraging
him/her to take the next several steps into the attractive area of applied electrochem-
istry. Our aim is to guide him/her through the selected topics of contemporary life in
which electrochemistry progressively improves our knowledge and quality of life.

Acknowledgments: Krystyna Jackowska dedicates her part of this book to Polish

and valued Russian friends (Dr. R.V. Ivanova and Prof. M.N. Rodnikova from Moscow)
for their encouragement and support. Special thanks are due to Prof. Magdalena
Skompska and Prof. Marek Szklarczyk from Warsaw University, and to Prof. G.A.Tsir-
lina and Prof. O.A. Petrii from Moscow State University for some reviews and referen-
ces. She is also very grateful to her family for patience.
Paweł Krysiński highly acknowledges the understanding and forbearance of his
family, with special thanks due to his wife, Hanna, for her extreme patience, kindness,
support, and help.
The authors thank Dr. Magdalena Osial for her skills and time devoted for prepar-
ing all the graphics in this book. Without her work and willingness to share some of her
experimental data, this book will be largely delayed, to say nothing about its form.
The authors are also greatly indebted to the editorial staff of De Gruyter.
Preface to the 2nd edition
In the still growing awareness of the necessity of minimizing the deteriorating impact
on the environment by all fundamental activities of mankind, electrochemistry can
provide tools for environmental control, pollution avoidance, and destruction of pol-
lutants as well as clean energy conversion and storage. Moreover, the use of electrons
in electrochemical processes is considered a “green” model in its preventive mode
(electrosynthesis, hydrogen production, power storage, metal recovery/recycling) and
the control/remedial mode (wastewater treatment, air pollution control, destruction
of organic pollutants). In all these processes, the electron as a reagent frequently
leads to cleaner technology by simple avoidance of the use of hazardous chemicals
and allows for subsequent recovery, recycling, or at least pollution control.
Therefore, in this second edition of the Applied Electrochemistry textbook, we
added a brief review of current and emerging approaches utilizing electrochemistry
for a cleaner environment, the so-called “green” electrochemistry. This review is con-
tained in Part IV, entitled Electrochemistry in Environmental Science – selected
topics.
This part entails the utilization of electrochemistry to monitor the environment
for certain most commonly polluting gases (Chapter 13. Sensing the environment).
This chapter classifies and shows the applications of electrochemical sensors in moni-
toring the safety of life through the development of rapid, selective, and miniaturized
detection techniques that can trigger further remediation processes. As described in
Chapter 14 entitled Green fuel – hydrogen production. Environmental pollution,
and global warming, have triggered vast research for new, promising sources of
clean, “green” energy, such as hydrogen. One of the simplest methods for hydrogen
production is water electrolysis. Therefore, this chapter encompasses methods and
technologies currently used in water electrolysis, stressing out its thermodynamics,
efficiency, and energy losses. A brief description of a new method of water electrolysis
that is based on bioelectrochemical systems, such as microbial electrolysis, is also in-
cluded. Furthermore, this chapter presents an outline of solar energy utilization in
water electrolysis, including the semiconducting nanostructured systems and again,
thermodynamics and efficiency for such devices.
Then, we will describe the electrochemical processes and methods in the elimina-
tion of the most common pollutants (Chapter 15. Electrochemical and photocata-
lytic methods in pollutant removal). An outline is given to the electrochemistry in
the removal and recovery of heavy metals, followed by the removal of organic pollu-
tants via the electrochemical treatment. This chapter finally will summarize the appli-
cation of solar energy for a cleaner environment with a description of the processes
utilizing this renewable and sustainable source of natural energy in the removal and
degradation of some inorganic and organic pollutants.
To better understand all processes involved in the application of electrochemistry
for a greener environment, some modifications were also necessary to previous chap-

https://doi.org/10.1515/9783111160986-203
VIII Preface to the 2nd edition

ters, particularly to the Basic concepts part. And so, the Basic Concepts part, as be-
fore, consists of Chapter 1.1 Structure of interfaces, which introduces a brief ther-
modynamic background for understanding the electrified interfaces and electrolyte
solution under equilibrium and steady-state conditions. As in the first edition it con-
tinues with a basic introduction to the structure of interfaces, followed by the charge
transfer processes occurring at metal/electrolyte interface from the point of view of
equilibrium and non-equilibrium phenomena, defining the so-called “reversible” and
“irreversible” electrodes. However, the Basic concepts part in the current edition
was expanded to include Chapter 1.2 Structure of the bulk of electrolytes. Conduc-
tivity, followed by Chapter 1.3 Nonaqueous electrolytes, describing the structure of
the bulk of aqueous electrolytes, conductivity, as well as nonaqueous electrolytes, in-
cluding ionic liquids, molten salts, and solid electrolytes. Finally, Part I was amended
with a brief background to Membranes and membrane potentials (Chapter 1.4),
both under equilibrium conditions and non-equilibrium conditions, necessary for a
better understanding of the design of batteries and other storage devices.
We hope that this 2nd edition, completed with new information contained in the
additional chapters will be found interesting and useful, by giving the background
knowledge and directions in the applications of electrochemistry for cleaner, “greener”
and sustainable environment and quality of life.
Contents
Preface V

Preface to the 2nd edition VII

Part I: Basic concepts

1 Basic concepts 3
1.1 Structure of interfaces 3
1.1.1 Electrical double layer at interfaces: metal/electrolyte 7
1.1.2 Electrochemical potential – potentials at interfaces: internal, surface,
external potential 14
1.1.2.1 Metal–solution interface at equilibrium: Nernst equilibrium
potential 18
1.1.2.2 Electron work function 20
1.1.3 Charge transfer processes across the metal/electrolyte interface 21
1.1.3.1 Basic concepts of nonequilibrium thermodynamics 21
1.1.3.2 “Reversible” electrode processes 27
1.1.3.3 “Irreversible” electrode processes: basic concepts of electrochemical
kinetics 32
1.1.3.4 Briefly on Marcus, Hush, Levich, Dogonadze (MHLD) theory of
electrode processes 37
Bibliography 39
1.2 Structure of the bulk of electrolytes: conductivity 40
1.2.1 Interactions in aqueous electrolytes 40
1.2.1.1 Ion–solvent interaction; solvation 40
1.2.1.2 Ion–ion interactions 41
1.2.2 Conductivity of electrolytes 44
1.3 Nonaqueous electrolytes 48
1.3.1 Ionic liquids 48
1.3.2 Molten salts 50
1.3.3 Solid electrolytes 51
Bibliography 53
1.4 Membranes and membrane potentials 53
1.4.1 Equilibrium potentials 53
1.4.2 Nonequilibrium potentials 56
Bibliography 63
X Contents

2 Selected electrochemical methods applied in analytical chemistry

and material science 64
2.1 Transient methods 64
2.1.1 Capacitive current 65
2.1.2 Faradaic current 67
2.2 Electrochemical impedance spectroscopy 70
2.2.1 Data presentation 74
2.3 Electrochemical quartz crystal microbalance method 82
Bibliography 87

Part II: Electrochemistry in material science – selected topics

3 Corrosion 91
3.1 General remarks 91
3.2 Corrosion – what does it mean? Mechanism of corrosion 91
3.3 Characterization of corrosion: corrosion potential, corrosion
current 97
3.3.1 Stability of materials: potential/pH (Pourbaix) diagrams – the
thermodynamic aspect 98
3.3.2 Stability of materials: current–potential (Evans) diagrams – the kinetic
aspect 103
3.4 Evaluation of corrosion rate from electrochemical measurements 105
3.4.1 Linear scan voltammetry 106
3.4.2 Electrochemical impedance spectroscopy in corrosion 107
3.5 Localized corrosion: pits, crevices, intergranular corrosion – oxygen
reduction as accompanying cathodic reaction 109
3.6 Hydrogen evolution as accompanying reaction – role in corrosion:
embrittlement and cracking 111
3.7 Protection against corrosion 113
3.7.1 Electroplating 116
Bibliography 117

4 Electrocatalysis 118
4.1 General remarks 118
4.2 How to compare the activity of catalysts in electrochemical
reactions? 120
4.3 Electrocatalysts 122
4.3.1 Metals, alloys, and oxides 122
4.3.2 Carbon catalysts and supports 124
4.4 Catalyst activity 125
4.4.1 Electron work function effect 126
 Contents XI

4.4.2 Adsorption impact 126

4.4.3 Size effect 129
4.5 Electrocatalysts in hydrogen–oxygen fuel cells 131
4.5.1 Oxygen reduction reaction 132
4.5.2 Hydrogen evolution and oxidation reactions 135
Bibliography 136

5 Electrodeposition 138
5.1 General remarks 138
5.2 Electrocrystallization: nucleation and growth 139
5.2.1 Critical size of nuclei 141
5.2.2 Instantaneous and progressive nucleation 142
5.2.3 Analytical approach to experimental data 144
5.3 Deposit morphology 148
5.4 Practical aspects of electrodeposition 149
5.5 Electrodeposition of binary alloys and semiconductor compounds 152
Bibliography 156

6 Underpotential deposition (UPD) 157

6.1 General remarks 157
6.2 Experimental examples – UPD features 158
6.3 Underpotential deposits – catalytic properties 161
Bibliography 163

7 Electrochemical methods in the formation of nanostructures 164

7.1 General remarks 164
7.2 Template-assisted electrodeposition of nanostructures 165
7.2.1 Active template-assisted electrodeposition 165
7.2.2 Restrictive template-assisted deposition 166
7.2.2.1 Porous anodic alumina (PAA) matrices 167
7.2.2.2 Porous anodic titanium dioxide matrices 168
7.2.3 Concluding remarks 168
7.3 Applications of nanostructured materials 170
7.3.1 Catalysis 170
7.3.2 Batteries: supercapacitors 171
7.3.3 Photoelectrochemical devices: solar cells 171
7.3.4 Biological applications: nanozymes 171
Bibliogaphy 172
XII Contents

8 Electrochemistry in energy conversion and storage 175

8.1 Batteries 175
8.1.1 Electrochemical cell – fundamentals 175
8.1.2 Characteristics of batteries 179
8.1.3 Classification of batteries and examples 183
8.1.3.1 Primary batteries 183
8.1.3.2 Secondary batteries 184
8.1.3.3 Lithium batteries 186
8.1.3.4 Fuel cells 188
8.2 Supercapacitors 192
8.2.1 General remarks 192
8.2.2 Capacitor versus electrochemical cell/battery 193
8.2.3 Working characteristics of supercapacitors 198
Bibliography 200

9 Interfacing applied electrochemistry and biology 202

9.1 General remarks 202
9.2 Bioelectrocatalysis 203
9.3 Biosensors 208
9.4 Biofuel cells 212
9.4.1 Design and enzymes used in biofuel cells 214
9.4.1.1 Biocatalysts and design of biocathodes 214
9.4.1.2 Biocatalysts and design of bioanodes 216
9.4.2 Energy and power of the biofuel cells 217
Bibliography 221

Part III: Photoelectrochemistry in materials science – selected

topics

10 Semiconductors electrochemistry and photoelectrochemistry:

fundamentals 229
10.1 Basic characteristics of semiconductors 229
10.1.1 Semiconductor under illumination 232
10.1.2 Recombination 235
10.1.3 Carriers and electric field 236
10.2 Semiconductor–electrolyte interface 237
10.2.1 Model of semiconductor–electrolyte interface 237
10.2.2 Space charge layer 239
10.2.3 Polarization of semiconductor–electrolyte interface 242
10.3 Fundamentals of electrochemical reactions on the SC electrode 243
10.4 Photoeffects at/in semiconductor electrode in electrolyte 250
 Contents XIII

10.5 Influence of surface states 254

10.6 Corrosion, photocorrosion, and stability 258
Bibliography 262

11 Solar energy conversion in photoelectrochemical cells 264

11.1 General remarks 264
11.2 Efficiency and key parameters 265
11.3 Photoelectrochemical cells: classification – principle of operation 267
11.3.1 Liquid photovoltaic cells 267
11.3.2 Dye-sensitized solar cells 269
11.3.3 Quantum dot-sensitized solar cells 271
11.3.4 Photoelectrosynthetic cells 272
Bibliography 278

12 Semiconductor particles in photocatalysis 280

12.1 General remarks 280
12.2 Size effect and dimensionality 281
12.3 Photocatalytic degradation of organic/inorganic pollutants 284
Bibliography 288

Part IV: Electrochemistry in environmental science – selected

topics

13 Sensing the environment 291

13.1 General remarks 291
13.2 Electrochemical sensors 292
13.2.1 Amperometric sensors 293
13.2.2 Potentiometric sensors 302
13.2.3 Conductometric sensors 307
13.3 Future perspectives and challenges 310
Bibliography 312

14 Green fuel: hydrogen production 313

14.1 General remarks 313
14.2 Water electrolysis 314
14.2.1 Thermodynamics, energy losses, and efficiency 314
14.2.2 Methods and technologies 321
14.2.2.1 Alkaline water electrolysis (AWE) 322
14.2.2.2 Polymer electrolyte membrane electrolysis (PEM) 323
14.2.2.3 Solid oxide electrolysis (SOE) 325
14.2.2.4 Seawater electrolysis 326
XIV Contents

14.2.2.5 Microbial electrolysis (MEC) 329

14.3 Solar energy in photo water splitting 331
14.3.1 Thermodynamics, energy losses, and efficiency 333
14.3.2 Photoelectrolysis of water: trends 340
14.3.3 Photocatalysis of water: trends 345
References 350

15 Electrochemical and photocatalytic methods in pollutant

removal 352
15.1 General remarks 352
15.2 Removal and recovery of heavy metals: electrochemical
treatment 353
15.2.1 Electrodeposition and electrodialysis 354
15.2.2 Electrocoagulation and electroflotation 357
15.3 Degradation and removal of organic pollutants: electrochemical
treatment 361
15.3.1 Fenton and electro-Fenton oxidation 361
15.3.2 Anodic oxidation 363
15.4 Solar energy: pollutant degradation and removal 367
15.4.1 Inorganic pollutants 368
15.4.2 Organic pollutants 370
References 372

List of abbreviations 375

List of Symbols 379

Index 385
Part I: Basic concepts
1 Basic concepts

1.1 Structure of interfaces

To begin with, it is necessary to establish a solid foundation for all topics presented in
the following chapters. This foundation can be derived from the laws of thermody-
namics that provide tools not only for qualitative and quantitative description of sys-
tems and processes but also capabilities to predict their further development. There
are four state functions in thermodynamics, namely, the internal energy, U, enthalpy
H, entropy, S, Gibbs’ free energy, G (also called the thermodynamic potential), and the
Helmholtz’ free energy, F. Together with their parameters of state – V, p, T, ni, these
functions describe precisely the state of a given system or process. For the purpose of
this book, let us choose Gibbs’ free energy G (the thermodynamic potential) for subse-
quent chapters of this book. As mentioned earlier, this function is the state function,
meaning that its change depends only on the initial and final state of the system.
From the mathematical point of view such extremely small change can be written as
the total differential versus the state parameters of Gibbs’ free energy: G = G(p, t,
ni ). Thus,




∂G ∂G ∂G
dG = dp + dT + Σ dni (1:1)
∂p T, n ∂T p, n ∂ni p, T
i i

Subscripts next to parentheses show that the remained parameters of state are constant.
Thus, the meaning of this total differential is that we can sum partial differential of G
versus p, keeping T and ni constant, controlling the change dp, and so on, and then sum
all partial differentials to get the overall change of G, dG. As long as we do not assign the
physicochemical meanings of all three partial differentials in the above equation, it re-
mains purely mathematical. However, taking into account the laws of thermodynamics,
we can identify the meanings of these partial differentials, rewriting the above equation
as follows:

dG = Vdp − SdT + Σμi dni , (1:2)

where V is volume, S – entropy, and μi is the chemical potential of species “i” in the
system under consideration.
The detailed arguments behind this transformation is beyond the scope of this
book. Interested reader is directed to the textbooks on thermodynamics.
Now, let us identify the system that is used in electrochemistry – the electrode. Typi-
cally, an ideal conductor (semiconductor electrodes will be discussed in Section 10.2) and
its properties do not depend on the bulk (volume), but surface behavior, where all of the
excess charges originating, for example, from the external polarization, are localized. An
experimentally measurable parameter that can be initially used for an interface with
zero electric charge is the so-called surface or interface tension. The latter is more ap-

https://doi.org/10.1515/9783111160986-001
4 1 Basic concepts

propriate, because there are no free surfaces, but surfaces separating two phases
in contact. For our purposes, we will use the term “surface tension” throughout
the text. Quantitatively, a work δw, required to increase a surface area by δA, is
given by the following equation:

δw = γdA, (1:3)

Proportionality coefficient γ [ J/m = N/m] is called the surface tension (interfacial

2

tension)
Thermodynamics of such an interface can be described with the help of the change
of selected state function like Gibbs’ free energy, which will depend only on the initial
and final state. For an open, multicomponent system of a surface area A, we can write:

G = Gðp, T, ni , AÞ, (1:4)

and its total differential:





∂G ∂G ∂G ∂G
dG = dp + dT + Σ dni + dA (1:5)
∂p T, n , A ∂T p, n , A ∂ni p, T, A ∂A p, T, n
i i i

or

dG = Vdp − SdT + Σμi dni+γ dA (1:6)

Therefore, at constant p, T, ni, we can obtain the thermodynamic definition of surface

tension:

dG = γ dA (1:7)

Until now we have considered the total value of Gibbs’ free energy change (thermody-
namic potential change) of a system with the surface. Now let us turn to the descrip-
tion of the surface (interface) itself. In doing this we will use the so-called Gibbs
model, in which the ideal surface has no volume and the two phases separated by this
surface are at equilibrium. This is illustrated in Fig. 1.1.

interface σ

Fig. 1.1: The Gibbs model of two contacting phases A and

B at equilibrium, represented by two slabs of different
A B colors and separated by an ideal surface σ of Vσ = 0. See
text for details.

All these regions are at equilibrium, and so chemical potentials of any component “i”
that can distribute itself between these three regions are also at equilibrium:

μAi = μBi = μσi = μi (1:8)

 1.1 Structure of interfaces 5

This model formally assumes that the total internal energy of such a system can be
described by a sum of selected constituents: phase A, phase B, and their contact sur-
face σ:

Utot = UA + UB + U σ (1:9)

The total volume is due solely to a sum of volumes A and B, Vσ = 0, so

Vtot = VA + VB (1:10)

Moreover, any component ni of the system partitions itself between phase A, phase B,
and surface σ. All these regions are at equilibrium, and the equilibrium condition is
fulfilled also by the equilibrium of chemical potentials of species “i” in the system:

μAi = μBi = μσi = μi . (1:11)

With the help of some thermodynamic relations that are beyond the scope of this book,
we can finally write the expression for the Gibbs’ free energy change of the surface σ:

dGσ = −Sσ dT − Adγ + Σμi dni σ (1:12)

This equation describes the equilibrium change of Gibbs’ free energy of the interface
(surface) separating two phases, relating it to the change of parameters T, γ, and ni. Set-
ting T = const and dividing by A to obtain the specific values per unit surface, we finally
get after some rearrangement:

dγ = −Σ Γσi dμi (1:13)

or, in a more convenient form for the case of adsorbing species on a solid surface:

Γ = −ða=RT Þδγ=δa (1:14)

where Γ stands for the surface concentration of adsorbing species [mol/m2], and a is
the activity of this species in the solution, often replaced by its bulk concentration,
c [mol/dm3]. What we just derived is the quantitative relationship between the amount
of a substance accumulated at the surface with its activity (or concentration) in the
bulk of the solution. This equation, known as Gibbs’ adsorption isotherm, plays a cru-
cial role in, for example, separation techniques and catalysis.
Now, having all necessary instruments provided by the thermodynamics of the
uncharged systems, it is time to utilize the above knowledge for the systems, in
which the surface can be charged in a controlled manner – the electrodes [1–6].
Most if not all of the areas of applied electrochemistry contain electrode/electrodes as
an integral part of the systems, determining the systems’ properties, applications, and
performance. Thus, let us introduce here the new state parameter – charge, q. At con-
stant temperature, by virtue of the approach described earlier in this chapter, we can
write:
6 1 Basic concepts

σ
σ ∂Gσ
dG = − Adγ + Σμi dni + γ, ni (1:15)
∂q

The last term describes Gibbs’ free energy (work at const. T), required for bringing a
charge q from infinity into the system.1 As we should remember from physics classes,
this is the definition of the electrical potential E. Therefore,

dGσ = − Adγ + Σμi dni σ + Edq (1:16)

Because simultaneously the integral form for Gibbs’ surface free energy is

Gσ = Σμi ni σ + E × q (1:17)

and therefore

dGσ = Σμi dni σ + Σni σ dμi + Edq + qdE, (1:18)

we finally get the relationship between the changes of surface tension, concentration
changes in the bulk of a solution, surface concentration of species “i” and the surface
charge of a metal electrode, σ Me:

dγ = −σ Me dE − ΣΓi dμi (1:19)

Before we go deeper into the models describing the structure of the interface between
the metallic electrode and electrolyte, we should place these efforts in a broader pic-
ture. First, we should stress that the metal/solution interface should be considered as
model interface, easy to control experimentally (e.g., by controlling the polarization of
the electrode), and the theory derived from the investigations of such an interface
should be useful also for other types of interfaces, including the nanostructured interfa-
ces and biointerfaces. But why the interfaces should be charged, is this process a spon-
taneous one? Yes, it is spontaneous, except for the case of external polarization. The
reasons for the appearance of a charge on the interfaces can be summarized below:
– external polarization (electrodes);
– adsorption and orientation of solvent dipoles;
– ionic adsorption from the solution;
– dissociation of surface functional groups
Of course, all these effects can occur simultaneously.

Essentially, one can say that there are no uncharged interfaces. The consequences of this
charge will be manifested in chapters that follow. Here, we will present the model of
metal/solution interface that allows us to get insight into the structure of the interfacial

 Since our charged phase (metal electrode) behaves like an ideal conductor, this charge will be accu-
mulated on its surface.
 1.1 Structure of interfaces 7

region, but more important from the point of view of applications – its utilization in
the construction and design of modern power sources and electrode processes.

1.1.1 Electrical double layer at interfaces: metal/electrolyte

What is the electrical double layer (e.d.l.)? It is a region of molecular dimensions at the
interface between the two phases/substances, in which the electric field is created (e.g.,
by polarizing an electrode or by dissociation of surface groups). Of course, both phases in
contact have to contain charged species (electrons, ions, or polar molecules).
In the e.d.l., charges of opposite sign attract themselves and have the tendency to ac-
cumulate at the interface. Finite sizes and solvation result in the separation of these par-
ticles. Charges of the same sign repel themselves leading to an unequal distribution of
opposite charges with respect to the charges of the same sign within the interfacial re-
gion. Taken together, these interactions generate the electric field in the interfacial region,
regardless of the nature of two phases in contact (with the proviso discussed above).
In order to describe quantitatively the profiles of potential and the distribution of
charges in the interfacial region, we will take advantage of the relation introduced
above for the metal/solution interface (eq. 1.19). For the sake of simplicity, let us as-
sume that we do not change the electrolyte concentration nor its content so that dμi = 0.
Then, the above equation can be simplified to

dγ = −σ Me dE (1:20)

Let us now focus on the model electrified interface of ideally polarizable electrode/aque-
ous electrolyte solution. By an ideally polarizable electrode, we will understand an elec-
trode that within a given potential range can only accumulate charge (being polarized
or charged) with no redox reaction (electron transfer across the interface). Changing its
polarization by dE, we also change the interfacial tension by dγ, with proportionality co-
efficient being the charge σMe, introduced onto the electrode surface (ideal conductor) as
a result of its polarization. The negative sign tells us immediately that the interfacial ten-
sion decreases with an introduction of charge onto the surface. This is intuitively under-
standable, because the charges of the same sign on the surface repel each other with
Coulomb forces, thus decreasing the cohesion interactions between the surface atoms.
Furthermore, one can predict that at zero surface charge, the surface tension will be the
largest. Such interfacial behavior has its further consequences, for example, in the elec-
trocatalysis (Chapter 4), corrosion (Chapter 3), or in the underpotential deposition (Chap-
ter 6), and will be discussed in the appropriate chapters. Here, we will continue to
develop a model describing the interfacial property responsible for charge storage and
electrical power generation.
8 1 Basic concepts

The above equation can be rearranged, so we can get the following:

dγ=dE = −σ Me (1:21)

So, if we know the experimental dependence of γ versus E, from the slope of this de-
pendence E = f(γ) we can obtain the value of surface charge σMe of a metallic surface
for each value of the applied potential E. The best known and thoroughly experimen-
tally elaborated dependence of γ versus E, also known as the electrocapillary curve, is
that for the interface of mercury/aqueous electrolyte solution. However, other interfa-
ces were also thoroughly investigated. These include air/solution interface or the two
immiscible electrolyte solutions. All such interfaces were investigated not only from
the point of view of basic research in order to achieve some insight into the structure
of various interfacial systems, but also from the point of view of applications in the
adsorption and separation techniques. The shape of the electrocapillary curve can
be drawn as in Fig. 1.2(a).

(a) (b) (c)

ɣ σM cedl

Epzc E Epzc E Epzc E

Fig. 1.2: Graphs representing: (a) the electrocapillary curve γ versus polarizing potential E,
(b) surface charge σ, and (c) differential capacitance, Ce.d.l.. Please note the maximum on the
electrocapillary curve corresponding to the potential of zero surface charge, Epzc (inflection point on curve
b and to the minimum on curve c).

Second derivative of the equation dγ/dE = – σMe at constant chemical potential µ

yields

d2 γ=dE2 = −dσ Me =dE = Ce.d.l. (1:22)

where Ce.d.l. is the differential capacitance of the electrical double layer formed in the
interfacial region. The shape of the relations σMe versus E and Ce.d.l. versus E is shown
in Fig. 1.2(b) and (c), respectively. Now we see that the interface behaves as a system
that stores the electric charge. The corresponding device is well known from electron-
ics – a dielectric capacitor, in which the charge accumulated on its plates results in a
potential gradient between these plates. Also the change of surface charge density by a
factor dσMe will result in a change of potential drop by dE (and vice versa), proportional
to the capacity.
 1.1 Structure of interfaces 9

In the case of metal/solution interface, one of the condenser plate is easily identi-
fied – this is the electrode, whereas the description of a second “plate” that is formed by
the ions of the opposite sign that are present in the solution depends on the model of
the electrical double layer mentioned above. One condition, regardless of the accepted
model, has to be fulfilled: all the interfacial region has to be electrically neutral, meaning
that the charges on the metal surface (i.e., electrons) are compensated by the charges in
the solution part of the e.d.l. (i.e., ions). This electroneutrality condition reads:

− σ Me = σ s ðsuperscript “s” stands for “solution”Þ (1:23)

According to the simplest model proposed by Helmholtz (1879), in the solution the
ions of the opposite charges with respect to the charge of an electrode (counterions)
adsorb electrostatically via the Coulomb forces, directly on the electrode surface, neu-
tralizing its charge. They form firm so-called Helmholtz’ layer, with linear potential
drop within its space, normal to the electrode surface. This is shown schematically in
Fig. 1.3. Obviously, the electric field within such interfacial region is limited to the
thickness this layer. Beyond it, inside the solution there is no electric field. Also, since
we treat the metal as an ideal conductor, within bulk of the metal there is no electric
field, too. This model corresponds directly to the flat capacitor.

Metal Solution

–
+ qH = –qs
– Outer CH,specific = εε0/d
Helmholtz’
– ε0 = 8.8*10–12 F/m
Ψ plane
+ ε = 78 for water
– d = 2*10–10m – radius of adsorbed ions
(solvated or not)
– Cedl~340 *10–6F/cm2 – experimentally
+ ~20–40*10–6 F/cm2
–

OHP OHP x

Fig. 1.3: The Helmholtz model of the electrical double layer with solvent (water) molecules and solvated
ions at their closest approach to the electrode surface, forming the so-called outer Helmholtz plane (OHP).

This model predicts that the capacitance should be independent of the concentration of
ions and electrode polarization (charge density). As it is found experimentally, it describes
relatively well the electrical double layer behavior for large concentrations and electrode
polarization.
Accounting for the thermal motions of ions in the electrolyte led independently
L.G. Gouy and D.L. Chapman (1920) to the so-called diffuse electrical double layer. Due
to the thermal motions, the ions cannot form the compact, firm layer, but are dif-
10 1 Basic concepts

fused, dispersed out of the interface forming layer of ions with exponentially chang-
ing concentration as a function of distance. These thermal motions counteract and
overcome the organizing force of Coulomb interactions. Other assumptions underly-
ing the Gouy–Chapman model are listed as follows:
– The ions are point charges (no dimensions).
– Solvent – dielectric continuum, characterized by electric permittivity ε (no molecu-
lar structure).
– Charge distribution in the volume dV in the radius dr from the interface is given
by the Maxwell–Boltzmann distribution.
– The thermal motions are stronger than the electrostatic forces.
– The relationship between the space charge density ρ and the potential distribu-
tion in this region is given by the Poisson equation:

d2 ψ ρðxÞ
=− (1:24)
dx2 εε0
X
ρð x Þ = F zi c i ð x Þ (1:25)
i

For an infinite, flat, ideally polarizable electrode, the potential profile along the nor-
mal to the electrode into the bulk of the electrolyte, ψ(x), is given as follows:

ψðxÞ = ψ0 e − κx (1:26)

where ψ0 is the electrical potential at the surface and κ−1 = λD is called the Debye
screening length (Debye thickness), at which the potential ψ0 decreases e-times (orange
line in the inset in Fig. 1.4). Practically, beyond this distance, the effect of electrode
polarization is negligible, screened by the counterions; however, within the e.d.l.
region, the counterions are concentrated, while the coions diluted.

Ψ(x) For: 0.1 M KCI λD = 0.96 nm

Ion concentration

H2O λD = 680 nm, in practice: ca.100 nm

Ψ0 Counterions Blood plasma: λD = 0.8 nm (ca.)

Coions
x

Distance from the charged x

surface (e.g., electrode)

Fig. 1.4: Potential profile (distribution) along the normal from the electrode surface, according to the
Gouy–Chapman model. Inset shows the coions and counterions distribution according to this model. The
orange line marks the Debye thickness (length).
 1.1 Structure of interfaces 11

This distance strongly depends upon the ionic concentration, dielectric permittivity of
the solvent and valency of ions according to the following:

− 0.5
e2 X 0 2
λD = c z ½m (1:27)
εε0 kB T i i

where ρðxÞ, volume charge density of ion “i” at a distance x [C m−3]; ψ(x), the electrical
potential at distance x [V]; ψH , the electrical potential at Helmholtz layer [V]; ε, rela-
tive dielectric permittivity [no units]; ε0 , dielectric permittivity of vacuum [J−1 C2 m−1];
z, valency of ion; e, elementary charge [C]; kB, Boltzmann constant [J K−1]; T, tempera-
ture [K]; F, Faraday constant [C mol−1].
The charge density at the surface is related to the potential at x = 0 by


dψ
σ = −εε0
Me
(1:28)
dx x=0


2κεε0 kB T zeψðxÞ
σðxÞ ≈ sinh (1:29)
ze 2kB T

where σðxÞ is the surface charge density at a distance x [C m−2].

Finally, the recognition that joint models of Helmholtz and Gouy–Chapman
explain better the observed experimental data, such as the behavior of the e.d.l. as
capacitor, led O. Stern to the following (Fig. 1.5):
– e.d.l. contains two parts (but this is not the reason why it is called “double,” we will
discuss this later);
– Helmholtz layer – ca. of thickness of adsorbed ions or polar molecules, essentially
bound to the surface, linear potential decay;
– Gouy–Chapman layer – from the Helmholtz layer to the bulk, diffuse, with expo-
nential potential decay.

It is necessary to remind at this point that, regardless of the model of the e.d.l. potential
profile, its value in the solution bulk (ψS) always equals zero. Now the interface has two
regions in series that can accumulate charge: the Helmholtz layer with linear potential
drop and the Gouy–Chapman layer with exponential potential decay. Thus, it can be
represented as the two capacitors in series (Fig. 1.6).
As it is well known from physics, the inverse of total capacitance of the two ca-
pacitors in series equals to
1 1 1
= + (1:30)
Ce.d.l. CH CG−Ch

Now, let us divide formally the overall potential decay within the e.d.l., ψMe – ψS into
two parts: within the Helmholtz layer ψMe – ψH, and within the Gouy–Chapman layer
(ψH – ψS); thus, ψMe – ψS = (ψMe – ψH) – (ψH – ψS); keeping in mind that ψS = 0.
12 1 Basic concepts

Metal Solution

ΨMe

ΨH

ΨS

OHP OHP x

Fig. 1.5: The Stern model of e.d.l., being a joint model of Helmholtz and Gouy–Chapman models.

CH CG-Ch

Fig. 1.6: Two capacitors in series representing two separate regions of charge aggregation and
separation: the Helmholtz and Gouy–Chapman regions. This approach explains the experimental results
in Fig. 1.7.

Then

Δψe.d.l. = ΔψH + ΔψG−Ch (1:31)

Dividing this last relation by a total charge on a surface and calling upon the definition of
a capacitor, we can get the relation for two capacitors in series, as described earlier:

∂Δψe.d.l. ∂ΔψH ∂ΔψG−Ch

= + (1:32)
∂σ Me ∂σ Me ∂σ Me

or
1 1 1
= + (1:33)
Ce.d.l. CH CG−Ch

The first term describes the differential capacitance of Helmholtz layer (if present)
with a linear potential decay, generally independent of the ionic concentration,
whereas the second is the capacitance of Gouy–Chapman layer with an exponential
potential decay, being strongly affected by the concentration of ions (Fig. 1.7), as
shown in eqs. (1.27)—(1.29).
 1.1 Structure of interfaces 13

CG-Ch

0.01 M
C (F/m2)

0.001 M

CH 0.1 M

0.01 M
0.001 M

0.0 E (V)

Fig. 1.7: Experimental data for differential capacitance on Hg electrode at various concentrations of
aqueous 1:1 electrolyte (e.g., KCl, solid lines) and predictions of Gouy–Chapman theory (dotted lines).
Green frame – the Gouy–Chapman region, blue oval – the Helmholtz region. (Adapted from: D.C. Graham,
Chem. Rev. 41 1947 441–501).

The total differential capacitance, Ce.d.l., is a very important quantity allowing to ex-
perimentally verify our model and subsequently utilize it in the designing of new
charge storage devices, such as the e.d.l. supercapacitors. Therefore, let us evaluate its
specific value per unit area, assuming some real quantities characteristic for aqueous
electrolyte solutions. For this purpose we model the CH as a plate capacitor formed by
the metallic surface and hydrated ions adsorbed on it. Assuming the radius of such
ions as ca. 2 Å, the capacitance of this layer would be CH = εHε0/rion. The question
under continuous dispute is: how to estimate the value of the dielectric permittivity
of the Helmholtz layer, εH. For the purpose of this example, let us assume its value
equal to 10. This in turn yields the value of 0.45 Fm−2 as the estimated capacitance CH
(45 µFcm2), whereas the experimental values are typically 20–150 µFcm2. The same
time, the values of CG−Ch depend on the concentration, varying from 7.2 µFcm2 for
1 mM to ca. 200 µFcm2 for 1 M monovalent aqueous electrolytes (the experimental val-
ues for the case of mercury electrode are 6.0 µFcm2 to ca. 23 µFcm2, respectively).
Regardless of these discrepancies, this brief insight discussed above explains why
the theory of the electrical double layer has found its applications in designing the
charge storage, energy conversion, and power devices for microelectronics. The e.d.l.
applications are not limited to microelectronics, but also in analytical separation, recti-
fication, and miniaturized field analytical devices, nanofluidics, leading finally to the
so-called lab-on-a-chip, in which phenomena related to the properties of e.d.l. play cru-
cial role in their operational capabilities and performance. Below we will exemplify the
role of the e.d.l. for the field-effect control in nanofluidic transistors [7]. This example in
14 1 Basic concepts

the cited work shows the modulation of channel conductance that mimics the behavior
of classical metal-oxide semiconductor field effect transistors (MOSFET). MOSFETS’ typi-
cal applications are electronic switches, amplifiers, logic gates, and so on. In the discussed
example, where the e.d.l. is put to work, under a certain source–drain bias, “gate”
voltage induces concentration gradients switching directions following changes in
gate-voltage polarity. Another example [8] shows the rectifying behavior upon the
ionic conductance of nanochannels with charged walls. This time the negative
charges originate from the functional groups present on the channel walls. At suffi-
ciently small channel diameter, the electrical double layers formed on both walls
start to overlap, repelling the coions (in the cited case – Cl−) from the channel and
allowing only the passage of counterions (in this case – K+ ions) [7, 8] through the
channel, making it highly selective.

K+
K+
–
Cl
Cl–

Nanochannel Microchannel

Fig. 1.8: Surface charge and e.d.l. effects in nanochannels and microchannels. Left panel:
A scheme of two monolayers overlapping in a nanochannel of negatively charged walls with sufficiently
narrow diameter, smaller than two Debye lengths of the e.d.l. In this case a rectifying behavior is
observed as shown in the scheme with the red and green arrows for Cl− and K+ ions conductance,
respectively. The reason of this behavior is the coions and counterions distribution in the e.d.l. electric
field (see inset in Fig. 1.4). Right panel: Both ions freely moving through the microchannel that is either
uncharged or much wider that the two Debye lengths.

The theory of e.d.l. is also useful in gaining insights into the surface properties of
nanomaterials, such as colloidal suspensions, where surface charge can result either
from the adsorption of ions or ionizable molecules stabilizing such suspension via the
Coulomb repulsion forces, or the charge can originate from the structural surface
functionalities (e.g., carboxylic groups), being inherent part of chemistry of nanopar-
ticles. The first case, namely, the e.d.l. formation due to the charges adsorbing on a
colloidal nanoparticle, is illustrated in Fig. 1.9.

1.1.2 Electrochemical potential – potentials at interfaces: internal, surface,

external potential

The electrostatic potential inside a given phase, called the inner potential, φ, is defined
by the work, we, performed by external forces, required to bring an electric, positive
 1.1 Structure of interfaces 15

(a) (b)
Ψ Ψ

Ψ0 Ψ0
ζ

λD d d
λD
ζ

Fig. 1.9: e.d.l. structure according to Stern and potential distribution in the e.d.l., for (a) colloidal particle,
positively charged and suspended in an electrolyte, (b) the same particle, positively charged but with
strongly adsorbed negative ions (e.g., halogen ions); their adsorption changes the resultant surface
charge (σ) and zeta potential (ς) of the nanoparticle, measured at the so-called plane of shear, appearing
only when the nanoparticle moves in relation to the suspending medium.

test charge qt from infinity and from a vacuum into the bulk of this phase, far from
the phase boundary, divided by the value of this charge:
we
φ= (1:34)
qt

Please note that the units of electrostatic potential are Volts [V], because J/C = V.
The assumption underlying this process is that the test charge is sufficiently
small, so it does not affect the electrostatic field associated with this phase, and the
work performed is only that necessary to overcome the electrostatic forces. All other
forces, such as the chemical forces, remain unaffected during the process.
By analogy to the electrostatic potential, the chemical potential for particles “i”
inside the same phase is defined as the work, wch, performed against all chemical
forces during the transfer of ni moles of particles “i” into such phase, divided by the
number ni of these particles. Again, the assumption is made that this virtual transfer
does not affect the total concentration of these particles:
wch
μi = (1:35)
ni

Note that, in this case, the units of µi are Joules [J] and if ni corresponds to the number
of moles, then the units of a chemical potentials are J/moles.
The sum of chemical and electrostatic energies of charged particles “i” in a given
phase (per mole, with charge ziF) is called the electrochemical potential of species “i”:

μeι = μi + zi Fφ (1:36)

here F is the Faraday constant and zi is the valency of ion.

16 1 Basic concepts

Figure 1.10(a) shows two metals Me1 and Me2 (treated as ideal conductors) brought
into contact and at equilibrium, pictured by the arrow pointing in two opposite
directions.

(a) (+, ∞ vacuum) (b)

φ1 Ψ1 Ψ2 φ2

Vacuum

χ1 χ2 Metal Surface
Me1 Me2

Fig. 1.10: (a) Two metal slabs at equilibrium, illustrating the definitions of inner, surface, and outer
potentials. (b) This intends to explain the “image” forces appearing at close approach of a charge to the
surface of an ideal conductor. See text for details.

The only charged components of these two metals that can partition between the two
phases, achieving finally an equilibrium are the electrons (the nuclei do not mix under
ambient conditions, preserving the identities of the two different metals). We now can
use the definition of the electrochemical potential to quantitatively formulate this equilib-
rium as the state where the electrochemical potentials of electrons are the same in the
two contacting metals:

μee ð1Þ = μee ð2Þ (1:37)

We return to the consequences of eq. (1.37) later in the text. Now, let us focus on the
inner (internal) potential φ. The transport of the test charge from infinity in vacuum
can take place formally in two steps, shown in Fig. 1.10(a). First, the charge is brought
close to the interface (red dot in Fig. 1.10(a)). The potential defined by this work is
called the external potential ψ. (Please note that while discussing the electrical double
layer we also used the symbol ψ. In this case, however, the reference point is not
in vacuum but in the bulk electrolyte solution, far away from the interface, where
ψS = ψ(∞) = 0). Since the point of close proximity and the reference point in infinity
are located in the same phase (in this case: the vacuum), the external potential ψ
can be measured.
The statement “close to the interface” needs to be defined more precisely. When
the charge is being moved from infinity toward the surface of charged conductor, the
work is performed, primary due to the electrostatic field produced by the charge of the
 1.1 Structure of interfaces 17

conductor. As the test charge approaches the surface, image forces begin to affect the
test charge, distorting also the electrostatic field produced by the conductor. The term
“image forces” reflects the induction of an image of the test charge, negative in sign,
because the approach of positive test charge induces an accumulation of electrons at
the surface of conductor due to Coulomb interactions, forming an image charge of op-
posite sign. Thus, the surface of ideal conductor acts as a “mirror,” reflecting the ap-
proaching charge (Fig. 1.10(b)). Such induction creates additional dipolar potential drop
across the surface affecting the original field of the conductor. Calculations of the cutoff
distance of the image forces between the surface and the test charge give the value of
such distance of ca. 1 μm.
In the next step the charge is brought through the interface and into the phase inte-
rior. This step defines the so-called surface potential χ. It is determined by surface
charges and by dipoles aligned at the interface of the conductor. So finally we have:

φ=ψ+χ (1:38)

As can be immediately seen, the surface potential χ defined across the boundary (sur-
face) can be related to the electrical double layer potential drop. One has to keep in
mind, however, that the definitions of φ, ψ, and χ were given for an ideal conductor in
contact with vacuum (ideal dielectric), whereas the e.d.l. models describe the interfaces
between the two phases containing freely moving charges (ions, electrons, holes). In
contrast to the external potential, the values of surface potential and inner potentials
refer to points in different phases, and therefore cannot be measured.
Let us now return to eq. (1.37). Taking into account eq. (1.36), the equilibrium con-
ditions for electrons in contacting metals Me1 and Me2 can be written as follows:

μe, 1 − Fφ1 = μe, 2 − Fφ2 (1:39)

where μe, 1 and μe, 2 denote the chemical potentials of electrons in phase 1 and 2, re-
spectively, a minus sign is due to the negative charge of electron (ze = −1). After re-
grouping, we get

Δφ = φ2 − φ1 = ðμe, 2 − μe, 1 Þ=F (1:40)

The potential difference, Δφ, between the two phases defined by this equation is called
the Galvani potential (the term recommended by IUPAC) and cannot be measured di-
rectly, because it refers to the points in different phases; in the case of Fig. 1.10(a), these
are the Me1 and Me2. Taking into account eq. (1.38), we can rewrite the eq. (1.40) in a
form, showing its two components:

Δφ = φ2 − φ1 = ðψ2 + χ2 Þ − ðψ1 + χ1 Þ (1:41)

or
18 1 Basic concepts

Δφ = Δψ + Δχ (1:42)

Δψ is called the Volta potential, whereas Δχ, the surface potential difference, is fre-
quently called the surface potential. So, how to obtain the value of Δφ between the two
metals Me2 and Me1? We already said that Δψ is measurable, but how to account for Δχ
– the surface potential difference between the two inherently different metals (phases)?
Let us play a trick and add one more set of interfaces, adding the same metal Me1 on
the other end of metal Me2 as in Fig. 1.11. For clarity in the description that follows, let
us mark it as Me1’. This Me1’ possesses the same surface properties as Me1; there-
fore, χ1 = χ01 = χ.

χ1 χ2 χ1

Fig. 1.11: Slab model representation of a system

Me1 Me2 Me’1
enabling measurements of the Galvani potentials.

So now, at equilibrium, we will have:

μe, 1 − Fφ1 = μe, 2 − Fφ2 and μe, 2 − Fφ2 = μ0e, 1 − Fφ01 , but also μe, 1 − Fφ1 = μ0 e, 1 − Fφ01

Therefore, the value of φ2 – φ1 can be expressed in terms of φ01 − φ1 :

μ0e, 1 − μe, 1
φ01 − φ1 = (1:43)
F
Let us now use eq. (1.41) and we obtain:

φ01 − φ1 = ψ01 + χ01 − ðψ1 + χ1 Þ (1:44)

When Me1 and Me1ʹ are the same metals, we can easily assume that their surface prop-
erties are the same. Therefore, χ01 = χ1 and under such conditions:

φ01 − φ1 = ψ0 1 − ψ1 (1:45)

the Galvani ðφ01 − φ1 Þ potential is equal to Volta potential ðψ01 − ψ1 Þ. Moreover, being
localized in the same phase it is now is measurable.

1.1.2.1 Metal–solution interface at equilibrium: Nernst equilibrium potential

We are now ready to apply the above information to the system commonly found in
electrochemistry: the interface between a metal (electronic conductor) and electrolyte
solution (ionic conductor). For the sake of clarity, we introduce here the simplest pos-
sible system that is the metal at equilibrium with aqueous solution of its ions, such as
Cu|Cu2+ or Zn|Zn2+. The counterions in the solution (anions) can be of any soluble
salts of metal under consideration (e.g., SO42−, NO3−). Let us rewrite eq. (1.37) describ-
 1.1 Structure of interfaces 19

ing the equilibrium condition for electrons in two contacting metals, in order to de-
scribe the equilibrium condition of Mez+ in the solution with the same metal cation
inside the metal phase:

Mz + ðsolutionÞ $ Mz + ðmetalÞ (1:46)

~ðMz + ÞS = μ
μ ~ðMz + ÞM (1:47)

or

μðM z + ÞM + z+FφM = μðMz + ÞS + z+FφS (1:48)

After regrouping, we have

φM − φS = ΔφM=S = ½μðM z + ÞS − μðM z + ÞM =z+F (1:49)

Since μðM z + ÞS = μ0 ðM z + ÞS + RT ln aMz + , and the chemical potential of metal cation in

its metal phase defines its standard chemical potential: µ(Mz+)M = µ0(Mz+)M, we can
finally write:

ΔφM=S = ½μ0 ðM z + ÞM − μ0 ðM z + ÞS =z+F + ðRT=z+FÞlnaMz + (1:50)

The internal potential difference ΔφM/S between metal (M) and solution (S) is called the
Galvani potential, as was discussed above, whereas the difference between the standard
chemical potentials (square brackets in eq. (1.50)) is the change in standard molar Gibbs
free energy, ΔrG0, of redox reaction eq. (1.46). This equation, in a form,

RT
Eeq = E0 + lnaM z + (1:51)
nF
is known as the Nernst equation, or electrode equilibrium potential, Eeq. Similar con-
siderations lead to a more general equation, describing the Nernst equilibrium poten-
tial for redox couple of activities aox and ared for oxidized and reduced forms,
respectively, exchanging only electrons with metal electrode. The latter does not par-
ticipate chemically in this process and is sometimes called the redox electrode. In
these cases, the Nernst equilibrium potential will read:
RT aox
Eeq = E0 + ln (1:52)
nF ared

Equation (1.52) shows that the Nernst equilibrium potential is equivalent to Galvani
potential (internal potential difference) of metal/solution interface at equilibrium.
It depends on the activities of ions participating in the electrode processes. The activ-
ity of ion “i” in the solution, ai, is related to its concentration, ci, via the activity coeffi-
cient, γi: ai = γici, and so, eq. (1.52) can be written as follows:
20 1 Basic concepts

RT cox
Eeq = Ef + ln (1:53)
nF cred

where Ef is the so-called formal potential accounting for the activity coefficients (the
effect of interactions of solution components on the standard electrode potential E0),
and as such it does not have thermodynamic meaning, contrary to E0, being equal to:

Δr G0
E0 = − (1:54)
zF
As we discussed earlier, the absolute values of Galvani potential or Nernst potential
are not accessible experimentally. However, they can be assessed in relation to the
reference electrodes. This will be addressed in Chapter 8.

1.1.2.2 Electron work function

A wealth of information on surface properties of conducting or semiconducting materi-
als in modern applied electrochemistry and electronics is needed for pursuing the tech-
nological development of various devices and processes. One of such information is the
electron work function, or simply the work function, Φ. It is defined as the minimum
amount of work that has to be performed by the external forces in extracting one elec-
tron from a metal, across its surface and into vacuum, where it becomes a free electron
with no kinetic energy. It is worth to realize that the work functions of all metals are
positive; otherwise the electrons would escape from the material spontaneously. Be-
cause of the effect of surface properties, the work function depends on the position of
the point to which the electron is transferred. We addressed this problem defining the
internal, external, and surface potential, and as previously, the point in the vacuum,
“just outside” the metal, is regarded as the final point of electron transfer. For such con-
ditions, the work function Φ is equal to the difference in electrochemical potentials of
the electron at this point in vacuum, μ ~0 , and any point in the metal μ~M . Since chemical
potential of electron in the vacuum is equal zero: μe = 0, therefore the electrochemical
0

potential of an electron in vacuum e μ0e = −eψ. Minus sign in this relation is due to the
negative valency of electron. Therefore,
 M 
Φ=e μ0e − e
μM
e = −eψ − μe − eφ = −μe + eχ
M M
(1:55)

The same relation can be obtained directly from the definition of the electrochemical
potential of the electron in a metal:

~M
μ M M M M M M M
e = μe − e φ = μe − e ðψ + χ Þ = ðμe − eχ Þ − e ψ = α − e ψ
M M M
(1:56)

According to the above equation,

e +e·χ
Φ = −α = −μM M
(1:57)
 1.1 Structure of interfaces 21

and

α= −Φ (1:58)

As we stated above, Φ, depends on the material and the state of its surface (including
crystallographic structure). The following table exemplifies the values of work func-
tions for several metals. Since this is the energy required to withdraw the electron, its
units are also the energy units in eV.

Pt . – . eV (or . for Pt())

Au .–. eV

Ag .– eV

Pd .–. (Pd()) eV

W .–. eV

Ni .–. eV

The measurements of work functions of conducting and semiconducting materials

find their applications in various areas of applied science, for example, electrocatalysis
(cf. Chapter 4), materials characterization, semiconductor science, polymer science,
electrical characterizations, nanostructured materials, soft samples, and even bioscien-
ces; various techniques have been developed to obtain such information. These include
the photoemission spectroscopy, measurements of current induced by electron beam,
scanning electron microscopy, and Kelvin probe/vibrating plate capacitor techniques.
Interested reader is directed to the review work, describing the latter electrochemical
technique [9].

1.1.3 Charge transfer processes across the metal/electrolyte interface

1.1.3.1 Basic concepts of nonequilibrium thermodynamics

The previous paragraph provided some basic thermodynamic definitions and concepts
for the systems at equilibrium. Even in the case of the e.d.l. formation, one can easily
assume that from the point of view of other processes of interest in the applied electro-
chemistry, the structure of the interface is attained so fast that can be considered as
being at “equilibrium.” This will be shown in the text below. However, most of the elec-
trochemical processes that are currently employed or observed in the energy storage
and transformations, electrodeposition, corrosion, or photoelectrochemistry are related
to the transport of charge, mass, and energy (e.g., heat), and, as long as this transport
continues, we cannot describe it using the equilibrium formalism [10]. Or, perhaps we
can, modifying appropriately such formalism?
22 1 Basic concepts

Let us summarize briefly the systems at equilibrium:

– It is characterized by uniform values of intensive state parameters and “poten-
~ ; in other words, all intensive state
tials” throughout the system, such as p, T, µ, μ
parameters (X) are constant and their gradients are zero: dX dt = 0, ∇X = 0.
– Densities of conjugate extensive quantities, for example, V, Ni, S, can be inhomo-
geneous at equilibrium (here: Ni and S are the number of molecules of species “i”
and the entropy of the system, respectively).
– S (entropy) is maximized at constant internal energy U of the system.
Q
– For the closed system di S = dS − Tel , where diS is called the entropy production,
and Qel is the heat exchange between the system and its surroundings during
the equilibrium process. It is immediately evident that at equilibrium where
the second law of thermodynamics is fulfilled, diS = 0.

Generally, thermodynamics describes what cannot happen (the first and the second
law of thermodynamics). But how to apply its concepts for systems that are not at
equilibrium?
Irreversible thermodynamics applies the principles of thermodynamics for
nonequilibrium systems and gives the relaxation rules for a system to achieve the
equilibrium state or stationary state. Let us introduce two terms that are basic for the
formalism that follows: the flux Ji of energy, current, mass, and so on and the ther-
modynamic force Xi. Its impact results in the flux J. For the case of mass flux,
Jm = dm
dt · A. The fluxes are vectors, whereas the thermodynamic forces can be either
1

vectors or scalars. If there is only one thermodynamic force responsible for a given
flux, it is called the conjugated force. Before we put all the above at work, let us for-
mulate below several postulates that will be useful for further discussion.

Postulate 1
Generally, the beauty of thermodynamics lays in the fact that it is possible to apply
the state functions defined for reversible systems and to use principles of continuum
limit to define meaningful, useful, local values of various thermodynamic quantities,
for example, chemical potential μi. Therefore, useful theories can be derived by as-
suming a functional relation between the rate of a process and the local deviation
from equilibrium (the “driving force”).

Postulate 2
For any irreversible process, the rate of entropy production is everywhere ≥ 0

di S
≥0 (1:59)
dt
 1.1 Structure of interfaces 23

Postulate 3
For small perturbations we can assume a linear coupling between fluxes Ji and ther-
modynamic forces Xi:

Ji = Lii Xi (1:60)

Here, “i” or subscripts denote the different types of fluxes, for example, heat Q, charge
q, current I, and so on, and Lii are the so-called Onsager phenomenological coeffi-
cients. Now, if there are multiple fluxes in the system and each one is somehow de-
pendent on the other, then for a selected flux “i” we can write:
X
Ji = Lii Xi Xj (1:61)
i, j

Developing the above for many fluxes,

Ji = Lii Xi + Lij Xj + ... + Lin Xn (1:62)

Jj = Lji Xi + Ljj Xj + ... + Ljn Xn (1:63)

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..
Jn = Lni Xi + Lnj Xj + ... + Lnn Xn (1:64)

or
Xn
Ji = Lii Xi + k=1 Ljk Xk (1:65)
k≠i

For a better understanding of this approach, let us consider only two fluxes: the flux
of charge (current) and the flux of heat. It is a common knowledge that when a cur-
rent flows through a conductor (wire), this conductor warms up in a way related to
the magnitude of current: the larger the current, the warmer the wire. In the above
formalism, we can describe these two interrelated fluxes as

Jq = Lqq Xq + LqQ XQ (1:66)

JQ = LQq Xq + LQQ XQ (1:67)

Here Lqq and LQQ are called the direct coefficients, because they conjugate the thermody-
namic force directly to the flux it causes, whereas LqQ and LQq are the coupling coeffi-
cients because they indirectly couple all other possible fluxes with the selected single
flux. In our case the flux of current is coupled to the flux of heat and vice versa: if one
heats up the metallic conductor, its resistivity increases, and so the flux of current de-
creases. The latter observation leads us directly to Onsager symmetry postulate.
24 1 Basic concepts

Postulate 4
Onsager symmetry postulate: Lij = Lji (microscopic reversibility). This postulate is
equivalent to
∂Jq ∂JQ
= (1:68)
∂XQ ∂Xq

Table below summarizes some examples of fluxes and their conjugated thermody-
namic forces.

Process Flux J Conjugated force Empirical law

Energy (heat) transport


1 dQ 1 ∇T JQ = −K∇T Fourier
JQ = ∇ =− 2
A dt T T

Charge transport 1 dq − ∇φ Jq = −ρ∇φ Ohm

Jq =
A dt
Mass transport 1 dni − ∇μi ðat constant TÞ Ji = −Mi ci ∇μ i Fick
Ji =
A dt

Let us now put all the above together for some “edible” applications in electrochemis-
try. First, we focus only on mass flux of charged species “i,” identifying it as indepen-
dent of other possible fluxes and assigning to it the thermodynamic driving force Xi
equal to the electrochemical potential gradient of this species. Moreover, the empirical
Onsager proportionality coefficient Li equals to the product of concentration of species
“i”, ci, and its mobility, ui, defined as the ability of charged particles to move through a
medium in response to an electric field. Thus, we can write

Ji = Li · Xi
~i ;
Xi = −gradμ Li = ui · ci (1:69)
~i
Ji = −ui · ci · gradμ

Minus sign in the description of the thermodynamic driving force is due to the inher-
ent mathematical definition of a gradient which is positive from “low” values to
“high” values, whereas flux J occurs from “high” to “low.”
Assuming only one direction of flux along the “x” axis and substituting e
μi with its
definition (eq. 1.36), we will get (at uniform, constant temperature throughout the
system):
 
d ln ci dφ
Ji = − ui ci RT + zi F (1:70)
dx dx

where all symbols have the same meaning as before. This equation, called the Nernst–
Planck equation for flux, can be rewritten as follows:
 1.1 Structure of interfaces 25

dci dφ
Ji = − ui RT + ui ci zi F (1:71)
dx dx

This is one of the fundamental equations that can be applied to various areas of elec-
trochemistry where the movement of charges (or mass) is involved. To show the use-
fulness of the Nernst–Planck equation for solving various problems, we now put it
under the test, setting forth different limiting conditions describing different pro-
cesses in electrochemistry.
1. Let us assume that there is no electric potential gradient in our system, that is,
dϕ
dx = 0. This condition describes the situation, where the charged species (e.g., ion) is
far away from the electrified interface (e.g., electrode). This is also a situation where
the ions are beyond the thickness of the e.d.l. discussed in the previous section.
Then, if there is still transport of the charged species, under the isothermal condi-
tions, this transport (flux) will be described by

dci dci
Ji = −ui RT = −Di (1:72)
dx dx

where Di is the diffusion coefficient of species “i.” If we consider 1 mole of

charged species, then uiRT = ziFD
So, from the Nernst–Planck equation, we obtained the well-known Fick’s law for
mass transport due to the concentration gradient. Such systems work under diffu-
sion control.
But what we could get if there is no concentration gradient, but these are the elec-
trons and/or ions that flow due to the potential gradient?
dc
2. Let us require that only electrical potential gradient is of interest. Then dxi = 0, and

dφ
Ji = − ui ci zi F (1:73)
dx

Taking into account Faraday’s law of equivalence of mass and charge transport:
X
jel = zi FJi (1:74)
i

and merging these two equations, we will have

X dφ X dφ
jel = − Ui ci z2i F 2 =− Ui ci zi F (1:75)
i
dx i
dx

The right-hand side of this equation is exactly the first Ohm’s law if we set the
resistance R = −κ − 1 . For the case of ion transport, however, we should take into
account the stoichiometry coefficients, νi, of electrolyte dissociation, and so we
have the straightforward definition of electrolyte conductivity:
26 1 Basic concepts

X
κ= jzi jci FUi , where Ui is called the electrolytic mobility. (1:76)
i

We can multiply the examples of applications of the Nernst–Planck equation, but the
two above examples will show us its usefulness in applied electrochemistry. Let us elab-
orate a little more on the first example: the first Fick’s law and diffusion-controlled pro-
cesses. This approach finds its applications in the electroanalytical chemistry, because
the solutions of differential equations of Fick’s first and second laws (the latter we will
discuss in, e.g., Chapter 2), under appropriately chosen boundary conditions, lead to the
direct functional relations between the measured current and concentration of the elec-
troactive species in the solution, food or environment. Therefore, this provides precise
control over the composition or possible contamination of the investigated samples or
pollutants. In many cases, it is also possible to determine the identity of an investigated
electroactive substances. The monitoring capabilities of electrochemistry is utilized fre-
quently in pharmacy and medicine (e.g., body fluid analyses).
The second example is more obvious: Ohm’s law is one of the fundamentals of
electronics. On the other hand, the case of charge transport in the electrolyte describ-
ing its conductivity is one of the essential factors in constructing the charge storage
and converting devices. A proper choice of charges moving between the electrodes in
batteries and cells, providing their high mobilities and high conductivities, is one of
the technological problems continuously requiring improvement as the demand for
such devices grows.
Having all this basic information introduced in the sections above, we now de-
scribe processes that can be involved in the electron transfer across the electrode/so-
lution interface – the redox process. They are visualized in the graph below, with the
proviso that there is no adsorption/desorption of electroactive species, nor the conse-
cutive reactions take place. In the description below, Ox represents the oxidized form
of an electroactive species (e.g., Fe(CN)63−) while Red represents the reduced form of
an electroactive species (e.g., Fe(CN)64−).
In Scheme 1.1:
– Step 1 describes the diffusion of the electroactive species (Ox for the reduction
and Red for the case of oxidation) to/from the Helmholtz plane (OHP) at the elec-
trode (compare Section 1.1), this transport is characterized by the mass transfer
coefficient kD [cm/s] and will be discussed below.
– Step 2 approximates a reorientation of ionic “atmosphere” that accompanies a
charged electroactive species and formed of nonelectroactive counterions always
present in the solution (e.g., K+ ions). Typically, this process takes ca. 10−8 second
(tens of nanoseconds).
– Step 3 reorganization of solvent dipoles, ca. 10−11 second (tens of picoseconds).
– Step 4 accounts for the distance changes between the “central” ion and its “ligands”
(e.g., Fe3+/4+ and CN− in our example below), time frame ca. 10−14 second.
– Step 5 is the electron transfer itself, ca. 10−16 second.
 1.1 Structure of interfaces 27

Ox diffuses to/from
the electrode (step 1)
Steps 2–4 Ox OHP Ox∞
kd,Ox

Electron
e transfer kRed kOx
(step 5)

kd,Red
Red OHP Red∞
Red diffuses to/from
Electrode the electrode (step 1)

Scheme 1.1: A scheme showing processes that can be involved in the electron transfer across the electrode/
solution interface – the redox processes. No adsorption/desorption of electroactive species take place, and
no the consecutive reactions are involved. More details are given in the text.

For further description of the electrode processes, we will always follow the conven-
tion (Stockholm convention) to write the redox reaction toward reduction; therefore,

Ox + ne ! Red

Cu2+ + 2e ! Cu0

Zn2+ + 2e ! Zn0

or

FeðCNÞ6 3− + 1e ! FeðCNÞ6 4−

where ne stands for a number of electrons involved in the unit redox process (half-
reaction).

1.1.3.2 “Reversible” electrode processes

By “reversible” electrode processes we understand the electrode processes that are
limited by the rate of diffusion of a redox species to the electrode (Step 1 in the graph
above) and not by the rate of electron transfer between the redox species and the
electrode (or other processes Steps 2–5). From this point of view, the redox process
itself is at the dynamic equilibrium, meaning that the rate of oxidation is equal to the
rate of reduction. There are two obvious conditions of reversibility: reversibility of
mass and reversibility of energy. The first one states that the mass converted in the
“forward” process has to be equal to the mass converted back during the “backward”
process. The second condition means that the energies necessary for “forward” and
“backward” processes are equal. At constant pressure p and temperature T it means
that the Gibbs’ free energy change of the redox reaction, ΔGp,T = 0.
28 1 Basic concepts

Some quantitative considerations: Let us call upon the Nernst–Planck equation

for flux for the case of diffusion processes (reduction of Ox):
dcOx dcOx
JOx = −uOx RT = −D (1:77)
dx dx

If we now set that the volume containing the electroactive species is sufficiently large,
we can assume that the concentration gradient is constant and independent of time
(stationary state), which gives us immediately that the JOx is constant, too (of course
within the limited time frame). Then, introducing Faraday’s law, the equation for cur-
rent, with cOx dependent only on the distance from the electrode, cOx(x), will read:
 
dcOx ðxÞ
i = −nFDA (1:78)
dx

Further assumption of a linear concentration profile in a constant diffusion region of

thickness δ (see Fig. 1.12) can be written as follows:

dcOx ðxÞ cOx,x=0 − c0Ox

= (1:79)
dx δ

where cOx, x = 0 is the concentration of Ox at the electrode, c0Ox is the concentration of

Ox in the bulk of the solution, and δ is the thickness of the diffusion layer, as men-
tioned earlier.
Thus the equation for current is now
 0 
c − cOx,x=0
i = −nFDA Ox (1:80)
δ

󰛿
0
COx

COx, x=0
Electrode x

Fig. 1.12: Graph representing a “frozen in time” concentration profile versus the distance x from the
electrode surface (x = 0). As a consequence, a constant diffusion region of thickness δ is developed. Bulk
0
concentration of Ox is denoted here as cOx , whereas its concentration at the electrode surface is cOx,x=0 .
This means that the potential applied to the electrode is sufficiently large to reduce immediately all
arriving Ox form.
 1.1 Structure of interfaces 29

It should be obvious for the reader that there exist two limiting conditions for the
dc ðxÞ
magnitude of the concentration gradient Ox dx :
dcOx ðxÞ
– dx = 0, meaning that no resultant flux appears to/from the electrode. This situ-
ation reflects the equilibrium conditions; no reduction (oxidation) reaction takes
place at the electrode surface or, more correctly, the rate of reduction equals the
rate of oxidation. This means that the concentration of Ox in the bulk (x = ∞)
equals to its concentration at the surface of the electrode (x = 0), c0Ox = cOx,x=0 .
– cOx,x=0 = 0, meaning that the redox reaction is so fast that it literally consumes all
of Ox form diffusing to the electrode from the bulk, converting it into Red form.
This corresponds to the maximum value of the concentration gradient, and ac-
cording to the eq. (1.80), this results in the maximum value of ionic flux JOx and
(Faraday law) yields constant, limiting value of current, il:

0 
c
il = nFAD Ox (1:81)
δ

The equilibrium conditions of redox processes at the metallic electrodes are quantita-
tively described by the well-known Nernst equation, relating the equilibrium poten-
tial, Eeq, of such electrodes with the concentration ratio of forms Ox and Red present
in the solution:
RT cOx
Eeq = Ef + ln (1:82)
nF cRed

where Ef is the so-called formal potential of the electrode if the ratio cOx/cRed = 1 (see
also Section 1.1.3); other symbols have their usual meanings.
But what drives the formation of a concentration gradient, finally reaching its maxi-
mum value for the case of immediate reduction (oxidation) of the electroactive species at
the electrode surface, so that cOx,x=0 = 0 ðcRed,x=0 = 0Þ? The reason is the external polariza-
tion of the electrode, for example, from an external voltage source, such as potentiostat.
To drive the reduction, the electrode should be polarized negatively with respect to its
equilibrium, Nernst potential. Under such conditions, the electrode acts as a source of
electrons for the reduction process to proceed. The more negative the electrode potential,
the more Ox form is being consumed at the surface. Quantitatively, this is described by
an overpotential, η. The overpotential is the potential difference between the potential
at which the redox event proceeds experimentally, E, and the thermodynamically
determined equilibrium potential (Nernst equation), Eeq:

η = E − Eeq (1:83)

Since the diffusion to the electrode is the slowest process, the electrode polarization
develops the concentration gradient, the overpotential is called the concentration
overpotential, ηc. Until now we were discussing the situation of disturbing the elec-
trode from its equilibrium by polarizing it from an external source, changing the
30 1 Basic concepts

concentration of electroactive species in the electrode close vicinity. In this case, we

can control the value of ηc by controlling the value of E. However, it is necessary to
say, at this moment, that the concentration overpotential can also be created in the
situation when our electrode is connected via an external resistance to the other
electrode of different potential. Thus, we build an electrochemical cell being dis-
charged by the resistance. Current i flows through the circuit causing the concentra-
tions of the electroactive species at the electrode surfaces to also change: on the
anode the oxidation process will increase the concentration of OxI decreasing the
concentration of RedI, whereas, on the cathode, the reduction will build the concen-
tration of RedII, decreasing the concentration of OxII. Thus the concentration over-
potential will be created whether we externally polarize the electrode (electrolysis),
or the electrode will spontaneously depolarize itself performing electrochemical
work in the cell.
Let us now return to the diffusion-limited electrolysis on the electrode. The prob-
lem we are now facing is: how does the concentration profile developed with the ap-
plied overpotential correspond to the measured electronic current flow. In other
words, can we use the current measured during the electrode polarization as a “fin-
gerprint” of the type of redox process occurring at this electrode?
To answer this question, we relate the electrode potential E to the concentration
of Ox at the electrode surface cOx,x=0 . We then can get
RT cOx,x=0
E = Ef + ln (1:84)
nF cRed

Now, after subtraction from the above relation the equilibrium potential Eeq defined
by the Nernst equation, we get the relation of the concentration overpotential with
the natural logarithm of the ratio of the concentration at the surface of the electrode
to the concentration in the bulk. Because cOx,x=0 < c0 when the reduction proceeds, the
value of overpotential for the reduction will be negative:
R cOx,x=0
η= ln 0 < 0 (1:85)
nF c

Introducing to this equation the expressions for current i and limiting current il we
will finally obtain the desired relation between the concentration overpotential and
current measured experimentally:

RT il − i
ηc = ln (1:86)
nF il

or


nFηc
i = il 1 − exp (1:87)
RT
 1.1 Structure of interfaces 31

i Oxidation

il,a

Eeq E Fig. 1.13: Current–potential profile resulting from eq.

il,c (1.87), applied for both the reduction and oxidation
processes. Characteristic points, enabling immediate
identification of Eeq and il for the investigated process,
Reduction are shown.

Of course, the above considerations can be carried out for the anodic processes, replac-
ing Ox with Red as the electroactive species that diffuse to the electrode surface.
There are two characteristic values that can be immediately obtained from the
above dependence of current versus potential, which are characteristic of the investi-
gated redox process:
– the equilibrium potential, Eeq of the investigated redox couple, for the case of
zero current measured in an external circuit (η = 0), and
– the limiting current, il for large overpotentials, both cathodic and anodic. The
last one is related directly to the diffusion coefficient D and the thickness of the
diffusion layer δ by eq. (1.81). The ratio D/δ = kD is the so-called mass-transfer
coefficient.

Scrupulous reader will easily see that the above considerations are valid only for the
steady-state conditions, where JOx (or JRed for oxidation reaction) are constant because
in the above description of the diffusion-limited, linear, semi-infinite processes we did
not account for the time of the electrolysis process. In fact, the graph representing the
“frozen-in-time” concentration profile should be supplemented with other graphs,
showing the evolution in time of the diffusion layer at constant applied potential, and/
or graph showing the potential-dependent concentration change at the electrode sur-
face (Fig. 1.14).
As it is shown in this graph, as the time of electrolysis increases at constant elec-
trode potential, the thickness δ of the diffusion layer also increases, expanding from the
electrode surface toward the bulk of the electrolyte, so δ1 < δ2 < δ3. With this, the con-
centration gradient, being inversely proportional to δ, decreases (compare eq. 1.81),
slowing down the diffusion of the electroactive species to the electrode surface. This, in
turn, decreases the current flow through the external circuit.
All this can be accounted for if we include in the formula for the current flow
during polarization of the electrode not only the overpotential domain but also the
time domain. We will discuss it in the next parts, presenting transient methods used
in applied electrochemistry.
32 1 Basic concepts

󰛿3
t3 > t2 > t1
󰛿2
0
󰛿1 COx
COx

t increases

COx, x = 0
Fig. 1.14: Time evolution of concentration profiles
E constant
of the diffusion layer at the electrode surface at
x constant applied potential.

1.1.3.3 “Irreversible” electrode processes: basic concepts of electrochemical

kinetics
Let us now assume that Step 5 – the electron transfer itself – in the scheme (Scheme 1.1)
of processes involved in the redox reaction is the sluggish one. The main reason validat-
ing such an assumption is the kinetic limitation of the heterogeneous electron transfer
rate. However, one should take into account Steps 2–4 in the detailed description. Such
considerations, including changes of distances or geometry in the solvation and/or coor-
dination shells of the two chemical species, led to the development of Marcus, Hush,
Levich, Dogonadze (MHLD) theory (R.A. Marcus, Nobel Prize in chemistry 1992).
In other words, there exists an activation energy barrier limiting kinetically the
transfer of electrons across the metal/solution interface. If so, we can draw upon the
kinetic equations that should account for the electron transfer processes under such
conditions.
First, keeping in mind the accepted convention of redox processes: Ox + ne → Red,
and bearing in mind that the overall “concentration” of electrons in their source (e.g.,
metallic connectors) during such reaction is independent of time, we can consider the
redox process as described by the first-order kinetics:

dcOx dcRed
ν= − = = kc (1:88)
dt dt

As we stated before, Faraday’s law relates the flux of charges in time (the rate of the
reaction, ν), with current flow: i = nFAv, or, introducing current density j = i/A.

νRed = k1 cOx and νOx = k2 cRed (1:89)

and therefore

jC = nFk1 cOx and jA = nFk2 cRed (1:90)

where k1 and k2 are the reduction and oxidation rate constants, respectively. Other
symbols were defined before. At equilibrium Ox + ne ↔ Red, the potential of such
 1.1 Structure of interfaces 33

electrode reaches the Nernst’ value and the rate of reduction, νRed, equals the rate of
oxidation, νOx, and so the cathodic and anodic currents are equal: jC = jA = j0. The last
value, j0, is called the exchange current density and characterizes the redox process for a
given system (electrode, electrolyte solution, and redox species). Considering the electro-
chemical processes occurring on the electrode, both the reduction and oxidation pro-
cesses should be considered at any time and the overpotential value, even though one
predominates the other (except at equilibrium). During the reduction, the cathodic cur-
rent is larger than the anodic one, and vice versa, during the oxidation, the anodic cur-
rent is larger. What is measured during the experiment in the external circuitry is the
resultant, net current. According to the IUPAC convention, the net current is

j = jA − jC (1:91)

Since the concentrations of Ox (or Red) forms in the bulk of electrolyte do not change
in time in the applied potential range, the presence of net current corresponds to the
changes of the reaction rate constants k1 and k2. Let us now apply the Eyring–Polanyi
theory (transition-state theory) for the electrode process to account for the depen-
dence of rate constants of the energetics of the electron transfer rate. We also intro-
duce the concept of standard rate constant k0 and standard Gibbs free energy of
activation, ΔG0#:


kB T ΔG0#
k0 = exp − (1:92)
h RT

Here, kB is the Boltzmann constant, h denotes the Planck constant, and other symbols
have their usual meanings.
If the net observed current is a sum of cathodic and anodic ones, in light of
eq. (1.90), it will depend on the two rate constants: k1 for reduction and k2 for
oxidation:
!
0 kB T ΔG0#
k1 = exp − 1
(1:93)
h RT
!
0 kB T ΔG0#
k2 = exp − 2
(1:94)
h RT

where ΔG10# and ΔG20# are the standard Gibbs free energies of activation for reduc-
tion and oxidation reactions, respectively.
According to Fig. 1.15, ΔG10# > ΔG20#, so the reduction rate constant, k10, is smaller
than the oxidation rate constant, k20, and the reduction current is smaller than the
oxidation one. The electric field ΔGE = nFE decreases the value of ΔG20# to ΔG2,E0# by a
fraction (1−α)nFE accelerating even further the oxidation reaction, slowing down the
reduction by an increase of ΔG10# by a fraction αnFE. The coefficient α, called the
charge transfer coefficient or barrier symmetry coefficient, falls within the range
34 1 Basic concepts

0 ≤ α ≤ 1 reflects the influence of electrode polarization on the Gibbs free activation

energy of the reaction. Taking this into account, we can rewrite eqs. (1.93) and (1.94)
in form:

GE#
ΔG0#
G0# 𝛼nFE

0#
ΔG2,E ΔG2
0# (1–𝛼)nFE

0# 0#
ΔG1 ΔG1,E
nFE

X Distance from
Red Ox the electrode
Helmholtz plane
Metal ions Centers of ions
in solid electrode in the solution

Fig.1.15: Schematic drawing of the standard Gibbs free energy of activation in the absence of electric field
E (black) and in the presence of electric potential, E (red), polarizing the electrode anodically. Detailed
description is given in the text.

!
kB T ΔG0# + αnFE
k1 = exp − 1 (1:95)
h RT

and
!
kB T ΔG0#
2 − ð1 − αÞnFE
k2 = exp − (1:96)
h RT

These two equations can be rewritten to separate the part with no electric field
(“chemical”), described by k10 and k20 , respectively, from the part containing the effect
of polarization of the metal/solution interface (“electrical”):


αnFE
k1 = k10 exp − (1:97)
RT

and


ð1 − αÞnFE
k2 = k20 exp (1:98)
RT

Now, introducing these equations into eq. (1.90), we will get the equations for cathodic
and anodic currents that flow through the electrode at potential E. At equilibrium, the
 1.1 Structure of interfaces 35

potential of an electrode is equal to Nernst equilibrium potential, E = Eeq. Moreover, the

rates of oxidation and reduction reactions are equal (eq. 1.89), and so

νRed = νOx = k1 = k2 (1:99)

Thus, under equilibrium conditions we now have



αnFEeq ð1 − αÞnFEeq
k1 0 exp − = k2 0 exp = ks (1:100)
RT RT

Rewriting the above two equations in terms of current densities (eq. 1.90) and keeping
in mind that the resultant current flow is a net flow of anodic and cathodic ones
(eq. 1.91), we will have



ð1 − αÞnFE αnFE
j = nFcRed k2 0 exp − zFcox k1 0 exp − (1:101)
RT RT

Now, calling upon the definition of the overpotential (this time called the kinetic over-
potential) η = E – Eeq, or E = Eeq + η, we can rewrite the above equation in a form:



ð1 − αÞnFEeq ð1 − αÞnFη
j = nFcRed k20 exp exp
RT RT



αnFEeq αnFη
− nFcox k1 exp −
0
exp − (1:102)
RT RT

Finally, after some rearrangement we have the so-called Butler–Volmer equation:




ð1 − αÞnFη αnFη
j = j0 exp − exp − (1:103)
RT RT

where

j0 = nFks cRed α cOx 1−α (1:104)

is the so-called the exchange current density, describing the rate of oxidation and re-
duction currents at equilibrium, and ks is the standard rate constant described above
(eq. 1.100).
Equation (1.103) is valid when the electrode reaction is controlled by the kinetics
of electron transfer at the electrode (not the diffusional mass transfer described ear-
lier), taking advantage of Arrhenius and Eyring transition state theory (eqs. 1.93 and
1.94). The utility of the Butler–Volmer theory in electrochemistry is wide and we can
show it later in the text for the case of, for example, corrosion processes. The overpo-
tential η appearing in this equation is called the activation overpotential and is differ-
ent from the concentration overpotential in its origin, as discussed above for the case
of diffusion-limited currents (Section 1.3.2). The most important parameter describing
the rate of electron exchange across the electrode interface is the exchange current i0
(or exchange current density j0). It is characteristic for a given electrode material,
36 1 Basic concepts

redox system present, and electrolyte solution; the higher its value the faster the sys-
tem reaches equilibrium state and equilibrium potential, approaching the limiting
conditions described by the diffusion of electroactive species (eq. 1.87). If the exchange
current is small, we can assume the irreversibility of the electrode process and the
equilibrium state of the electrode cannot be achieved. The exchange current is the
equilibrium current (electron self-exchange between the Ox and Red states or elec-
trode material and Red or Ox) and does not require any additional energy due to the
polarization of the electrode). The effect of electrode polarization is contained in the
exponential part of the Butler–Volmer equation. Therefore, let us consider that we
are measuring the current density at high activation overpotentials, either for a ca-
thodic, reduction process (ηC) or anodic, oxidation process (ηA). For sufficiently high
overpotentials eq. (1.103) will take the form:
 
− αnFηc
Cathodic process: jC = j0 exp (1:105)
RT

 
ð1 − αÞnFηA
Anodic process: jA = j0 exp (1:106)
RT

Frequently, the above equations are shown in a linearized form of lnjC or lnjA:

αnFηc
lnjC = lnj0 − (1:107)
RT

ð1 − αÞnFηA
lnjA = lnj0 + (1:108)
RT

This notation often called the Tafel relations allows one for easy evaluation of the ex-
change current j0, as shown in Fig. 1.16.
For small values of activation overpotential (η < 25 mV), eq. (1.103) can also be
linearized, because for small x, the exponential ex ≈ 1 + x. Thus, we can get the linear
dependence of current density and η:

nFη RT j
j = j0 or η= (1:109)
RT zF j0

RT 1
Because the units of fragment zF j0 are Ohms, the right-hand side of this equation
can be rewritten in the form:

η = Re j (1:110)

where Re is called the resistance of the electrode reaction or polarization resistance

(for corrosion process discussed later).
 1.1 Structure of interfaces 37

(a) jA (b)
j In j
𝛼 = 0.5
In j0
j0

j0 𝜂 = E–Eeq

Eeq Eeq E

jC

(c) j = jA – jC
j

𝜂 = E–Eeq
𝛼 = 0.25 oxidation proceeds easier (blue line)
𝛼 = 0.5 symmetric process (black line)
𝛼= 0.75 reduction proceeds easier (red line)

Fig. 1.16: (a) Graphical representation of Butler–Volmer equation (eq. 1.103), showing also the anodic ( jA)
and cathodic ( jC) contributions to the measured current, as well as the nonmeasurable exchange current
j0 for the case of charge transfer coefficient α = 0.5. (b) Linearization of eq. (1.103), showing the evaluation
ð1 − αÞzF
of the exchange current j0. The slopes are − αzFRT and RT for the cathodic and anodic currents,
respectively. From these slopes we can evaluate the value of charge transfer coefficient, α. Typically, its
value falls within the range of 0.25 < α < 0.75, and its effect on the redox current is exemplified in c).

Summarizing the above mechanisms of redox reaction at the electrodes that are re-
sponsible for current flow in the external circuit, the latter can be represented graphi-
cally (Fig. 1.17), with separate regions assigned to different limiting conditions: kinetic
and diffusional.

1.1.3.4 Briefly on Marcus, Hush, Levich, Dogonadze (MHLD) theory of electrode

processes
One may notice that in the above description of electrode processes, Steps 2–4 shown
in Scheme 1.1, were completely neglected. Although in Eyring–Polanyi theory (transi-
tion-state theory) for the electrode process the electrode and the Ox or Red forms be-
come strongly coupled forming the so-called activated complex, the electrode can be
only weakly coupled with the substrates. In such a case the electroactive forms (Ox or
38 1 Basic concepts

Concentration
polarization
C (diffusion)
Activation
polarization
(kinetics)
Fig. 1.17: Anodic (oxidation) current summarizing
A
the mechanisms of redox reaction with separate
regions assigned to different limiting conditions:
Eeq E
kinetic and diffusional.

Red) retain their individual chemistries. It is the reorganization of the surroundings

(induced thermally), the solvent (outer sphere), and the solvent together with li-
gands (inner sphere) that create favorable situation before the electron transfer.
This concept takes advantage of Marcus electron transfer theory (published 1956,
Nobel Prize in Chemistry 1992), which employs a classical, purely electrostatic
model. It was further developed, extended, and refined significantly by Hush, Lev-
ich, and Dogonadze to address a variety of different types of charge transfer reac-
tions. We address here only brief insight into this theory.
The effects of such structural changes were accounted for by the introduction of
the reorganization energy, λ, corresponding to the energy required to change the
substrate (Ox) surroundings into the structure characteristic for the product (Red) at
equilibrium. It is equivalent to the energy required for the products to achieve the
configuration of substrates (or vice versa). If we assume that the energy curves can be
approximated by quadratic functions (parabolas) as in Fig. 1.18, then the Gibbs free
energy of activation, ΔG#, is equal to
1
ΔG# = ðλ + ΔGr Þ2 (1:111)
4λ

Moreover, an assumption is made that these parabolas are identical in shape for sub-
strates and products. Thus, the rate constant of electron transfer during the unit
redox process, kct, can be written in a form equivalent to eq. (1.92), but including the
reorganization term λ:

 !
kT ΔG# kT ðλ + ΔGr Þ2
kct = exp − = exp − (1:112)
h RT h 4λRT
 Bibliography 39

+
λ

ΔG#
+
ΔGr

Reaction coordinate ξ

Fig. 1.18: Left panel: Outer sphere electron transfer (upper part): reaction between the electrode and the
electroactive species does not require changes in the original sphere and the e.d.l. Inner sphere (lower
part): Reaction with the formation of an activated complex with structural changes of the original
coordination sphere and e.d.l. structure. Note that even if there is no strong interaction with the
electrode, an outer sphere reaction can depend on the electrode material because of the double layer
effects and the effect of the metal on the structure of the Helmholtz layer. Right panel: Description of
symbols used in eq. (1.111).

MHLD theory found its implications not only in electron transfer between inorganic,
organometallic, and organic molecules, but also in bioelectrochemistry of proteins,
enzymes, and nucleic acids.

Bibliography

[1] Bocris J O’M, Reddy AKN. Modern electrochemistry. V2 New York, USA, Plenum Press, 1970.
[2] Bagotsky V.S. Fundamentals of electrochemistry. Hoboken, USA, J Wiley & Sons Inc., 2006.
[3] Gileadi E. Physical electrochemistry. Fundamentals, techniques and applications. Weinheim, FRG,
Willey-VCH Verlag GmbH & Co, 2011.
[4] Koryta J, Dvorák J, Kavan L. Principles of electrochemistry, John Wiley & Sons Inc, ed.,
Chichester, 1993.
[5] Bard AJ, Inzelt G, Scholtz F, eds., Electrochemical dictionary, 2nd Edition, Springer-Verlag, 2008, 2012.
[6] Brett CMA, Oliveira Brett AM., Electrochemistry: Principles, methods, and applications, Oxford
Science Publ., 1993.
[7] Karnik R, Fan R, Yue M, Li DY, Yang PD, Majumdar A. Electrostatic control of ions and molecules in
nanofluidic transistors. Nano Letters 2005, 5, 943–948.
[8] Wang C, Wang S, Chen G, Kong W, Ping W, Dai J, Pastel G, Xie H, He S, Das S, Hu L. Flexible, bio-
compatible nanofluidic ion conductor. Cem. Mater. 2018, 30, 7707–7713.
[9] Melitz W, Shen J, Kummel AC, Lee S. Kelvin probe force microscopy and its applications, in Surface
Science Reports, 2011, 66, 1–27.
[10] Kjelstrup S, Bedeaux D, Johannessen E, Gross J. Non-equilibrium thermodynamics for engineers, 2nd
edition. World Scientific,. Singapore, 2017.
40 1 Basic concepts

1.2 Structure of the bulk of electrolytes: conductivity

1.2.1 Interactions in aqueous electrolytes

1.2.1.1 Ion–solvent interaction; solvation

For the purpose of this text, we will discuss primarily the aqueous electrolyte solu-
tions, but the potential reader should keep in mind that the electrolyte solutions can
use also other solvents of sufficiently high relative dielectric permittivity (dielectric
constant), ε, allowing for the dissociation of solute molecules into ions. And so the
electrolyte solutions are the systems containing a solvent and a substance that is dis-
sociated partially (weak electrolytes) or totally (strong electrolytes) into ions. These
solutions conduct the electric current and the charge carriers are the dissociated ions
moving in the external electric field. Of course, the electroneutrality condition has to
be fulfilled. The choice of whether the electrolyte can be classified as “strong” or
“weak” depends on the so-called dissociation coefficient (constant), α, being equal to 1
for all strong electrolytes, such as ionic salts, hydrochloric acid, alkaline metals’ hy-
droxides, etc. In the case of weak electrolytes, the dissociation constant is below 1 and
depends on the concentration of such electrolytes. The lower the concentration, the
higher the dissociation constant, approximating 1 for infinitesimally small concentra-
tions. It is also worth mentioning, that the dissociation constant depends strongly on
the properties of a solvent affecting the interactions between the ions and solvent
molecules. In the absence of an external electric field, one should consider at least
two types of interactions: ion–solvent molecules interactions and ion–ion interac-
tions. In the classical theories of electrolyte solutions, ions are considered as particles
moving in a dielectric continuum of a solvent.
Therefore, the main deviation from an ideal gas behavior was due to the Coulomb
electrostatic interactions, determining the nonideal behavior of electrolytes. In general,
one should consider also the electrostatic and/or Van der Waals interactions between
the ions and solvent molecules, contributing largely to the behavior of electrolyte solu-
tions. These interactions are generally described as solvation (hydration in the case of
aqueous solutions) of ions and lead to the formation of a solvation “sphere” around an
ion. The importance of such interactions becomes evident when comparing the hydra-
tion energies of monovalent ions and neutral molecules: in the former case, the en-
thalpy of hydration is in the range of several hundred kJ/mol, whereas for simple
nonpolar molecules, such as methane, it is in the range of tens of kJ/mol. Therefore, it is
evident that such interactions should lead to the structuring of electrolyte solutions.
The simplest description of ion–solvent interactions was proposed by Max Born (ca.
1920). It allows us to estimate the electrostatic component of Gibbs free energy of solva-
tion of an ion, by treating it as a sphere of an “effective” radius r, of zie unit charge in a
continuous dielectric medium of a relative dielectric constant ε. This estimation is
based on the following three-step rationale:
 1.2 Structure of the bulk of electrolytes: conductivity 41

Ð0
a) discharging the ion in a vacuum requires a work W1 = z e ψ1 dq, where ψ1 is the
i
electrical potential at the surface of an ion: ψ1 = 4πεq r. After integration, we get:
0

ðzi eÞ2
W1 = − (1:113)
8πε0 r

b) discharged ion is now introduced to the solvent of a relative dielectric constant εr

without any work loss or gain,
c) re-charging of an ion in this new environment (solvent), to the electric potential
q Ð zi e ðzi eÞ2
at the surface ψ2 = 4πεε r. This process requires a work W2 = 0 ψ2 dq = 8πεε r.
0 0

Under constant pressure and temperature conditions, the work required for both pro-
cesses carried out for 1 M of ions is equal to the change of the Gibbs free energy:


N A z2 e 2 1
ΔG = − 1− (1:114)
8πε0 r0 εr

Even though this interpretation is oversimplified, it is frequently used when compari-

sons of different solvation behavior of solvents are considered, as can be the case in the
evaluation of interfacial electrochemical processes and selection of solvents during the
construction of fuel cells. On the other hand, for a given solvent, the “effective” ionic
radius and ionic charge are factors leading sometimes to the so-called structure break-
ing of a solvent. This effect increases with ionic radius, for example, Li+ < Na+ < Rb+ < Cs+
or, Cl− < NO3− < I− < ClO3− and with an increasing charge, for example, Li+ < Be2+ < Ca2+.
The solvation of ions has to be considered when discussing the electrolyte’s electric con-
ductivity/ionic resistance in batteries and fuel cells because it plays a more important
role in determining electrolyte properties than the ion–ion interactions described
below.

1.2.1.2 Ion–ion interactions

By eq. (1.35) we have already defined the chemical potential of species “i,” µi. Now, we
can use it for our considerations of ion–ion interactions. For an ideal solution, the
i ; can be described as:
chemical potential of its component, μid

i = μi + RTlnci ,
μid 0
(1:115)

where μ0i is the standard chemical potential of species “i” and ci is its concentration.
However, due to the ion–ion interaction in the real solutions the “real” concentration
of species “i” is different and is reflected by the activity coefficient γi. Thus,

i = μi + RTlnci γi ;
μre 0
(1:116)
42 1 Basic concepts

where μre i is the chemical potential of “i” in a real solution. Therefore, the difference
between μre i and μi can be considered as molar Gibbs free energy loss for the electro-
id

static interactions between ions, resulting in the first approximation, in the deviations
of electrolytes from ideal solutions:

RTlnγ i = μre
i − μi
id
(1:117)

Activity coefficients of single ions cannot be measured experimentally because an

electrolyte solution must contain both positively charged ions and negatively charged
ions. Instead, a mean activity coefficient for the strong, totally dissociated electrolyte
AnBm is defined as:

γ± = ðγ −n γ +m Þ1=ðn+mÞ (1:118)

And so, the multiple-charge generalization from electrostatics gives the expression for
the potential energy of the entire solution

RTlnγ ± = μre − μid (1:119)

Some thermodynamic considerations, easily found in other textbooks, can lead to the
equation, named the Debye‒Hückel limiting equation (Peter Debye, Erich Hückel,
1923). The underlying theory assumes a simplified model of electrolyte solution, nev-
ertheless giving quite good predictions of mean activity coefficients for ions in dilute
solutions. Under similar assumptions as in the case of the Gouy‒Chapman model de-
scribed earlier in this chapter, it is a linearized Poisson‒Boltzmann model (eqs. (1.24)
and (1.25)) and provides a good starting point for modern approaches to the descrip-
tion of real electrolyte solutions. However, what differs the Debye‒Hückel model
from the Gouy‒Chapman electrical double layer model is, that instead of a flat, uni-
form metal|solution interface, now we have a “central” point charge (be it a cation or
anion), surrounded by a spherical cloud of counterions’ contribution, balancing to
zero the net charge of a central ion and its ionic cloud. The central ion and its ionic
cloud are electrically neutral, and the radius of such a cloud is described by the very
same equation as the Debye length of the e.d.l., vide supra, eq. (1.27).
The evaluation of the electrostatic energy of interactions between the central ion
and its ionic cloud (Coulomb interactions) is the key point of the D‒H theory. As a
consequence, it allows for the evaluation of such interactions with the limiting equa-
tion, valid for strong electrolytes under infinite dilution, containing monovalent ions
(e.g., KCl):
p
RTlogγ ± = −Ajz + z −j I (1:120)

where A is a temperature-dependent constant characteristic for the solvent at a given

temperature T, z+ and z− are the valencies of ions, and I is the ionic strength of the
solution:
 1.2 Structure of the bulk of electrolytes: conductivity 43

A δ- B

δ+
+ –
Fig. 1.19: An ionic atmosphere at rest (A). A spherical cloud of charge δ+ (or δ−) surrounds point charge
ions in solution. Its radius is identical to the Dybye length of the e.d.l., eq. (1.27). (B). Asymmetrical cloud
in an external electric field; positively charged central ion and net-negatively charged ionic cloud.

1X 2
I= c i zi (1:121)
2 i

For a more concentrated solution, the minimum distance between the ions, a, is intro-
duced (the distance of closest approach of ions), an empirical parameter, changing
the form of the limiting Debye‒Hückel equation:
p
Ajz + z−j I
log γ ± = − p ; (1:122)
1 + aB I

here B is also a constant dependent on the solvent parameters and T. Finally, for even
higher ionic strength, one more empirical parameter C is added, accounting for the
effect of ionic strength on the dielectric constant of a solvent.
p
Ajz + z−j I
log γ ± = − p + CI (1:123)
1 + aB I

All these modifications are illustrated in Fig. 1.20.

0
logϒI

2 Fig. 1.20: Graphical representations of subsequent modifications

1 of the D‒H theory: curve 1 ‒ the limiting equation (straight line),
2 – including the distance of closest approach, 3 – influence of
√I the ionic strength on the dielectric behavior of a solvent.

Therefore, the activity coefficients are themselves functions of concentration as the

amount of inter-ionic interaction increases as the concentration of the electrolyte
increases.
44 1 Basic concepts

1.2.2 Conductivity of electrolytes

It was pointed out in the previous section, that apart from the mass transport during
the electrode reactions (reversible and irreversible), the Nernst‒Planck equation (eq.
1.70) describes also a substantial part of ion mass/charge transport through the bulk
of the electrolyte solution under the influence of the external electric field, called mi-
gration, Jm,i. In addition to this, this equation should be expanded to include also the
convective transport of ions, defined as the transport of substances with a moving me-
dium (e.g., in a liquid flow due to the thermal gradient – convection). Under typical
conditions, the total flux of species “i” is then the algebraic (vector) sum of densities
of all flux types:

Ji = Jd;i + Jm;i + Jcv;i (1:124)

Here: Jd,i describes the diffusion transport (1st Fick’s law, eq. (1.72)) and Jm,i is the mi-
gration flux, eq. (1.73). The convection flux density, Jcv,i, description is beyond the
scope of this textbook. Since we have described in detail the limiting case of diffusion
current control of the electrode processes (Section 1.1.3.2), let us now focus on the mi-
gration flux of ions. Taking into account eq. (1.73) (vide supra), we can get the straight-
P
forward definition of a specific conductivity of electrolyte solution, κ = F i jzi jci Ui ,
(compare eq. (1.76)), where F zi corresponds to the charge transferred by 1 mol of “i”
ions with valency zi. Other symbols were described previously. What can be immedi-
ately seen from this relation is that all effects occurring in the electrolyte that
decrease the ionic mobility will also decrease the conductivity of the electrolyte. Keep-
ing the ionic conductivity as high as possible is of crucial importance in designing
power sources and storage systems, fuel cells in particular (Chapter 8, cf. eq. (8.11a)).
One may also think that since the specific conductivity scales linearly with concentra-
tion, it might be sufficient to increase the conducting behavior of a specific electrolyte
by simply increasing its concentration. Let us verify this assumption by introducing of
molar conductivity of an electrolyte, Λ:
κ
Λ= (1:125)
c
X
Λ=F jzi jUi (1:126)
i

By doing this, we eliminate the effect of electrolyte concentration, so the molar con-
ductivity should be constant, and concentration-independent. This type of behavior
can be observed only at almost infinite dilutions of electrolytes, where at constant
temperature T, the electrolyte ions can move independently of one another in the
external electric field. This principle is often referred to as Kohlrausch’s law of inde-
pendent migration. Under such conditions, each ionic species’ contribution to the
 1.2 Structure of the bulk of electrolytes: conductivity 45

conductivity of the electrolyte solution depends only on the nature of that particular
ion. This approximation is valid also for weak electrolytes because, in the infinite
dilution, even these electrolytes are totally dissociated into the respective ions.
Thus, the Kohlrausch law can be written as:
X
Λ0 = νi λ0i (1:127)
i

where Λ0 is the limiting molar conductivity of the electrolyte, νi is the stoichiometric

coefficient of an ion “i,” and λ0i is its individual limiting conductivity.
Comparing the above two equations (eqs. (1.125) and (1.126)), it is evident that the
molar conductivity of an individual ion “i” can be written as:

λi = zi FUi (1:128)

where Ui denotes the mobility of ion “i” in the solution at a given concentration (Ui0 will
be its mobility at infinite dilution). The mobility of ion at a given external electric field
strength, E [V/m], is given by:
vi
Ui = ; (1:129)
E

where vi is the velocity of an ion [m/s].

Therefore, apart from the concentration of an electrolyte, also the individual
ionic mobilities (conductivities) at a given concentration do contribute to the overall
charge-transfer properties of an electrolyte. The above discussion (and eqs. (1.76),
(1.125–1.127)) gives a background to understanding the conducting behavior of aque-
ous electrolytes illustrated by the graphs in Fig. 1.21A, B.
There are several information that can be drawn from this graph:
1. The specific conductivities of weak electrolytes are much smaller compared to
the strong ones. This is a direct effect of smaller concentration of “free” ions con-
cerning the analytical concentration of such electrolyte, due to the incomplete dis-
sociation of, for example, acetic acid molecules into the respective ionic species.
2. For very dilute electrolytes, regardless of whether it is strong or weak, the depen-
dence of κ vs. c approaches linearity, as predicted by eq. (1.76), because the num-
ber of ions increases with the concentration.
3. For weak electrolytes, one can observe that the specific conductivity reaches a
shallow plateau and then decreases, because of a decrease in its dissociation con-
stant (for a more detailed description, see Ostwald dilution law).
4. Similar behavior can be observed for strong electrolytes, however, the maximum
value at these curves is much higher and attained much later. This observation can
be explained by a smaller “effective” concentration of ions forming associates of
different stoichiometries and charges, for example, the ionic pair has a net charge
46 1 Basic concepts

a) b)

Specific conductivity [ Ω–1cm–1] 1.0

stro
ng e
lect
roly

ᴧ[ Ω cm mol ]
–1
HCl te

0.5

2
KOH

–1
KCl

LiCl
MgSO4 weak electrolyte
CH3COOH
2 4 6 8 10 12
√ concentration [ M1/2 dm–3/2]
concentration [M dm–3]

Fig. 1.21: A presents the dependence of specific conductivity, κ (eq. (1.76)), of concentrations of aqueous
solutions of strong (HCl, KOH, KCl, and LiCl, dissociation constant = 1) and weak (MgSO4, CH3COOH,
dissociation constant < 1). B – molar conductivities of strong, medium, and weak electrolytes vs. square
root of their concentrations (see text for details).

zero and therefore does not contribute to the overall conductivity, whereas the ion-
triplet has a net charge ± 1, but its dimensions are much larger than that of contrib-
uting ions and so it moves slower in the external electric field. The ion-pair forma-
tion due to the electrostatic interactions was quantitatively formulated by Niels
Bjerrum (1926).

However, a drastic decrease in the specific conductivity of strong electrolytes after

reaching a maximum cannot be explained by this relatively simple kind of interaction
between ions of opposite charge. As the electrolyte concentration increases, the aggre-
gates formed can be held together also by chemical forces, thus attaining special fea-
tures, such as shorter interionic distances, partial desolvation of ions contained in such
structures, or even forming a new solvent sheath. These structures can be seen dis-
tinctly in spectroscopies (IR or Raman). All these give rise to the two effects that are
responsible for a drastic decrease of conductivity in concentrated strong electrolytes:
the relaxation effect and the electrophoretic effect, both impeding strongly the ionic
mobilities. The first one, in its simplest form, results from the presence of an ionic atmo-
sphere around a “central” ion. Under the influence of an external electric field, the cen-
tral ion moves to the electrode of the opposite sign, whereas its ionic cloud moves
toward the other electrode. Therefore, during its movement, the central ions lose and
rebuild their ionic cloud in specified time, called the relaxation time, which de-
pends on the type of ions, solvent, external electric field strength, and temperature.
 1.2 Structure of the bulk of electrolytes: conductivity 47

This also leads to a temporal deformation of the ionic cloud (see Fig. 1.19B) and such
structure generates a local electric field of opposite direction with respect to the exter-
nal electric field in which the ions move, conducting current. This effect hinders the
movement of ions, decreasing their mobilities and decreasing electrolyte conductivity.
The other effect – the electrophoretic effect, is related to the movement of solvated
ions: cations and anions in opposite directions against each other in the external elec-
tric field. The friction between these moving solvated ions diminishes the ionic mobility,
scaling with the square root of ionic strength and inversely proportional to the solvent
parameter – the so-called dynamic viscosity (internal friction coefficient), η, of a sol-
vent. The higher the value of this parameter, the “thicker” the solvent imposing higher
resistance to the moving ions, and of course – higher resistance of the electrolyte solu-
tion. The consequence of both effects on the specific conductivity of electrolyte solutions
is their dependence on concentration, solvent viscosity, dielectric constant, and temper-
ature. These parameters have to be taken seriously into account, for example, in the
construction of new power sources or power storage: The lower the specific conductiv-
ity of the electrolyte, the higher the internal resistance in a such device, affecting its
performance, costs, and usability (Chapter 8).
For better visualization of the ionic effective mobilities in solutions, Fig. 1.21B
shows the dependence of molar conductivity of strong to weak electrolytes in an
aqueous solution as a function of the square root of the ionic strength. The form of
this function predicting linear behavior for strong univalent (1:1) electrolytes was de-
rived empirically by Kohlrausch:
p
Λ = Λ0 − K I (1:130)

where K is the empirical constant, sometimes named the Kohlrausch coefficient. This
equation allows for the evaluation of limiting molar conductivities of strong electro-
pffiffiffi
lytes by extrapolation of Λ as a function of c to c→0. This approach is valid only for
strong electrolytes, as illustrated by Fig. 21B. A more detailed description, allowing for
the assignment of a physical meaning of parameter a, led to the Debye‒Hückel‒Ons-
ager (D‒H‒O) theory. This theory considers also the parameters characterizing not
only the electrolyte, such as the valencies of ions and their individual limiting conduc-
tivities, λ0i , but also those of the solvent, such as dielectric conductivity ε, and its dy-
namic viscosity, η. The latter parameter is necessary to describe the motion of ions in
the solvent. We stress again that the D‒H‒O theory is limited only to strong electro-
lytes and cannot be used for evaluation of limiting conductivities of weak electro-
lytes. One has to consider the degree of dissociation of an electrolyte, α, and its
dissociation constant, Ka. The interested reader is encouraged to get a deeper under-
standing of the interrelations between the Debye‒Hückel theory (ions are immobile)
and the Debye‒Hückel‒Onsager theory (ions are moving in the external electric
field) for both types of electrolytes, by reading some basic electrochemistry text-
books referred to at the end of this chapter [1].
48 1 Basic concepts

1.3 Nonaqueous electrolytes

Until now, we were discussing primarily the aqueous electrolyte solutions that are ob-
tained by the dissolution of salts in water. In such systems, the energy necessary for
the dissociation of salt into ions was provided primarily by the solvent–salt lattice in-
teractions. However, the lattice dissociation into ions can be achieved also by provid-
ing heat: the salt can be melted down without any additional component acting as a
solvent. Such a system is called molten salt for a temperature well above 100 °C or
ionic liquid (IL) if the melting temperature is around or even below ca. 100 °C. Any-
way, as time goes by, definitions of both liquids tend to overlap with each other. Cur-
rently, however, the term molten salts is usually applied to melts of inorganic
compounds, composed of cations and anions, being in a solid state at room (and
slightly above) temperature. With an increase in temperature, the short-range and
long-range ordering of solids is overcome by thermal energy, preserving only the
long-range ordering in a molten, liquid state. In this state the ions possess certain mo-
bilities, keeping approximately constant density of liquid due to the cohesion forces
between the constituent particles. However, before going further, we have to stress
that the theories describing the structure of the interface between the electrode and
aqueous electrolyte, as well as those referring to the conductivity behavior of the ion‒
ion and solvent‒ion interactions cannot be applied directly to ionic liquids and mol-
ten salts. First of all, there is no solvent involved in dissociation and solvation. Sec-
ondly, the “infinite dilution” limit and dimensionless charge approximation (point
charge limit) that we were using in Section 1.1.1 and in describing the structure of
aqueous electrolyte, do not hold in the case of ILs and molten salts. The sizes of ions
are not to be ignored (the so-called excluded volume), particularly for bulky constitu-
ents of ionic liquids, nor the high viscosity of such electrolytes, affecting their con-
ducting properties. The interested reader is referred to the literature at the end of this
chapter, also for the structures of most common ILs [2, 5].

1.3.1 Ionic liquids

Exploration of environmentally friendly green solvents as alternatives to aqueous,

high-temperature molten salts or volatile organic compounds in electrochemical en-
ergy storage, and electrolytic processes has been persistently pursued. A class of elec-
trolytes, called ionic liquids (ILs) is widely studied from the point of view of replacing
aqueous electrolytes for such purposes. Ionic liquids are typically defined as com-
pounds totally composed of ions in their pure forms, meaning that they are overall
neutral assemblies of charged particles. They are characterized by melting points
below or around the boiling point of water under ambient pressure. Their hybrid or-
ganic-ionic behavior and intermolecular interactions give rise to a set of complex phe-
nomena that are beyond the scope of this book. Therefore, here we will briefly
 1.3 Nonaqueous electrolytes 49

outline only those features that are directly related to the electrochemical aspects
of ILs.
Ionic liquids (ILs) possess extraordinary properties such as nonflammability, and
relatively high ionic conductivity at a level of ca. 10–30 mS/cm, low toxicity, the wide
electrochemical window of ca. 4 V, and electrochemical and thermal stability; all
these features make them ideal candidates for electrolytes in electrochemical devices
like batteries, double-layer capacitors, fuel cells, photovoltaics, actuators, and other
electrochemical devices [2–4].
In addition, ILs have been widely used in electrodeposition, electrosynthesis, elec-
trocatalysis, and electrochemical capacitors, making them ideal solvents for a wide
range of electrochemical applications. Since ionic liquids contain only charge carriers,
means that when they are used as solvents, they also act as supporting electrolytes,
minimizing costs and wastes by reducing the use of hazardous and polluting organic
solvents for a greener environment. The use of ILs in so many different applications
is determined by their intrinsic properties. There are many feature reviews dealing
with different aspects of ILs, in all these areas of applied electrochemistry and the
reader is directed to these excellent publications [2–5]. Of particular interest are ILs
existing in the liquid phase around room temperature. A typical room temperature IL
with a bulky organic cation (e.g., alkylammonium, N-N-dialkylimidazolium, alkylimi-
dazolium, N-alkylpyridinium, and alkyl‐phosphonium) is weakly coordinated to an or-
ganic or inorganic anion, such as BF4−, Cl−, I−, PF6−, AlCl4−, etc. The ionic conductivity
of ILs (at least ten times smaller than that of conventional aqueous electrolytes) is
highly dependent on temperature (as is the case for all solvents) because as viscosity
increases, the charge carriers must overcome a greater frictional force, decreasing
their mobilities, and, as a result, conductivities. These relations can be represented in
Arrhenius-type plots, but instead of a straight line, ILs show a certain curvature that
can be explained by the temperature-dependent changes in the structural properties
of bulky ions. Nonetheless, the choice of appropriate cation and anion of the ILs can
alleviate to some extent such dependency, improving their transport properties, ad-
justing also the potential window of IL’s stability and its capability to dissolve salts
with metallic cations for, for example, metal-ion batteries using ionic liquids as sol-
vents. It is also worth mentioning that the specific conductivity, κ, of conventional
strong electrolyte solutions is proportional to the number of charge carriers. How-
ever, the assumption of ILs consisting entirely of ions as charge carriers does not hold
true. Rather, as in the case of the strong aqueous KCl electrolyte specific conductivity
graph shown in Fig. 1.21A, where after reaching maximum value a decrease in con-
ductivity is observed with increasing concentration, one should expect, that IL ions of
opposite sign may form relatively stable aggregates of nullified resultant charge. If so,
these aggregates will not participate in the transfer of charges. Such ion-ion interactions
may vary for different ionic liquids and their mixtures. Therefore, it is difficult to define
the number of charge carriers by simply knowing the IL chemical structure. Literature
data estimation leads to the conclusion that for many commonly used RTILs (room-te-
50 1 Basic concepts

meprature ionic liquids), the degree of ion pair dissociation at room temperature varies
between 50% and 70%.
The features of RTILs that make them special as electrolytes can be summarized
below:
– RTILs are electrolytes without solvents that stay liquid in a broad temperature
interval around the room temperature.
– The ions of RTIL are typically large organic ions (at least one of them), containing
a substantial amount of hydrophobic constituent functionalities. These features
reduce the freezing temperature of the ILs, because the Coulomb interactions are
weaker in the range of their action, whereas their complicated shape hinders
reaching ordered crystal structures.
– If RTILs are composed of less reactive ions, they are more inert electrochemically
compared to many standard electrolyte solutions; therefore they can sustain
higher voltages (larger potential window).
– Various RTILs can be mixed with each other to obtain desired electrochemical be-
havior, such as lower viscosity and thus increased ion mobility.

1.3.2 Molten salts

Generally speaking, molten salts consist of cations and anions with various valencies,
being regarded as extremely concentrated solutions or solutes without solvents (such
as high-temperature ILs) at elevated temperatures and include alkali halides and salts
such as nitrates, sulfates, and carbonates. Currently, this is often expanded to com-
prise silicates, borates, and organic salts. Of course, one can easily find many more
potential candidate cations and anions for molten salt; the only limitation in such a
creative approach is their melting temperature and conductivity. If binary and ter-
nary systems are considered, this number is, in fact, quite large. Molten salts dissoci-
ate completely into ions and therefore they are considered strong electrolytes. The
high mobility of the ions at elevated temperatures results in high ionic conductivity.
Therefore, molten salts are good conductors exhibiting 10–50 times higher conductiv-
ity than the aqueous electrolyte or ionic liquids. This in turn results in a remarkable
decrease in the internal resistance of the electrochemical power sources and storage
devices described later in this book (Chapter 8). Additionally, molten salt electrolytes
are characterized by wide electrochemical windows, low raw material costs, and non-
flammability. The features used to characterize molten salts are mostly the same as
those of ionic liquids mentioned above, except for the ions that are mainly inorganic
with no additional chemical functionalities, and therefore a thermal energy is re-
quired to overcome the lattice association forces solidifying the electrolyte structure.
However, the need to keep the high temperature of a device utilizing the molten salt
as an electrolyte, requires special construction of such a device, additional energy lost
for heating, as well as some safety issues and thermal management related to the re-
 1.3 Nonaqueous electrolytes 51

gime of high-temperature operation. This may lead to increased weight and cost. Nev-
ertheless, the elevated temperatures required for electrochemical devices utilizing
molten salts can be advantageous for operation in harsh environments, such as metal-
lurgy, the oil and gas industry, internal combustion engines, hybrid transportation ve-
hicles, or even space exploration.

1.3.3 Solid electrolytes

Since there has been a continuously growing demand for more efficient, reliable, and
environmentally friendly materials for use as electrolytes, particularly for solid-state
energy storage systems and environmental sensors described later in this book, solid
electrolytes (SEs) are considered one of the key components for these systems. Their
advantages over the electrolytes described above are thermal and chemical stability,
lower toxicity, wide electrochemical window, and good mechanical properties. How-
ever, the main issue is the ionic conductivity, especially at ambient temperatures.
Solid electrolytes are generally classified into two main groups: inorganic electrolytes
(IE), and polymer electrolytes (PE). Further, the inorganic electrolytes incorporate ox-
ides, and sulfides (both amorphous and crystalline), whereas the polymer electrolytes
include, for example, polymer-salt complexes, gel polymer electrolytes, and compo-
sites. Comparing their conductivities, the inorganic electrolytes provide much higher
ionic conductivities (up to 10–2 S/cm for sulfide-based electrolytes), comparable to
those of organic liquid electrolytes, whereas polymer electrolytes show conductivities
still not sufficient for their subsequent wide applications. Therefore, we will focus
below on the conductivity behavior of inorganic electrolytes, outlining the migration
of ions in a solid electrolyte at different scale lengths: from the atomic to a macro-
scopic scale.
Atomic scale. At the atomic scale, mobile ions, in the case of oxides and sulfides –
cations (e.g., Li+, Na+, or Mg2+) diffuse in solids along the lowest energy migration
pathways and can be envisioned as ion hoping between stable sites and/or intermedi-
ate metastable sites of the crystalline framework of anions (e.g., O2−, S2−). In a crystal-
line case, cationic vacancies and interstitial sites are also considered mobile-charged
species. Three main migration mechanisms can be distinguished: i) ion migration into
a neighboring vacant site (vacancy migration in the opposite direction, ii) interstitial
direct migration between not fully occupied sites, and iii) concerted mechanism in
which the migrating ion from the interstitial site displaces its neighbor in the lattice
into the next metastable cation site. This is shown graphically in Fig. 1.22 below
The conductivity, σ; of an ion in a solid electrolyte can be defined as:

σ=q×n×u (1:131)

Here, q is the charge, n is a concentration of mobile charge carriers in a solid, and u is

their mobility defined as before. We also should notice, that typically only a fraction
52 1 Basic concepts

lon migrates to neighboring Interstitial direct migration Concerted (displacing) migration

vacancy

Occupied stable cation Occupied metastable

site cation site
Unoccupied stable Unoccupied metastable
cation site cation site

Fig. 1.22: Ion migration pathways in a crystal lattice of a solid electrolyte.

of total charges is mobile, thus a precise determination of n is at least ambiguous. As

can be easily understood, ion conduction is a thermally activated process, usually de-
scribed by a modified Arrhenius equation:
− Ea
σ = σ 0 × T m e kB T , (1:132)

where σ 0 is the pre-exponential factor, m is typically equal to − 1, kB is the Boltzmann

constant, T is temperature and Ea is a characteristic activation energy for ion conduc-
tion. In the simplest case, Ea includes the energy of forming the mobile defects and
energy for their migration (highest energy barrier along the conduction path). Of
course, the charge transfer, regardless of the type of electrolyte (liquid or solid) is
linked to the mass transfer. In the case of solid electrolytes, it is also useful to define
the diffusivity of charges, D, according to the relationship similar to eq. (1.72):
u
D= kB T; (1:133)
q

and therefore, after substituting into the equation for conductivity:

Dq2
σ= ×n (1:134)
kB T
These basic equations can be utilized to understand and describe simple, monocrys-
talline materials. However, solid electrolytes currently used in various electrochemi-
cal devices, show an amorphous character, at least partially. Therefore, various
concepts and theories have been developed for the description of such materials.
Some of the problems that have to be accounted for are outlined below.
 1.4 Membranes and membrane potentials 53

Macroscopic scale. At a macroscopic scale, the information on the ionic conductivity

of solid electrolytes comes mainly from the impedance spectroscopy measurements
(see Section 2.2). These measurements are sensitive to structural features, however,
they provide only the total ionic conductivity. Resolving the contributions of each pos-
sible component to the obtained results requires some empirical hypotheses, careful
modeling of each possible charge transfer mechanism, and fitting procedures. Despite
continuous efforts, a theory for conduction in amorphous solids is not yet satisfacto-
rily established. This is mainly because compositional and/or structural inhomogenei-
ties, including impurities, can dominate the macroscopic ionic conductivity. The
obvious examples of such inhomogeneities in polycrystalline materials are grain
boundaries between identical crystals or contact surfaces between crystals of differ-
ent orientations or misaligned grains. Another feature that strongly affects ionic con-
ductivity in solid electrolytes is poor physical contact between neighboring solid
particles, leading to porosity and inhomogeneous current densities, detrimentally af-
fecting the conductivity of the sample. To get better insight into the fundamentals of
solid-state electrolytes, the interested reader is referred to [6, 7].

Bibliography

[1] Bagotsky VS. Fundamentals of electrochemistry. Wiley & Sons Inc., 2006.
[2] Fedorov MV, Kornyshev AA. Ionic liquids at electrified interfaces. Chem. Rev. 2014, 114(2), 978–2036.
[3] Lee CP, Ho KC. Handbook of ionic liquids, properties, applications and hazards. New York, Nova Sci
Publishers, Inc., 2012.
[4] Matsumoto H. Electrochemical aspects of ionic liquids. Ohno H, ed. Hoboken, NJ, John Wiley & Sons,
Inc., 2005.
[5], Dong K, Liu X, Dong H , Zhang X, Zhang S. Multiscale studies on ionic liquids. Chem. Rev. 2017, 117,
6636–6695.
[6] Famprikis T, Canepa P, Dawson JA, Saiful Islam M, Masquelier C. Fundamentals of inorganic solid-
state electrolytes for batteries. Nat. Mater. 2019, 18, 1278–1291. doi: 10.1038/s41563-019-0431-3.
[7] Karabelli D, Birke KP, Weeber M. Performance and cost overview of selected solid-state electrolytes:
Race between polymer electrolytes and inorganic sulfide electrolytes. Batteries 2021, 7, 18. doi:
10.3390/batteries7010018.

1.4 Membranes and membrane potentials

1.4.1 Equilibrium potentials

The formation of a potential drop across metal|electrolyte solution interphase – the

electrical double layer, can be considered a straightforward example of equilibrium
potential drop. This is because the movement of electrons in the metal phase and ions
in the electrolyte is so fast that any perturbation introduced to the system, such as
polarization change, will readjust the structure of the e.d.l. much faster than the diffu-
54 1 Basic concepts

sion-controlled electrochemical process (see Scheme 1.1 and its description in the
text). Even if the electrode process is irreversible, the substrate and product species,
by virtue of their concentrations, do not affect significantly the electrical double-layer
structure – so it remains essentially at equilibrium.
Apart from this situation, we can imagine and build systems in which, even
though there is no external polarization or redox reactions, the potential drop across
an interface will be formed. The key word here is the “interface.” Thus, let us take
into consideration a system, in which there are two aqueous electrolytes with equal
concentrations of KCl (or other salt, acid, or hydroxide) that are separated by a semi-
permeable membrane. The ions of this salt can move freely through the membrane,
so in this particular case, such a membrane is called a diaphragm. On the contrary, a
semipermeable membrane means that it allows only certain ions to pass through it,
but other ions cannot pass through such a membrane. The membrane can be porous
and its semipermeability (permselectivity) may stem from the structural features,
such as the functional groups embedded inside its matrix, the pore size, or the charge
of the membrane surface and pores. Well-known examples are the dialytic mem-
branes used in clinics and laboratories. So, let us put them to work, as in Fig. 1.23.

I II

KCl KCl

K+n Xn-

Xn-

Fig. 1.23: The semipermeable membrane separating two aqueous solutions of KCl of initially equal
concentrations in compartments I and II. The initial equilibrium condition is established. Next, the
potassium salt of a polyvalent anion, K+nXn−, is added to compartment II. See text for details.

We have already discussed above the equilibrium conditions under constant p, and T.
This was done, for example, in the case of electrons in the contacting metal 1 and
metal 2, eq. (1.37). This condition is valid also for any ionic species “i” capable of pene-
trating the semipermeable membrane. For these, we can write for our KCl salt:

μei I = μei II (1:135)

Using the definition of electrochemical potential, we will have for both potassium and
chloride ions on both sides (I and II) of the semipermeable membrane:

μK +I + zK + FφI = μK +II + zK + FφII (1:136)

 1.4 Membranes and membrane potentials 55

and for anions:

μCl−I + zCl − FφI = μCl−II + zCl − FφII (1:137)

Introducing next the definition of chemical potentials for both ionic species and keep-
ing in mind that zK + = 1, whereas zCl− = −1, we will have after some regrouping:

RT cIK + RT cIICl−
φII − φI = ln = ln ; (1:138)
F cIIK+ F cIICl−

or, more generally

RT cIIi
φII − φI = − ln (1:139)
zi F cIi

Please note, that because there are two aqueous solutions on both sides, therefore, the
standard chemical potentials of each ionic species, μi 0 , are the same on both sides and
therefore they cancel out. However, if we consider the interface between the two im-
miscible solutions of the same salt that can be dissolved in both solvents (e.g., water|
nitromethane system), we cannot state that the standard potential of species “i” is the
same in both solvents. Therefore, the μi 0;I ≠μi 0;II; and the above equation should in-
clude also a term taking into account such a situation. The difference in standard
chemical potentials defines the partition coefficient, Kp of the extraction process fre-
quently used in purification procedures:

μ0;II − μ0;I RT cIIi

φII − φI = − − ln (1:140)
zi F zi F cIi

μ0;II − μ0;I
Kp = − (1:141)
zi F

Anyway, in our system of aqueous solutions separated by a semipermeable mem-

brane, the above equilibrium equation is called the Donnan equation. Since the KCl
aqueous concentrations are equal on both sides of a membrane, the resultant poten-
tial difference is zero.

CKI + = CCl
I
−

CKII+ = CCl
II
−

and

φII − φI = 0

This initial situation will change if we introduce to the right-hand side compartment a
given concentration of potassium salt of multivalent, large anion, such as polymeric
56 1 Basic concepts

anion or protein, K+nXn−, whose anion Xn− is large enough to be blocked by the mem-
brane. For this ion, the membrane is impermeable, whereas K+ and Cl− still can move
through the membrane. The electroneutrality condition will require the influx of K+
ions with simultaneous efflux of Cl− ions from compartment II, to achieve finally:
n −, II
CKII+ = CCl
II
− +X

From this, one can easily get that CKII+ > CKI + ; and CCl
II
− < CCl − .
I

Therefore, the Donnan equilibrium potential established across a semipermeable

membrane in such a system will be negative, using the notation: φII − φI < 0.
Of course, mirroring the system from Fig. 1.23, that is, adding the salt K+nXn− to
the left compartment (instead of the right one) will result in a positive value of the
Donnan potential. Please note that in the discussion above, the structure of a semiper-
meable membrane is unimportant. Its role is to mechanically separate polyanion Xn−
from its penetration through the membrane.
Let us go now a step further. What happens right after introducing K+nXn−, but
before the equilibrium potential is established? This situation is described below.

1.4.2 Nonequilibrium potentials

For such a situation, we should use the full form of the Nernst‒Planck equation, solv-
ing it with the help of Faraday’s law under the electroneutrality conditions, for
charged species of valence zi ≠ 0, able to move across the membrane

dci dϕ
Ji = −ui RT − ui ci zi F (1:142)
dx dx
dci dφ
ji = zi FJi = −zi ui RTF − ui ci z2i F 2 (1:143)
dx dx

Electroneutrality requires that the ionic current passing across the membrane at any
time equals zero, therefore:
X X
Ji = ji = 0 (1:144)
i i
X dlnci X dφ
− RTF zi ui ci − F2 z2i ui ci =0 (1:145)
i
dx i
dx

After regrouping:
X dlnci X dφ
RTF zi ui ci = − F2 z2i ui ci (1:146)
i
dx i
dx
 1.4 Membranes and membrane potentials 57

If we now assume that the potential drop across the membrane is the dependent vari-
able (one can equally well solve it with concentration profile as the dependent vari-
able), we will get:


dφ RT X dlnci X
=− zi ui ci z2i ui ci (1:147)
dx F i
dx i

Let us now define the so-called ionic transference number (ion transfer number),
ti, as:

jzi jUi ci
ti = P ; (1:148)
j jzj jUj cj

which describes how much charge is carried out by a given ion, “i,” compared to the
sum of all other ions “j” present in the solution, or, in other words, what is the frac-
tion of the total electric current carried in an electrolyte by a given ionic species “i.”
Here Ui = |zi|Fui is called electrolytic mobility.
After integration of the above differential equation from side I to side II, and as-
suming constant values of ci and ti, being independent of distance from the mem-
brane, for the univalent electrolyte we can obtain:


 II
RT t+ t c
φ − φ = ΔφL = −
II I
+ ln I (1:149)
F z+ z− c

or, keeping in mind the negative sign of z− ; in the case of monovalent ions:

RT cII
ΔφL = − ½t+ − t− ln I (1:150)
F c
The above equation, defines the so-called liquid junction potential, ΔφL , and tells us
that the value and sign of the potential developed across the semipermeable mem-
brane depend on the difference between the ionic transference numbers (or their mo-
bilities) in given electrolytes. If cII > cI and cations are moving faster across the
membrane compared to anions (t+ > t− Þ, the value of ΔφL will be negative. Moreover,
the properties of such a membrane are irrelevant according to this approach. How-
ever, most of the membrane separators used in the construction of electrochemical
batteries, fuel cells, and electrochemical sensing devices require separators whose
performance strongly depends on their structure and are specially tailored for their
applications.
Let us have a closer look at the interface between the aqueous solutions II and I,
differing in their composition. It seems logical, that a transition layer will develop
within the separating membrane, within which the concentrations of each component
“i” can exhibit a smooth, linear change from their values, ci;o and ci;i (see Fig. 1.24):
58 1 Basic concepts

ci,x = ci,o + (ci,i – ci,o )x/dm

I o i II

cx

dm

Fig. 1.24: A scheme of a permselective membrane of thickness dm, separating compartments I and II.
Linear concentration gradient is assumed between sides “o” and “i” of this membrane (blue arrow), with
two Donnan equilibrium potentials developed across both interfaces (green arrows). An ion exchange
reaction between bound and free ions within the membrane is also shown (red arrow). Other
mechanisms can also be involved in the performance of such a system. See the text for a detailed
description of this scheme.

ci, x = ci, o + ðci, i − ci, o Þx=dm : (1:151)

Here, ci, x is the concentration of “i” at point x inside a membrane, dm is the thickness
of a membrane, whereas ci, i and ci, o are the concentrations of “i” at the membrane
surface “i” and “o.” This summarizes the so-called Henderson approach, and intro-
duced to eq. (1.150) for univalent electrolytes on both sides of such a membrane, yields
the Lewis‒Sargent equation:
RT u + + u −, II RT ΛII
ΔφL = + ln = ln (1:152)
F u + + u −, I F ΛI

However, this approach does not account for the role of membrane structure in its
selectivity. Here, we may want to recall a rectifying behavior of a nanoporous mem-
brane containing charged nanochannels. Because of the presence of the electrical
double layer due to the charges on the channel walls, if such membrane separates
two electrolytes, at sufficiently small channel diameters, the co-ions will be repelled
from the channels, allowing only the flow of counterions through the channel, thus
providing very high selectivity of ionic conductance for such separator (see Section
1.1.1, Fig. 1.8)
Let us now go a step further and consider a membrane of a given thickness with
its chemical structure capable of exchanging ions between the membrane constitu-
ents and only selected ions from the aqueous electrolyte solutions. These are the so-
called ion exchange membranes. The unique properties of ion exchangers can be ex-
plained by their structure – a framework carrying functional groups bearing positive
or negative charge, compensated by ions of opposite charge (counter-ions). The
counter-ions can move freely, and therefore they can easily be replaced by other ions
with the same sign and at least similar affinity to the functionalities of the structural
 1.4 Membranes and membrane potentials 59

framework. Depending on the interactions and affinities between the functional

groups and counter-ions, important characteristics of an ion exchange membrane can
be designed – namely – its selectivity, defined as the ability of an ion exchange mem-
brane to distinguish between different species of counter-ions.
The best-known and used example of such a system is a glass, pH-sensitive elec-
trode. It is beyond the scope of this book to describe in detail the functioning of glass
electrodes as pH-sensing (proton exchange) devices. The interested reader is referred
to the literature examples at the end of this section [1, 2]. Here, we will only outline
processes involved in the functioning of a tiny, glassy, noncrystalline solid membrane,
leading to the potential development responsive to the concentration of hydrogen
ions in the solution. This solid, based on a silicon oxide framework, is doped with a
relatively high concentration of alkali and/or alkaline metal oxides. The metal ions
possess significant mobilities within the framework, whereas the negatively charged
functional groups are bound within the silicon oxide framework, and therefore, their
mobility equals zero. In contact with water, the glass hydrates with metal ions quanti-
tatively being exchanged for hydrogen ions. Therefore, the following system and pro-
cesses can be exemplified in the case of a typical “sodium” glass membrane, and Na+,
and H+ ions present in an aqueous solution:

Phase I o membrane i Phase II

Na+ (cNa+ ;I ), H+ (cH + , I )|Na+(cNa), H+ (cH)|Na+(cNa+ , II ), H+ (cH + , II )

Both types of ions can diffuse into the membrane with different mobilities uNa and
uH. The membrane has a defined number of negatively charged, structural sites,
where Na+ and H+ are the counterions and can be bound to these sites. The total po-
tential developed across such a system is a sum of contributing potential differences:
   
ΔφM = φII − φI = φII − φi + ðφi − φo Þ + φo − φI (1:153)

Moreover, the equilibria between ions present in the solutions and those bound in the
membrane can be described by two equilibria: the Donnan equilibria between the
ions in the solutions and those free, unbound within the membrane, and the ion ex-
change equilibrium between bound and free ions within the membrane.
The Donnan equilibria:





RT cNa+ , II RT cH + , II
φ − φi = −
II
ln =− ln (1:154)
F cNa+ , i F cH + , i





RT cH + , I RT cNa+ , I
φo − φ =
II
ln = ln (1:155)
F cH +, o F cNa + , o
60 1 Basic concepts

where cNa + , i , cH + , i , cH + , o , and cNa + , o are the concentrations of free ions in the mem-
brane, respectively, whereas cNa+ , II , cH+ , II , cH + , I , and cNa + , I are the concentrations of
these ions in solutions II and I (phase II and I), respectively.
The ion exchange equilibria of the type Na+ (free) ↔ Na+ (bound) and H+ (free) ↔
+
H (bound) are:

KNa+ = CNa+ , o =cNa+, o = CNa+ , i =cNa + , i , (1:156)

KH + = CH + , o =cH + , o = CH + , i =cH + , i , (1:157)

where CNa+ , o , CNa + , i , CH + , o , and CH + , i denote the concentrations of bound ions, assum-
ing their ideal behavior. Moreover, the sum of CNa+ and CH + is constant as the constant
is the number of ion exchange sites in the membrane (negatively charged, structural
sites).
The ratio KNa+ =KH + is obviously the equilibrium constant of the ion exchange
reaction:

Na+ ðfreeÞ + H+ ðboundÞ $ Na+ ðboundÞ+H+ ðfreeÞ

Now we have to describe the potential difference inside the membrane, between
sides “i” and side “o” (φi − φo Þ. This is the diffusion potential and we will use again the
Nernst‒Planck equations:



dCNa+ dφ
JNa + = −uNa+ RT − uNa+ CNa+ F (1:158)
dx dx



dCH + dφ
JH + = −uH + RT − uH + CH + F (1:159)
dx dx

Since we assumed that there is no net current passing through the membrane and
that the mobilities of structural anions (functional negative groups of the glass ma-
trix) are equal to zero, u − = 0, we can write that JH + + JNa+ = 0. Therefore:


RTdðuNa+CNa+ + uH + CH + Þ dφ
+ ðuNa+ CNa+ + uH + CH + ÞF =0 (1:160)
dx dx

Integrating the above equation from x = o to x = i, we will finally get the diffusion po-
tential inside the membrane:

RT uNa+ CNa+ , i + uK + CH + ,i
φi − φo = − ln (1:161)
F uNa+ CNa+ , o + uK + CH + , o

Now we are ready to substitute all terms in eq. (1.153) for ΔφM with eqs. (1.154), (1.155),
(1.161), and ion exchange equilibria (eqs. (1.156) and (1.157)). This will yield the follow-
ing relation for the membrane potential, ΔφM :
 1.4 Membranes and membrane potentials 61

RT uNa+ KNa+ cNa+ ,II + uH + KH + cH + , II

ΔφM = − ln (1:162)
F uNa+ KNa+ cNa+ , I + uH + KH + cH + , I

By dividing the nominator and denominator of this equation by uNa+ KNa + , and setting
uH + KH +
u + K + = KH =Na , we get the so-called Nikolsky‒Eisenman equation:
+ +
Na Na

RT cNa+ ,II + KH + =Na+ cH+ ,II

ΔφM = − ln (1:163)
F cNa+ ,I + KH + =Na+ cH+ ,I

KH + =Na+ is called the selectivity constant of an ion-selective membrane for ions H+ ver-
sus ions Na+ as in our case, or in general, for ions of one type with respect to ions of
cH + , I
different species. In the case of large KH + =Na+ and sufficiently high, we can get
cNa+ ,I
Nerstian slope of ΔφM as a function of logcH + ,I :
2:303RT cNa+ ,I
ΔφM = const: + logcH+ ,I ðKH+ =Na+  Þ (1:164)
F cH+ ,I

The membrane is then specifically selective to H+ ions. Thus, using the example of
glass and pH electrodes, we were able to explain the selectivity behavior of ion ex-
change membranes. Of course, depending on the design and purpose of such mem-
branes, the mechanisms involved in their performance will differ. The interested
reader is directed to the literature already mentioned above [1, 2].
Here, we should mention other types of solid-state membranes widely used in in-
dustrial electrochemical applications, energy conversion, and storage technologies,
and environmental monitoring systems. Examples of such membranes include metal
oxides, metal sulfides, and mixed systems. Among metal oxides with high ionic con-
ductivity, a prerequisite for their applications as all-solid-state electrolytes described
in the previous chapters, zirconia-based membrane separators (ZrO2) are widely uti-
lized. For better stability and electrochemical performance zirconium oxide is fre-
quently stabilized with yttrium (YSZ, yttria-stabilized zirconia) or scandium oxides
(SSZ, scandia-stabilized zirconia). In these solid crystalline structures, the mobile
charges are oxygen dianions in the crystalline network. At sufficiently high tempera-
tures, the mobile oxygen ions are responsible for charge transfer within the solid-
state membrane and the overall system separating two compartments of, for example,
partial oxygen pressures pOI2 and pOII2 ; can be outlined in a way similar to that al-
ready shown above:

Compartment I o membrane i Compartment II

O2 (pOI2 , gas) |YSZ(or SSZ)mobile O2- ions| O2(pOII2 , gas)
62 1 Basic concepts

In this case, the developed membrane potential will be given by:

RT pOII2
ΔφM = t ln
2nF pOI2

where t is an average ionic transference number (eq. (1.148)), n is the valence of gas
being responsible for potential development (in this case oxygen with 4e redox reac-
tion), and other parameters have their usual meanings.
Let us now summarize the picture that emerges from the above section, from the
point of view of utilizations of such ion-exchange/ion conducting membranes in ap-
plied electrochemistry, including these applications that are involved in sustainability
and a cleaner environment (“green” electrochemistry).
1. First, the ion exchange processes can be driven by concentration gradients, an
electric field applied across the membrane (electrodialysis), or both. In such situa-
tions, the ions are not only exchanged but also cross the membrane.
2. Depending on their applications, the semipermeable membrane can be inorganic
or organic ion-conducting solids (e.g., in fuel cells, electrolyzers), organic-inor-
ganic composites, metal-organic frameworks, hydrogen-permeable metal mem-
branes, but sometimes also a thin layer of supported solution separating two
electrolytes (e.g., in ion-selective sensor electrodes).
3. When the membrane separates two compartments, it can work continuously.
This is different from processes used in analytical applications and separations,
where the ion exchange column or bed operates discontinuously with the need
for regeneration.
4. In the case of membranous systems, some synthetic applications are possible, such
as the recovery of compounds from wastewater or desalination of salt water.
5. When such a membrane is placed directly between the anode and cathode in an
electrolyzer or a battery it acts as an electrolyte, sometimes the only electrolyte
between electrodes.
6. Power can be generated in fuel cells by supplying suitable substrates/reactants to
the electrodes separated by such a membrane.
7. Power can be stored when the two electrodes separated by such membrane are
charged (rechargeable batteries, capacitors, supercapacitors, and hybrid cells).
8. These membranes can be used for various types of sensors for real-time control
of the environment, industrial processes, and pollution detection/prevention (Part
IV, Chapter 13).

Among the above applications, fuel cells and industrial electrolysis for hydrogen and
oxygen “green” generation are the most known and demanding for the ion exchange
membranes from the point of view of low electrical resistance, mechanical and chem-
ical resistance to pressure, temperature, and other operating conditions. The design
of each membrane’s characteristics strongly differs per its application, due to the
 Bibliography 63

working conditions of the device utilizing such a membrane. Lately, the so-called bi-
polar membranes (BPMs) are gaining more interest, particularly in energy conversion
and storage technologies. These membranes are composed of cation- and anion-ex-
change layers laminated together with the abrupt transition of ion-exchange behavior
in an interfacial layer. The BPM provides conditions to operate in different electro-
lytes on either side, whereas the interfacial layers provide space for nanostructured
catalysts for energy conversion and storage processes. The examples of applications
of membranous systems designed for various domains of electrochemistry, including
“green” electrochemistry, will be given later in this book.

Bibliography

[1] Ion-selective electrodes: Overview; Eric Bakker. In: Encyclopedia of analytical science (third
edition). 2019.
[2] Głąb S., Hulanicki A. Ion-selective electrodes: Glass electrodes. In: Reference module in chemistry,
molecular Sci. Eng. 2013.
2 Selected electrochemical methods applied
in analytical chemistry and material science

Applied electrochemistry utilizes numerous methods in its efforts to improve progres-

sively our quality of life through the understanding of charge transfer phenomena.
These efforts lead to the construction and design of new devices and systems that can
be used in industry (e.g., electrical energy storage and supply, catalysis and corrosion),
and also in more personalized applications, such as batteries for smartphones, laptops,
and solar panels. Electroanalytical techniques provide unique information on mecha-
nisms and systems that are currently involved in the necessary development of the
above areas. Both the instrumentation and theoretical fundamentals have been ad-
vanced such that nonspecialists can easily use them in search of further development
in technology, industry, and everyday life. Below we lay a brief background of selected
electrochemical methods, commonly used in material science for:
– characterization of interfaces;
– characterization of electrode processes, kinetics, mass transfer;
– corrosion studies;
– coatings and paints;
– investigations of membrane transport phenomena;
– electrocatalysis, adsorption, and electrosorption studies;
– studies of cells, batteries, fuel cells, supercapacitors;
– conductive polymers;
– semiconductors;
– photoelectrochemistry and others.

2.1 Transient methods

Before we start discussing selected electrochemical methods in this chapter, we have

to summarize briefly some basics of the design of electrochemical cell that is typically
used for such studies, as well as the electrode response to the applied potential, E, or
introduced charge, Q, in terms of the measured current, i. First of all, since we already
discussed that at the electrolyte/solution interface the electrical double layer is
formed, accumulating and separating ionic charges, it behaves as a condenser. This is
because if one polarizes the condenser with potential, E, it will accumulate/separate
charges, Q, proportionally to its capacitance, C:

Q=C× E (2:1)

https://doi.org/10.1515/9783111160986-002
 2.1 Transient methods 65

Assuming the constant value of C and differentiating the above versus time, we
can get:

dQ dE
= ic = C × (2:2)
dt dt

Since the derivative of Q versus time is current, this equation tells us that any change
 
in potential in time dE
dt will result in charging current, ic, passing through the electro-
chemical cell. This current overlaps and adds to the so-called Faradaic current, being
the result of the charge transfer reaction at the electrode. Therefore, it is necessary to
separate these two currents in order to gain insights into the mechanisms of charge
storage (e.g., supercapacitors) and charge transfer (e.g., electrocatalysis and electrode-
position). For now, it is sufficient to keep this in mind, but we will return to this in
more detail later in the text.
Thus, in the absence of a charge transfer reaction (ideally polarizable electrode),
the current flow results in charging (or discharging) the capacitance of the electrical
double layer and is also called the capacitive current.

2.1.1 Capacitive current

The electrochemistry in material science usually begins in the electrochemical cell.

Its design is dictated by the organization and the number of electrodes used, as well
as their purpose. Typically, there are three electrodes: working electrode (WE), ref-
erence (Ref) electrode, and the counter/auxiliary electrode (CE). The CE should not
affect the investigated flow of current, therefore the area of this electrode should
be much larger than that of the working electrode. The working electrode is polar-
ized against the Ref of constant potential, independent of the process being investi-
gated on the working electrode. The current flow (both charging and Faradaic) as a
response to the WE polarization/charging is measured between the WE and CE elec-
trodes. Because the capacitance of other components of an electrochemical cell is
much larger, they do not contribute to the investigated current flow. Therefore, this
type of situation can be represented by an electrical equivalent circuit, containing
the capacitance of the e.d.l. of WE electrode, Cdl, in series with the ionic (ohmic)
resistance of the solution, in which all three electrodes are immersed, Rs. Now we
can introduce the relationships between the electrode potential and capacitive cur-
rent, ic, for ideally polarizable electrode in the three most frequently used analyti-
cal methods as follows:
(a) Potential step (chronoamperometry), assuming that at time = 0 there is no charge
at the electrode surface:
66 2 Selected electrochemical methods


E
ic = expð − t=Rs Cdl Þ (2:3)
Rs

This equation tells us, that after applying a potential step of magnitude, E, an expo-
nentially decaying current will flow with its time constant. This is observed as current
“spike” within the first few tenths of a second after application of a given potential
value.
(b) Current step, assuming that at t = 0, i = 0 and then the RC circuit is charged by a
constant current, ic, giving rise to the potential E:


t
E = ic Rs + (2:4)
Cdl

This relation allows for the determination of Cdl. This is also the principle of the deter-
mination of chemisorbed hydrogen on platinum or other catalysts by its oxidation.
(c) Potential scan (linear scan), where the potential is scanned from its initial value,
Ei, to a final value, Ef, with a rate v = dE
dt over the time t (the final potential value is
determined by the duration of scan, t):


 
E
ic = vCdl + − vCdl expð − t=Rs Cdl Þ (2:5)
Rs

It is easily seen that the current response for the ideally polarizable electrode (no
Faradaic current) contains two components: a steady-state, vCdl , and a transient
one. Proportionality of ic to v enables determination of not only the capacitance of
e.d.l. but also any charge accumulated at the surface, such as electrochemically ac-
tive surface layer, adsorbed atoms. This proportionality is of importance also for
the thin layer cells. It may happen that during cyclic scans and cyclic voltammetry
(CV) experiments, particularly at high scan rates, the capacitive current may ex-
1
ceed the Faradaic response, being proportional to v =2 as we will see next.
The most important current response directly related to the redox processes on
WE electrode is the Faradaic response. In this case, in the equivalent circuit represent-
ing our experimental setup, we should introduce the charge-transfer resistance, Rct,
connected in parallel to the WE capacitance, allowing for the current flow across the
electrode/solution interface. Now, we are ready to arrange all equivalent elements of
our experimental setup in a simple scheme shown in Fig. 2.1.
 2.1 Transient methods 67

Counter
electrode

Rs
Reference
electrode

Cedl Rct Fig. 2.1: A simple electronic scheme (equivalent circuit) of the
electrochemical cell. Rs – electrolyte resistance, Rct – charge transfer
resistance related to the redox reactions (faradaic current) occurring at
Working the working electrode (WE), Cedl – differential capacitance of the electrical
electrode double layer at the WE electrode.

2.1.2 Faradaic current

Most of the voltammetric techniques used in applied electrochemistry are based on a

controlled change of the potential of the working electrode. Therefore, it is not possible
to attain steady-state conditions, even for a longer period of time. In the case of tran-
sient techniques, such as chronoamperometry, linear scan, or cyclic voltammetry (CV),
the steady-state conditions can never be reached. So, we cannot rely directly on the
Nernst–Planck equation, nor on the first Fick’s law, discussed in the previous part. How-
ever, we can use the second Fick’s law that considers the flux of matter in time domain,
t. For the case of a semi-infinite diffusion along the x-axis, normal to the electrode sur-
face (x is also the distance from the electrode surface), the second Fick’s law will read:

∂cOx ðx, tÞ ∂2 cOx ðx, tÞ

= DOx (2:6)
∂t ∂x2

This partial differential equation can be solved by setting boundary conditions that
are appropriate for a selected electrochemical technique.
As mentioned earlier, the most frequently used transient techniques are based on
controlled potential; therefore, we later focus on chronoamperometry and CV. How-
ever, the interested reader is encouraged to search in the suggested readings at the
end of this paragraph.

(a) Potential step (chronoamperometry)

This is illustrated in the case of a reduction reaction in Fig. 2.2. This figure contains
also the boundary conditions necessary to solve second Fick’s law in order to get the
current response over time.
68 2 Selected electrochemical methods

|E| cOx reacts on the electrode iF

Boundary conditions same for:
being converted totally chronoamperometry
to cRed chronopotentiometry
chronovoltametry
0 (no electrode
t = 0, x ≥ 0, cox = cox
reaction)
t ≥ 0, x ∞, cox cox 0 (bulk)
t=0
Chronoamperometry
t > 0, x = 0, cox = 0

t t
no reaction at this potential

Fig. 2.2: Chronoamperometry. Left: potential step of magnitude sufficient to reduce (or oxidize) all
substrate arriving to the electrode surface. Right: current response to this potential step described by the
Cottrell equation. Blue curve – the effect of semispherical diffusion, eq. (2.9). See text.

Solving eq. (2.6) with these conditions will yield the so-called Cottrell equation:
1
iF ðtÞ = nFADc0 (2:7)
ðπDtÞ1=2

The term found in the denominator of this equation,

δD = ðπDtÞ1=2 (2:8)

describes the time evolution of the diffusion layer into the bulk of the solution; it in-
creases with t1/2.
The above equations are valid only for planar electrodes of “infinite” size from the
point of view of the diffusion layer thickness, δD . For microelectrodes and spherical
electrodes, we should change to radial coordinates and the current response will read:
" #
1 1
iF ðtÞ = nFADc 0
1
+ (2:9)
ðπDtÞ2 r0

where r0 is the radius of the spherical electrode or microelectrode. This equation will
become the usual Cottrell equation for r0 → ∞. It is also important to note that for
short t, the spherical contribution is minimal, whereas for longer times the first term
in square brackets can be neglected and a steady-state current will flow.

(b) Cyclic voltammetry

CV is a type of potentiodynamic technique. In a CV experiment, the working elec-
trode potential is changed linearly versus time and after the set potential is reached,
the WE potential is reversed linearly to return to the initial potential. These cycles
of potential scans may be repeated as many times as needed. The current at the
working electrode is plotted versus the applied voltage (the WE potential) to give
 2.1 Transient methods 69

the cyclic voltammogram graph. CV can be used to study qualitative and quantita-
tive information about electrochemical processes under various conditions, such as
the presence of intermediates in oxidation–reduction reactions, or the reversibility
of a reaction. It can be very useful in the developmental studies of power sources,
for example, resolving the processes responsible for power storage and generation.
Therefore, in the next paragraph, we will consider in more detail this technique.
Assume that only Ox is present initially; therefore, we will consider the cathodic po-
larization only (convention). The transport of Ox obeys the Fick’s law. The reduction reac-
tion will produce Red species, and then, upon the reverse scan (anodic polarization), the
oxidation reaction will also obey the Fick’s law:

∂cOx ðxÞ ∂2 cOx ðxÞ ∂cRed ðxÞ ∂2 cRed ðxÞ

= DOx ; = DRed (2:10)
∂t ∂x2 ∂t ∂x2

Faradaic current, iF, is always “conjugated” with capacitive current, iC, as discussed
earlier, because of the charge accumulation at the electrode surface (e.d.l.):

i = iC + iF = Ce.d.l. × ðdE=dtÞ + iF = Ce.d.l. × v + iF

In order to solve the above partial differential equations, we have to set the following
boundary conditions:
1. At t = 0, at the electrode surface, x = 0, concentration of cOx is equal to the bulk
concentration, while cRed is equal to zero: cOx, x = 0 = c0Ox, cRed, x =0 = 0
2. At t > 0 and far away from the electrode, cOx approaches the bulk concentration,
whereas cRed approaches zero: x → ∞ cOx → c0Ox, cR → 0
3. At t > 0, at the electrode surface (x = 0) both fluxes of Ox and Red compensate
(what arrives, upon reaction must leave):



∂cOx ðxÞ ∂cRed ðxÞ
DOx + DRed =0
∂t x=0 ∂t x=0

4. At the time between the starting point (t = 0, Ei) and time of the reversal of the
scan, λ, the potential sweep is linear with rate υ: 0 < t ≤ λ, E = Ei – υ t
5. At t > λ, the linear scan proceeds in the opposite direction with the same rate:

E = Ei − υλ + υðt − λÞ

For reversible, diffusion-controlled processes, one more boundary condition should

be introduced, derived directly from the Nernst equation:
 
cOx, x = 0 nF  
= exp E − E0 (2:11)
cRed, x = 0 RT

These boundary conditions can be used to derive the theoretical expression for the
current–potential–time relationship, yielding the shape of a reversible voltammo-
70 2 Selected electrochemical methods

gram. It is beyond the scope of this book to present more theory; it can be found in
the suggested readings. What is important, however, is the resultant shape of a dif-
fusion-controlled cyclic voltammogram, presenting maximum and minimum peak
currents for the reduction and oxidation reaction of a given redox species (Fig. 2.3).

i (A) Ep,a

Faradaic
current
ip,a
Capacitive
current
Ep,c = E 0 – 0.0285/n
½
0
ip,c = – 2.69*105 *n3/2 *A*COx*DOx *󰜈½
|Ep,c – Ep,a| = 59/n [mV] at 298 K
ip,c
i~󰜈½
|ip,a/ip,c| = 1

Ep,c E (V)

Fig. 2.3: Cyclic voltammogram showing the summary of characteristics for reversible electrode processes.
The contributions of capacitive and Faradaic currents to the overall response are distinguished, as well as
the so-called background current in the absence of a redox couple (green curve), due to the capacitive
contribution of ideally polarizable WE. Reversibility criteria are also summarized.

Any change in CV shape, peak-to-peak separation, or peak currents ratio indicates

that the observed electrode process changes from the diffusion-controlled, reversible
regime into the irreversible, kinetically controlled one.
Below is the graph summarizing the above techniques and illustrating graphi-
cally the shape of the applied potential stimulus (first column), the measured cur-
rent response (second column), and the concentration profile during the time of
the applied potential stimulus. The first row shows the steady-state conditions
(first Fick’s law), independent of time, with constant diffusion layer thickness.

2.2 Electrochemical impedance spectroscopy

This section describes the basic principles of the electrochemical impedance spectros-
copy (EIS), a very powerful electrochemical technique that found its applications in
almost all areas of applied electrochemistry.
The classical electrochemical techniques measure charge, current, potential as a
function of time, the latter can be expressed in terms of applied potential in poten-
tiodynamic techniques, such as CV. Contrasting to these, EIS reports the measured
impedance as a function of frequency (frequency domain) at a constant, selected po-
 2.2 Electrochemical impedance spectroscopy 71

Potential profile Current response Concentration profile

E i c
Steady state

𝛿 = const

t E x
E i c
Potential step

𝛿 increases

t t x
i

E c
Cyclic coltammetry

𝛿 increases
t

i
t 𝛿 x

Fig. 2.4: Plots of potential profiles applied to the working electrode (left column), current–potential or/and
current–time responses to the applied potential (middle column), and concentration profiles as a function of
distance from the electrode surface (right column), for different types of electroanalytical techniques. Please
note that in the case of cyclic voltammetry, the X-axis can be either time (t) or potential (E), because both are
related by the potential scan rate.

tential. Understanding EIS may pose some difficulties, as it requires some knowledge of
Laplace and Fourier transforms and complex numbers; nevertheless, the advantages of
EIS are numerous. At each frequency and applied potential, EIS provides a large num-
ber of useful information that can be analyzed with the help of the electrical equivalent
circuits, representing different electrochemical processes at the electrode interface and
its vicinity. Impedance measurements allow us to determine the components of this cir-
cuit and their values. These electrical components are related to the physicochemical
characteristics of the system. This will be clearly seen from the text and examples that
follow.
72 2 Selected electrochemical methods

Let us recall the well-known Ohm’s law that relates the resistance of an electrical
circuit element known as a resistor, to the current flow I under the influence of ap-
plied voltage E:
E
R= (2:12)
I

This relation, however, is valid only for the ideal resistor R. The properties of an ideal
resistor are as follows:
– The Ohm’s law is fulfilled at all voltages E and currents I
– The value of R is independent of frequency of the applied voltage E
– The alternating current (AC) and voltage signals are in phase with each other (see
Fig. 2.5).
Magnitude

phase shift = 0

t Fig. 2.5: Current and voltage signal profiles in the

AC (alternating current) electrochemical
impedance spectroscopy (EIS) in the case of pure
electrical resistance, R, in the studied system.
No phase shift between the two is observed.

However, real electrochemical cells and redox processes cannot be represented by

the Ohm’s law, but require a more general parameter – impedance, Z:
EAC
Z= (2:13)
IAC

As in the case of Ohm’s law, the impedance Z is a measure of the relation between the
applied voltage, E, and current response, IAC,, but unlike the resistance, R, it is not
limited by the properties of an ideal resistor. And so, the electrochemical impedance
is usually measured by applying an AC sinusoidal voltage of small amplitude and fre-
quencies varying from 106 Hertz to 0.01 Hertz, to an electrochemical cell and measur-
ing the current response of the investigated system. This current response can then
be analyzed as a sum of sinusoidal functions (Fig. 2.6).
When an AC flows through a circuit, the relation between current and voltage
across a circuit element is characterized not only by the ratio of their magnitudes, but
also the difference in their phases. For example, in an ideal resistor, the moment
when the voltage reaches its maximum, the current also reaches its maximum, as
shown in the graph (Fig. 2.5) (current and voltage are oscillating in phase). But for a
 2.2 Electrochemical impedance spectroscopy 73

Magnitude

Voltage
Current

Fig. 2.6: Phase shift of current response for a

t
sinusoidal potential applied to the electrode in
the presence of a capacitive component to the
studied system. The amplitude change is
neglected.

capacitor, the maximum current flow occurs as the voltage passes through zero and
vice versa (current and voltage are oscillating 90° out of phase, see Fig. 2.6). Complex
numbers are used to keep track of both the phase and magnitude of current and volt-
age. The excitation signal, expressed as a function of time, can be written as follows:

Et = E0 sinðωtÞ (2:14)

Here, Et is the potential at time t, E0 is the amplitude of the potential (its value at a
maximum), and ω is the radial frequency (in radians), related to the signal frequency
f in Hertz, as follows:

ω = 2πf (2:15)

Assuming the linearity or pseudolinearity of current–voltage relation, valid only for a

small excitation signals, the current response for the above excitation will also take
sinusoidal form and should read:

It = I0 sinðωt + φÞ (2:16)

So, the current response It is shifted in phase by a phase angle φ and has a different
amplitude than I0. As mentioned earlier, the impedance Z is potential-to-current ratio
(as in Ohm’s law), so:

Et E0 sinðωtÞ sinðωtÞ
Z= = = Z0 (2:17)
It I0 ðsinωt + φÞ sinðωt + φÞ

Thus, as potential and current, the impedance is expressed in terms of magnitude Z0

and phase shift φ. With the help of Euler’s relationship:

expðjφÞ = cosφ + jsinφ (2:18)

we can replace the trigonometric forms of eqs. (2.14)–(2.17) in terms of imaginary

relations:

Et = E0 expðjωtÞ (2:19)
74 2 Selected electrochemical methods

and the current response:

It = I0 expðjωt − φÞ (2:20)

Finally, the impedance Z(ω) is a complex number:

Et
Z ðωÞ = = Z0 expðjφÞ = Z0 ðcosφ + jsinφÞ (2:21)
It

Equation (2.21) tells us that as a complex number, the total impedance of a system, Z
(ω), is a sum of real, Zre, and imaginary, Zim, part:

ZðωÞ = Zre + Zim (2:22)

or
pffiffiffiffiffiffiffi
ZðωÞ = Z′ + jZ′′, where j = −1 (2:23)

2.2.1 Data presentation

Typically, the experimental data obtained during EIS experiments can be displayed in
complex plane (imaginary part Z″ vs. real part Z′), called the Nyquist plot, but they can
also be displayed as a vector, defined by the impedance magnitude, |Z|, and phase
angle, φ (phase shift), described above. The vector and the complex quantity are differ-
ent representations of total impedance and are mathematically equivalent. The graph
below shows the representations of EIS data as vector and complex plane for single
frequency (Fig. 2.7).

Vector Complex plane

Imaginary Impedance Z’’

de
nitu
mag

phase angle
I

Real Impedance Z ’

Fig. 2.7: The representations of EIS data as a vector and a complex plane for a single frequency.

Now, let us consider an ideally reversible electrode that can be represented as a sim-
ple resistor, R. Then, since in the case of a resistor there is no phase shift, the total
impedance will read:
 2.2 Electrochemical impedance spectroscopy 75

Et E0 sinðωtÞ
Z= = = Z0 = R (2:24)
It I0 ðsinωtÞ

In complex plane Z′ versus Z′, it will be represented by a single point, as shown in the
simulation for R = 2,000 Ohm (Fig. 2.8). The same data can be also represented as the
so-called Bode plots to visualize the effect of the frequency of the applied sinusoidal
perturbation on both Z′ and φ (phase shift) (Fig. 2.8), respectively. All EIS graphs were
obtained with help of a freeware EIS Spectrum Analyser, by Genady Ragoisha, Re-
search Institute for Physical-Chemical Problems, Belarusian State University.
However, in the case of an ideally polarizable electrode, there will be only charge
accumulation at the interface. In this case, the circuit element should be a capacitor C.
Then, the expression of the total impedance should be derived as follows. First, we take
eq. (2.12) defining the potential polarization, then, we take the definitions of a capacitance
C, and resultant charging and discharging current, IC,, assuming the independence of C of
polarization:

dQ
Et = E0 sinðωtÞ, Q = C · E and IC = (2:25)
dt

This will give us finally:

 π
IC = ωCE0 cosðωtÞ = ωCE0 sin ωt + (2:26)
2

This equation shows that for an ideal capacitor the flow of current, IC, is shifted to 90°
with respect to the applied sinusoidal perturbation Et. Equation (2.26), per analogiam
to Ohm’s law, can be rewritten as follows:
E0  π
IC = sin ωt + (2:27)
XC 2

where XC = 1/ωC is called the capacitive reactance, which, unlike the resistance, R, de-
pends not only on the capacitance, C, but also on the angular frequency, ω, of the AC
current flow through this capacitor. Now, the total impedance contains only the imagi-
nary part:

Z = j · Z′′ = j=ωC (2:28)

It is obvious from the above equations that the larger the ω, the lower the impedance
of an ideal capacitor. Again, this can be visualized with the help of Nyquist and Bode
plots (Fig. 2.9) (C = 1 μF):
Let us now put together the above two circuit elements, R and C, in series. Then, the
total impedance Z will be equal:
76 2 Selected electrochemical methods

Fig. 2.8: The Nyquist and Bode graphs; simulations for the resistor (2 kOhm).
 2.2 Electrochemical impedance spectroscopy 77

Fig. 2.9: The Nyquist and Bode plots for a single capacitance. Black arrow indicates the direction of ω
increase. Please note that the Y-axis in the right-hand side panel is in radians, so the value of phase shift
φ = 1.570796 radians corresponds to π/2 = 90°.
78 2 Selected electrochemical methods

j
Z = R + jXC = R + (2:29)
ωC

The next step in our representation of an electronic equivalent circuit should be

the circuitry shown in Fig. 2.1, modeling the typical electrode/electrolyte interface.
Here, the interface is not only charged (ideally polarizable interface), but also some
part of this charge can participate in the redox reaction and can be transferred
across this interface. In other words, this reaction is partially short-circuiting the
charge accumulation with a charge transfer resistance, Rct. This is shown in Fig. 2.1
as a parallel mesh of a capacitor, Cdl, and a resistor, Rct. Moreover, we should in-
clude also the ionic resistance of the solution, Rs, that does not participate in the
redox reaction itself. Such a solution, usually of concentration at least 10 times
larger than that of the redox species, is usually called the supporting electrolyte
and its role is to maintain possibly high conductance of the solution and to keep
the ionic strength constant, so the structure of the electrochemical double layer
should remain unaffected by the electrode processes (Fig. 2.10).

Cdl

RS
Fig. 2.10: A simplified Randles’ electronic equivalent
Rct circuit.

Here, the total impedance Z of such circuit is a sum of RS in series with the imped-
ance, Zp, of parallel connection of Cdl and Rct. So
For a parallel mesh:
1 1 Rct
= + jω Cd l , Zp = (2:30)
Zp Rct 1 + jωRct Cdl

For total impedance:

Rct
Z = R S + Zp = R S + (2:31)
1 + jωRct Cdl

It is important to note at this point that in the case of real systems, Rct should be re-
placed by the Faradaic impedance, ZF, aiming to account for all possible electrode pro-
cesses, such as diffusion, adsorption, surface structuring/evolution, electrodeposition,
coating, and corrosion. In chapters that follow, we will exemplify some of the simplest
models of electrochemical processes; however, the interested reader is directed for
further readings listed at the end of this chapter. For such a circuit, called the simplified
Randles circuit, the Nyquist plot is shown in Fig. 2.11, together with the Bode plots.
 2.2 Electrochemical impedance spectroscopy 79

Fig. 2.11: Nyquist and Bode plots for RS (100 Ohm) in series with parallel connection of Cdl (2E-5 F) and Rct
(1,000 Ohm), as in Fig. 2.7. Black arrow indicates the value of total impedance for frequency of
approximately 2 Hz, whereas its length is the impedance magnitude, according to the Pythagorean law:
Z2 = (Z′2 + Z″2)0.5. This is the vector representation of the impedance jZ j. The angle between this vector and
80 2 Selected electrochemical methods

The Nyquist plot also shows three very important points, denoted a, b, and c. These are
the characteristic points allowing us to immediately retrieve the information on RS, Rct
and Cdl, because the impedance value at point a equals to RS, at point b it is equal to RS +
Rct (the imaginary part at both points equals zero), whereas at point c (the maximum of
the semicircle) ωCdlRct = 1. The values of RS (point a) and RS + Rct (point b) can also be
found on the Bode plot of Z′ versus log(frequency).
As stated earlier, in the case of real electrochemical systems, in order to account
for all possible electrode processes, the charge transfer resistance, Rct, should be re-
placed by the Faradaic impedance, ZF. The most common (but not limited to) elements
that constitute part of ZF are the Warburg impedance, W, and the constant phase ele-
ment (CPE). Warburg impedance represents a resistance to mass transfer, showing the
extent of diffusion control in the investigated system. Typically, it shows a 45° phase
angle (phase shift). The diffusion control has to be considered, for example, in studies
of corrosion processes, where there is limited access to oxygen (discussed later in chap-
ters that follow). The CPE approximates the “real” surface that is not an ideally flat, uni-
form surface with “smeared” charge uniformly distributed over the whole surface.
Normally CPE shows a 70°–90° phase shift. The latter phase angle is characteristic of a
perfect capacitor. Generally, the lower the phase angle, the more “imperfect” the inter-
face is with diffusive processes taking control over the whole electrochemical process.
This can happen for instance in porous electrodes, conductive polymer coatings, or lith-
ium batteries. Figure 2.12 shows the Nyquist and Bode plots for Warburg impedance, W,
in series with Rct, both in parallel with Cdl, and the whole mesh is in series with RS.
One may question why we presented the impedance data in both the complex plane
(Nyquist plot) and Bode (Z′, Z″ vs. logarithm of frequency). Therefore, let us compare the
information that can be immediately obtained from both types of plots.
Nyquist plot
– Individual charge transfer processes are resolvable.
– Frequency is not obvious, apart from its increase toward the lower frequencies.
– Small impedances can be hidden by large impedances.

The two last points are the disadvantage of a Nyquist plot. Therefore, to make the re-
sponse of the system more specific, the Bode plots are used.
Bode plot
– Individual charge transfer processes are resolvable as in Nyquist.
– Frequency is explicit.
– Small impedances can be identified easily even in the presence of large
impedances.

Fig. 2.11 (continued)

the X-axis, commonly called the “phase angle, ’” and tan’ = Z′′
Z′ . Black circular arrow shows the direction
of the increase of angular frequency, ω.
 2.2 Electrochemical impedance spectroscopy 81

Fig. 2.12: An example of EIS simulation for the “full” Randles’ equivalent circuit that includes Warburg
impedance W (left panel – Nyquist; middle and right-hand side – Bode) for RS = 100 Ohm, Rct = 1,000 Ohm,
Cdl = 2E-5 F, and W = 50 Ohm.
82 2 Selected electrochemical methods

Finally, a word of caution should be said with respect to EIS. Even though EIS can provide
very useful and precise information about the studied processes and their mechanism
and therefore become more and more popular among the electrochemists, several crite-
ria have to be fulfilled before the appropriate conclusions could be inferred from the re-
sults and model fitting:
A. The system under study
1. The investigated electrochemical system obeys Ohm’s law, that is, the value
of impedance is the proportionality coefficient of potential stimulus and cur-
rent response: E = Z.I. Moreover, the value of Z is independent of the magni-
tude of potential perturbation.
2. The system is stable during the EIS experiment and does not change in time
since some of the experiments can last quite long. For instance, if a wide fre-
quency range is used with several steady-state DC potentials chosen for each
frequency range, the experiment can take up to hours.
3. The observed response of the system is due only to the applied potentials
(both DC and AC).

B. Data presentation and modeling

1. It is a must that each element of the electrical equivalent circuit corresponds to
a specific mechanism or process in the studied system.
2. One should not simply multiply the equivalent circuit elements until a good fit
is obtained, but use the simplest model that fits the data (the Occham’s razor
principle). This model should comply with point B.1, of course.

At the end of this section, let us now summarize what was said above in the form of a
flowchart graph. We should state, however, that this flowchart, with some modifica-
tions, should be used in all applied electrochemistry experiments.

2.3 Electrochemical quartz crystal microbalance method

As we discussed in the previous section, numerous electrochemical methods are avail-

able for measurements of characteristic parameters (current, potential, and charge)
of electrochemical processes and systems.
In the last years, many hybrid methods were developed for the investigations of elec-
trochemical systems. They are “in situ” techniques that give us the possibility to examine
the state of electrode surface and reaction products in real time of electrochemical pro-
cesses. One such technique is electrochemical quartz crystal microbalance (EQCM), a very
powerful tool for the investigations of thin films and interfacial phenomena. EQCM
method gives us reach information about deposition, dissolution of metal films, redox
processes in polymer films, formation of self-assembled monolayers, and changes at the
electrode/solution interface resulting from adsorption or reaction. EQCM method is based
 2.3 Electrochemical quartz crystal microbalance method 83

Sample

Experiment

Physical model Electronic

of processes equivalent circuit

Calculations &
Simulations

Verification with
experiment

Characterisation
No of system and Yes!
processes done

Scheme 2: Flowchart graph summarizing a critical approach to the electrochemical results obtained with
different techniques.

on simultaneous detection of frequency changes and, in consequence, of mass changes

on the electrode during its electrochemical response.
Figure 2.13(a) shows a schematic diagram of an EQCM apparatus. It contains three
electrodes cell. Resonator (working electrode) consists of AT-cut quartz crystal (QC) in
a form of a thin disk of approximately 320 μm thickness, covered on both sides by a
thin layer of gold (200 nm thickness). Other coinage metals and carbon coatings are
also available commercially. Figure 2.13(b) shows the top view of QC with the depos-
ited Au layer. One side of the Au-covered crystal is in contact with the electrolyte
and works as a typical electrode. The Ref electrode and C counter electrode are as
usual. All equipment used, that is, potentiostat/galvanostat, oscillator, and frequency
counter, are controlled by a computer. The typical experiment would involve the
application of the potential waveform, or the potential step, to the working elec-
trode, and simultaneous measurements of current flowing through the cell and the
oscillation frequency of the crystal. In many cases, the oscillator and frequency counter
is replaced by the impedance analyzer.
84 2 Selected electrochemical methods

(a) CE (b)
Au
RE Potentiostat Computer

quartz

Frequency
electrotyle Oscillator
counter

Quartz crystal

WE

Fig. 2.13: (a) Schematic diagram of apparatus for electrochemical QCM measurements. (b) Schematic top
view of quartz crystal with deposited gold electrode.

QCM and in consequence EQCM methods are based on the reverse (converse) piezo-
electric effect. The primary piezoelectric effect involves the formation of the charge
on a crystal surface caused by mechanical strain. In consequence, one can observe a
potential drop across such a crystal that is proportional to the magnitude of strain.
Such an effect is characteristic for acentric crystals, such as quartz and other crystals
(topaz, sucrose). The converse effect means that the application of an external electric
field across the crystal will generate the strain and crystal deformation. The applica-
tion of the alternating potential across quartz crystal causes changes in the strain po-
larity, in consequence, the vibrational motions and further formation of acoustic
transverse wave. This transverse acoustic wave propagates across the crystal depth
between the two crystal surfaces (Fig. 2.14).
The standing wave conditions are attained when acoustic wavelength, λ, is equal
to 2dq ðλ = 2dq Þ, where dq is the thickness of the quartz crystal. In the resonant condi-
tion, the resonant frequency of acoustic wave f0 is given by the equation:
!
νtr μq
1=2
 
f0 = = 1=2
: 2dq (2:32)
2dq ρq

where νtr is the transverse velocity of acoustic wave in AT-cut quartz, ρq is the density
of quartz ( ρq = 2.648g=cm2 ), μq is the shear modulus of quartz ( μq = 2.947 × 1010 N=m).
Under the resonant conditions, such specially prepared quartz crystal works as a reso-
nator. Let us suppose that on this resonator we deposited compact, uniform layer of
some foreign (unknown) material with acoustic properties identical to those of quartz.
In this case, the change of the thickness resulting from the deposited layer can be
treated as the change in the thickness of the quartz crystal. The acoustic wave will now
propagate across quartz and deposited layer, but with different frequency (lower)
caused by an increase in thickness. The fractional changes in the frequency Δf caused
by fractional changes in thickness Δd are described by the equation:
 2.3 Electrochemical quartz crystal microbalance method 85

(a)

Q+
(b)

(+) Fig. 2.14: Schematic representation of piezoelectric

(–) effects: (a) quartz crystal (QC) is not affected by
external forces. The distribution of dipoles in QC is
Q– chaotic (no piezoelectric effect). (b) Quartz crystal
~
under the influence of stress. Stress causes the
E orientation of dipoles resulting in the formation of
(c) electric field (primary piezoelectric effect). (c) Quartz
crystal under the influence of the electric field. The
field causes reorientation of dipoles resulting in the
lattice strain and shear deformation of crystal
(converse piezoelectric effect).

Δf Δd Δd
= = − 2f0 (2:33)
f0 dq νtr

Δm
Using this equation, and taking into account that Δd = Aρ q
, where A is piezoelectrically
active area, we can obtain the well-known Sauerbrey’s equation:

2f 2 Δm
Δf = −  0 1=2
= − Cf Δm (2:34)
A μq ρq

Cf is called the integral mass sensitivity of a resonator.

Because the film is treated as an extension of quartz crystal, the Sauerbrey’s
equation applies only to systems in which the following conditions are met: the depos-
ited mass must be rigid and distributed evenly over the surface, so the deposited film
has a uniform thickness across the active region of resonator, and the mass of the
film does not exceed 2% of the mass of quartz crystal, that is, typically lower than 20
μg/cm2. The integral mass sensitivity has a value of 56.6 Hz.cm2/μg for 5 MHz crystal
in air, and in aqueous solution, it is lower (about 42 Hz.cm2/μg).
In practice, depending on the measured films, there are some effects that disturb the
Sauerbrey’s relation. We should mention here such effects, as the changes in viscoelastic-
ity of films due to swelling and ionic exchange. The other effects that can influence the
results are microsurface roughness of QC resulting in trapping of liquid in cavities of the
electrode surface, too high mass load on the electrode, temperature changes that affect
density, and viscosity of the solution in which the crystal is operating.
86 2 Selected electrochemical methods

As mentioned earlier, EQCM method is widely used in the investigations of differ-

ent types of thin films. Let us consider the simple deposition of thin film of metal on a
QC electrode. In this case, the frequency changes Δf result only from mass changes
(Δm) caused by the reduction of metal ions, and they are related proportionally to the
charge Q ðC=cm2 Þ,

106 MCf Q MQ
Δf = because Δm = (2:35)
nF nF
Here M is the apparent molar mass (g/mol), n is the number of electrons involved in
the unit process, and F is the Faraday constant. The factor 106 completes the conver-
 
sion from units of Cf Hz=μg cm2 to the units of M ðg=molÞ.
A plot of Δf versus Q will give the apparent molar mass of deposited metallic spe-
cies. Any deviation from linearity will point out that the process is more complicated
and can involve the exchange of solvent or ions or both between the film and solu-
tion, as it is observed for thin film of conducting polymers. Figure 2.15(a) presents the
CV voltammogram and frequency change of QCM during the deposition of Ag film on
Au-QC electrode.
CV scan shown in Fig. 2.15(a) begins at +0.6 V where only Ag+ ions exist, and the
frequency of resonator (electrode) is then set to zero (by the equipment circuitry).
Upon Ag+ reduction, the QCM frequency decreases as a result of Ag deposited mass
increase. Upon the oxidation of metal Ag during the reverse scan, the frequency of
QCM is going back to zero value. The relation between frequency changes and charge
flowing during the reduction and oxidation of Ag+, Ag species on the Au-QCM elec-
trode is shown in Fig. 2.15(b). Based on eq. (2.35), the relation is linear and points out
that there is no additional reaction at the resonator electrode. From the slope of Δf
versus Q, this relation allows us to determine the molar mass of Ag. As molar mass of

(a) (b)
0.6
0.0 0.0
0.4
1 –0.4
Δf (kHz)
Δf (kHz)

–0.5
i (mA)

0.2
–1.0 –0.8
0.0 1
–1.5 –1.2
–0.2
–0.2 0.0 0.2 0.4 0.6 –1.2 –0.8 –0.4 0.0
E (V) Q (mC)

Fig. 2.15: (a) Cyclic voltammogram (red curve) and frequency change (blue curve) recorded on QCM|Au
electrode in solution containing 0.002 M AgClO4 and 0.1 M HClO4; (b) Plot of frequency shift versus charge
passed during the measurements in the solution containing 0.002 M AgClO4 and 0.1 M HClO4.. Unpublished
data, courtesy of M. Skompska (Faculty of Chemistry, Warsaw University).
 Bibliography 87

Ag is well known, the deposition of Ag on Au-QCM electrode is often used for calibra-
tion and experimental determination of Cf .
To learn more about the theory and applications of the EQCM method, an inter-
ested reader will find in refs. [5–7].

Bibliography

[1] Bard AJ, Faulkner LR. Electrochemical Methods, Fundamentals and Applications, John Wiley & Sons,
Inc., 2001, 87–416.
[2] Marken F, Neudeck A, Bond AM. Cyclic Voltammetry. In: Scholtz F., ed., Electroanalytical Methods.
Guide to experiments and applications. Springer-Verlag, 2010.
[3] Retter U, Lohse H. Electrochemical impedance spectroscopy. In: Scholtz F., ed., Electroanalytical
Methods. Guide to experiments and applications. Springer-Verlag, 2010.
[4] Lasia A. Electrochemical impedance spectroscopy and its applications, Springer-Verlag, 2014.
[5] Gileadi E. Physical electrochemistry. Fundamentals, techniques, applications. Weinheim, Germany,
Wiley-VCH Verlag GmbH & Co. KGaA, 2011, 253–264.
[6] Buttry DA, Ward MD. Measurements of interfacial processes at electrode surface with the
electrochemical quartz crystal microbalance. Chem. Rev 1992, 92, 1355–1379.
[7] Buttry DA. Application of the quartz microbalance to electrochemistry. In Bard AJ, ed.
Electroanalytical Chemistry: A Series of Advances. New York, USA, Marcel Dekker, 1991, 17, 1–85.
Part II: Electrochemistry in material science –
selected topics
3 Corrosion

3.1 General remarks

Because corrosion causes spontaneous and uncontrolled degradation of materials in

their environment, it is one of the largest challenges in materials science and industry.
The economic costs of corrosion are extremely high and include the prevention of cor-
rosion and replacement of corroded materials. It is estimated that approximately 5–7%
of global income is spent on corrosion-related costs. Corrosion occurs naturally and
reflects the tendency of a material to achieve a more chemically stable form. Even
though this term can be applied to all materials, such as metals, alloys, semiconduc-
tors, conducting polymers, and even dielectrics, here we will focus on metals and their
electrochemical reactions with the environment, leading to the destruction of metal
construction to more environmentally stable products, such as oxides, hydroxides, or
sulfides.

3.2 Corrosion – what does it mean? Mechanism of corrosion

Corrosion is caused by oxidation of metals paralleled by the reduction of electroactive

species present in the environment. Both types of reactions require charge transfer;
therefore, electrochemistry is a technique of choice in the investigations of corrosion
processes. Corrosion reaction consists of oxidation and reduction reactions of the redox
couples being different from each other. Thus, it differs from the equilibrium electro-
chemistry, where the redox reaction at equilibrium is described by the Nernst equa-
tions. Under this condition, the reduction (cathodic) and oxidation (anodic) currents are
the same, though different in directions of electron flow, and the net current is zero. In
corrosion, however, usually, more than two different redox couples are involved in the
overall process. The open-circuit potential or the potential that is formed in the system
is determined by the potential at which the net current is zero, as in the equilibrium
electrochemistry, but the anodic current is related to the constant dissolution of metal,
while the cathodic current is related to the reduction of different species from the
environment.
We can classify the corrosion processes, for example, with respect to the
environment:
– Atmospheric corrosion: it is caused by air and air pollutants present in the air. It
depends on the presence of water (rain, dew, humidity, or melting snow) that dis-
solves various salts from the environment (e.g., salt used in icy roads). Atmo-
spheric corrosion is predominant of all the other forms of corrosion and the cost
of protection against atmospheric corrosion is approximately 50% of the total cost

https://doi.org/10.1515/9783111160986-003
92 3 Corrosion

of all other corrosion types. It affects all structures from small articles and utili-
ties to skyscrapers, industrial plants, and bridges.
– Soil corrosion: soil corrosion affects metallic constructions (such as oil pipelines) in
direct contact with soil or bedrock. It is not limited to metals, such as iron, steel, or
copper, but also constitutes a hazard to concrete structures. The rate of soil corrosion
depends greatly on the type of soil due to the fact that some soils are more corrosive
than the others because of their composition and water content.
– (Micro)biological corrosion: because living organisms can provide local corrosive en-
vironments (such as biofilm), for instance, under the aerobic conditions, some bac-
teria can oxidize sulfur, producing sulfuric acid: 2 S + 3 O2 + 2 H2O → 2 H2SO4,
which in turn can corrode ferrous structures, such as in-ground pipes or other un-
derground steel constructions. On the other hand, some anaerobic bacteria can re-
duce sulfates to sulfides obtaining hydrogen from organic wastes: SO42− + 4 H2 →
4 H2O + S2−. Both processes can be promoted in wet environments.

Or
– Wet corrosion (usually in an aqueous environment): it appears when two differ-
ent metals located in an electrolytic solution are in contact with each other. This
environment creates an electrochemical cell with potential between the two met-
als: the more reactive metal acting as an anode (oxidation), while the other one
acts as a cathode (reduction). When metals are exposed to air at temperatures
lower than 100 °C, water vapor can condense on their surface forming an aque-
ous layer, promoting the electrochemical corrosion reactions.

One can also classify corrosion from the point of view of its mechanism:
– Chemical – in dry gases, nonconducting liquids
– Electrochemical – in electrolyte solutions (aqueous, wet soil, humid atmosphere)

Although this chapter is devoted to the electrochemistry of corrosion, it is worth to

mention also that the mechanical action can facilitate the corrosion and resulting mate-
rial damage. The first type of mechanical wear (fatigue) results from mechanical stress –
the stress corrosion cracking (SCC). During SCC, the initiation of small defects – cracks,
pits, and their subsequent growth on metal or protective layer in an aqueous solution
can enhance the destructive effect of corrosion. It concentrates at the front of the de-
fects, stimulating the crack and pits growth and corrosion progress under the stress.
We will discuss this later in the text. Other mechanisms stimulating corrosion are due
to the flow of suspension of solid particles or even turbulent flow of gas bubbles crash-
ing and destroying the rust layer or passive film, thus exposing bare metal surface to
the fluid. These types of corrosion are called erosion corrosion and cavitation corrosion,
respectively.
As can be seen, the classifications criteria frequently interleave; therefore, for the
purpose of this handbook, we will focus on electrochemical models of corrosion.
 3.2 Corrosion – what does it mean? Mechanism of corrosion 93

Corrosion of metal starts when the oxidation of metal leading to the anodic disso-
lution and a formation of hydrated metal oxide begins. This can be described by the
following reactions:

Me ! Men + + ne −
Me + x · H2 O ! MeOy ðOHÞx − y + 2x · H + + 2x · e −

The dissolution rate is exponentially increased with an increase in the anodic poten-
tial of the metal following the Butler–Volmer formalism discussed in Chapter 1
(eq. (1.103)).
In order to maintain neutrality condition, a reduction reaction of a substance pres-
ent in the environment must occur simultaneously. In an aqueous solution, this is typi-
cally the hydrogen evolution due to the reduction of protons (acidic media) or water
reduction (neutral to alkaline media):

2 H + + 2 e − ! H2
2 H2 O + 2 e − ! H2 + 2 OH −

Other typical reduction is due to the presence of oxygen dissolved in the aqueous
solutions:

O 2 + 4 H + + 4 e − ! 2 H2 O
O2 + 2 H2 O + 4 e − ! 4 OH −

Obviously, the first reduction will take place preferentially in acidic media, whereas
the latter takes place in alkaline ones. Both reactions are controlled by the diffusion
of oxygen; therefore, their rates are similar. As already discussed at the beginning
of this chapter, corrosion consists of oxidation and reduction reactions proceeding
simultaneously of the redox couples being different from each other. The potential
that appears naturally in the system, the open-circuit potential, is determined by
several participating redox reactions whose net current is zero. This corrosion
potential is called the mixed potential and is not the equilibrium potential deter-
mined by the equilibrium conditions of each of the reactions, but results from their
kinetic relations. If only two redox couples are involved in corrosion, the open-
circuit potential will be between the two equilibrium potentials, even though it is
impossible to predict its precise value based on the thermodynamics.
Let us consider corrosion processes of Zn (valid also for other metals, such as
Fe, but at different potential regime) in acidic media, as shown in Fig. 3.1. At thermody-
namic equilibrium, the Nernst potential for zinc redox reaction (A – anodic) and hydro-
gen electrode reaction (C – cathodic), respectively, will be given as follows:
94 3 Corrosion

H2 – 2e H+
ia Zn – 2e Zn2+
ia,Me

ia,H

Ecorr
Eeq, H/H+ Eeq, Me –E

2H+ + 2e(Zn) H2
ic
Zn2+ + 2e Zn

Fig. 3.1: Schematic representation of the formation of corrosion potential, Ecorr, for two different redox
couples under wet, acidic conditions (solution, Butler–Volmer formalism only for clarity). Dissolution of Me
(Zn, Fe) in the oxidation (anodic) reaction coupled to the reduction of protons with hydrogen evolution in
the cathodic reaction. Arrows show the shifts of the i–V curves until the Ecorr is achieved with the net
current equal to zero. Please note the sign of E on x-axis.
No contribution of the reduction of O2 is considered.

RT
A : Eeq, Zn=Zn2 + = E0 + lnaZn2 + (3:1)
2F

′ RT pH 2
C: Eeq, H2 =2 H + = E0 − ln (3:2)
2F ðaH + Þ2

Under equilibrium, ia. Zn = ic, Zn = i0,Zn for metal and ia,H = c,H = i0,H.
It is immediately seen that in the corrosion (oxidation) process, the concentration
(activity) of Zn2+ ions in the solution will increase, shifting the potential of metallic Zn
toward more positive values (to the left in Fig. 3.1, shown by an arrow), whereas the
hydrogen evolution (reduction) coupled to this process will decrease the concentra-
tion of H+ ions, shifting the potential of H2/2 H+ redox couple (redox system 2) toward
more negative values (to the right, as marked by an arrow). Preservation of charge
and energy will finally equilibrate the rates of dissolution of metal and evolution of
hydrogen, so the oxidative and reductive currents will equilibrate at the potential
value Ecorr, being established between the two equilibrium potentials of both redox
couples:

ia, Me = − ic, 2 = icorr

where ia,Me and ic,2 are the anodic dissolution current of metal Me and cathodic cur-
rent of the redox couple 2, respectively, whereas icorr is the resultant corrosion cur-
rent, anodic positive, cathodic negative. At the potential Ecorr, metal will corrode at a
constant rate, providing the presence of a sufficient amount of coupled redox system
2 in the environment.
 3.2 Corrosion – what does it mean? Mechanism of corrosion 95

Other reactions that can participate in the overall corrosion process of Zn metal
are exemplified as follows:

O 2 + 4 H + + 4 e ! 2 H2 O
O2 + 2 H2 O + 4 e ! 4 OH −
Fe3 + + e ! Fe2 +
−
NO3 + 3 H + + 2 e ! HNO2 + H2 O

The corrosion example discussed above does not require any additional metal to be
present in the environment: the corrosion will proceed only due to the environment
surrounding the metal. In the above case of Zn (but also for Fe), it was an aqueous
solution of acidic pH.
Now we consider other possible cases of electrochemical corrosion, introducing
to the aqueous solution two metal slabs of different redox potentials such as Zn and
Fe Fig. 3.2. When these metals are separated from each other in the solution, upon the
open-circuit conditions, no corrosion will proceed, apart from the possible environ-
mental corrosion. If these two slabs are brought into contact, the electrochemical po-
tentials of electrons of one metal will equilibrate with the electrochemical potential of
the second one; hence, the metal of more negative Nernst potential will act as anode
and start to dissolve (Zn), whereas the metal of more positive Nernst potential will act
as cathode and will decrease greatly its corrosion (Fe). The electric contact between
the two metals induces a potential change in the more negative direction from the
equilibrium value at Fe and in the more positive direction from the equilibrium value at
Zn. The positive shift of the potential of Zn2+/Zn redox couple from Eeq, Zn=Zn2 + will result
in an increase of the anodic dissolution current of Zn, which can be described by the
electrochemical kinetic formalism of Butler–Volmer (eq. (1.103)). Contrary to this corro-
sion process, the negative shift from the equilibrium potential of Fe2+/Fe redox couple
from Eeq, Fe=Fe2 + will decrease the dissolution of Fe. The Zn/Fe system is therefore used
for the corrosion protection of steel structures. The zinc metal coupled with steel is
called the sacrificial anode. This is illustrated in Fig. 3.2, where initially both metals cor-
rode in acidic media with hydrogen evolution, and, after being brought into contact, the
potentials shift as described earlier to finally attain the corrosion potential of Zn/Fe pair,
Ecorr,Zn/Fe.
The above discussion applies also to the situation depicted in Fig. 3.3, where Fe and
more “noble” metal, such as Cu in contact, submerged, for example, in seawater at neu-
tral to alkaline pH. For this Fe/Cu pair, iron will act as anode, whereas Cu as cathode,
facilitating the corrosion of iron, particularly next to the Fe/Cu contact, etching Fe. This
graphics illustrates also the formation of the corrosion macrocell (contact) that forms
always when two metals of different redox potentials contact in a corrosion-promoting
environment. Metal of more negative redox potential will always undergo anodic disso-
lution, “protecting” its counterpart with more positive redox potential.
96 3 Corrosion

i Zn → Zn 2++ 2e Fe → Fe 2++ 2e

Ecorr(Zn) Ecorr(Fe)
Ecorr(Fe-Zn)
2H++ 2e(Fe) → H2
2H++ 2e(Zn) → H2

Fig. 3.2: Example of electrochemical corrosion in the case of two metal slabs of different redox
potentials such as Zn and Fe immersed in an aqueous solution of slightly acidic pH. Upon electrical
connection of these metals, they will tend to equilibrate their electrochemical potentials (compare
Fig. 1.10) and (Zn) starts to dissolve, whereas Fe decreases its corrosion.
This shifts the potentials of both metals along the potential axis toward more positive values (Zn) and
negative (Fe) toward Ecorr(Fe–Zn).

O2

NaCL solution

Fe(OH)2 + rust
OH–
Fe2+ Fe2+ OH–
Fe2+ Fig. 3.3: Formation of the corrosion macrocell. Here, Fe/Cu contact
e pair is submerged in, for example, seawater. Iron acts as an anode,
Fe Cu whereas Cu as a cathode, facilitating the corrosion of iron,
particularly next to the Fe/Cu contact, etching Fe.

Electrochemical corrosion can also appear within the same metal or alloy having in-
homogeneous grain structure, such as steel. This type of corrosion, called the intergran-
ular corrosion, appears because the energy of the interior of grains is smaller than that
of the grain boundaries (Fig. 3.4. Here Zn granular structure is exemplified). Then, the
grain interior will act as multiple cathodes and boundaries as multiple anodes, forming
a series of microcells. Corrosion due to the formation of microcells is generally hard to
control and therefore dangerous, mainly because it is within the metallic structure and
no rust is visible on the surface until the structure failure and breakdown. The above
two examples of electrochemical corrosion are classified based on the spatial locations
of anodic and cathodic sites (macro- and microcells) but they differ also with respect to
participants: a macrocell is formed when anode and cathode are formed from different
metals, while a corrosion microcell is formed when multiple cathodic and anodic sites
are within the same metal or alloy due to their granular structure.
 3.3 Characterization of corrosion: corrosion potential, corrosion current 97

H2SO4
H2↑

Zn2+ Zn2+

Zn
Fig. 3.4: Corrosion microcell formed with multiple cathodic and
Inclusions
anodic sites existing within the same metal or alloy due to their
Zn Zn granular structure.

3.3 Characterization of corrosion: corrosion potential,

corrosion current
Corrosion potential (open-circuit potential) was discussed earlier. The second impor-
tant parameter describing the corrosion process is the corrosion current, icorr. The
higher the corrosion current, the greater the corrosion rate. Its value depends on all
participating species: their exchange currents, equilibrium potential differences, pH,
and so on. Let us consider the simplest possible system with only two participants, such
as corroding metal Me and the reaction counterpart – redox system 2. Then, according
to the Butler–Volmer equation (eq. (1.103)) we can write the following set of equations:
For corrosion of Me:
 
ð1 − αÞzFηMe − αzFηMe
ia, Me = i0, Me exp − exp (3:3)
RT RT

For the developing overpotential of Me:

ηMe = Ecorr − Eeq, Me (3:4)

For the redox system 2:

 
ð1 − αÞ zFη2 − αzFη2
ic, 2 = i0, 2 exp − exp (3:5)
RT RT

For the overpotential of 2:

η2 = Ecorr − Eeq, 2 (3:6)

For the case of charge transfer coefficient α = 0.5, the above set of equations will yield:
  
1 zF Eeq, 2 − Eeq, Me
icorr = ði0, Me × i0, 2 Þ × exp
2 (3:7)
RT
98 3 Corrosion

This equation shows that the measured corrosion current is a function of

i0, Me, i0, 2, ΔEeq:

icorr = f ði0, Me , i0, 2 , ΔEeq Þ (3:8)

3.3.1 Stability of materials: potential/pH (Pourbaix) diagrams – the thermodynamic

aspect

In 1966, M. Pourbaix proposed a system allowing classifications of various regions for

different metals in aqueous solution, depending on pH and potential versus standard
hydrogen electrode, assuming a concentration of Me ions of 10−6 M.

aÞ O2 + 4 H + + 4 e $ 2 H2 O
bÞ 2 H + + 2 e $ H2
cÞ Men + + ne $ Me

E (v) 1.8
1.6 H2O2
1.4 O3 O2 thermodynamically stable
with respect to water
1.2 O2 (in case of corrosion oxygen
1.0 evolution and acidifying)
0.8
0.6
0.4
0.2
0.0 H3O+/H2 H2O thermodynamicaly
–0.2 stable with respect
–0.4 to O2 and O2
–0.6
–0.8
H2 thermodynamicaly
–1.0 stable with respect
–1.2 to H2O and O2
–1.4 (pH increase)
–1.6
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH

Fig. 3.5: The Pourbaix graph identifying several regions and boundaries on the potential/pH plane for the
case of aqueous solution and a simple redox couple Men+/Me.

The Pourbaix graph (Fig. 3.5) distinguishes several regions on the potential/pH plane
for the case of aqueous solution and a simple redox couple Men+/Me, characterized by
its Nernst potential. The upper dashed line corresponds to the oxygen four-electron
 3.3 Characterization of corrosion: corrosion potential, corrosion current 99

reduction at equilibrium (reaction a) and its slope reflects its pH dependence, accord-
ing to the equation:

RT a4H + · pO2
Eeq = EO0 2 =H2 O + ln (3:9)
4F aH2 O 2

At a pressure of 1,013 hPa and temperature 25 °C, it will read:

Eeq = 1.23 − 0.059pH + 0.012logpO2 (3:10)

Therefore, the resultant slope of line a is −59 mV per unit pH.

Above this equilibrium line, O2 is thermodynamically stable with respect to
water. If corrosion appears, it will be accompanied by oxygen evolution and acidifica-
tion of the solution. Below this equilibrium line, water is thermodynamically stable
with respect to H2 and O2.
The lower dashed line (line b) corresponds to the reduction of protons at
equilibrium:

RT a2H +
Eeq = E20 H + =H + ln (3:11)
2 2 pH 2

At constant H2 pressure and at 25 °C, this equation will give

Eeq = − 0.059pH − 0.029logpH2 (3:12)

Thus, this line will also have a slope of –59 mV per unit pH. Below this line H2 is thermo-
dynamically stable with respect to water and oxygen. If corrosion appears, it will be
accompanied by hydrogen evolution and alkalization of the solution.
The last line, line c, is parallel to abscissa because the Nernst equation for Me re-
duction/oxidation is independent of pH:
RT
E = EMez + =Me +
0
lnaMez + (3:13)
zF

We are now ready to consider two examples of Pourbaix diagrams for the case of Zn and
Fe. In the potential–pH diagrams, Pourbaix arbitrary chose 10−6 M for the concentration
of metal ions and in the next examples, we follow this choice and T = 298 K.
In the case of zinc, keep in mind the possible amphoteric equilibria.
The four equilibrium lines 1–4 describe the following Nernstian equilibria of zinc
with respect to pH:
1. Zn2+ + 2 e ↔ Zn
RT
Eeq, 1 = EZn
0
2 + =Zn + lnaZn2 + (3:14)
2F
100 3 Corrosion

2. Zn(OH)2 +2 H+ + 2 e ↔ Zn + 2 H2O

RT
Eeq, 2 = E20 + lnaH +
F

since activities of solid Zn(OH)2, metal Zn, and H2O are equal to unity. Thus, this equa-
tion provides the equilibrium line 2 with slope −59 mV per unit pH:

E2 = − 0.439 − 0.0591pH (3:15)

3. HZnO2− + 3 H + 2 e ↔ Zn + 2H2O
+

3RT
Eeq, 3 = E30 + ln aHZnO − · aH + (3:16)
2F 2

E3 = 0.054 − 0.0886pH + 0.0295logaHZnO − (3:17)

2

4. ZnO22− + 4 H+ + 2 e ↔ Zn +2 H2O

RT
Eeq, 4 = E40 + 2 lnaZnO2 − · aH + (3:18)
F 2

E4 = 0.441 − 0.1182pH + 0.0295logaZnO2 − (3:19)

2

Below all these four lines (denoted as solid, thick line), metallic zinc will be immune
against corrosion. The line separating zinc stability against corrosion from the rest
of this Pourbaix diagram is shown as a thick solid line in Fig. 3.6.
As one probably notice at this point that there are several pH-dependent forms of
oxidized zinc such as Zn(OH)2, HZnO2−, and ZnO22−, the equilibrium lines separating
the stability area of these forms are independent of potential:

ZnðOHÞ2 + 2 H + $ Zn2 + + 2 H2 O; pKa = 8.5

−
ZnðOHÞ2 $ HZnO2 ; pKa = 10.7
− 2−
HZnO2 $ ZnO2 ; pKa = 13.1

Therefore, they are marked as lines parallel to the ordinate (thin, solid lines). Thus,
this Pourbaix diagram distinguishes, apart from the stability region of Zn metal, addi-
tional four different stable Zn oxidation products, depending on pH and Eeq.
As more important, but also a more complicated diagram, taking into account vari-
ous possible equilibria, we introduce now the diagram for iron (Fig. 3.7). For soluble ionic
species, there are Fe2+, Fe3+, and also various oxoacid ion species in high pH, alkaline sol-
utions. The two iron oxides, Fe2O3 and Fe3O4, are also considered. Nevertheless, having
the appropriate data, it is not difficult to draw the equilibrium lines corresponding to the
reactions taking place in the Fe/H2O system.
 3.3 Characterization of corrosion: corrosion potential, corrosion current 101

E (v) 1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
4
0.2
0.0
–0.2 3
–0.4 2
–0.6
–0.8 1
–1.0
–1.2
–1.4
–1.6
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
pH

Fig. 3.6: Construction of the Pourbaix diagram in the case of redox reactions of Zn in aqueous media at
different pH.

E (v) 1.8
1.6
1.4 Fe3+
1.2 Electrode potential
1.0 D O2/OH–
0.8
Corrosion
0.6 Formation
Fe2+
0.4 of Fe2O3
0.2 B on Fe
A
0.0 Electrode potential
–0.2 H2/H–
[Fe2+] increases
–0.4 above 10–6
–0.6 [Fe2+] = 10–6
–0.8 C
[Fe2+] decreases
–1.0 –6
below 10
Formation
–1.2 of Fe3O4
–1.4 Fe resistant Corrosion
on Fe
against corrosion HFeO2
–1.6
–2 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 16
pH

Fig. 3.7: The Pourbaix diagram of iron in aqueous media at different pH.
102 3 Corrosion

In this diagram there are three distinct regions:

1. Passivation, immunity region of Fe against corrosion: the potential is sufficiently
low, and bare iron is stable.
2. Corrosion region: Fe2+ and Fe3+ are stable, continuous corrosion of metallic iron.
3. Passivity region: stable iron oxide films are formed, inhibiting corrosion by sepa-
rating bare iron from the environment.

As mentioned earlier, Pourbaix assumed the concentration of iron ions equal to 10−6 M.
If it is higher, iron undergoes corrosion; if lower, iron is in the anticorrosive state. This
is shown in Fig. 3.7.
Pourbaix diagrams can be used to predict corrosiveness of the metal. Let us use
the above diagram to predict qualitatively the corrosion of Fe in aqueous solution. To
do this, the following data will be required:
1. pH of the solution
2. Corrosion potential, Ecorr
3. Redox potential ERedox of the solution, measured with the help of the so-called
redox electrode (Pt, glassy carbon electrode) with respect to the saturated hydro-
gen electrode as a reference electrode.

From the value of Ecorr and solution pH, we can judge in which of the regions iron metal
(as electrode) is localized. If Ecorr is more negative than that of Eredox of the solution with
environmental redox species, then the corrosion of Fe (anodic dissolution) can be pre-
dicted to proceed to a great degree, and vice versa. In other words, when the Ecorr of
iron sample is within the corrosion region of the Pourbaix diagram, a high corrosion
rate of such a sample should be expected. Then, one should consider the protection
of this sample. Consider, for example, point A on the diagram. If pH of solution is
weakly acidic, then iron is in the corrosion region. There are several ways to protect
this sample. One is obvious but hardly practical, namely, we should change the solu-
tion pH to a more alkaline, resulting in shifting our sample into the passivity region
(point B). But as we said, this method is unsuitable in reality. Another possibility is to
shift the potential of iron in a negative direction to place it inside the immunity region.
This method is called cathodic protection and is effected by contacting directly our Fe
sample with other metals that can act as sacrificial anode, such as Zn, Al, and Mg. This
will shift the potential of Fe to a more negative value (point C). The mechanism of such
protection was described earlier in the text. Another possible way to shift the potential
of iron toward the immunity region is to polarize it negatively from an external source
(point C). The third mechanism is called the anodic protection and this protection is
achieved by polarizing the iron sample in a positive direction in order to place it in the
passivation region (point D). In the case of anodic protection by a passivating layer, at-
tention should be paid to the existence of defects, even pinhole type in the continuous
layer. Their existence in the presence of aggressive ions in the solution, for example,
 3.3 Characterization of corrosion: corrosion potential, corrosion current 103

Cl−, will lead to localized corrosion. The passive layer will be etched by adsorption of
these ions forming small pits of large (few millimeters) penetrating depths.

3.3.2 Stability of materials: current–potential (Evans) diagrams – the kinetic

aspect

The potential–pH (Pourbaix) diagrams are based on the thermodynamic equilibria or

redox reactions of metals (oxidation) and environmental counterparts (reduction) partici-
pating in corrosion reactions. As we said earlier, the corrosion potential does not corre-
spond directly to the equilibrium redox potentials of participants. The actual corrosion
potential is a result of a mixed potential system and its value is between the equilibrium
potentials of all redox systems present. At this potential, the anodic positive current of
metal dissolution is coupled to the cathodic, negative current of environmental counter-
parts. From the electrochemical measurements of the current–potential relationship, one
can estimate the corrosion current of metal under study. Keeping in mind that the cur-
rent values are directly related to the corrosion rate, let us plot the potential–current dia-
grams showing the effects of the exchange current, i0Me, and the equilibrium potential of
corroding metal, Eeq,Me, on the resultant corrosion current icorr and corrosion potential,
Ecorr. The initial system is presented in Fig. 3.8(a), in which ordinate shows the potential
values: equilibrium for Me and environmental redox system 2, with resultant corrosion
potential in between the two. Abscissa presents the corresponding current values in the
logarithmic scale, resembling the Tafel plots (Chapter 1, Fig. 1.16) for the evaluations of
the exchange current and transfer coefficient characteristics of single redox couple in the
solution. We will leave unchanged the electrochemical properties of the environmental
couple 2, for clarity.
Figure 3.8(a) shows the values of Nernst potential of Me, Eeq,Me, its exchange current,
0
i Me, as well as similar values for the environmental redox couple 2. The point where the
anodic dissolution current of Me is equal to the cathodic current of 2 marks the corrosion
potential, Ecorr, and corrosion current, icorr. Now, consider the effect of replacing metal
Me, with more noble metal Me′, with its equilibrium potential E’eq,Me’, more positive than
that of Me, that is, being closer to Eeq,2 (Fig. 3.8(b)). This will result in a drastic (logarithmic
scale) decrease of corrosion current, i’corr. Figure 3.8(c) shows the effect of lowering the
exchange current of corroding metal by, for example, modification of its surface proper-
ties, such as by introduction of a passivating layer. In this case, without changing the
Nernst equilibrium potential, we introduce a kinetic barrier against the electron transfer,
decreasing the resultant icorr and shifting the Ecorr toward the equilibrium potential of the
environmental redox counterpart 2.
In the previous cases, we kept unchanged the electrochemical behavior of the en-
vironmental couple 2, assuming its kinetic-type redox properties. However, in neutral
pH, the reduction of dissolved oxygen gas can be considered as the preferential ca-
thodic reaction coupled to the anodic corrosion reaction of iron. Since the oxygen re-
104 3 Corrosion

(a) (b)
E E

Eeq,2 Eeq,2
'
Ecorr
Ecorr E'eq,Me'
Ecorr

Eeq,Me E'eq,Me'

log |i20| log |i| ' |

log |icorr log |i|
log |i0Me| log |icorr|
log |icorr|
(c)
E
Eeq,2
'
Ecorr
Ecorr

Eeq,Me 0' |
log |iMe
log |i0Me| log |i|
'|
log |icorr
log |icorr|

Fig. 3.8: The Evans kinetic diagram for metal: (a) Me and environmental redox system 2; (b) more noble
Me′ replaces Me (e.g., being electroplated on Me), being now in contact with 2. This results in a decrease of
corrosion current. (c) A decrease of the exchange current of Me also decreases the corrosion rate. See text.

duction current is usually diffusion-controlled one in the stagnant solution in equilib-

rium with oxygen, it reaches the limiting current value. Figure 3.9 shows an example
of an oxygen-rich solution after its concentration was limited (e.g., by deaeration or
changing the environmental conditions).

E
Er,2
Me Mez++ ze
Ecorr limiting the oxygen level
O2 + 4 H++ 4e 2 H2O
'
Ecorr
Er,Me
' |
log |icorr log |icorr| log |i|

Fig. 3.9: The effect of decreasing the activity of environmental redox couple 2, for example, depletion of
oxygen in the previously oxygen-rich environment. Such situation can occur near the bottom of a
stagnant water.
 3.4 Evaluation of corrosion rate from electrochemical measurements 105

Let us now consider the Pourbaix diagram (potential–pH) of iron in light of the Evans ki-
netic diagram (current–potential). To do this we will plot current measured in the external
circuit with respect to the polarization potential of an iron electrode in a deaerated solu-
tion. We will go along the constant pH line at pH = 5 and pH = 9 (Fig. 3.7, dashed lines pH =
5 and 9). Here, we will explain in more detail potential–current behavior for solution pH =
5. Similar considerations can be carried out at pH = 9.
Initially, at sufficiently low potential Eeq,Fe = −0.7 V (for concentration of Fe2+
lower than 10−6 M), iron will be in its immunity state, resistant to corrosion at pH 5.
Corrosion begins at a potential more positive (anodic) with rapid exponential rise of
corrosion current until it is equalized by the reduction of environmental couple 2, to
achieve Ecorr and icorr. Further positive polarization of Fe up to approximately 0.2 V
results in moving to the passivation region, when the surface is covered with iron
oxide film. At this potential, called the passivation potential, roughly corresponding to
the formation potential of Fe2O3, the corrosion current drops to very small values,
constant within the passivity region of Fe. Subsequent positive polarization of iron
causes the breakdown of passivity and steep rise of current. In general, three regions
can be distinguished: active (corrosion), transition, and passive regions, as shown in
Fig. 3.10.

Breakdown of passive layer

E and/or dissolution region
Passive region

Transition region
Epassivation

Eeq,Me Active region

log |ipas| log |i|

Fig. 3.10: Potential–current diagram of Fe identifying different corrosion processes.

3.4 Evaluation of corrosion rate from electrochemical

measurements
By combining the electrochemical kinetics with the corrosion process, we have now
tools to estimate the corrosion current and the corrosion mechanism.
106 3 Corrosion

3.4.1 Linear scan voltammetry

The simplest way to evaluate the corrosion process is to measure the value of the open-
circuit voltage EOCP that can be approximated as the corrosion potential, Ecorr, and then,
to measure the so-called polarization resistance, Rp, defined as the resistance of the metal
to oxidation during the application of an external potential in a given corrosive condition.
The Rp value is typically measured by polarizing linearly the sample in a very narrow
potential window around the Ecorr, typically within +/−5 mV to +/−10 mV range. Data anal-
ysis should provide a line in the imeas versus Epol coordinates, as shown in Fig. 3.11.

imeas

η
I Rp =
i
η

−10 mV Ecorr Epol

10 mV Fig. 3.11: Evaluation of corrosion parameters from the

linear scan voltammetry at narrow potential range
around Ecorr.

You may want to recall now the similarity of this approach to the evaluation of the
electrode polarization discussed in the case of the Butler–Volmer theory of electrode
kinetics for low overpotentials η values.
Next, the corrosion current can be used to calculate the corrosion rate with the
help of Faraday’s law. At the corrosion potential, Ecorr, both the cathodic reaction in
the environment and the anodic dissolution of metal will have the same corrosion
current, icorr. At the corrosion potential, the metal is “self”-polarized to Ecorr, and the
overpotential of the metal for this situation is

η = E − Eeq, Me = Ecorr − Eeq, Me

The dissolution current will follow the Butler–Volmer kinetic equations (see Figs. 1.16
and 1.17, and eq. (1.108)). Thus, the anodic dissolution of a metal electrode can be writ-
ten as follows:
 
ð1 − αÞzFη 0 ð1 − αÞzF Ecorr − Eeq, Me
iA, Me = i0Me × exp = iMe × exp (3:20)
RT RT
We now introduce a small perturbation ΔE to this overpotential in the range of
+/−5 mV to +/−10 mV, as discussed earlier, for the evaluation of polarization resistance
Rp. Thus, due to this perturbation, a new anodic dissolution current will flow:
 3.4 Evaluation of corrosion rate from electrochemical measurements 107

Ecorr − Eeq, Me + ΔE
iA, Me = i0Me × exp (3:21)
bA

and, after some rearrangement:

Ecorr − Eeq, Me ΔE
iA, Me = i0Me × exp × exp (3:22)
bA bA
RT
where bA = . Finally, we will have
ð1 − αÞzF


ΔE
iA, Me = icorr × exp (3:23)
bA

the same time, on the cathode:


ΔE RT
iC, 2 = icorr × exp ; bC = (3:24)
bC αzF

As already discussed, the current measured in such an experiment will be the

difference between the anodic current and the cathodic one; therefore:


ΔE ΔE
i = iA, Me − iC, 2 = icorr exp − exp − (3:25)
bA bC

In the case of small ΔE, we can develop the exponential function into a series and
taking only the first term, we will get


bC + bA ΔE
i = icorr × ΔE × , where = Rpol (3:26)
bC × bA i

and, finally, we will have a simple relation between the polarization resistance Rpol
and the corrosion current icorr:


1 bC + bA
icorr = × (3:27)
Rpol bC × bA

where bA and bC are slopes of plot log|i| versus applied potential E, for E larger than
approximately 0.1 V and −0.1 than Ecorr for anodic and cathodic polarization, respec-
tively. The value of current i can be measured by electrochemistry in a classical three-
electrode cell, with metal of interest as an anode. Even though the icorr (Fig. 3.12) can-
not be directly measured, it can be indirectly estimated with the help of the Tafel plot
(see Chapter 1, eq. (1.108)).

3.4.2 Electrochemical impedance spectroscopy in corrosion

Electrochemical impedance spectroscopy is very well suited to study corrosion. In

particular, it is a very powerful tool to study various coatings used to prevent corro-
108 3 Corrosion

In |i|
In |icorr|

Fig. 3.12: Tafel plot allowing for the evaluation of

Ecorr E
icorr.

sion. An ideal metal surface resistant to corrosion (ideal polarizable electrode), as

judged for instance from the Pourbaix diagram, is a pure capacitor, C, in series with
the solution resistance, Rs. Capacitance of such capacitor is related to the electrical
double layer capacitance, discussed in Section 2.2. The equivalent electrical circuit,
corresponding to this approach, was also shown there. However, the real metal/solu-
tion interfaces rarely behave in an ideal manner; they undergo corrosion – oxidation
with charge transfer across the interface. Then, the equivalent circuit has to be modi-
fied to include the charge transfer resistance, Rct, or polarization resistance, RP, short-
circuiting the interfacial capacitance, C. This is shown in Fig. 2.10 as simplified Ran-
dles model.
As in the case of plots in Section 2.2, the EIS Nyquist and Bode plots presented in
this chapter were simulated using freeware EIS Spectrum Analyser, by Genady Ra-
goisha, Research Institute for Physical–Chemical Problems, Belarusian State Univer-
sity. If we now include the diffusion contribution to the corrosion process, for
example, oxygen diffusion, then we will have to add the Warburg coefficient to the
circuit. Thus, we obtained the Randles full equivalent circuit of mixed kinetic and dif-
fusion process (Fig. 2.12).
In the real corrosion system, the metal is often already covered with corrosion
products, such as an oxide layer (may be passive), the adsorption layer of inhibitors,
protective passivating layer, or paint. In time, the protective layers may deteriorate and
cracks or other types of defects may be formed. Therefore, we now consider an even
more complex system, which involves the passivation or protection layer, for example,
iron oxide layer, again in reference to the Pourbaix diagram (Fig. 3.7) at various stages
of integrity: from the totally passivating one to that containing defects. First, let us con-
sider ideally passivating/protective coating thin layer on a metal surface. Intuitively, the
equivalent electric circuit for such a system should be represented by a solution resis-
tance, RS, and the capacitance of metal/coating/solution interface, CC, in series with RS.
The value of dielectric capacitance coating, well adhered to the metallic surface is usu-
ally one to two orders of magnitude lower than that of the e.d.l. capacitance, C, so now
even though the circuit looks identical to that describing the ideally polarizable elec-
trode, it will differ in the capacitance value, for the same solution resistance. You may
not see this difference in the Nyquist plot, but it will become evident if the data are
plotted in the form of Bode graphs that separate the effects of imaginary and real parts
 3.5 Localized corrosion 109

of the data (capacitance and resistance, for short). During the long-term exposure of the
coating to solution, corrosion is triggered in the cracks and pinholes. In the beginning,
it is localized but can spread around in time. Thus, the circuit parameters will change
in time, too, providing means of controlling corrosion progress. A pinhole or crack pen-
etrating through the coating produces a double-layer capacitance of the clean metal in-
terface, Cedl, greater than that of the coating: Cedl > CC. The Faradaic reaction that
proceeds is represented by the resistance, RPC. And so, now we have the following equiv-
alent circuit: RS in series with a mesh containing CC, parallel to RPC that is in series with
Cedl short-circuited with the polarization resistance, RP. This circuit is shown in Fig. 3.13,
with its Nyquist and Bode representation for the following data: CC = 1 x 10–8 F, Cedl = 1 x
10–6 F, RS = 100 Ohm, Rp = 10 kOhm, RP 5 kOhm.
The complex plane (Nyquist) plot of this circuit displays two semicircles with low-
frequency impedance decreasing with further deterioration of the protective coating.
Electrochemical impedance spectroscopy has proven to be ideally suited for studies of
the quality of protective coatings (passivating or organic layers), resulting in the pub-
lication of ISO norms. Other electrochemical studies discussed above are not sensitive
enough, although much simpler and easy to use. For the case of EIS, however, one has
to assure that the experiments and modeling are carried out correctly. Also, the equip-
ment has to be capable of measuring at high frequencies and resistances that are re-
quired for thin dielectric films [10].

3.5 Localized corrosion: pits, crevices, intergranular corrosion –

oxygen reduction as accompanying cathodic reaction
Localized corrosion is observed in small, restricted, local areas on metallic surfaces. It
is dangerous in transferring pipelines and storage tanks, particularly in chemical
processing plants. Contrasting to general corrosion affecting whole areas of metallic
constructions and equipment, localized corrosion penetrates the metal very rapidly,
with rates often several orders of magnitude higher than the corrosion rates for general
corrosion of the very same metal or alloy. It is also very difficult to localize in advance,
causing a substantial threat to life and damage to the installations and engineering sys-
tems. Localized corrosion is mainly due to heterogeneities and defects in metal, metal
coatings, or environment and takes place wherever the anodic and cathodic sites on a
metal surface can form corrosion microcells. Three main types of localized corrosion
can be distinguished, such as pitting corrosion, crevice corrosion, and intergranular cor-
rosion; the latter was described earlier, in the microcell corrosion section.
Pitting corrosion is caused by local dissolution or defects of passive film on the
metal surface. Cavities are formed in this process, surrounded by an intact passivated
surface [5, 6]. Pitting can be initiated by various factors and the most common are:
– Localized chemical or mechanical damage to the protective oxide film. Acidity of
aqueous solution, low dissolved oxygen concentration or high concentrations of
110 3 Corrosion

Fig. 3.13: Electrochemical impedance spectroscopy with simulated results for a metallic surface covered
with protective (passivating) coating with a small pinhole/crack penetrating through the passivating layer.
The electric equivalent circuit and its simulated response are shown in the form of Nyquist and Bode
plots. Circuit parameters are described in the text.
 3.6 Hydrogen evolution as accompanying reaction – role in corrosion 111

chloride (such as in seawater) or other aggressive compounds found in the chemi-

cal processing plants can account for majority of pitting damage.
– Localized, poor application of protective coatings.
– Localized nonuniformities and inclusions in the metal surface.

Pitting corrosion is most difficult to detect, predict, and protect against, because, apart
from a very small area of corrosive attack, corrosion products may often cover the
pits with semipermeable “plugs” that do not form passivating layer. Pitting corrosion
can also be harmful due to the formation of stress areas with fatigue and stress crack-
ing that may be stimulated at the bottom of corrosion pits.
Crevice corrosion appears in confined areas of the metal surface – crevices,
where access to the environment is restricted [6]. This type of corrosion is usually
associated with a stagnant solution in the microenvironment of the site under at-
tack. Such stagnant surrounding, where mass transport is limited, typically appears
in narrow crevices (e.g., under gaskets, insulation material and slightly detached
coating layers) or under deposits. Crevice corrosion is usually initiated by changes
in local chemistry within the crevice, such as depletion of oxygen, inhibitor, changes
in pH, and increase of concentration of aggressive ions (e.g., Cl− and S2−). For exam-
ple, metal ions produced from the anodic reaction can accumulate in the crevice.
The increase of the metal ion concentration in the stagnant region promotes, in
turn, the hydrolysis reaction between the metal ions and water. The progress of this
reaction will gradually lower the local pH in the crevice, facilitating further corro-
sion and increasing its rate. The most common form is oxygen depletion cell corro-
sion. It occurs because the stagnant solution inside the crevice has lower oxygen
content as compared to the surface of a metal or alloy immersed in an aqueous envi-
ronment. Thus, the crevice site forms an anode at the metal surface, whereas the
metal surface in contact with a normally aerated aqueous environment forms a
cathode. The oxygen differentiation corrosion is not limited to crevices. It appears
everywhere where the concentration gradient of oxygen is formed. All metal con-
structions that are designed for seawater, or penetrating wet soil, are susceptible to
the differential oxygen concentrations, as illustrated in Fig. 3.14.

3.6 Hydrogen evolution as accompanying reaction –

role in corrosion: embrittlement and cracking
Hydrogen evolution may occur during the various steps of manufacturing of metal
structures and constructions (e.g., cathodic protection, electroplating, and welding) or
proton reduction process, accompanying the primary corrosion reaction. Hydrogen
may also appear over time through environmental exposure, such as soil, salt, water,
or gradual corrosion of coatings. In the practical use of iron-derived material, such as
steels, atmospheric corrosion plays an important role in hydrogen generation and
112 3 Corrosion

Air Air
Construction Construction
element of steel element of steel

O2 concentration OR O2 concentration
decreases decreases
Corrosion: O2 reduction Corrosion: O2 reduction
oxidation and oxidation and
dissolution of Me Sea water dissolution of Me Wet soil

Fig. 3.14: A scheme of metal corrosion due to the difference in oxygen concentration.

partition into the corroding construction, even though the primary cathodic reaction
is the reduction of oxygen. The cathodic reaction of hydrogen evolution results from
the combination of several elementary steps. In acidic media the following reactions
will proceed:

Volmer reaction: H3 O + + e $ Had + H2 O

Tafel reaction: Had + Had $ H2
Heyrovsky reaction: Had + H3 O + + e $ H2 + H2 O

Whereas in alkaline solutions:

Volmer reaction: H2 O + e $ Had + OH −

Heyrovsky reaction: Had + H2 O + e $ H2 + OH −

Obviously, the source of electrons is the anodic dissolution (corrosion) of the metal.
Most of the hydrogen atoms present at the surface of a metal desorb from the
surface as H2 formed in Tafel and Heyrovsky reactions. However, small part of hydro-
gen atoms being adsorbed, Had, produced during Volmer reaction become absorbed,
Hab, by the surface layer of the metal and can diffuse through the metal structure,
accumulating at the grain boundaries, defective microscopic voids, creating pressure
from within the metal, for instance, in iron, the specific volume of the absorbed hy-
drogen, VH, can reach approximately 2.6 cm3/mol corresponding to the pressure build-
up to approximately 105–107 atm. This pressure can easily reduce metal strength up to
the point where it cracks – the so-called hydrogen embrittlement or hydrogen-
assisted cracking. Hydrogen embrittlement is the process by which metals become
brittle and susceptible to fracture due to the introduction and subsequent diffusion of
hydrogen into the metal. Metals become more susceptible to hydrogen embrittlement
under stress: cracking will initiate faster when the metal structure is subjected to
stress. Although hydrogen atoms embrittle a variety of substances, including steel and
other metals and alloys, hydrogen-assisted cracking of high-strength steel is of the
most importance.
Hydrogen embrittlement can be prevented by several methods, mainly relying on
the minimization of the contact between the metal and hydrogen, particularly during
 3.7 Protection against corrosion 113

fabrication and use. Often preheating or postheating can be applied to force hydrogen
to diffuse out of a structure before the latter can be used. However, one has to con-
sider that if high temperature is applied, hydrogen can react with carbon atoms pres-
ent in steel, forming methane. Methane molecules will not diffuse out of steel but
accumulate in tiny voids and grain boundaries at high pressures, also initiating crack-
ing of the material.
Much more about corrosion fundamentals, as well as about some other topics de-
scribed in the text here, the interested reader can find in [1–9].

3.7 Protection against corrosion

Protection against corrosion is a must nowadays because anything made of iron or

steel is a prime target of corrosion processes. Therefore, we will focus now on corro-
sion inhibitors. They are chemical substances used as additives in very small quanti-
ties to the environment. These substances can inhibit and either minimize or prevent
corrosion. An efficient inhibitor should be environmentally friendly, economical for
application, and producing stable effect even if used in small concentrations. The effi-
ciency of inhibitors, P, is evaluated by making a direct comparison of corrosion rate
with and without the presence of the inhibitor:

P = ðv0 − v=v0 Þ · 100

where v0 is the corrosion rate in the absence of inhibitor, and v is the corrosion rate
in the same environment in the presence of inhibitor.
A qualitative classification of inhibitors is shown in Fig. 3.15, yet for the sake of
clarity and to stay within the scope of this book, we will discuss on cathodic and an-
odic inhibition; however, it can be seen that, for example, cathodic inhibitors can also
be classified as scavengers, decreasing oxygen concentration (hydrazine – N2H2) or
increasing pH (Sb2O3) (Fig. 3.15).
Let us now recall eq. (3.7), relating the corrosion current with the exchange cur-
rents of metal, the i0,Me, environmental couple, i0,2, and the equilibrium potential dif-
ference of both participating couples, Eeq,2 – Eeq,Me:
  
1 zF Eeq, 2 − Eeq, Me
icorr = ði0, Me × i0, 2 Þ × exp
2
RT

Close inspection of this equation provides us immediately with tools that rely on the
decrease of the product of the exchange currents and can be used to protect the metal
against corrosion.

1. Corrosion inhibitors I. Inhibition of corrosion by the addition of cathodic inhibitors;

decrease of i0,2 no change of i0,Me.
114 3 Corrosion

Classification of inhibitors

Inetrface inhibitors Environmental conditioners

Vapor phase Liquid phase

Cathodic Mixed (adsorption)

Anodic (passivator)

Poison Precipitator Chemical Film forming

Physical

Fig. 3.15: Classification of corrosion inhibitors.

This type of cathodic inhibition can be achieved in two ways, all of them lead to a
shift of Ecorr toward the equilibrium potential of metal:
– by inhibiting the reduction of H+ as cathodic reaction, for example:
by adding Sb2O3. Then, the following reaction will proceed:

Sb2 O3 + 6 H + + 6 e ! 2 Sb + 3 H2 O

– by inhibiting oxygen reduction as cathodic reaction, for example:

2 N2 H2 + 6O2 ! 4 NO3 − + 4 H +

or

2 SO3 2 − + O2 ! 2 SO4 2 −

Graphically, this type of protection can be presented on Evans diagram (Fig. 3.16).

E
Er,2 (a)
(b)
Ecorr(a)
Ecorr(b) Ecorr

Eeq,Me

log |i| (b) log |i| Fig. 3.16: Evans diagram in the case of the presence of a
log |i| (a) cathodic inhibitor.
 3.7 Protection against corrosion 115

In this diagram, Ecorr(a) and Ecorr(b) stand for corrosion potentials before and after
inhibitor was added, respectively, and i(a) and i(b) are corrosion currents before and
after the inhibition, respectively.

2. Corrosion inhibitors II. Inhibition of corrosion by the addition of anodic inhibitors;

decrease of i0,Me no change of i0,2. Ecorr increases toward Eeq,2.
Anodic inhibitors are chemical substances that alter the anodic reaction of met-
als, forcing their surface into the passivation region (see Pourbaix diagram). They also
form the protective oxide films on the surface of metal; therefore, they are frequently
referred to as passivators. In the kinetic Evans diagram, the oxidation reaction cur-
rent diminishes (from curve a to curve b in log scale), shifting also the corrosion
potential Ecorr toward the equilibrium, Nernstian potential of the environmental
counterpart, Eeq,2, as shown in Fig. 3.17.

E
Er,2 (b)
Ecorr (b)
(a)
Ecorr (a)
Ecorr

Er,Me

log |i| (b)

log |i| (a) Fig. 3.17: Evans diagram change induced by an anodic inhibitor.

Examples of anodic inhibitors include

– chromates,
– molybdates,
– tungstates,
– nitrites,
– orthophosphates,
– organic compounds containing heteroatoms (S, O, N), double bonds, and/or polar
functional groups.

Due to the chemical characteristics of anodic inhibitors, they are considered environ-
mentally unsafe.
Most authors suggest that synthetic organic inhibitors work by being adsorbed
onto the metal surface through electrostatic interactions or interactions with empty d-
orbitals of iron atoms. Nevertheless, there exist still major drawbacks of such com-
pounds related to their toxicity, non-biodegradability, or production costs. Therefore,
to overcome these problems, the researchers turned to cheap, eco-friendly, green cor-
rosion inhibitors, derived from plant extracts – seeds, fruits, roots, barks, and leaves.
These extracts contain a large number of phytochemicals with their green anticorro-
sion activity due to the structural presence of nitrogen, oxygen, or sulfur heteroatoms,
116 3 Corrosion

as well as unsaturated organic rings and double bonds which can adsorb on the sur-
face of the metal and provide effective inhibition against the corrosive environment.
Moreover, the source plants for such inhibitors are renewable, of relatively high
availability, and reasonably simple and low-cost extract processes, utilizing environ-
mentally safe solvents like water or alcohol. However, this approach is still far from
being commercialized. Much more on the principles, design, and utilization of eco-
friendly corrosion inhibitors can be found in [11].

3.7.1 Electroplating

Electroplating is the method that uses electrochemical deposition of a metal to form a

thin, coherent metal coating layer on another metal that serves as an electrode. Elec-
troplating can be used to protect iron and iron-derived metals (e.g., steel) against cor-
rosion, but more generally, it can be used to change the surface properties (functional
and/or decorative) of metal, such as wear resistance, lubricity, abrasion, or even the
aesthetic qualities. As it forms an additional layer of metal, it may also be used in elec-
troforming to build up forms on a model or to add size on a metallic substrate. The
metal to be plated is connected as the cathode in a DC circuit, whereas the metal to be
electrodeposited can be dissolved in the electrolyte or can be the anode that is sacri-
ficed by its electro-oxidation. The choice of metal for the protective layer depends on
its purpose:
– functional/decorative, or
– sacrificial coating.

A functional or decorative layer is used primarily to enhance the functionality of the

metal or its attractiveness. Common metals used for such a purpose include nickel,
copper, chromium, zinc, and tin; less frequently, for special purposes (e.g., electronics
and jewelry) these will be gold, silver, platinum, ruthenium, rhodium, palladium, and
iridium. Let us present some of the electroplating examples below.
Nickel electroplating is widely used for its corrosion-resistant properties. Such cov-
erage is an excellent replacement for chrome electroplating because the overall proce-
dure of nickel electroplating is less harmful to the environment. Additionally, the nickel
layer is aesthetically pleasing. The layer of nickel also provides good lubricity.
The silver coating layer in engineering is done mainly to increase the electronic
and heat conductivity of the base metal, a feature necessary for electronic devices, as
well as for its excellent corrosion resistance. One has to consider also the enhance-
ment of the aesthetic look of the product coated with polished silver layer. This fea-
ture is due to the very good light reflectance of silver.
We now address the so-called sacrificial coating, where the coating layer serves
not only as a mechanical barrier against corrosion of underlying iron or steel but
participates actively in its protection. The metal used for such coating is sacrificial,
 Bibliography 117

that is, it is used up in the reaction. Common examples are zinc and cadmium, the
latter being gradually prohibited because of the environment. The remarkable effec-
tiveness of zinc electroplating and relatively low cost of such procedure have made
it the most common and widespread choice for protecting surfaces in all types of
processes. Almost one-third of all zinc metal is used for such protective coatings.
When zinc is electroplated onto the surface of ferrous metals, it is exposed in most
cases to atmospheric oxygen and humidity, producing zinc hydroxide:

2 Zn + O2 + 2 H2 O ! 2 ZnðOHÞ2

Then, in the presence of atmospheric carbon dioxide (greenhouse gas), the hydroxide
combines to produce an insoluble layer of zinc carbonate:

ZnðOHÞ2 + CO2 ! ZnCO3 + H2 O

This insoluble, gray layer of zinc carbonate adheres strongly to the underlying zinc,
sealing it from further rusting. This way also the iron or steel construction covered
with a protective zinc layer is safe from corrosion.
Finally, electroplating creates a much stronger bond with the surface of the mate-
rial to protect it from corrosion when compared with other methods, such as painting.

Bibliography

[1] Revie RW, Uhlig HH, Corrosion and corrosion control. An Introduction to corrosion science and
engineering. John Wiley & Sons, Inc., 2008.
[2] Landolt D. Corrosion and surface chemistry of metals. CRC Press, Taylor & Francis Group, LLC, Boca
Raton, FL, 33487, 2007, first edition, EPFL Press.
[3] Pourbaix M. Lectures on electrochemical corrosion. Plenum Press, New York-London, 1973.
[4] Pedeferri P. Corrosion Science and engineering. Engineering Materials book series.
Lazzari L, Pedeferri MP, Springer Nature Switzerland AG, 2018.
[5] Szklarska-Smialowska Z. Pitting corrosion of metals. Natl. Assn. Corr. Eng.. 1986.
[6] Szklarska-Smialowska Z. Pitting and crevice corrosion. Houston Texas, NACE International Corrosion
Soc., 2005.
[7] Tait WS. Electrochemical corrosion basics. In: Kutz M., ed., Handbook of environmental degradation
of materials. 3rd ed., William Andrew Applied Sci. Publishers., Elsevier Inc. 2018, Chapter 5,
pp. 97–115.
[8] Popov BN, Lee J-W, Djukic MB, Hydrogen permeation and hydrogen-induced cracking.
In: Kutz M., ed., Handbook of environmental degradation of materials. 3rd ed., William Andrew
Applied Sci. Publishers., Elsevier Inc. 2018, Chapter 7, pp. 133–162.
[9] Ohtsuka T, Nishikata A, Sakairi M, Fushimi K. Electrochemistry for corrosion fundamentals.
Springer, 2018.
[10] Lasia A. Electrochemical impedance spectroscopy and its applications, Springer-Verlag, 2014.
[11] Guo L.,Verma Ch., Zhang D., (Eds.), Eco-Friendly Corrosion Inhibitors. Principles, Designing and
Applications, Elsevier, 2022.
4 Electrocatalysis

4.1 General remarks

Electrocatalysis is a type of heterogeneous catalysis that results in the increase of the

rate of a reaction occurring at the electrode–electrolyte interface. In this type of catal-
ysis, the electrode surface or electrode surface covered with catalytic species (e.g.,
metal particles and bimetal particles) play the role of catalyst. The catalyst assists in
transferring the electrons between the electrode and reactants or/and facilitates the
chemical transformation of adsorbed intermediate formed during reaction. Before
going further we will briefly characterize the catalysis, looking for differences in both
types.
Catalysis is the process in which the addition of some substance increases the
rate of chemical reaction. This substance named catalyst is not consumed and not
changed chemically in the overall reaction. Catalyst reacts with one or more reactants
to form an intermediate, which subsequently gives the final reaction product and re-
generated catalyst:

A + B ! P, A + K $ A    K, A    K + B ! K + P

where A, B – reactants, P – product, K – catalyst, A    K – intermediate.

In the presence of a catalyst, chemical reactions occur faster because the catalyst
provides an alternative reaction pathway with lower activation energy than in the
case of a noncatalyzed reaction. Figure 4.1(a, b) illustrates the changes in energy and
reaction pathway during catalyzed reaction.

(a) (b)

Activated
complex (AK....B)*
Ea (A....B)*
Potential energy
Potential energy

(A....K)*
Activation
energy

AK + B

A+B A+B+K

P+K
P

Reaction coordinate Reaction coordinate

Fig. 4.1: Schemes showing the changes in energy and reaction pathways between the noncatalyzed (a)
and catalyzed (b) chemical reactions. A, B – reactants; P – product; K – catalyst.

https://doi.org/10.1515/9783111160986-004
 4.1 General remarks 119

Catalysts have no effect on the chemical equilibrium and do not change the equilib-
k
rium constant K = k1 , since the rate constant of both the forward k1 and the backward
2
k2 reactions are affected in the same way. In consequence, the catalysts cannot change
the standard Gibbs free energy of reaction Δr G0 . Two types of catalysis can be distin-
guished – homogeneous and heterogeneous one, depending on whether a catalyst is
in the same phase as the substrate or not. Homogeneous catalysis means that catalyst
and reagents exist in the same phase. Such kinds of electrocatalysis can be found
when, for instance, soluble organometallic compounds are used as catalysts in the so-
lution together with reactants. In heterogeneous catalysis, the solid-state catalysts are
used and reactions take place on their surface. The reactions occur when the reagents
are strongly adsorbed, chemisorbed on the catalyst surface, and some bonds are
formed between a catalyst and adsorbed molecules, resulting in intermediate forma-
tion. On the other hand, if the strength of bonds becomes too great, the activity of
catalyst deteriorates because the adsorbed molecules block the active sites of the cata-
lyst. The activity of the catalysts is indicated by the “volcano” curve and will be dis-
cussed later. In heterogeneous catalysis, the total surface area of the catalyst has a
great impact on its activity. Because of that, some materials with very large surfaces
are used as the carriers for catalyst species (e.g., zeolites or other mesoporous materi-
als). The other way to improve the catalytic activity and increase the reaction rate is
to minimize the size of catalyst particles.
The electrocatalysis has many common features with classic heterogeneous cataly-
sis. Many catalytic materials used in both types of catalysis are the same such as Pt, Pd,
Ni, or oxides. In both cases, the rate of reaction depends on the catalyst properties such
as chemical composition and surface area of catalyst. Also in both cases, the catalyst
increases the rate of reaction by the acceleration of the slow step or by changing the
reaction pathway. What does distinguish the electrocatalysis from chemical catalysis?
The main difference is the dependence of the electrocatalytic reaction rate on electrode
potential, more exactly on the potential difference across the interface, which can be
modified by an external source of voltage. Following equations show the differences in
the rates (ѵ) of chemical and electrochemical reaction and roughly in the catalytic and
electrocatalytic reactions of first order, as well:


k1 ΔG0#
chemical catalysis: A ! B, vr = cA k1 , vr = cA Pexp − (4:1)
RT

k2 j
electrocatalysis: An + + ne ! B, vr = nFcA k2 , vr =
nF



ΔG0# αnFE
vr = cA Pexp − exp − (4:2)
RT RT
120 4 Electrocatalysis

where ΔG0# is the Gibbs free energy of activation (activated complex theory), P is the
preexponential factor, E the electrode potential, j the current density, and α the trans-
fer coefficient.
From eq. (4.2), it is clearly seen that small changes in electrode potential cause
the tremendous increase of reaction rate, even by several orders of magnitude.
The other features that differentiate the electrocatalytic and catalytic reactions are:
(i) the participation of electrons in the electrocatalytic reaction (they must be supplied to
or withdrawn from the catalyst surface), (ii) the presence of nonreactive species in solu-
tion such as solvent and ions, (iii) the electric field existing at the electrode–solution in-
terface can significantly change the surface properties of catalyst and affect the rate of
reaction, and (iv) temperature of electrocatalytic reactions is much lower than the chem-
ical catalytic reaction.

4.2 How to compare the activity of catalysts in electrochemical

reactions?
In practice, the problem that arises is the selection of well-determined quantitative
parameters characterizing the activity of electrocatalyst. Generally in electrochemical
experiments, the relation current–potential is measured. To eliminate the influence
of the differences in the surface area of electrodes, the results are presented as cur-
rent density–potential plots. In many cases, the geometric area of the electrode is
used for current density calculations, but in the case of electrocatalysis, we have to
use the “true” (real) working surface area. This real surface area can be determined
by in situ or ex situ methods, such as hydrogen adsorption, voltammetry, BET (Bruna-
uer S, Emmet P, Teller E isotherm), and porosimetry. If metal nanoparticles are used
as catalyst on conducting substrate, it is possible to determine their amount by elec-
trochemical quartz microbalance (see Section 2.3).
Let us consider the same oxidation reaction occurring with electrocatalysts 1 and
2. The current density in both cases is described by the Volmer–Butler equation
(Chapter 1):

 
ð1 − α1 ÞnF E − Eeq
j1 = j0, 1 exp (4:3a)
RT

 
ð1 − α2 ÞnF E − Eeq
j2 = j0, 2 exp (4:3b)
RT

η = E − Eeq

If the pathway of reaction is the same in both cases, we may assume that transfer
coefficients α are equal: α1 = α2 . Then the ratio of current density is expressed by the
ratio of exchange current densities:
 4.2 How to compare the activity of catalysts in electrochemical reactions? 121

j1 j0, 1
= (4:4)
j2 j0, 2

Since the exchange current density is related to the standard rate constant (Chapter 1),
we may also use this parameter for comparison of different electrocatalysts. The
higher the exchange current density value, the better the electrocatalyst for reaction
is. We can also fix the current density values, for example, 50 mA/cm2 and compare
the overpotential values at which such current density is reached. The lower the over-
potential value, the better the electrocatalyst is. In practice, if the reaction is irrevers-
ible or very complicated (Eeq , ƞ are not determined), we use working Tafel expression
E = a + blogj jj and compare the catalyst at chosen potential or current density. It is il-
lustrated in Fig. 4.2(a, b) for the same reaction occurring at two different catalysts
(electrodes).

(a) (b)

E(V)
j (mA/cm2)

Catalyst (1) Catalyst (1)

Catalyst (2)
Catalyst (2)
E2

E1

A B

j (mA/cm2) E(V)

Fig. 4.2: Schemes illustrating how to compare the activity of catalysts in an electrochemical reaction.
Scheme (a) presents potential–current density curves for two catalysts. The best catalyst has a higher
value of current density at the same potential value. Scheme (b) presents current density–potential curves
for two catalysts. The best catalyst has a lower value of potential at the same current density value.

It is clearly seen that in the case (a), in the region (A) the first (1) and in the region (B)
the second (2) catalyst are more active showing a higher value of current density.
At constant current density (case b), this catalyst is better, for which the same current
value is attained at lower potential (catalyst 1) [1, 2].
122 4 Electrocatalysis

4.3 Electrocatalysts

The optimum electrocatalytic material containing the catalyst species and the support
should be characterized by large, catalytically active surface area with active catalytic
sites distributed uniformly across the support surface. The support should prevent the
agglomeration of catalytic species, exhibit the activating interaction with them, and pro-
vide good electrical conductivity to catalytic centers. In heterogeneous electrocatalysis,
plenty of materials of various compositions and sizes are applied as catalysts. Below we
will briefly describe some materials used as catalysts and supports.

4.3.1 Metals, alloys, and oxides

The metal electrodes are widely used in electrochemical laboratories and in techno-
logical processes: electrolysis, electroplating, as sources of energy. Their usefulness is
only limited by the potential range in which they are electrochemically stable, and do
not undergo dissolution or oxidation. The catalytic activity of metals is mostly gov-
erned by the electron structure of their atoms, more exactly by the existence of free
unpaired electrons in d-orbitals and their ability to form bonds with adsorbed reac-
tants. Catalytic activity of metals varies from metal to metal and from reaction to re-
action. The high catalytically active metals belong to subgroup VIIIB of elements.
These are irons: Fe, Ru, Os; cobaltous: Co, Rh, Ir; platinums: Ni, Pd, Pt. Platinum is the
most universal catalyst and is stable in wide potential range in acidic and alkaline
solutions. Platinum and the metals of its group are used in many reactions such as
hydrogen evolution (HER), oxygen reduction (ORR), and organic compound reduction,
for example, nitrobenzene and benzofuroxan.
All these materials are very expensive and their amount is limited and because of
that the scientific works went and still continue in two directions: (i) the formation of
bi- or multicomponent catalysts with very high activity and (ii) to minimize the catalyst
amount by decreasing particle size of a catalyst (nanoparticles, ad-atoms are used) or
by application of thin films as electrode. Catalysts containing two metals were exam-
ined in many reactions taking place in fuel cells, such as oxidation of formic acid (FA),
oxidation of methanol, and HRE. Bimetallic catalysts can be prepared in many ways.
The most straightforward way is to prepare a bulk alloy by metallurgy, that is, melting
of two components together, followed by homogenization. The other methods used are
chemical precipitation and further annealing or the electrochemical deposition (Chap-
ters 5 and 6). Some surface alloys were prepared by using layer-by-layer deposition in
the UHV (ultra-high vacuum) system and their catalytic activity was tested in the oxi-
dation reaction of CO. The subject of intensive studies is the difference between the
bulk and surface composition of alloys [3]. It is well documented that the annealed
surface of alloy is enriched in Pt amount, when Pt bulk alloys composition is also Pt
rich, for example, Pt3Me (Me – Sn, Fe, Co, Ni, Ru, Ag). The most popular and highly
 4.3 Electrocatalysts 123

investigated catalyst is Pt–Ru alloy. In the late 1960s, the strong synergy effect was
found for this system. The increase of three order magnitude in catalytic activity was
observed at anodic oxidation of methanol in comparison with pure Pt, whereas pure
Ru is inactive for this reaction. This alloy–catalyst has also another feature; in contrast
to pure Pt, it is not poisoned by COads species. The mechanism currently accepted is the
adsorption of methanol molecules at Pt sites of the surface and their partial oxidation:

Pt + CH3 OH ! PtðCH3 OHÞads

Pt−CH3 OHads ! Pt−ðCOÞads + 4 H + + 4 e

The oxidation of COads is accelerated by molecules containing oxygen atom and ad-
sorbed on Ru sites close to Pt sites:

Ru + H2 O ! Ru−OHads + H + + e

Ru−OHads + Pt− ðCOÞads ! Pt + Ru + CO2 + H + + e

The catalytic activity Pt–Ru depends on the surface concentration of Ru. The effect is
significant even at very low Ru concentration. For these pathways, the quantitative
model was developed connecting the oxidation rate of methanol with a surface con-
centration of Ru in alloy. Generally, the methanol oxidation reaction may be de-
scribed by the so-called bifunctional mechanism.
The other organic compound with potential application in the fuel cells is FA. The
mechanism of electrooxidation of FA on Pt and other Pt-group metals was widely in-
vestigated. Dual pathway of reaction was proposed:
dehydrogenation
HCOOH ! CO2 þ 2 Hþ þ 2e

dehydration
HCOOH ! COads + H2 O

The mechanism of FA electrooxidation on Pt electrode and Ru differs. In the first case,

the overall reaction of HCOOH on Pt is dehydrogenation reaction with the dehydra-
tion reaction being the blocking reaction. On Ru electrode, oxidation of HCOOH pro-
ceeds via dehydration and formation of adsorbed CO species that are finally oxidized
to CO2 ðCO + H2 O ! CO2 + 2 H + + 2 eÞ. The rate of HCOOH electrooxidation on Ru
electrode is about 20 times lower than on the Pt electrode. The application of Pt–Ru
alloy increases a few times the rate of FA oxidation in comparison with Pt as being a
result of the increase of COads oxidation rate. On the Pt–Ru alloy, two pathways are
included in the oxidation reaction. The best results were obtained for the alloy with
surface Ru concentration of 50%. To know much more about bimetallic electrocata-
lysts, interested readers will find in refs. [2, 3].
124 4 Electrocatalysis

The surface of many metals is covered with spontaneously formed oxides, and
other more noble metals become covered during anodic polarization. The composition
and properties of oxides obtained electrochemically depend on the potential and elec-
trolyte used. In their structure, vacancies can be found as a result of nonstoichiometric
formation and lack of oxide or metal ions in a crystal lattice or on the surface. Many of
oxides are semiconductors characterized by energy band >2.5eV and low intrinsic con-
ductivity. The special doping by donor or electron acceptors increases the conductivity
and changes the type of conductivity from intrinsic to n or p (Chapter 10). The semi-
conducting oxides are widely used in the photocells for water splitting and in photo-
catalysis (Chapters 11 and 12). The application of oxides as electrodes is limited. They
are not stable, undergoing the reduction during cathodic reactions and often dis-
solving in acidic solution. The best-known oxide applied in batteries is PbO2 (Chap-
ter 8). The electrocatalytic properties have nickel oxide widely used in the oxidation
of organic compounds. The oxide Co3O4 was tested as a catalyst in ORR [2]. In the
last decade, the mixed oxides are of scientific interest as catalysts in chemical and
electrochemical catalysis. The term mixed oxides is applied to solid ionic compounds
containing the oxide anion O2– and two or more types of cations. Typical examples are
FeTiO3 (a mixed oxide of iron (Fe2+) and titanium (Ti4+) cations) and compounds with
spinel structure such as cobaltites MCo2O4 or perovskite structure RTO3, where R is a
rare earth element (La) and T is the transition metal (e.g., Ni, Co, and Mn). Since the
perovskites are not conducting they have to be doped (usually with BaO) before being
used as the electrode. The perovskites catalyze the oxygen evolution reaction from alka-
line solutions. Many of the perovskites were tested in the reaction of oxygen reduction
for future application in fuel cells [4, 5].

4.3.2 Carbon catalysts and supports

There are plenty of carbon materials that are applied in heterogeneous chemical cataly-
sis as well as in the electrocatalysis [2]. Materials such as black carbon, activated carbon,
and Vulcan carbon have a very large specific surface area (from 200 for Vulcan to
3,000 m2/g for activated C), but the high electrical resistance limited their usefulness in
electrochemistry. However, when mixed and sintered with conducting carbons (graph-
ite powder), they are often applied as a support for catalyst particles. Glassy carbon
(GC) often named vitreous carbon has large conductivity, almost smooth ideal surface,
and the high values of overpotentials of water oxidation and reduction. The potential
window of GC is large, the material is stable in acidic and alkaline solutions, because of
that GC is widely used as an electrode material in electrochemical investigation. Vitreous
carbon produced as the foam is called reticulated vitreous carbon. It is a good conduct-
ing material with developed surface and large volume and is often used in electrochem-
istry as three-dimensional electrode for nanoparticles, conducting polymer deposition.
Graphite is another carbon material used in the form of paste as electrode material.
 4.4 Catalyst activity 125

Graphite is a good conductor, is stable, and its surface area is large but unfortunately
should be determined before experiments. Highly oriented pyrolytic graphite (HOPG) is
an analog of metal single crystal (in carbon chemistry) with preferential crystallographic
orientation. HOPG is often used for the in situ determination of adsorption or underpo-
tential deposited (UPD) on adatom structure by means of AFM (atomic force micros-
copy) and scanning tunneling microscopy (STM). In recent years, the most intensively
investigated material is graphene (Nobel price 2010, Andre Geim, Konstantin Novoselov),
a two-dimensional carbon nanomaterial. The graphene (G) exhibits interesting and reach
features that the perspectives of application are very wide. Structurally, graphene is a
one-atom-thick planar sheet of sp2 bonded carbon atoms closely packed in a honeycomb
lattice. This material has unique physicochemical properties including excellent elec-
trical and thermal conductivity, fast charge transfer mobility, mechanical flexibility,
and very large specific surface area (theoretically 2,630 m2/g for the single layer). Besides
graphene, some of its derivatives can be obtained during synthesis such as graphene
oxide (GO) and reduced graphene oxide (RGO). They contain oxygenated functional
groups: − O, = O, − OH, or others, which can be involved in various chemical reactions.
The graphene, GO, and RGO can be applied in electrocatalysis in two ways: as the cata-
lytic support or as carbocatalyst. Graphene, RGO-supported Pt, Au particles, binary alloy
composed of Pt, Au, Co, Ni, and oxides such as Co3O4, MnO2, and Cu2O were explored as
ORR catalyst for fuel cell application [6].
Recently, nitrogen (N)-doped carbon nanomaterials have been developed. It was
found that the incorporation of nitrogen in graphene (NG) or its derivatives results in
enhancement of catalytic activity of metal particles due to improvement in metal–
graphene binding. What is interesting is that NG itself exhibits the enhanced electrocata-
lytic activity toward the reduction of oxygen in comparison with undoped graphene. Re-
cently, efforts were made to apply carbocatalyst (metal-free catalyst) such as graphene
and graphene doped with N or other elements (e.g., B, S, Se, and P) as catalyst for oxygen
reduction. However, the application of other elements than N has a small impact on the
catalytic activity of graphene. The rich literature on this topic can be found in the review
[6]. The other carbon nanomaterials that were used as a support for catalysts are carbon
nanotubes (CNTs) and carbon nanofibers (CNFs).

4.4 Catalyst activity

Why some metals have catalytic properties? If we compare the electron structure of
Pt, Ru, Ir, Co, and Ni atoms, we see common features. All these atoms have unpaired
electrons in d-orbitals. Atoms of Pt and Ru have also unpaired electrons in s orbitals:
Pt – 5s2p6d96s1 and Ru – 4s2p6d75s1. Such atoms can interact with species that are sour-
ces of electrons for creating the electron pairs. This interaction results in catalytic ac-
tivity in the formation of intermediates. One may expect that we can predict the
catalytic behavior of a metal and its catalytic activity on the basis of the electron
126 4 Electrocatalysis

structure of atoms, we can – but only to some extent. The large effort was made to
find the correlation between bulk properties of a catalyst such as an electron work
function (Chapter 1), heat of sublimation, melting temperature, or other physicochemi-
cal parameters and the activity of a catalyst. We will briefly describe here the electron
work function (simply, work function) effect and after that, since the heterogeneous
electrocatalytic reaction is a surface process, we will concentrate on the adsorption and
size effect [1, 2, 7].

4.4.1 Electron work function effect

In the electrochemical reaction, electrons are transferred from the metal to reactants in
the solution, so the rate of reaction should increase with lowering of the work function.
But it is not the case. Two work functions should be distinguished: one term defines the
work that must be done by the external force in transferring the electron from the
metal into vacuum ΦM and other term means the work of electron transferring from
the metal to the contacting electrolyte ΦE . It was shown [2] that work function ΦE is
independent of metal nature, when it is determined at the same electrode potential.
However, this is true if some simple redox reaction is considered such as Fe 3+ + e!Fe
2+
. The measurements of activation enthalpy ΔHa0# (activation heat) carried out for this
reaction on various electrodes have shown independence of ΔHa0# of ΦM [1, 2]. It means
that the activation enthalpy of the reaction that involves only charge transfer is inde-
pendent of the electron structure of the electrode. The decisive role in electrocatalytic
reactions plays the surface process – adsorption.

4.4.2 Adsorption impact

The extent of the adsorption effect is dependent on the bond strength between catalyst
and adsorbed reactant or intermediate species. Therefore, in consequence, it is depen-
dent upon the Gibbs free energy of adsorption ΔG0ads , degree of surface coverage, θ, and
influence of potential on adsorption parameters. If we consider, very roughly, few cata-
lysts and an increase of bonding strength between them and adsorbed species, we
should expect both the increase of θ and increase of the rate of the formation of reac-
tion products. If the increase of θ in a series of catalysts is continued, their free surface
for adsorption decreases causing also the decrease of reaction rate. It is the clash of
opposite tendencies; in consequence, the reaction rate will increase until it reaches its
maximum and then it decreases. The plots of the rate of reaction (i.e., exchange current
density) against the enthalpy of adsorption (the heat of adsorption is connected with
the strength of adsorption bond) for different catalyst and the same reaction will show
the volcano (bell) shape (Fig. 4.3) [8, 9].
 4.4 Catalyst activity 127

Pt
Rh, Ru
log j0

Ir

Au
Fe Ni
W
Co
Mo
Zn Ti
Sn Bi
Ag

Pb
Tl Cd
In

Me-H strength bond (kJ/mol)

Fig. 4.3: Volcano plot for the hydrogen evolution reaction, showing exchange current density dependence
on metal–hydrogen strength bond. Adapted from [9].

Let us consider in more detail the simple two-step reaction A ! I ! B with moderate
adsorption of intermediate I, so the equilibrium between adsorption and desorption
can be established. Further, we assume that (i) the surface of catalyst (electrode) is
homogeneous, which means that the Gibbs free energy of adsorption is coverage-
independent, (ii) the active sites are well defined with one molecule occupying a single
site, and (iii) the active sites are energetically equal and there is no interaction between
the adsorbed species. At such assumptions, the Langmuir isotherm is applicable:

kads cA θ
θ= , or in the form = Kads, 0 cA (4:5)
kdes + kads cA 1−θ

where kads , kdes are rate constants of adsorption, desorption, respectively, ci is bulk
kads
concentration of reactant; in this case cA , Kads, 0 = is the equilibrium constant for
kdes
adsorption and is related to the standard Gibbs free energy of adsorption ΔG0ads ,


ΔG0ads
Kads, 0 = exp − (4:6)
RT

In steady state, the rate of reaction: A ! I ! B, vr = kdes θ,

kdes kads cA
vr = (4:7)
kdes + kads cA

Keeping in mind that rate constants can be described by the Arrhenius equation and
depend upon the activation energy of adsorption (or desorption), Ea , or on the enthalpy
of activation (heat of activation) of adsorption ΔHa0# , one can write the equations:
128 4 Electrocatalysis

0# 
ΔHa
kads = Cexp (4:8a)
RT


′ ΔHa0#
kdes = C exp − (4:8b)
RT

ΔHa0# = Ea − RT

Putting kads , kdes expressions from eqs. (4.8a) and (4.8b) in eq. (4.7), we obtain the vol-
cano shape dependence between the rate of reaction and parameters characterizing
energetically the adsorption [2].
Langmuir isotherm was derived under assumptions that adsorption sites are equiva-
lent, independent from each other and there is a lack of interaction between the ad-
sorbed species. There are many experimental results where the Langmuir isotherm is not
valid or is valid in a limited range of θ. The other isotherms were introduced, which
take into account the inhomogeneity of catalyst surface (Temkin isotherm) and lateral in-
teraction between adsorbed molecules (Frumkin isotherm) [7]. Both these isotherms were
delivered under assumption that ΔG0ads depend linearly on θ, more exactly the ΔHads 0
is θ
dependent:

ΔG0ads, θ = ΔG0ads + rθ (4:9)

The r parameter in Temkin isotherm is connected with inhomogeneity factor f, whereas

in Frumkin isotherm, parameter r = fint RT, where fint is the interaction factor. This fac-
tor has a positive value when repulsion interaction is involved or negative for attrac-
tion interaction between the adsorbed species. Applying the expression ΔG0 = − RTlnK
and eq. (4.9) one can easily derive the equation describing the equilibrium constant
Kads,θ at coverage θ:


rθ
Kads,θ = Kads, 0 exp − = Kads,0 expð − fint θÞ (4:10)
RT

going further Frumkin’s isotherm (1928) has the form:

θ
= Kads,θ ci (4:11a)
1−θ

or

θ
= Kads,0 ci expð − fint θÞ (4:11b)
1−θ
 4.4 Catalyst activity 129

while Temkin’s isotherm (1939) can be written as follows:

1 1
θ= lnKads,θ + lnci (4:12a)
f f

or

expðf θÞ = Kads,θ ci (4:12b)

If θ is small, the difference 1 − θ ≈ 1 and Frumkin’s isotherm becomes Temkin’s iso-

therm, however, with different meanings of f factor.
The coverage of catalyst surface depends not only on the bond strength, electron
structure of catalyst, and surface inhomogeneity but also on the applied polarizing po-
tential. The influence of potential upon coverage is shown below for Frumkin and Tem-
kin isotherms, and the expressions are derived by using basic equations (Chapter 1):

ΔG = ΔG#0 ± ΔGE , ΔGE = − nFE


θ nFE
Frumkin isotherm: expðfint θÞ = Kads,0 ci exp (4:13)
1−θ RT


nFE
Temkin isotherm: expðf θÞ = Kads,0 ci exp (4:14)
RT

Both isotherms show a linear dependence of coverage on potential:

2.3 nF
θ= logKads,0 ci + E (4:15)
f fRT

However, Frumkin isotherm can be used in this form only in such coverages where
θ
the assumption ≈ 1 is valid.
1−θ

4.4.3 Size effect

Except for the surface energy and adsorption, the surface factors that influence the activ-
ity of catalyst are the surface composition, the orientation of crystals creating the surface,
the surface defects, and catalyst size and shape. All these factors may be changed during
catalyst preparation and surface pretreatment; therefore, the surface state should be care-
fully controlled before catalyst application by spectroscopic and microscopic methods.
In the heterogeneous electrocatalysis, the catalysts are used in various forms, such
as planar electrodes, thin solid films, particles, nanoparticles, adatoms, monolayers,
and others. All these forms vary in size. The application of high-resolution microscopy
for surface characterization points out that they are nanostructured materials. Even the
130 4 Electrocatalysis

surface of the planar electrode is covered by many defects of nanoscale size. Some ex-
perimental results point out that the size of catalyst particles influences the catalytic
activity causing its increase or decrease. This effect is referred to as size effect [2, 10, 11].
The investigations of size effect is an important goal because the minimization of
catalyst size and its dispersion on the substrate to form very high surface area de-
crease the cost of manufactured batteries, fuel cells, or other reactors where cata-
lyst are used. The investigation results are often unexpected. For instance: (i) Pt
particles larger than 5 μm cause the oxygen reduction by 4e direct pathway to
H 2 O, while for particles lower than 5 μm, the dominant product of reduction is
H2O2 (see Section 4.5.1); (ii) CO is oxidized to CO2 at polycrystalline Au electrode
with high overpotential of 0.7 V as a result of AuO formation, 3 nm Au particles
are much more active, and oxidation of CO takes place at lower overpotential
0.3 V; (iii) at platinized platinum, the rate of H2 evolution is about order magnitude
lower than at Pt electrode; however, there are some examples in literature point-
ing out to the opposite effect. The problems with the comparison of catalytic activ-
ity of particles of various sizes are caused by difficulties in the preparation of
uniform particles and by the interaction between particles and support.
The mechanisms called bifunctional catalysis and other called “third body effect”
are connected with size impact on catalysis [3, 10]. The oxidation of methanol is an
example of a bifunctional mechanism (see the beginning of this section). This mecha-
nism corresponds to the reactions where the rate of reaction is related to the two-
component of catalyst (e.g., Pt–Ru) selectively chemisorbing the species taking part in
the slowest step of reactions.
Up till now, the “third body effect” is not well understood. As an example, the oxi-
dation of FA at Pt modified by UPD (Chapter 6) metals of groups III, IV, or VA can be
considered.
Let us assume that the processes involving the organic substances occur in two
parallel ways (e.g., dehydrogenation and dehydration of FA):
dehydrogenation
HCOOH ! CO2 + 2 H + + 2 e

dehydration
HCOOH ! COads + H2 O

The adsorbed COads inhibited the catalytic reaction on Pt. To improve the rate of reac-
tion, the foreign adatoms were deposited, called third body, which selectively sup-
pressed CO adsorption by blocking neighboring adsorption sites [12].
Not only the size of crystallites forming catalyst surface or particles but also crys-
tallographic surface structure influenced the catalytic activity. Many results on the in-
fluence of face structure on catalytic activity are contradictory. It was found that the
order of decreasing catalytic activity for H2 evolution at different faces of gold and
wolfram tungsten is as follows: (110) > (100) > (111) and (111) > (110) > (100), respec-
tively. The influence of crystallographic face depends also on the solution. It was re-
 4.5 Electrocatalysts in hydrogen–oxygen fuel cells 131

ported that for Pt single crystals, the catalytic activity against ORR changes in order
(110) > (111) > (100) in HClO4 (110) > (100) > (111) in H2SO4 [2].
The investigations of disperse catalyst are of special interest [2]. They contain
the crystallites and crystalline aggregates of different shapes and sizes. Disperse catalyst
can be obtained on foreign supports, for example, various C, oxides, zeolites, conducting
polymers, and others. The catalytic particles can be also deposited on a substrate con-
sisting of the same metal, for example, platinized platinum or on foreign metals like
UPD metal adatoms. The important parameter for dispersed catalyst is specific surface
area (true working surface area), which can be determined by measurements of low-
temperature argon adsorption (by BET method). Experimentally, it was found that, in
some cases, the specific catalytic activity (referred to the true working surface) of electro-
des with metal disperse catalyst is lower than the catalytic activity of the same electrode
without the disperse catalyst of the same metal. The reasons for such facts are the porous
structure of disperse catalyst resulting in the increase of concentration overpotential,
loses connected with boundaries of particles grains, and with the boundaries of particles
and substrate. However, as the total surface area of dispersed catalysts increases very
much, the overall rate of reaction is larger in such systems but not so much as expected
in comparison with compact electrodes.
Nowadays, there is a new field, which is developing very widely in chemical catal-
ysis and electrocatalysis. It is the so-called single-atom catalysis. Single atom catalyst
contains isolated metal atoms singly dispersed on a support. This topic is a subject of
reviews [13].

4.5 Electrocatalysts in hydrogen–oxygen fuel cells

From the mid-twentieth century, the huge development of investigations of new electro-
chemical sources of energy is observed. Among these sources, hydrogen–oxygen fuel
cells play the most important role. There, the electrical energy is produced as the result
of a conversion of chemical energy of the oxidation of H2 (HOR) at the anode and reduc-
tion of O2 (ORR) at the cathode. A very important reaction from the technological point of
view is the electrochemical production of pure H2 (HER), the fuel used in the hydrogen–
oxygen cell. The reader can find much more about the electrochemical energy converters
in Chapter 8. Below we will mostly concentrate on the ORR and electrocatalysts improv-
ing its efficiency. Also, we will very briefly describe the reaction of oxidation (ionization)
and the evolution of hydrogen.
132 4 Electrocatalysis

4.5.1 Oxygen reduction reaction

In acidic aqueous solution oxygen reduction occurs by two overall pathways: a direct
four- electron reduction and a peroxide two-electron, which involve H2O2 intermedi-
ate formation:

O2 + 4 H + + 4 e ! 2 H2 O, E0 = 1.299 V vs SHE

O2 + 2 H + + 2 e ! H2 O2 , E0 = 0.69 V vs SHE
Peroxide can further undergo reduction or decomposition

H2 O2 + 2 H + + 2 e ! 2 H2 O, E0 = 1.776 V vs SHE

2 H2 O2 ! 2 H2 O + O2

In alkaline solution, four- or two-electron reduction of O2 occurs with ions formation

O2 + 2 H2 O + 4 e ! 4 OH − , E0 = 0.401 V vs SHE

O2 + H2 O + 2 e ! HO2 − + OH − , E0 = − 0.065 V vs SHE

The HO2– ion is unstable and can undergo further reactions

HO2 − + H2 O + 2 e ! 3 OH −

2 HO2 − ! 2 OH − + O2

The standard potential E0 values of these reactions were obtained from thermody-
namic data. For the first reaction, the equilibrium potential is not established be-
cause of the very small value of exchange current density (10−8–10−7 mA/cm2). The
measured open-circuit potential (OCP) in the presence of oxygen at the Pt elec-
trode is about 1 V versus NHE (normal hydrogen electrode), which is 0.2–0.3 V
more negative than the equilibrium potential. This difference and a very small
value of exchange current density pointed out that the reaction of O2 direct reduc-
tion is slow even on Pt electrode, catalyst of the reaction. One of the reasons is the
stability of O − O bond in O2 molecule, as the dissociation energy of this molecule is
494 kJ/mol. The mechanisms of oxygen reduction on Pt were intensively investi-
gated. They are described in [14] and references there. According to the experi-
mental and theoretical studies, the OR reaction on Pt is usually explained by two
mechanisms: dissociative and associative [15]. They consist of chemical and elec-
tron transfer steps as shown below.
 4.5 Electrocatalysts in hydrogen–oxygen fuel cells 133

Dissociative mechanism:

2 Pt + O2 ! 2PtO

2 PtO + 2 H + + 2 e ! 2PtOH

2 PtOH + 2 H + + 2 e ! 2Pt + 2 H2 O

Associative mechanism:

Pt + O2 ! PtO2

PtO2 + H + + e ! Pt − O2 H


H2 O + PtO
Pt − O2 H + H + + e !
Pt + H2 O2

Pt − O + H + + e ! Pt − OH

Pt − OH + H + + e ! Pt + H2 O

If the OR reaction proceeds according to the dissociative mechanism, the formation of

H2O2 is not possible, because the O–O bonds of adsorbed molecules are broken. How-
ever, if the oxygen bonds are not broken, H2O2 can be produced undergoing further
decomposition to water (associative mechanism). The decomposition reaction is very
quick so the formation of H2O2 is not observed experimentally at Pt electrode. The
first step in ORR is the adsorption of O2 on the electrode surface. Three models for
oxygen adsorption on metal surface are considered: (i) O2 interacts by two bonds with
the single substrate atoms (Griffiths model), (ii) O2 interacts by one bond with a single
substrate atom (Pauling model), (iii) O2 interacts by two bonds with two substrate
atoms (Yeager bridge model, Fig. 4.4).

Me Me Me Me

Fig. 4.4: Models for O2 adsorbed on metal active sites. Adapted from [14].

The last model was proposed for the reaction on Pt metals. It appears that dual ad-
sorption sites more likely are involved in the dissociation of O2, while the single ad-
134 4 Electrocatalysis

sorption sites are preferable in hydrogen peroxide formation. The direct four-electron
pathway is characteristic for Pt and other Pt group metals. The peroxide pathway was
found for mercury and graphite electrodes that are inactive in the catalytic decompo-
sition of H2O2. At other metal electrodes, the ORRs occur in parallel following two
pathways, one of which dominates based on the experimental conditions such as sur-
face treatment and electrolyte composition.
There are plenty of experimental results obtained for different electrodes mostly in
acidic media but also in alkaline solution. They are described and discussed by Adžić
[14]. As was mentioned, the ORR is slow and requires suitable catalysts eliminating the
peroxide pathway, having an impact on catalytic efficiency. Catalysts should also have
a large surface area and be stable chemically and electrochemically. Pt is the only com-
mercially available catalyst with sensible activity and stability. However, the cost of this
catalyst is high and huge effort is made to replace Pt by other materials or minimalize
the amount of it by applying nanoparticles. Except Pt other metals from group VIII B
such as Pd, Ru, Ir, and Rh have been tested as cathode materials. These metals in terms
of activity in ORR follow the trend Pt > Pd > Ir > Rh > Ru. Pt alloys with transition metals
such as Fe, Ni, Co, and Cr have been extensively studied on different carbon substrate.
The amount of catalyst was minimalized by applying Pt and Pt alloy nanoparticles. Al-
though bimetallic catalysts (Pt–Fe, Ni, Co, etc.) show improvement in catalytic activity
in comparison with Pt, there is unfortunately a drawback, since these metals are solu-
ble in acidic solution in the potential range from 0.3 to 1.0 V versus NHE. To overcome
this problem, some multicomponent catalysts were synthesized on the Pt basis such as
PtTiCo, PtTiFe, PtCuCo, and PtCuCoNi [15]. Metal oxides such as NiO, ZrO2, TiO2, and
SnO2 were tested as a support for metal particles and as catalysts in both acidic and
alkaline solutions. Many metal oxides are unstable in acidic media; therefore, the layers
of conducting polymer–polypyrrole were used for protection. Apart from metal and
metal alloy nanoparticles, the nanostructured carbon materials were applied as support
and catalyst as well. CNFs, single-walled CNTs or multiwalled CNTs have been studied
extensively as a support of metal nanoparticles, due to their very high surface area and
high electric conductivity. They were used for Pt or Pd nanoparticle dispersions and
shown enhanced catalytic activity as compared with commercial Pt/C. The application
of graphene and its derivatives in ORR catalysis is described earlier. The noble metals
can be replaced as catalysts also by transition metal macrocyclic complexes. The role of
ligand type, ligand structure, and type of central metal cation was determined in the
catalytic activity of O2 reduction. For instance, the catalytic activity toward oxygen re-
duction decreases in this sequence [14]: Co phthalocyanine (Pc) > Co tetraphenyl por-
phyrin > Co tetramethyl porphyrin. The four-electron reduction of O2 has been found
for Fe–Pc and Mn–Pc, while Co–Pc two-electron reduction was observed.
 4.5 Electrocatalysts in hydrogen–oxygen fuel cells 135

4.5.2 Hydrogen evolution and oxidation reactions

The cathodic evolution reaction of H2 (HER)

2 H 3 O + + 2 e ! H2 + 2 H 2 O

is the most common and popular electrochemical reaction. It takes place in water
electrolysis in the production of chlorine. It is important in the corrosion of metals
(Chapter 3) and in metallurgy. In the reverse reaction, the anodic oxidation of molecu-
lar H2 (HOR) is utilized in hydrogen–oxygen fuel cells.
Let us consider both reactions in equilibrium: 2 H3 O + + 2 e $ H2 + 2 H2 O. The
aH +
equilibrium potential Eeq = E0 + RTF ln pH is established in acidic solution, if platinized
2
Pt is used as an electrode. For Pt-platinized electrode, the standard potential E0 is ac-
cepted to be 0 in all temperatures under conditions: p = 1,013 hPa and acid solution con-
centration = 1 N. The Pt|H2, H+ electrode is commonly used as the reference electrode
called NHE and all the potentials may be referred to it. Note that in literature you can
find another hydrogen electrode applied as reference: standard hydrogen electrode (SHE)
and reversible hydrogen electrode. The standard potential of SHE is zero, but in this case,
Pt is dipped in an ideal 1 M acid solution (aH + = 1). The electrochemical potential of elec-
trons for SHE is − 4.5 ± 0.1eV. We will use this standard electrode (SHE) in Chapters 10
and 11 for the correlation of energy levels of semiconductors and electrolytes containing
redox systems. The reversible hydrogen electrode does not require the proton concentra-
tion to be unity ERHE = − 0.059pH ðvs SHEÞ, at 298 K.
For the equilibrium reaction, the exchange current density is high, which is equal
to 10−2–10−4 A/cm2 and depends on the solution composition and its purity. As Pt and
metals of its group are sensitive to impurities, the OCP often differs from equilibrium
potential. There are some metals such as Hg, Pb, Bi, Cd, and In, known as “low ex-
change current density metal,” at them the equilibrium potential is not established. Ca-
thodic hydrogen evolution is a two-electron reaction occurring through the intermediate
steps. Some of them are referred to after the name of the scientist who suggested the type
of rate-determining step (rds) in the overall reaction. The steps that govern the reaction:
2 H3 O + + 2e ! H2 + 2 H2 O are:
(1) the transport of the hydronium ion to the electrode,
(2) the discharge of the ion by reaction:

H3 O + + e $ Hads + H2 O ð2aÞ ðVolmer reaction, discharge stepÞ,

or by the Heyrovsky reaction (electrochemical desorption step)

Hads + H3 O + + e ! H2 + H2 O ð2bÞ,

(3) the recombination of adsorbed hydrogen atoms via the Tafel reaction

2 Hads $ H2 ð3Þ ðTafel stepÞ.

136 4 Electrocatalysis

Further steps are (4) desorption of hydrogen molecules from the electrode surface to
the solution through the electrical double layer and, finally, (5) the transport of H2
from the electrode via diffusion or gas evolution. The main step in anodic oxidation

H2 + 2 H2 O ! 2 H3 O + + 2 e

is adsorption of H2 molecule on the electrode surface followed by dissociation and ad-

sorption of hydrogen atoms (Tafel reaction). The other important step is the ionization
of adsorbed H atoms and (H3O+) formation (Volmer reaction). The rates of individual
steps 2a, 2b, and 3 are dependent on the bond energy of the hydrogen adsorbed on
the metal surface. If this energy is low or intermediate, the reaction of the discharge
of hydronium ion (Volmer reaction) followed by electrochemical desorption (Heyrov-
sky reaction) is the rds. Such a route was found for metals such as Hg, Cd, In, Pb, Ag,
and Au. For metals such as W, Mo, and Nb, the discharge of hydronium ions is quick,
and the rds is electrochemical desorption (Heyrovsky reaction). For Pt and metals of
the Pt group, the rds is recombination resulting from the large value of Me–H energy
bonds. In hydrogen oxidation, the slowest step is the dissociation of H2 molecules ad-
sorbed on the electrode surface. The common catalysts for HOR is Pt. Unfortunately, Pt
is very sensitive for impurities in the solution and in H2 gas. Commercially applicable
H2 is mainly produced by steam reforming of hydrocarbons and may contain CO. Car-
bon monoxide adsorbs very strongly on Pt poisoning the catalyst surface. To overcome
this problem, Pt alloys (Me–Ru, Rh, Ni, or others) were tested [3]. It was found that
the second compound decreases CO adsorption or catalysis the oxidation of adsorbed
CO. The best catalyst used in HOR is Pt–Ru (Section 4.3). Much more about hydrogen
evolution, processes, and catalysts you can find in review [16] and in references there.

Bibliography

[1] Bocris JO`M, Reddy AKN. Modern electrochemistry. V2 New York, USA, Plenum Press, 1970.
[2] Bagotsky VS. Fundamentals of electrochemistry. Hoboken, USA, J Willey & Sons Inc., 2006.
[3] Ross Jr. PN, The science of electrocatalysis on bimetallic surfaces. In Lipkowski J, Ross PN ed.
Electrocatalysis, New York, USA, Wiley-VCH, 1998, 43–74.
[4] Hwang J, Rao RR, Giordano L, Katayama Y, Yu Y, Shao-Horn Y. Perovskites in catalysis and
electrocatalysis. Science 2017, 358, 751–756.
[5] Risch M. Perovskite electrocatalysts for the oxygen reduction in alkaline media. Catalysts 2017, 7,
154–185.
[6] Zhou X, Qino J, Yang L, Zhang J. A review of graphene-based nanostructural materials for both
catalyst support and metal-free catalyst in PEM fuel cell oxygen reduction reaction. Adv. Ener.
Mater. 2014, 1301523, 1–25.
[7] Gileadi E. Physical electrochemistry. Fundamentals, techniques and applications. Weinheim, FRG,
Willey-VCH Verlag GmbH&Co, 2011.
[8] Quaino P, Juarez F, Santos E, Schmicler W. Volcano plots in hydrogen electrocatalysis – uses and
abuses. Beilstein J. Nanotechnol. 2014, 5, 846–854.
 Bibliography 137

[9] Bockris JO`M`, Reedy AKN, Gamboa-Aldeco M. Modern electrochemistry 2A. Fundamentals of
electrodics. Second Edition. New York, USA, Kluwer Academic. Plenum Publisher, 2000.
[10] Petrii OA, Tsirlina GA. Size effects in electrochemistry. Russ. Chem. Rev. 2001, 70,
285–298.
[11] Wieckowski A, Neurock M. Contrast and synergy between electrocatalysis and heterogeneous
catalysis. Adv. Phys. Chem. 2011, ID 907129 18p.
[12] Leliva E, Iwasita T, Herrero E, Feliu JM. Effect of adatoms in the electrocatalysis of HCOOH oxidation.
A theoretical model. Langmuir 1997, 13(23), 6287–6293.
[13] Su J, Ge R, Dong Y, Hao F, Chen L. Recent progress in single-atom electrocatalysts: concept,
synthesis, and applications in clean energy conversion. J. Mater. Chem. A 2018, 6, 14025–14043.
[14] Adzic R. Recent advances in the kinetics of oxygen reduction. In Lipkowski J, Ross PN ed.
Electrocatalysis. New York, USA, Wiley-VCH, 1998, 197–242.
[15] Khosteng L. Oxygen reduction reaction. In Ray A, Mukhopadhyay I, Pati RH ed. Electrocatalysis for
fuel cells and hydrogen evolution. Theory to design. London, United Kingdom, Intech Open, 2018,
25–50.
[16] Dubouis N, Grimaud A. The hydrogen evolution reaction: from material to interfacial description.
Chem. Sci. 2019, 10, 9165–9181.
5 Electrodeposition

5.1 General remarks

Plenty of electrolytic processes are applied in industry for manufacturing many of re-
agents and materials such as chlorine, sodium hydroxide (electrolysis of NaCl), hydro-
gen (electrolysis of water), and metals by electroextraction from the proper salt (e.g.,
Zn, Cd) or by electrorefining [1, 2]. Electrorefining means anodic dissolution of contami-
nated metals and further cathodic deposition of purified metals. By using electrolytic
deposition (shortly named electrodeposition), we can obtain not only metals but also
alloys, semiconductors, and conducting polymers. All these materials can be deposited
on the surface of conducting substrate in the form of layers, thin films, or nano-objects.
Electrodeposited thin layer of metal (electroplating) are widely used as protective or
decorating coatings. Thin layers of semiconductors and conducting polymers found ap-
plication in the construction of solar cells, alloys can be used as catalyzers in fuel cells,
and all these materials are recently applied in microelectronic and nanotechnology.
Comparing the electrodeposition with vapor deposition one clearly sees some advan-
tages of previous one, such as (i) the equipment for electrodeposition is more simple
and not so expensive, (ii) the electrodeposition can be easily controlled by electrode po-
tential or current density, and (iii) the electrodeposition occurs at low temperature.
In practice (industry), the electrodeposition is carried out in the two-electrode
cell. The material is deposited on the cathode during the reduction of ions in the case
of metals, alloys, or on the anode during oxidation of the electrode or oxidation of
monomers resulting in the formation of oxides, conducting polymers. The deposition
can be carried out at constant potential or constant current from an electrolytic bath.
Bath contains solvent with appropriate species (like metal ions, metal complexes, and
organic monomers) and some salt to increase the conductivity of the solution. Addi-
tionally, buffers for pH stabilization and some additives for improvement of deposi-
tion kinetics and morphology of deposits can be used.
Generally, the electrolysis and thus also electrodeposition processes are governed
by two Faraday’s laws. They described (i) the relation between the amount of depos-
ited mass m and the electrical charge Q passing during deposition reaction (first law),
and (ii) the relation between the mass of deposited material and its equivalent weight
(second law). The equivalent weight of a substance is its molar mass divided by elec-
trons transferred per ion in the reaction ðM =nFÞ.
Faraday’s laws may be summarized in the equation:

M
m= Q (5:1)
nF

https://doi.org/10.1515/9783111160986-005
 5.2 Electrocrystallization: nucleation and growth 139

If the current density j changed with time, the charge is described by the integral
Ðt
Q = 0 ηF jAdt; if the current density, surface area A, and Faradaic efficiency ηF is
not time t dependent, then

Q = ηF jAt = ηF it (5:2)

The thickness of deposited layer d can be estimated by using the equation:

m ηF Mit
d= , d= (5:3)
ρA ρAnF

ρ is the density of deposit and i is current.

For a proper determination of the thickness of the deposit, it is necessary to take
into account the reactions, which occur simultaneously with the main reaction of de-
position (side reactions). In such a case, the Faradaic efficiency of reaction is lower
than 1.
The investigations of both deposition reactions and mechanisms of deposit forma-
tions are carried out in the three-electrode cell, where the working electrode can be
cathode or anode depending on the measured system. The electrochemical transient
techniques are used for this purpose, mainly step potential method (Section 2.1).

5.2 Electrocrystallization: nucleation and growth

During the electrodeposition, the formation of a new solid phase and crystal growth
take place at electrode–liquid interface. In literature, the term electrocrystallization is
often used for steps involved in these phenomena [3–5]. The considerations below deal
with metal deposition; however, some of them are so general that they may be used for
other materials such as alloys, semiconductors, and even conducting polymers. Sche-
matically some electrocrystallization steps are shown in Fig. 5.1.
Let us consider the solvated (hydrated) metal ion (Men+) which will be reduced at
the electrode surface. In the first step, the solvated ion should get close to the elec-
trode to permit an electron transfer from the electrode. Then the intermediate species
are formed, which have a partial charge and some solvent molecules, so they can be
considered as a kind of adsorbed ions, the ad-ions. These ad-ions have higher energy
than atoms incorporated into metal lattice; therefore, they cannot remain on the sur-
face too long. They move from the surface site, where charge transfer took place and
where they were created, to the surface defects with a lower energy of incorporation
of new atoms. Ad-ions mostly integrate with the crystal lattice at kink sites and step
sites, the process is finished when the edge of the surface is reached. This step-growth
mechanism is schematically shown in Fig. 5.2.
The other mechanism is associated with screw dislocations and leads to spiral growth
of crystal. The screw dislocations result from a shift of one atom with respect to a perfect
140 5 Electrodeposition

Electrode Electrolyte

e Diffusion solvated
Charge
Mn+ ion
transfer

Ad-ion

Nucleation

Cluster

Hemisphere

Growth
Cone

Dendrite

Fig. 5.1: Simplified scheme showing some steps involved in the electrocrystallization of metal on foreign
substrate. Adapted from [3].

arrangement of others. This shift causes the formation of the edge at which new atoms
(ad-ions and ad-atoms) can be stacked [2].
The ad-ions have to move from the place of their formation to incorporation sites.
The concentration gradient of ad-ions between the place of their formation and elimi-
nation at defects is a driving force for the ad-ions motion. Such type of transport is
named surface diffusion. The electrodeposition of metal or semiconductors mostly oc-
curs at the potential more negative than the equilibrium potential (Nernst equation) of
the investigated system. The difference between these two potential ðη − overpotentialÞ
is the driving force for the electrocrystallization.
If the electrode surface is planar with a small amount of defects or deposition is
carried out on the substrate foreign to deposited metal, the ad-ions accumulate to
form a stable nucleus (crystal). The nucleus is stable and continues to grow farther if
it reaches a critical size.
 5.2 Electrocrystallization: nucleation and growth 141

Ion in Ion in
solution solution

Direct Transfer
transport to ad-ion Ad-ion

m
k ato
Ad-ion Kin
Surface diffusion

Fig. 5.2: Scheme illustrating the consecutive stages involved in incorporation of ad-ion and lattice
building. Adapted from [6].

The process of nuclei formation is named nucleation and it is the beginning of the
formation of the condensed phase. From nuclei further crystal growth proceeds, the
relation of the rates of both processes influence the morphology of the deposit. The
nuclei can be two or three dimensional (D) and their dimensionality depends on the
interaction between substrate and deposited species, as well as on the relation be-
tween crystal lattice of a substrate and that of deposit. The nucleus is stable and can
further grow only when it reaches a critical size.

5.2.1 Critical size of nuclei

In the classical approach, we assume that electrocrystallization has many features in

common with the growth of the solid phase from its supersaturated vapor. Based on
the thermodynamic description of homogeneous nucleation from the vapor [7], by
analogy we can obtain the expression for the Gibbs free energy of electrochemical
condensed phase formation:
nFηρ
ΔG = (5:4a)
M

M is molecular weight, ρ is the density of deposit, and ƞ is overpotential.

The total Gibbs free energy of the formation of a disk shape nucleus (2D) with
radius ðrÞ and height ðhÞ is the sum of two Gibbs free energies of formation, the bulk
ΔGbulk and the surface ΔGsurf :
142 5 Electrodeposition

πr2
ΔGT = ΔGbulk + ΔGsurf = hρnFη + 2πrhγ (5:4b)
M
γ is molar surface free energy (interfacial tension, Chapter 1).
For spherical nucleus (3D), this equation is written as follows:

4πr3
ΔGT = ΔGbulk + ΔGsurf = ρnFη + 4πr2 γ (5:4c)
3M
Differentiations of these equations with respect to r show that free energy of nucleus
formation ΔGT passes through a maximum at the critical radius rc . The nuclei with
radii < rc are therefore unstable, whereas those with radii > rc are stable and can
grow further. The free energy in maximum is called critical Gibbs free energy ΔGc of
nucleus formation and has the meaning of activation energy of homogeneous nucle-
ation. In both cases, 2D and 3D nuclei, the critical free energy and critical radius depend
on overvoltage:
γM
2D nuclei rc = − (5:5a)
nFρη

πhMγ2
ΔGc = − (5:5b)
nFρη
2Mγ
3D nuclei rc = − (5:6a)
nFη

16πM 2 γ3
ΔGc = (5:6b)
3πρ2 n2 F 2 η2

note that ƞ < 0.

The above expressions pointed out that an increase of overvoltage causes a de-
crease in the critical radius. It means that at the higher overpotentials we should ex-
pect a growth of small nuclei forming a fine-grained polycrystalline deposit.

5.2.2 Instantaneous and progressive nucleation

The formation of critical nuclei does not depend only on overpotential but also on
thermal fluctuations, so the rate constant of nucleation ðBÞ is related to the activation
energy of nucleation ΔGc :


ΔGc
B = F exp − (5:7)
kT

F is a preexponential factor.
In another approach, the formation of clusters from ad-atom or ad-ions is consid-
ered as a growth center, and ΔGc, n is Gibbs critical energy of clusters formation of
 5.2 Electrocrystallization: nucleation and growth 143

n-atoms. In contrary to vapor deposition, where the homogeneous nucleation oc-

curs, the nucleation is a heterogeneous process, since the nuclei form on the active
sites of the electrode surface.
In the simple approach to the kinetics of nucleation, it is accepted that the process
is the first-order reaction [3, 7]. As nucleation proceeds, the number of nuclei changes
with time:

NðtÞ = N0 ½1 − expð − BtÞ (5:8)

N0 is the initial number of active sites (kinks, steps, etc.), and B is the nucleation rate
constant.
There are two limits to this equation:
i) If Bt  1, then expð − B tÞ ! 0 and NðtÞ = N0 ðtÞ. This equation points out that the
number of nucleation sites remains constant at its initial value. This type of nucle-
ation is referred to as instantaneous nucleation.
ii) if Bt is very small, expð − B tÞ ≈ 1 − Bt; therefore, NðtÞ = N0 Bt; we can write further
N(t) = B’t, where B’ is the modified nucleation rate constant. In this case, the num-
ber of nuclei increases in time. This type of nucleation is referred to as progres-
sive nucleation.

The processes of nucleation and growth of 2D and 3D deposits were widely investi-
gated by using potential step experiments. The experimental current–time curves
show characteristic features like the current density maxima. Their time position tm
depends on the potential (overvoltage) (Fig. 5.3).

–0.90 V
–0.4 –0.88 V
–0.86 V
–0.3 –0.84 V
–0.82 V
I (mA)

–0.81 V
–0.2
–0.80 V
–0.70 V
–0.1

0.0
0.1 0.2 0.3 0.4
t (s)

Fig. 5.3: Deposition of Cd on glassy carbon electrode at different potentials from 0.01 M
Cd(ClO4)2 + 0.1 M LiClO4 solution. Transients show maxima at different potential values. Unpublished
data: M. Osial, K. Jackowska (Faculty of Chemistry, Warsaw University).
144 5 Electrodeposition

The quantitative analysis of current transient curves can provide information con-
cerning the mechanism and kinetics of both nucleation and growth processes.

5.2.3 Analytical approach to experimental data

There are plenty of books and review articles dealing with the theoretical and practi-
cal aspects of electrodeposition. Interested readers will find some of them in [3–9].
Such studies deepened our understanding of electrocrystallization processes and
their influence on the morphology of deposits.
All theoretical equations given below regarding i − t curves are based on the as-
sumption that the nucleation and growth are the slowest steps of deposit formation. It
is convenient to consider two limiting cases of growth of 2D (disks and cylinders) and
3D (hemispheres). Let us concentrate on the models describing 2D (cylinders) growth
and formation of a monolayer. It is assumed that the incorporation of ad-atoms (ad-
ions) occurs at the edges of growing nuclei. The current i flowing to the single nucleus
before their overlapping into a monolayer is described:

i = nFkA (5:9a)

k is the rate constant of incorporation (or growth parallel to the plane of substrate), A
is a surface area, A = 2πrh for cylinders, so the current and the charge for the forma-
tion of single cylinder can be written as

i = 2πrhnFk (5:9b)

πr2 hnFρ
Q= (5:9c)
M

The derivative of this equation leads to expression:

2πhnFρrðtÞ drðtÞ
iðr, tÞ = (5:9d)
M dt

As drðtÞ
dt = ρ , the integration gives us rðtÞ = ρ t.
Mk Mk

Substituting rðtÞ into eq. (5.9d) and assuming the instantaneous nucleation of nu-
clei N ðtÞ = N0 we obtain the equation for current:

2N0 πhMnFk2
i= t, 2D instantaneous (5:10)
ρ

If the nucleation of nuclei (growth center) is progressive, then NðtÞ = N0 Bt, and the
equation is given as follows:
 5.2 Electrocrystallization: nucleation and growth 145

2BN0 πhMnFk2 2
i= t, 2D progressive (5:11)
ρ
Both equations predict that current increases for all times, which is unacceptable and
does not agree with experimental curves recorded in the potential step experiments.
These equations were obtained under the assumption that nuclei grow independently.
It may be true in a very early stage of nucleation when a low number of nuclei occur.
When the number of nuclei increases and growth proceeds, neighboring nuclei will
come in contact and overlap. The overlap problem has been considered theoretically
by applying the concept of “extended area” Aex . The extended area means the total
area of nuclei without an overlap. Figure 5.4 illustrates how the overlap of nuclei
changes the real surface area A and how the extended area may be presented.

(a) (b)

(c)
Aext

Fig. 5.4: Scheme presenting overlapping of growth

centers and influence on surface area.

The relation between A and Aex is given by Avrami theorem:

A = 1 − expð − Aex Þ, when Aex is very large A ! 1; and all surface is fully covered
with 2D deposit.
The current–time equations for the overlap case [3] are as follows:


2N0 πhMnFk2 t πN0 M 2 k2 t2
i= exp − , 2D instantaneous, overlap (5:12)
ρ ρ2


2BN0 πhMnFk2 t2 πM 2 BN0 k2 t3
i= exp − , 2D progressive, overlap (5:13)
ρ 3ρ2

Considering the mathematical form of these equations it is clearly seen that a maxi-
mum should be observed on i − t curves and that the current should go to zero. The
latest is understandable as during growth all active sites (edges) are filled and the
growth stops once the layer (monolayer) is completed. The current–time curves illus-
trating such features are shown in Fig. 5.5(a).
146 5 Electrodeposition

(a) (b)
I 1 j/jmax
Progressive (2)
2
Instantaneous (1)

t
t/tmax
(c)

1.0

0.8
Instantaneous
(I/Imax)2

0.6
–0.79 V
0.4

0.2 Progressive
–0.81 V
0.0
1.0 2.0 3.0 4.0
t/tmax

Fig. 5.5: (a) Plot illustrating the changes in the shape of transient curves. Curve 1 presents 2D
instantaneous growth with overlap (eq. (5.12)) and curve 2 presents 2D progressive growth with overlap
(eq. (5.13)). (b) Dimensionless plots illustrating differences between 2D instantaneous and 2D progressive
growth (eqs. (5.14) and (5.15)). (c) Dimensionless plots presenting a comparison between calculated and
experimental data. Experimental data were obtained for Cd deposition on a graphite paste electrode.
Unpublished data: A. Frydrychewicz, M. Osial, K. Jackowska (Faculty of Chemistry, Warsaw University),
G. A. Tsirlina, (Faculty of Chemistry, M.V. Lomonosov Moscow State University).

It is convenient to compare the experimental curves with calculated ones using nondi-
mensional quantities imaxi t
and tmax , and imax , tmax mean the values in maximum. The
expressions in terms of nondimensional parameters are:


i t ðt2 − tmax
2
Þ
= exp − 2
, 2D instantaneous (5:14)
imax tmax 2tmax


i t2 2ðt3 − tmax
3
Þ
= 3
exp − 3
, 2D progressive (5:15)
imax tmax 3tmax

In Fig. 5.5(b), dimensionless plots according to the above equations are presented for
monolayer deposit (2D).
The comparison of calculated and experimental curves gives us the possibility to
distinguish between these two mechanisms (Fig. 5.5(c)). It should be noted that for
 5.2 Electrocrystallization: nucleation and growth 147

such comparison the theoretical current values have to be recalculated to current

density values.
Let us describe very briefly the 3D growth of nuclei in the form of cones [3,6]. It is
assumed that the nuclei may grow simultaneously in two directions parallel to the
plane (substrate) and perpendicular to plane. When at short times the number of nu-
clei is small and they do not overlap the equations describing current–time depen-
dence as follows:

k2 M 2 k12 2
i = nFπN0 t, 3D instantaneous (5:16)
ρ2

k2 M 2 k12 B 3
i = nFπ t, 3D progressive (5:17)
3ρ2

k1 and k2 are rate growth constants parallel to substrate (plane) and perpendicular,
respectively. The coalescence and overlapping of nuclei lead to the expressions:


πM 2 N0 k12 2
i = nFk2 1 − exp − t , 3D instantaneous, overlap (5:18)
ρ2



πM 2 B1 k12 3
i = nFk2 1 − exp − t , 3D progressive, overlap (5:19)
3ρ2

In both cases for a long time, the current reaches the value of nFk2, which means that
perpendicular growth is only permitted. All these theoretical equations were derived
under the assumption that nucleation and growth of nuclei are controlled by kinetics.
In many cases, however, diffusion may be the slowest step in the growth of nuclei.
Linear and hemispherical diffusion were considered. It is assumed that in most cases
the further growth of crystallites (3D established nuclei, beyond their initial formation)
is under diffusion control with hemispherical symmetry. The current–time equations
for instantaneous and progressive nucleation and independent growth of crystallites
are as follows:

nFπM 1=2 N0 ð2DcÞ3=2 1=2

i= t , 3D instantaneous (5:20)
ρ1=2

2nFπM 1=2 BN0 ð2DcÞ3=2 3=2

i= t , 3D progressive (5:21)
3ρ1=2

D is the diffusion coefficient of electroactive ions, c is its concentration, and other

symbols have the meanings as above. The slope of j (current density) versus t1=2 gives
the nuclei number density N0, if the Dc product is known.
Because during deposition the number of nuclei increases, their diffusion zone also
increases and finally overlaps. The effect of overlapping of diffusion zones and its influ-
148 5 Electrodeposition

ence on i − t relationship was considered by Scharifker et al. [10, 11]. Avrami theorem
was adopted in the form: θ = 1 − exp ð − θex Þ, where θex is a fraction of the area covered
by the diffusion zone without overlap. The equations derived by Scharifker are:

nFD1=2 c
i= ½1 − expð − N0 πKDtÞ, 3D instantaneous, overlap (5:22)
π1=2 t1=2

K = ð8πcM=ρÞ1=2


nFD1=2 c BN0 K ✶ Dt2
i = 1=2 1=2 1 − exp − , 3D, progressive, overlap (5:23)
π t 2

K ✶ = ð4=3Þð8πcM=ρÞ1=2

In both cases, the current passes a maximum and then approaches the limiting cur-
rent for the diffusion planar to the electrode surface.
The time and current in maximum may be extracted from these equations by set-
di
ting dt = 0.
3D instantaneous:

tmax = 1.26=N0 πKD, imax = 0.638nFDcK 1=2 N0 1=2

3D progressive:

tmax = ð4.67=BN0 πK ✶ DÞ1=2 , imax = 0.462nFD3=2 cðK ✶ Þ1=4 ðBN0 Þ1=4

The product i2max tmax does not depend on K, K ✶ , B, N0 ; therefore, it can be used as a
diagnostic criterion of the type of nucleation.
The behavior for 2D nuclei is qualitatively similar to that shown by 3D nuclei.
However, in 2D growth, metal deposits in the form of 2D monolayer. This type of nu-
cleation is expected when metal deposition takes place on foreign metal substrate, ex-
hibiting a strong metal–substrate interaction. It can lead to underpotential deposition
described in Chapter 6.
The above equations were applied mostly for the description of metal deposition;
however, rarely they were also used to determine the mechanism of nucleation and
growth of semiconductors and conducting polymers.

5.3 Deposit morphology

As pointed earlier, the crucial electrochemical parameter influenced nucleation and crys-
tal growth is overpotential. Of course, there are also other factors having an impact on
morphology such as differences in the reaction mechanism, side reactions, methods ap-
plied in electrodeposition (steady-state method and pulse method), and conditions of de-
position. Let us concentrate on the overpotential influence. At low overpotentials, the
 5.4 Practical aspects of electrodeposition 149

deposition takes place mostly at lower energy sites of surface (e.g., kinks). The energy for
nucleation is not sufficient, so the formation of new nuclei occurs very rarely. During
growth, the large crystals are formed on the surface and deposits are mostly in the form
of layers and blocks. As the overpotential increases, the concentration of ad-ions in-
creases too, so crystal growth is possible at less energetically favorable sites (e.g., steps).
Deposits are formed as the bunched layers and ridges. When the overpotential is high
enough and further increases, the dominant process is the nucleation leading to a large
number of crystal grains. Grains coalesce forming the deposit with some defects present.
The granularity of deposit (the number of crystals) is determined by the relation between
nucleation and growth rates. If the nucleation rate is low compared to the rate of crystal
growth, one will obtain the deposit with a small number of large grains. On the contrary,
high nucleation rates lead to a fine granular deposit. At higher overpotentials, the mass
transport of active species from the bulk of solution to the electrode surface becomes im-
portant. At mass transport control, deposits often grow in the form of powders and den-
drites. In Fig. 5.6, the relation between deposit morphology, electrochemical parameters
(j, E), and various regions of kinetic and mass transport control are shown [2, 9].

Men+ + ne Me 2 H+ + 2 e H2

j
Mass transport
Kinetic control +
mass transport

(diffusion)
(mix region)

constant
(charge transfer)
Kinetic control

Powders
Nodules, dendrites
Polycrystalline growth
Bunched layers, ridges
Spirals, layers, blocks (well formed)

Eeq –E

Fig. 5.6: Scheme presenting relations between deposit morphology, electrochemical parameters, and
various regions of kinetic and mass transport control. Adapted from [9].

5.4 Practical aspects of electrodeposition

The industry of electroplating was developed for decades. The plating of metals and
alloys plays an important role in the deposition of protective or decorative coatings
and deposition of circuits in microelectronic.
150 5 Electrodeposition

Above, we briefly described mechanisms of early stages of deposition and the influ-
ence of overpotential on morphology of metal deposits. But in practice, not only mor-
phology is important but also adhesion, mechanical properties, uniformity of coating,
and its thickness. Good quality deposits are formed at current densities below the mass-
transport limit, at a fraction of limiting current. Let us consider other factors that can
influence the properties of deposits, such as side reactions, the additives in the plating
bath, and distribution of current density [2, 12]. We will concentrate first on the side reac-
tion. This reaction does not contribute to the main deposition reaction and decreases the
Faradaic efficiency ηF . When the reaction of metal deposition is carried out from aque-
ous solution, the reduction of hydrogen ions is the side reaction, resulting in H2 evolution
at the cathode. Possibility of H+ ions reduction and H2 evolution during metal deposition
depends on the relation between standard potentials of H+/H2 (E0 = 0 V) and Me/Men+
couples. Metals such as Pd, Ag, and Cu with positive standard electrode potential (E0 =
+0.91 V, E0 = +0.8 V, E0 = +0.34 V vs SHE, respectively) can be deposited even from acidic
solution with 100% efficiency. Metals such as Cd, Fe, and Zn having the standard
electrode potential more negative than SHE (E0 = −0.4 V, E0 = −0.44 V, E0 = −0.76 V)
can be deposited also, but with much lower efficiency than 100%. There are some
metals (Al and Mg), which cannot be deposited from aqueous solution.
The hydrogen bubbles produced during metal deposition may cover parts of the
electrode surface leading to a nonuniform deposit. To improve and make the depos-
ited layer smooth or even bright and sheen, different compounds are used as an addi-
tion to the plating bath. Additives may influence the deposition process in a variety of
ways. They may form a complex with metal ions and change the deposition potential,
as well as reaction mechanism. The organic additives may adsorb strongly on the elec-
trode surface resulting in a decrease in the rate of metal deposition. They may alter
the growth of nuclei by blocking the growth sites, changing the concentration of nu-
clei, the rate of surface diffusion, and so on. According to their main function, the ad-
ditives are referred to as carriers, brighteners, and wetting agents. The carriers
influence the nucleation and growth of the forming phase, refining the deposit grains
and producing brighter coating. The main role of brighteners is to make the deposi-
tion preferential on defects of the surface, resulting in the smoothing of the surface.
The wetting agents decrease the surface tension, facilitating the detachment of gas
bubbles generated by hydrogen evolution. There are plenty of additives used in the
plating bath, such as coumarin, dextrin, saccharin, thiazole, uric acid, sodium sulfo-
nate, and benzaldehyde.
If there is no side reaction, and only metal deposition takes place, the variations
in the thickness of the deposit are connected with no uniformity of current distribu-
tion on/at an electrode during deposition [2, 12]. The uniformity of current distribu-
tion is described by the Wagner number Wa defined as
 5.4 Practical aspects of electrodeposition 151


δη=δj RF
Wa = = (5:24)
ρL Rs

ρ is the specific resistivity of the solution (Ωcm), L is a characteristic length, RF is Fara-

daic (kinetic) resistance, and Rs is a resistance of solution.
The choice of L is not obvious. It may be, for instance, the difference between two
electrodes: cathode (working electrode) and anode or counter electrode. The Faradaic
resistance (Chapter 1) at potential near to the equilibrium potential is
RT 1
RF = (5:25)
nF j0

The large values of Wagner number correspond to more uniform current distribution
and a more uniform thickness of deposit. Deposit uniformity increases with solution
conductivity increase and decreases with current density increase. Therefore, in prac-
tice, the plating is often carried out initially at a lower current density to obtain uni-
form current distribution at thin layer coating (to increase Wagner number) and after
that, the higher current density is used to achieve the thickness need. The concept con-
nected with Wagner number is the “throwing power” T of plating bath. It is a measure
of bath ability to produce coating of uniform thickness on samples having complicated
geometry. In the bath with a higher T, more uniform deposit can be obtained. To quan-
tify the throwing power of a bath, the Haring–Blum cell is often used. The cell is shown
in Fig. 5.7. The cell consists of an anode and two cathodes, one on either side of the
anode. Only the cathode surfaces facing anode are active. The cathodes are placed at
different distances from anode x1 and x2.

Cathode 1 Cathode 2
Anode

X1 X2

Fig. 5.7: Scheme of Haring–Blum cell.

152 5 Electrodeposition

Throwing power can be determined using the equation:

100ðK − DÞ
Tð%Þ = (5:26)
K +D−2
x w
where K = x1 is the ratio of cathode distances from anode, and D = w1 is the weight
2 2
ratio of deposits on the two cathodes.
The best uniformity of deposit is reached when the weight of the deposits on both
electrodes is the same ðD = 1Þ. It corresponds to 100% of T and is possible when Wa  1.

5.5 Electrodeposition of binary alloys and semiconductor

compounds
In this section, we will briefly describe some features of electrodepositions of binary al-
loys and semiconductor compounds [6, 12–17]. If the metal A is electrodeposited from a
solution containing An+ ions on the same substrate (A), the deposition is going at a poten-
tial slightly higher than the equilibrium potential of A/An+ couple, Eeq = E0 + RT
nF ln aAn +
(aA is assumed 1) .
On a foreign substrate, the deposition potential (onset potential) of A can be sig-
nificantly more negative than the redox potential of A/A n+ couple. It results from
higher overpotential ƞ needed for the formation (called sometimes nucleation over-
potential) of a new phase A. The situation is more complicated when equilibrium is
established between metallic ions An+, Bn+ in solution and binary A–B alloy. Then we
can write the equations describing the equilibrium potential of systems Aalloy/An+
and Balloy/Bn+. Since these two potentials are equal, we can derive the equilibrium
potential of alloy A–B:
RT a n+
Eeq ð AÞ = E0, A + ln A (5:27)
nF aA, alloy

RT a n+
Eeq ðBÞ = E0, B + ln B (5:28)
nF aB, alloy


RT xA lnaAn + + xB lnaBn +
Eeq ðA − BÞ = E +
0
(5:29a)
F nxA + nxB

ΔG0mix xA EA0 + xB EB0

E0 = − + (5:29b)
F xA + xB

where ΔG0mix is standard Gibbs free energy of the alloy mixing, and xA , xB are the
molar fractions of the components.
 5.5 Electrodeposition of binary alloys and semiconductor compounds 153

If alloy forms a stable solid solution, the ΔGmix is negative. It causes a positive
shift of the redox potentials of A and B. The phenomenon is referred to as underpo-
tential codeposition and was observed, for instance, during the Au–Cu alloy electrode-
position. The presence of two metal ions (or more) during the alloy formation affects
also the kinetics, causing the variations in the exchange currents and limiting diffu-
sion currents of both metals. These and the differences in the concentration of A, B
ions in the solution may result that one component is deposited under mass control
while the other is under activation control. In consequence, we may obtain a deposit
of no uniform thickness and no uniform composition. Let us consider two cases: the
first one in which the reduction reactions of An+ and Bn+ ions are under kinetic control
(under control of activation overpotential). By using Volmer–Butler equation we ob-
tain the expressions for current densities of individual reactions:


αA FðE − Eeq, A Þ
jA = j0, A exp − (5:30)
nRT


αB FðE − Eeq, B Þ
jB = j0, B exp − (5:31)
nRT

For simplicity, we assume further that number of electrons taking part in reactions
(nÞ = 1 and transfer coefficients ðαA , αB Þ are equal to each other αA = αB = α, then the
ratio of current densities is

 
jA j0, A αF Eeq, A − Eeq, B
= exp − (5:32)
jB j0, B RT

It is clearly seen that the amount of metals in alloy A–B will depend on their exchange
current density and difference of their equilibrium potentials. The higher the value of
j0,A and more positive Eeq, A in comparison with the same parameters of metal B, the
greater the amount of metal A in the alloy (Fig. 5.8).
In the second case, we assume that the reduction of An+ ions is under mass trans-
port control (i.e., diffusion), while the Bn+ ions reduced under kinetic control. The
equations describing the current density of both processes are as follows:

nFDA c0A
jA = jL, A = − (5:33)
σ

 
αB F E − Eeq, B
jB = j0, B exp − (5:34)
nRT
154 5 Electrodeposition

E
log j0,A
Eeq,A

log j0,B
Eeq,B

log jA

log jB

log j (deposition current density)

Fig. 5.8: Potential–current density plots presenting the influence of differences in the exchange current
density ðj0, A , j0, B Þ on alloy composition AB.

The ratio of current density is

 
jA jL, A αB F E − Eeq, B
= exp (5:35)
jB j0, B nRT

The expression points out that when the current density jB is higher than limiting cur-
rent density jL, A , the amount of metal B exceeds the amount of metal A in alloy even if
the equilibrium potential of B/Bn+ system is more negative than Eeq, A , Fig. 5.9.
Electrodeposition being widely applied for metals and metallic alloys, it is also
developed for semiconductor (SC) thin film formation. In this way, binary III–V, II–VI
(e.g., GaAs, CdS, ZnO), ternary (CuInSe2), and even quaternary (Cu2 ZnSnS4) compounds
are formed [15, 16]. The electrodeposition of semiconductor compounds exhibits some
specific features. The most of semiconductor compounds obtained electrochemically
are binary compounds containing at least one metallic component such as Cd, Cu, Zn,
In, and Ga and one nonmetallic component: S, Te, Se, As, P, and O. The equilibrium po-
tentials and deposition potentials of nonmetals are very different from those of metals.
For instance, the standard potentials of Cd/Cd2+ and Zn/Zn2+ are –0.4 V and –0.76 V ver-
sus NHE, respectively, while Eo of Te/Te4+ is +0.55 V versus NHE and Eo of Se/Se4+
is + 0.86 V versus NHE. Most of the semiconductor compounds have also the large val-
ues of negative Gibbs free energy of formation, for example, −154.7, −135.4, and –99.5 kJ/
mol for CdS, CdSe, and CdTe, respectively. Similarly as in the case of alloy, this results
in the shift of the cathodic deposition potentials of less noble component (Eeq potential
of less noble component is more negative than Eeq of noble component) to the more
positive value. The condition to be met to obtain the binary compounds of well-defined
 5.5 Electrodeposition of binary alloys and semiconductor compounds 155

Eeq,A log j0,A

Eeq,B

log j0,B

log jA log jA

log jB
log jL,A

log j (deposition current density)

Fig. 5.9: Potential–current density plots presenting the influence of relation between limiting current
density ðjL, A Þ and current density jB on alloy composition AB.

stoichiometry has been discussed by Krӧger [17]. This is behind the scope of this book.
However, it is worth to mention that two component classes have been distinguished:
class I with large difference in deposition potentials of individual components and class
II with a smaller difference. In the first case, the deposition potential is determined by
the less noble component over the whole range of metal–nonmetal composition. This is
the case of CdX (X = S, Te, Se). For class II, the deposition potential may be determined
by one component or the other, depending on the composition. Deposited SC films have
much higher resistivity than films of metals or alloys. Therefore, after the deposition of
the first layer, the distribution of charge and uniformity of current density may be
changed resulting in composition changes. It was observed for CdTe. During the electro-
deposition, the value of deposition potential and concentration of components are the
crucial parameters that determine the type of conductivity and bandgap energy [18].
The richness of CdTe in Te causes changes in the type of CdTe conductivity from n-type
to p-type. The same is observed in many cases for more complicated semiconductor ma-
terials I–III–VI2 (CuInGaSe2). The solubility of nonmetal salts used for the formation of
semiconductor compounds is very low, and complicated equilibrium may be estab-
lished in aqueous solutions, so very often the deposition kinetic is under mass-transfer
control.
156 5 Electrodeposition

Bibliography

[1] Bagotsky VS. Fundamentals of electrochemistry. Hoboken, USA, John Willey & Sons Inc, 2000.
[2] Fuller TF, Harb JN. Electrochemical Engineering. Hoboken, USA, John Willey & Sons Inc, 2018.
[3] Pletcher D, Greff R, Peat L, Peter LM, Robinson J. (Southampton Group). Electrocrystallization. In
Instrumental methods in electrochemistry. Amsterdam, The Netherlands, Elsevier, 2001, 286–316.
[4] Milchev A. Electrocrystallization: Fundamentals of nucleation and growth. Dordrecht,
The Netherlands, Kluwer Academic Publisher, 2002.
[5] Budevski E, Staikov G, Lorentz WJ. Electrochemical phase formation and growth:
an introduction to the initial stages of metal deposition. Weinheim, FRG, VCH, 1996.
[6] Bocris JO`M, Reddy AKN, Gamboa-Aldeco M. Modern electrochemistry. Fundamentals of electrodics.
V 2A, New York, USA, Kluwer Academic Publisher, 2000.
[7] Harrison JA, Thirsk HR. The fundamentals of metal deposition. In Electroanalytical chemistry: a
series of advance. Ed. Bard AJ. New York, USA, Marcel Dekker, 1971, 67–147.
[8] Milchev A. Electrocrystallization: nucleation and growth on nano-clusters on solid surfaces. Russ.
J. Electrochem. 2008, 44, 619–645.
[9] Walsh FC, Herron ME. Electrocrystallization and electrochemical control of crystal growth:
fundamental considerations and electrodeposition of metals. J. Phys. D: Appl. Phys 1991,
24, 217–225.
[10] Scharifiker B, Hills G. Theoretical and experimental studies of multiple nucleation. Electrochim. Acta.
1983, 28, 879–889.
[11] Scharificer B, Mostany J. Three dimensional nucleation with diffusion controlled growth. Number
density of active sites and nucleation rate per site. J. Electroanal. Chem. 1984,
177,13–23.
[12] Gileadi E. Physical chemistry. Fundamentals, techniques, applications. Weinheim, FRG, Willey VCH
Verlag GmbH, 2011.
[13] Plieth W. Electrochemistry for material science. Amsterdam, The Netherlands, Elsevier, 2008.
[14] Zangari G. Electrodeposition of alloys and compound in the era of microelectronics and energy
conversion technology. Coatings 2015, 5, 195–218.
[15] De Mattel RC, Feigelson FS. Electrochemical deposition of semiconductors. In Mc Hardy J, Ludwig
F. ed Electrochemistry of semiconductors and electronics. Processes and devices. NJ, USA, Noyes
Publications 1992, 1–52.
[16] Bouroushian M. Electrochemistry of metal chalcogenides. Berlin, FRG, Springer Verlag, 2010.
[17] Krӧger FA. Cathodic deposition and characterization of metallic or semiconducting binary alloys or
compounds. J. Electrochem. Soc. 1978, 125, 2028–2034.
[18] Osial M, Widera J, Jackowska K. Influence of electrodeposition conditions on the properties of CdTe
films. J. Solid State Electrochem. 2013, 17, 2477–2486.
6 Underpotential deposition (UPD)

6.1 General remarks

According to thermodynamics, the reduction of metal ions Men+ on the corresponding

Me substrate can take place at an equilibrium potential of Me/Men+ couple ðEeq Þ, but
in practice, it occurs at a potential slightly negative than Eeq (e.g., deposition of Ag+ on
Ag). Reduction of Men+ ions on foreign metal substrate (S) and formation of bulk de-
posit usually require more energy and occur at much more negative potentials than
Eeq . If the deposition potential is more negative than the equilibrium potential, the
deposition process is named “overpotential deposition” (OPD). Plenty of experimental
results pointed out that the reduction of metal ions on foreign metal substrate, result-
ing in the formation of monolayer, takes place at potentials more positive than the
equilibrium potential of Me/Men+. In contrary to the OPD, such a process is referred
to as underpotential deposition (UPD) Fig. 6.1. UPD is a self-limiting process and is ter-
minated when a complete monolayer is formed.
j (mA/cm2)

Men+ + ne Me

UPD OPD
Eeq –E (V)

Fig. 6.1: Current density–potential plot showing the regions of

underpotential deposition (UPD) and overpotential deposition
(OPD).

What are the reasons of UPD?

UPD is the result of attractive interactions (Me–S) between the depositing metal ions
n+
Me and the substrate S, being stronger than the atomic interaction Me–Me, S–S. As a
result of these interactions, a metal is deposited in the UPD region as a monolayer (one
atomic layer). Formation of such monolayer involves partial or complete discharging of
Men+ ions (partial or complete charge transfer) and potential-dependent adsorption re-
ferred to as Faradaic adsorption. If the UPD process is carried out on polycrystalline sub-
strate or single crystal real surface, the nucleation and growth of monolayer occur
preferentially at defect sites and expanded from there (see 2D nucleation and growth in
previous chapter). Now and then, if the surface of the substrate is very rough, the forma-
tion of two atomic layers is observed.

https://doi.org/10.1515/9783111160986-006
158 6 Underpotential deposition (UPD)

At first look, it appears that the UPD process taking place at more positive potential
than Eeq defines the laws of thermodynamics, but it does not, as we will see later [1, 2].
Let us consider the same two redox reactions of Men+ ions, the first one on the
corresponding Me and the other on a metal substrate S. In the first case:

RT aMen +
Men + + neðMeÞ $ Me ðMeÞ, Eeq = E0 + ln
nF aMe

further

RT cMen + fMen +
Eeq = E0 + ln
nF aMe

for the pure metal, we can assume that aMe = 1,

then
RT
Eeq = E0 + lncMen + fMen + (6:1)
nF

In the second case: Men+ + ne(S) $ Me (S), the surface of the substrate is only partially
covered, so we can assume that aMe = fMe θ, where θ is the degree of coverage of the
substrate (S) surface. The product fMe θ < 1, since θ can be only ≤ 1 and fMe < 1.
The potential of monolayer formation EML is

RT cMen + fMen +
EML = Eeq ðθÞ = E0 + ln (6:2)
nF fMe θ

and can be further written as

RT
EML = Eeq ðθÞ = Eeq − ln fMe θ (6:3)
nF

so EML = Eeq ðθÞ > Eeq .

Conclusions from these equations are as follows:
i. the potential of the formation of metal monolayer, EML , is more positive than the
equilibrium potential Eeq ;
ii. if θ is constant, the tenfold increase of Men+ ions concentration shifts EML in the
positive direction of 59/n mV;
iii. if θ is equal to 1, the formation of monolayer is completed.

6.2 Experimental examples – UPD features

In Fig. 6.2 we show the shape of cyclic voltammogram obtained during the deposition
of Men+ ions on S metal substrate (e.g., Pb2+ on polycrystalline silver). Two cathodic
and two anodic peaks of current density are recorded. Peak 1 corresponds to a reduc-
 6.2 Experimental examples – UPD features 159

jp,OPD
3
jp,UPD
j (mA/cm2)

3 2 1
0 Fig. 6.2: Current density–potential curve recorded for
the deposition of a metal. The end of the UPD region
is marked by the red curve. At more negative
2 1 potentials, the overpotential deposition takes place.
E(V) The UPD shift is pointed on the potential axis. Adapted
Eeq ΔEUPD EUPD from [3].

tion of Men+ ions at more positive potential (UPD) and formation of a monolayer, and
peak 2 corresponds to bulk deposition of metal (OPD). The anodic current density
peaks 3 and 4 correspond to striping the bulk deposit and monolayer, respectively.
The difference between the potentials of the peak current density of formation of UPD
monolayer (1) and the equilibrium potential of the same metal (2) is noted as UPD
shift: ΔEUPD = Ep, UPD − Eeq . When Ep, UPD is determined in solution with concentration
(1 M) and UP deposition is reversible, this potential is chosen as the standard potential of
the formation of UPD monolayer EML (under the assumption that fMe θ, fMen + are equal
to 1). Since, Ep, UPD and Eeq are described by the same Nernst equation, the ΔEUPD shift
does not depend on the concentration of metal ions in solution when the UPD process is
reversible. In this case, the potentials of cathodic and anodic peaks on CV curves are not
dependent on the scan rate up to some ѵ value, characteristic for each system [3].
The UPD shift depends on the type of substrate and may range from 0.1 V even up
to 0.5 V. For instance, when Ag is deposited on polycrystalline Pt, Pd, and Au, the UPD
shift is equal to 0.44, 0.30, and 0.52 V, respectively. When different metals are depos-
ited on the same substrate, the UPD shift varies too. In the case of the deposition of Tl,
Cd, on polycrystalline Ag, it is equal to 0.28 and 0.16 V, respectively. The shift values
depend also on the crystallographic structure of the single-crystal substrate.
Kolb [4] studying UPD phenomena on a large amount of M–S samples derived ex-
perimental correlation between the underpotential shift ΔEUPD and work function Φ
of metal and substrate:

ΔEUPD = 0.5ΔΦ (6:4)

Let us go back to Fig. 6.2 to the part of the CV curve concerning UPD of the monolayer
formation and removal. As potential is changed with time: EðtÞ = E + νt (ѵ is a sweep
rate of potential), we can plot the current–time curve and easily calculate the charge:
160 6 Underpotential deposition (UPD)

ðE
q= jF dt (6:5)
Ein

where jF is the current density associated with the UPD process.

The surface under the curve (red color) is now divided into several parts 1–3.
These under lines 1 and 2 correspond to the formation and removal of partial mono-
layer while the surface under line 3 corresponds to the formation and striping of the
complete monolayer. Let us denote the charge required for complete monolayer for-
mation by qi and fractional coverage of substrate (electrode) by θ, then the Faradaic
charge consumed at any potential during formation of the partial monolayer is
qF = qi θ. As qF and in consequence θ depends on the potential, we may now introduce
the new notation – the adsorption pseudocapacitance, CF :



dqF dθ
CF = , CF = qi (6:6)
dE μ dE μ

If we sweep the potential from its initial value Ein ðθ = 0Þ to E ðθ = 1Þ and back to Ein
value, we record a Faradaic current jF of the formation and removal of a monolayer:





dθ dθ dE dE
jF = qi = qi = CF = CF ν (6:7)
dt dE dt dt

The equation points out that the current density of formation and removal of UPD
monolayer depend linearly on the sweep rate, ν.
The UPD was observed also for nonmetals such as hydrogen, oxygen, and halo-
gens. Let us concentrate on a very important reaction from a scientific and techno-
logical point of view – on H2 evolution (HER) [5]. In the first step, the hydrated ions
H3O+ move to the electrode double layer and undergo discharge with the formation
of electroadsorbed H on the electrode surface. Two types of electroadsorbed H are
distinguished: the underpotential deposited HUPD and overpotential deposited HOPD.
Experimentally, the electroadsorbed HUPD is observed at potential E more positive
than E0 ðH + jH2 Þ, ½E0 ðH + jH2 Þ = 0, while HOPD is formed when E < E0 ðH + jH2 Þ:

Me + H3 O + + e ! Me-HUPD + H2 O, E>0
+
Me + H3 O + e ! Me-HOPD + H2 O, E<0

It is known that UPD of H takes place at Pt, Pd, Rh, and Ir, but not at Ag and Au elec-
trodes, while OPD of H occurs on all metallic and other conducting substrates at
which HER occurs. Both electroadsorbed HUPD and HOPD can undergo further change
to the absorbate state, if the substrate (electrode) is capable of absorbing hydrogen
atoms:

HUPD ! Habs , E < 0; HOPD ! Habs , E>0

 6.3 Underpotential deposits – catalytic properties 161

The formation of HUPD at polycrystalline Pt is demonstrated in Fig. 6.3 In this figure,

the HUPD, 1 and HUPD, 2 ascribe the peaks of desorption and adsorption of underpoten-
tial deposited hydrogen.

0.2 PtO formation

HUPD,1

0.0
j/mA/cm–2

–0.2
HUPD,2 PtO reduction

–0.4
0.0 0.4 0.8 1.2
E vs NHE/V

Fig. 6.3: Cyclic voltammogram recorded for Pt in 1 N H2SO4. It shows the regions of HUPD, electrosorption
(HUPD,2), electrodesorption ( HUPD,1), and PtO formation and reduction. Unpublished data, courtesy of
B. Maranowski, M. Szklarczyk (Faculty of Chemistry, Warsaw University).

It is worth to mention that only HOPD species participate in the reaction of H2 evolution:

Me-HOPD + H3 O + + e ! Me + H2 ðgÞ + OH − , Volmer-Heyrowsky mechanism

2 Me-HOPD ! 2Me + H2 ðgÞ, Volmer mechanism

The fact that UPD and OPD of H take place in different range of potentials (UPDH, E > 0;
OPDH, E < 0) points to the differences in their Gibbs free energy of electrosorption (for
UPDH Δec−ads G < 0, for OPDH Δec−ads G > 0). It implies further that these two species oc-
cupy two distinct adsorption sites on the electrode surface. Plenty of spectroscopic ex-
periments were carried out but they are not fully conclusive. Interested readers will find
much more about UPD of H in the review [5].

6.3 Underpotential deposits – catalytic properties

We can apply many electrochemical methods together with microscopic techniques

for UPD, controlling in this way the amount, distribution, and structure of UP depos-
its. In fact, we can obtain the metal substrate decorated with nanostructural objects.
It is worth to mention the two methods based on UPD: surface-limited replacement
reaction (SERR) and electrochemical atomic layer epitaxy (ECALE) [6]. The SERR
method has been developed by Brankovic et al. [7]. The process consists of the depo-
sition of metal Me (2) on a metal substrate Me (1) under UPD condition and after
that the replacement of M (2) monolayer by more noble metal ions, N, takes place,
162 6 Underpotential deposition (UPD)

forming one atomic layer of N atoms. The driving force for such replacement is the
difference in the redox potentials of Me/Me (2)n+ and N/Nn+ couple, so the process
may take place at open-circuit potentials. An example is the replacement of Cu UPD
monolayer on Pd single crystal by Au. The second method was introduced by Stick-
ney [8]. This method enables the formation of semiconductor nanofilms “layer by
atomic layer,” each layer being deposited under the UPD control. This method can
be used to obtain a wide range of semiconductors composed of metal and one of the
nonmetal elements such as S, Se, Te, and As. The schemes of both methods SLRR and
ECALE are presented in Fig. 6.4(a, b).

(a)
Au Cu
Au Cu Cu
Au Au Cu Cu

Cu Cu Cu Cu Cu Cu Cu Au Au Au Au Au Au Au

Pd Pd

(b)

S
Cd Cd
S S S
Cd Cd Cd Cd

Fig. 6.4: (a) Scheme illustrating SERR (surface-limited replacement reaction) method. Adapted
from [6]. (b) Scheme illustrating ECALE (electrochemical atomic layer epitaxy) method showing formation
of thin semiconducting film in membrane. The formation may be carried out by using electrochemical
methods or chemical vapor deposition.

UPD and derivative methods can be applied for the formation of nanostructures:
nanoparticles and monolayers that are on the top of metal substrate changing its
properties and playing an important role in electrocatalysis. The foreign ad-atoms
may modify the electrocatalytic activity of substrate by (i) changing local electron
density on the surface and work function of the substrate, (ii) preventing the poison-
ing of substrate surface by strong adsorption of intermediates, (iii) creating the en-
ergetically different active sites for adsorption, and so on. The enhancement of
electrocatalytic properties manifests itself in the decrease of the overpotentials, in
the improvement of reversibility of reaction, or in selectivity changes. There are
many electrochemical reactions that are important from a practical point of view,
such as oxygen reduction, hydrogen evolution, and oxidation of organics fuels. In all
of them, the UP deposits were used as electrocatalysts to improve the efficiency and
selectivity of the reactions. The reduction of oxygen was influenced by the UP deposit.
This influence depended on the type of substrate [2]. The catalytic effects were observed
 Bibliography 163

in acidic solution when PbUPD and BiUPD were deposited on Au and PbUPD on glassy car-
bon, and also in alkaline solution for PbUPD, TlUPD, BiUPD, and CuUPD deposited on Pt.
These effects mainly result from the decrease of overpotential of charge transfer and
from changes in the mechanism of O2 reduction from 2e to 4e (Chapter 4). However, in
acidic solutions, PbUPD, TlUPD, BiUPD, and CuUPD deposited on Pt inhibit the reduction of
O2. Some experimental results also pointed out that the reaction of H2 evolution is in-
hibited when Pt is covered with ad-atoms of heavy metals such as PbUPD, SnUPD, and
TlUPD [2]. It was mentioned in Chapter 4 that the platinum has very good electrocatalytic
properties, and they are strongly enhanced if Pt is modified with UP deposit. Underpo-
tential deposited PbUPD and BiUPD on Pt catalyze the oxidation of HCOOH, HCHO, and
CH3OH, the potential fuel in the fuel cells (Chapter 8).
Interested readers can find much more about the theory and application of UPD
in [9, 10].

Bibliography

[1] Sudha V, Sangaranarayanan MV. Underpotential of deposition of metals: structural

and thermodynamic consideration. J. Phys. Chem. B 2002, 106, 2699–2707.
[2] Kokinidis G. Underpotential deposition and electrocatalysis. J. Electroanal. Chem. 1986, 201, 217–236.
[3] Gileadi E. Physical electrochemistry, fundamentals, techniques and applications. Weinheim, FRG,
Wiley-VCH Verlag GmbH & Co.KGaA, 2011.
[4] Kolb DM. Advances in electrochemistry and electrochemical engineering. Eds. Gerischer H, Tobias
CW. New York, USA, J. Wiley & Sons, 1978 (v.11).
[5] Jerkiewicz G. Electrochemical hydrogen adsorption and absorption. Part I. Underpotential
deposition of hydrogen. Electrocatalysis 2010, 1, 179–199.
[6] Zangari G. Electrodeposition of alloys and compounds in the era of microelectronics
and energy conversion technology. Coatings. 2015, 5, 195–218.
[7] Brankovic SR, Wang JX, Adzic RR. Metal monolayer deposition by replacement of metal adlayers on
electrode surface. Surf. Sci. 2001,474, L.171–L.179.
[8] Stickney JL. Electrochemical atomic layer epitaxy (ECALE). Nanoscale control in the electrodeposition
of compound semiconductor. In Advances in electrochemical science
and technology. Eds. Alkiire RC, Kolb DM. New York, USA, Willey & Sons, 2009, 7, 1–105.
[9] Mayet N, Servat K, Kokoh B, Napporn TW. Probing the surface of noble metals electrochemically by
underpotential deposition of transition metals. Surfaces. 2019,
2, 257–276.
[10] Oviedo OA, Reinaudi L, García SG, Leiva EPM. Underpotential deposition. From fundamentals and
theory to applications at the nanoscale. Springer International Publishing AG, Switzerland, 2016.
7 Electrochemical methods in the formation
of nanostructures

7.1 General remarks

The so-called dimensionality plays a crucial role in determining the properties of

nanostructures obtained via the electrochemical deposition, such as quantum phe-
nomena in the electron transport, photonic behavior of nanocrystals, and even mag-
netic properties of nanostructured transition metals [1–8]. In addition, the catalytic
behavior of deposited nanoparticles, noble metals in particular, resulted in the tre-
mendous resources being devoted to the development of new methods for preparing
nanometer and micron-scale structures of various topologies.
As was introduced in the previous chapters, nanostructures can be formed by
the two electrochemical methods: UPD and OPD. In both cases such synthesis starts
at the surface of a conducting substrate (working electrode), S. This process is also
referred to as heterogeneous nucleation and its mechanism has been described in
Chapters 5 and 6. Because UPD is a self-limiting process and is terminated when a
complete monolayer is formed, the deposition is limited to the surface of conducting
substrate, S. Nevertheless, under controlled UPD potentiostatic or galvanostatic con-
ditions, the deposition of nanostructures, sometimes called 0D nanostructures, can
be effected. The UPD, offering a “bottom-up” approach, has been widely used to con-
trol the growth of nanostructures of different types and shapes. This includes Au octa-
hedral nanocrystals, nanorods, bimetallic Au–Pd hexoctahedra, Ag “flakes,” Pt–Cu
hierarchical structures, and Pd concave nanostructures. For more detailed library of
different surface grown structures electrodeposited for catalytical purposes, the inter-
ested reader is referred to excellent resource books [1, 2].
These nanostructures can subsequently act as “seeds” under the OPD conditions
for further growth of 2D structures: nanowires, nanotubes, or nanorods. For these
types of depositions, the nucleation and growth can be controlled separately by vary-
ing the potential or current values, their duration, and profile. In order to obtain well-
ordered and uniform in size and shape nanostructures, it is necessary to use tem-
plates at the electrode surface or the structural characteristics of the electrode surface
itself. The following sections will describe briefly such an approach yielding 0D and
2D nanostructures (Fig. 7.1).

https://doi.org/10.1515/9783111160986-007
 7.2 Template-assisted electrodeposition of nanostructures 165

Reactant Reactant

Catalyst Nanowire Nanowire

nanocluster nucleation growth

Fig. 7.1: 0D nanocluster acting as a “seed” for further nucleation and growth of 1D nanowire.

7.2 Template-assisted electrodeposition of nanostructures

Generally, the template-assisted electrodeposition can be divided into two groups [9]:
– active template-assisted electrodeposition and
– restrictive template-assisted electrodeposition.

7.2.1 Active template-assisted electrodeposition

Surface template-assisted electrodeposition is widely used to synthesize various kinds of

nanostructures of tailored morphology and topology. The key issue in using such electro-
deposition is the preparation of proper templates because the surface characteristics of
the electrode surface affects – “templates” – the obtained nanostructures. The nucleation
and subsequent growth of nanostructures is influenced by Gibbs’ surface free energy of
the substrate and crystallographic lattice mismatch at the boundary between the surface
and nucleus of reduced metal. For instance, electrodeposited metal nanostructures, such
as Ag, Au, or Pd, tend to form small nanoclusters on surface defects, such as step edges
at the surface (see schematic drawing, Fig. 7.2).

Fig. 7.2: Nanostructures deposited on the energetically

favored step edges in the crystalline surface of a substrate
electrode.

Since for this type of electrodeposition it is required to have a detailed control over
the electrode surface, numerous works were devoted to the highly oriented pyrolytic
graphite (HOPG) and graphene surfaces, as well as single crystal metal surfaces as
templates. For instance, it was found that in the case of HOPG, palladium deposition
reaction at early stages proceeds through progressive nucleation under diffusion con-
trol. Palladium islands and the development of their dense distribution indicated the key
166 7 Electrochemical methods in the formation of nanostructures

role of step edge energy barrier in determining the growth pattern of Pd electrodeposi-
tion [10]. Similar results were obtained in the case of atomically flat terraces of HOPG
and Ag deposits. Preferential deposition of silver nanostructures was observed at step
edges and other surface defects [11, 12]. Other metals, like Pt nanocrystals, were depos-
ited on the basal plane of HOPG with the help of pulsed potentiostatic method with con-
trolled pulse amplitude, duration, and temporal resolution. Similar approach was used
for one-step, pulsed potentiostatic electrodeposition of Ag on Au(111) single crystal via
the UPD. The analysis of scanning tunneling microscopy images revealed that stable Ag
nanoclusters are formed all over the surface of Au(111) single crystal, without preferen-
tial tendency of deposition at the step edges, nor diffusional agglomeration on the defect
sites in time. Moreover, by controlling the potential step it was possible to obtain Ag clus-
ters of monoatomic height [13, 14].

7.2.2 Restrictive template-assisted deposition

As we have shown above, the template-based electrochemical synthesis of various con-

ductive nanostructures is a very general and relatively simple, low-cost method. The
subsequent growth of nanostructures on “seed” nanoclusters to form nanowires or
nanotubes will continue as long as the electric field is applied. Current density on the
tips of growing 2D structures will increase with their elongation, because of decreasing
distance between these tips and counter electrode. Therefore, the new deposits will be
more likely formed on the tips, resulting in the propagation of growth. However, the
use of surface topology of a working electrode as a template for such synthesis has its
limitations, since it is almost impossible to control the surface distribution, organiza-
tion, and length of the deposited structures. Therefore, this method is hardly used in
practice for the synthesis of nanowires or nanotubes. Instead, porous 3D matrices with
desired channel distribution and length are used to control the growth of nanowires
and nanotubes via the electrochemical deposition. Figure 7.3 illustrates the scheme of
the typical setup used for the electrochemical deposition of 2D nanostructures with the
help of an inert, nonconducting porous membrane. Examples of such matrices include,
but are not limited to, track-etched polycarbonate filtration membranes, porous alu-
mina, porous titanium dioxide (TiO2, anatase), mesoporous silica, conducting polymers,
and so on. Some of them can be prepared electrochemically, such as porous alumina
and porous titanium dioxide, finally yielding well-organized 2D nanotubes, and there-
fore we will outline their formation in more detail. They can serve as electrochemically
fabricated arrays of nanotubes – nanostructures themselves, but, of course, they can be
used also as inert matrices (porous alumina in particular) for electrochemical deposi-
tion of conducting nanorods and nanowires, as shown in Fig. 7.3.
The most facile electrochemical synthesis of vertically aligned metal oxide nano-
tubular structures is the so-called self-ordering electrochemical anodization (SOA). It
can be carried on a suitable metal, for example, Al, Ti, W, Zn, and Zr. Exposing such
 7.2 Template-assisted electrodeposition of nanostructures 167

I (mA)
I
Porous
membrane III
II

t (s)
Metal electrode

Fig. 7.3: Typical experimental setup for the restricted template-based electrochemical growth of 2D
nanostructures. Right-hand side panel shows the current–time profile for the deposition of metallic
nanowires growing inside the nanometer-sized pores in a nonconducting, templating membrane.

metal to a sufficiently high anodic voltage, an oxidation reaction will be triggered.

The reaction path can proceed through different mechanisms, depending mainly on
the electrolyte content and anodization parameters, that is, voltage and/or current
densities. The first path is obvious: the metal ions are formed, and the metal is dis-
solved continuously (corrosion, electropolishing). The second path, for example, in
the presence of oxygen-saturated solution, involves the formation of insoluble oxide,
and a passivation layer is formed. However, under specific conditions of competition
between oxide formation and dissolution, a self-organized nanotube or nanorod ar-
rays can be formed. The best of investigated systems, in which almost perfect organi-
zation was achieved, was the self-ordered nanotubular arrays of aluminum oxide
structures. It is formed in acidic electrolytes and has tunable porous morphology and
mechanical and chemical strength. Its formation triggered hundreds of research work
and numerous technological applications, including templating and nanostructuring
of metal, semiconductor, or polymer materials [15, 16]. Many of the principles and
mechanisms described for porous alumina SOA can be adapted for the formation of
self-organized 1D structures on other metals, taking into account some specifics of
electrochemical behavior of their oxides.

7.2.2.1 Porous anodic alumina (PAA) matrices

The porous structure of aluminum oxide was first reported in 1953 [17]. Since that time
a whole gallery of nanotubular arrangements of different pore geometries has been
proposed, such as hexagonal, square, and triangular openings [18]. When PAA matrix is
used for the formation of metal or carbon nanowires, the pattern of their growth inside
the pores of PAA under potentiostatic conditions can be followed by measuring the cur-
rent versus time curve. Three zones can be distinguished, as shown in the right panel
of Fig. 7.3: zone I corresponds to the initial deposition of metal inside the pores, and
zone II reflects the complete filling of some pores which gives rise to the formation of
168 7 Electrochemical methods in the formation of nanostructures

an array of hemispherical nanoelectrodes. They finally coalesce into a continuous

metal layer, completely covering the PAA surface (zone III). The current saturates due
to semi-infinite diffusion control. Before zone III is reached, the cathodic deposition of
metal can be stopped by disconnecting the applied voltage. Further careful chemical
dissolution of PAA template leaves the metal nanowire “forest” protruding from the
substrate surface. Similar nanowires can be formed from conducting polymers, semi-
conductors, and oxides [19].

7.2.2.2 Porous anodic titanium dioxide matrices

Hollow titanium dioxide (TiO2) nanostructures, such as nanotubes, have attracted a
broad range of scientific attention, not only as templates but also as nanostructured
components for solar energy conversion and as possible nanocarriers for drug deliv-
ery systems [20]. As templates for growing well-aligned 1D nanostructures, titania
nanotubes were successfully used for coating or filling the active materials on/into
TiO2, for fabrication of electrodes in supercapacitors and batteries with an enhanced
capacitance due to the enlargement of electrode area [21–24].
Formation of compact tubular layers of TiO2 is usually carried out in aqueous elec-
trolytes, as in the case of porous alumina, at a constant potential in the range of 10–30
V, applied to pristine Ti electrode. In the absence of competing factors, such as fluoride
ions, this will result in a compact TiO2 layer. However, in the presence of F− ions that
form water-soluble complexes TiF62−, complexation of Ti4+ ions formed during the an-
odic process as well as dissolution of TiO2 under the influence of fluorides occurs, both
processes competing with the formation of compact TiO2 layer. Under the appropriate
electrochemical conditions balance [23], well-organized tubular array of TiO2 is ob-
tained (Fig. 7.4c).

7.2.3 Concluding remarks

As described earlier, electrochemical methods are extremely suitable for the preparation
of various nanostructured materials. Both cathodic and anodic polarization can be ap-
plied to obtain such structures of controlled shape, distribution, dimensions, and proper-
ties. While the reduction procedures are predominantly used for the electrodeposition of
metals, alloys, bimetallic structures, polymers, semiconductors, and oxides with precise
control over their shapes, oxidation processes are mainly used to prepare oxides; porous
alumina and titania being their most frequent representatives.
 7.2 Template-assisted electrodeposition of nanostructures 169

(a) (b)

(c) (d)

(e) (f)

Fig. 7.4: Panel a: Silver nanoneedles electrodeposited on HOPG electrode (SEM image, 500 nm frame). These
nanoneedles were formed randomly at the flat surfaces of HOPG. Their needlelike growth was controlled by
mercaptopropionic acid (MPA) preferential adsorption. Courtesy of Prof. M. Mazur, Faculty of Chemistry,
University of Warsaw. Panel b: SOA of porous alumina matrix deposited electrochemically on Al electrode (top
view). Courtesy of Dr A. Brzózka, Faculty of Chemistry, Jagiellonian University, Kraków. Panel c: P.K. own SEM
image of TiO2 nanotube array, ca. 50 nm in diameter. Somehow tilted bundles resulted from nonoptimized
anodization process. Panel d: ZnO nanorods grown electrochemically onto FTO electrode from Zn(NO3)2
aqueous electrolyte. Courtesy of Dr K. Zarębska, Faculty of Chemistry, University of Warsaw. Panels e, f:
Template-assisted electrochemical growth of conducting polymer (polyindole) nanowires from its monomer
solution. Courtesy of Dr M. Osial, Faculty of Chemistry, University of Warsaw.
170 7 Electrochemical methods in the formation of nanostructures

7.3 Applications of nanostructured materials

Every time, when the high surface area versus small dimensions for efficient catalysis,
energy conversion, and storage is required, present-day engineering turns to nanoma-
terials. This is because conducting and semiconducting materials at the nanometer
scale exhibit unusual and frequently advantageous behavior compared to their bulk
counterparts. The origin for these differences stems from the fact that a significant
fraction of the total number of atoms and molecules in the nanostructures reside on
the surface and therefore the surface energy plays the dominant role in their proper-
ties. Nanostructured materials offer an extremely broad possibilities of applications,
ranging from electronics, photonics, energy storage devices, catalysis, medical appli-
cations, and so on. It is unrealistic to summarize all possible applications or foresee
the new ones within this section. Therefore, here we will focus on several, most com-
mon uses of electrochemically synthesized nanostructures. More broad information
can be found in [1–8], as well as in [25, 26].

7.3.1 Catalysis

Electrocatalysis was described in detail in Chapter 4. Here, we will only outline the nano-
particle-based electrocatalysis. Noble-metal-based nanomaterials have attracted much
interest in their promising potentials in fields of energy-related and environmental ca-
talysis. Tailoring and controlling the surface/interface structure of such metal-based
nanomaterials at the atomic scale have great significance for optimizing the performan-
ces in practical catalytic applications. Nanostructures are used to control electrochemical
reactivity, such as catalytical oxygen electroreduction on several reactions, exemplified
here as O2 electroreduction (ORR), CO2 electroreduction, and ethanol electro-oxidation.
The presence of low-coordinated sites or facets, particle size, shape, and composition in
nanostructures can be used to tune reactivity and selectivity. Since the price of noble
metals, platinum, in particular, is skyrocketing, while their availability is decreasing,
other nanostructured catalysts began to emerge, with low or no Pt content. And so, in
recent years, many alloy systems and core–shell systems with lower Pt content was re-
ported and used, particularly for ORR in fuel cells and for decreasing the emission of
CO2 and CO. Systems such as PtPd, PtAu, PtCu, PtNi, and PtW have been studied of vari-
ous morphologies and compositions. Much research has been devoted to Pt-free nanoca-
talysts, from which palladium-, ruthenium-, and iridium-based nanostructures were
expected to at least replace Pt-based electrocatalysts. However, even though Pd is ca.
three times more expensive than Pd, the catalytic efficiency of palladium, the best of
other metals studied, was at least five times lower than that of Pt [27–29]. Therefore, the
efforts to design nanostructures with better activity than Pt in catalysis are still required
and in progress.
 7.3 Applications of nanostructured materials 171

7.3.2 Batteries: supercapacitors

As will be discussed in Chapter 8, nanostructuring of electrodes plays a crucial role in

lithium-ion batteries and supercapacitors. Moreover, commercialization requirements di-
rect the researchers toward environmentally safer and cheaper electrode materials with
good capacitive behavior, such as oxides and hydroxides of Ni, Co, Mn, W, Ru, and V, all
in a form of nanostructures. In addition, carbon-derived materials, such as carbon nano-
tubes, have attracted much attention because of their advantageous electronic and me-
chanical properties. As an example, a simple method of direct electrodeposition of
hexagonal arrays of Ni(OH)2 on titanium substrate has been proposed for high specific
capacitance material for supercapacitors [25]. In the case of Li-ion secondary batteries,
also discussed later in this book, the improvement of storage lifetime and the number of
charging/discharging cycles can be achieved by nanostructuring of the electrode material.
This approach takes an advantage from the experimental and theoretical observations
that the smaller the dimensions of the electrode materials that react with lithium (nano-
structuring), the better the electrochemical performance of the cell due to the improved
Li intercalation densities [26].

7.3.3 Photoelectrochemical devices: solar cells

Similar considerations as in the case of batteries and supercapacitors can be applied

for nanostructures used for solar energy conversion, discussed in Chapter 11. Here,
we will mention titanium-derived nanostructures (nanoparticles, nanotubes, and
nanorods) as their use in the photoelectrochemical devices seem to be most explored.
These nearly translucent structures can be easily dye-sensitized, tailoring their spec-
tral/electrochemical behavior for more efficient light energy conversion. More de-
tailed information can be found in excellent review [23] and references therein. TiO2
nanotubes show a much higher electron diffusion length compared to nanoparticles.
It was proved experimentally that 15–20 µm long nanotubes show the maximum solar
cell efficiency. Along the same line, the organized anatase (thermally annealed form
of rutile TiO2) nanotube layers are more efficient than the rutile phase, as the electron
transport is faster, leading to high-efficiency solar cells [18].

7.3.4 Biological applications: nanozymes

In recent years a broad range of electrosynthesized nanostructures have been consid-

ered as possible theranostic tools. In particular, due to the luminescent and superpar-
amagnetic behavior of certain metal oxides (e.g., Gd2O3, Fe3O4), they were tested as
tissue imaging tools. Biocompatibility of oxides allowed for their promising use as a
drug delivery platform for multimodal treatment of numerous diseases, including
172 7 Electrochemical methods in the formation of nanostructures

cancer [30, 31]. However, it should be stressed that for this type of applications, tech-
niques other than electrochemical synthesis are employed, more suitable for obtain-
ing large amounts of nanostructures suspended in liquids, such as coprecipitation,
thermal decomposition of precursors, bulk redox reactions, and so on. So, here we
will outline the applications of those nanostructures that can be obtained electro-
chemically on the electrode surface. Along this line, it was reported lately that anodic
titanium dioxide nanotubular structures can be used for intracellular drug delivery.
When compared to solid-core nanostructures, such hollow nanotubes offer substan-
tially higher capacity for drug loading and can be optimized for the appropriate
drug release kinetics. Additionally, they can be loaded with magnetic nanoparticles
for site-specific guidance in an external magnetic field [32, 33]. Metallic nanostruc-
tures of different shapes can also be used in biomedical applications. Among them,
gold nanoparticles received broad interest, because they are known to be stable,
easily electrosynthesized (even though they are obtained mostly in a form of suspen-
sion). Their shape can be controlled experimentally (e.g., gold nanorods); they pos-
sess tunable optical properties and tailored surface chemistry. For more detailed
biological applications of metallic, carbon-derived, oxide, and semiconductor nano-
particles, interested reader is directed to Ref. [34].
Nanozymes: Although we already mentioned the electrocatalytic activity of vari-
ous nanostructures, noble-metal nanoparticles in particular, their broad applica-
tions were generally focused on energy-related and environmental catalysis areas.
However, some nanomaterials have been found to exhibit unexpected enzymelike
activities, and great advances have been made in this area due to the unique charac-
teristics of such nanomaterial-based artificial enzymes (nanozymes). From the elec-
trochemical point of view, four types of redox enzymes have been reported to be
mimicked by either metallic (e.g., Au) or oxide nanoparticles (e.g., Fe3O4, CeO2).
These redox enzymes include superoxide dismutase, peroxidase, oxidase, and cata-
lase. In all these cases, the enzymelike activities are due to the intrinsic behavior of
nanoparticle cores instead of the functional groups present on their surfaces. For
more comprehensive information regarding nanozymes, their structure, and prop-
erties, we suggest two successive review papers showing the research progress and
references therein [35, 36].

Bibliogaphy

[1] Parak WJ, Manna L, Simmel FC, Gerion D, Alivisatos P. Quantum dots. In Nanoparticles.
From theory to application. Schmid G. Wiley-VCH, Germany, 2004, 4–362.
[2] Nasirpouri F. Electrodeposition of nanostructured materials. Springer AG, Switzerland, 2017.
[3] Petri OO, Electrosynthesis of nanostructures and nanomaterials. Russ. Chem. Rev. 2015, 84, 159–193.
[4] Zhang YW. Ed. Bimetallic nanostructures: Shape-controlled synthesis for catalysis.
VCH Wiley, Germany, 2018.
 Bibliogaphy 173

[5] Efekhtari A. Ed. Nanostructured materials in Electrochemistry. Wiley-VCH, Germany, 2008.

[6] Losic D, Santos A. eds. Electrochemically engineered nanoporous materials. Springer,
Switzerland, 2015.
[7] Cao G. Nanostructures and nanomaterials. Synthesis, properties and applications. London, Imperial
College Press, 2006.
[8] Rauh A, Carl N, Schweins R, Karg M. Role of absorbing nanocrystal cores in soft photonic crystals: A
spectroscopy and SANS study. Langmuir 2018, 34, 54–867.
[9] Bera D, Kuiry SC, Seal S. Synthesis of nanostructured materials using template-assisted
electrodeposition. JOM 2004, 56, 49–53.
[10] Gimeno Y, Hernandez Creus A, Carro P, Gonzalez S. Salvarezza RC, Arvia AJ. Electrochemical
formation of palladium islands on HOPG: Kinetics, morphology, and growth mechanisms. J. Phys.
Chem. B 2002, 106, 4232–4244
[11] Potzschke RT, Gervasi CA, Vinzelberg S, Staikov G, Lorenz WJ. Nanoscale studies of electrodeposition
on HOPG (0001). Electrochim. Acta 1995, 40, 1469–1474.
[12] Zoval JV. Electrochemical deposition of silver nanocrystallites on the atomically smooth graphite
basal plane. J. Phys. Chem. 1996, 100, 837–844.
[13] Mazur M. Electrochemically prepared silver nanoflakes and nanowires, Electrochem. Comm. 2004, 6,
400–403.
[14] Borissov D, Tsekov R, Freyland W. Pulsed electrodeposition of two-dimensional Ag nanostructures
on Au(111). J. Phys. Chem. B 2006, 110, 15905–15911.
[15] Sulka GD. Highly ordered anodic porous alumina formation by self-organized anodizing.
In: Eftekhari A. ed. Nanostructured materials in electrochemistry. Wiley VCH, Germany, 2008,
pp.1–97.
[16] Brzózka A, Budzisz A, Hnida K, Sulka GD, Losic D., Santos A. eds. Electrochemically engineered
nanoporous materials. Springer, Switzerland, 2015, 219–272.
[17] Keller F, Hunter MS, Robinson DL. Structural features of oxide coatings on aluminium.
J. Electrochem. Soc. 1953, 100, 411.
[18] Masuda H, Fukuda K. Ordered metal nanohole arrays made by a 2-step replication of honeycomb
structures of anodic alumina. Science 1995, 268, 1466–1468; Masuda H, Asoh H, Watanabe M, Nishio
K, Nakao M, Tamamura T. Square and triangular nanohole array architectures in anodic alumina.
Adv. Mat. 2001, 13, 189–192.
[19] Choi J, Sauer G, Nielsch K, Wehrspohn RB, Gosele U. Hexagonally arranged monodisperse silver
nanowires with adjustable diameter and high aspect ratio. Chem. Mater. 2003, 15, 776–779.
[20] Kafshgari MH, Mazare D, Distaso M, Goldmann WH, Peukert W, Fabry B, Schmuki P. Intracellular
drug delivery with anodic titanium dioxide nanotubes and nanocylinders.
ACS Appl. Mater. Interfaces 2019, 11, 14980–14985.
[21] Global A, Faraji M. Fabrication of nanoporous nickel oxide by de-zincification of Zn–Ni/
(TiO2-nanotubes) for use in electrochemical supercapacitors. Electrochim Acta 100, 2013, 133–139.
[22] Wang Y-G, Wang Z-D, Xia Y-Y. An asymmetric supercapacitor using RuO2/TiO2 nanotube composite
and activated carbon electrodes. Electrochim. Acta 2005, 50, 5641–5646.
[23] Roy P, Berger S, Schmuki P. TiO2 nanotubes: Synthesis and applications, angew. Chem. Int. Ed. 2011,
50, 2904–2939.
[24] Lee K, Mazare A, Schmuki P. One-dimensional titanium dioxide nanomaterials: Nanotubes. Chem.
Rev. 2014, 114, 9385–9454.
[25] Zhao DD, Bao, SJ, Zhou WH, Li HL. Preparation of hexagonal nanoporous nickel hydroxide film and
its application for electrochemical capacitor. Electrochem. Comm. 2007, 9, 869–874.
[26] Beaulieu LY, Eberman KW, Turner RL, Krause LJ, Dahn JR. Colossal reversible volume changes in
lithium alloys. Electrochem. Solid-State Lett. 2001, 4, A137–A140.
174 7 Electrochemical methods in the formation of nanostructures

[27] Nie Y, Li L, Wei Z. Recent advancements in Pt and Pt-free catalysts for oxygen reduction reaction.
Chem. Soc. Rev. 2015, 44, 2168–2201.
[28] Duan S, Du Z, Fan H, Wang R. Nanostructure optimization of platinum-based nanomaterials for
catalytic applications. Nonomaterials, 2018, 8, 949.
[29] Mistry, H.; Varela, A.S.; Strasser, P.; Cuenya, BR. Nanostructured electrocatalysts
with tunable activity and selectivity. Nature Rev. Mater. 2016, 1, 1–14.
[30] Kafshgari M, Harding F, Voelcker N. Insights into cellular uptake of nanoparticles.
Curr. Drug Deliv. 2015, 12, 63–77.
[31] Steinhauser I, Spaenkuch B, Strebhardt K, Langer K. Trastuzumab-modified nanoparticles:
optimization of preparation and uptake in cancer cells. Biomaterials 2006, 27, 4957–4983.
[32] Kafshgari MH, Mazare A, Distaso M, Goldmann WH, Peukert W, Fabry B, Schmuki P. Intracellular
drug delivery with anodic titanium dioxide nanotubes and nanocylinders.
ACS Appl. Mater. Interfaces, 2019, 11, 14980–14985.
[33] Kafshgari MH, Kah D, Mazare A, Nguyen NT, Distaso M, Peukert W, Goldmann WH, Schmuki P, Fabry
B. Anodic titanium dioxide nanotubes for magnetically guided therapeutic delivery. Nature Research
Scientific Reports 2019, 9, 13439.
[34] Ashikbayeva Z, Tosi D, Balmassov D, Schena E, Saccomandi P, Inglezakis V. Application
of nanoparticles and nanomaterials in thermal ablation therapy of cancer, Nanomaterials, 2019, 9,
1195. doi:103390/nano9091195.
[35] Wei H, Wang E. Nanomaterials with enzyme-like characteristics (nanozymes): next-generation
artificial enzymes. Chem. Soc. Rev. 2013, 14, 6060–6093.
[36] Wu J, Wang X, Wang Q, Li S, Zhu Y, Qin L, Lou Z, Wei H. Nanomaterials with enzyme-like
characteristics (nanozymes): next generation of artificial enzymes (II). Chem. Soc. Rev. 2019, 48,
1004–1076.
8 Electrochemistry in energy conversion
and storage
8.1 Batteries

8.1.1 Electrochemical cell – fundamentals

The simplest electrochemical cell contains two electronic conductors (electrodes, mostly
metals) and ionic liquid conductor containing ionic species (electrolyte) between them.
The electronic conductor and its interface with electrolyte serve as the place where the
electrochemical reactions occur. To prevent any unwanted reaction with electrolyte
species, it is often necessary to apply diaphragm dividing the cell into two parts (half
cells). In this case, the additional potential is formed, called the liquid junction potential,
which can be limited by means of a salt bridge or binary electrode. The schemes of ex-
emplary cells without (A) and with junction (B) are as follows:

Ag Ag2 O, KOHaq , HgO Hg, Zn ZnSO4, aq CuSO4, aq Cu ðAÞ

Zn ZnSO4, aq ða1 Þ CuSO4, aq ða2 Þ Cu, Zn ZnCl2, aq ða1 Þ AgNO3, aq ða1 Þ Ag ð BÞ

The other kind of electrochemical cells are the concentration cells, in which the two
half cells differ only in the concentration of the same electroactive species taking part
in electrochemical reaction:

AgjAgCl, HClaq ða1 Þ, H2 ðp1 ÞjPt − PtjH2 ðp2 Þ, HClaq ða2 Þ, AgCljAg ðAÞ

Ag AgCl, HClaq ða1 Þ HClaq ða2 Þ, AgCljAg ðBÞ

As the voltage of concentration cells is low, they are not used practically in energy
production and will not be considered further.
Note that a slash | represents a phase boundary, a double slash || represents the
phase boundary whose potential is negligible as a result of application of separator be-
tween the two electrolytes, and coma separates two components in the same phase.
The overall chemical reaction in an electrochemical cell is formed by two inde-
pendent half-reactions taking part in half cells, each of them characterized by the in-
terfacial potential difference Δφ (Galvani potential, see Chapter 1). It is accepted that,
in the notation of cells, the electrode with a lower value of potential (Δφ1 , E1 ) is placed
on the left side of such notation. At this electrode, being negatively charged, named
anode, the oxidation reaction takes place. At another electrode, which is positively
charged, cathode, (Δφ2 , E2 ), the reduction reaction occurs:

https://doi.org/10.1515/9783111160986-008
176 8 Electrochemistry in energy conversion and storage

anode ðoxidation reactionÞ: Red1 ! Ox1 + ne

cathode ðreduction reactionÞ: Ox2 + ne ! Red2

The overall reaction taking place in the cell is the sum of these two reactions:

Red1 + Ox2 ! Ox1 + Red2

Electrochemical cells can operate in the three modes: equilibrium mode, galvanic
mode, and electrolytic mode.
Let us consider the behavior of the cell being in equilibrium mode (at equilib-
rium). The most convenient for laboratory purpose is the Daniell cell (Fig. 8.1(a)):
Zn ZnSO4, aq ða1 Þ CuSO4, aq ða1 Þ Cu

(a) (b)
ΔEeq = Eeq(cu/cu2+) – Eeq(Zn/Zn2+) Rex
A e
V Salt bridge V Salt bridge

Zn Cu Zn Cu

Anode Cathode Anode Cathode

Zn Zn2+ + 2e Cu2+ + 2e Cu Zn Zn2+ + 2e Cu2+ + 2e Cu
Zn2+ Cu2+ Zn2+ Cu2+
SO2–
4 SO2–
4 SO2–
4 SO2–
4
H2O (a ) (a ) H2O H2O (a ) (a ) H2O
1 1 1 1
Overall Overall
Zn + Cu2+ Zn2+ + Cu Zn + Cu2+ Zn2+ + Cu
reaction reaction

(c) Voltage
Source
VS
A e
V Salt bridge

Zn Cu

Cathode Anode
Zn2+ + 2e Zn Zn2+ Cu2+
Cu – 2e Cu2+
SO2–
4 SO2–4
H2O (a ) (a ) H2O
1 1

Overall Zn + Cu2+
Zn2+ + Cu
reaction

Fig. 8.1: (a) Scheme illustrating Daniell cell in the state of equilibrium. (b) Scheme illustrating Daniell cell
in galvanic mode. Discharge of the cell takes place. The potential of the cell decreases in time.
(c) Scheme illustrating Daniell cell in electrolytic mode. Note that comparing with galvanic mode, in the
electrolytic mode, the anode becomes the cathode.
 8.1 Batteries 177

If we put metallic Zn in one part of the cell containing Zn2+ ions and Cu in the other
part containing Cu2+ ions, some kind of equilibrium is established at the interfaces:
Zn|Zn2+, Cu|Cu2+, resulting of the following reactions:

Zn $ Zn2 + + 2 e

Cu2 + + 2 e $ Cu

The equilibrium of the whole system is characterized by the equilibrium potentials

of anode and cathode: Eeq, 1 ðZn2 + jZnÞ and Eeq, 2 ðCu2 + jCuÞ. Putting the electronic volt-
meter of very high resistance between the electrodes or compensating the equilib-
rium potentials of electrodes by an external potential source of the same absolute
value, but of opposite sign, we can measure the equilibrium potential of the cell:

ΔEeq = Eeq, 2 − Eeq, 1 (8:1)

This difference was called previously “electromotive force (emf)” but now the term
“zero-current cell potential” is often used.
Assuming that the conditions of thermodynamic reversibility at equilibrium are
fulfilled and that the volume work ð− pdV Þ is zero, we can apply the equation relating
the changes in Gibbs free energy with nonvolume work, wrev , at constant pressure
and temperature:

dG = dwrev (8:2a)

As the processes are reversible, the work done by such a system has its maximum
value (wrev = wmax ), so further we can write:

ΔG = wmax (8:2b)

In the electrochemical cell T, p are constant, and the only source of energy (work) is
electrode reactions, energetically characterized by changes in Gibbs free energy
(ΔrG). These reactions can produce electrical work pushing the electrons through an
external circuit. We can calculate the maximum electrical energy ( − nFΔEeq ), which
can theoretically be produced by the cell, using the following equations:

wmax = Δr G = − nFΔEeq (8:3)

v
Δr G = Δr G0 + RTlnQ, Q = Πi ai i (8:4)

− nFΔEeq = nFΔE0 + RTlnQ (8:5)

RT
ΔEeq = ΔE0 − lnQ (8:6)
nF

Δr G0 is standard Gibbs free energy of reaction, ΔE0 is standard cell potential, and Q is
the reaction quotient.
178 8 Electrochemistry in energy conversion and storage

The reaction quotient Q means the product of activities ai of all substrates and
reaction products raised to the power equal to their stoichiometric numbers ѵi (they
are positive for reaction products and negative for substrates).
The value of Δr G gives us the value of chemical energy, which we can obtain from
the reactions; this value decreases to zero in time when the reaction is terminated
and the activities of substrates and products attain the equilibrium values. Then

Δr G = 0 and

Δr G0 = − RTlnK (8:7)

In this equation, K is equilibrium constant, ΔEeq = 0, and, according to eq. (8.5)

RT
ΔE0 = lnK (8:8)
nF

The cell, in which the overall reaction: Red1 + Ox2 ! Ox1 + Red2 has not reached equi-
librium state, and Δr G of reaction is lower than 0 ðΔr G < 0Þ, operates in galvanic
mode, producing electrical energy during the spontaneous discharging of the cell. The
chemical reaction drives the electron through the external circuit (load, device) con-
necting the two electrodes (Fig. 8.1(b)).
Maximum amount of energy that can be produced depends on the thermody-
namic driving force, meaning the equilibrium potential of the cell ΔEeq . In practice,
for the cell working in galvanic mode, the terms open circuit potential (OCP) or dis-
charging cell potential Ud are often used. Note that they show the potential difference
between the two electrodes (half-cells) and do not have thermodynamic meaning. As
was described earlier, the cell working in the galvanic mode converts the chemical
energy only into electrical energy.
The third mode of the electrochemical cell is the electrolytic mode. In this mode,
the flow of current is forced through the cell by an external source of potential (volt-
age). In this case, the anode in galvanic mode becomes cathode in the electrolytic cell
(reduction reaction) and cathode becomes anode (oxidation reaction; see Fig. 8.1(c)).
In the electrolytic mode, the electrical energy from an external source is converted
into the chemical energy and is stored in the reaction products at the electrode – the
cell is being charged and works as an electrochemical energy storage. It should be
mentioned that the cell operates in electrolytic mode only when the applied voltage is
larger than that of the equilibrium potential of this cell.
Understanding how the electrochemical cells work gives us the necessary back-
ground for understanding how the electrochemical sources of energy (EPS – electro-
chemical power sources) function, how to characterize them, and what parameters are
important for comparison and for the decision of further application and improve-
ments. Similar to the simple electrochemical cell, the EPS can provide the electrical en-
ergy while working in galvanic mode discharging themselves (primary batteries, fuel
cells), and subsequently, in the case of secondary batteries [1, 2], the EPS can be re-
 8.1 Batteries 179

charged by an external voltage source, to store the electrical energy in the form of
chemical energy for future use.

8.1.2 Characteristics of batteries

The main parameter that should be considered before construction and practical ap-
plication of a battery is its intrinsically available work (energy) of a chemical reaction.
The maximum energy available for conversion to electrical work is equal to Δr G and
is given by eq. (8.3). Based on this equation, one can conclude that the intrinsic maxi-
mum efficiency εmax of the electrochemical converter working reversibly is equal to
100%. However, the better expressions determining the intrinsic maximum efficiency
of electrochemical converter are:

Δr G Δr G
εmax = = (8:9)
Δr H Δr G + TΔr S

or
ΔEeq
εmax =  (8:10)
∂ΔEeq
ΔEeq − T ∂T

∂ΔEeq
∂T is the temperature coefficient of cell.
Equation (8.10) points out that the theoretical intrinsic efficiency can depend on the
sign of the temperature coefficient of a cell and may be lower or higher than 1. When
this coefficient is positive, the working cell takes some amount of heat from the sur-
roundings (equal TΔS when working reversibly) and changes it to work.
When the cell undergoes discharge, the decrease of the starting potential
(ΔEeq , OCP, Ud ) is observed. The degree of potential changes is described by the po-
U
tential efficiency, εp = ΔEdeq . For the characterization of the degree of energy conversion,
the current efficiency is also considered. In many reactions the current efficiency is
equal to 1; it means that all substrates are fully converted to products, and so there is no
alternative reaction consuming the electrons.
During work in galvanic mode, the cell is characterized by the dependence of the
discharge potential Ud upon flowing current (discharge currentÞ, Id :
X
Ud = Id Rex = ΔEeq − jηj − Id Rin (8:11a)

Ud = Id Rex = ΔEeq − ðηaC, + ηcC + ηaA + ηcA Þ − IRin (8:11b)

Rex is external resistance, Rin is internal resistance of a cell, usually the electrolyte
P
resistance and other resistances resulting from the cell design, jηj − means the sum
of concentration (ƞc,C, ƞc,A) and activation overpotentials (ƞa,C, ƞa,A) at both electrodes:
cathode ( C) and anode ( A) (see Chapter 1).
180 8 Electrochemistry in energy conversion and storage

Based on such dependences, the power delivered by a cell during its discharge
can be calculated using the equation: P = Ud Id . The common plots Ud versus Id and
P versus Id are presented in Fig. 8.2. It is clearly seen from the equation and the
plots that the power output of a battery is low when discharging current Id is low, but
the power is also low at the highest value of current, as a result of Ud drop. Convenient
parameters that can be determined from such plots are admissible discharge current
Iadm , maximum admissible power Padm , and cut-off potential – the minimum available
voltage. The battery cannot operate above Iadm or below cut-off potential.

Ud P (W)

Pmax
Padm
Cut-off
voltage

Iadm Id

Fig. 8.2: Voltage (Ud Þ–current (Id ) curve during discharging and power (P) dependency on discharging
current.

Considering Equation (8.11b) and taking into account the expressions determining the ac-
tivation and concentration overpotentials of cathode and anode (see Chapter 1), this equa-
tion can be written as follows:


RT Id =AC RT Id =AC RT Id =AA
Ud = Id Rex = ΔEeq − In + ln 1 − + ln
αnF j0, C nF jL, C βnF j0, A


RT Id =AA
+ ln 1 − − Id Rin (8:12)
nF jL, A

j0 is the exchange current density, jL is the limiting current density, A is the electrode
surface, indices A, C mean anode and cathode, and α and β are the transfer coeffi-
cients, different for anode and cathode, respectively.
This equation points out how the hindrance in electrode reactions and the inter-
nal resistance of the cell affect the shape of potential – discharging current plot
(Fig. 8.3(a, b, c)).
The curve (1) in Fig. 8.3(a, b, c) presents the idealized Ud Versus Id dependence. The
shape of this curve is changed by the large activation overpotentials, meaning the low
value of exchange currents (curve 2, Fig. 8.3(a)) The changes of the curve (1) in Fig. 8.3(b)
 8.1 Batteries 181

(a) (b) (c)

Ud (V) Ud (V) Ud (V)
–1.2 –1.2 –1.2
1 1 1
2 2
2

Id(A) Id(A) Id(A)

Fig. 8.3: Influence of different factors on idealized Ud − Id dependence.

Idealized behavior – curve 1 (a,b,c), the influence of large activation overpotential – curve 2 (scheme, a),
influence of large concentration overpotential – curve 2 (scheme, b), and the influence of large internal
resistance – curve 2 (scheme c).

result from the large values of concentration overpotentials and low values of limiting
currents. Figure 8.3(c) presents the influence of the internal resistance of the cell on the
shape of the curve (1). The results of calculations and resulting plots can be found in [1].
During discharging the dependence of discharge voltage, Ud , upon discharging time td is
also recorded. The typical discharging curve for a primary battery is shown in Fig. 8.4.
The usefulness of battery can be judged by the length of plateau being long for good
battery (changes of Ud are very small for many hours under given load), while for worse
battery the time of reaching cut-off potential is short.

Ud (V)
OCP

Cut-off
voltage

Useful
Useful time
time

Time (h)

Fig. 8.4: Comparison of discharging curves for good and worse batteries.

As was mentioned earlier, during the process of charging the secondary batteries, the
electrical energy is converted into chemical energy and stored in the reaction prod-
ucts. In analogy to Ud , the charging potential Uch is given by eq. (8.13), where Ich
means the current flowing during the charging process:
182 8 Electrochemistry in energy conversion and storage

X
Uch = Ich Rex = ΔEeq + jηj − Ich Rin (8:13)

Comparing Eqs. (8.13) with (8.11a) it is clearly seen that the charging of battery de-
mands more electrical energy than one can obtain during discharging. This additional
energy is needed to overcome the overpotentials of backward electrode reactions. The
example of discharging and charging curves for a secondary battery are shown in
Fig. 8.5.

U (V)
2.4
Charging
2.0 Discharging
1.8

Fig 8.5: An example of discharging and charging curves of

secondary battery. Note that the charging of battery demands
Time (h) more electrical energy than we can obtain during discharging.

Another parameter that can be determined from Id versus td curves is the total charge
Qd = Id td delivered by the battery during full discharge. It is called the capacity (Ah,
1 Ah = 3,600 C) of a battery. The product of capacity and potential (voltage) determines
the total energy (J, Wh) delivered by a battery. Since the capacity of the battery depends
upon a magnitude of external resistance, cut-off voltage, and temperature, the total en-
ergy is not a characteristic quantity. The practical quantities that give us the possibility to
compare batteries are as follows: specific energy – the energy delivered by unit mass of
battery (Wh/kg) and specific power (W/kg) – the power delivered by a unit mass of bat-
tery in a brief period of time. The Ragone diagram, the dependence of specific power
upon specific energy, is presented in Fig. 8.6. From this diagram plots, it is clearly seen
that some batteries that can provide high power do not provide a high amount of energy
per unit weight and vice versa.
Specific power (W/kg)

Fig. 8.6: Dependence of specific power of battery on its specific

Specific energy (Wh/kg) energy. The Ragone diagram.
 8.1 Batteries 183

The other important parameters (“life” parameters) for primary and secondary batteries
are as follows: the rate of self-discharge, shelf life, and time of useful work (time of full
discharging and the number of charge-discharge cycles). Self-discharge of the battery is a
result of additional reactions that are going in a battery, like corrosion, redox reactions of
electrolyte impurities, and so on. The rate of self-discharge of battery influences the shelf
life – the maximum time between production and utilization of battery. The number of
charge–discharge cycles determines how many times the battery can be charged/dis-
charged before losing its usefulness for practical applications.

8.1.3 Classification of batteries and examples

Batteries, as was shown above, operate like single electrochemical cells. However, to
meet the operational requirements and conditions, their constructions are much more
complicated. When the voltage of the single battery is not high enough to run the devi-
ces, the number of batteries is connected in series and works as a unit. Below we briefly
have described some examples of different kind of batteries: primary batteries, second-
ary (rechargeable) batteries, and fuel cells, focusing mainly on secondary batteries and
fuel cells. For much more detailed information about fundamentals, constructions, and
applications of different types of batteries, interested readers will find in books, reviews,
and references therein [1–11].

8.1.3.1 Primary batteries

Primary batteries often called single-discharge batteries contain a finite amount of ac-
tive reagents. When these reactants are consumed to some level during the battery
discharge, such a battery cannot be used again. They produced the electrical energy
when a load (external resistance, a device) is connected between the two electrodes of
a cell, and discharging reaction takes place. For quite a long time in the past, the
zinc–carbon battery has been used as a major primary battery, first patented in 1868
by G.L. Leclanché ( Leclanché battery). A more modern version of this battery con-
tains a metallic Zn body working as an anode. The cathode is made from a carbon rod
(collector of electrons) in direct contact with paste containing MnO2, mixed with car-
bon powder. The electrolyte is NH4Cl or ZnCl2 or a mixture of both, with a small
amount of water, in the form of a paste.
The electrode processes occurring in zinc–carbon batteries Zn|NH4Cl, MnO2|C
are quite complicated, and in a rough approximation, the discharging reaction, which
is the source of electrical energy, can be written as

Zn + 2 MnO2 + 2 H2 O ! ZnðOHÞ2 + 2 MnOðOHÞ

In some books, the other overall reaction is proposed:

184 8 Electrochemistry in energy conversion and storage

Zn + 2 MnO2 + 2 NH4 Cl ! ZnðNH3 Þ2 Cl2 + 2 MnOOH

Freshly manufactured batteries have the OCP of about 1.55 V. These batteries have
some disadvantages: the OCP strongly depends on the composition of electrode and
electrolyte, which decreases during discharging at the shelf storage. The specific en-
ergy (∼70–80 Wh/kg) is relatively low. Taking advantage of the construction and of
the electrode materials utilized in the Leclanché cell, much more advanced battery
was made, the so-called alkaline manganese dioxide battery: Zn|27%–40% KOH aq|
MnO2|C. In this battery, Zn is in the form of a paste with a high surface area, and so
the specific energy is much higher than in the case of Leclanche battery. The manga-
nese dioxide can be replaced by other oxides (HgO, Ag2O). All these batteries called
“alkaline” are successors of the original Leclanche battery. The main advantage of
modern Leclanche batteries is their low costs, and so they are widely used for simple
applications as power devices.

8.1.3.2 Secondary batteries

Secondary batteries (rechargeable) after their discharge can be recharged by forcing
a flow of electric current in the opposite direction. This should regenerate the dis-
charging products to the original substrates. Secondary batteries do not only produce
the electrical energy during discharging, but also can store this energy in a form of
chemical energy. (products of the electrochemical reaction that proceeds during
charging). The efficiency of energy storage in rechargeable batteries is high (about
60–90%) and properly manufactured batteries can be charged and discharged thou-
sands of times (charging/discharging cycles) without any substantial loss of their
properties.
Currently, the most widely used secondary battery is the lead–acid battery, first in-
vented by G. Planté in 1859. The original battery contained two lead plates separated by
special materials and immersed in sulfuric acid. The layer of lead dioxide was electro-
deposited on Pb during the first charging cycle. Further development of this battery led
to the changes in electrode construction and composition, resulting in substantial mass
decrease and much better performance [5, 6]. The present lead–acid battery contains
Pb anode in the form of small Pb particles deposited onto the Pb–Sb grid and lead diox-
ide cathode (active material) deposited onto Pb (collector of electrons) screen. As the
electrolyte, sulfuric acid of various concentrations is used. The acid anions are involved
in electrode reactions, and so the OCP of batteries depends on acid concentration and is
equal to 2.15 V at acid concentration 5.2 M (25° C) and decreases during discharging
(anions are consumed in the reactions).
The scheme of lead–acid battery can be written:

PbjPbSO4 , H2 SO4 ð25 − 40%Þ, PbO2 jPb

at such concentration, sulfuric acid practically exists in form of H+ and HSO4− ions.
 8.1 Batteries 185

The reactions during discharging at the anode and cathode are as follows:

anode ðnegative electrodeÞ: Pb + HSO4 − ! PbSO4 + H + + 2 e

cathode ðpositive electrodeÞ: PbO2 + HSO4 − + 3H + + 2e ! PbSO4 + 2H2 O

The reactions during charging of lead–acid battery from an external source of poten-
tial are as follows:

cathode ðnegative electrodeÞ: PbSO4 + H + + 2e ! Pb + HSO4 −

anode ðpositive electrodeÞ: PbSO4 + 2 H2 O ! PbO2 + HSO4 − + 3 H + + 2 e

Overall reaction in this battery is

Pb + PbO2 + 2 HSO4 − + 2 H + $ 2 PbSO4 + 2 H2 O

During discharge, the dense layer of lead sulfate is deposited on electrodes (Pb, PbO2),
resulting in the passivation of electrode active materials and in a decrease of discharge
voltage. The other effect influencing the battery performance is a decrease of sulfuric
acid concentration to about 12–24% during a discharge. This is due to water forma-
tion and causes a decrease in electrolyte density, compromising the specific energy of
acid–lead battery. The theoretical specific energy is 171 Wh/kg, but in practice, the spe-
cific energy is much lower and does not exceed 40 Wh/kg. The lifetime of acid–lead
battery allows for ca. 100–500 discharging–charging cycles. The specific energy is really
low, but for application as a car battery, it is sufficient. The main advantage of
lead–acid batteries lay in its low cost compared to other available secondary batteries.
Other rechargeable batteries include nickel–cadmium batteries or nickel–iron
batteries. They contain Cd (Fe) anode and NiO(OH) cathode, its active material being
pressed in the pores of a sintered nickel plate (metallic Ni is a collector of electrons).
Concentrated KOH solution is used as an electrolyte. The scheme of Ni–Cd battery is
Cd| 9M KOHaq| NiO(OH). The cell reactions are as follows:

anode: Cd + 2 OH − ! CdðOHÞ2 + 2 e, discharging

cathode: 2 NiOOH + 2 H2 O + 2 e ! 2 NiðOHÞ2 + 2 OH − , discharging

cathode: CdðOHÞ2 + 2 e ! Cd + 2 OH − , charging

anode: 2 NiðOHÞ2 + 2 OH − ! 2 NiOOH + 2 H2 O + 2 e, charging

The overall reaction in this battery is

Cd + 2 NiOOH + 2 H2 O $ CdðOHÞ2 + 2 NiðOHÞ2

The average OCP (cell voltage) is 1.29 V, and the theoretical energy density is 208 Wh/
kg. The main advantage of Ni–Cd as compared to lead–acid batteries is their longer
cycle life – about 3,500 cycles, in comparison with 500 cycles of lead–acid battery. The
Ni–Cd batteries can last 5–10 years depending on their configuration and applications,
186 8 Electrochemistry in energy conversion and storage

whereas the lead–acid batteries do not last more than 5 years. As the cost of Ni–Cd
batteries decreased over the last years, they replaced the alkaline primary batteries.
However, because of cadmium toxicity, the Ni–Cd battery will be probably replaced
by nickel–metal hydride (NiMeH) batteries, which have the same potential character-
istics but higher capacity. The overall reaction occurring in NiMeH cell is

MeHðsÞ + NiOOHðsÞ $ MeðsÞ + NiðOHÞ2 ðsÞ

As a cathode, like in Cd–Ni batteries, the nickel oxide hydroxide NiOOH is used. The
anode (instead of Cd) consists of hydrogen-absorbing alloy – an intermetallic com-
pound. The compositions of the alloys are patented but typically two components are
used. The one which possesses a high affinity to hydrogen incorporation in its struc-
ture is a mixture of La with rare earth elements or Zr and Ti, whereas the second
component is the metal that forms weak hydrides (Ni, Co, Al). NiMH battery contains
a 30% solution of potassium hydroxide as the electrolyte, like in Cd–Ni cell. A fully
charged battery has the cell voltage of ca. 1.25 V, the energy density in the range
40–110 Wh/kg, and the capacity of such battery is much higher than that of Cd–Ni
battery.

8.1.3.3 Lithium batteries

The huge development of portable electronics results in the need for a new type of
batteries that have higher voltage, larger capacity, and can be conveniently formed in
various sizes and shapes. Numerous experiments were carried out on applications of
alkali metal in cells because of their high negative potential in the electrochemical
scale. Lithium has the most negative value of electrode potential, so the application of
Li /Li+ as an anode in the batteries automatically increases the OCP of lithium cell to
4.5 V. This value is two times higher than for batteries described above. However, Li
is very active chemically and decomposes water; therefore, in lithium cell, only apro-
tic organic solvents can be used. The special cleaning of solvent and lithium salt used
as a source of lithium ions increases the cost of lithium cells. The lithium salts and
aprotic organic solvents widely applied in lithium cells are: LiBr, LiAsF6, LiPF6, AC
(acetonitrile), DMF (dimethylformamide), EC (ethylene carbonate), and PC (propylene
carbonate). To choose proper salt and solvent for lithium cells, one should consider
the conductivity of the solution and electrochemical stability of the solvent. The elec-
trochemical stability of the solvent is defined by the potential range in which the sol-
vent does not decompose. Li is aggressive metal and after the first contact with an
aprotic solvent, the passivation of Li occurs resulting from the formation of thin film
on the electrode surface. This film, on one hand, protects the Li anode against corro-
sion and, on the other hand, it causes a decrease of potential in the first phase of bat-
tery discharging. The lithium batteries can be designed and work as primary or
secondary batteries. The examples of primary lithium batteries include batteries with
 8.1 Batteries 187

liquid cathode such as Li-SOCl2 battery, Li-SO2 battery, or batteries with solid cathode
(metal oxide, sulfide).
The scheme of Li-thionyl chloride cell is: Li|Li AlCl4, SOCl2|C, chloride thionyl
works in cell as liquid cathode and also as a solvent for Li salt. In such types of cells
instead of thionyl chloride, sulfuryl chloride or phosphoryl chloride can also be used
as a liquid cathode. The reactions that occur during discharging of the cell are as
follows:

anode ðnegative electrodeÞ: Li ! Li + + e

cathode ðpositive electrodeÞ: 2 SOCl2 + 4 e ! SO2 + S + 4 Cl −

and the overall reactionis : 4 Li + 2 SOCl2 ! SO2 + S + 4 LiCl

Theoretical reversible potential and specific energy of the cell were estimated as 3.5 V
and 1.48 × 103 Wh=kg, respectively; however, in the commercial cells, the value of spe-
cific energy of ca. 700 Wh/kg was reached. Similar batteries containing SO2 were man-
ufactured. The schemes of these cells are: Li|SO2, LiBr, AN|C, and the reactions of
discharging of these cells can be written: 2Li + 2SO2 ! Li2 S2 O4 . These batteries are
characterized by good value of energy density – 333 Wh/kg, stable discharging poten-
tial (2.8–2.9 V) even at high discharging currents. Many of lithium primary cells have
solid cathode of metal oxides, metal sulfides, and oxy-salts. Compounds such as CuO,
Bi2O3, MnO2, CuS, FeS, Ag, Cu oxy-salts were tested in lithium batteries. Very popular
is the cell with Ag2CrO4 cathode:

LijLiClO4 , PCjAg2 CrO4 , C

The discharging reaction is

2 Li + Ag2 CrO4 ! 2 Ag + Li2 CrO4

The OCP of this battery is about 3 V and its specific energy is ca. 200 Wh/kg or even
higher. The battery is manufactured in a button shape and is widely used to power
the pacemakers.
In 2019 year, John B. Goodenough, M. Stanley Whittigham, and Akira Yosino re-
ceived the Nobel prize for their contribution to the development of rechargeable lith-
ium-ion battery. Contrary to the primary Li batteries, in the secondary (rechargeable)
batteries no metallic Li is used. Graphite intercalated with Li ions serves as anode in
these batteries, whereas cobalt oxide, manganese oxide, or other oxides intercalated
with Li ions are used as cathodes. The electrolyte is the solution of lithium perchlorate
or lithium hexafluorarsenate, LiAlF6, in mixed organic solvents. The scheme of an ex-
emplary cell is as follows:

CjLiClO4 , ECjMn2 O4 , EC − ethylene carbonate,

overall reaction in the cell is:

188 8 Electrochemistry in energy conversion and storage

LiC6 + Mn2 O4 $ 6 C + LiMn2 O4

As may be seen, lithium ions do not participate in overall reaction but take place in
partial electrode reactions:

anode: LiC6 ! Li + ðsolvÞ + 6 C + e

cathode: Li + ðsolvÞ + Mn2 O4 + e ! LiMn2 O4

The Li-ion battery operates in a voltage range from about 4.0 to 2.5 V, and its specific
energy is about 150 Wh/kg. Some disadvantages are connected with relatively unattrac-
tive shelf life and safety conditions. The information about the last developments and
new challenges in the lithium batteries field can be found in the bibliography [7–9].

8.1.3.4 Fuel cells

A fuel cell is an electrochemical device, which, unlike to other devices, can be continu-
ously fed with substrates, and the reaction products are also removed continuously. A
fuel cell consists of an electrolyte sandwiched in some matrix (membrane) between
thin porous anode and cathode.
Fuel cells emerged as promising green substitutes for fossil fuels. They are unique
in terms of the variety of fuels and feedstocks for energy conversion into electricity.
The potential to reduce overall energy consumption, carbon emission, and high energy
density makes these cells suitable for a wide range of applications, across multiple sec-
tors, including transportation, industrial, commercial, residential buildings, and long-
term energy storage for the grid in reversible systems. In the last few years, research
interest in fuel cells has been boosted by the growth in green energy requirements.
In the most known hydrogen–oxygen fuel cell, hydrogen is delivered to the anode
where the oxidation reaction takes place, and the oxygen is reduced at the cathode. In
the fuel cell, hydrogen gas (or hydrogen-rich fuels) and oxygen (pure or from the air)
are directly converted into electrical energy and heat through the electrochemical reac-
tion of water formation. The reactions that take place in the cell are as follows:

anode: H2 ! 2 H + + 2 e,
cathode: 1=2 O2 + H2 O + 2 e ! 2 OH −

The overall reaction is

H2 + 1=2 O2 ! H2 O, Δr G0 = − 237 kJ=mol, E0 = 1.229V

Depending on the electrolyte used in hydrogen–oxygen cells, we can distinguish:

1. Alkaline fuel cells (AFC), containing 30–70% solution of KOH. The development of
alkali fuel cells started in the thirties of the twentieth century when the British sci-
entist Francis T. Bacon demonstrated the first fuel cell, now the so-called Bacon
 8.1 Batteries 189

cell. About 20 years later, the stack of Bacon cells was demonstrated, the technology
was patented, and stacks were applied in Apollo spacecraft. The electrical power
was 5 kW.
In the modern hydrogen–oxygen fuel cell, gas-diffusion electrodes are used,
consisting of a porous layer of carbon black and of catalytic layer. Different cata-
lysts, such as carbon, manganese, nickel mesh, and so on, are tested and applied.
The working temperature of cells is in the range from 60 to 240° C. One of the
main problems with AFC is the degradation of electrodes by carbonate, formed in
the reaction: CO2 + 2 OH– ↔ CO32− + H2O (carbon dioxide is the contaminator of
gas reactants used). The crystals of carbonate (K2CO3) block the pores of gas-
diffusion electrodes, affecting the performance of the cell. To avoid this prob-
lem, stable solid alkaline polymers are applied – mainly polymers containing
methyl-ammonium or methyl-pyridine groups. Such solid-state cell can work at
temperatures up to 120 °C.
2. Phosphoric acid fuel cells (PAFC) contain 85–95% solution of phosphoric acid as the
electrolyte. The acid fills the silicon carbide matrix. The reactions at the anode, cath-
ode, and the overall reaction are the same as in other hydrogen–oxygen cells. The
reactions take place on highly dispersed catalyst Pt or Pt alloys with Cr, Ti on porous
carbon. The operating temperature of the cell is in the range 150–200 °C.
3. Membrane fuel cell (MFC) – in these cells, the polymeric ion-exchange membrane
as an electrolyte is used. These fuel cells are also named the solid polymer fuel cell
or proton-exchange membrane fuel cell (PEMFC). Modern PEMFCs are built of mem-
brane electrode assemblies (MEA), which include the electrodes, solid electrolyte,
catalyst, and gas diffusion layers pressed together. The Nafion polymer is mainly
used as a solid electrolyte. This polymer contains sulfonic acid groups − SO3H, and
serves as an acid electrolyte, a separator, and a cation-exchange membrane. As sul-
fonic acid is a strong acid, the sulfonic group dissociates, and therefore the proton
conductivity of the membrane is high (0.1 S/cm at 80 °C). For the same reason, the
membrane allows for the passage of protons but stops the passage of other ions.
The MEA is made as follows: an ink of catalyst suspension (mainly Pt or Pt–Ru),
carbon, and electrode material are sprayed or painted onto the solid electrolyte,
and then the carbon paper is hot pressed on either side to protect the inside of the
cell, acting also as electrodes. Carbon paper can also play the role of a gas diffusion
layer. Their operating range of temperatures is 60–100 °C. Splitting of the hydrogen
molecules is relatively easy by using a Pt catalyst. Unfortunately, splitting the oxy-
gen molecule is more difficult and this causes significant electric losses, about 20%
in the efficiency of energy conversion. Difficulties in the reduction process can
lead to a two-electron reduction of oxygen to hydrogen peroxide, instead of four-
electron reduction to water or OH– (Chapter 4). Up till now, there is a huge effort
for finding the appropriate catalysts for this process. The scheme of MFC is shown
in Fig. 8.7.
190 8 Electrochemistry in energy conversion and storage

Rex

Anode Cathode

Fuel gas H2 O2 Air (O2)

(H2) + H+
H2 H O2
H+ H+
H2 H+ +
O2
H+ H
H2 H+ O2
H+ H+
H2 H+
H+
Unused H 2O Unused(O2)
H2
fuel gas H2O
(H2)

Anodic gas Cathodic gas

diffusion layer Anodic Cathodic diffusion layer
catalyst layer catalyst layer
Proton-exchange membrane

Fig. 8.7: Scheme of membrane fuel cell.

As we mentioned at the beginning of this chapter, a double slash ||, appearing in the
schemes of the cells, represents the phase boundary whose potential is negligible as a
result of the application of a separator between the two electrolytes. This is theory; in
real cells and fuel cells operating under galvanic or electrolytic conditions, the pres-
ence of a separator, being necessary to separate physically (but not electrically) the
contents of the anodic and cathodic compartments (as it is shown in Fig. 8.7 in the
case of proton exchange membrane (PEM)), introduces an additional component to
the Ud, (eq. 8.12), diminishing it (or increasing the Uch). Therefore, the design of such a
separator is a key factor in the construction of high-performance fuel cells. The most
desired properties of membraneous separators are:
– high permeability for counterions, impermeable to coions;
– low electrical resistance – the permeability of an ion-exchange membrane for the
counterions under the driving force of an electrical potential gradient should be
as high as possible;
– good mechanical stability – mechanically strong, should have a low degree of swell-
ing, shrinking or embrittlement, and cracking during the long-term operation;
– high chemical stability – the membrane should be stable over the entire operation.

These properties determine to a large extent the technical feasibility and the econom-
ics of the MFCs.
The basic processes and mechanisms involved in the functioning of the ion-ex-
change membranes were briefly described in Chapter 1.4 of this book.
 8.1 Batteries 191

Ion-exchange membranes can be classified by their function as separators:

– cation-exchange membranes (CEM) that contain fixed negatively charged ions and
have a selective permeability for cations. A special case of such membranes utilized
widely in the MFC is the proton exchange membrane (PEM). With their develop-
ment, PEM fuel cells emerged as promising substitutes for fossil fuels with great
potential to reduce overall energy consumption, zero carbon emission, and high en-
ergy density. In the last few years, research interest in PEM fuel cells has been
boosted by the growth of green energy requirements for the transportation sector
and mobile energy frameworks. However, the commercialization of PEM fuel cells
suffers from two major hurdles, that is, high cost and low durability,
– anion-exchange membranes (AEM) that contain fixed positively charged ions and
which have a selective permeability for anions. Certain reports prove the suitabil-
ity of AEM compared to CEM, where high performance can be related to reduced
proton accumulation and pH splitting.

More complex structures involve amphoteric ion-exchange membranes in which both

negatively and positively fixed ionic groups are randomly distributed, bipolar mem-
branes consisting of a cation- and an anion-exchange membrane laminated together,
and mosaic ion-exchange membranes which are composed of macroscopic domains
of negatively fixed ions and those of positively fixed ions, randomly distributed in a
polymer matrix.
Another classification is based on the structure and basic material with the repre-
sentative example being a solid oxide fuel cell, operating at high temperatures and
mentioned briefly below.
All these fuel cells operate in moderate temperatures range from 60 to 240 °C. At
higher temperatures, molten-carbonate (MCFC) and solid oxide fuel cells (SOFC) are
used. MCFC contains a molten mixture of carbonates (sodium, potassium, lithium) as
an electrolyte; the operating temperature of such a cell is about 650 °C. Solid oxide
fuel cell with the solid electrolyte – zirconium oxide doped with other metal oxides
can work at temperatures up to 1,000 °C.
For the last few decades, a huge development in science and technology in designing
fuel cell is observed, as a result of wide applications in power plants, automotive trans-
portation, space vessels, and so on. There is a strong interest in the development of fuel
cells working at temperatures < 100 °C, like the PEMFC described above, or direct metha-
nol fuel cell (DMFC), where hydrogen is replaced by liquid methanol. They are very
promising cells and leading candidates for replacing the alkaline batteries. DMFCs,
like PEMFCs, are built of MEAs that include the electrodes, proton-exchange membrane,
and the porous carbon layers. The fuels are methanol solution in water (concentration
between 2 and 5 M/L) and oxygen (from air). The electrodes in DMFC consist of noble
metal catalysts, Pt–Ru alloy is used as the anode, and Pt as the cathode. Reactions pro-
ceeding during discharging of DMFC are as follows:
192 8 Electrochemistry in energy conversion and storage

anode: CH3 OH + H2 O ! CO2 + 6 H + + 6 e

cathode: 3=2 O2 + 6 H + + 6 e ! 3 H2 O

The overall reaction:

CH3 OH + 3=2 O2 ! CO2 + 2 H2 O

Δr G0 = − 706.2 kJ=mol, Eeq 0 = 1.22V

The main disadvantages of DMF cells are the processes called “methanol crossover” and
slow electro-oxidation of methanol at the anode. The crossover process means the pene-
tration of methanol across membrane separator to the cathode compartment, resulting in
losses of methanol and in decrease of the working potential of the cell. The decrease in
the rate of methanol electro-oxidation is caused by the formation of various intermedi-
ates (CO, HCO, HCOO), poisoning the catalyst surface. However, the cells possess a num-
ber of advantages, such as easy transport, storage, low cost of methanol, and the
possibility of compact design, small size, and weight that make them suitable for future
potential applications. At present, the challenge for scientists is to find a new type of cata-
lysts that will yield higher values of exchange currents not only for electro-oxidation of
methanol but also for electro-oxidation of other organic compounds like hydrazine, for-
mic acid, and ethanol. There is a huge literature concerning fuel cells; a few you will find
in the bibliography [10, 11].

8.2 Supercapacitors

8.2.1 General remarks

Nowadays, it is difficult to imagine the everyday life without mobile phones, tablets, or
laptops – in fact any kind of small, portable electronic devices. All of them require fre-
quent charging, boosting the development of novel, sustainable, and efficient energy
storage systems, fulfilling sometimes also demands of high power and energy of per-
sonal and commercial transportation systems. It is well known that all these require-
ments can be met by designing devices with high density of stored energy and power,
large number of charging/discharging cycles, capable of miniaturization and compati-
ble with modern electronics. Supercapacitors perfectly fulfill the above requirements
[12–15]. Therefore, they find a wide range of applications of high-power sources in
short time intervals, for example, for start/stop systems in electrical and hybrid ve-
hicles, but also as energy sources for medical devices like defibrillators (e.g., AED) or
pacemakers [16–18]. The latter, however, does not require high power, but a backup
is needed for short-time powering, before battery replacement. Recently, due to a
strong dependence of solar cell operation on weather conditions, as well as fluctua-
tion in solar radiation, a new approach, based on supercapacitors, was proposed to
 8.2 Supercapacitors 193

provide uninterrupted use of solar energy-generated electricity. Usually, such unin-

terrupted supply was realized by secondary batteries connected in parallel to the
solar panel. Replacing them with supercapacitors provides the advantage of long-
term stability and fast dynamics upon repeated charging/discharging cycles, relatively
low energy loss, and effective buffering of voltage fluctuations in solar radiations
[19–22].

8.2.2 Capacitor versus electrochemical cell/battery

A capacitor (condenser) is a passive electronic device, with two conducting terminals,

separated by a dielectric, that stores electrical energy. The relation between the charge
Q stored between the conducting terminals (electrodes) and the potential developed
across the dielectric, U, defines the capacitance C of the capacitor:

Q
C= (8:14)
U

The unit determining the capacitance is Farad [F; 1 F = 1 C/1 V] relating the charge
accumulated within the capacitor and potential developed between the conducting
terminals.
The physical form and construction of capacitors vary widely, depending upon
their applications as many types of capacitor are in common use. Most capacitors con-
tain metallic surfaces as conducting terminals separated by a dielectric medium. The
conducting terminal can be in a form of foil or thin film or an electrolyte in the case
of electrolytic capacitors. The nonconducting materials commonly used as dielectrics
include ceramic, polymer film, paper, mica, air, and oxide layers. The relationship be-
tween the electric field intensity, E, developed within the dielectric, and the potential
across this dielectric, U, is given by
ðd
U= EðxÞdx (8:15)
0

The energy stored within the capacitor is equivalent to the electric work of charge
separation within the dielectric, Wel, and therefore is given by
ð ð
Q Q2 CU 2 QU
Wel = UdQ = dQ = = = (8:16)
C 2C 2 2
As we discussed earlier, the energy stored in the electrochemical systems (cells, bat-
teries) is due to the faradaic reactions that are characterized by changes in Gibbs free
energy, ΔrG. The maximum electrical energy that can be produced due to these reac-
tions is (eq. 8.3)
194 8 Electrochemistry in energy conversion and storage

wmax = Δr G = − nFEeq (8:17)

Please note that the units of both U and Eeq are Volts, but while Eeq is defined by the
equilibrium potentials of redox electrodes, U is the potential difference between the po-
larized conducting terminals and not the reversible redox reactions. Now, keeping in
mind that the product nF is the charge that can flow through the cell as a result of fara-
daic reaction, it is easily seen that the energy stored in the electrochemical cell is twice
that of the capacitor charged with the same amount of Q = n F, with potential U equal
to the zero-current cell potential Eeq:

wmax = 2Wel (8:18)

So, if the energy stored in the conventional capacitor is half that of the energy stored in
the electrochemical cell, then what are the advantages of practical use of capacitors or
supercapacitors over batteries or cells as power supplies in nowadays electronics? First
of all, batteries and cells are characterized by totally different mechanism of charge
storage. This mechanism is based on the conversion of chemical into electrical energy
as a result of electrochemical reactions on the electrodes (anode and cathode). Some-
times, it is also accompanied by the processes of insertion/disinsertion of electroactive
ions within the whole electrode material, as in the case of lithium batteries. All these
mechanisms result in relatively slow charging/discharging processes. Limited kinetics
of the electrode reactions additionally determines the relatively low power of batteries
and electrochemical cells. In the case of capacitors, however, the mechanism of charge
storage is different and depends upon the dielectric and electrode materials used. The
charge is electrostatically stored between the conducting material (electrodes) upon the
external polarization. The overall capacitance C of, for example, parallel plate capacitor,
scales proportionally to the electrode area, A, and dielectric permittivity, ε, of the mate-
rial between the electrodes, and inversely to the distance between the conducting
plates, d. Consequently, the dielectric capacitance of such capacitor is small. However, if
we use for the “plates” of conducting materials of large specific surface area, for exam-
ple, porous or nanostructure conductive electrodes separated by a few nanometers of
electrolyte, then we will design the so-called supercapacitor, where the charge is
stored within the electrical double layers at the interfaces of the electrolyte and the
electrodes [23]. The performance difference between a dielectric and supercapacitor is
considerable. Typical specific capacitance of large capacitors is up to 15–50 µF/cm2, but
a supercapacitor has been shown to have capacitance as high as 1,500 F/g. Figure 8.8)
summarizes graphically the differences between the three types of energy storage devi-
ces: dielectric capacitor, supercapacitor, and battery.
The overall construction scheme of a supercapacitor relies on the two nano-
structured or porous electrodes in contact with current collectors. The electrodes
are separated with a thin membrane saturated with concentrated electrolyte solu-
tion. The membrane enables free movement of ions, but protects against the elec-
 8.2 Supercapacitors 195

Capacitor Supercapacitor Battery

Dielectric Electrolyte Electrolyte

d Electroactive
d
materials

Fast Fast Slow

Long lifetime Long lifetime Short lifetime
Low energy Medium energy High energy

Fig. 8.8: Graphical representation of differences between the three types of energy storage devices:
dielectric capacitor, supercapacitor, and battery.

trode short-circuit. Depending on the mode of electrical energy storage, there are
different types of supercapacitors:
– EDLC supercapacitors (electrical double layer capacitors) where the energy is
stored via the reversible adsorption of ions. This process is purely electrostatic.
– Pseudocapacitors, in which the electrical energy is additionally stored with help
of fast redox reactions localized on the surfaces of the electrodes.

If both electrodes are of the same material, then we have the symmetric system, but, of
course, they can be made of different conducting materials, yielding the asymmetric
supercapacitors, where each electrode has different specific capacitance.
In the case of EDLC supercapacitors, apart from the electrode area, their perfor-
mance depends on the surface concentration of adsorbing ions, and therefore the
charging/discharging cycle of this device depends on ionic mobilities and solvent re-
orientation/reorganization. Moreover, they are simple to load and their lifetime is
very long with extremely high number of charging/discharging cycles. Since there are
no electrochemical reactions, the potential and charge accumulated within the super-
capacitor are not limited by these reactions (Fig. 8.9).
For pseudocapacitors, their charge storage capability is additionally enhanced by
nanostructured transition metal mixed oxide materials, such as RuO2 or MnO2. Other
oxides include TiO2, VO2, MoO2, NiO2, CoO2, SnO2, and LiO2. In addition, thin layers of
electroactive polymers or polymers with metallic centers embedded within (such as
metallocenes) can be included in the design of pseudocapacitors. Interesting structures
include also polyoxometallates of W, Nb, and Ta.
Pseudocapacitance (or capacitance redox) appears when the adsorbed ions of
mixed valency can undergo fast and reversible surface redox reactions with no diffu-
sion involved. In acidic media, such reactions can be written as follows:
196 8 Electrochemistry in energy conversion and storage

Charging Charged supercapacitor Discharging

Electrolyte

Electrode Electrode Electrode Electrode Electrode Electrode

Electrolyte Electrical Electrolyte
double layer

IRint

Fig. 8.9: EDLC supercapacitor: charging, storage, and discharging processes. Below the middle capacitor
is its electrical equivalent circuit.

RuOx ðOHÞy + δH + + δe − $ RuOx − δ ðOHÞy + δ

MnIV O2 + xe − + xH + $ Hx MnIII IV
x Mn1 − x O2

The stoichiometric coefficients describe only the surface layer, because the overall re-
actions of the mixed oxide layers are more complex. In the case of nanostructured,
ruthenium-based oxides, the specific capacitance can reach up to 900 F/g; however,
these materials are quite expensive and toxic. Less costly are manganese-based mixed
oxides, but they give lower specific capacitance of ca. 200 F/g.
In certain cases, these processes can be accompanied by intercalation of cations
within the mixed oxide matrix. For instance,

MnIV O2 + e − + H + $ MnIII OOH intercalation=reduction

Or, more generally (C+ stands for a cation):

MnIV O2 + e − +C + $ MnIII OOC intercalation=reduction

MnIV O2 + e − +C + $ MnIII OO − C + surface adsorption=reduction

Supercapacitors have a very long life cycle of almost one million charging/discharging
cycles with cycle efficiency of about 95%, and are charged and discharged to the maxi-
mum limit in seconds. They have a high specific power of up to 15 kW/Kg as compared
to, for example, lithium ion battery (up to 2 kW/Kg). As in the case of cells and batter-
ies, specific energy and power of supercapacitors are most frequently defined for unit
mass. However, contrary to batteries and cells, this is typically the mass of electrodes,
because they are the major contribution to the total mass. Due to their design, super-
capacitors are also characterized by low internal resistance, Rint, which allows for
high currents to be supplied by such device. This resistance depends on the ionic con-
ductivity of electrolyte, ion diffusion within the porous structure of electrodes and
electrical contact between the grains of electrode material.
 8.2 Supercapacitors 197

Electrode materials used for the construction of supercapacitors should be of large

surface (even though of small external dimensions), of high conductance, thermally, me-
chanically, and chemically stable and of low costs and ease of manufacturing. Nanostruc-
tured, conducting materials are of primary use, such as carbon materials (graphene,
nanotubes, activated carbon, carbon fiber-cloth, carbide-derived carbon, carbon aerogel,
graphite, nanofibers), conducting polymers, transition metal oxides, polyoxometallates,
their mixtures, and the like. Similar requirements should be fulfilled by the electrolyte
filling the space within the porous structure of supercapacitors. For this purpose aqueous
electrolytes are of lesser use because of their relatively narrow potential window of up
to ca. 1.2 V in the neutral pH. Other types of electrolytes include nonaqueous, organic
electrolytes, ionic liquids, solid gel, and polymer electrolytes [24–29].
Table 8.1 compares the major performance parameters for ECDL capacitors, pseu-
docapacitors, lithium ion battery, and hybrid systems. It is easily seen that although the
specific power supply of ECDL and pseudocapacitors is high, their energy density is ca.
10% of a lithium ion battery. On the other hand, a battery, even though it stores more
specific energy, has limited number of charging/discharging cycles, shorter lifetime, lim-
ited power delivery, and significant impact on the environment. It is obvious then that
there should exist a device combining features of rechargeable batteries and supercapa-
citors. These are the hybrid structures in which a supercapacitor is coupled with a bat-
tery, forming a kind of “super battery” [27–29]. This hybrid structure provides more
effective energy supply in connection with faster charging. Their performance is pre-
sented in the last column (Hybrid) of Table 8.1.

Table 8.1: Comparison of general performance parameters of different energy storage systems.

Parameter ECDL Pseudocapacitor Li/ion battery Hybrid

Potential /V .–. .– .–. .–

Charging/Discharging cycles – – – –

Capacitance /F .– – – –

Specific energy .–. – – –

Wh/kg

Specific power kW/kg – – .– –

Lifetime /years – – – –

Self-discharge weeks weeks months months

Even better comparison of various energy storage systems currently available in the
market is shown in a form of Ragone diagram discussed earlier (Fig. 8.10).
It is clearly seen from this plot that charge storage devices that could fulfill all
needs of current electronics, having high specific energy as well as high specific
power output, are yet to be designed.
198 8 Electrochemistry in energy conversion and storage

103 h
10 h 1h 0.1
Energy density (Wh/kg)
102 Lithium 36
h
NiCads

101 Lead-acid 3.6

h
L
EDC citors
rc a pa h
100 e
sup ybrids 0.3
6
h

10–1 ms
Aluminum 36
electrolytics
10–2
101 102 103 104
Power density (Wh/kg)

Fig. 8.10: The Ragone diagram of specific energy density versus specific power density of energy storage
and conversion devices. It is worth to note that the electrolytic capacitor was developed in 1896 by Polish
engineer, Karol Pollak.

8.2.3 Working characteristics of supercapacitors

The relationship between the capacitance of, for example, a parallel plate capacitor
and its geometry can be described by
εε0
C= A (8:19)
d

where ε0 and ε are the dielectric permittivities of vacuum and dielectric (inside the
capacitor), respectively, A is the area of current collector, and d is the distance be-
tween the two plate collectors. This distance is typically in a nanometer scale and is
limited by the thickness of the adsorbed ions in the e.d.l. at the electrode/electrolyte
interface. The area A should be very large due to the nanoporous or nanostructured
surface of the electrodes of the supercapacitor. As it is evident from the above equa-
tion, a decrease of the distance d and an increase of the area A results in an increase
of capacitance C. These two parameters determine the conducting materials that can
be used for the design of supercapacitors. The third parameter that affects the charge
storage capability of supercapacitor (and capacitors in general) is the dielectric per-
mittivity, ε, of electrolyte between the electrodes/collectors.
Even in the case of symmetric supercapacitors, due to the differences in size of
cations and anions in the electrolyte, both electrodes can have different capacitances.
This is shown in the middle panel in Fig. 8.9. Then the overall capacitance C is related
to capacitances of both electrodes according to the equation:
1 1 1
= + (8:20)
C C+ C−
 8.2 Supercapacitors 199

where C is the total capacitance of supercapacitor [F], and C+, C− are the capacitances
of + and – electrodes, respectively [F]. According to the above equation, the electrode
of smaller capacitance will determine the overall system capacitance.
As we discussed above (eq. 8.16) the total electrical energy, Wel, accumulated
within the supercapacitor is directly related to its capacitance, C, and working poten-
tial, U:

CU 2
Wel = (8:21)
2
If we express the value of Wel per unit mass (or volume) of the electrode material, we
will get the specific energy.
Maximum specific power depends also on the working potential U:

U2
Pmax = (8:22)
4Rint
where Rint is the internal resistance described above. As in the case of specific energy
it can be expressed per unit mass or unit volume.
Another important parameter determining the performance of supercapacitors
as charge storage and energy supply systems is their rate of charging and discharging.
This is described by the charging/discharging time constant, τ [s]:

τ = Rint · C (8:23)

Its values, shown in the Ragone diagram as skewed, blue lines, describe the dynamics
of charging/discharging of electrical power sources and for the case of supercapacitors
are very short, in the range of 0.1 to 20 seconds. It is worth to note that although the
fast charging rate is very advantageous for a charge storage device, the fast discharge
characteristics is disadvantageous from the point of view of supporting relatively con-
stant potential over an extended period of time. Therefore, supercapacitors find their
primary use where the large charge has to be delivered over a short period of time. The
potential-time discharge characteristics of battery and supercapacitor are shown in
Fig. 8.11. Please keep in mind that these curves are intended solely to show the different
shapes of these two characteristics, but neither the potential nor time axes are compa-
rable for both systems.
Nevertheless, since supercapacitors can be constructed at low cost practically in
any size, they find extensive use as short-time power support or for applications
where the buffering of fluctuating loads are required, such as hand-held devices,
laptops, photographic flashes, to name a few. On a larger scale, they can stabilize
voltage and buffer unwanted power fluctuations on the grids. They are also suitable
as temporary energy harvesting and storage systems [17, 18, 22]. Low internal resis-
tance of supercapacitors triggered the applications requiring short-term high cur-
rents, such as in the case of engine start-up [30]. It is worth to note that the first
200 8 Electrochemistry in energy conversion and storage

U (V)
Battery discharging

Electrochemical supercapacitor
discharging

t (s)

Fig. 8.11: Discharging curves for a battery and electrochemical supercapacitor. See text for details.

hybrid bus in Europe, partially utilizing supercapacitors, was constructed and put
into operation in 2012 under the nickname Ultracapbus in Nuremberg, Germany.

Bibliography

[1] Bocris J O`M, Reddy AKN. Modern electrochemistry. Fundamentals of electrodics. V 2B. New York,
USA, Kluwer Academic Publisher, 2000.
[2] Bagotsky VS. Fundamentals of electrochemistry. Hoboken. USA. J. Willey & Sons Inc, 2006.
[3] Linden D, Reedy TB. Handbook of batteries (3 ed). New York. USA. Mc. Graw-Hill, 2002.
[4] Fuller TF, Harb JN. Battery fundamentals. In Electrochemical Engineering. Hoboken. USA. J. Willey&
Sons Inc, 2018.
[5] Davidson A, Monahov B. Lead batteries for utility energy storage: a review. J. Energy Storage 2018,
15, 145–157.
[6] Lach J, Wróbel K, Wróbel J, Podsadni P, Czerwinski A. Application of carbon in lead-acid batteries a
review. J. Solid State Electrochemistry 2019, 23, 23693–705.
[7] Etacheri V, Marom R, Elezari R, Salitra G, Aurbach D. Challenges in the development of advanced Li-
ion batteries: a review. Energy & Environmental Science 2011, 4, 3243.
[8] Schipper F, Ericson EM, Erk Ch, Chesneau FF, Aurbach D. Review – recent advances and remaining
challenges for Lithium ion battery. Nickel rich, LiNixCoyMnzO2. J Electrochem Soc 2017, 164, A 6620–A
6628.
[9] Heidari EK, Kamyabi-Gol A, Sohi HM, Ataie A. Electrode materials for lithium ion batteries: a review.
Journal of Ultrafine Grained and Nanostructured Materials 2018, 1, 1–12.
[10] Carrette L, Friedrich K, Stimming U. Fuel cells-fundamentals and applications. Fuel Cells 2001, 1,
5–39.
[11] O`Hayre R, Suk-Wan Cha, Collela W, Prinz FB. Fuel cells. Fundamentals. Hoboken. USA. J. Willey&
Sons Inc, 2016.
[12] Zhang Y, Feng H, Wu X, Zhang A, Xia T, Dong X, Li X, Zhang L. Progress of electrochemical capacitor
electrode materials: A review. Int. J. Hydrogen Energy 2009, 34 4889–4899.
[13] Gonzalez A, Goikolea E, Barrena JA, Mysyk R. Review on supercapacitors: Technologies and
materials. Renewable & Sustainable Energy Reviews. 2016, 58, 1189–1206
[14] Raza W, Ali FZ, Raza N, Luo YW, Kim KH, Yang JH, Kumar S, Mehmood A, Kwon EE. Recent
advancements in supercapacitor technology. NanoEnergy 2018, 52, 441–473.
 Bibliography 201

[15] Poonam SK, Arora A, Tripathi SK. Review on supercapacitors: Materials and devices. J. Energy
Storage 2019, 21, 801–825.
[16] Slesinski A, Fic K, Frackowiak E. New trends in electrochemical supercapacitors. In: VanEldik R., Macyk
W. eds. Materials for sustainable energy. Book Series: Advances in Inorganic Chemistry. Elsevier Acad.
Press Inc., San Diego CA, USA . . . 2018, 72, 247–286
[17] Huang Y, Zhu MS, Huang Y, Pei ZH, Li HF, Wang ZF, Xue Q, Zhi CY. Multifunctional energy storage
and conversion devices. Adv. Mat. 2016, 28, 8344–8364
[18] Baptista JM, Sagu JS, Wijayantha UKG, Lobato K. State-of-the-art materials for high power and high
energy supercapacitors. Performance metrics and obstacles for the transition from lab to industrial
scale – A critical approach.
[19] Meng H, Pang S, Cui G. Photo-supercapacitors based on third-generation solar cells. ChemSusChem
2019, 12, 3431–3447
[20] Xu X, Li S, Zhang H, Shen Y, Zakeeruddin SM, Graetzel M, Cheng YB, Wang M. A Power Pack Based
on Organometallic Perovskite Solar Cell and Supercapacitor. ACS Nano 2015, 9, 1782–1787.
[21] Yang PH, Xiao X, Li YZ, Ding Y, Qiang PF, Tan XH, Mai WJ, Lin ZY, Wu WZ, Li TQ, Jin HY, Liu PY, Zhou J,
Wong CP, Wang ZL. Hydrogenated ZnO core-shell nanocables for flexible supercapacitors and self-
powered systems. ACS Nano 2013, 7, 2617–2626.
[22] Lau D, Song N, Hall C, Jiang Y, Lim S, Perez-Wurfl I, Ouyang Z, Lennon A. Hybrid solar energy
harvesting and storage devices: The promises and challenges. Materials Today Energy 2019, 25,
UNSP 100852.
[23] Fuller TF, Harb JN. Electrochemical double-layer capacitors. In: Electrochemical Engineering, Wiley &
Sons, Inc. 2018, pp.251–273.
[24] Pal B, Yang S, Ramesh S, Thangadurai V, Jose R. Electrolyte selection for supercapacitive devices: a
critical review. Nanoscale Advances. 2019, 1, 3807–3835.
[25] Choi H, Yoon H. Nanostructured electrode materials for electrochemical capacitor applications.
Nanomaterials 2015, 5, 906–936.
[26] Wang G, Zhang L, Zhang J. A review of electrode materials for electrochemical supercapacitors.
Chem. Soc. Rev. 2012, 41, 797–826.
[27] Muzaffar A, Ahamed MB, Deshmukh K, Thirumalai J. A review on recent advances in hybrid
supercapacitors: Design, fabrication and applications. Renewable & Sustainable Energy Reviews
2019, 101, 123–145.
[28] Dubal DP, Ayyad O, Ruiz V, Gomez-Romero P, Hybrid energy storage: the merging of battery and
supercapacitor chemistries. Chem. Soc. Rev., 2015, 44, 1777–1790.
[29] Xie J, Yang PP, Wang Y, Qi T, Lei Y, Li CM. Puzzles and confusions in supercapacitor and battery:
Theory and solutions. J. Power Sources. 2018, 401, 213–223.
[30] Kouchachvili L, Yaici W, Entchev E. Hybrid batter/supercapacitor Energy storage system for the
electric vehicles. J. Power Sources. 2018, 237–248.
9 Interfacing applied electrochemistry and biology

9.1 General remarks

The tight and direct relation between biology and current flow has been known since
the works of Luigi Galvani in his villa at the suburban hills of Bologna at ca. 1780. Then
he discovered that static electricity generator connected to the severed frog’s leg stimu-
lated its twitch. This discovery gave a new thrust to the understanding of the workings
of nervous system. The acknowledgment of such connection culminated in the famous
phrase by Albert Szent-Györgyi: “Life is nothing but an electron looking for a place to
rest” (Nobel Prize, Physiology or Medicine, 1937). And so, the direct transformation of
chemical energy into electrical energy via the electrochemical reactions involving trans-
fer of electrons through the biochemical pathways become the focus of bioelectrochem-
istry. In the sections that follow we will focus on applied electrochemistry that involves
enzymatic catalysis for anodic and cathodic processes defined previously in the text.
The main stream of research and applications of enzymatic catalysis focuses on
the electroactive enzymes – oxidoreductases, in search for the best, stable, and re-
producible constructs for an extended and low-cost, disposable use. Among the oxi-
doreductases, glucose oxidase (GOx), laccases, bilirubin oxidase (BOx), peroxidase,
tyrosinase, dehydrogenases, for example, cellobiose dehydrogenase, and catalases
play key roles. Their substrates can be oxygen, hydrogen peroxide, phenols, qui-
nones, sugars, lignins, and so on. Enzymes as biological electrocatalysts are unique
as reaction accelerators, because for some of them the number of elementary charge
transfer reactions occurring at the so-called active center of the enzyme reaches up
to 106/s. This shows their advantage over the inorganic catalysts currently used in
industry. Another advantage of enzymes as bioelectrocatalysts is their specificity in
the reactions of a single target compound or a very small group of compounds. More-
over, enzymes work at ambient temperatures and pH from slightly acidic (e.g., 4.5) to
slightly alkaline (e.g., 8.5), whereas commercially used inorganic catalysts require ele-
vated temperature, above 200 °C and, very often highly acidic media, posing potential
threat to the user and environment. The mechanism of catalytic activity of the enzymes
also differs them from that of inorganic catalysts. Even though the first step is always the
substrate binding, for the case of enzymes this step is already very selective and specific.
This is achieved by structural binding pockets, complementary to the particular sub-
strates with respect to shape, charge, and hydrophilic/hydrophobic balance. These fea-
tures allow the biocatalysts to distinguish between very similar molecules. The specificity
of enzymes was initially explained by the so-called lock and key model that assumed that
both the enzyme and its substrate must have specific complementary shape, so that the
substrate can fit exactly into the binding pocket. This model can explain the specificity of
enzyme–substrate interactions, but does not account for the energetic stabilization of the
transition state of the resulting complex. A modification of this rigid model – the induced

https://doi.org/10.1515/9783111160986-009
 9.2 Bioelectrocatalysis 203

fit model – acknowledges that both the enzyme and substrate are flexible and during
mutual interactions can reshape, and adapt to each other to the final precise positions
that enable the enzyme to catalytically convert its substrate into product. At this precise
position, the final shape and charge distribution of the enzyme–substrate complex is at-
tained with lower standard Gibbs free energy of activation, ΔG0# (see the Eyring–Polanyi
theory, eq. (1.92)) when compared with noncatalyzed reaction).
The sequence of the amino acids specifies the structure of the enzyme, thus
determining its catalytic activity; however, only a small part of enzyme structure
(ca. 2–6 amino acids) is directly involved in catalysis, forming the catalytic site. The cat-
alytic site in turn is located next to one or more binding sites with functionalities orient-
ing properly the substrates with respect to the catalytic sites. The catalytic and binding
sites together form the active site of the enzyme. The remaining majority of the enzyme
structure protects and maintains the precise orientation and dynamics of the active
site. In some enzymes, no amino acids are directly involved in catalysis; instead, the
enzyme contains sites to bind and orient catalytic cofactors. There are a large variety of
natural oxidoreductase inorganic and organic cofactors, such as metal ion centers,
iron–sulfur clusters, hemes, pyrolloquinoline quinone, nicotinamide adenine dinucleo-
tifde (NAD+), nicotinamide dinucleotide phosphate (NADP+), and flavin adenine dinucle-
otide (FAD). Their role is to change the oxidation state during the substrate catalysis.
“Wiring” these cofactors or enzymes to the electrode surface, either directly or with the
help of linker molecules or nanostructures, is a challenge that underlines the effective-
ness and performance of the biocatalysts either as biosensors or as electrodes in biofuel
cells (BFCs). Since the more detailed description and classification of enzymes is well
beyond the scope of this chapter, interested reader is directed instead to several excel-
lent literatures [1–5].

9.2 Bioelectrocatalysis

Bioelectrocatalysis, similarly to electrocatalysis, is a type of heterogeneous catalysis

that results in the acceleration of electrochemical reaction rate at the electrode/elec-
trolyte interface due to the biological catalysts: enzymes, or even whole cells [5]. The
applications of bioelectrocatalysis range from the preparations of electrodes for BFCs
to biosensors and bioelectronic devices [6].
Until recently, the enzymes have not been used in electrocatalysis for a number
of reasons, such as inherent instability of enzymes in artificial environment, on the
electrode surface in particular, and the absence of pure enzyme preparations on a
larger scale. However, due to the improvements in methods of immobilization of en-
zymes in their active forms on electrodes, as well as in the methods of enzyme isola-
tion and purification, the above-mentioned problems are largely reduced. Therefore,
the following advantages of enzymatic bioelectrocatalysis play a crucial role in its de-
velopment and applications:
204 9 Interfacing applied electrochemistry and biology

– Possibility of electrochemical reactions at ambient temperatures and neutral pH,

where no effective inorganic catalysts are currently available
– Possibility of electrochemistry of complex organic reactions with high efficiency
at moderate overpotentials and ambient conditions
– Possibility of utilization in BFCs
– Possibility of the development of highly selective bioelectrochemical sensors

Last, but not least, with the progressive depletion of natural sources of noble metals
as catalysts and rapid increase of their costs, the relatively cheap production of en-
zymes for modification of the electrode surface becomes more and more attractive,
particularly when the usage of enzymes for the above purposes is sufficient to merit
the large-scale production, decreasing further their costs.
Some examples of redox reactions that can be catalyzed by enzymes immobilized
on the electrode surface, largely exploited from the point of view of BFCs and biosen-
sors, are:

O2 + 4H + ! Laccase
2H2 O

This cathodic reaction can be catalyzed by laccases (as in the above reaction) or, for
example, BOx. The direct, four-electron reduction of oxygen to water is most favor-
able from the point of view of BFCs and is discussed in Section 8.1.
Peroxidases (e.g., horseradish peroxidase) catalyze the two-electron reduction of
oxygen, energetically less favorable cathodic reaction:

H2 O2 + 2H + + 2e !Peroxidase
2H2 O

Hydrogenases catalyze the reduction of protons:

!
Hydrogenase
2H + + 2e H2

There exist several approaches to bind the biocatalysts with the electrodes for appli-
cations in bioelectrocatalysis:
– Mechanical/physical entrapment into the nonconducting or conducting polymeric
network on the electrode
– Covalent binding of the enzyme to the modified surface, for example, via the car-
bodiimide procedure
– Enzyme cross-linking
– Noncovalent (e.g., electrostatic, affinity) binding to the electrode modified with
the appropriate functional groups
– Direct adsorption on the electrode.
 9.2 Bioelectrocatalysis 205

The last approach is less frequently used, because the enzymes tend to undergo dena-
turation in contact with metals, losing their activity. There are several conditions to
be fulfilled for successful immobilization of biocatalysts: they should be stable, enzy-
matic activity unaffected, and with proper orientation of the enzyme’s active center
toward the electrode – the latter feature indispensable for the case of the direct elec-
tron transfer (DET) between the enzyme and the electrode. The last condition is less
important for the case of mediated electron transfer (MET), where a redox couple – a
mediator – transfers the electrons between the active center and the electrode in con-
cert with enzymatic conversion of a substrate S into product P. Moreover, the access
of the substrate to the active center of the immobilized enzyme should not be compro-
mised [7–10].
The differences between DET and MET is shown in Fig. 9.1.

Direct electron transfer (DET) Mediated electron transfer (MET)

Substrate Product

Product –e
Substrate
enzyme Mediator +e Mediator
oxidized reduced
form form
Enzyme
Enzyme
active site Electron
e Regenerated
tunneling
oxidized form
distance
Electrode Electrode

Fig. 9.1: The differences between the direct electron transport (DET) and mediated electron transport
(MET) of an enzyme at, or in the vicinity of, the electrode surface.

In the direct bioelectrocatalysis (DET), the rate of catalyzed reaction is determined by

the activity of an enzyme and/or by the thermodynamics of the reaction; however, in
mediated bioelectrocatalysis (MET), apart from the enzymatic activity, it is the reaction
of mediator molecules that may screen the substrate–enzyme reaction of interest.
In order to assess the bioelectrocatalytic properties of enzyme-modified electrodes,
we will use the Michaelis–Menten model of enzymatic reaction. We have to keep in
mind, however, that this model was developed for free enzyme and substrate in the
solution, whereas here we have a surface-confined enzyme and such confinement af-
fects its accessibility by a substrate molecule. Moreover, it may affect the enzyme’s ac-
tivity, too. Therefore, the obtained kinetic parameters of such immobilized system
cannot be compared with the very same enzyme in the solution. Thus, now the scheme
of reaction can be written as follows:
k1 kcat
E + S Ð ES ! E + P
k−1
206 9 Interfacing applied electrochemistry and biology

where E and S are the enzyme and substrate, respectively, ES is the enzyme–substrate
complex, P is the product, k1 and k‒1 are the rate constants of complex formation and
dissociation, respectively, and kcat is the catalytic rate constant of product P formation.
Assuming a steady-state condition, that is, d½ES=dt = 0, and using kinetic equations:

d½ES
= k1 ½E½S − ½ESðk − 1 + kcat Þ = 0 (9:1)
dt

Defining the rate of the enzymatic reaction as v = ðd½P=dtÞ = −kcat ½ES, we can get
v = kcat ½E0  K ½S+ ½S, where kcat ½E0  defines the maximum rate of catalytic reaction, νmax .
M
Here, square brackets are used to denote concentrations and [E0] is the total concen-
tration of an enzyme ([E0] = [E] + [ES]). To find the maximum value of a biocatalytic
reaction, the substrate concentration is increased until a constant rate of product for-
mation is achieved. At the maximum reaction rate, all the enzyme active sites are oc-
cupied by the substrate, and the amount of ES complex is the same as the total
concentration of enzyme. The catalytic rate constant, kcat, is also known as the turn-
over number, describing the number of substrate molecules converted by one active
site of an enzyme per second. KM is the Michaelis–Menten constant:

k − 1 + kcat
KM = (9:2)
k1

The Michaelis–Menten constant characterizes the affinity of an enzyme to the sub-

strate; the lower the value of KM, the greater the enzyme affinity to the substrate.
Each enzyme has a characteristic KM for a given substrate. The efficiency of a biocata-
lyst is frequently expressed by a ratio of kcat/KM, called the specificity constant, reflect-
ing the affinity and catalytic capability of an enzyme toward a given substrate.
Therefore, it is useful for comparing catalytic properties of the same enzyme for dif-
ferent substrates or different enzymes with the same substrate. The theoretical maxi-
mum value for the specificity constant is limited by the diffusion rate of a substrate
toward the enzyme and can reach up to 109 [M–1s–1] for a free enzyme in the solution.
Since the electrocatalytic reaction of the immobilized enzyme requires the trans-
fer of electrons between the enzyme active center and the electrode (even in MET),
then, calling upon the Faraday’s law relating the rate of the reaction with charge
being transferred (eqs. (1.31) and (1.91)), we can finally get the catalytic current due to
the enzymatic reaction on the electrode:

½S
icat = nFAΓkcat (9:3)
½ S  + KM

where Γ is the surface concentration of the enzyme and A is the electrode area. This is
the electrochemical version of Michaelis–Menten equation relating the current measured
on the enzyme-modified electrode with the enzyme electrocatalytic properties: kcat and
KM. The maximum current, imax, at the plateau of icat versus [S] is equal to nFAΓkcat . This
 9.2 Bioelectrocatalysis 207

is a key equation used for amperometric biosensors. In a form of Lineweaver–Burk plot

(double reciprocal plot), it predicts the linearity of 1/icat versus 1/[S]:
1 1 KM 1
= + × (9:4)
icat nFAΓkcat nFAΓkcat ½S

By measuring current as a response to the controlled addition of substrate and plot-

ting its reciprocal 1/icat versus 1/[S], we can get all characteristics of biocatalyst-
modified electrode: imax, kcat, and KM. This is shown in Fig. 9.2.

1/icat

Slope
KM/nFAΓkcat

1/nFAΓKcat
–1/KM

1/[S]
0

Fig. 9.2: Lineweaver–Burk plot, showing characteristic features for retrieving the electrocatalytic
properties of enzyme-modified electrode.

Then, if needed, we can apply the Eyring–Polanyi theory (transition-state theory) for
the bioelectrocatalytic process, to account for the dependence of rate constants on the
energetics of the electron transfer rate (eq. (1.92)).
Above, we have described several methods of enzyme immobilization on the elec-
trodes for further use as anodes and cathodes for BFCs and/or biosensors. Now, we
shall briefly focus on materials that can be utilized as electrode support for biocata-
lyst immobilization. The proper selection is essential for the biocatalyst applications
[4–10]. These materials should fulfil several requirements:
– They should be inexpensive, reproducible, and with highly developed active sur-
face for immobilization and electron transfer.
– The immobilization procedure for such support should be simple and without sig-
nificant compromise of biocatalyst activity and specificity.
– Should be chemically stable in the analytical environment and inert to the en-
zyme, detected analytes, and mediator (for MET).
– If they contain functionalities mediating in the electron transfer, these functional-
ities should be reversible electrochemically.
– The conductivity of such materials should be high.
208 9 Interfacing applied electrochemistry and biology

Many efforts have been made to improve the properties of conducting supports for
enzyme deposition. The use of conducting porous and mesoporous materials as well
as nanoscale materials has been widely explored. Examples of such electrode materi-
als are metal or oxide nanoparticles, carbon nanowires, carbon nanotubes, modified
carbon nanotubes, carbon felt, carbon paper, carbon paste, graphene, graphite in
form of felt, woven felt, graphite cloth platinized, and/or polyaniline/modified polya-
niline conducting polymers [12, 13]. Table 9.1 explores some of the most interesting
solutions:

Table 9.1: Materials utilized for enzyme immobilization.

Material(s) for catalytic bioelectrode formation Immobilized biocatalyst Reference

Citric acid-coated superparamagnetic iron oxide nanoparticles Fructose dehydrogenase []

on gold electrode

Carbon paper disk coated with carbon nanotubes Laccase, bilirubin oxidase []

Carbon nanomaterials, nanofibers, nanotubes Laccase, bilirubin oxidase, []

cellobiose oxidase, glucose
oxidase

Metal nanoparticles (Au, Ag) tyrosinase, sensors  []

Au/Pt/Pd/TiONT/Ti

Hybrid, osmium redox copolymer/carbon Glucose oxidase, bilirubin []

oxidase

Chitosan, TiO/CeO/chitosan/glassy carbon electrode Tyrosinase []

The catalytic properties of oxidoreductases immobilized on conducting supports can

be used either in biosensors or BFCs.

9.3 Biosensors

Monitoring of phenolic compounds in the food industry and for environmental (e.g.,
to detect polyphenols) and medical applications (monitoring the levels of glucose or
neurotransmitters) has become increasingly relevant in recent years. Enzymatic bio-
sensors based on oxidoreductases represent a fast method for online and in situ
monitoring of these compounds. They rely mostly on the amperometric detection of:
– the product(s) of enzymatic reaction, or
– direct enzyme-MET between the electrode and the analyte.

The first type of detection is used, for example, in the case of re-reduction of phenols,
polyphenols, and neurotransmitters, oxidized by the enzyme–tyrosinase immobilized
 9.3 Biosensors 209

on the electrode. The presence of enzyme improves the selectivity versus the interfer-
ing molecules (ascorbic acid, uric acid, etc.). The direct detection is used when the en-
zyme transfers the electrons between the conducting support and the analyte, causing
its reduction or oxidation. This can be seen in the laccase biosensors “wired” to the
electrode surface and monitoring the oxygen level by its direct reduction to water.
For both cases, of course, the measured catalytic current should be proportional to
the analyte concentration.
Figure 9.3 presents a relatively simple biosensor layout consisting of a detection
layer of enzyme(s) immobilized onto a working transducer electrode. As discussed
earlier, enzymes are optimal catalysts/recognition elements because they provide ex-
cellent selectivity and specificity for their substrates and have high catalytic activity.
However, at the same time, enzymes are the shortest lived components of biosensors,
BFCs, because they gradually lose activity in time, even if unused, determining there-
fore the lifespan of the enzyme-based electronic devices.

Enzymatic
detection
layer
Analytes

icat
Signal

Enzymatic Transducer
detection (electrode)

Fig. 9.3: A typical design of the enzyme-coupled electrochemical biosensors.

During the sensing event, a controlled voltage is applied to the electrode inducing
redox reaction of the analyte, generating a signal that scales with the analyte’s con-
centration. In the potentiometric mode, a change in the electrode potential as a result
of redox reaction can also be used as the measurable response. Finally, an appropri-
ate electronic device (e.g., computerized potentiostat) collects and displays the resul-
tant signal. This type of electrochemical detection also offers additional selectivity of
biosensor by controlling the potential applied to the electrode, because different elec-
troactive molecules can be oxidized/reduced at different potentials. The overall system
is compatible with modern miniaturization trends and can be also made biocompatible
for implanted bioelectronics. Biocompatibility is often an important issue, since biologi-
cal fluids, such as blood, are the most common sample matrices for enzyme electrodes
210 9 Interfacing applied electrochemistry and biology

in clinical applications. The components of such matrices, mostly proteins, can affect or
even deteriorate the sensor performance by fouling its surface.
Apart from the above-mentioned food industry and environmental monitoring,
the electrochemical biosensors are most commonly used in health care, by monitoring
the glucose level in blood being the dominant application. Using a glucose biosensor
as an example, we will present the evolution of biosensors, based on the sensing
event and bioelectrochemistry of signal transduction between the electroactive en-
zyme and the electrode. This evolution is shown in Fig. 9.4.

Substrate Product Substrate Product Substrate Product

Enzyme Enzyme Enzyme

MedOX MedRed
O2 H2O2

Electrode Electrode Electrode

Fig. 9.4: Schematic representation of differences between the first through the third generation of
biosensor design and characteristics.

Figure 9.4(a) shows schematically the main characteristics of the so-called first-gener-
ation glucose biosensor: the trapped GOx, oxidizes β-D-glucose (S) to β-D-gluconolactone
(P), with a simultaneous reduction of GOx cofactor FAD to FADH2. FADH2 is subsequently
re-oxidized to FAD by its natural substrate – dissolved O2 producing H2O2. Appropriate
voltage applied to the electrode oxidizes H2O2 at the electrode surface, yielding an elec-
tric signal as a result of biosensing event. Unfortunately, the first generation of sensor
has several disadvantages and shortcomings. The active site and FAD are buried deeply
within the protein shell, hindering drastically the diffusion of reagents [11]. Additionally,
as we discussed earlier, the Marcus theory states that the electron transfer requires reor-
ganization energy, λ, and also decays exponentially with increasing distance. The active
sites of many oxidoreductases are hidden within the proteins. Moreover, O2 has a limited
solubility in aqueous media and is the reagent of unstable concentration (due to the fluc-
tuations in partial pressure in the air), leading to oxygen deficiency at higher glucose
concentrations affecting the sensor response. The oxidation of hydrogen peroxide as a
sensing event also proceeds at relatively high potential, leading to the simultaneous elec-
tro-oxidation of interfering species typically present in biomatrices, for example, ascor-
bic acid, uric acid, and even possible drugs such as paracetamol.
 9.3 Biosensors 211

Second-generation biosensor is shown schematically in Fig. 9.4(b). Its design solved

many of the first-generation biosensor problems by the incorporation of a redox medi-
ator molecule that shuttles the electrons between the enzyme and electrode. As stated
earlier, DET is not possible, so the role of mediator molecules is to facilitate the trans-
fer between deeply buried FAD redox center of GOx. In the mediator redox cycle, its
oxidized form MedOx regenerates FAD by oxidation of FADH2. The resultant reduced
form of mediator, MedRed, is re-oxidized at the electrode generating the electric signal.
Ideal mediator molecules should show fast and reversible redox kinetics with both re-
duced and oxidized forms being nontoxic, chemically stable, and ideally should main-
tain constant concentration within the biosensor, that is, they should not diffuse out of
the sensor construct. Mediators such as poly(vinylimidazole) and poly(vinylpyridine)
linked with osmium complexes, or Os+/2+ redox couples co-immobilized in a hydrogel
[18] serving as electron relays provided close proximity for the redox center of the pol-
ymers and the FAD redox center of the enzymes resulting in fast sensor response and
high current output [20]. This approach was reported also for immobilized dehydro-
genases with azine/polyazine relays between the electrode and NAD+/NADH [21, 22].
The use of synthetic mediators in second-generation biosensors eliminated also the
problems related to oxygen and allowed also for lower applied potentials. However,
the observed gradual decrease of sensors performance due to the steady leaching of
mediators to the surrounding solution stimulated the design of the third generation
of biosensors.
The design of third-generation sensors involves “wiring” of the enzyme to the mod-
ified transducer surface in such a way to provide a DET from the redox site to the elec-
trode or through the immobilized polymeric relays (Fig. 9.4(c)). This approach is
typically achieved by direct co-immobilization of the enzyme and mediator on the elec-
trode surface or within the conductive polymer matrix [23] (in the case of some peroxi-
dases having peripheral redox centers, a DET can be achieved by depositing the enzymes
directly on the electrode). The mediators cannot escape from the sensor and their concen-
tration effectively facilitates the transport of electrons between the enzyme active and
the electrode. Of particular interest is, of course, direct electrical contact between the en-
zyme and electrode by connecting the active center through the tailored molecular com-
munication path. Such configuration should greatly increase the charge transfer kinetics,
resulting in larger sensing currents. This design is also advantageous from the point of
view of BFCs, where the efficient electron transfer at the bioelectrodes is required. Other
enzyme and mediator immobilization methods include the layer-by-layer deposition of
polyelectrolytes, entrapment in hydrogels, and electropolymerization in the presence of
the enzyme and the mediator, such as carbon or osmium copolymer “wires” [24–26].
212 9 Interfacing applied electrochemistry and biology

9.4 Biofuel cells

Even though the concept of BFC dates back to the beginning of the past century, the
real progress in this area began in the 1960s with NASA space program and the
demand for alternative power sources. However, the initial results were poor and
the BFC concept was dormant until the energy crisis of the late 1970s. And currently,
due to the permanently increasing demand for cheap, renewable, and sustainable fuel,
the enzymatic BFCs represent an emerging advantageous technology that can generate
electrical energy from biologically renewable catalysts and fuels, addressing the above
issues. In principle, a BFC can be considered as a special type of an electrochemical fuel
cell, where inorganic catalysts (mainly metals, such as Pt, Pd, Ru, and Rh) are replaced
with biocatalysts in the form of microorganisms, organelles, or enzymes. In a conven-
tional fuel cell, an oxidation reaction occurs at the anode/solution interface and the re-
duction reaction takes place at the cathode/solution interface. Both reactions use
relatively simple inorganic chemistries (Chapter 8), releasing the electrons at the anode
that subsequently travel through the external circuit producing electrical energy that
can be used for different engineering purposes. However, those fuel cells that operate
in the so-called low-temperature regime (ca. 80–100 °C) typically require expensive me-
tallic p-group catalysts. Contrary to the above, the BFCs use enzymes or whole cells as
catalysts and therefore they can operate under mild, physiological conditions: tem-
perature in the range of 20–40 °C and near-neutral pH. However, these conditions
are not their limitations, though advantageous from the point of view of BFC appli-
cations, because extremophilic organisms and enzymes are capable of operation
under a wide range of external conditions. In the sections that follow, we will briefly
describe the principles and basic concepts of enzymatic BFCs only, directing the
reader interested in microbial fuel cells to the review papers [27, 28]. We will con-
sider recent advances in mediated and DET-based BFCs capable of producing rela-
tively significant power while being quite stable for a long period of time. A wide
variety of redox enzymes have been tested and employed to create electrodes (both
cathodes and anodes) for use in BFCs as energy sources powering small electronic
devices, self-powered sensors, and implantable power sources. Here we will limit
our discussion to the oxidoreductases – a group of enzymes that catalyze oxidore-
duction reactions by transferring electrons from an oxidant to the reductant. Some
of these enzymes were presented in the preceding sections.
From the point of view of their applications and competition with fuel cells, the
requirements for successful BFCs can be summarized as follows:
– BFC should convert chemical energy into electrical with the help of low-cost,
easy-accessible, stable redox enzymes as catalysts.
– They should work under mild conditions, neutral pH, and ambient temperatures.
– Should use natural sources for fuel, such as ethanol, methanol, glucose, and oxygen.
– BFC should be renewable and cheap (biomass is cheap).
– The only products of their reaction should be water and carbon dioxide.
 9.4 Biofuel cells 213

Further comparison of both types of currently used fuel and BFCs is given in Table 9.2:

Table 9.2: Comparison of fuel cells and biofuel cells focusing on catalyst used
and operating conditions.

Fuel cell Biofuel cell

Catalyst Noble metals Microorganisms or enzymes

pH Highly acidic or basic Between . and .

Temperature Above  °C –

Efficiency (%) – 

Fuel H, methanol, O Carbohydrates, methanol, O, etc.

As any of the electrochemical cell, the BFC contains cathode and anode made out of
well-conducting material, most frequently carbonaceous, placed in separate compart-
ments – cathodic and anodic – typically separated by a semipermeable membrane to
avoid mixing of the electroactive species (fuel and biocatalysts) in the electrolyte. If
the biocatalysts are immobilized on the electrodes separately, there is no strict de-
mand for the use of separating membranes. There are some exemptions, of course,
and the most common is when GOx is used on the anode for oxidation of glucose and
oxygen used as fuel on cathode interferes with anodic processes or both cathodic and
anodic processes. For example, reactive oxygen species (ROS, such as O2.–), produced
by possible incomplete reduction of O2, can seriously affect the activity of enzymes.
However, it has been shown that suitably immobilized GOx gives an anodic current
(rate of the reaction) very close to the theoretical maximum [29], indicating that the
presence of oxygen in the anodic compartment does not affect the biocatalytic activity
of this enzyme. This BFC employed also cytochrome C oxidase immobilized on cathode.
The selectivity of the immobilized enzymes separated the reduction and oxidation reac-
tions. Such an approach eliminates issues of crossover. By removing the membrane sep-
arator, such design allows both electrodes to be as close to each other as possible,
decreasing the internal resistance of the cell without short-circuiting, and allowing for
further miniaturizations. This leads to subsequent research on improvement of the sys-
tem performance [open circuit potential (OCP), power output] to make it more econom-
ically effective as portable energy devices and compatible with current electronics and
also directing the technology toward the implantable systems [30].
Finally, the generalized schemes of both fuel and BFCs is shown in Fig. 9.5, exem-
plifying the latter in the case of glucose/oxygen BFC – the Holy Grail of current
research:
This scheme reveals advantageous feature of BFC discussed above: the fuel cell re-
quires the proton exchange membrane (PEM), with H+ as connecting ion, whereas BFC
214 9 Interfacing applied electrochemistry and biology

e e e e

C6H12O6
H2 PEM O2 fuel
fuel oxidant
C6H10O6

H + H2O O2
product
H2O
Anode Cathode
Anode Cathode

Noble metal catalyst

Biocatalysts

Fig. 9.5: Generalized schemes of both fuel and biofuel cells. Biofuel cell is shown as a
glucose/oxygen biofuel cell.

does not necessarily require any membrane separating the anodic compartment from
the cathodic one.

9.4.1 Design and enzymes used in biofuel cells

Obviously, in light of the above discussion, the enzymes that show high catalytic activ-
ity, the ability to DET, and low overvoltage of catalytic reactions are typically used in
BFC studies. Although DET is highly desirable for enzymatic BFCs, many of the currently
investigated systems rely on the MET, because of higher current densities that can be
achieved. The incorporation of organometallic redox mediators into the conducting
redox polymers pioneered by the Heller group [31 and all subsequent related works]
provided a means of “indirect wiring” of the enzymes. These redox mediators – osmium
complexes – mentioned earlier in the text, characterized by high stability, facile, and
reversible kinetics of the electron transfer, were connected with polymer chains, typi-
cally polyvinylpyrrolidone, polyacrylamide, or polyvinylimidazole. Osmium redox cen-
ters in such molecules by tuning the ligands can mediate cathodic or anodic reactions
and therefore were utilized for biocathodes and bioanodes. In summary, the engineer-
ing approaches to create well-performing bioelectrodes are mainly focused on tailoring
the properties of scaffolds for enzyme immobilization and improving the electron trans-
fer rates between enzymes and electrodes.

9.4.1.1 Biocatalysts and design of biocathodes

Multicopper oxidases (MCO) such as laccase (Lac) or BOx have been intensively inves-
tigated as bioelectrocatalysts for cathodic reduction of oxygen to water. Even though
 9.4 Biofuel cells 215

BOx shows better tolerance against chloride inhibition, the availability of laccases re-
sults in more frequent utilization of this enzyme for the design of cathode in BFC.
Laccases have oxidative activity toward ortho- and para-diphenols and therefore
are also used for biosensor design. However, for the use as a part of biocathode in
BFC, it is imperative that the enzyme takes four electrons necessary for the reduction
of oxygen to water only from the electrode material. It is instructive to follow the elec-
tron path inside this multicopper enzyme, from the electrode material to oxygen. The
catalytic mechanism of the laccase starts with the entry of an electron at the T1 copper
site, followed by an internal electron transfer from the reduced T1 to the T2 and T3
copper sites. The T3 copper in the presence of the T2 site functions as an acceptor in
the aerobic oxidation process. The reduction of oxygen to water takes place at the T2
and T3 clusters [32]. Now, we should recall the term “internal resistance” and all its
components that were discussed earlier in the case of power outputs and perfor-
mance of cells in general (Chapter 8). In the case of BFCs, the overpotentials of the
electron pathways within the biocatalyst and its immobilizing layer should be consid-
ered as part of this inner potential drop, diminishing the BFC performance, and as
such should be minimized by proper design.
Laccases have defined Nernst (equilibrium) redox potentials. However, some of
them are branded as “low” potential, because for T1 and T3 copper sites, this potential
falls within the range of approximately 0.4–0.5 V (vs NHE), whereas for the so-called
high potential laccases EN is approximately equal to 0.8 V. Regardless of these values,
for the case of T2 Cu, the equilibrium potential, Eeq, is approximately 0.4 V for both
types of laccases [32]. Of course, the higher the overall potential established on enzy-
matic cathode, the better from the point of view of the BFC performance, in particular
from the point of view of the OCP, discussed in Chapter 8.
The design and charge transfer mechanism of MCO biocathodes is directly related
to the technology of enzyme immobilization. The use of redox mediators (MET) im-
proves the current output and stability of BFCs; however, the redox mediators com-
monly used are of limited choice, due to the fact that they need to have similar redox
potential as the active site of immobilized enzymes, have small overpotential (good re-
versibility), should be nontoxic, biocompatible, and stable. Moreover, they should not
leak from the cathodic space. We addressed these issues earlier in the case of biosen-
sors. Due to the above limitations, other constructs are being developed, such as con-
ductive polymers that can serve as both the immobilization scaffold for the enzyme and
conducting wires between the copper centers and electrode. Lately, an extensive effort of
researchers has been directed toward the nanomaterials, such as carbon nanomaterials
and metallic (gold) and metal-oxide nanoparticles. These materials greatly improve the
catalytic performance of biocathodes, exhibiting high conductivity, good biocompatibility,
and enhancement of DET by addressing/approaching directly the electron entry catalytic
site of MCOs, that is, the T1 site of laccases. Among these nanomaterials, multiwalled car-
bon nanotubes, either surface-modified or pristine, play a crucial role. High surface area
216 9 Interfacing applied electrochemistry and biology

provides also a statistically improved orientation of MCOs at such modified cathodes, en-
hancing also this way the electron transfer from the active site to the electrode.
The generalized scheme of oxygen enzymatic (laccase) electroreduction can be
written in the following steps:
1. The formation of “active complex” laccase – oxygen in the presence of water
molecules
laccase + O2 + H2 O ! laccase − HO2 OH
2. Transfer of two electrons from the electrode to the complex

laccase − HO2 OH + 2e + 2H + ! laccaseðOHÞ2 + H2 O

3. Transfer of the next two electrons followed by dissociation

laccaseðOHÞ2 + 2e + 2H + ! laccase + 2H2 O

This scheme also provides an explanation of the pH dependence of laccases and BOx:
they require a slightly acidic medium with the highest activity at approximately pH
5.5 (laccase).

9.4.1.2 Biocatalysts and design of bioanodes

On the other part of the BFC shown in Fig. 9.5 (right panel), the anode is designed to oxi-
dize glucose. Most of the oxidoreductase enzymes used for the design of anodes contain
redox cofactors changing their oxidation state during substrate catalysis. One of such co-
factors is nicotinamide adenine dinucleotide phosphate NAD(P)+, in NAD(P)-dependent
dehydrogenases. This is a diffusional mediator being reduced to NAD(P)H upon oxidation
of fuel (sugar). Subsequently, it has to be re-oxidized/regenerated at the electrode. How-
ever, NAD(P)H tends to quickly passivate the electrode surface. Additionally, it has large
overpotential, between 0.5 and 0.9 V, depending on the electrode material, so the NAD(P)-
dependent dehydrogenases are not the best choice for anode design; nevertheless, they
are still commonly used because they are oxygen insensitive comparing to oxidase-based
bioanodes [30, 33]. Other bioanodes used for the construction of BFCs rely on flavopro-
tein oxidases, such as GOx or glucose dehydrogenase (GDH), carrying a flavin cofactor,
(FAD), tethered within the protein shell. The role of FAD, as in the case of NAD(P)+, is to
change its oxidation state during the anodic reaction of sugars. Now, the proposed elec-
trochemical oxidation of glucose can be written as follows:
In the case of GOx:

glucose + GOxðFADÞ ! gluconolactone + GOxðFADH2 Þ;

Regeneration at the electrode: GOxðFADH2 Þ ! GOxðFADÞ + 2H + + 2e

 9.4 Biofuel cells 217

In the case of GDH:

glucose + GDHðFADÞ ! gluconic acid + GDHðFADH2 Þ

Regeneration at the electrode: GDHðFADH2 Þ ! GDHðFADÞ + 2H + + 2e

The electrons then flow through the external circuit to the cathode, generating the
power output of BFC. It should be mentioned here that FAD utilizes O2 as an electron
acceptor, producing another ROS species – H2O2 – which can impart a negative effect
on both enzymes in the system. However, FAD-dependent GDH, GDH(FAD), appears to
be oxygen insensitive, unlike FAD-dependent GOx, GOx(FAD) [33, 34].
The engineering strategies to create efficient bioanodes, as in the case of bio-
cathodes, are mainly focused on tailoring the properties of scaffolds for enzyme
immobilization and improving the electron transfer rates between enzymes and
electrodes. Among such strategies, nanomaterials and carbon nanomaterials in par-
ticular that improve direct electron communication between the enzyme and elec-
trode emerge as crucial materials [30, 33–38].

9.4.2 Energy and power of the biofuel cells

Critical factor and challenge of BFCs in their applications as “green,” sustainable, and
renewable energy sources are their output voltages and power that are generally in-
compatible with the operational requirements of commercially available microelec-
tronic devices. Such devices are typically built to operate at a minimum of 1–3 V voltage
supply; therefore, BFC will need a step-up power system [39–41]. It is easy to evaluate
the magnitude of the equilibrium potential, ΔEeq, also called the OCP, at standard condi-
tions of the electrochemical reaction, using its relation with the standard Gibbs free en-
ergy, ΔrG0, derived previously for the general case (eq. (4.1)):

Δr G0 = − nFΔEeq
0
(9:5)

All symbols used here have their meaning defined previously.

In the example of glucose as an anodic fuel and oxygen as a cathodic one, we can
write:
1
C6 H12 O6 ðglucoseÞ + O2 ! C6 H10 O6 ðgluconolactoneÞ + H2 O (9:6)
2

At 25 °C and p = 1 bar, 0.1 M glucose concentration at pH 7, the modulus of ΔrG0 is

approximately 227.2 kJ/mol. Then, with n = 2 the value of ΔEeq (OCP) is equal to ap-
proximately 1.2 V. This is also the thermodynamically maximal value of OCP in an
aqueous media; otherwise, we begin to split water at higher potential values.
218 9 Interfacing applied electrochemistry and biology

As in the case of batteries and fuel cells, the typical polarization curve in galvanic
mode is a superposition of contributions due to bioanode and biocathode kinetic overpo-
tential regions, mass transport limitations, and ohmic losses (compare Fig. 8.2):
X X
Ud = Id Rex = ΔEeq − ηa, c, Cathode + ηa, c, Anode − Id Rin (9:7)

where Ud is the potential of BFC under a given load (Rex) biasing the cell from its equi-
librium potential value, ΔEeq, and inducing a discharge current flow Id (Fig. 9.6).
X
ηa, c, Cathode is the sum of activation and concentration overpotentials at the bioca-
X 
thode, while ηa, c, Anode describes the sum of activation and concentration overpo-
tentials at the bioanode. Rin is the internal BFC resistance of ionic transport within the
electrolyte, ohmic drop in the polymer scaffold of the enzyme on the electrode, if pres-
ent, but also the resistance of molecular “wires” transferring electrons between the
electrode and the enzyme.
Discharge potential Ud

Power
of BFC
OCP
Kinetic
Ohmic loss

Mass transport

Cell current Discharge

current Id
Short Maximum cell current
curcuit

Fig. 9.6: Schematic graph of the BFC discharge curve at decreasing load, Rex, and increasing the discharge
current, Id, identifying different mechanisms responsible. Each point of this discharge curve, Ud = f(Id),
where Ud and Id are the discharge potential and current, respectively, can be described as in the case of
battery (compare Fig. 8.2).

It is important to understand that in the case of BFCs, these overpotentials are directly
related to the electrochemical mechanisms of redox reactions of enzymes deposited
on the electrodes. For biocathodes, MCO-based bioelectrodes undergoing DET with
low overpotentials are widely adopted, whereas for bioanodes operating through the
cofactors, their low overpotentials determine the choice of the enzymes. Of course,
these should also correlate with possibly large ΔEeq . For a better understanding of basic
issues related to the performance of the BFC, let us split the BFC into anodic and ca-
thodic parts, and examine each of the electrodes separately. This is shown in Fig. 9.7.
As we have shown before, in an ideal case, the OCP is determined by the difference
between the thermodynamic equilibrium potentials between the cathodic reduction
and anodic oxidation reactions, Eeq,c and Eeq,a, respectively (for our BFC in Fig. 9.5,
 9.4 Biofuel cells 219

Electrocatalytic oxidation of biofuel at bioanode

Electrocatalytic
current
Overpotential Eeq,c
for onset of at the
fuel oxidation bioanode
OCP
Cell voltage E of BFC
Eeq,a Overpotential for
and current
at the onset of oxidant
giving maximum
bioanode power reduction

Electrocatalytic reduction
of oxygen (oxidant) at the biocathode

Fig. 9.7: Schematic graph of cathodic and anodic electrocatalytic current/voltage characteristics plotted
along the overall BFC voltage (cathodic current is negative). Green arrows show the value of current Id at
maximum power of BFC, Pmax, under a given load Rex.

these are the reduction of oxygen and oxidation of glucose), taking into account the
nonstandard conditions of the BFC. At very low discharge currents, the measured Ud is
determined by the difference between the onset potentials for catalysis on cathode and
anode, respectively (red arrow). Then, the Ud decreases with decreasing load, until the
maximum value of the product Ud∙Id is achieved, which is equal to the maximum
power of a given BFC, according to the definition of power: PBFC = Ud∙Id . Finally, with
still decreasing load Rex, we reach the so-called short-circuit conditions (Fig. 9.6), where
Rex = 0, and anode and cathode are directly connected electrically. Then, again as in the
case of the equilibrium state, no useful electrical work is done. Useful power can be
delivered by a cell, also by a BFC cell, at voltages between these two limiting cases. The
magnitude of current is determined by the electrode with the lowest electrocatalytic
rate (in Fig. 9.6, this is the reduction of oxygen at the cathode, as indicated by the lower
catalytic “saturation” current, limited by the slow diffusion and low concentration of
oxygen).
In contrast to the conventional fuel cells with noble metal catalysts, BFCs are of
very low power outputs. Typical power outputs of BFCs are in the range of up to milli-
watts per square centimeter of their geometric area (see Table 9.3), compared to the
kilowatt outputs from fuel cells. Therefore, their applications are limited to the wear-
able or implantable short-term devices in medicine or self-powered sensors, such as
glucose sensors, harvesting energy from biofuel inside the body, or from the environ-
ment that is monitored. These applications include self-powered portable electronic
devices for point-of-care measurements [41].
In order to even better visualize the challenges that have to be overcome for future
commercial utilization of BFC, let us present the graphical representation of power out-
puts for various electrochemical power sources [46]. Similarly to the Ragone diagram, it
shows the current status of research on performance of BFCs (Fig. 9.8).
220 9 Interfacing applied electrochemistry and biology

Table 9.3: Power outputs of several enzymatic BFCs.

Anodic enzyme Cathodic enzyme Maximum power, Ref.

Pmax (mW/cm)

GOx BOD . []

GDH Laccase . []

GOx Laccase . []

GOx Laccase . []

GOx BOD . []

Batteries

Solar cells

Redox
Power flow
Biosensors gap cells

Biofuel cells Fuel cells

10–12 10–9 10–6 10–3 100 103 106 109 P (W)

Fig. 9.8: A sketch of power output ranges of biofuel cells and other cells (including solar cells), illustrating
the power gap between the fuel and biofuel cell range, challenging further biofuel cell research for
practical applications (adapted after [46]).

However, one should take into account a tremendous development in this area,
and therefore, even at the stage of writing this book, the above diagram seems to
be obsolete (c.f. [27, 30, 33]). Due to the fact that both oxygen and glucose as bio-
fuels are quite abundant in the environment, the development of BFCs is still pro-
gressing and the introduction of new electrode materials, especially conducting
nanomaterials, constantly improves the performance of such power sources. Also,
modern miniaturization strategies can provide charge pump systems boosting the
final output voltage of BFCs to the levels compatible with current electronics [24,
30, 39, 40, 47].
 Bibliography 221

Bibliography

[1] Suzuki H. Active Site Structure. In: Suzuki H., ed. How Enzymes Work: From Structure to Function.
CRC Press, Taylor & Francis, 2015, pp. 117–140.
[2] Stryer L, Berg JM, Tymoczko JL. Biochemistry (5th ed.). San Francisco: W.H. Freeman.
2002.
[3] Petsko GA, Ringe D. Chapter 1: From sequence to structure. In: Lawrence E, Robertson M, eds.,
Protein structure and function. London: New Science Press Ltd. 2003, pp. 2–27.
[4] Boyer R. Enzymes I: Reactions, Kinetics, and Inhibition. In: Molloy K, ed., Concepts in Biochemistry.
3rd ed. John Wiley & Sons, Inc. 2006, pp. 131–162.
[5] Rusling JF, Wang B, Yun S. Electrochemistry of Redox enzymes. In: Bartlett PN. ed.,
Bioelectrochemistry. Fundamentals, Experimantal Techniques and Applications, John Wiley & Sons
Ltd. 2008, pp. 39–76.
[6] Masa J, Schuhmann W. Electrocatalysis and bioelectrocatalysis – Distinction without a difference.
Nano Energy. 2016, 29, 466–475.
[7] Tarasevich MR. Bioelectrocatalysis. In: Srinivasan S, et al., eds. Comprehensive Treatise of
Electrochemistry. Springer, 1985, vol. 10, pp. 231–295.
[8] Willner B, Willner I. Reconstituted redox enzymes on electrodes: From fundamental understanding
of electron transfer at functionalized electrode interfaces to biosensor and biofuel cell applications.
In: Willner I, Katz E, eds. Bioelectronics. From Theory to Applications, Wiley VCH, 2005, pp. 35–123.
[9] Leech D, Kavanagh P, Schuhmann W. Enzymatic fuel cells: Recent progress Electrochimica Acta.
2012, 84, 223–234.
[10] Putzbach W, Ronkainen NJ. Immobilization Techniques in the Fabrication of Nanomaterial-Based
Electrochemical Biosensors: A Review. Sensors. 2013, 13, 4811–4840.
[11] Tsujimura S. From fundamentals to applications of bioelectrocatalysis: bioelectrocatalytic reactions
of FAD-dependent glucose dehydrogenase and bilirubin oxidase. Biosci. Biotechnol. and Biochem.
2019, 83, 39–48.
[12] Jackowska K, Krysinski P. New trends in the electrochemical sensing of dopamine. Anal. Bioanal.
Chem. 2013, 405, 3753–3771.
[13] Florescu M, David M. Tyrosinase-Based Biosensors for Selective Dopamine Detection. Sensors. 2017,
17, 1314–1330.
[14] Kizling M, Rekorajska A, Krysinski P, Bilewicz R. Magnetic-field-induced orientation of fructose
dehydrogenase on iron oxide nanoparticles for enhanced direct electron transfer. Electrochem.
Comm. 2018, 93, 66–70.
[15] Majdecka D, Draminska S, Janusek D, Krysinski P, Bilewicz R. A self-powered biosensing device with
an integrated hybrid biofuel cell for intermittent monitoring of analytes. Biosens. Bioelectron. 2018,
102, 383–388.
[16] c.f., Bollella P, Ludwig R, Gorton L. Cellobiose dehydrogenase: Insights on the nanostructuration of
electrodes for improved development of biosensors and biofuel cells. Appl. Mat. Today. 2017, 9,
319–332.
[17] Mahshid S, Li C, Mahshid SS, Askari M, Dolati A, Yang L, Luo S, Cai Q. Sensitive determination of
dopamine in the presence of uric acid and ascorbic acid using TiO2 nanotubes modified with Pd, Pt
and Au nanoparticles analyst. 2011, 136, 2322–2329.
[18] Gregg BB, Heller A. Cross-linked redox gels containing glucose-oxidase for amperometric biosensor
applications. Anal. Chem. 1990, 62 258–263.
[19] Njagi J, Ispas C, Andreescu S. Mixed ceria-based metal oxides biosensor for operation in oxygen
restrictive environments. Anal. Chem. 2008, 80, 7266–7244.
222 9 Interfacing applied electrochemistry and biology

[20] Zhu Z, Garcia-Gancedo L, Flewitt AJ, Xie H, Moussy, F, Milne WI. A critical review of glucose
biosensors based on carbon nanomaterials: Carbon nanotubes and graphene. Sensors. 2012, 12,
5996–6022.
[21] Karyakin AA, Karyakina EE, Schmidt HL. Electropolymerized azines: A new group of electroactive
polymers. Electroanalysis. 1999, 11, 149–155.
[22] Karyakin AA, Ivanova YuN, Karyakina EE. Equilibrium (NAD(+)/NADH) potential on poly(Neutral Red)
modified electrode. Electrochem. Commun. 2003, 5, 677–680.
[23] Wang J. Electrochemical glucose biosensors. Chem. Rev. 2008, 108, 814–825.
[24] Harper A, Anderson MR. Electrochemical glucose sensors – developments using electrostatic
assembly and carbon nanotubes for biosensor construction. Sensors. 2010, 10, 8248–8274.
[25] Heller A. Miniature biofuel cells. Phys. Chem. Chem. Phys. 2004, 6, 209–216.
[26] Barton SC, Pickard M, Vasquez-Duhalt R, Heller A. Electroreduction of O-2 to water at 0.6 V (SHE) at
pH 7 on the ‘wired’ Pleurotus ostreatus laccase cathode. Biosens. Bioelectron. 17, 2002, 1071–1074.
[27] Santoro C, Arbizzani C, Erable B, Ieropoulos I. Microbial fuel cells: From fundamentals to
applications. A review. J. Power Sources. 2017, 356, 225–244.
[28] Rinaldi A, Mecheri B, Garavaglia V, Licoccia S, Di Nardo P, Traversa E. Engineering materials and
biology to boost performance of microbial fuel cells: a critical review. Energy Environ. Sci. 2008, 1,
417–429.
[29] Katz E, Willner I, Kotlyar AB. A non-compartmentalized glucose vertical bar O-2 biofuel cell by
bioengineered electrode surfaces. J. Electroanal Chem. 1999, 479, 64–68.
[30] Rasmussen M, Abdellaoui S, Minteer SD. Enzymatic biofuel cells: 30 years of critical advancements.
Biosens. Bioelectron. 2016, 76, 91–102.
[31] Ohara TJ, Rajagopalan R, Heller A. Wired enzyme electrodes for amperometric determination of
glucose or lactate in the presence of interfering substances. Anal. Chem. 1994, 66, 2415–2457.
[32] Rodriguez-Delgado MM, Aleman-Nava GS, Rodriguez-Delgado JM, Dieck-Assad G, Martinez-Chapa
SO, Barcelo D, Parra R. Laccase-based biosensors for detection of phenolic compounds. Trends Anal.
Chem. 2015, 74, 21–45.
[33] Xiao XX, Xia HQ, Wu RR, Bai L, Yan L, Magner E, Cosnier S, Lojou E, Zhu ZG, Liu AH. Tackling the
challenges of enzymatic (bio)fuel cells. Chem. Rev. 2019, 119, 9509–9558.
[34] Suzuki A, Ishida K, Muguruma H, Iwasa H, Tanaka T, Hiratsuka A, Tsuji K, Kishimoto T. Diameter
dependence of single-walled carbon nanotubes with flavin adenine dinucleotide glucose
dehydrogenase for direct electron transfer bioanodes. Jap. J. Appl. Phys. 2019, 58, art.
no. 051015.
[35] Sakthivel M, Ramarai S, Chen SM, Chen TW, Ho KC. Transition-metal-doped molybdenum
diselenides with defects and abundant active sites for efficient performances of enzymatic biofuel
cell and supercapacitor applications. ACS Appl. Mat. Interf. 2019, 11, 18483–18493.
[36] Bollella P, Ludwig R, Gorton L. Cellobiose dehydrogenase: Insights on nanostructuration
of electrodes for improved development of biosensors and biofuel cells. Appl. Mat.
Today. 2017, 9, 319–332.
[37] Bollella P, Fusco G, Stevar D, Gorton L, Ludwig R, Ma S, Boer H, Koivula A, Tortolini C, Favero G,
Antiochia R, Mazzei F. A glucose/oxygen enzymatic fuel cell based on gold nanoparticles modified
graphene screen-printed electrode. Proof-of-concept in human saliva. Sens. Actuat. B: Chem. 2018,
256, 921–930.
[38] Leech D, Kavanagh P, Schuhmann W. Enzymatic fuel cells: Recent progress. Electrochim. Acta. 2012,
84, 223–234.
[39] Wey AT, Southcott M, Jemison WD, MacVittie K, Katz E. Electrical circuit model and dynamic analysis
of implantable enzymatic biofuel cells operating in vivo. Proc. IEEE 2014, 102, 1795–1810.
 Bibliography 223

[40] Majdecka D, Draminska S, Janusek D, Krysinski P, Bilewicz R. A self-powered biosensing device with
an integrated hybrid biofuel cell for intermittent monitoring of analytes. Biosens. Bioelectron. 2018,
102, 383–388.
[41] Montes-Cebrián Y, Álvarez-Carulla A, Colomer-Farrarons J, Puig-Vidal M, Miribel-Català PL. Self-
powered portable electronic reader for Point-of-Care amperometric measurements. Sensors. 2019,
19, 3715.
[42] Shitanda I, Momiyama M, Watanabe N. Toward wearable Energy storage devices: Paper-based
biofuel cells based on a screen-printing array structure. ChemElectroChem. 1017, 4, 2460–2463.
[43] Hou C, Liu A. An integrated enzymatic biofuel cells and supercapacitor for both efficient electric
energy conversion and storage. Electrochim. Acta. 2017, 245, 295–300.
[44] Kang ZP, Jiao KL, Cheng J, Peng RY, Jiao SQ, Hu ZQ. A novel three-dimensional carbonized PANI
(1600)@CNTs network for enhanced biofuel cell. Biosens. Bioelectron. 2018, 101, 60–65.
[45] Reuillard B, Le Goff A, Agnes C, Holzinger M, Zebda A, Gondran C, Elourzaki K, Cosnier S. High
power enzymatic biofuel cell based on naphthoquinone-mediated oxidation of glucose oxidase in a
carbon nanotube 3D matrix. Phys. Chem. Chem. Phys. 2013, 15, 4892–4896.i PCCP 15 2013 4892
[46] Bullen RA, Arnot TC, Lakeman JB, Walsh FC. Biofuel cells and their development. Biosens.
Bioelectron. 2006, 21, 2015–2045.
[47] Cosnier S, Gross AJ, Giroud F, Holzinger M. Beyond the hype surrounding biofuel cells: What’s the
future of enzymatic fuel cells? Curr. Opin. Electrochem. 2018, 12, 148–155.
Part III: Photoelectrochemistry in materials
science – selected topics
Generally, the photoelectrochemistry deals with an influence of electromagnetic
irradiation (light illumination) on the electrode properties and on the electro-
chemical reactions taking place at the electrode. We can distinguish mainly two
types of photoelectrochemical behavior:
1. Illumination of the electrode and its photoexcitation results in the subsequent
electrochemical reaction of neutral reactants at electrode. Such behavior is char-
acteristic of the semiconductor (SC)–electrolyte interface.
2. Illumination of nonphotoactive electrode and its vicinity results in photoexcita-
tion of reactants molecules and their further reactions. Such behavior is mostly
characteristic for photocatalysis and photoelectrocatalysis, but also may be ob-
served at SC–electrolyte interface with some dyes as reactants.

This part of the book we dedicate to the photoelectrochemistry of SCs and its appli-
cations in solar energy conversion. In the beginning, we will briefly describe the na-
ture of SCs, unusual properties of SC–electrolyte interface, simple charge transfer
reactions, and the behavior of SC under illumination (Chapter 10). Photoelectro-
chemical cells and their application in solar energy conversion will be discussed in
Chapter 11. Going further, we will also characterize some photocatalytic processes at
SC particles (Chapter 12).

https://doi.org/10.1515/9783111160986-010
10 Semiconductors electrochemistry and
photoelectrochemistry: fundamentals

10.1 Basic characteristics of semiconductors

Semiconductors (SCs) are materials with intermediate electronic conductivity between

the conductivity of metals and insulators. Typical SCs are germanium (Ge) and silicon
(Si), elements of group IV. SC compounds consist of the elements of other groups, such as
(i) AIIIBV (e.g., GaAs, GaP, InSb) and (ii) AIIBVI (e.g., ZnTe, CdTe, CdSe, CdS, ZnO). The com-
pounds containing rare-earth metals and elements of group VI (e.g., TiO2, WS2, MoSe2)
also have properties of SCs. The efforts are made to synthesize the more complicated
compounds such as ZnSeP2 or perovskites.
For the description of SC properties, the band theory of solid state is applied [1–3].
It is beyond the scope of this book to present in detail this theory; however, some infor-
mation is necessary to understand this chapter. For the solid state, it is assumed that
the energy of electron levels of corresponding atoms constitute the bands. Between the
bands with energy of the so-called allowed electron state is a band with a forbidden
energy region. The most important are the bands called the valence (VB) and the con-
duction band (CB). The VB is formed with occupied energy levels of valence electrons
and energetically is characterized by the highest occupied level of energy EV . The CB is
unoccupied at T = 0 K and can be filled with electrons when additional energy is sup-
plied to solids (heat, electric, or electromagnetic field). This band is characterized by the
energy of the lowest unoccupied energy level EC . In metals, the VB overlaps with the
CB. In SCs and insulators (dielectrics), these bands are separated by the band of for-
bidden energy, characterized by band gap energy Eg

Eg = EC − EV

SCs can be divided into two categories: intrinsic and doped SC. In the intrinsic SCs at
T > 0, the current carriers are generated as a result of the thermal excitation of elec-
trons from the VB to the CBs. At the same time, the equal number of holes, positively
charged carriers, is created in the VB. When another external stimulus (electric field,
light) is absent, the equilibrium concentration of electrons n0 and holes p0 is estab-
lished, so in the intrinsic SC: n0 = p0 = ni , and n0 p0 = n2i .
Both electrons and holes participate in the current flow; their concentration in
room temperature in intrinsic SC is low 1010−1011 cm−3. To increase the conductivity, the
foreign atoms (impurities) are induced in the SC lattice. These foreign atoms can work
as donors or acceptors of electrons. They have very low ionization energy (0.01 eV), and
so they are completely ionized at room temperature and become charged. If ionized,
the donors of concentration ND supply electrons to the CB, creating electron conductiv-
ity and becoming positively charged ND+ . In this type of SC (n-SC), the concentration of

https://doi.org/10.1515/9783111160986-011
230 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

electrons is much higher in the CB than holes in the VB: n0  p0 The electric current in
bulk of SC is then carried mostly by electrons, and because of that, the electrons are
called the “majority carriers” and the holes “minority carriers.” When acceptor atoms
are induced in the lattice, they withdraw the valence electrons creating holes. In this
type of SC, the concentration of positively charged holes in the VB is much higher than
the concentration of electrons in CB: p0  n0 . It is the so-called p-type semiconductor
(p-SC), where “majority carriers” are holes and “minority carriers” are electrons. The
acceptors in the SC lattice are charged negatively, and their concentration NA = NA− .
For the doping of Ge or Si, the elements of group V (P, As, Sb) are used as donors
and the elements of group III ( B, Al, Ga, In) as acceptors are utilized.
There are some parameters resulting from the band theory of solid state, which
are important for the description of SC properties. We will briefly describe some
of them.
The probability of the occupancy of the state of energy E by electron is given by
Fermi–Dirac function:
1
f ð EÞ =  (10:1)
E − EF
exp k T + 1
B

where EF is energy of Fermi level, kB is Boltzman constant, and T is temperature (K).

The occupancy of electron levels and their distribution in an intrinsic SC is shown
in Fig. 10.1.

E E

Ec T > OK
Eg T = OK
EF
EV

f (E)
1/2 1

Fig. 10.1: Occupancy of electron levels and Fermi–Dirac function for intrinsic semiconductor in thermal
equilibrium at T > 0 K. The Fermi energy remains approximately at the center of the bandgap.

As one can see that at T = 0 K, the f ðEÞ = 1 if E < EF and f ðEÞ = 0, if E > EF . It means that
at T = 0, all electrons are located in the VB. At the higher temperatures, there are as
many electrons excited into the CB as there are empty states in the VB, and so EF re-
mains at 21 Eg . When the energy of electron level E is much higher than EF and
 10.1 Basic characteristics of semiconductors 231

ðE − EF Þ  kB T, the Fermi–Dirac distribution function became Boltzman distribution

function:
EF − E
f ðEÞ = exp (10:2)
kB T

Another important term is the function of density electron states in the conduction
and VB:

3
mef 2 1
DC ðEÞ = 4π e
· ðE − Ec Þ2 , if E > EF (10:3a)
h2

3
mef
p 2 1
DV ðEÞ = 4π · ðEV − EÞ2 , if E < EF (10:3b)
h2

mef ef
e is the effective mass of an electron at the bottom of the CB, mp is the effective
mass of a hole at the top of the VB, and h is Planck’s constant.
Based on both functions, Boltzman function and the density state function, the
concentration of the charge carriers can be described in the thermal equilibrium by
the following equations:


EF − Ec
n0 = NC exp (10:4a)
kB T


EV − EF
p0 = NV exp (10:4b)
kB T

where NC , NV are the density of electron state in the CB and the VB, respectively.
The term Fermi level is very important and is defined as the energy of level for
which the probability of occupation by electron is ½. In electrochemistry, the Fermi
level is identical with an electrochemical potential of electron (Chapter 1):

EF = e
μe = μe − Fφ (10:5)

μe is a chemical potential of electron, and φ is inner potential in particular phase.

The EF value depends on the choice of the reference state. In physics, the energy
of electron in vacuum is taken as reference and is accepted to be zero.
For intrinsic SC, the Fermi level is located near the center of the forbidden band:

EV + EC kB T NV
EF = + ln (10:6a)
2 2 NC
232 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

For doped SCs, the Fermi level is described by the following expressions:
ND n0
EF = EC + kB Tln = EC + kB Tln , n0 ≈ ND , nSC (10:6b)
Nc Nc

NV NV
EF = EV + kB Tln = EV + kB Tln , p0 ≈ N A , pSC (10:6c)
NA p0

where ND , NA are the concentrations of donors or acceptors in the bulk of SC.

The scheme of the band energy and the location of energy levels EF, ED, EA for
intrinsic, n-SCs, and p-SCs is presented in Fig. 10.2(a–c).

(a) (b) (c)

f (E) f (E) f (E)

Conduction Conduction Conduction

band band band
EC EC EC
EF
EF ED EA
EF
EV EV EV
Valence Valence Valence
band band band

Fig. 10.2: Scheme of the energy bands for the intrinsic (a), n-type (b), and p-type (c) semiconductors. The
Fermi energy level EF and energy levels of donors ED and acceptors EA are marked.

10.1.1 Semiconductor under illumination

Two types of bandgap in SCs may be considered: direct and indirect band. According to
this, we distinguished SCs with direct and indirect bandgap energy. In the direct band
SC, the top of the VB occurs at the same value of the crystal electron momentum k, as
the bottom of the CB. In indirect band SC, the top and the bottom of SC bands are
shifted and electron momentum is not for them the same. Figure 10.3(a, b) illustrates
both types of SC and optical transitions from the valence to CB during illumination.
Let us briefly consider phenomena during the illumination of SC. When the
energy of photons Eph = hv < Eg , the photons may be absorbed only by lattice or free
carriers, then the absorption coefficient α is low. Absorption coefficient α is the rate
of decay of light incident intensity I0 and is a characteristic parameter, appearing in
Lambert–Beer’s law:

I ðxÞ = I0 expð − αxÞ (10:7a)

Here, I0 is the intensity of incident illumination, and I ðxÞ is the illumination intensity
at the distance x, at which the absorption is considered.
 10.1 Basic characteristics of semiconductors 233

(a) (b)
E E
Conduction Conduction
band Ef band
Ec Ec Ef
Eg h󰣍 Eg
hν
Ev Ei
Valence Ei Valence Ei
band band
K K

Fig. 10.3: Excitation of electrons across the bandgap by photon absorption. (a) Direct process, no phonon
is required. (b) The indirect process necessitates a phonon interaction.

If the photon of energy ðhv ≈ Eg Þ is absorbed in the “direct” SC, electron–holes pairs are
generated: electrons at the bottom of the CB and holes at the top of the VB. The excited
electrons and holes are not stable and their annihilation results in photon emission (the
so-called radiative recombination). The increase of photon energy ðhv > Eg Þ results in
an increase in kinetic energy of generated electron and holes. This excess energy is lost
to the lattice, such process is called thermalization. In direct transition from band to
band (interband transition), the absorption coefficient shows a step rise with the pho-
ton energy increase, up to 104 cm−1. The penetration depth x = α1 is then about 1
μm. Examples of “direct” SCs are GaAs, CdTe, CdSe, InP, ZnO. When the “indirect” SC is
irradiated with photon energy close to Eg , the electron transition between the maxi-
mum of the VB and the minimum of the CB is not possible. The transitions occur only
with the absorption of a photon and simultaneous absorption or emission of a phonon
(interaction with the lattice vibrations). The band-to-band transition, in the “indirect”
SC, occurs when ðhv > Eg Þ. Absorption coefficient for “indirect” SC is much smaller than
that for ”direct” SC; it means that the penetration depth of photon is higher in “indirect”
SC. Examples of such SCs are Si, Ge, GaP, and α-SiC.
The absorption coefficient α for direct and indirect transitions is given by
 n
α = A hν − Eg (10:7b)

where A is a proportionality coefficient and n = ½ or n = 2 for direct and indirect tran-

sition, respectively.
Before going further, we introduce the other term applied in the SC physics: ex-
citon. In the interband transitions when electrons and holes are generated they
formed the electrically neutral, two particles state electron–hole referred to exciton.
The excitons result from electron–hole Coulomb attraction. They move through a
crystal as a single entity. The lifetime of excitons is limited. They may annihilate in
the recombination process or dissociate in the electric field to the “free” electron
and holes, moving in the opposite direction.
234 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

When the SC is no longer in thermal equilibrium being perturbed by an external

stimulus, for example, illumination, the concentration of electrons n and holes p are
not equal to the equilibrium value n0 , p0 . In this case, the concentrations of carriers
in the bands are as follows:

nðxÞ = n0 + Δn (10:8a)

pðxÞ = p0 + Δp (10:8b)

Under illumination, the relative changes in the concentration of minority carriers are
much higher than the changes in the concentration of majority carriers. These expres-
sions illustrate the relative changes in the concentration:
Δp Δn Δn Δp
for nSC:  and for pSC: 
p0 n0 n0 p0

The changes in the concentration of carriers under illumination disturb the equilib-
rium. The terms named quasi-Fermi levels are introduced for the characterization of
the energy of charge carriers in non-equilibrium conditions. The quasi-Fermi levels
for electrons n EF✶ and holes p EF✶ are not equal as can be seen from the following
equations:

✶ n
n EF = Ec + kB Tln (10:9a)
NC

✶ NV
p EF = EV + kB Tln (10:9b)
p

The changes in concentration and energy of electrons cause the split of Fermi level as
is shown schematically in Fig. 10.4(a, b).

(a) (b)
E (ev) E (ev)
Ec Ec
*
nE f nE f
*
EF
Photon Photon
h󰜈 EF h󰜈
* *
pE f
pE f
Ev Ev
n-type p-type

Fig. 10.4: Quasi-Fermi levels for semiconductors under illumination showing significant changes in Fermi
energy levels of minority carriers (a) holes (n-SC) and (b) electrons (p-SC).

As you can see, the changes in Fermi energy levels are much higher for minority car-
riers because of the significant increase in their concentration under illumination. The
bands in both Figs. 10.2 and 10.4 are flat. This is a strong simplification and means that
the surface of the SC is not charged. In practice, we described, investigated, and used
 10.1 Basic characteristics of semiconductors 235

junctions (interfaces) such as Me|SC, n-SC|p-SC, SC|electrolyte. They are charged and
vary in their properties from the bulk properties. The behavior and proper energy
scheme of the SC–electrolyte interface will be described later.

10.1.2 Recombination

As was mentioned above, when the SC is illuminated with the energy E ≥ Eg , a certain
amount of the electron–hole (e–h+) pairs is generated: electrons in the conduction and
holes in the VB, resulting in the perturbation of equilibrium. The recombination is the
process of the equilibrium restoration as a result of the annihilation of e–h+ pairs.
The nonequilibrium concentration of electrons n and holes p is determined in the
bands by equations:

dðΔnÞ
= Gn − Rn (10:10a)
dt
dðΔpÞ
= Gp − R p (10:10b)
dt

Gn , Gp are generation rates, and Rn , Rp are recombination rates of electrons and

holes, respectively.
The recombination rates are described by equations:

n − n0
Rn = (10:11a)
τn
p − p0
Rp = (10:11b)
τp

Equations (10.11a, b) describe the mean lifetime of electrons τn generated in the CB

and of holes τp generated in the VB. If the deviations from equilibrium are small, we
can assume that τn , τp are constant. Under this condition when the external stimulus
is switched off, the lifetime of carriers is defined as the time at which the excess con-
1
centration of electrons or holes decreases to exp its initial value.
We can distinguish several types of recombination:
(i) the direct band-to-band recombination,
(ii) the recombination with participation of impurities, defects that work as the traps,
often called trap-assisted recombination or Shockley–Read–Hall (SRH) recombination,
(iii) the surface recombination.

In direct band-to-band recombination, the generated free electrons go back from CB to

VB. It may occur with photon emission (radiative recombination), or the energy excess
may be transferred to SC lattice as phonons. The recombination rate for band-to-band
radiative recombination is described by equation Rr = Br ðnp − n0 p0 Þ. Since during illu-
236 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

mination, the concentration of majority carriers is slightly changed, the radiative re-
combination depends only on the concentration of minority charge carriers. For n-SC,
the recombination rate is described by Rp ; for p-SC by Rn , eqs. (10.11b) and (10.11a), re-
spectively. Such type of recombination is dominant in narrow bandgap SCs and in
wide bandgap SCs with direct transition (CdS, GaAs).
In the second case, SRH recombination occurs via the intermediate bulk local
energy level created within the bandgap by dopants and defects in the crystal lat-
tice. Such energy levels are named traps. If their energy lays close to the CB they are
electron traps; if it lays close to the VB they become the hole traps. This type of re-
combination is non-radiative; the excess of energy is lost in the form of thermal en-
ergy in the lattice vibrations. SRH recombination is dominant in indirect bandgap
SCs. Much more about the trap-assisted recombination and the equations describing
the recombination rate and lifetime of charge carriers can be found in [3]. The bulk
recombination decreases the efficiency of solar energy conversion and should be
limited. Among the methods that are applied is the reduction of the SC bulk size by
nanostructuring. The surface recombination will be described later in Section 10.5.

10.1.3 Carriers and electric field

When the electric field is not applied to SCs, the concentration of charge carriers is
determined by equilibrium values n0 and p0 eq. (10.4); the average velocity of the
movement of the carriers is zero. In the electric field, the velocity (the so-called drift
velocity) of the electrons and holes differs resulting from differences in mobility of
electrons μn and holes μp . The external electric field causes the changes in the energy
of carriers and in consequence influences their concentration:

eðφ − φB Þ
n = n0 exp (10:12a)
kB T

− eðφ − φB Þ
p = p0 exp (10:12b)
kB T

Here, φ − φB is the potential drop in SC; further in the text it will have the meaning of
potential drop in the space charge layer of SC.
The ordered movement of electrons and holes in the electric field is called drift
current idrift or in electrochemistry – migration current. The other kind of current that
flows results from variation in concentration of carriers. It is diffusion current idiff
similar to that described for electrochemical kinetics (Chapter 1). Generally, the den-
sity of total current j is the sum of drift current density and diffusion current density
of both carriers:
 10.2 Semiconductor–electrolyte interface 237

j = jn + jp (10:13a)

dn
jn = jn, drift + jn, diff = enμn 2 + eDn (10:13b)
dx
dp
jp = jp, drift + jp, diff = epμp 2 − eDp (10:13c)
dx

where e is elementary charge (e = 1.6 * 10−19 C), Dn and Dp are diffusion coefficients of
electrons and holes, μn , μp are mobility of electrons and holes, and 2 is the intensity
of the electric field. As there is a large difference between the concentration of major-
ity and minority carriers, only one part of eq. (10.13a) is important.

10.2 Semiconductor–electrolyte interface

Before going further, we should correlate the energy scale used by physicists and
electrochemists. In thermodynamic electrochemical scale, all oxidation–reduction
potentials or equilibrium potentials are referred to the standard hydrogen electrode
(SHE); its standard electrode potential is accepted to be 0 V ðE0 = 0Þ in an ideal 1 M
solution of acid, at any temperature. In practice, the values of potentials are referred to
different reference electrodes: normal hydrogen electrode (NHE), standard potential of
which is accepted to be 0 V ðE0 = 0Þ in 1 N acid solution, SCE – saturated calomel elec-
trode, silver chloride electrode Ag|AgCl|Cl –, and others.
It is clearly seen that for the description of the SC properties more convenient is
∼
to use physical scale, where energy of electron level EF = μ, Ec , Ev and other energy
e
levels are referred to as the value 0 eV in vacuum. If we want to consider the proper-
ties of a SC electrode from an electrochemical point of view, we should correlate both
scales. Such a correlation is shown in Fig. 10.5. The correlation is based on some ex-
perimental and theoretical results pointing out that the value of the Fermi energy
level (electrochemical potential of electrons) for the SHE electrode is − 4.5 ± 0.1 eV
(more exactly − 4.44 ± 0.2 eV) [4].

10.2.1 Model of semiconductor–electrolyte interface

When the metal is immersed in the electrolyte solution, the double layer is created spon-
taneously at the interface, which consists of the Helmholtz layer and the Gouy–Chapman
layer (Chapter 1). The concentration of electrons in a metal is very high, ca. 1022 cm−3;
therefore the charge of accumulated ions at the metal surface is compensated within a
few Å in the metal vicinity. In the SC, the electron concentration is much lower, in intrin-
sic SC is not higher than 1011 cm−3, and in doped SC it is not higher than 1015–1016 cm−3.
Because of that, the charge of electrolyte ions or other charged species existing on/at the
238 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

E (eV)
CdTe
–3.0 –1.5 GaAs
CdS TiO2
–1.0 CdSe
–4.0 –0.5 1.5 1.4
0.0 0.0 V
–5.0 1.7 eV eV
+0.5 2.25
+1.0 eV eV
–6.0 +1.5
+2.0 3.2
–7.0 +2.5 eV Conduction band
+3.0
–8.0 +3.5 Valence band

E (V), SHE

Fig. 10.5: The position of conduction and valence band edges for some semiconductors is shown on the
vacuum scale and with respect to the SHE reference.

SC surface is compensated in the vicinity of SC. This charged region in SC is called space
charge layer and is characterized by the potential drop φSC , charge QSC , capacity CSC , and
thickness LSC . The existence of the space charge layer in SC results in the differences
between the behavior of SC electrodes and metal electrodes. The electric field of the
space charge layer leads to the changes in the concentration of electrons, to the changes
in their potential energy resulting in the band bending. Since the electron is nega-
tively charged, the bands are bent downward if φSC > 0 and upward when φSC < 0.
The schematic model of SC–electrolyte, together with potential changes, is shown in
Fig. 10.6. Note that the x-axis is directed toward the SC bulk and that at the interface
SC–electrolyte x = 0.
The Galvani potential φG of SC–electrolyte interface is described as follows:

φG = φðSCÞ − φðelÞ = ½φð∞Þ − φð0Þ + ½φð0Þ − φð− LH Þ + φð− LH Þ − φð−∞Þ

(10:14)
= − φSC + φH + φG − Ch

φSC , φH , φGCh are potential drops in the space charge layer of SC, in Helmholtz, and in
Gouy–Chapman layers (ionic part of double layer), respectively.
Note that the potential of the ionic part of the double layer and space charge
layer have different signs. It results from the assumptions that potential (electric
field) decreases to 0 in the bulk of SC and for ionic part of the layer it drops to 0 in the
bulk of electrolyte.
The Galvani potential between two conductors cannot be measured, as discussed
in Chapter 1. What we can measure is the electrode potential E. In practice, the term
“electrode potential” is used to denote the difference between two electrodes, the
studied electrode, and the reference electrode. In reality, the electrode potential is the
algebraic sum of all individual Galvani potentials of interfaces. If the electrode is in
 10.2 Semiconductor–electrolyte interface 239

Semiconductor Electrolyte

Space –
charge e + +
Helmholtz
layer layer
–
e
Bulk + +
SC –
e
Gouy–chapman
–
e layer
+ –
Qsc < 0
Lsc

𝜑G-ch

𝜑H

𝜑SC

x 0 –x

Fig. 10.6: Scheme illustrating the electric double layer at the semiconductor–electrolyte interface. The
distribution of potentials within the layer is shown.

equilibrium with the redox system, the electrode potential can be calculated by the
Nernst equation with respect to the NHE electrode.
In the part of the text that follows, we will denote the electrode potential as E,
and potential applied from an external source as U.

10.2.2 Space charge layer

Three states of the space charge layer can be distinguished: depletion, inversion, and
accumulation. They are presented in Fig. 10.7.
Depletion occurs when the charge in the space charge layer derives from the ion-
ized atoms of donors or acceptors. In the case of n-SC, as the donors are positively
charged, QSC > 0, the potential drop φSC < 0. In p-SC, acceptors are negatively charged
then QSC < 0 and the potential drop φSC > 0 When in the n-SC, the positive charge QSC
increases and the potential drop φSC became more negative; the space charge layer is
in the inversion state. In this case, the charge is formed by mobile minority carriers –
holes. In the p-SC, if the space charge layer is in the inversion state, the charge is also
formed by mobile minority carriers, but in this case, electrons. Therefore, QSC became
more negative and φSC more positive in comparison with the depletion state. The ac-
240 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

n–type n–type n–type

E (eV) E (eV) E (eV)

EC
EC φsc
EF EC EF
EF
EV
EV
EV
Qsc >0, φsc <0 Qsc >>0, φsc <0 Qsc <0, φsc >0
Depletion Inversion Accumulation

p-type p-type p-type

E (eV) E (eV) E (eV)
EC EC
φsc E

EF EF
EV E
EV
E
Qsc <0, φsc >0 Qsc <<0, φsc >0 Qsc >0, φsc <0
Depletion Inversion Accumulation

Fig. 10.7: Diagrams for different states of space charge region: depletion, inversion, and accumulation.

cumulation state in the space charge layer exists, when the charge is created by ma-
jority carriers. Then for n-SC QSC is negative, while for p-SC QSC is positive and so
φSC > 0, φSC < 0 for n-SC and p-SC, respectively.
If φSC = 0, the bands are not bent; they are flat to the surface. Such a state can be
forced by the external potential. Potential of SC electrode (measured against reference
electrode) at which such state is reached is called the flat band potential Efb . The posi-
tions of Efb for some SCs are shown in Fig. 10.5.
The expressions describing QSC , φSC , LSC , CSC are complicated and their forms de-
pend on the type of SC and level of doping. Their derivations are beyond the scope of
this book and can be found in [3]. Here we describe the simple case for n-SC when the
depletion state exists in the space charge layer. As you will find later, this depletion
state is very useful when the conversion of light energy is considered.
In the depletion state of the space charge layer of n-SC, QSC > 0; the charge is de-
termined by the concentration of ionized donors ND = ND+ . The potential distribution
can be obtained in this case by solving eq. (10.15), setting the following boundaries for
integration: φðLSC Þ = 0, φð0Þ = φSC , x = LSC

d2 φ eND
=− (10:15)
dx 2 ϵ0 ϵSC
where ϵSC is relative dielectric permittivity of SC, and ϵ0 is dielectric permittivity of
vacuum.
 10.2 Semiconductor–electrolyte interface 241

The solution of this equation is as follows:

eN D
φðxÞ = − ðx − LSC Þ2 (10:16)
2ϵ0 ϵSC

at the interface where x = 0,


eN D
φðxÞ = φSC = − LSC 2 (10:16a)
2ϵ0 ϵSC


2ϵ0 ϵSC jφSC j 1=2
LSC = (10:17)
eN D

It is clearly seen that the thickness of the space charge layer LSC depends on the dop-
ing. In the lightly or moderately doped SC, LSC is in the range of 10–1,000 nm.
The other term that is important is the so-called Debye length (screening length):


ϵ0 ϵSC k B T 1=2
LD = , ðn − SC, depletionÞ (10:18)
2e2 N D

This term is defined for all SC states and types, by equation:

1=2
ϵ0 ϵSC kB T
LD = (10:19)
e2 ðn0 + p0 Þ

and is also applied in the description of Gouy–Chapman layers in electrolyte (Chapter 1).
The differential capacity of the space charge layer can be delivered from the
dQ
equation CSC = dφSC , and for the depletion state of n-SC is given by
SC

1=2
ϵ0 ϵSC φSC
CSC = pffiffiffi −1 (10:20)
LD 2 kB T

Using eq. (10.18), we can write (10.20) in the form:


−2 2 kB T
CSC = 2 2 jφSC j − (10:21a)
ϵ0 ϵSC N D e

or


−2 2 kB T
CSC = E − Efb − (10:21b)
ϵ20 ϵ2SC N D e

where Efb , E are flat band potential and electrode potential, measured against the ref-
erence electrode, respectively.
−2
The dependence of CSC versus E is linear, and such a plot is called a Mott Schottky
plot and measurements of CSC as a function of SC electrode potential give us the possi-
242 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

bility to determine Efb and ND . Mostly the impedance spectroscopy is applied for CSC
determination (Chapter 2).
The other important parameters are the concentration of electrons and holes in
the space charge layer. They vary from equilibrium values in the bulk of SC and de-
pend on the potential drop in the space charge layer:

φSC = φs − φb (10:22)

Here φs is potential on the SC surface and φb is potential in the bulk of SC referred to as 0:


eφSC
n = n0 exp (10:23a)
kB T


− eφSC
p = p0 exp (10:23b)
kB T

10.2.3 Polarization of semiconductor–electrolyte interface

As was mentioned above, the Galvani potential of SC–electrolyte interface is the sum
of three potential drops:

φG = φðSCÞ − φðelÞ = − φSC + φH + φG − Ch

In consequence, the differential capacitance of SC–electrolyte interface C also consists

of three capacities CSC , CH , CG − Ch connected in series:
1 1 1 1
= + + (10:24)
C CSC CH CG − Ch

If we compare the thickness of these three regions in the middle concentrated electro-
lyte, the inequalities LSC > LH and LSC > LGCh are fulfilled in the majority of cases.
Note that LGCh = λD (Chapter 1). Therefore, the main interfacial potential drop φG is
realized in the space charge layer of SC, jφSC j  jφH j, jφG − Ch j; and also CSC has the
main contribution to capacitance C of the electrical double layer.
When the potential of SC electrode E is changed by externally applied voltage, U,
the potential profiles should vary in every part of double layer:

U − E = ΔE = −ΔφSC + ΔφH + ΔφG−Ch (10:25)

Let us consider the case when jΔφSC j  ΔφH , Δ φG − Ch . It means that the change of SC
electrode potential takes place in the vicinity of the space charge layer, so ΔE = jΔφSC j.
In this situation, the positions of the band edges at SC surface EC, S , EV, S are constant
with respect to the reference electrode and electrolyte. The band edges are “pinned”
to the SC surface (Fig. 10.8(a)). If the SC is very strongly doped, the Fermi energy levels
are close to the conduction or VB. The SC begins to behave as the metal. In this case,
 10.3 Fundamentals of electrochemical reactions on the SC electrode 243

we referred to the term “metalized” SC. It results in a situation, jΔφH j  jΔφSC , and
means that the changes in electrode potential take place in the Helmholtz layer. In
the energy diagram, the energy levels of the surface are shifted with respect to the
electrolyte and the bands are “unpinning.” However, with respect to the Fermi level
of SC, the band edges EC, S , EV, S maintain the same relative position (Fig. 10.8(b)).

(a) E (ev) E (ev)

eφsc
EC –eΔφsc
EF EC eΔE
EF

EV
EV
SC Electrolyte SC Electrolyte

(b) E (eV)

eΔφH

EC
EF

EV
SC Electrolyte

Fig. 10.8: (a) Energy schemes for semiconductor–electrolyte interface showing influence of applied
external potential. Scheme illustrates the pining of band edges EC, EV at surface. The application of
external potential change Fermi energy EF and φSC . ΔE = ΔφSC . (b) Energy scheme for
semiconductor–electrolyte interface to which the external potential is applied illustrates the pining of
Fermi level and unpinning of EC , EV . The polarization changes the position of EC, EV and φH . ΔE = ΔφH .

The second case, in which the changes in electrode potential take place mostly in
Helmholtz layer jΔφH j > jΔφSC j, is when the concentration of the surface state is large.
The influence of the surface state on the behavior of the SC–electrolyte interface will
be described in Section 10.5.
Much more about features of SC–electrolyte interface can be found in [3, 5, 6].

10.3 Fundamentals of electrochemical reactions

on the SC electrode

In the previous sections, the specific features of SC and their influence on SC–electrolyte
interface behavior were described. These features such as the bandgap energy, two types
of charge carriers, the existence of space charge layer in SC, and the surface states influ-
244 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

ence the electrode kinetics as well and differ it from electrochemical reaction at the
metal electrodes. More about the kinetics of the reactions at SC electrodes can be found
in the books [3, 5, 7–9]. Here we will concentrate on the reactions occurring at SC electro-
des in the solution containing the redox couple. The properly selected redox systems are
used in electrochemical solar cells to create the depletion state in the space charge layer
of SC and also to protect the SC electrodes against corrosion and photocorrosion.

Semiconductors in the solution with the redox system

If we immerse SC in the electrolyte containing redox couple, the equilibrium is estab-
lished. This equilibrium is described by equality of electrochemical potentials of elec-
trons in SC and in redox couple in electrolyte:

μ ~e ðRedoxÞ
~e ðSCÞ = μ (10:26a)

Since the electrochemical potential of electrons in solid state is equal to the Fermi en-
~e , the concept of the introduction of the Fermi energy level of redox
ergy level EF = μ
couple and description of the redox system in terms of occupied and unoccupied en-
ergy levels was developed by Gerisher [7]. This model will be described very briefly
later. The equilibrium between SC and electrolyte solution with redox couple can be
described then by relation:

EF = EF, Redox , EF, Redox = − 4.5eV − eERedox (10:26b)

cRed
EF, Redox = EF,
0
Redox + RT ln (10:26c)
COx

EF, Redox is the Fermi energy level of the redox couple, and ERedox is the equilibrium
potential of the redox couple.
If EF > EF, Redox , the equilibrium is established by the movement of electrons from
n-SC to Ox form of the redox system. As a result, in the space charge layer of SC, the
depletion state is formed ðQSC > 0Þ. It is also true, when Efb < ERedox .
In p-SC the depletion state is formed, when EF < EF, Redox or Efb > ERedox In this case,
QSC < 0. Described situations are illustrated in Fig. 10.9(a, b). To obtain in the space
charge layer of SC, the inversion state, or accumulation state, one has to choose redox
couple with proper equilibrium potential ERedox .
To create the inversion state in the space charge layer, we have to use the redox
couples which fulfill the relations EF  EF, Redox for n-SC and EF  EF, Redox for p-Sc,
respectively. The accumulation state is obtained when EF < EF, Redox for n-SC and
EF > EF, Redox for p-SC.
Note that the units of EF , EF, Redox are eV but the units of Efb , ERedox are V.
Let us now consider the charge transfer processes in the dark taking place at n-SC
electrode immersed in the electrolyte with a redox system [8]. The established equilib-
rium is dynamic. It means that the electrons flow from the CB of n-SC to the oxidized
form of redox couple in the electrolyte solution and electrons flow from the reduced
 10.3 Fundamentals of electrochemical reactions on the SC electrode 245

(a) E (eV) E (eV)

φ0 <0
sc
EC EC
EF
EF,Redox EF EF,Redox

EV EV
Qsc>0
n-sc E (V) n-sc E (V)

(b) E (eV) E (eV)

EC EC
φ0 >0
s
sc

EFF,Redox
,Redox EF EF,Redox
EF
EV EV
Qsc<0
p-sc E (V) p-sc E (V)

Fig. 10.9: Influence of redox system in electrolyte on the state of space charge layer in semiconductor:
(a) n-SC and (b) p-SC. The redox system generates the depletion state in the space charge layer of
semiconductors.

form of redox system in solution to the CB of n-SC. These are described by the reac-
tions (i) and (ii):
kf
i) eðSCÞ + OxðelÞ ! RedðelÞ, forward reaction
kf
ii) eðRed, elÞ + vðSCÞ ! OxðelÞ, reverse ðbackwardÞ reaction

Here v represents the empty electron state in the CB of n-SC and is assumed to be
constant, independent of potential, and e(SC) and e(Red, el) are electrons from SC and
from Red form in the electrolyte solution, respectively.
The rate of both reactions can be expressed by simple equations:

rateðiÞ = kf nS0 cOx , rateðiiÞ = k′r cRed , k′r = kr v

nS0 is the surface (space charge layer) concentration of electron in the CB of n-SC in
eφ0
equilibrium, nS0 = nexpð− k SCTÞ, and φ0SC is a potential drop in the space charge layer
B
of SC when equilibrium is established between SC and the redox system in the
electrolyte.
The flux of electron in forward and reverse reaction creates the cathodic and an-
odic current, in CB, respectively. The currents are very low and the net current in
equilibrium is zero like for the case of a metal electrode in the electrolyte with a
redox system. Considering the above, it is clearly seen that for n-SC only forward reac-
246 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

tion (cathodic current) will be a potential-dependent one since the electron concentra-
tion in the space charge layer is also potential dependent. Applying the external po-
tential (bias), U, we can write the equation describing the cathodic current density:


eðφ0SC + UÞ
jC, CB = − eFkf cOx ns , where ns = nexp − (10:27a)
kB T



eðφ0SC + UÞ eU
jC, CB = − eFkf cOx nexp − = − eFkf cOx nS0 exp − (10:27b)
kB T kB T


eU
jC, CB = j0, CB exp − (10:27c)
kB T

The anodic current density is constant, not U-dependent, and in equilibrium it is

jA, CB = − eFk′r cRed ffi jC, CB = − eFkf cOx nS0 = j0, CB

eq eq
(10:28)

Therefore, the current that flows in the dark across the CB of n-SC is described by the
equation:


− eU
JCB = jA, CB − JC, CB = j0, CB 1 − exp (10:29a)
kB T


− FU
JCB = j0, CB 1 − exp (10:29b)
RT
eq
Please note that jA, CB ffi jA, CB .
Now we consider the p-SC. Here, the VB is practically full, and so the electrons
transfer to the solution is constant and independent of potential. However, the elec-
tron transfer from the reduced form of redox couple in the solution to the VB depends
on the density of holes, empty space at the surface (space charge layer). This density
φ0
is the potential dependent pS0 = pexpð− k SCT Þ, and so using a similar procedure as
B
above we can obtain the density of current across the VB:


eU
JVB = j0, VB exp −1 (10:30a)
kB T


FU
JVB = j0, VB exp −1 (10:30b)
RT

The j0, CB and j0, VB have a meaning of exchange current density (Chapter 1); however,
they do not depend only on rate constant k and cOx , cRed but also on nS0 or pS0 , respec-
tively. One can change these concentrations by changing the equilibrium condition.
For instance in the case of n-SC, the application of a redox system with EF, Redox much
more negative than EF of n-SC (ERedox is more positive) causes the φ0SC increase and
thus a decrease of nS0 .
 10.3 Fundamentals of electrochemical reactions on the SC electrode 247

From eqs. (10.29a) and (10.30a), or (10.30b) and (10.30b), it is clearly seen that
significant currents can flow only in one direction. This behavior is shown in
Fig. 10.10 for both types of SC.

n-sc j (mA/cm2) p-sc j (mA/cm2) Anodic

current
Reverse bias (forward bias)

U (V) U (V)
Cathodic Reverse bias
current
(forward bias)

Fig. 10.10: Current density–potential characteristics of semiconductor electrodes in an electrolyte

solution, in dark.

It is the current of majority carriers, electrons in n-SC and holes in p-SC. For n-SC, the
cathodic current increases exponentially for negative U. For positive U the current
is much lower in magnitude and almost independent of potential. For p-SC, the an-
odic current increases exponentially while U is positive and at negative U current is
saturated and potential independent. From above it can be clearly seen that the
SC–electrolyte interface has rectifying properties similar to p–n junctions. In SC elec-
tronics, when the current magnitude is large, the applied potential is called a forward
bias, and when the current is small the notion reverse bias is used. The model discussed
above is similar to that applied for the description of current–voltage behavior of n–p
junction [8].
Assuming that the potential of SC electrode in the equilibrium with electrolyte so-
lution containing the redox couple is determined by ERedox , we can exchange the U in
the above equations by the overpotential ƞ. In Chapter 1, η was defined as η = E − Eeq .
In this case, the equivalent description of overvoltage (overpotential) is η = U − ERedox
and the current densities can be written:


− F j ηj
JCB = j0, CB 1 − exp (10:31a)
RT


F j ηj
JVB = j0, VB exp −1 (10:31b)
RT

then Tafel equation and plot (Chapter 1) can be used to determine the j0, CB and j0, VB .
In the electrochemistry of SCs, the Gerisher model is often used for the descrip-
tion of electrode kinetics with charge transfer mechanism [7]. The model can be called
“macroscopic” since the charge transfer is considered between the energy state of
electrons of the SC electrode and redox couple in the solution.
The basic concepts of this model are as follows: (i) the electrons can be exchanged
only between the same energy levels of SC and redox couple, without any energy
248 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

losses. (ii) The electron energy levels of redox couple are characterized by empty and
occupied states; the empty, unoccupied states come from the oxidized form of redox
couple, while occupied states come from reduced form. (iii) For the energy description
of the redox couple and its electron states, the terms applied in solid state are used,
such as energy of Fermi level EF, Redox (eq. 10.26b) and function of the density of elec-
tron states in the redox system DRedox ðEÞ and DOx ðEÞ, DRed ðEÞ for the unoccupied and
occupied energy electron levels, respectively. Due to the thermal fluctuation of the
solvation shell, the occupation of these energy states in the redox electrolyte is given
by Boltzman distribution function f ðEÞ (eq. 10.2). In this equation EF is exchanged by
EF, Redox . Note that E means here the energy.
Density of states DRedox is the sum of unoccupied and occupied states (electron
energy levels):

DRedox ðEÞ = DOx ðEÞ + DRed ðEÞ

they are equivalent to the density of state function in SC in CB and VB (eqs. 10.3a, b).
For the redox system,

DOx ðEÞ = cOx WOx ðEÞ (10:32a)

DRed ðEÞ = cRed WRed ðEÞ (10:32b)
" 2
#
ðE − EF,
0
Redox + λÞ
WOx ðEÞ = W exp
0
(10:32c)
4kB Tλ
" 2
#
ðE − EF,
0
Redox − λÞ
WRed ðEÞ = W exp
0
(10:32d)
4kB Tλ

cOx , cRed are concentration of oxidized and reduced forms of redox system, EF, 0
Redox is
the energy of standard Fermi level of redox system (eV), corresponding to standard po-
0
tential ERedox (V), and λ is the reorientation energy (Chapter 1, Marcus theory).
Above equations can be illustrated by Fig. 10.11(a, b), where the dependence of
electron energies of a redox system versus density of state is shown for different con-
centration of Ox and Red forms.
Applying these equations and remembering that the flux of charge carriers elec-
trons and holes depend not only on the redox system but also on the concentration of
carriers and their distribution in SC, the equations for currents were delivered for SC
electrode with redox system and under polarization. It is beyond this book; the inter-
ested reader will find much more about Gerischer’s model and its application in [5, 7].
Here, using this model, we will only illustrate the correlation between the energy
state of the redox system and SC when the electrode is in the equilibrium and polar-
ized by an external potential source. Figure 10.12 illustrates the equilibrium estab-
lished between n-SC electrode and the redox system in the electrolyte. The Fermi
energy EF and EF, Redox are equal and lay near the CB. The electrons transfer will occur
 10.3 Fundamentals of electrochemical reactions on the SC electrode 249

(a) (b)
E (eV) E (eV)
Dunocc Dunocc

0
Eox
2λ
EF,Redox EF,Redox
0
ERed

Docc Docc

Density of states Density of states

Cox = CRed Cox > CRed
0
EF,Redox = EF,Redox

Fig. 10.11: Diagram showing the distribution of the electron energy levels of the oxidized and reduced
forms of a redox couple: cOx = cRed (a) and cOx > cRed (b).

E (eV)

Dox

φ°
sc Efb
EC
EF EF,Redox

DRed

EV Redox system
E (V) in electrolyte
n-Sc

Fig. 10.12: Semiconductor–electrolyte interface at equilibrium showing overlap of the energy levels of Ox
form of redox couple with conduction band. The space-charge layer is in a depletion state. The exchange
current that flows between the conduction band and Ox form in the electrolyte is very small.

E (eV)
EC e Dox

EF,Redox

EV DRed
Redox system
E (V) in electrolyte

Fig. 10.13: Semiconductor electrode under cathodic polarization. Cathodic polarization generated the
accumulation state in the space-charge layer rich in electrons. In this situation current that flows between
the conduction band and Ox form in the electrolyte is large.
250 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

between the CB and oxidize form of redox couple. Additionally the position of flat
band potential and φ0SC are marked. The exchange current will be very low since the
concentration of electrons in the space charge layer of SC (depletion state) is low and
energy levels of the oxidize states are empty. Figure 10.13 presents the changes in this
scheme caused by the cathodic polarization. The cathodic polarization was used to
change the state of the space charge layer from depletion to accumulation. In this
case, the cathodic current (flow of electrons between the CB and oxidized form) will
be large since the concentration of electrons is potential dependent.

10.4 Photoeffects at/in semiconductor electrode in electrolyte

It was mentioned above that when n-SC is immersed in electrolyte containing redox
couple and Fermi energy EF of SC is larger than Fermi energy of the redox system in
electrolyte solution EF, Redox ðEF > EF, Redox Þ, the equilibrium is established and the de-
pletion state is formed in the space charge layer of SC. Now we will illuminate n-SC,
characterized by absorption light coefficient α, with the light energy hν ≥ Eg . During
the illumination, the pairs e–h+ are generated and may be separated by the electric
field of the space charge layer in the depletion state. The majority carriers (electrons)
are directed to the bulk of SC and when the electrical circuit is closed to the opposite
electrode. Minority carriers (holes) move to the interface and take place in the oxida-
tion reaction of reduced species of redox couple: Red + h + ! Ox. This reaction is
equivalent to the reaction of electrons of reduced forms with holes in SC. The changes
in the energy of electron levels and in the potential of the space charge layer under
illumination of n-SC surface are shown in Fig. 10.14.

E (eV)

Ec*
Ec
*
nEF
EF* eEph
EF EF, Redox
*
pEF
Ev*
Ev

Fig. 10.14: Energy scheme of semiconductor–electrolyte interface, in dark and under illumination (red
symbols). The conduction and valence band edges are pinned. During illumination the energy of Fermi
level, EF , is changed together with EC, EV in bulk of SC.
 10.4 Photoeffects at/in semiconductor electrode in electrolyte 251

Since during illumination concentration of charge carriers increases, EF , φSC are

changed (Sections 10.1.1 and 2). The EF or φSC changes are measured to the reference
electrode as the open circuit potential under illumination:
OC
Eph = EdOC + Eph (10:33)

EdOC is the OCP (open circuit potential) of electrode measured in the dark, and Eph is
the photopotential. If in the dark, the equilibrium is established between SC and
redox couple present in electrolyte, the EdOC is equal to the equilibrium potential
ERedox . The photopotential is described by equation:

kB T
Eph = − lnð1 + bI0 Þ (10:34)
e

b is parameter dependent on diffusion coefficient, diffusion length, the concentration

of carriers, and absorption coefficient. The delivery of (10.34) equation and the equa-
tion determining b parameter can be found in the book [3]. Equation (10.34) points
out that Eph < 0 since b and I0 > 0. If the depletion layer exists in SC and there is no
surface recombination, the measured Eph OC
is always more negative than EdOC . Equation
OC
(10.34) also shows the influence of the intensity illumination I0 on Eph ; at low inten-
sities, this dependence should be linear, and at higher intensities it should be
logarithmic.
During illumination, when the SC electrode is connected through an external cir-
cuit with another electrode in electrolyte solution, we can measure also changes in
the current:

jlight = jd + jph (10:35)

jlight , jd , jph are the current densities: under illumination jlight , in the dark jd , and pho-
tocurrent jph .
To derive the expression for photocurrent, we should consider the depth of light
penetration x = α − 1 . Two cases are shown in Fig. 10.15(a, b) for n-SC.
In the case, (i) α − 1 < LSC , the light is absorbed in the narrow vicinity of the space
charge layer. Photogenerated pair e–h+ is separated and carriers moved in the electric
field. The flux of holes to the interface SC–electrolyte with the redox system is con-
stant and maximal. It means that the changes in φSC do not influence the photocurrent
and the photocurrent should be independent of the potential. In the other case, (ii)
α − 1 > LD , where LD is Debye length;

LD = LSC + Ldiff ; Ldiff is the diffusion length,

the carriers are generated in all vicinity of deep light absorption. The direction of the
electric field at the interface of SC–electrolyte is such that the minority carriers gener-
ated in the space charge layer move to the interface, participating in the reaction with
252 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

(a) (b)

Ec Photon Ec
h𝜈 Photon
h𝜈

Ev
Ev

x = ∝–1

Ldiff LSC Ldiff LSC

LD = LSC + Ldiff LD

Fig. 10.15: Correlation between the depth of light absorption x, the thickness of the space-charge
layer, LSC , and diffusion length Ldiff . In case (a) x < LSC ; in case (b) x > LD . Adapted from [3].

a redox couple. The minority carriers generated beyond the depletion layer LSC , but
in the vicinity of LD , move in the direction of interface by diffusion, and those that
were generated even deeper than LD recombine. The characteristic length Ldiff (in
short cut Ld ) is equal to Lp for n-SC, or Ln for p-SC and is defined by the equations:
pffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Lp = Dp τp = kB Tμp τp (10:36)
pffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Ln = Dn τn = kB Tμn τn (10:37)

where Dp , Dn are diffusion coefficient of holes and electrons; τp , τn are lifetime of

holes and electrons; and μp , μn are mobility of holes and electrons, respectively.
There are few models applied for the description of the photocurrent generated
at the SC–electrolyte interface. We will apply the Gärtner model introduced first for
the description of photocurrents generated under the illumination of SC–metal junc-
tion [9, 10]. In this model, the flux of minority carriers, generated under illumination,
to the interface (junction) is the sum of two fluxes, one in the depletion region of
SC, jSC , and the other in the diffusion zone, jdiff , where the minority carriers near the
depletion layer diffuse to this zone before the recombination. Then we can write an
equation for the density of photocurrent:

jph = jSC + jdiff (10:38)

The photocurrent generated in the depletion state of the space charge layer is de-
scribed by the equation:
 10.4 Photoeffects at/in semiconductor electrode in electrolyte 253

jSC = eI0 ð1 − exp − αLSC Þ (10:39)

The photocurrent originated from the diffusion is given for n-SC (minority carriers –
holes) by the expression:


αLp
jdiff = − eI0 exp − αLSC (10:40)
1 + αLp

The photocurrent (eq. 10.38) generated at n-SC–electrolyte interface:


exp − αLSC
jph = − eI0 1 − (10:41)
1 + αLp

for p-SC, Ln should be used in eqs. (10.40 and 10.41).

The photocurrent depends upon LSC . Taking into account the dependence of LSC
on φSC (eq. 10.17) and the relation between the potential drop in the space charge
layer of SC and the electrode potentials φSC = E − Efb , we obtain the relation between
the photocurrent and n-SC electrode potential:
2 3
− α CðE − Efb Þ1=2
exp
jph = − eI0 41 − 5 (10:41)
1 + αLp


2ε0 εSC 1=2
where C=
eND

The total current that flows across SC electrode under illumination is the sum of photo-
current and the current of majority carriers. The latter is described by eq. (10.29b, 10.30b)
since the changes of minority carriers concentration are very small.
The expressions are as follows:


− FU
j = jph + j0, CB 1 − exp , n−SC (10:42a)
RT


FU
j = jph + j0, VB exp −1 , p−SC (10:42b)
RT

The photogenerated positive (anodic) current jph results from the flux of generated holes
and dominates; when the n-SC electrode is reverse biased (polarized), at forward bias a
small cathodic current is observed. For p-SC the photocurrent would be negative (ca-
thodic), dominating at reverse bias and at the forward bias a small anodic current would
be observed. The shape of the current–potential curve is shown in Fig. 10.16.
254 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

n-SC j (mA/ cm2) p-SC j (mA/ cm2)

Efb Efb
U (V) U (V)

Fig. 10.16: Current density–potential characteristics for semiconductor electrode under illumination.
Note that we used symbols Efb and Ufb interchangeably; they have the same meaning and value.

In photoelectrochemistry the term “quantum yield” ∅ is applied. ∅ is defined as the

ratio of photocurrent jph to the intensity of light I0 illuminated the SC electrode:

jph
∅= (10:43)
I0

We can introduce this notation in eq. (10.41) and rewrite the expression in logarithmic
form [10]:

− lnð1 − ∅Þ = αCðE − Efb Þ1=2 + lnð1 + αLd Þ (10:44)

The equation pointed out that the graph − lnð1 − ∅Þ versus ðE − Efb Þ1=2 should be lin-
ear and the intercept of the plot on the potential axis when − lnð1 − ∅Þ = 0 gives the
2ε0 εSC 1=2
value of Efb . The slope of the graph gives the value of αC, where C = ð eN x
Þ ,
Nx is ND or NA depending on the type of doping. If α is known from other measure-
ments, the concentration of dopants can be determined. If the concentration of dop-
ant is known from capacitance measurement (Section 10.2), we can determine the
absorption coefficient. For such purpose, the SC–electrolyte interface should be il-
luminated with monochromatic light. Upon such measurements, using the light of
different wavelengths λ, we can determine the dependence of α on λ. From the in-
tercept on − lnð1 − ∅Þ (y-axis), we can obtain a magnitude of lnð1 + αLd Þ and calcu-
late the diffusion length of minority carriers, Ld , and by using eqs. (10.36) and
(10.37) the lifetime of carriers can be assessed.
The photocurrent measurement and application of the Gärtner model give us an
opportunity to determine many important parameters. The main weakness of the
Gӓrtner model is that it neglects the surface recombination.

10.5 Influence of surface states

So far, we have assumed that the free charge carriers are directly transferred from
the conduction or VB to the redox couple in the electrolyte and that all the changes in
the potential at the interface SC|electrolyte occur in the space charge layer of SC.
 10.5 Influence of surface states 255

However, in these processes at the interface, the surface state can play an important
role. Two types of surface states can be distinguished: intrinsic and extrinsic surface
states. The intrinsic surface states are due to the SC solid-state nature and result from
discontinuity and termination of crystal structure at the surface (Tamm state). The
other intrinsic states (Shockley) arise as a result of the covalent bond breaking, the
creation of dandling electron bonds and radical states. The extrinsic surface states
issue from surface defects, adsorbates, oxide layer, and interface formation. In the
case of the SC|electrolyte interface, the ions, reaction products, oxides, and other sur-
face species may be responsible for such states. The distribution of surface states may
be described usually by three models, two of them consider the fluctuation of distri-
bution of energy (Gaussian model, exponential model). We will consider the simplest,
monoenergetic model. In this model, the surface states are described by one localized
energy level, Ess , of electrons, which can lose or capture one electron. The energy
level lay in the bandgap energy (Fig. 10.17) and can be filled with electron (donor
level) or be empty (acceptor level).

E (eV)

EC
EF
Ess

Fig. 10.17: Simplified energetic model of surface states

EV
characterized by one localized energy level ESS of electrons.

The surface states are characterized by the density of surface states NSS , the density
of charge in the surface states Qss , and Fermi distribution function fss .The density
charge is given by
Nss
Qss = eNss fss =  (10:45)
E −E
1 + exp ssk T F
B

In the presence of surface states, the total charge of SC part of the double layer is the
sum = QSC + QSS , so the resultant capacitance of SC part of the interface is equal to the
two capacitors connected in parallel C = CSC + CSS . The scheme of the equivalent electri-
cal circuit describing the SC|electrolyte capacitance is shown in Fig. 10.18.

CSC

CG-Ch CH

Fig. 10.18: Scheme of equivalent electrical circuit showing

CSS semiconductor–electrolyte capacitance.
256 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

It was mentioned earlier that a large amount of surface states change the potential
distribution at SC–electrolyte interface and can cause the “metallization” of SC elec-
trode ð jΔφH j  jΔφSC j Þ. The relation points out that any changes of the potential of
SC electrode occur in the Helmholtz layer. To observe such changes in the potential
distribution at SC|electrolyte interface, the density of surface states should be larger
than 1013 cm−2 [3, 4]. Except for the changes in the potential distribution, the surface
states can affect the kinetics of electrode reactions, changing the pathway of reaction.
They can participate in the recombination and charge carriers trapping, decreasing
the photocurrent and in consequence the efficiency of conversion of solar energy in
photelectrochemical cells (PEC). In Section 10.4, the Gӓrtner model, which neglects the
surface and space charge recombination, was applied for photocurrents determination.
Many efforts have been made to improve the Gӓrtner model; here we will briefly de-
scribe Peter’s model [11]. This model is presented in Fig. 10.19.

E(eV) h𝜈
Jn
Ec kind,tr
krec
Ox/Red
ktrap kd,tr Fig. 10.19: Scheme illustrating the involvement of surface
Ev states in recombination, trapping and direct and indirect
Jh
(V) charge transfers. Semiconductor electrode under
Semiconductor Electrolyte illumination. Adapted from [11].

Let us consider the n-SC electrode in an electrolyte solution with redox couple, which
causes the formation of the depletion state in the space charge layer of SC. SC has the
surface states with energy levels in the vicinity of bandgap energy. We illuminate the
electrode (such electrode works as photoanode in the PEC) and generate charge carriers:
electron and holes. The flux of minority carriers (holes) photogenerated within Debye’s
length, LD , ðLD = LSC + Ldiff Þ, arrives at the SC–electrolyte interface (more detailed de-
scription you can find in Section 10.4). Furthermore, there are few steps that should be
considered: (1) direct transfer of photogenerated holes from VB to Red form of a redox
couple, and (2) indirect transfer of photogenerated holes from the VB to Red form of a
redox couple via the surface states, which in this case work as traps. The photogenerated
holes can also recombine via the surface states with electrons from the CB. In this case,
the surface states play the role of recombination centers. All these processes: charge
transfer, trapping, and recombination occur with different rates, characterized by rate
constants: kd, tr , kind, tr , ktrap , krec and compete with each other. Let us consider two lim-
iting cases. In the first case, all photogenerated holes that arrive at the surface partici-
pate in direct charge transfer between the VB and reduced species in the solution. The
rate of direct charge transfer is characterized by the rate constant kd, tr and the photo-
current is determined by flux of holes dp dt = Gp (generation rate of holes) and expressed
by the Gӓrtner equation. In the second case, considered by Peter [11], all holes that ar-
 10.5 Influence of surface states 257

rive at the surface are rapidly trapped by surface states and further participate in the
indirect charge transfer or in the recombination. The flux of holes across the interface is
then given by

dps
= Gp − kind, tr ps − krec ps (10:46a)
dt
where ps is the density of holes at surface equal to the density of trapped holes, kind, tr
is rate constant of indirect charge transfer from traps (surface states) to reduced form
in solution, kind, tr = ktrap cRed , krec is the recombination rate constant, and ktrap is rate
constant of trapping.
dps
Under the condition of steady-state illumination, = 0, taking into account eq.
dt
(10.46a)

Gp
ps = (10:46b)
kind, tr + krec
kind, tr
jph = ekind, tr ps = eGp (10:46c)
kind, tr + krec

This equation points out that trapping and recombination at the surface states de-
crease the photocurrent.
The above models do not consider the recombination, which can occur in the vi-
cinity of the space charge layer of SC as a result of dopants and lattice defects.
Taking into account all recombination processes (bulk and surface) it is clearly
seen how important is (i) the determination of parameters characterizing the minor-
ity charge carriers, such as the lifetime, mobility, and the diffusion length, and (ii)
the determination of kinetic rate constants of recombination and charge transfer
processes.
Some methods were lately developed: the intensity-modulated photocurrent spec-
troscopy and open-circuit voltage-decay, which provide some insights into the surface
carrier dynamics. Interested readers will find more on this topic in references [12, 13].
In the last years, many efforts have been made to avoid recombination processes. Few
types of approaches were developed: (i) nanostructuring of SC material on a scale com-
parable, or even smaller than, the width of space charge layer, (ii) application of cocata-
lyst to remove the kinetic limitations, and (iii) optimization of deposition procedure and
surface treatment (etching, thermal treatment). However, the electrochemical treat-
ment can, on the one hand, decrease the surface recombination but, on the other
hand, can facilitate the photocorrosion. The photocorrosion is another process that
should be eliminated in photoelectrochemical cells. The description of the photocorro-
sion processes and methods of protection against them will be addressed in the fol-
lowing section.
258 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

10.6 Corrosion, photocorrosion, and stability

Besides the trapping and recombination of minority carriers via surface state, the sta-
bility of the SC electrode is a crucial problem in photoelectrochemical cells. The SCs
like the metals can go spontaneous decomposition (oxidation, dissolution) as a result
of chemical, electrochemical reactions, that is, corrosion. The approach used for the
description of metal corrosion can be also applied to the SC materials (Chapter 3). The
main difference in corrosion of metals and SCs results from the possibility of partici-
pation of two charge carriers: electron and holes in corrosion of SC [3]. Let us consider
the decomposition of MX SC. We can write the reactions of decomposition with partic-
ipation of electrons from the CB (cathodic reaction) and holes from the VB (anodic
reaction) as follows:
n−
MX + ne + solv ! M + Xsolv (10:47a)

MX + nh + + solv ! Msolv
n+
+X (10:47b)

In the corrosion, the conjugated reactions with those above are: H2 − 2 e ! 2 H +

(with 10.47a) and 2 H + + 2 e ! H2 (with 10.47b).
Based on the thermodynamic data and calculation of Gibbs free energy changes
of reactions (10.47a, b), we can determine the standard equilibrium potential of de-
0 0
composition of MX with participation of electrons Edec, n and holes Edec, p . These reac-
tions can be also described in energy scale by electrochemical potential (Fermi
energy) EF, dec, n and EF, dec, p , respectively.
The thermodynamic criteria point out that corrosion or the cathodic decomposi-
tion of the SC compound occurs, with the participation of electron from the CB, if the
0
potential of the electrode will be lower than Edec, n . In the case of decomposition with
holes from the VB, the corrosion or anodic decomposition is expected, if the potential
0
of the electrode is higher than Edec, p:

E < Edec,
0
n, E > Edec,
0
p

Schematically, the thermodynamic stability related to the decomposition of SC, in the

condition of flat band potential, is shown in Fig. 10.20.
In this figure, four cases are presented, the SC is (a) stable with respect to cathodic
and anodic decomposition, (b) stable to cathodic decomposition, (c) stable to anodic
decomposition, and (d) unstable and can undergo cathodic and anodic decomposition.
The decomposition of material significantly increases, if the SC electrode is illumi-
nated. In Fig. 10.21, the influence of illumination on SC corrosion rate is presented.
It is clearly seen that illumination results in an increase in corrosion cur-
rent density, ðlnjcorr, d , lnjcorr, l Þ – means corrosion current density in dark and under
illumination.
 10.6 Corrosion, photocorrosion, and stability 259

(a) E (eV) (c) E (eV)

EF,dec,n EF,dec,n
EC EC
EF EF

EF,dec,p
Ev Ev
EF,dec,p

(b) E (eV) (d) E (eV)

EC EC
EF EF
EF,dec,n EF,dec,n

EF,dec,p
Ev Ev
EF,dec,p

Fig. 10.20: Correlation between the position of band energies EC , EV and decomposition potentials in
energetic scale EF, dec, n , EF, dec, p for n-type semiconductor: (a) electrode is resistant to decomposition by
electrons and holes, (b) electrode is not resistant to decomposition by electrons and holes, (c) electrode is
not resistant to decomposition by holes, and (d) electrode is not resistant to decomposition by electrons.

n-type p-type
E (V) E (V)

Edec,p 2 ERedox 3' 3 4

1'
1
Ecorr, I Ecorr, d
Ecorr, d Ecorr, I

ERedox Edec,n

Inj Inj
Injcorr, d Injcorr, I Injcorr, d Injcorr, I

Fig. 10.21: Influence of illumination on the corrosion of n-type and p-type of SC. Curves 1 and 3 show
corrosion of both semiconductors in dark. Curves 2 and 4 show corrosion of semiconductor under
illumination. Note that corrosion current density under illumination is large than corrosion current density
in dark. Adapted from [3].

Let us consider the few cases concerning the n- and p-SCs immersed in the electrolyte
solution with the redox system. The redox system was chosen in such a way that the
depletion layer was formed in the space charge layer of SC. Such a situation is spe-
cially created in SC electrodes working as photoanodes and photocathodes in the pho-
toelectrochemical cells. Note that the behavior of SC electrodes in redox electrolyte
and under illumination was described Sections 10.3 and 10.4.
260 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

In Fig. 10.22 four cases are presented for n- and p-SC.

(a) E (ev) (b) E (ev)

e e

EC Ec
EF,Redox EF,dec,p
EF
EF EF,Redox
Eg *
pEF
EF,dec,p Eg *
pEF

Ev Ev
n-type + E (v) n-type E (v)
h h+

(c) E (ev) (d) E (ev)

e e
Ec Ec

* *
nEF EF,dec,n n EF
Eg
Eg EF EF,Redox
EF EF,Redox
Ev EF,dec,n
Ev
p-type p-type
+ h+
h
E (v) E (v)

Fig. 10.22: Energetic diagrams presenting the thermodynamic stability of n- and p-SCs and their
protection against photocorrosion by the redox system. The n-SC is protected against photocorrosion
when EF, Redox > EF, dec, p (a); p-SC is protected when EF, Redox < EF, dec, n (b). Parts c, d point out that
semiconductors undergo photocorrosion.

In the case of n-SC (a, b) only holes can participate in the reactions. They oxidize ei-
ther the SC or the Red form of a redox couple in the solution. The reaction with the
participation of electrons is not possible since their concentration in the CB, under
such condition, is low. Which reaction dominates at SC–electrolyte interface depends
on the relations between Fermi energies of redox systems and decomposition of SC,
or on the relations between the corresponding potentials. In the case (a) and (b) the
following reactions can occur:

MX + nh + + solv ! Msolv
n+
+X
Red + nh + ! Ox

If EF, Redox > EF, dec, p , or ERedox < Edec,

0
p , the oxidation of reduced specious is preferable
(a), and if EF, dec, p > EF, Redox , or Edec, p < ERedox , a decomposition of n-SC takes place (b).
0
 10.6 Corrosion, photocorrosion, and stability 261

For p-SC, since the concentration of holes is very low in the depletion layer, the
electrons participate in the reduction of the oxidized form of redox couple or in the
reduction of SC materials:
n−
MX + ne + solv ! M + Xsolv
Ox + ne ! Red

If EF, Redox < EF, dec, n , or ERedox > Edec,

0
n the reduction of Ox form takes place (c); if
EF, dec, n < EF, Redox , or Edec, n > ERedox the decomposition of p-SC occurs (d).
0

Thus, we can conclude that for the protection of SC against photocorrosion, we

have to apply for photoanode the redox systems: ERedox < Edec, 0
p and for photocathode
the redox system: Eredox > Edec, n .0

The diagram of electrochemical potentials for few SCs with marked decomposi-
tion potentials is shown in Fig. 10.23.

E (eV)
CdSe CdS CdTe GaAs TiO2
–3.0 –1.5
–1.0
–4.0 –0.5
0.0 1.4 0.0 V
1.5
–5.0 +0.5 1.7 2.25 eV eV
+1.0 eV eV
–6.0 +1.5
+2.0 3.2
–7.0 +2.5 eV Conduction band
+3.0
–8.0 +3.5 Valence band

E (V), SHE

Fig. 10.23: Correlation between the position of conduction and valence band edges EC ,EV and the
decomposition potentials (black short line). All presented semiconductors are unstable and may be
oxidized by holes during illumination.

As examples of the thermodynamic approach, we will consider the CdX (X = S, Se, Te)
electrodes that are applied as photoanodes in photovoltaic liquid cells (Chapter 11). The
data used here are taken from books [3, 14] and references therein. The photocorrosion
reaction can be written:

CdX + 2 h + ! Cd2 + + X, 2 H + + 2 e ! H2
0
The calculated decomposition potentials Edec, p are equal to 0.32 V (n-CdS), 0.12 V
(n-CdSe), −0.08 V (n-CdTe) versus NHE. It was proved experimentally that such redox
systems such as S2 − jS2n − , Se2 − jSe2n − ,Te2 − jTe2n − with standard redox potentials −0.48 V,
−0.71 V, −0.81 V versus NHE almost fully protect the electrodes against photocorrosion.
In these redox systems, the values of flat band potentials, Efb are very negative and
lay in the potential range from −1.3 V (n-CdS) to −1.7 V (n-CdTe) versus NHE. Since the
262 10 Semiconductors electrochemistry and photoelectrochemistry: fundamentals

difference between redox potentials of these couples and the flat potentials of CdX is
large, the φSC and in the consequence Eph oc
should be large, too. However, the solutions
of these redox systems are colorful, and so a significant part of light energy is lost.
Other solutions containing SO23 − , S2 O23 − , FeðCNÞ46 − ions were tested. It was found that
their efficiency in the protection of the CdS photoanode is about 95–85%.
The quasi-thermodynamic model can be used for description of all types of the SC
decompositions: corrosion in dark, dissolution under the external applied potential
and photocorrosion. However, this approach does not provide full insight into the de-
composition phenomena. We have to keep in mind the kinetics of these processes. It
is difficult to make some generalization since the kinetics depends on the crystal
structure, electronic structure, and surface states, their influence should be deter-
mined individually.
Like in the case of metals, some SCs can protect themselves against corrosion under-
going passivation. A good example is silicon, which in the majority of solution oxidizes
and is covered with a protective oxide films. What is interesting is its ability to go from a
passive state to an active state under illumination [3]. Apart from usage of redox systems
and passivation, we can also protect the SC electrodes against decomposition by deposit-
ing the thin layers of conducting polymers on their surface. Polymers such as polypyrrole,
polythiophene, polyaniline, and others were applied. The main problem is the deposition
of CP on the SC surface. Polymers can be deposited chemically, electroless-deposited, or
electrochemically. In the two first methods, the strong oxidants have to be used for
monomer oxidation and further polymerization. In electrochemical deposition, the
anodic potentials of monomer oxidation are large and can oxidize and dissolve the
SC. For instance, the oxidation potential of pyrrole, thiophene, and 3-methylthiophene
are 0.8 V, 1.6 V, 1.35 V versus SCE in acetonitrile solution, respectively. This problem was
partially solved by the application of soluble polymers such as polyaliphatic-thiophenes
and polyalkoxy-thiophenes. Interested readers can find more about conducting polymers,
their synthesis, and applications in [15]. All this fundamental information introduced in
this chapter will be applied in the description of the electrochemical photocells and their
applications in the solar energy conversion.

Bibliography

[1] Kittel C. Introduction to solid state physic. 8 ed. Hoboken, USA, John Wiley & Sons Inc, 2005.
[2] Rosenberg HM. The solid state. 2 ed. Oxford, UK, Clarendon Press, 1978.
[3] Pleskov YV, Gurevich YY. Semiconductor photoelectrochemistry. NY,USA, Plenum Publishing
Corporation, 1986.
[4] Bard AJ, Faulkner LR. Electrochemical methods. Fundamentals and applications. NY, USA, John Wiley
&Sons Inc, 2001.
[5] Meming R. Semiconductor electrochemistry. 2 ed. Weinhaim, FRG, Wiley VCH Verlag GmbH, 2015.
 Bibliography 263

[6] Rajeshwar K. Fundamentals of semiconductor electrochemistry and photoelectrochemistry. In ed.,

Licht S. Semiconductor electrodes and photoelectrochemistry. Vol 6. In Bard AJ, Stratman M., ed.
Encyclopedia of electrochemistry. Weinheim, FRG, Wiley VH Verlag GmbH, 2002, 1–53.
[7] Gerischer H. Semiconductor electrode reactions. In: ed., Delahay P, Tobias CW. Advances in
electrochemistry and electrochemical engineering, Vol 1. NY, USA. Interscience, 1961, 139–232.
[8] Fuller TF, Harb JN. Electrochemical Engineering. Hoboken, USA. John Willey& Sons Inc, 2018.
[9] Peter L.M. Semiconductor photoelectrochemistry. In: Pletcher D. Electrochemistry. Vol.9, London,
United Kingdom, RSC, 1984, 66–100.
[10] Sharon M. The photoelectrochemistry of semiconductor/electrolyte solar cell. In: Licht S., ed
Semiconductor electrodes and photoelectrochemistry. Wiley-VCh. Verlag Gmbh, Weinheim, FRG,
2002, 287–316.
[11] Peter LM. Photoelectrochemical kinetics at semiconductor electrodes. In ed., Compton RG,
Hancoock G. Application of kinetic modeling. Vol.37 Comprehensive chemical kinetics. Amsterdam,
The Netherlands, Elsevier, 1999, 223–280.
[12] Gurudayal, Peter LM, Wong LH, Abda FF. Revealing the influence of doping and surface treatment
on the surface carrier dynamics in hematite nanorod photoanodes. ACs Appl. Mater. Interfaces,
2017, 9, 41265–41272.
[13] Zaban A, Greenshtein M, Bisquert J. Determination of the electron lifetime in nanocrystalline dye
solar cells by open-circuit voltage decay measurements. Chem. Phys, Chem., 2003, 4, 859–864.
[14] Bouroushian M. Photoelectrochemistry and application. In: Electrochemistry of metal
chalcogenides. Berlin, FRG, Springer Verlag, 2010, 207–308.
[15] Skotheim TA, Reynolds JR ed Conjugated polymers. Theory, properties and characterization CRC
Press USA 2007.
11 Solar energy conversion
in photoelectrochemical cells

11.1 General remarks

There are two main reasons for the rapidly increasing interest in the application of solar
energy in our everyday life: first, the exhaustion of fossil fuel resources (e.g., oil, coal, and
gas), and second, the problem with environmental pollution resulting mostly from the
application of natural fuels for electric energy production. The Sun is the cleanest, most
abundant, and practically inexhaustible energy source. The power that the Sun continu-
ously delivers to the Earth is about 1.2 × 105 TW (terawatts), while our civilization produ-
ces and uses currently 13 TW, and about 80% of it comes from fossil fuels. Sunlight
energy can produce heat or can be converted to electricity (solar to electricity) by apply-
ing photovoltaic (PV) cells. The best PV commercial solar cell based on crystal silicon has
a conversion efficiency of about 18%, and for the laboratory silicon cell 25% efficiency
was reached, being close to the theoretical limit of 31%. This cell has a single p–n junction,
so it can capture only a small part of the solar spectrum. The effort is made to construct
stacked cells with different bandgaps that enable to capture a greater fraction of solar
energy. The calculated efficiency limit is then 43% for two p–n junctions, and increases
reaching 66% for infinite numbers of junctions [1]. The cheaper solar cells can be made
from other materials but their conversion efficiency is not so satisfied. It is worth to men-
tion the DSSC (dye-sensitized solar cells). The laboratory solar cells based on dye sensitiza-
tion of oxide semiconductors are typically 10–11% efficient. We will describe this type of
cell in Section 11.3. In other solar cells, the junctions of semiconductor nanoparticle – con-
ducting polymers – are applied. Up till now, their efficiency is not so large, about 3–4%,
but they are flexible, can have large active surface, and their formation is not expensive
[2, 3]. The surprising increase of efficiency comes from quantum-dot phenomena result-
ing in the multiplication of electron–hole pairs for a single incident photon [4].
The cells for solar energy storage are also intensively developed (STF – solar to
fuel). In these cells, solar energy is converted into chemical energy and stored in the
reaction products. The photocells where the splitting of water or the reduction of CO2
occurs are such examples.
We can distinguish two types of cells: solid cells and liquid cells. The solid cells
are PV cells containing n-SC/p-SC junctions or Me/SC junctions or multijunctions of dif-
ferent types. They convert solar energy into electrical energy. The liquid cells are re-
ferred to as photoelectrochemical cells (PEC) and will be described in detail later.
They can convert solar energy into electricity (PV cells) or can store the energy in re-
action products (photoelectrosynthetic cells).

https://doi.org/10.1515/9783111160986-012
 11.2 Efficiency and key parameters 265

11.2 Efficiency and key parameters

There are plenty of factors that influence the efficiency ε of solar energy conversion
in the solar cells, such as thermodynamic efficiency, threshold efficiency, charge car-
riers separation efficiency, and storage efficiency. It is very difficult to determine
these factors experimentally. Because of that, other quantities that are measured and
oc
determined are open-circuit photopotential Eph , short-circuit current density under
ph
illumination jsc, c , fill factor ff, and external quantum conversion efficiency EQE. The
basic parameter determining the optional energy conversion is the thermodynamic
efficiency, coming from the Carnot cycle. For solar energy conversion, εtherm is ap-
proximately 0.7. The conversion of solar energy is the threshold process, which
means that only a part of quantum energy hν = Eg is absorbed by the direct semicon-
ductor, and for indirect SC hν > Eg . The term threshold efficiency, εthresh , means the
ratio of the number of absorbed light quanta of a proper energy to the total number
of light quanta striking the surface of SC. The important quantity is quantum yield ∅,
which characterizes the conversion efficiency of the incoming photon flux of energy
hν ≥ Eg into electrical current (Section 10.4). Quantum yield is connected with the effi-
ciency of separation of charge carriers generated under illumination. This ability
manifests itself in the value of short-circuit current density, jph sc, c , under illumination.
The efficiency of the carriers collection is determined by the changes of electrochemi-
cal potential of minority carriers and manifests itself as open-circuit photopotential,
OC
Eph . Another important parameter is the external quantum conversion efficiency
EQE. This value indicates the photocurrent produced by the cell when illuminated
with photons of a particular wavelength and does not take into account the losses as-
sociated with reflection of incident photon and recombination of charge carriers. The
quality of the solar cell is characterized by a fill factor:

max max
jph Eph
ff = ph
(11:1)
OC
jsc, c Eph

All these parameters can be obtained from the working characteristics of a photocell
(j vs. U) as shown in Fig. 11.1(a). The fill factor is near unity when the working charac-
teristic is rectangular (case 1, Fig. 11.1(b)). The shape of the characteristics is changed
when the internal resistance of the cell is large (case 2, Fig. 11.1(b)).
The power conversion efficiency can be determined by the expression
ph
Pmax jmax max
ph Eph
OC
jsc, c Eph
ε= = = ff (11:2)
Pin Pin Pin

To compare different solar cells, these crucial parameters (eq. (11.2)) should be deter-
mined in the standard conditions (STC – standard test conditions).
266 11 Solar energy conversion in photoelectrochemical cells

(a) (b)

j (mA/cm2) ph
j (mA/cm2) jsc,c,1

max Pmax,1

nt
jph,1

rre
cu
jsc,c,2

rk
Da
OC
Eph U (V) Pmax,2
max
jph,2

jph
max max
Eph , jph max ph max ph
ph Eph,2 Eoc,2 Eph,1 Eoc, 1 U (V)
jsc, c

Fig. 11.1: (a) Current density–potential characteristics of a photocell in dark and under illumination.
Parameters for fill factor determination are marked. (b) Scheme illustrates the influence of the internal
resistance of photocell on the current–potential characteristics. If the internal resistance is low, the power
Pmax, 1 is high. Increase of internal resistance of the photocell causes decrease of power − Pmax, 2 .

The STC represents a temperature of 25 °C and an irradiance of 100 mW/cm2 with an air
mass 1.5 spectrum (AM 1.5). The standard reference solar spectrum AM 1.5 defined by the
American Society for Testing and Materials characterizes the effect of the Earth’s atmo-
sphere on the solar radiation and corresponds to a solar zenith angle of 48.2°. Figure 11.2
shows the dependence of spectral irradiation on wavelength λ and energy.

W/m2 nm UV vis IR
Spectral irradiance

500 850
300 700 1,000
λ (nm)
E (eV)
4.1 1.8 1.0
2.5 1.2

Fig. 11.2: Distribution of solar spectrum irradiance upon wavelength (λ, nm) and energy (eV).

It is clearly seen that semiconductors applied in a photocell should have the bandgap
energies in the visible range of light. All these parameters described above characterize
the PV cell, where solar energy is converted to pure electrical energy.
 11.3 Photoelectrochemical cells: classification – principle of operation 267

The solar energy can also be transformed into chemical energy and stored as fuel.
If solar energy is used only for fuel formation (STF), the power efficiency can be deter-
mined by the equation:
ph
Pout jsc, c ΔE0 ηF
εSTF = = (11:3)
Pin Pin

Pout is the output power of the photoelectrosynthetic cell and Pin the input solar power,
c the short-circuit photocurrent density, ηF Faradaic efficiency, and ΔE the
ph 0
jsc,
difference of the standard potentials of half-reactions occurring at the anode and
cathode.
To determine the solar to hydrogen conversion efficiency εSTH of photo-driven
water splitting, we applied in eq. (11.3), the difference of standard potentials of the
hydrogen and oxygen evolution ðΔE0 = − 1.23 VÞ or the Gibbs standard free energy of
reaction: H2 O ! H2 + 1=2 O2 , Δr G0 = 237 kJ=mol [5].

11.3 Photoelectrochemical cells: classification – principle

of operation

It was mentioned in Section 10.3 that when SC is immersed in an electrolyte solution

containing a redox system, the space charge layer is created in SC at the interface.
The state of this layer depended on the correlation between the Fermi level of SC
and Fermi level of the redox system. Let us suppose that the depletion state is
formed, which is the best state for energy conversion. Generated under illumina-
tion, the electron–hole pairs are separated in the electric field of space charge layer
(n-SC), and electrons flow to the counter (auxiliary) metal electrode and holes to the
interface SC/electrolyte, where electrons reduce the Ox form and holes oxidize the
Red form of redox couple. Depending on the reactions that take place in the cells
and their Gibbs free energy, Δr G; we distinguish usually three types of photocells:
(i) Electrochemical PV cells, Δr G = 0
(ii) Photoelectrosynthetic cells, Δr G > 0
(iii) Photocatalytic cells, Δr G < 0

11.3.1 Liquid photovoltaic cells

The electrochemical PV cell called sometimes “liquid junction photocell” or “regenera-

tive photocell” may contain one or two SC electrodes. In the first case, when n-SC or p-
SC is used under illumination, electrodes work as photoanode or photocathode. In
the second case, two electrodes are photoactive, so the total open-circuit photopoten-
OC
tial of the cell is the sum of two electrode photopotentials. The maximum value Eph
268 11 Solar energy conversion in photoelectrochemical cells

may be predicted from the differences of flat band potential of both SC electrodes if
the band edges are pinned:
OC
Eph = Efb ðnSCÞEfb ðpSCÞ (11:4)

In Fig. 11.3(a) and (b), the energy diagram for PV cell in dark and under illumination
is shown.

(a) E (ev)

–eφsc

EF EF, Redox EF , Me
EF, dec, p

Electrolyte
E(V) Metal
n-type

(b)
e E(eV) e
e
–eφsc
Ec*
EF* eE e
ph c EF, Me
EF *
pEF A EF, Redox

Ev* h+ Electrolyte
E(V)
n-type Metal

Fig. 11.3: Energy diagram for liquid photovoltaic cell (regenerative cell, Δr G = 0) in the dark (a) and under
illumination (b). The band of SC is “pinned.” During illumination, the Fermi energy level ðEF Þ and potential
drop in the space-charge layer of SC ðφSC Þ are changed; ηA , ηC mean the overpotentials for anodic and
cathodic reactions, respectively.

In dark, equilibrium is established when Fermi energy levels of semiconductor, metal

electrodes, and redox system become equal: EF, SC = EF, Me = EF, Redox . As a result of equili-
bration of Fermi energy levels, the depletion state of the space charge layer is created,
and the energy bands are bent (look in Sections 10.3 and 10.4). Under illumination, the
bands unbend, the potential drop, φSC , decreases, and the Fermi level of SC is shifted,
which leads to the change in electrode potential. This change is equal to photopotential,
Eph (marked in Fig. 11.3(b) as eEph , in energy scale). When the edges of the energy band
are pinned and φH remains constant, the maximum Eph = Eph OC
and is determined by dif-
ference ðERedox − Efb Þ or φSC in dark. To obtain the large values of Eph OC
in photocell, we
 11.3 Photoelectrochemical cells: classification – principle of operation 269

should choose the proper redox couple and semiconductor to make this difference as
large as possible. The position of decomposition potential of SC ðEdec,
0
p Þ or Fermi en-
ergy level of decomposition by minority carriers ðEF, dec, p Þ is shown in the diagram,
ηA , ηC mean the overpotential for anodic and cathodic reactions. As mentioned in
Section 10.6, the redox system can protect n-SC against the photocorrosion when:

EF, Redox > EF, dec, p ðeVÞ or ERedox < Edec,

0
p ðVÞ

The general schemes of regenerative photocells can be written as

electrolyte electrolyte electrolyte

1Þ n − SC MeðPtÞ; 2Þ p − SC MeðPtÞ; 3Þ n − SC p − SC:
Ox, Red Ox, Red Ox, Red

The reactions that are going under illumination are as follows:

1) n − SC✶ ; Red1 + h + ðSCÞ ! Ox1 ; Pt; Ox1 + e ! Red1

2) p − SC✶ ; Ox1 + eðSCÞ ! Red1 ; Pt; Red1 − e ! Ox1
3) n − SC✶ ; Red1 + h + ðSCÞ ! Ox1 ; p − SC✶ ; Ox1 + eðSCÞ ! Red1

In all cases, the overall reaction is: Red1 + Ox1 $ Ox1 + Red1 , Δr G = 0.
The typical examples of such cells are chalcogenide electrodes such as n-CdTe, n-
CdSe, and n-CdS immersed in the solution containing redox couples: Te2−/Ten2−, Se2−/Sen2−,
or S2−/Sn2−. These redox couples not only form the depletion state in the space charge
layer of SC, but also protect the semiconductor against photocorrosion (Section 10.6).
The highest reported conversion efficiency of about 16% was obtained for single
crystals of n-CdSe in solution containing Fe(CN)63− ions. The efficiencies of 12% and
11% were determined for n-GaAs in Se2−/Sen2− solution and n-CdTe in Te2−/Ten2− solu-
tion, respectively. However, the application of single crystals is not economical;
therefore, plenty of methods such as vapor deposition, chemical deposition, electro-
chemical, deposition, and sol–gel are applied for the formation of thin solid poly-
crystalline or amorphous semiconducting films. These methods have been used to
prepare thin polycrystalline films of n-Cd(S, Te, Se), p-CdTe, p-InP, n-GaAs, and
others. We may treat the liquid PV cell described earlier as the classical cell. More
broad information about such cells can be found in [6–8] and references therein.
Let us now concentrate on the new generation of electrochemical PV cells:
(i) DSSC
(ii) Quantum dot-sensitized solar cells (QDSSC)

11.3.2 Dye-sensitized solar cells

These types of cells were first invented by Grätzel in the late 1990s of twentieth
century [9]. The first laboratory cell contained a titanium sheet covered with the
“fractal” TiO2 film of a high surface area. The surface of the film was further cov-
270 11 Solar energy conversion in photoelectrochemical cells

ered with the ruthenium dye RuL3 (L-2,2′-bipirydyl-4,4-dicarboxylate). All system

Ti/TiO2/RuL3 worked as photoelectrode. As a counter electrode, Pt wire was used.
Cell was filled with an aqueous, slightly acidic solution containing bromide and a
small amount of bromine. The open-circuit photopotential reached 1 V, and the
overall conversion efficiency in full sunlight was 1–2%. It was the beginning. At
present, the efficiency of conversion in dye-sensitized PV cells is 10–12% [10, 11].
How this cell operates and what was changed to reach such efficiency? Figure 11.4
presents the principle of cell operation.

e e
E (eV)
D*
e
e
CB
Ox

Red
+/D
D e
VB

n-type Electrolyte
Dye Metal

Fig. 11.4: Operating principle and energy diagram of a dye-sensitized solar cell. D+/D, dye in a ground and
oxidized state; D* dye in an excitation state; Ox/Red, mediator in electrolyte.

When the SC with high value of Eg is illuminated with visible light, the dye (sensitizer)
molecules absorbed the light and become excited.
If the energy of the excited state of the dye lies above the energy Ec of conduction
band (CB) of SC, the photoexcitation of dye is followed by electron injection into CB.
The oxidized dye molecules are regenerated by a reduced form of redox couple in the
electrolyte solution. The redox system itself is regenerated at a counterelectrode by
electrons flowing by the external circuit from the SC electrode. Let us consider the
cell containing the I−/I3− redox couple in the organic solvent such as acetonitrile or
propylene carbonate with lithium salt as a supporting electrolyte.
The basic reactions that occur during the illumination are as follows:

D + hv ! D✶ ; D✶ ! D + + eðSCÞ
3 − 1
I + D + ! D + I3− ; I3− + 2e ðMeÞ ! 3I −
2 2

The other reactions that take place are: (i) the radiative deactivation of excited dye,
(ii) recapture of electrons from CB by oxidized dye molecules, and (iii) recapture of
 11.3 Photoelectrochemical cells: classification – principle of operation 271

CB electrons by oxidized species of the redox couple. All these additional reactions
decrease the efficiency of DSS cells and should be suppressed.
As mentioned earlier, the first DSS cells contain a planar crystalline TiO2 elec-
trode covered with adsorbed dye. Because of the low surface area and low concentra-
tion of dye, the conversion efficiency was low. The situation was changed when not
only the nanostructured metal oxide, mainly TiO2, but also ZnO, Nb2O3, and SnO2
were applied. The high porosity of mesoscopic semiconductor films enables the in-
corporation of dyes in large concentrations. Nowadays, working DSS cells contain
photoelectrode made from TiO2 (anatase) nanoparticles, deposited by hydrothermal
method on flexible plastic support covered with a transparent conducting layer of
fluorine-doped thin oxide or thin-doped indium oxide. The average size of the
nanoparticles is 20 nm. They are covered and shelled by specially designed light-
harvesting dye molecules. Mostly, Ru(II)bisbipyridyl complexes are used [10]. They are
known as being very stable and possessing the light absorption coefficient α in the
range 2–4·103 cm−1 so the absorption length of these sensitizers, 1=α, is about 2.5–5 μm.
The light-harvesting efficiency (LHE) of these dyes is over 90% for the wavelength
range near the absorption maximum for the thickness of the nanocrystalline film (d)
equal to 6 μm. The LHE is determined by expression:

LHEðλÞ = 1 − 10 − αd (11:5)
As other PV cells, the quality of DSS cell is determined by fill factor ff, the power con-
version efficiency ðεÞ of DSSC (see eqs. (11.1) and (11.2)), and by the external quantum
efficiency sometimes referred to as incident photon to current conversion efficiency.
This parameter depends upon the LHEðλÞ, quantum yield ∅, and collection efficiency
εcoll (EQC). In this case, the quantum yield ∅ determined the electron injection from
excited dye to the CB of SC:

EQCðλÞ = IPCEðλÞ = LHEðλÞ∅εcoll (11:6)

The collection efficiency of charge carriers (here electrons) depends on the relation-
ship between the electron diffusion length Ln , the thickness of the nanocrystalline
layer of SC (d), and the absorption coefficient ðαÞ. The good collection is obtained
1
when Ln > d > . The best conversion efficiency is about 11–12%, so DSS cells can com-
α
pete with conventional solar cells. One can find much more about DSS cells in articles
[10–12] and references therein.

11.3.3 Quantum dot-sensitized solar cells

In the last decade, the huge development of nanotechnology is observed. It creates the
possibility of the application of nanostructural materials in solar cell design. As men-
tioned earlier, in the DSS cells the nanocrystalline metal oxide semiconductors of high
value of bandgap energy were used with the organic dyes as a sensitizer. In QDSSC,
272 11 Solar energy conversion in photoelectrochemical cells

semiconductor oxide nanoparticles, mostly TiO2, are also used. However, as sensitizer,
the other semiconductors are applied (PbS, CdSe, and CdS) [12]. They have lower band
gap energy than oxide support and are used in the form of quantum dots. During illu-
mination, in quantum dots (QDs) the electron–hole pairs are also generated. The main
problem here is the effective separation of charge carriers. For the efficient electron
separation and injection, the bottom level of the CB of QDs should have higher energy
than the bottom level of the CB of oxide SC. After separation, electrons from QDs are
injected to CB of oxide and the holes are transferred to electrolyte oxidizing the re-
duced form of the redox couple. There are some advantages of the application of QDs
as sensitizer when compared with dyes. Most of the dyes have narrow absorption
band, while the QDs can absorb the energy of solar spectrum hν ≥ Eg . Other advan-
tages of QDs are the amount of light that they can absorb and the tunability of the
bandgap energy. When QDs become smaller, their bandgap widens due to the quan-
tum confinement effect. In consequence, it is possible to form the cascade SC photo-
electrode containing the quantum dots with controlled sizes, covering all solar
spectrum. There are some specific features of QDs, which can find application in solar
cell but need special efforts. One feature is the generation of “hot” carriers when QDs
absorbed light of hν > Eg , whereas the other is the carrier multiplication process. It
means that one photon can excite two or more electrons. These two effects may in-
crease the open-circuit photopotential and photocurrent; however, the lifetime of
”hot” carriers (electrons) is very short so some methods should be developed to dis-
solve this problem. Up till now, the QDSSC show smaller fill factors, lower photocur-
rents, and lower power conversion efficiency (about 4%) in comparison with their
dye analogs [12].

11.3.4 Photoelectrosynthetic cells

In these photocells the light energy is applied to initiate and carry out the reac-
tions, resulting in the formation of the new products storing the chemical energy.
The Gibbs free energy of overall reactions that take place in such cells is larger
than 0 ðΔr G > 0Þ. This energetic barrier can be overcome with the help of light en-
ergy. However, in many cases, additional energy of the electric field is required,
supplied by the external source. The most popular reaction in which the rich en-
ergy products are formed is the splitting of water to hydrogen and oxygen. The cell
where the splitting of water occurs is called photoelectrolytic cell. The first photo-
electrolysis of water with the application of the SC electrode was carried out by
Fujishima and Honda in 1972. They used n-TiO2 photoanode and Pt counter elec-
trode in a two-compartment cell separated with diaphragm. During illumination
with ultraviolet (UV) radiation, they observed both O2 and H2 evolution at photo-
anode and counter electrode, respectively. As the energy of light was not suffi-
cient, they had to apply the external voltage to facilitate the reaction. That work
 11.3 Photoelectrochemical cells: classification – principle of operation 273

initiated the huge development of research in the field of the photoelectrolysis of

water and is continued in our days [13, 14]. The overall reaction of water split-
ting is

2 H2 O ! 2 H2 + O2

The standard Gibbs free energy of this reaction ðΔr G0 = 237 kJ=molÞ is equal to 1.23 eV
per electron. In photocells, this reaction is the sum of two half reaction taking part at
semiconductor photoanode (n-SC) and at metal cathode or semiconductor photocath-
ode (p-SC):

2 H2 O + 4h + ! 4 H + + O2 , E0 = + 1.23 V ðNHEÞ, n − SC

4 H + + 4 e ! H2 , E0 = 0 V, Me ðPtÞ or p − SC

Thus, the difference of standard potentials of the water-splitting cell

ΔE0 = EC0 − EA0 = − 1.23 V, at pH = 0ðH + = 1 MÞ

The water oxidation reaction is not spontaneous. To oxidize the water by applying
solar energy the photoexcited electrons must have the energy higher than 1.23 eV.
Such energy is also required for the photons; in the wavelength scale, the energy is
equal to 1,200 nm. Therefore, all solar spectra up to this value can theoretically be
used in the water splitting. In practice, these are the losses in energy resulting from
losses in absorption of light, dissipation of energy in heat form, or some hindrances in
the kinetics. These losses are estimated to be 0.8 V, so the minimal energy of photons
required for the water photodecomposition is about 2.1 eV, that is, <620 nm.
Two types of photocells for water splitting can be distinguished: one is referred to
as direct photoelectrolysis and the other as assisted photoelectrolysis [6, 7]. Their en-
ergy/potential diagrams under illumination are presented in Fig. 11.5(a) and (b). Both
cells contain n-SC electrodes working as photoanodes and Pt cathode. The acid solu-
tion is used as electrolyte. During illumination, the electrons and holes are photogen-
erated and instantly separated in the electric field of the space charge layer. The
electric field causes the flow of the holes to the n-SC/electrolyte interface, where H2O
molecules are oxidized to O2. The electrons flow through the external circuit to the Pt,
where they reduce H+ ions. This is only possible when the value of flat-band potential
Efb versus NHE is more negative than the potential of hydrogen electrode (Eeq of H+/H2)
(see Fig. 11.5(a)).
If the value of Efb is more positive than Eeq of H+/H2 system, the potential for hy-
drogen evolution is not achieved. The additional external potential (UÞ has to be used
to move the potential of Pt cathode to the more negative value than the potential of
the hydrogen electrode (see Fig. 11.5(b)).
In this case, for the splitting of water, not only the light energy is used, but also
the electrical energy. Such a process is called the assisted photoelectrolysis. The other
274 11 Solar energy conversion in photoelectrochemical cells

(a)
e e
E(eV)

Efb
–eφsc e
E *c H+/ H2 EF, Me
c
E*
F *
pE F A O2H2O

hν
E *v h+ Electrolyte
E(V)
n-type Metal

(b)
E(eV) e
Efb H +/ H 2 c EF, Me

U
–eφsc
E *c
E *F O2/H2O
*
pE F A

h+
E *v Electrolyte hv
E(V)
n-type Metal

Fig. 11.5: (a) Schematic energy diagram of photocell under illumination for direct water splitting. The
value of flat band potential ðEfb Þ in a potential scale should be more negative than the redox potential of
H+/H2 couple. (b) Schematic energy diagram of photocell under illumination for assisted water splitting.
The value of flat band potential ðEfb Þ in a potential scale should be more positive than the redox potential
of H+/H2 couple. External potential U is applied.

cells that can be used for photoelectrolysis of water contain two semiconductor elec-
trodes. In such a cell, Pt cathode is replaced by a p-type SC electrode working as pho-
tocathode. The applied n- and p-type semiconductors differ in their bandgap energies:
band gap energy Eg of n-SC should be greater than Eg of p-SC. Under illumination, the
holes generated in photoanode (n-SC) flow to SC/electrolyte interface oxidizing the
water molecules, while the electrons flow to the photocathode, where they recombine
with the holes of the valence band of p-SC. The minority carriers–electrons generated
in photocathode (p-SC) flow to the interface reducing the H+ ions, whereas holes re-
combine with electrons flowing from n-SC. The energy/potential diagram of such cells
is shown in Fig. 11.6.
As described earlier, the photocells for splitting of water contain one or two semi-
conductor electrodes. The choice of semiconductor materials for such purpose is limited
by several requirements. Let us consider these requirements to be fulfilled by the pho-
toanodes (n-type SC):
 11.3 Photoelectrochemical cells: classification – principle of operation 275

E(eV) e E *c
H+/ H 2 Efb
e
Efb nEF
*

E*c
E *F, Me
EF*
O2H2O E *v
*
pE F
h+
E *v Electrolyte
E(V)
n-type p-type
hv

Fig. 11.6: Schematic energy diagram of a photocell under illumination for water splitting. Photocell
contains two photoelectrodes: photoanode (n-SC) and photocathode (p-SC).

1) The bandgap energy of SC should be at optimal values, which means at the energy
that allows for the efficient absorption of visible light. The range of visible light
changes from about 400 to 700 nm, and the corresponding energy ranges from 1.8
to 3.1 eV. It is worth to mention that about 30–40% of solar energy reaching the
Earth`s surface lays in the visible region.
2) The absorption coefficient α should be large, and α is large in the direct band
semiconductors.
3) The doping of SC should be substantial, securing LSC > α − 1 , then the separation of
photocarriers is effective.
4) The diffusion length and lifetime of minority carriers should be large.
5) SC material should be chemically stable in water and in the acid solutions.
6) SC electrodes should be resistive against photocorrosion, so

EF, Redox > EF, dec, p or EF, Redox < Edec,

0
p , Redox − O2 H2 O

7) For the direct photoelectrolysis, the flat band potential of semiconductor electrode
Ufb , or the potential of the bottom of the CB should be more negative than the
redox potential of H+/H2 (vs NHE). The potential of the top of the valence band
should be more positive than the corresponding redox potential of O2/H2O.

Actually, such semiconductor materials that fulfill all these requirements do not exist.
The semiconductors such as SrTiO3, KTiO3, LaTi2O7, and TiO2 have Efb potentials more
negative than the potential of hydrogen electrode (at pH = 0). However, their bandgap
energy is large: 3.3, 3.5, 4, and 3.2 eV, respectively, so they can absorb only UV part of
solar spectra. Unfortunately, UV represents only a small fraction (about 4.5%) of the
sunlight reaching the Earth. The other semiconductors materials such as metal chalco-
genides have appropriate values of the energy gap, for instance, ZnTe (2.3 eV), CdS
276 11 Solar energy conversion in photoelectrochemical cells

(2.4 eV), and ZnSe (2.6 eV). All of them are direct band semiconductor and their flat
band potential is much more negative than the redox potential of H+/H2 (vs NHE).
However, they are not stable in the aqueous acid solution and undergo photocorro-
sion during illumination. For the water splitting, the semiconducting metal oxides are
widely used as the photoanodes. These materials are chemically stable and avoid pho-
tocorrosion during illumination in acid solutions. Unfortunately, most of them have
large bandgap energy and they do not absorb the light efficiently in the visible range.
For instance, the Eg value of TiO2 and ZnO, often applied in the PEC, is 3.2 eV. Better
absorption ability has WO3 and Fe2O3 with Eg equal to 2.7 and 2.2 eV. However, the
flat band potentials of those oxides are about 200 mV or more positive than the poten-
tial of hydrogen electrode.
As mentioned earlier, for the water splitting, instead of Pt cathode we can use semi-
conductor photocathode (p-type SC). In this case, the flat band potential of photocathode
should be more negative than the potential of hydrogen electrode Eeq of H+/H2, since in
the reduction of H+ ions participate the electrons photogenerated in p-SC (Fig. 11.6).
Additionally, the p-type of SC electrode should fulfill the requirements (1–4) de-
scribed above and to be resistive against the photocorrosion, the decomposition po-
tential of p-SC electrode should be lower than the potential of hydrogen electrode:

EF, Redox < EF, dec, n or 0

Edec, n < EF, Redox , Redox − H + jH2

Such cells with two SC photoelectrodes do not require an external source of potential.
The main problem there is the protection of photocathode against the photocorro-
sion, since the semiconductors with low Eg have to be applied to fulfill the condition
Eg (n-SC) > Eg (p-SC). The p-Cu2O (2.2 eV), p-GaP (2.3 eV), and p-WSe2 (1.7 eV) were
tested as photocathodes.
It was pointed out that presently available semiconducting materials cannot af-
ford direct splitting of water in the photocell containing metal cathode. Therefore, the
external source of potential is required. Instead of applying the electrical battery we
can use solid or liquid PV cell, connected in series with PEC [15]. The STH (solar to
hydrogen) efficiency in such tandem cells depends on the design. Up to date, the high-
est STH efficiency reached in the tandem PV-PEC system containing one semiconduc-
tor/electrolyte junction (PEC) is 12.4%.
Up till now, plenty of the semiconducting metal oxides have been reported to be
active for the water splitting; however, most of them absorb UV light. Many efforts
were made to improve the absorption properties of oxides and to shift the onset of
absorption toward the visible range of light. One of the methods is the application of
so-called Z-scheme photocatalysts [16]. A “Z-scheme” can be considered as a special
tandem cell, in which two semiconductors are connected by an electrolyte containing
redox system as mediator. Two photocatalysts are illuminated simultaneously, the
electrons are generated in SC with lower CB, and holes from SC with higher VB react
with redox couple, while the remaining electrons and holes participate in the reaction
 11.3 Photoelectrochemical cells: classification – principle of operation 277

of water splitting. The correlation of energy bands of SC and the equilibrium potential
of the redox system is shown in Fig. 11.7.

e
CB
H+/H 2 H+ H2
–0.41
e
CB

Ox/Red Ox Red
Red Ox

VB
O2/ H2O
+0.82 H2O h+

VB
E(V), NHE O2 h+
pH 7

Fig. 11.7: Schematic energy diagram of the two-step photocatalytic water splitting system (photocatalytic
Z-scheme).

Efforts are made to construct Z-scheme without a liquid redox system, replacing it by
a solid electron mediator (metal nanoparticles). The other approach for improvement
of the absorption properties is the doping of metal oxides with metal or nonmetal
ions (V, Mo, Ru, Li, Si, or others). However, the doping induces the additional energy
levels in the bandgap energy, which may work as recombination centers. There are
several methods for the reduction of the recombination, such as the modification of
SC materials with metal particles (Ag, Au, Pt), the modification of semiconductor
metal oxide with other semiconductor particles, and the application of nanostructures
for photoanodes formation. There is very rich literature concerning the splitting of
water and modifications of applied semiconductors. More can be found in the reviews
[16–19] and references therein.
The solar energy can be exploited in other important reaction – the reaction of
CO2 photoreduction. This reaction can be carried out in the PEC with semiconductor
photocathode, or on the semiconductor particles.
The mechanism of CO2 reduction is complicated and may be multielectron pro-
cess, resulting in the formation of many carbonate compounds [20, 21]. For instance,
eight or 12 electrons are involved during formation of methane or ethanol, respec-
tively. Let us consider only two-electron processes of formation of CO and formic acid
(HCOOH) and six-electron process of methanol (CH3OH) formation. The last two com-
pounds may be used as fuel in the fuel cells. The CO2 reduction can be written as
follows:
278 11 Solar energy conversion in photoelectrochemical cells

1) CO2 + 2 H + + 2 e ! CO + H2 O, Δr G0 = 19.9 kJ=mol, E0 = − 0.1 V,

Eeq ðNHE, pH7Þ = − 0.52 V.

2) CO2 + 2 H + + 2 e ! HCOOH, Δr G0 = 33 kJ=mol, E0 = − 0.17 V,

Eeq ðNHE, pH7Þ = − 0.58 V.

3) CO2 + 6 H + + 6 e ! CH3 OH + H2 O, Δr G0 = 9kJ=mol, E0 = − 0.046V,

Eeq ðNHE, pH7Þ = − 0. 37V.

The values of Δr G0 were calculated using thermodynamic data from physical chemis-
try [22].
The values of redox potentials of reactions are close to each other, so from the
thermodynamic point of view during the CO2 reduction, the mixture of compounds
may be formed. However, the main problem that should be solved is the competition
between the reduction of CO2 and the reduction of hydrogen ions that can occur at
photocathode. The equilibrium potential of the hydrogen electrode at pH 7 is equal to
−0.41 V (NHE). Since the potential of hydrogen electrode is more positive than the
equilibrium potentials of reactions 1 and 2 and is only slightly more negative than for
reaction 3, the minority carriers (electrons) photogenerated in photocathode may par-
ticipate in the reduction of H+ ions. The flat band potential, Efb , for the reduction of CO2
should be therefore more negative than the equilibrium potential of reactions (1, 2, 3) at
a given pH. There are few semiconductors that fulfill such condition: p-Cu2O, p-GaP,
and p-CdTe. However, these semiconductor materials have low values of Eg , so they are
susceptible to photocorrosion. Applying mixed SC oxides with small and large Eg , we
can suppress the photocorrosion.
In the third type of cells (photocatalytic cells), the reactions can occur even in
dark, since their Δr G < 0. However, they proceed so slowly that light energy is utilized
for acceleration. The suspensions of semiconductor particles may be considered as
photo microcells working in the open-circuit conditions.
In the next chapter, we will consider briefly the properties of such systems and
their applications.

Bibliography

[1] Crabtree GW, Lewis NS. Solar energy conversion. Phys. Today 2007, 37–42, www.physicstoday.org
[2] Skompska M. Hybrid conjugated polymer/semiconductor photovoltaic cells. Synth. Met. 2010, 160,
1–15.
[3] Saunders BR, Turner ML. Nanoparticle-polymer photovoltaic cells. Adv. Colloid. Interface Sci. 2008,
138, 1–23.
[4] Nozik AJ. Quantum dot solar cells. Physica E. 2002, 14, 115–120.
 Bibliography 279

[5] Chen Z, Dinh HN, Miller E. Efficiency definitions in the field of PEC. In Photoelectrochemical water
splitting: Standards, experimental methods and protocols. Berlin, FRG, Springer, 2013, 7–16.
[6] Pleskov YV, Gurevich YY. Semiconductor photoelectrochemistry. NY, USA, Consultant Bureau,
Division of Plenum Publishing Corporation, 1986.
[7] Pleskov YV. Solar energy conversion. A photoelectrochemical approach. Berlin, FRG, Springer, 1990.
[8] Sharon M. The photoelectrochemistry of semiconductor/electrolyte solar cell. In : Licht S., ed.
Semiconductor electrodes and photoelectrochemistry. Weinheim, FRG, Wiley-VCh. Verlag Gmbh,
2002, 287–316.
[9] Vlachopoulos N, Liska P, Augustynski J, Gratzel M. Very efficient visible light energy harvesting and
conversion by spectral sensitization of high surface area polycrystalline titanium dioxide films. J. Am.
Chem. Soc 1988, 110, 1216–1220.
[10] Gratzel M. Solar energy conversion by dye-sensitized photovoltaic cells. Inorg. Chem. 2005, 44,
6841–6851.
[11] Mc Evoy AJ, Gratzel M. Dye sensitized regenerative cells. In: Licht S. ed. Semiconductor electrodes
and photoelectrochemistry. Weinheim, FRG, Wiley-VCh. Verlag Gmbh, 2002, 397–406.
[12] Kamat PV, Tvrdy K, Baker DR, Radich JG. Beyond photovoltaic: Semiconductor nanoarchitectures for
liquid-junction solar cells. Chem. Rev. 2010, 110, 6664–6688.
[13] Licht S. Photoelectrochemical solar energy storage cells. In: Licht S. ed. Semiconductor electrodes
and photoelectrochemistry. Weinheim, FRG, Wiley-VCh. Verlag Gmbh, 2002, 317–345.
[14] Sharon m, Licht S. Solar photoelectrochemical generation of hydrogen fuel. In: Licht S. ed.
Semiconductor electrodes and photoelectrochemistry. Weinheim, FRG, Wiley-VCh. Verlag Gmbh,
2002, 345–357.
[15] Prevot MS, Sivula K. Photoelectrochemical tandem cells for solar water splitting. J. Phys. Chem. C
2013, 117, 17879–17893.
[16] Ismail AA, Bahneman DW, Photochemical splitting of water for hydrogen production by
photocatalysis: A review. Sol. Energy. Mater. Sol. Cells. 2014, 128, 85–101.
[17] Kudo A, Miseki Y. Hetergeneous photocatalyst material for water splitting. Chem. Soc. Rev. 2009, 38,
253–278.
[18] Augustynski J, Aleksander BD, Solarska R. Metal oxide photoanodes for water splitting. Top Curr.
Chem. 2011, 303, 1–38.
[19] Li J, Wu N. Semiconductor-based photocatalyst and photoelectrochemical cells for solar fuel
generation: a review. Cat. Sci. Technol. 2015, 9, 1360–1384.
[20] Bockris J O M, Reddy AKN. Modern electrochemistry 2B. Electrodics in chemistry, engineering,
biology and environmental science. New York, USA, Kluwer Academic/Plenum Publisher, 2000.
[21] Fujishima A, Tryk DA. Fundamentals of photocatalysis. In: Licht S. ed. Semiconductor electrodes and
photoelectrochemistry. Weinheim, FRG, Wiley-VCh. Verlag Gmbh, 2002, 497–535.
[22] Atkins P. Physical chemistry. NY, USA, Freeman WH and Company, 1994.
12 Semiconductor particles in photocatalysis

12.1 General remarks

In the previous chapter, we have described how by using photoelectrochemical cells

the solar energy can be converted in the electrical or chemical energy. The reactions
that we considered in detail were the splitting of water and reduction of CO2. There are
many others important reactions for human being such as oxidation or reduction of
inorganic/organic pollutions in water and air. Some of these reactions were investigated
on illuminated, conventional semiconductor electrodes. Now, we will describe the appli-
cations of semiconductor particles in photocatalysis.
Photocatalysis is a kind of heterogeneous catalysis, where the light energy is ap-
plied for acceleration of reaction occurring on semiconductor surface. Depending on
the place where photoexcitation occurs, we can distinguish two types of processes. If
the initial photoexcitation of electrons takes place in the molecules adsorbed on SC sur-
face, followed by injection of photogenerated electrons to the catalyst, the process is
referred to “sensitized photoreaction.” The example of such processes is the photoexci-
tation of dye molecules adsorbed on SC particles (Fig. 12.1(a)). In the second type, the
initial photoexcitation occurs in the catalyst – semiconductor particles. The photogener-
ated charge carriers flow to the SC particle | solution interface, interface and react with
acceptors and donors (Fig. 12.1(b)).
Such process is referred to as the “catalyzed photoreaction” and will be described
briefly below.

(a) (b)

D* h𝜈
e
h𝜈

CB A CB
D
A–

D+
VB D VB
+
D+

SC Dye SC

Fig. 12.1(a, b): Scheme illustrating photocatalytic processes: sensitized photoreactions (a) and catalyzed
photoreactions (b).

https://doi.org/10.1515/9783111160986-013
 12.2 Size effect and dimensionality 281

In the photocatalytic studies, the colloid semiconducting particles or particle suspen-

sions are often used. Their physicochemical properties depend on their nature and size.
Their charge and potential drop across the particle–solution interface is influenced by
the solution properties (ions, pH, kind of solvent). Much broader information about col-
loids, suspensions, micro-, and macroemulsions can be found in bibliography [1, 2].

12.2 Size effect and dimensionality

Since the properties of particles are size dependent [3], their behavior differs from
that of bulk conventional semiconductor electrodes. Some important features of the
reduction of particle size into nanometer scale can be described as follows:
1) Quantum size (QS) effect is observed when the size of particle is significantly re-
duced. It results in an increase in band gap energy. For instance, by changing the
size of CdS particles its band gap energy can be tuned between 2.4 and 4.5 eV. How-
ever, there are some semiconductors with large effective electron mass, in which
the QS effect occurs only at a very low value of particles radius. Such example is
TiO2. Only when particle size is < 3 nm the QS effect appears. The band widening
causes the shifting of potentials of electrons (bottom edge of CB) to more negative
and holes (upper edge of VB) to more positive values. In consequence, the reduction
and oxidation ability of charge carriers increases.
2) The size is also important for the light absorption. In a suspension of isolated, in-
dividual nanoparticles, light is absorbed by spherical particles and also scattered
on them. The relationship between optical penetration depth x = α1 , the width of
space charge layer Lsc (if exists), and the particle size should be appropriate. If
not, the recombination may dominate in the vicinity of particles.
3) The development of band bending and the formation of space charge layer de-
pend on the particle size. When the dimension of particle (d) is low d ≤ LD ,
ðLD ≥ Lsc , LD means Debye length, Sections 10.2 and 10.4), the band bending does
not exist. In this case the energy bands are flat. For d > LD the band bending is
developed.
4) The reduction of particle size leads to on an increase of surface area and surface
to volume ratio, causing the significant increase of the number of active catalytic
centers.

Let us compare the processes taking place during the illumination of semiconductor
particles and conventional (bulk) n-SC electrode (Fig. 12.2(a–c)) immersed in electro-
lyte solution containing redox couple. In dark, the state of bulk SC and particles de-
pends on the relation between EF and EF, Redox .
If EF > EF, Redox , the space charge layer is formed in the depletion state (bulk n-SC)
and proper band bending is developed (Fig. 12.2(a)). What is going on in the particles
depends on their size. If the particles are small, the energy bands are flat (see point 3
282 12 Semiconductor particles in photocatalysis

(a) (b)
E(eV)
e
Efb
e
CB Ox CB

Red h𝜈
Red
Ox + +
VB
VB
Flattening of band Red
energy
n-type

(c)
Flattening
of band
energy
Ox

h𝜈
Red

+ + +
h

Fig. 12.2(a–c): Illustration of differences in behavior under illumination between n-SC electrode (a) and SC
particles: small particle with dimension d < LD (b) and large particle with d > LD (c); CB, VB – conduction,
valence band, respectively; Ox, Red – oxidized, reduced form of a redox couple in a solution.

above; Fig. 12.2(b)). If the particles are large and n type, they lose their electrons to the
Ox form of a redox couple and exist in the depletion state (Fig. 12.2(c)). They can also
exist in accumulation state (EF, Redox > EF ), as a result of electron flow from Red form
of a redox system to the conduction bands of n-SC particles (such case is not consid-
ered here). The equilibrium in dark is characterized, for SC electrode and particles, by
equality of Fermi energy levels EF = EF, Redox and equality of anodic and cathodic cur-
rents. Note the difference between the two systems: the potential of SC electrode can
be changed using external potential source (potentiostat), while in the case of par-
ticles their potential is determined by redox systems in the solution.
In practice, the particles may be treated as microelectrodes at open circuit poten-
tial, with the reduction and oxidation reactions proceeding simultaneously at the
same surface.
If we illuminate the n-SC electrode that possesses the space charge layer in the
depletion state, with the photon energy hv ≥ Eg , the generated excitons (bound
electron–hole pairs) instantly dissociate in the electric field of space charge layer,
forming free electrons and holes. The minority carriers, holes, flow to the SC |
electrolyte interface reacting with Red form of a redox couple. The majority car-
riers flow to the back contact of SC electrode. It results in the unbending of energy
 12.2 Size effect and dimensionality 283

bands (Fig. 12.2(a)) and generation of photopotential and photocurrent. The behav-
ior of SC electrodes in dark and under illumination was described in details in
Chapter 10 (Sections 2–6). Now, if we illuminate the small particle, the excitons are
also generated; however, they cannot dissociate since the space charge layer does not
exist. They diffuse to the interface and dissociate in the electric field of the interface SC
Ps | electrolyte or recombine before that. Furthermore, electrons and holes take part in
the reactions with Ox and Red forms of a redox couple (Fig. 12.2(b)).
During the illumination of large n-SC particles (n-SC Ps), which exist in the deple-
tion state (Fig. 12.2(c)), the generated excitons dissociate in the electric field of space
charge layer. The holes and electron migrate to the interface taking part in the reac-
tions with Red and Ox forms of a redox couple. However, in the case of n-SC particles
in depletion state, the reaction of holes (minority carriers) with the Red form will be
more preferable. In consequence, the reaction of holes causes the negative charging
of particles and flattening of the energy band. The particles are negatively charged,
and the oxidation reactions take place under illumination, so such particles may be
treated as micro-photoanodes.
Let us summarize the main processes which occur in/at particles:

SC + hv ! eðCBÞ + h + ðVBÞ, photoexcitation, charge separation (1)

eðCBÞ + A ! A − , eðCBÞ + Ox ! Red (2)
+ +
h ðVBÞ + D ! D + , h ðVBÞ + Red ! Ox (3)

There are also other processes that can take place in the semiconductor particles after
photoexcitation, such as trapping of charge carriers, relaxation of trapped carriers,
and recombination:

eðCBÞ ! etrap ; hþ ðVBÞ ! hþ trap ; trapping

þ þ
etrap ! e; h trap !h ; relaxation
þ þ
eþh trap ; eþh ðheat; photonÞ; recombination

Such processes, as trapping and recombination, decrease the photocatalytic efficiency.

All the above reactions have different time scale. Their time scale depends on the na-
ture of particles, nature of trap centers, reagents, and so on. However, some generali-
zation may be made. The separation of carriers and their trapping are the quickest
processes; their time scale is femtoseconds. For relaxation, the time scale is in picosec-
onds. For the recombination, the time scale depends on its type. The radiative recom-
bination is quick (nanosecond or less). If the trapped carriers take part in the
recombination, the time scale can increase up to microseconds. The time scale for
charge transfer reactions changes from nanosecond to microsecond depending on the
type of reagents and charge carriers participating in the reaction, whether they are
free or trapped. In the last case reaction (2) and (3) can be written:
284 12 Semiconductor particles in photocatalysis

etrap + A ! A − , etrap + Ox ! Red

h + trap + D ! D + , h + trap + Red ! Ox

Much more about the electron transfer dynamics you will find in the review [4] and
references therein.
Up till now, we have considered the freestanding 0D (zero-dimensional) nanopar-
ticles; they are typically applied as photocatalysts. The SC nanoparticles assembly can
be used for the formation of photoelectrodes. This type of electrodes has a large specific
area for reactions, but unfortunately are also characterized by slow diffusion and high,
interparticles recombination of charge carriers. For the photoelectrode construction,
the so-called 1D nanomaterials, such as nanowires, nanorods, nanotubes, vertically
aligned on the transparent conducting substrates, as well as 2D thin films are more suit-
able. They are characterized by increased light path length, higher charge mobility, and
lower charge recombination. Nanophotocatalysts, as well as nanostructured photo-
electrodes, formed by an arrangement of 0D and 1D photocatalysts are studied inten-
sively. For more broad information about the fundamentals of photocatalysis, strategy
of development, and perspectives, their interested reader can find in the excellent
book [4].

12.3 Photocatalytic degradation of organic/inorganic pollutants

In the last decade, the heterogeneous photocatalysis has been proved to be efficient
method for environmental remediation. Among wide spectrum of semiconductors,
TiO2 (anatase and rutile) has attracted attention due to its excellent properties such as
the sufficient band gap energy, photostability, nontoxicity, and low cost. TiO2 was
used in PEC for water splitting (Chapter 11, section 11.3), where photogenerated holes
and electrons initiate the oxygen and hydrogen evolution. These charge carriers can
also be involved in degradation of various organic pollutants, creating very reactive
radicals according to the below reactions [5–8]. Light absorption and photogeneration
of holes and electrons initiate the chain reactions involving charge carriers, water
(moisture), and pollutants:

TiO2 + h v ! eðCBÞ + h + ðVBÞ

holes:

h + ðVBÞ + H2 Oads ! OHads + H + , radical may be also free in solution (1)

h + ðVBÞ + OH − ads ! OHads (2)

 12.3 Photocatalytic degradation of organic/inorganic pollutants 285

electrons:

eðCBÞ + O2 ! O2  − (3)

O2  − + H + ! HO2  (4)

2 HO2 ! H2 O2 + O2 (5)

2 O2  − + 2 H + ! H2 O2 + O2 (6)
H2 O2 + eðCBÞ! OH + OH − (7)
H2 O2 + O2  − ! OH + OH − + O2 (8)

organic pollutants + OH ! intermediate ! CO2 + H2 O (9)

Reactions (1) and (2) generate highly reactive hydroxyl radicals •OH, which are mainly
responsible for the degradation of organic pollutants (9). The reactions occur on the
particle surface, where radicals and pollutants may be adsorbed.
The superoxide radical O2•− formed in reaction (3) is not so highly reactive; how-
ever it may be involved in the formation of hydroperoxy radical HO2• or H2O2 (4, 6),
which can be further reduced to the adsorbed or free •OH radical. The oxidative and
reductive ability of radicals depend on their equilibrium redox potentials. Since H+ or
OH− ions are involved in the reactions, the potential is pH dependent and decrease (is
more negative) with an increase in pH. The same tendency is observed for SC metal
oxide. The position of the edges of the conduction and valence band of SC is shifted to
more negative potentials with an increase in pH. It results from acid/base equilibrium
for the metal oxide surface groups. Therefore, the Efb potential of semiconducting
metal oxides is pH dependent: Efb ðV Þ = Efb ðpH = 0Þ − 0.059pH
Figure 12.3 presents the correlation between positions of conduction and valence
bands of some SC oxides with the equilibrium redox potentials of important radicals
•
OH, O2•−. From the thermodynamic point of view, in order to initiate the pollutant
degradation, the bottom of CB must be located at more negative potentials than the
redox potential of O2|O2•− couple (−0.28 V vs NHE, pH 7). In addition, the top of VB
should be located at potential more positive than the redox potential of OH−| •OH cou-
ple (1.6 V vs NHE, pH 7). Looking at Fig. 12.3 we see instantly that oxides such as TiO2
and ZnO can be applied for pollutants degradation. Other oxides, such as SnO2, ZrO2,
Fe2O3, Nb2O5, and so on, have also been intensively studied.
In practice, to choose the best photocatalyst for given reaction, we have to deter-
mine and compare the activity of photocatalysts. Quantum yield ∅ is often used for
such comparison [9]. Quantum yield in the heterogeneous photocatalysis is defined as
the ratio of reaction rate and the absorption rate of photon flux:
v
∅= , Ia = I0 F, (12:1)
Ia
286 12 Semiconductor particles in photocatalysis

E(eV) pH 7

–3.0 –1.5 Cu2O

–3.5 –1.0 TiO2

–4.0 –0.5 ZnO

H+/H2
Fe2O3 O2/O•–
–4.5 0
O2/H2O2
–5.0 0.5
O2/H2O
–5.5 1.0

–6.0 1.5 OH–ads/•OHads

–6.5 2.0
H2O/•OH
–7.0 2.5

E(V), SHE

Fig. 12.3: Correlation between the conduction and valence bands of SC oxides (in the energy and
potential scales) with the equilibrium potentials of oxygen radicals.

ѵ is the photocatalytic reaction rate, Ia is absorbed photon flux, I0 is intensity of inci-

dent photon flux, and F is the integrated absorption fraction.
It is difficult to assess the absorption rate of photon flux experimentally. Serpone
et al. [10] had proposed a standardization protocol to evaluate relative photonic
efficiency ð=Þ, using phenol as a standard organic reagent and TiO2 (P-25 Degussa) as
a standard photocatalyst.

rate of organic compound degradation

== (12:2)
rate of phenol degradation

Both the initial rates in eq. (12.3) have to be determined under same experimental
conditions. Such procedure eliminates the influence of light intensity, reactor geome-
try, and TiO2 concentration on the photonic efficiency =.
The parameter that is used in chemistry for comparing the reaction rates is the rate
constant. In the heterogeneous catalysis of organic compounds, the Langmuir–Hinshel-
wood model is used for kinetic description. This model is based on the assumption that
only one reagent (in our case – the pollutant molecule) is adsorbed on the catalyst sur-
face. The Langmuir isotherm is applied for description of surface concentration of re-
agent (this isotherm is described in Chapter 4). Rate of the first order reaction, νr , is
expressed by the equation:

dc
νr = − = kΘ (12:3a)
dt
 12.3 Photocatalytic degradation of organic/inorganic pollutants 287

dc Kads c
νr = − =k (12:3b)
dt 1 + Kads c

k is the rate constant, Θ is degree of surface coverage, Kads is equilibrium adsorption

constant, and c is reagent concentration.
If the initial concentration is low, Kads c  1, the equation can be reduced to

dc
νr = − = kKads c = kapp c (12:4)
dt

kapp is apparent rate constant and may be easily evaluated from the plot − ln cc ver-
0
sus t. To compare the photoactivity of different catalysts, the apparent rate constant
should be determined for a model reaction. Decomposition of phenols or dyes, such as
methyl orange, rhodamine B, and methylene blue, belong to these reactions.
There are plenty of inorganic pollutants in air such as CO2, CO, NO2, SO2, and so
on. They are produced in the road traffic, in the domestic heating, or by industry. Im-
portant pollutant is carbon dioxide, whose concentration increases drastically, result-
ing from application of fossil fuel in the production of energy. There are few methods
of reduction of CO2 contaminations [11]. The most promising seems to be the photoca-
talytic reduction, in which solar energy is applied. The photoreduction of CO2 in PE
cell was described in Section 11.3, where the thermodynamic data were also presented
for the formation of different CO2 reduction products. In PE cell, the p-type of semi-
conductor has to be used as photocathode. There are at least two advantages in appli-
cation of SC particles in CO2 reduction. One is the increase of reaction efficiency, as
the surface area of particles is higher than the surface of conventional SC electrodes.
The second, very important, is the possibility of application of any kind (n or p type)
of semiconductor particles. It results in the widening of the spectrum of semiconduc-
tor materials. It also enables better correlation between CB and VB edges of semicon-
ductor and redox potentials of reaction of CO2 reduction. Such scheme for few SC is
shown in Fig. 12.4.
TiO2, ZnO, CdS, GaP, and other SC particles were tested as photocatalysts for CO2
reduction. More broad information on this topic can be found in review and referen-
ces therein [11]. Up till now, the best photocatalyst for degradation of organic and in-
organic pollutants is TiO2. Its photocatalytic activity depends on physicochemical
properties, including the primary particle size, degree of aggregation, crystalline
structure, morphology, surface area, and on the external conditions, such as pH, kind
of acceptors/donors in solution, and irradiation intensity. Commercially available TiO2
Degussa P25 is regarded as model catalyst, widely used as a reference to other photo-
catalysts. The main disadvantage of TiO2 and other SC oxides is their large Eg resulting
in the absorption of light energy in UV range. This problem was discussed briefly in
Section 11.3.
288 12 Semiconductor particles in photocatalysis

E(eV)

–3.5 NiO Ce2O3 –1.0

TiO2 3.5 2.4

ZnO eV eV
3.2 3.2 CO2/HCOOH –0.6
–4.0 eV eV CO2/CO –0.5
H+/H2 –0.4
CO2/CH3OH –0.3
Bi2O3 –0.2
–0.1
–4.5 2.8 0
eV

E(V), SHE

Fig. 12.4: Correlation between the conduction bands of some of SC oxides (in the energy and potential
scales) with the redox potentials of CO2 reduction reactions, pH-7. Arrows point to the position of valence
bands, and the bandgap energy, Eg , is marked at arrows.

Bibliography

[1] Butt H-J, Graf K, Kappl M. Physics and chemistry of interfaces. Weinheim, FRG, Willey- VCH
GmbH, 2003.
[2] Lyklema J. Fundamentals of colloids and surface science: Solid liquid interface. London, United
Kingdom, Academic Press, 1995.
[3] Cao G. Nanostructures and nanomaterials. Synthesis, properties and applications. London, United
Kingdom, Imperial College Press, 2006.
[4] Schneider J, Bahnemann D, Ye J, Li Puma G, Dionisiou (eds). Photocatalysis. Fundamentals and
perspectives. Cambridge, United Kingdom, RSC Publishing, 2016.
[5] Li J, Wu N. Semiconductor-based photocatalyst and photoelectrochemical cells for solar fuel
generation: A review. Catal. Sci. Technol. 2015, 5, 1360–1384.
[6] Schneider J, Matsuoka M, Takeuchi M, Zhang J, Horiuchi Y, Anpo M, Bahnemann DW. Understanding
TiO2 photocatalysis: Mechanisms and materials. Chem. Rev. 2014, 114, 9919–9986.
[7] Linsebigier AL, Lu G, Yates JT Jr.. Photocatalysis on TiO2 surfaces: Principles, mechanisms and
selected results. Chem. Rev. 1995, 95, 735–758.
[8] Fujishima A, Zhang X, Tryk DA. TiO2 photocatalysis and related surface phenomena. Surf. Sci. Rep.
2008, 63, 515–582.
[9] Zhang J, Chen G, Bahnemann DW. Photoelectrocatalytic materials for environmental application.
J. Mater. Chem. 2009, 19, 5089–5121.
[10] Serpone N, Sauve G, Koch R, Tahiri H, Pichat P, Piccinini P, Pelizzeti E, Hidaka H. Standardization
protocol of process efficiencies and activation parameters in heterogeneous photocatalysis: Relative
photonic efficiencies ζr. J. Photochem. Photobiol. A: Chem. 1996, 94, 191–203.
[11] Wang W-N, Soulis J, Yang YJ, Biswas P. Comparison of CO2 photoreduction system: A review. Aerosol
Air Qual. Res. 2014, 14, 533–549.
Part IV: Electrochemistry in environmental
science – selected topics
13 Sensing the environment

13.1 General remarks

There exist major concerns for the safety of life from a number of harmful pollutants
(anthropogenic or not) in the environment: atmosphere, ground (soil), and water.
Moreover, on top of the chemical pollution, biological micropollutants (bacteria, vi-
ruses, and others), coming from sources such as many human and livestock wastes,
enter the environment, groundwater in particular, causing many diseases. Some of
such microorganisms acquired multi-drug resistance as the undesired effect of phar-
maceutics abuse by humans. The large spectrum of pollutants present in the environ-
ment is usually categorized based on their mode of deteriorating effect, chemical
structure, or both.
Apart from different initiatives, social concerns, and legislative actions, means of
controlling and reporting such threats necessitate the development of sensitive, selec-
tive, and rapid detection techniques, screening the environment, and triggering the re-
mediation processes [c.f. 1]. For these purposes, two different approaches are usually
considered. The first one, more traditional, typically involves a multi-step approach,
such as sampling, handling of a sample, collection of samples, and transportation to
qualified laboratories. This time-consuming and costly procedure is nowadays being
challenged with in-situ monitoring of environmental threats with the help of electro-
chemical sensors. Thus, electrochemistry becomes a sustainable way to monitor and re-
solve the analytical complexity of environmental pollution in a mild, green, and real-
time way. Moreover, it can be used for the cost-effective prevention of the formation of
toxic materials during industrial processes and their leakage into the environment. A
very important from the point of view of “green” electrochemistry for improving the
quality and safety of life, is the area of the food industry, where electrochemical sensors
attract more and more interest to the industrial community. This is because the issue of
food safety has become a significant public concern and awareness worldwide due to
its relation to increased human diseases and economic burden. Therefore, there is a
need for real-time monitoring of food safety and quality during its production, packag-
ing, and storage, to screen and control at each manufacturing step the presence of con-
taminant residues, such as antibiotics, allergens, pesticides, toxins, pathogenic bacteria
or the microbial breakdown of the food products. In all these areas the advantages of
electrochemical sensors such as high selectivity and sensitivity, and minimal power re-
quirements for the related instrumentation and equipment are obvious. Moreover, the
use of electrons in electrochemical sensors for signal acquisition and transduction is
considered a “green” model of detection and analysis of environmental pollution with
no waste generation, low costs, fast analysis, and possibilities for miniaturization.
Therefore, we will discuss some basic information about electrochemical sensors, their
types, and applications for use in the detection of various environmental pollutants af-

https://doi.org/10.1515/9783111160986-014
292 13 Sensing the environment

fecting the everyday quality of life. It is worth keeping in mind, that the rapidly devel-
oping technologies and advances in microelectronics and microengineering coupled
with the new concepts in nanoscience, allow for the design of miniaturized sensors
with enhanced efficiency, sensitivity, selectivity, and durability for the real-time control
of the environment. For in-depth information, the interested reader is directed to sev-
eral publications [2–4]. All the needs mentioned above in environmental monitoring,
clinical and food diagnostics, paralleled with rapid development in biotechnologies,
also paved the way for the design of biosensors with biological sensing elements, such
as enzyme, DNA, aptamer and protein sensors, as well as immunosensors. The use of
biological molecules as sensing elements opened up a competitive area of research
aimed at selecting the proper element, defining its sensing behavior in the composite
matrices of the controlled environment (wastewater, soil, air), and its integration with
signal transducer and analyzing setup. For the sensor activities, numerous enzymes
from the group of oxidases are utilized, among them urease for metals, laccases, and
tyrosinases for phenolic compounds, acetylcholine esterase or organophosphorus hy-
drolase (anodic oxidation of p-nitrophenol as a product of the enzymatic reaction of the
latter enzyme). Some of the above-mentioned enzymatic biosensors for their proper
performance are coupled to the competitive immune reaction (Chlorsulfuron, Sima-
zine), thus forming the immunoassay-based biosensors. Formaldehyde and sulfur
dioxide biosensors rely on formaldehyde dehydrogenase or sulfite oxidase/cyto-
chrome c, respectively, for direct redox reaction generating the amperometric signal
proportional to the pollutant concentration in the air at equilibrium with aqueous
supporting electrolyte. However, as we discussed earlier in this book (Section 9.3),
most biosensors are based on reactions catalyzed by macromolecules (e.g., enzymes)
that are isolated from their original biological environment, the lifespan of these bi-
osensors is limited, creating a major drawback and motivating permanent efforts to
counteract such an issue [5].

13.2 Electrochemical sensors

Electrochemical sensors are by far most frequently employed for the detection of var-
ious pollutants, such as heavy metals, herbicides, pesticides, pharmaceutics, and dyes
in complex environmental matrices such as air, water, and soil. Taking the definition
by the IUPAC of a chemical sensor, the electrochemical sensors in the environment
can be defined as devices that produce an electrical, analytically usable signal propor-
tional to the concentration of pollutant as a result of pollutant(s) reaction with sensing
electrode, transducing/converting such reaction into an electrical signal. If an integral
part of the sensing electrode is of biological origin, then the sensor is called the bio-
sensor and was discussed earlier in this book (Part II, Chapter 9 (Interfacing applied
electrochemistry and biology)). Nevertheless, all of the electrochemical sensors can be
divided into several groups, depending on the type of transduced signal, such as am-
 13.2 Electrochemical sensors 293

perometric, potentiometric, conductometric (impedimetric), photoelectrochemical,

and electrogenerated chemiluminescence. The most reliable from the point of view of
environmental monitoring are the first three types of sensors, because of the ease of
relaying the signal (changes of current, potential, or conductivity) into the control
unit, sometimes far away from the sensor location. Therefore, below we will describe
the behavior of these three types of sensing electrodes (electrochemical sensors). Of
course, the design of those sensors varies, depending on the target pollutant and the
environment that is to be monitored. With the rapid technological progress, the sens-
ing processes – the “heart” of such sensors, were transferred from typical, aqueous
solutions to less volatile solvents, such as ionic liquids, described in the previous chap-
ter (Section 1.3.1) and to the all-solid-state sensors, utilizing various types of solid-state
ion conducting, ion selective or semiconducting membranes capable of monitoring
pollutants at high temperatures (e.g., the SOx, NOx, and COx monitoring sensors in the
car exhaust systems). Since this is beyond the scope of this book to cover all existing
solutions, we will focus mainly on the electrochemical reactions responsible for the
sensing performance in aqueous phase-containing systems. However, there are sev-
eral key parameters, describing the sensors’ performance, regardless of their design.
These are the sensor’s linear range, its sensitivity, and selectivity. The linear range
means the range of pollutant concentration for which the sensor’s output/response
scales linearly. The sensitivity is defined by the ability to detect minute changes in the
sensor environment, whereas selectivity is the sensor’s capacity to differentiate be-
tween various pollutants. With the sensor’s sensitivity also another parameter is re-
lated, namely the limit of detection, or LOD (LoD). According to IUPAC, the limit of
detection is the lowest sensor response (signal) that can be monitored with sufficient
statistical significance. However, the exact threshold (level of decision) used to decide
when a signal is significant above the background noise remains somehow arbitrary.
The third important parameter defining the sensor’s performance is its selectivity. It
is quantified by the selectivity coefficient as:


SInt
logkA,Int = log (13:1)
SA

where SA is the selectivity against a pollutant A, and SInt is the selectivity against a
particular interferent. The more negative values of this coefficient, the more selective
is the sensor.

13.2.1 Amperometric sensors

Amperometric sensors utilize controlled voltage applied between the sensing and ref-
erence electrode, measuring the resultant oxidation (or reduction) current, being in
direct quantitative relation with the pollutant concentration, given by the Cottrell
equation (see Chapter 2):
294 13 Sensing the environment

1
iF ðtÞ = nFADc0 (13:2)
ðπDtÞ1=2

Under semi-infinite linear diffusion control, if the applied potential is sufficiently nega-
tive (for reduction) or positive (oxidation), all pollutant species diffusing to the surface of
a sensing electrode are immediately and totally reduced or oxidized (depending on the
type of sensing reaction), this current is called the limiting current, iL , scaling linearly
with the pollutant concentration. This type of sensor is widely utilized to control/meas-
urements of concentrations of different pollutants, such as gaseous pollutants (in the gas
phase) or pollutants dissolved in a liquid (sewage, wastewater, groundwater, etc.). They
provide good sensitivity and a relatively large linear range. They are simple to use and
easy to manufacture by, for example, microfabrication technology. Apart from environ-
mental monitoring, these sensors found their use in industrial safety and medical appli-
cations. Such sensors can be used to control a wide range of substances in different
phases – gaseous and liquid). They are most frequently used to monitor the concentra-
tion levels of oxygen (O2, both in gaseous and liquid phase), nitric oxide (NO, gas phase),
sulfur dioxide (SO2 both liquid and gas phase), ammonia (NH3, gas phase), carbon mon-
oxide (CO, gas phase), hydrogen sulfide (H2S, gas) to name a few. The sensitivity range
for amperometric sensors varies from part per million (ppm) level to micromolar con-
centration, depending on the pollutant. Since gas detection is essential to everyday life;
the use of sensors for toxic and hazardous gases can set the difference between life and
death. Therefore, below we will now consider some examples of amperometric sensors
for some most important gas pollutants.

Nitric oxide (NO)

Nitric oxide together with nitrogen dioxide, among other toxic gases, are emitted to the
environment in combustion processes that occur, for example, in fossil fuel electric
power plants and transportation (diesel, gasoline) engines, affecting our health and the
health of animals and greeneries on earth as well as in waters. Nitric oxide is a colorless
֗
gas, a radical (N=O), and should not be confused with brownish gas – nitrogen oxide,
and a major pollutant. Nitric oxide in the atmosphere can react with ozone, forming ni-
trogen dioxide and oxygen, thus decreasing the level of ozone in the atmosphere. When
exposed to oxygen, nitric oxide also generates nitrogen oxide, according to the reactions:

3NO + O3 ! 3NO2 or NO + O3 ! NO2 + O2

2NO + O2 ! 2NO2 ,

However, in an aqueous or humid environment, these reactions proceed to form ni-

trous and nitric acids:
 13.2 Electrochemical sensors 295

4 NO + O2 + 2 H2 O ! 4 HNO2
4 NO2 + O2 + 2 H2 O ! 4 HNO3

Nitrous acid, nitric acid, and sulfuric acid contribute to acid rain precipitation, result-
ing in corrosive and otherwise deteriorating effects on the environment. However,
one can acknowledge also that nitric acid in rainwater is an important source of nitro-
gen for plants.
Hence, the development of a NO gas sensor for measuring NO levels is very im-
portant to monitor and control the evolution of NO gas. Electrochemical detection of
NO is typically based on its diffusion-controlled electrooxidation because it results in
higher sensitivity and better performance in the presence of interfering dissolved ox-
ygen. The electrooxidation proceeds according to the following reactions:

Sensing electrode (WE): NO ! NO+ + e

NO+ + OH− ! HNO2 (fast, highly irreversible chemical step)
HNO2 + H2 O ! NO3 − + 3H+ + 2e

Under the neutral or physiological pH, the reaction proceeding at the counter elec-
trode (the cathode) is:

O2 + 4H+ + 4e ! 2H2 O

The overall redox reaction will be:

4NO + 2H2 O + 3O2 ! 4 NO3 − + 4H+

Depending on the electrode material, nitric oxide oxidation must be performed at suf-
ficiently positive potentials, ranging from +0.7 to 1.0 V vs. Ag|AgCl electrode. Different
types of electrode materials have been reported over time, including platinum, gold,
glassy carbon, graphite, and other electrodes [6]. Modification of the sensor with gas-
permeable membranes significantly improved the sensor’s selectivity and overall per-
formance. These membranes, such as Nafion, PTFE, or cellulose acetate, not only facil-
itated the NO diffusion to the internal electrolyte but also prevented its leakage and
evaporation from the sensor. They were also designed to prevent interference of car-
bon monoxide, CO, but in the case of sensors operating under “wet” conditions, such
as wastewater or biological fluids, the interference of ascorbic acid, uric acid, and
acetaminophen can be diminished, too. The amperometric sensors can monitor very
low redox current (e.g., down to pA) produced by the oxidation of NO over time at a
fixed (poise) voltage potential. The response time of this amperometric sensor is only
a few seconds, and coupled with its high sensitivity, it provides a fast, quantitative
measurement of very small changes in NO concentration.
For a critical review of recent advances of NO sensors, particularly those operat-
ing in physiological fluids, the interested reader is referred to [7].
296 13 Sensing the environment

Sulfur dioxide, SO2

Monitoring and control of sulfur dioxide gas emission to the environment is a serious
global problem in relation to the atmospheric pollution and poses a serious threat to
population health. Its presence in the atmosphere and subsequently in groundwater
mainly comes from industrial petrochemical fuel combustion and coal burning, refin-
eries, and other chemical plants. In contact with humidity and rainwater, it produces
acidity in rainwater (acid rains, acid fogs), being a major source of corrosion of build-
ings and other constructions. The major health concerns associated with exposure to
high concentrations of SO2 are respiratory illness, alterations in the lungs’ defenses,
and intensification of existing cardiovascular disease.
The reaction that occurred on the sensor’s working electrode in the presence of
sulfur dioxide in the environment, and in the case of a strong acid as an internal elec-
trolyte, is the oxidation reaction:

Sensing electrode (WE): SO2 + 2H2 O ! SO42− + 4H+ + 2e

At the same time, at the counter electrode, the cathodic reaction proceeds:

CE: 2H+ + 2e ! H2

Thus, the overall reaction in the sensor is:

SO2 + 2H2 O ! SO42− + 2H+ + H2

Ammonia, NH3
The most simple ammonia sensor relies on a straightforward oxidation reaction of NH3
to nitrogen and hydrogen at the sensing electrode. As in the case of other amperometric
sensors, the redox current output is used to determine ammonia concentration.

Thus, at the sensing electrode:

ðWEÞ: 2NH3 ! N2 + 6 H+ + 6 e

At the counter electrode:

ðCEÞ: O2 + 4H+ + 4 e ! 2H2 O

And the overall reaction:

4NH3 + 3O2 ! 2N2 + 6H2 O

This sensor, even though simple, depends on the availability of oxygen in its electro-
lyte solution, as well as other electrolyte components. Once the diffusion of oxygen
becomes the limiting rate of the reaction, the sensor is no longer capable of detecting
ammonia. Therefore, electrochemical ammonia sensors should be used only when the
normal ambient background concentration of ammonia is sufficiently low to allow a
 13.2 Electrochemical sensors 297

reasonable operational life. Similar limitations for amperometric sensors are imposed
when ambient oxygen reaction proceeds on cathode.

Carbon monoxide, CO
Carbon monoxide is a product of an incomplete combustion of carbon fuel. It is a col-
orless, odorless, tasteless, and poisonous gas that even in small doses claims perma-
nent health damage or death. In households, ca. 80% of CO-lethal poisoning comes
from heating systems and engine-driven tools (including cars). Poisoning by carbon
monoxide results from its binding to red blood cells (hemoglobin), blocking their ca-
pabilities of transporting oxygen to our body.
In industry, carbon monoxide is important in the large-scale production of nu-
merous specialty compounds, including surfactants, drugs, and fragrances in the
processes called hydroformylation. In the Fischer‒Tropsch catalytic process, it is
used to produce liquid fuels in a series of chemical reactions of the overall scheme:
(2n + 1)H2+nCO → CnH2n+2+nH2O, with n typically above 10. However, upon emission
to the atmosphere, carbon monoxide can also contribute to climate change. It is also in-
teresting to note, that CO at very low concentrations can serve as an endogenous neuro-
transmitter, being highly toxic at higher concentrations, as pointed out above. Because of
its high toxicity, carbon monoxide sensors are widely used for process control and safety
applications. CO sensing is based on the oxidation reaction on the working electrode,
(WE), in the presence of water molecules (either from the supporting electrolyte or from
the surrounding air, depending on the sensor’s design), according to the scheme:

ðWEÞ: CO + H2 O ! CO2 + 2H+ + 2e

And, at the counter electrode, similar to the NO and NH3 sensors:

ðCEÞ: O2 + 4H + + 4 e ! 2H2 O

Thus, the overall reaction is:

2CO + O2 ! 2CO2

Oxygen, O2
The necessity of monitoring industrial oxygen related to the safety and health of
workers prompted the research on amperometric oxygen sensors. Perhaps the best-
known pioneering work in this area was the invention of the Clark electrode used for
monitoring oxygen levels in liquids, water, and blood in particular. Briefly, Leland
C. Clark’s oxygen-sensing concept relies on the use of a two-electrode cell with an oxy-
gen-permeable (breathable) membrane separating the sensor electrode (cathode)
from the electrolyte in the cell. Oxygen can diffuse through the membrane being sub-
sequently reduced on the cathode, typically bare platinum, generating current scaling
with the concentration of oxygen in the sample, according to the reaction similar to
298 13 Sensing the environment

those at counter electrodes for NO, NH3, and CO sensors: (please note the reaction for
DMFC, Chapter 8):

ðWEÞ: O2 + 4H+ + 4e ! 2H2 O

ðCEÞ: 4Ag + 4Cl− ! 4AgCl + 4e

As one can see, the counter electrode for this sensor can be made from silver, and
both electrodes: Pt and Ag are immersed in aqueous saturated KCl solution. Thus, the
overall reaction in this sensor is:

4Ag + O2 + 4H+ + 4Cl− ! 4AgCl + 2H2 O

It is important to note, that the so-called oxygen concentration in the sample is in fact
the partial pressure of oxygen (pO2) dissolved in the sample at equilibrium, and under
a given pH, as described in Section 3.3, eq. (3.10). If one uses these two electrodes only,
the precise quantification of oxygen level is frequently inaccurate, because the electro-
chemical sensing reaction consumes high oxygen due to the rapid oxygen diffusion.
Moreover, unintended reactions can interfere with the sensing reaction, altering the
readouts of the sensor. Therefore, the introduction of oxygen-breathable, nonconduc-
tive membrane mentioned above, was a major development for this sensor. Such mem-
branes can be structured to allow only oxygen to slowly diffuse through them, barring
nontarget reactions to occur on the sensing electrode. Yet another problem can affect
the readout of Clark’s electrode, namely that in an acidic pH the sensing reaction can
stop to yield the hydrogen peroxide:

O2 + 2H+ + 2e ! H2 O2

Typically, this problem is solved by using a buffer at the Pt electrode that stabilizes
the pH of the solution around the platinum electrode.
Apart from serving as a medical device, adaptations of Clark’s electrode are used for
measurements of the water quality in the oceans, seas, freshwater, underground, etc.,
with oxygen level as one of the indicators for the, for example, water treatment plants.
Careful readers will certainly recognize that many of the sensing reactions are
similar to those discussed in the case of fuel and biofuel cells. However, in those
cases, the reactions were carried out in reversed directions.

Some technical notes

A careful reader will immediately find that the electrochemical amperometric sensors
are based on the same theory, but different redox reactions, and therefore use differ-
ent materials for their construction and design, heavily depending also on the envi-
ronment of their work. They have to perform their sensing tasks, frequently in harsh
atmospheres, elevated temperatures, and pressure. It is impossible to cover all cur-
rently used solutions, nevertheless let us outline, even briefly, some key factors in the
construction of the amperometric electrochemical sensors.
 13.2 Electrochemical sensors 299

– Electrolyte. It has to be selected as a supporting electrolyte (both solvent and dissociated

solutes) to carry out the redox sensing reaction (electrolysis), rapidly and effectively at-
taining the limiting currents (diffusion-limited, as described in Section 2.1.2) at the work-
ing electrode. Obviously, then, the supporting electrolyte should not undergo the redox
processes within the potential window of sensing reactions. Moreover, it should not be
too volatile to avoid signal interference due to solvent evaporation. This brings forth the
issue of the sensor working at elevated temperatures when water as solvent will be un-
suitable because of its relatively low boiling point. Therefore, the current focus is on the
use of nonvolatile ionic liquids (ILs) as electrolytes. They are advantageous over aqueous
electrolytes because they have wide electrochemical windows, high chemical and ther-
mal stability, and intrinsic conductivity. ILs have also very good solvating properties nec-
essary for the pollutant solution, including gases, as well as wetting the electrodes. They
are also less leaky and corrosive, improving the lifespan of the sensor in dry or humid,
and hot environments. Even more robust and flexible sensors are made with all-solid-
state electrolytes (SSEs) placed between the sensing and counter electrodes. Even though
the sensing redox reaction is the same as in the case of liquid electrolytes, the volumetric
charge transfer mechanism and the mobile species transferring this charge within the
solid-state electrolytes are different. The mobile charges (ions) movement occurs in the
solid state. As already stated, the main advantage of such electrolytes is the complete re-
moval of liquid components, greatly enhancing the robustness, temperature of the sur-
rounding environment, and safety of the sensor device. However, there is some price
that has to be paid, namely, the conductivity of SSEs is much lower compared to the liq-
uid electrolytes. Depending on their chemical nature, all-solid-state electrolytes can be di-
vided into inorganic solid electrolyte (ISE), solid polymer electrolyte (SPE), and composite
polymer electrolyte (CPE). This is the most general classification and the interested reader
can find more detailed descriptions in the references at the end of this section.

Inorganic solid electrolyte (ISE). An inorganic solid electrolyte (ISE) is constituted by

an inorganic material in the crystalline or glassy state, and its ionic conductivity pro-
ceeds by diffusion through the crystalline or glassy lattice. The main advantages of
this class of solid-state electrolytes are high ionic conductivity and thermal, and chem-
ical stability. They are generally brittle and have relatively low compatibility with the
electrode materials, leading to a rapidly increasing interfacial resistance in time, thus
limiting their lifespan. The ISE materials can be based on oxides, sulfides, or phos-
phates, including their mixed crystalline structures like perovskites, garnets, and tita-
nates, to name a few. Frequently applied examples of ISE include LISICON (Li14Zn
(GeO4)4, lithium superionic conductor, NASICON (Na3Zr2Si2PO4, sodium superionic
conductor), LIPON (lithium phosphorous oxynitride, a family of materials with the
general formula of LixPO3Nz, etc.

Solid polymer electrolyte (SPE). Solid polymer electrolyte (SPE) contains salt solution
in a polymer host network. It can conduct ions through the polymer chains. Compared
to ISEs, SPEs are much easier to engineer by direct solution casting on the electrode
300 13 Sensing the environment

material, therefore they can be preferable for manufacturing processes on a large

scale. They seem to be more compatible with common electrode material, showing
also high plasticity at the electrode surface. Salts that are dissolved in the polymer
matrix are typically Li salts due to their high concentration being achieved in the
polymer. However, SPEs suffer from generally lower ionic conductivity compared to
the ISEs. Some polymeric matrices used for SPE are poly(ethylene oxide) – based
(PEO) and its polycarbonates, polyesters, polyamines, polysiloxanes, as well as some
eco-friendly biopolymers like chitosan, cellulose or lignin or their blends.

Composite polymer electrolyte (CPE). When SPE is modified with nanoparticles inert to ion
conduction, such as alumina, titania, or silica nanoparticles (extrinsic passive fillers),
mixed nanoparticulate oxides of lithium or other metals used in, for example, supercapa-
citors or fuel cells (extrinsic active fillers), then it is called a composite polymer electrolyte.
CPE can also be obtained by copolymerization, blending, and crosslinking with other poly-
mers and all these procedures aim at tailoring the properties of host SPE for better disso-
lution of ionic salts, its ion-conducting properties (ionic mobilities) or only increasing its
amorphous state by reducing crystallinity and compatibility with the electrode surface.
It is hard for one material to fulfill the criteria for a good all-solid-state electrolyte
to be used in solid-state gas sensors. Below, Tab. 13.1 exemplifies some materials ap-
plied for detecting three already discussed pollutant gases.

Tab. 13.1: Examples of solid electrolytes used for sensors of selected pollutant gases.

Pollutant gas Examples of solid electrolytes

NO Ba(NO), Na-β/β″-alumina/NaNO (or Ba(NO), NASICON/NaNO + LiCO

SO KSO, lithium aluminum titanium phosphate (LATP, Li+xAlxTi−x(PO)) modified with
LiSO

CO NASICON/LiCO-BaCO

CO, CO Yttria-stabilized zirconia (ZrO/YO)

– Electrode materials, catalysts. The electrodes used for gas sensors (as well as for any
other type of sensors) should fulfill the following conditions:
– They should be chemically and mechanically stable in the solution or on the
substrates
– They must provide excellent contact with the solid electrolyte
– They must be compatible with the breathable membrane/filter
– They must have a geometry that is suitable for sensor construction
– Ideally, they should possess catalytic properties toward the sensing reaction.

Therefore, several metals, such as Ti, Pd, Pt, and Au, carbon-based materials, such as
GCE, carbon nanotubes, and conducting polymers, for example, polyaniline, polyace-
 13.2 Electrochemical sensors 301

tylene, polypyrrole, etc. are used for the construction of sensors. To improve the sen-
sor performance, catalytic metallic nanostructures are also utilized. Table 13.1 exem-
plifies some metal nanostructured catalysts used for pollutant detection.

Tab. 13.2: Examples of metal catalysts used for

selected pollutant gases.

Pollutant gas Electrocatalyst

NO, NOx Au

SO Au, Au/C

CO Pt

O Au, Ag, Pt

– Breathable membrane. Typically it separates/covers the sensing electrode but allows

the target gas to diffuse to the electrode surface (see Fig. 13.1A below). It should also
filter out interfering species and unwanted particles (such as dust). In the case of liq-
uid electrolytes, it should also protect them from drying or leaking. Since condensed
water present in harsh environments or high-humidity human breath samples can
result in false detection, the breathable membrane should also be hydrophobic. To
achieve such properties, hydrophobic polyethylene (PE), polyurethane (PU), and poly-
tetrafluoroethylene (PTFE)-based membranes, among other polymeric structures, are
widely used. In particular, PTFE is of high usefulness because it is a highly hydropho-
bic, thermoplastic with excellent thermal stability and chemical resistance properties.
Moreover, it can operate within a wide range of temperatures, from −150 °C to 250 °C.
PTFE film is also an ideal host material for further modifications tailored for the spe-
cific sensor design improving also selectivity characteristics of gas sensors.
The technology described in this chapter can be modified according to the target
gas, working environment, pressure, and temperature, as well as other factors, such
as geometry. Apart from applying different materials for the sensors’ design, their
specificity and selectivity can be further improved by the application of bias poten-
tials specific to the redox reaction of pollutants of interest, and chemically selective
filters to block potentially interfering species.
Fig. 13 presents some possible configurations of amperometric gas sensors with
liquid or SSE, exemplified for CO and oxygen gas sensors.
302 13 Sensing the environment

A vent
case filter

breathable
working membrane
electrode
electrolyte
counter
electrode reference
electrode

pin

CO presence
B clean air (no CO) C CO2 CO H2O
e
working electrode

H+
no current flow

ion conductor (e.g. H+) A

A
solid/liquid
i

counter electrode
H2O e

Fig. 13.1: General scheme of a 3-electrode amperometric gas sensor with liquid electrolyte (A); scheme of
sensing mechanism for the amperometric gas sensor in sensing of CO (B), and oxygen (C).

13.2.2 Potentiometric sensors

A potentiometric sensor operates under conditions of near-zero current flow and meas-
ures the difference in potential between the working electrode (sensing) and a reference
electrode. Essentially, when these electrodes are immersed in the solution containing
pollutant molecules showing redox properties, even this simple configuration can pro-
vide potential signal scaling linearly with the natural logarithm of the concentration of
this pollutant according to the Nernst equation (compare Chapter 1, eq. (1.53)):
RT cp,ox
EN = E f + ln (13:3)
nF cp,red

Here, Ef is the formal potential, and cp,ox , cp,red are the pollutant’s concentrations in its
oxidized and reduced forms, respectively. Please note that to apply this equation, one
needs a redox-active pollutant and its liquid solution. Moreover, such a design, as in the
case of amperometric sensors using liquid or aqueous solution, lacks the necessary ro-
bustness and compatibility with in-situ or in-field monitoring. Therefore, the solid-state
potentiometric sensors were designed, in which liquid solutions are replaced by ion ex-
change membranes working as transducers. Ion-selective electrodes (ISEs), of which the
 13.2 Electrochemical sensors 303

pH electrode (Section 1.4) is the best-known example, are the most important of this
class of transducers, particularly in control of hazardous ions and compounds in
aquatic systems, such as groundwaters, seawaters, sewage, agricultural, industrial and
pharmaceutic wastes. Upon immersion in the pollutant solution, across such ion-
selective membrane a certain value of potential difference can develop, called the open
circuit potential (OCP), or if the quasi-equilibrium conditions are reached, the potential
at the sensing electrode can attain the Nernstian slope, according to the equation:

2.303RT
E = const. + logcMen+ (13:4)
nF

Referring to the scheme shown in Section 1.3, the origin of this equation can be ex-
plained by the following Scheme 13.1:

Reference electrode I Reference (contact) electrode II

pollutant solution (I) ISE internal solution (II)

E (I) ΔφM E(II)

Scheme 13.1: Simplified scheme of potentiometric sensor arrangement.

Here, E(I) and E(II) are the half-cell potentials of reference electrode I and reference
electrode II in their solutions, respectively, while ΔφM is the membrane potential.
Thus, the following potential difference will develop in the sensor at equilibrium:

E = EðI Þ − EðII Þ + ΔφM = const.′ + ΔφM (13:5)

Taking advantage of eq. (1.51), and keeping E(I) and E(II) half-cell potentials and concen-
tration of the internal solution (II) constant, we can easily derive that the membrane
potential in such a case can be described as dependent only on the concentration of
pollutant ion Men+ species in the solution (I).
Similar considerations based upon the equilibrium chemical potentials and par-
tial pressures of pollutant gas can lead to the analogous equation for gas sensors:
2.303RT
E = const. + logppol (13:6)
nF

where ppol is the partial pressure of pollutant gas to be measured, however not all
gaseous pollutants can be addressed this way, mainly due to the difficulties in defin-
ing the processes at the reference side of a membrane.
Specificity is imposed by the structure and properties of selective membranes,
which may be formed from metal salts, oxides, mixed compounds, or polymer mem-
branes containing ion exchangers or neutral carriers. Depending on its physical state
and chemistry, generally, there are three ISE membrane types and according to this
criterion the sensing electrodes are classified:
304 13 Sensing the environment

– Solid state/crystalline and mixed chemistry membrane electrodes. pH electrodes

can belong to this class. Other membranes are typically composed of an insoluble in-
organic salt, doped with ions to be detected. For instance, Ag2S/AgCl can be used for
Cl− detection, or Ag2S/CuS for the detection of Cu(II). The same Ag2S polycrystalline
matrix can be doped with “insoluble” sulfide of other cations to be detected. Ag2S has
a very low solubility product constant, pKsp ~ 50, therefore it is very stable in aqueous
solutions. Moreover, it is a relatively good ionic conductor, transducing the equilibra-
tion of ion exchange between the pollutant ion in the solution with the dopant ion in
the silver sulfide matrix. There are many other structures, such as fluorides of several
of the rare-earth metals (e.g., LaF3 with unusually low resistivities for F− flow) of ion-
selective solid-state membranes, depending on their target ions or even neutral mole-
cules that can be found in the relevant literature.

– Liquid membrane electrodes. This is an organic membrane material, very viscous,

and immiscible with water, inaccessible to hydrophilic ions from an aqueous solution.
Therefore, an ion-exchanger or ionophore molecules are dissolved in this membrane.
Mostly this membrane is not a free-standing one but is supported by a porous struc-
ture separating the analyte solution from the internal reference solution. In the case
of divalent alkaline earth metal cations, such as Ca2+ or Mg2+, the ion exchanger usu-
ally is an aliphatic diester of phosphoric acid which has a high affinity to this cation.
The ion exchange can be described as:

Ca2+ aq + 2ðROÞ2 PO−2 mebrane $ ½ðRO2 ÞPO2 2 Camembrane

where R is an aliphatic hydrocarbon chain containing 8 to 16 carbon atoms.

As in the case of all ISEs, equilibrium is established at both membrane|solution in-
terfaces. Other ion exchangers may include R‒S‒CH2‒COO− for Cu2+ and Pb2+, substi-
tuted phenantroline ligands for NO3− and BF4−. Ionophores that can be used for such
membranes, depending on analyte ions, include crown ethers, or valinomycin (for K+).
In contrast to charged ion exchangers, the important feature of ionophores is that they
are neutral carrier molecules whose structural cavities fit exactly the dimensions of the
analyte molecule or ion. An excellent example is the antibiotic – valinomycin which ef-
fectively chelates K+ ions into its electron-rich cavity presenting the outer lipophilic part
to the organic hydrophobic membrane. However, from the practical point of view of
monitoring the environment, such liquid membrane ISEs are now quite obsolete due to
the fast development of solid-state membranes as well as polymer membranes, which
provide much more robust sensors showing better selectivities than liquid systems.

– Polymer membrane electrodes. These electrodes constitute more robust and better-
performing systems compared to the “wet” membranes discussed above. In these
membranes, polymer works as a host matrix for ion exchanger or carrier. To improve
the exchange process and ionic mobility, a plasticizer is usually added. The perfor-
mance of these electrodes is highly selective and they have been used for environ-
 13.2 Electrochemical sensors 305

mental and health control purposes to determine ions such as K+, Ca2+, Cl−, and NO3−.
Particularly, measurements of blood potassium levels are carried out with polymer
membrane electrodes containing valinomycin. To date, there is a range of commer-
cially available ISEs that can detect specific ions (i.e., calcium, nitrate, potassium, cop-
per, barium, chloride, etc.). The environmental control applications of ISEs include
the monitoring of concentration levels of trace metals such as iron(III), mercury(II),
cadmium(II), copper(II), lead(II), and chromium(VI), as well as important anions such
as fluoride, phosphate, sulfate, nitrate, nitrite, chloride and cyanide, mostly in natural
waters, but also in tap water and agricultural and industrial drainage and sewage.
The elimination of liquid components in Scheme 13.1 above resulted in the develop-
ment of solid-contact material for ISE, where the reference, internal solution, and elec-
trode (II) were replaced by metal contacts, intermediated by a layer of conducting
polymer. The conducting polymers placed between an electronically conducting metal
contact and ionically conducting ion-selective membrane work as the ion-to-electron
transducers, completing the charge flow and thus connecting the device to the external
electronic circuits reporting the sensor response. Moreover, such configuration simpli-
fies the sensor design and offers unique possibilities for manufacturing miniaturized,
personal sensors also for health control and clinical analysis. The performance of such
solid contact is dictated by its redox state and ionic equilibrium which are affected by
the polymerization conditions and by the composition of the pollutant solution. This per-
formance has been significantly improved down to the nanomolar level, particularly in
terms of the detection limits of solid-contact ISE sensors. Conducting polymers applied
for sensing devices include polyaniline (PANI, pH sensing, used also in the nonaqueous
environment), polypyrrole (ISE for nitrates in waters), poly(3,4-ethylenedioxythiophene),
(PEDOT) or polypyrrole (PPy, e.g., for K+ ISE) and numerous other polymers, co-polymers
and their mixtures, some of them biocompatible, that can be found in the literature.
All these conducting polymer (CP) materials have to possess highly reversible
redox behavior and ease of penetration of the analyte ion to compensate for the
charge in the polymer. It is also obvious, that the polymer layer must form a stable
mechanical contact between the metal electrode (Me) and ISE. Taking an example of
PPy doped with X− anion, the ion-to-electron conversion process, in the case of, for
example, K+ analyte ions, can be illustrated as follows:

PPy+ X− ðCPÞ + K+ aq + eðMeÞ $ PPyðCPÞ + X− ðISEÞ + K+ ðISEÞ

where PPy+X−(CP) and PPy(CP) are the oxidation and neutral (reduced) states of PPy
in the CP phase, respectively, whereas K+aq and K+(ISE) are the concentrations of the
analyte ion (K+) in aqueous and ISE phase, respectively; X−(CP), X−(ISE) are the concen-
trations of X− anion in CP (polypyrrole) and ISE, respectively, whereas e(ME) are the
electrons being transferred between the metal and CP. Please note, however, that this
is the overall ion-to-electron reaction scheme, because, as the careful reader should
certainly notice, this reaction involves three charge transfers at equilibrium at three-
phase boundaries: (i) electron transfer at Me|CP interface, (ii) ion transfer of X− at the
306 13 Sensing the environment

CP|ISE interface and (iii) analyte ion transfer at the ISE|solution interface. All these
equilibria can be described by the Nernstian-type equations.
The potential of the solid-contact CP ISE is the sum of all three interfacial poten-
tials, as discussed above:
RT
E = EMejCP + ECPjISE + EISEjsol = const. + ln½K + aq , (13:7)
F

if the first two potential drops are fixed. Thus, this equation explains the sensitivity to
K+ analyte ions taken as an example. Of course, such an approach can describe the
ion-to-electron conversion process of other systems with ISEs and CPs appropriately
coupled to fulfill the demands of sensor performance in a given environment, pres-
sure, and temperature.

Some technical notes

All potentiometric sensors are based on the same principle of establishing a potential
difference between the sensing and reference electrodes at equilibrium or quasi-
equilibrium. The majority of technical requirements described above in the case of
amperometric sensors have to be met in the construction of potentiometric sensors.
In both types, the elimination of liquid components enables the sensors to work at
elevated temperatures. In many industrial processes, such as metallurgy, power
plants, or the petrochemical industry, high temperatures are a must. In these areas,
the sensors are utilized to analyze and control the emissions to reduce the released
hazardous gases by optimizing the combustion processes. If the sensor operates at ele-
vated temperatures, it usually uses solid ceramic electrolytes or ISEs, such as doped
zirconium oxide, that can work both in amperometric and potentiometric modes.
Such oxygen sensors combining both modes have been recently developed for auto-
motive applications. They are applied to obtain the maximum power from gasoline
combustion by controlling an optimum air-to-fuel ratio. However, the traditional de-
sign of potentiometric zirconia-based oxygen sensors is based on a thimble-type
shape and these sensors are often called the lambda sensors (see Fig. 13.2 below). The
sensors are based on yttria- or scandia-stabilized zirconia (YSZ or SSZ) placed between
two porous Pt electrodes. The thimble shape and sensor mounting in the exhaust sys-
tem provides separation of air (reference) and exhaust gas (measured) oxygen partial
pressures. At a sufficiently high temperature, the oxygen gas, the mobile oxygen ions in
the zirconia lattice, and the electrons in the electrodes are in thermodynamic equilib-
rium. To achieve it, electronic charge transfer occurs between Pt and the zirconia solid
electrolyte as oxygen is introduced or removed from the zirconia lattice. Moreover, the
electrochemical potential of oxygen ions has to be constant throughout the whole sys-
tem. Therefore, the whole system can be represented as an electrochemical cell:
O2 ðpO2 ðgasÞÞ, Ptj YSZ ðor SSZÞ matrix with mobile O2− ionsjPt, O2 ðpO2 ðreferenceÞÞ
 13.2 Electrochemical sensors 307

porous Pt porous Pt
reference sensing
air electrode
electrode
zirconia
exhaust ceramics,
gas flow oxygen ion
conductor

Fig. 13.2: A cross-section of a lambda sensor monitoring the oxygen content in an exhaust gas.

If the sensing electrode and reference electrode are exposed to different oxygen par-
tial pressures, at equilibrium, the electromotive force (EMF) of such sensor (electro-
chemical cell) can be described by the Nernst equation:

RT pO2 ðgasÞ
E = EMF = ti ln (13:8)
2nF pO2 ðreferenceÞ

where ti is the average ionic transference number of mobile ions in the zirconia ma-
trix (for only one ionic species it can be approximated as 1), all other symbols have
their usual meanings. Since this equation contains only thermodynamic quantities,
this sensor is very robust and stable. Nevertheless, it is important to keep in mind
that the above equation implies that only oxygen is involved in the potential deter-
mining reaction. However, the exhaust gases contain also NOx, CO, CO2, SOx, as well as
volatile organic compounds (VOCs), that can undergo a series of electrode reactions
interfering with the sensor readout. Usually, these reactions are nonequilibrium, so to
facilitate and improve the sensor performance; the sensing electrode also contains
catalytically active materials and is operated at temperatures above 600 °C. At the end
of this section it is also worth noting the similarity of the design of the potentiometric
lambda sensor with doped zirconia in between the Pt electrodes with the amperomet-
ric sensor design shown in Fig. 13.1C, where the ion-conducting ISE (can be zirconia)
was placed between the anode and cathode polarized from the external source and
generating the flow of electric current as a sensor output signal.

13.2.3 Conductometric sensors

Conductometric sensors rely on modulating their electrical conductivity in response to

the environment. Conductivity, σ, or specific conductance (reciprocal of resistivity, ρ) is
a material property normalized with respect to area, potential gradient, and time:
308 13 Sensing the environment

j
σ= (13:9)
E

where j is the current density passing through the sensor placed in the potential gra-
dient E. The conductometric sensors are also named chemiresistors to clearly indicate
their functions. Just to remind that ρ = 1/ σ is defined in the second Ohm’s law:

l
R=ρ (13:10)
A

where R is the resistance of the conducting sample, l and A are its length and contact
area, respectively, whereas ρ is the sample’s intrinsic property directly related to the
number of charge carriers within the sample volume. The inherent dependence be-
tween current and resistance/conductance (first Ohm’s law) sometimes causes prob-
lems in separating the amperometric and conductometric sensors. Both types of
sensors are electrochemical cells and as such they should be treated with general
rules of thermodynamics and electrochemistry. In considering the sensor response,
for the sake of simplicity, let us take into account only the pollutant/analyte influence
on the overall conductivity of the sensor material. A more detailed analysis can be
found in [8]. Let us consider further bipolar arrangements in the sensor construction,
in which the chemiresistive material separates two electrodes as in Fig. 13.3.

pollutant gas
contact
contact

sensor layer

Fig. 13.3: A chemiresisting sensor layer separates

two contact electrodes. The sensor is placed on a
insulating housing, heater housing/heater layer to improve its performance.

This arrangement is suitable for controlling the gaseous environment in nonconduct-

ing liquids because the conducting samples would short out the chemiresistor. The
conductivity of sensing layer for all types of conductors generally depends on the fol-
lowing parameters:
1. Charge carrier mobility (ui).
2. Number of charge carriers (ni).
3. Elementary charge (e).
4. The valency of the mobile charges (zi).

All these parameters are related to conductivity, σ, by the following equation:

 13.2 Electrochemical sensors 309

X
σ=e ui ni zi (13:11)
i

This equation is exactly the same as eq. (1.76), however, it is related to single charges,
not to the molar concentration of charges, as in the case of eq. (1.74).
Since the vast majority of conductometric sensors are employed for the detection
of hazardous gases at high temperatures and aggressive environments, the heater as
shown in the schematic graph above is not necessary, however, at ambient temper-
atures it is generally required to improve the sensors’ performance.
The best sensing layers are usually made of semiconducting solids and therefore,
the overall conductivity of the sensor depends on the net concentration of available
free charge carriers in the semiconducting material:

σ = e cn · μn + cp · μp (13:12)

where e is the elemental charge, μn, and μp are the mobilities of electrons and holes,
respectively, whereas cn , cp are the concentrations of electrons and holes, respectively.
The vast majority of conductometric gas sensors are based on oxide-type semicon-
ductors, such as ZrO2, ZnO, or SnO2 (n-type), or NiO, CoO (p-type). Although the exact
mechanism of conductometric changes of the semiconductor gas sensors under the
influence of pollutants is still under thorough examination, it is accepted that they are
caused by surface adsorption and catalysis. Thus, the principle of operation is based
on the presence of surface states forming the space charge layer (discussed earlier in
Section 10.2.2), that can exchange electrons with the bulk of the semiconductor. The
surface states can be formed by the adsorption of ambient (air) gas molecules (e.g.,
oxygen – an electron acceptor or hydrogen – donor). In the case of an n-type semicon-
ductor, the adsorbed oxygen species capture electrons from the inner of the semicon-
ductor forming a carrier depletion layer at the surface, reducing the semiconductor
conductivity. If the sensing reaction involves oxidation of a pollutant, the electrons
delivered to the semiconductor increase its conduction, whereas in the reduction pro-
cess, involving the electrons from the pollutant, the conduction decreases. For p-type
semiconductors, the situation will be the opposite. A more detailed description of the
operation principles of semiconducting sensors can be found in [8].
This simple case discussed above assumes a monocrystalline lattice of a semicon-
ducting sensing element, however, the real semiconducting sensors utilized for envi-
ronmental control are polycrystalline or even amorphous. Then, the morphology, the
grain contacts, and other parameters that are beyond the scope of this book become
also important.

Some technical notes

The tunable conducting behavior and high sensitivity of their surface states to changes
in the composition of the environment led to the great success of semiconductor con-
310 13 Sensing the environment

ductometric gas sensors. These devices can be simple, easy to manufacture on a large
scale, compact, and durable, therefore can be designed at reasonable production costs.
The semiconducting gas sensors have also some drawbacks. The most important one,
particularly for metal oxide semiconductors, is their typically high operation tempera-
ture, usually above 500 °C. The elevated temperature is required to mitigate their high
sensitivity to moisture and improve their response time and sensitivity. On the other
hand, such temperatures may cause poor stability and therefore induce poor energy
consumption efficiency. Nevertheless, since they are relatively simple and robust, they
are widely put to practical use when the sensing processes are necessary in demanding
environments and at elevated temperatures, (e.g., combustible hazardous gases from
metallurgy, refineries, or exhaust from transportation), where the carrier mobilities
and reaction kinetics are advantageous for sensors’ performance. In order to alleviate
the need to heat the sensing element and thus decrease the level of power consumption
without compromising the sensor performance, the development of chemiresistive gas
sensors, new gas sensor materials are being sought. For this purpose nonoxide semicon-
ductors, such as cadmium sulfide, cadmium selenide, and cadmium telluride, and their
composites have been explored. It was reported that these materials offer an additional
advantage of enhanced pollutant gas response, that is, the electrical conductivity
change ratio (gas sensitivity), under illumination compared to those in the dark. This
phenomenon is explained by the effect of light on gas adsorption/desorption and excita-
tion of the electrons in the semiconductor. Also, their operating temperatures are much
closer to the room temperature, decreasing the power consumption and diminishing
the operational safety issues.
Other possible directions that are based on rapidly developing nanotechnologies,
include nanostructurization of the sensor design (nanowires, nanobelts) [9], combin-
ing multiple types of semiconducting sensing materials [10, 11].

13.3 Future perspectives and challenges

Permanent heterogeneity in time and space of the pollutant in various environments,

such as aquatic, earth and air, complicates monitoring and appropriate, fast re-
sponses. This is particularly complicated in a largely unpredictable environment – at-
mosphere. Air pollution is one of the main challenges to health and environmental
sustainability. According to the World Health Organization, the vast majority of the
world’s population lives in places where air quality exceeds the limits recognized as
safe. The concentration of air pollutants depends on localization, population and in-
dustrial densities as well as time. Taking an urban environment as an example, its
local atmosphere is affected mainly by anthropogenic and natural processes and
varies in cycles related to urban activities and related emissions. Thus, better sensing
and local pollution assessment is a must. Therefore, the static, traditional monitoring
at fixed locations becomes frequently insufficient. In response to these needs, backed
 13.3 Future perspectives and challenges 311

by the new capabilities in microelectronics, microfabrication, and nanomaterials sci-

ence, new research in portable sensing device design and development is rapidly
growing. This may finally lead to personalized air pollutant sensors and monitoring
systems, assessing both the health and pollutant exposures of individuals. However,
proper assessment of the relation between the exposure and negative health effects
on humans (to say nothing on other organisms) is limited by the availability of reliable
data, collection and computational/modeling capabilities. One of the pathways to over-
come these difficulties is the development of miniaturized sensor arrays that can work
in liquid or atmospheric conditions. If each sensor in the array has different sensitivity
and selectivity, and the array is coupled to signal processing and pattern recognition
engine, then such system is called electronic tongue (e-tongue for liquids) or electronic
nose (e-nose for air). This is roughly shown by the scheme below (Scheme 13.2):

Sensors array Pattern

Array
with individual recognition Output
sensing processor
engine

Train Test

Knowledge base

Scheme 13.2: Diagram illustrating the functioning of e-tongue or e-nose. Adapted after [10].

Thus, the combination of modern electrochemical techniques with breakthroughs in

microelectronics and miniaturization allows the introduction of powerful analytical
devices for effective pollution control. This in turn, coupled with improvement of the
sensor response, development of machine learning and deep learning can pave the
way for pollution monitoring even at a personal, miniaturized level. However, regard-
less of the type of output signal being generated, the final decision-making for a
greener environment is in our hands.
One final word is worth adding: where environmental sensing and power supply
start to synergize, new opportunities for multifunctional devices based on similar
chemistry to store energy or detect environmental hazards (or both), are emerging
simply by the choice of electrodes and electrochemistry. One of such example can be
self-powered sensors with the ability to scavenge energy from the environment to
self-drive the sensor’s operation. They can harvest and convert into electric energy,
solar, mechanical, thermal energies, etc., that are available in the environment for
long-term sustainable, “green” operation [12].
312 13 Sensing the environment

Bibliography

[1] Kim M-Y, Lee KH. Electrochemical sensors for sustainable precision agriculture – a review. Front.
Chem. Sec. Electrochem. 2022, 10, 1–14.
[2] Simoes FR, Xavier MG. Electrochemical sensors. In: Nanoscience and its applications. Elsevier, 2017,
155–178.
[3] Hussain CM, Keçili R. Electrochemical techniques for environmental analysis. In: Modern
environmental analysis techniques for pollutants. Elsevier, 2020, 199–222.
[4] Baranawal J, Barse B, Gatto G, Broncova G, Kumar A. Electrochemical sensors and their applications:
A review. Chemosensors 2022, 10, 363.
[5] Electrochemical biosensors. In: Cosnier S, ed. Pan Stanford series on the high-tech of biotechnology,
Vol. 3. Pan Stanford Publishing Pte. Ltd, 2015.
[6] Xu T, Scafa N, Xu L-P, Su L, Li C, Zhou S, Liu Y, Zhang X. Electrochemical sensors for nitric oxide
detection in biological applications. Electroanalysis 2014, 26, 449–468. doi: 10.1002/elan.201300564.
[7] Brown MD, Schoenfisch MH. Electrochemical nitric oxide sensors: Principles of design and
characterization. Chem. Rev. 2019, 119, 11551–11575.
[8] Janata J. Principles of chemical sensors, Chapter 8, Conductometric sensors. Springer US, 2009. doi:
10.1007/978-0-387-69931-88.
[9] Neri G. Thin 2D: The new dimensionality in gas sensing. Chemosensors 2017, 5, 21. doi: 10.3390/
chemosensors5030021.
[10] Korotcenkov G, Cho BK. Engineering approaches for the improvement of conductometric gas sensor
parameters Part 1. Improvement of sensor sensitivity and selectivity (short survey). Sens. Actuators
B 2013, 188, 709–728.
[11] Ando M, Kawasaki H, Tamura S, Haramoto Y, Shigeri Y. Recent advances in gas sensing technology
using non-oxide II–VI semiconductors CdS, CdSe, and CdTe. Chemosensors 2022, 10, 482.
[12] Alagumalai A, Shou W, Mahian O, Aghbashlo M, Tabatabaei M, Wongwises S, Liu Y, Zhan J, Torralba
A, Chen J, Wang ZL, Matusik W. Joule 2022, 6, 1475–1500. doi: doi.org/10.1016/j.joule.2022.06.001.
14 Green fuel: hydrogen production

14.1 General remarks

As global warming and pollution of the environment are increasing, the investigation
and development of renewable energy sources are crucial. The most promising source
of clean energy, considered as the fuel of the future, is hydrogen.
Hydrogen plays an essential role in many industrial processes such as the produc-
tion of ammonia (Haber process), the production of methanol, oil refining, and others.
As a fuel, hydrogen can directly react with oxygen (air) in the fuel cells to provide
electrical energy or can be burned in combustion engines instead of fossil fuel. So it is
not surprising that the production of hydrogen increases every year and the global
hydrogen market is expected to grow to about US$220 billion by 2030. Hydrogen can
be produced from various nonrenewable and renewable energy sources. Currently,
the main sources for commercial production of hydrogen are fossil fuels: natural gas,
coal, and oil. Application of fossil fuels accounts for about 95% of total H2 production
worldwide. Plenty of methods are used for hydrogen generation from fossil fuels such
as steam reforming of natural gas, partial oxidation of heavy hydrocarbons, or coal
gasification. Other processes and technologies currently developed include direct
methanol reforming, reforming and pyrolysis of biomass, and fermentation of bio-
mass. However, in many methods besides hydrogen, carbon dioxide is formed, which
is emitted to the atmosphere, causing temperature increase and climate changes. To
reduce the amount of CO2 released into the environment, some technologies called
carbon capture and storage (CCS) have been developed. Currently, in industry, hydro-
gen is mainly produced from natural gas (methane) by using the steam methane re-
forming (SMR) method. In this way, about 50% of industrial H2 is generated. In the
first stage of the SMR method, methane (natural gas) is heated to the temperature of
700–1,000 °C at a pressure of 3–25 bar in the presence of steam and a proper catalyst
(nickel). In this step, except hydrogen, carbon monoxide is produced. In the second
stage, carbon monoxide and steam react at a lower temperature of about 360 °C, form-
ing carbon dioxide and hydrogen in a so-called water-gas shift reaction:

CH4 + H2 O ! CO + 3H2 first stage, endothermic reaction

CO + H2 O ! CO2 + H2 second stage, exothermic reaction

Finally, the carbon dioxide and other impurities have to be removed from the gas
stream. When CO2 is released into the atmosphere, the obtained H2 is often termed “gray
hydrogen,” and when it is captured (separated) by CCS processes, it is termed “blue hy-
drogen.” Comparing hydrogen as a fuel with common fossil fuels used, we can find some
advantages and disadvantages. The gravimetric energy density of hydrogen (141.8 MJ/kg)
is about 2–3 times higher than that of some fossil fuels, which means that hydrogen can
generate several times more energy than conventional fuels. For instance, the gravimet-

https://doi.org/10.1515/9783111160986-015
314 14 Green fuel: hydrogen production

ric density energies of methane and diesel are 55.6 MJ/kg and 45.4 MJ/kg, respectively.
Hydrogen has the lowest flash point (−231 °C), so even at very low temperatures enough
hydrogen evaporates for ignition. The flammability range of hydrogen is the widest
(4–75%) among other fuels. It is its disadvantage, but on the other hand, as water vapor
is produced during the combustion of hydrogen, flames are not toxic. An alternative
technology applied for the production of hydrogen from fossil fuel is water electrolysis.
Using these technologies, very pure hydrogen can be obtained (up to 99.99%), suitable
for application in fuel cells. More information about other methods of hydrogen produc-
tion, hydrogen economy, and technologies of production and storage can be found in [1].

14.2 Water electrolysis

Water electrolysis is the simplest method of hydrogen production. In the process, the
reactant, renewable H2O, is decomposed to pure H2 and O2 by utilization of the electri-
cal energy. Currently, only about 4% of total H2 world production is generated by
using water electrolysis. To date, the cost of hydrogen production through water elec-
trolysis is too high to be economically competitive with established technologies of
hydrogen production from natural fossil fuels (about 5 times). This is due to the high
electrical energy demand for electrolysis and low efficiency, resulting from energy
losses during the reaction of water splitting. Plenty of efforts were and are made for
the development of existing technologies of water electrolysis, to make them more ef-
fective and less expensive. Much research has been done related to the improvement
of the efficiency of electrolytic reactions, the finding of new low-cost catalysts, or the
application of alternative, renewable sources of electrical energy, such as wind en-
ergy or solar energy. Hydrogen obtained in this way is termed “green hydrogen.”
A few types of methods and electrolytic cells (electrolyzers) can be used for water
splitting. Two of them, alkaline water electrolysis (AWE) and polymer electrolyte mem-
brane electrolysis (PEM), are applied in the industry, however, on different levels. As
the technology of AWE is very well established, it is widely used for H2 production on a
big scale, while PEM is applied in small factories. Other methods such as solid oxide
electrolysis (SOE) and microbial electrolysis cell (MEC) are still developed. All these
methods are described briefly in Section 14.2.2. More comprehensive information re-
garding water electrolysis is given in reference [2], and others are cited in the text.

14.2.1 Thermodynamics, energy losses, and efficiency

On a laboratory scale, the electrolysis of water can be carried out in simple electro-
chemical two-chamber cells containing two planar electrodes: a cathode and anode
made from inert metal (mostly Pt), and an aqueous solution of electrolyte, by using a
DC electrical power source. Electrolysis of pure water is also possible, but as the con-
 14.2 Water electrolysis 315

ductivity resulting from hydronium (H+) and hydroxide (OH–) ions (5.7 × 10−6 S/m, Kw
= 1 × 10−14) is low, the process requires more electrical energy than in the case of aque-
ous electrolyte solution. Different electrolytes can be applied, but their ions cannot
undergo any competitive electrode reactions. Because of that, the standard redox po-
tentials of electrolyte cations should be lower than the standard redox potential of H+
and the standard redox potential of electrolyte anions should be higher than the stan-
dard redox potential of OH– ions. The chambers containing the cathode and anode
are separated by a diaphragm or an ion-conducting membrane. The reactions that
take place at the cathode and anode depend on the pH of the solution.
In acidic media, they are as follows:

Cathode: 4H+ ðaqÞ + 4e− ! 2H2 ðgÞ, E0 = 0 V (14:1a)

Anode: 2H2 OðlÞ ! O2 ðgÞ + 4H+ ðaqÞ + 4e−, E0 = 1.229 V (14:1b)

In neutral or basic solution, hydroxide ions participate in an oxidation reaction:

Cathode: 4H2 OðlÞ + 4e ! 2H2 ðgÞ + 4OH−ðaqÞ, E0 = − 0.828 V (14:2a)

Anode: 4OH−ðaqÞ ! O2 ðgÞ + 2H2 O ðlÞ + 4e−, E0 = 0.401 V (14:2b)

E0 means the standard electrode potential of an electrode in equilibrium, related to

SHE (Pt/H2, T = 298.15 K, H2 (p = 1 bar, 1 × 105 Pa), H+ (a = 1):

ΔE0 = EA0 − Ec0 = 1.229 V∼1.23 V

2 2
RT pH2 pO2
ΔEeq = ΔE0 − ln 2 (14:3)
4F aH O
2

In (14.1) and (14.2), the overall reactions are the same, gaseous hydrogen and oxygen
are generated at the cathode and anode, respectively:

2H2 OðlÞ ! 2H2 ðgÞ + O2 ðgÞ

The reaction is the inverse of the hydrogen–oxygen fuel cell reaction, where the chemi-
cal energy of the reaction is transformed into electrical and thermal energy. As a matter
of fact, some fuel cells can operate also as water electrolysis cells (electrolyzers), de-
pending on the current direction. The formation of H2 and O2 from water requires en-
ergy. This energy is equal to the enthalpy reaction ΔrH of water decomposition:

Δr H = Δr G + TΔr S (14:4)

where ΔrG is the Gibbs free energy of the reaction and ΔrS is the entropy of the reac-
tion, and can be supplied to electrolytic cells in the form of electrical and thermal
(heat) energy or only as electrical energy. The thermal energy compensates for the
losses connected with changes in the entropy of the reaction and is equal to product
TΔrS. The minimal electrical energy required for water decomposition is described by
316 14 Green fuel: hydrogen production

the Gibbs free energy of reaction ΔrG. When energy is delivered to the electrolytic cell
only in the form of electrical energy, the amount of electrical energy should be higher
as the entropy loss has to be compensated.
In standard conditions (T = 298.15 K, p = 1 bar), the total energy (work + heat) of
decomposition of 1 mol of water is described by the standard enthalpy of reaction
ΔrH0 equal to 285.83 kJ/mol. The contributions of the electrical energy (work) ΔrG0 and
thermal energy TΔrS0 (heat) are 237.13 kJ/mol and 48.7 kJ/mol, respectively.
Let us consider two cases. First is when the heat energy 48.7 kJ/mol is supplied to the
electrolysis cell (for instance, by thermostat) together with electrical energy, and the sec-
ond is when only electrical energy is used. In the first case, assuming that all the thermal
energy is consumed during the reaction, we can calculate the equilibrium potential of
electrolysis cell, ΔE0 , by using the equation w = ΔrG0 = nFΔE0 (compare equation (8.3)
and note the positive sign of work, because it is introduced into the cell), where n = 2,
F = 9.648 × 104 C/mol, ΔrG0 = 237.13 kJ/mol. ΔE0 is equal to 1.229 V and is also the standard
minimal potential of water splitting referred to the low heating value (LHV) of hydrogen,
equal to 241.8 kJ/mol or 120 MJ/kg. LHV is described as the thermal energy of hydrogen
combustion and formation of water in a vapor state and can be determined by subtrac-
tion of the heat of water vaporization (ΔvH0 = 44.0 kJ/mol) from the enthalpy of water
formation in a liquid state. When electric energy is only applied for water electrolysis,
without heat contribution, we used for potential calculation the value of total energy,
which means the enthalpy of reaction. In this case, ΔrH0 = nFΔEtn 0
, where n = 2, F = 9.648 ×
10 C/mol, ΔrH = 285.83 kJ/mol, ΔEtn is termed thermoneutral potential of electrolysis cell
4 0 0

and is equal to 1.481 V. The thermoneutral potential is also the minimal potential that
should be used for the electrolysis of liquid water in standard conditions. This potential
is referred to as the high heating value (HHV) of hydrogen equal to 285.83 kJ/mol or 142
MJ/kg. HHV is equal to the heat of hydrogen combustion and formation of water in its
liquid state (enthalpy of water formation in liquid state at standard conditions).
Comparing the standard values of free Gibbs energy and enthalpy of reaction, it is
clearly seen that about 20% more energy is required for electrolysis when only electri-
cal power is applied. The increase in temperature causes the decrease in ΔrG and influ-
ences the minimal potential of water splitting. For instance, at 150 °C, ΔrG✶ is equal to
221.49 kJ/mol and E✶ to 1.15 V (asterisk means not standard temperature). The influence
of T on water splitting is more significant at higher temperatures. The increase of tem-
perature up to 1,500 °C reduces the value of the reversible thermodynamic potential of
water decomposition E✶ to ∼0.7 V [3]. Considering only thermodynamic factors, we as-
sumed that reactions occur ideally and all energy losses in operating electrolytic cells
can be neglected. In practice, to overcome the energy losses and to maintain the elec-
trolysis of water, we have to apply higher values of potentials (higher than 1.23 V or
1.48 V). This difference in potentials called overpotential is described by the equation

η = ΔE − ΔEeq (14:5)
 14.2 Water electrolysis 317

where ΔE is the potential applied to the electrolysis cell and ΔEeq is the difference of
equilibrium potentials of electrodes in the cell described in equation (14.3), when cur-
rent does not flow (I = 0). If we consider the reaction at one electrode (cathode or
anode), this difference is the measure of polarization of electrode η = E − Eeq .
A few kinds of overpotentials can be considered in the operating electrolytic cell:
(i) ohmic overpotentials, ηohm, often termed Ohmic losses; (ii) activation overpotential
ηa originated from hindrances in the kinetics of electrode reaction at cathode and
anode; (iii) concentration overpotentials, ηc, resulting from problems with transporta-
tion of reactants or products to/from electrodes.
Therefore, the potential that has to be applied for water electrolytic cell is de-
scribed by the equation:
X    
ΔE = ΔEeq + jηj = ΔEeq + ηohm + ηa,C + ηa,A + ηc,C + ηc,A (14:6)

C and A represent cathode and anode, respectively.

In practice, the two overpotentials, ηohm and ηa, significantly increase the mini-
mal potential required for water electrolysis. The concentration overpotential is gov-
erned by the diffusion process, depending on the concentration gradient of reactants,
and can be eliminated to a greater extent by a continuous flow of reactant (H2O) to
electrodes and efficient removal of products H2 and O2. For these purposes, special
systems for water pumping and circulation, gas separators, and collectors are de-
signed for the different types of electrolyzers. Much more about the origin of concen-
tration overpotential, you can find in Section 1.1.3.2.
To carry out the electrolysis of water with good efficiency, we should minimize
the losses of energy caused by ohmic and activation overpotentials. The ohmic over-
potential depends on the internal resistance of the cell Rin :

ηohm = IRin (14:7)

where I is the current flowing through the cell.

There are a few components of internal resistance: (i) resistance of the electrolyte
solution depending on the conductivity of the electrolyte and on the volume of the
electrolyte between electrodes, and (ii) resistance of electrodes depending on the con-
ductivity of electrode materials, on the area of the electrode, and the special design of
electrodes with dispersed electrocatalysts. Note that the working, active surface of po-
rous electrodes loaded with dispersed catalyst is difficult to determine. If planar,
hemispherical, or disk electrodes are applied, the geometrical area of the electrode is
usually used for density current calculation. In other cases, special methods should be
used to determine the real, active electrode area. However, the main factor determin-
ing the internal resistance of the cell is the design of separators of the cathodic and
anodic parts of the electrolytic cell. Depending on the type of the electrolytic cell, the
polymer ion conducting membranes or inorganic diaphragms are used as separators.
Their resistance depends on many factors; for instance, the resistance of proton con-
318 14 Green fuel: hydrogen production

ducting membrane depends on hydration, the ionic state of the membrane, history of
membrane preparation, and treatment. In electrolyzer operating at high tempera-
tures, in the range 800–1,000 °C, the solid oxide (SO) electrolytes are used, their con-
ductivity depends on composition and temperature, and may be increased by special
treatment. By good engineering, the ohmic losses can be reduced to ∼0.1 V.
The activation overpotential ηa has its origin in the kinetics of reactions, their
slowness, their irreversibility, and the formation of new phases and is minimized by the
application of proper electrocatalysts. The relation between ηa and current density (j)
can be evaluated from the Butler–Volmer equation (see Section 1.1.3.3). For higher values
of applied potential or current, this relation is known as the Tafel equation:

ηa = a + blnj jj (14:8)

where j is the current density

RT RT
a= lnj0 , b= − for E  Eeq
αnF αnF
RT RT
a= − ln j0 , b= for E  Eeq
βnF βnF

R, T, F, and n have the usual meaning, α, β are transfer coefficients, j0 is the exchange
current density, and coefficient b is often called the Tafel slope.
Note that these relations are accomplished, when the reaction is going at one
working electrode and also when we consider the operating cell with different reac-
tions on the cathode and anode.
In this case

RT j
ηa,C = − ln (14:9a)
αnF j0,C

and
RT I
ηa,A = ln (14:9b)
βnF j0,A

The exchange current describes the oxidation and reduction currents, which flow
through the electrodes at equilibrium potential and is proportional to the standard
rate constant of the reaction (1.104). As the standard rate constant depends on the
standard Gibbs free energy of activation of the reaction (1.92), all factors that decrease
the activation energy of the reaction cause the increase of the rate constant and the
exchange current. In consequence, the activation overpotential decreases. Another pa-
rameter, the Tafel slope (b), gives us insight into the reaction mechanism providing
information on elementary steps and rate-determining steps (rds). To increase the re-
action rate, catalysts play a crucial role (see Chapter 4).
 14.2 Water electrolysis 319

In the case of water electrolysis, the HER (hydrogen evolution reaction) at the
cathode and OER (oxygen evolution reaction) at the anode are multistep reactions
with intermediate formation, influencing the rate of reactions and efficiency. Path-
ways of HER are much simpler than OER. In the first step of HER, in an alkaline solu-
tion, the water is split with the formation of hydrogen adsorbed (Volmer step) on the
free active site of electrode/catalyst (M):

H2 O + e $ Hads ðMÞ + OH−

This is followed by an electrochemical step of hydrogen formation (Heyrovsky step)

or by chemical recombination of adsorbed hydrogen atoms (Tafel step):

H2 O + Hads ðMÞ + e $ H2 + OH−

2Hads ðMÞ $ H2

The overall HER is a two-electron reaction. The crucial factor, which determined the
rate of HER and its activation overpotential is Gibbs free energy of H adsorption,
ΔG0ads,H – a measure of the binding energy of H with electrode surface. If the binding
energy is too low or too high, the exchange current densities of HER are low, in the
range of 10−9–10−7A/cm2. The moderate values of binding energy facilitate the HER.
This is the case of Pt and metals of Pt group – Ir, Rh, and also of Co and Ni. Their H
binding energy is in the range of 260–280 kJ/mol, and the values of exchange current
densities lay in the range of 10−4–10−2 A/cm2, depending on the catalyst and solution
used. Pt is considered the best catalyst for HER in acidic and alkaline media, but as
the first step in alkali media is the sluggish reaction of water dissociation, the rate of
HER is much lower than in acidic media (about 2 orders of magnitude). Generally, for
good HER electrocatalysts, the overpotential should be lower than 0.1 V.
The OER is a much more complex reaction than HER and requires much higher
overpotentials. To date, the mechanism of OER is discussed and further evaluated.
Generally, the adsorption evolution mechanism is accepted assuming the four elec-
tron transfer steps with the formation of oxygen-containing intermediates, their ad-
sorption, and binding with the active side (M) of the electrode/catalyst. The formation
of oxygen-containing radicals such as •OH, O•, and HOO• (HO2• notation is also used) is
postulated. In alkali media, the first step is the charge transfer step with the formation
of adsorbed •OH:

OH − ads ðMÞ $ 
OHads ðMÞ + e

It is followed by the further charge transfer and formation of adsorbed O•:

OH− +  OHads ðMÞ $ O ads ðMÞ + e + H2 O

320 14 Green fuel: hydrogen production

Thereafter, two pathways may be considered: one similar to the Tafel reaction of hy-
drogen formation:

2O ads ðMÞ $ O2 + 2e

And the other with the formation of HOO• (HO2•) radicals and O2 in the reactions

O ads ðMÞ + OH− $ HOO ads ðMÞ + e

HOO ads ðMÞ + OH− $ O2 + H2 O + e

The oxygen-containing intermediates (radicals •OH, O•, and HOO•) are adsorbed on
the catalyst surface through O atoms forming one (•OH, HOO•) or two bonds (O•). The
binding energy of these species with catalysts decides about the values of overpoten-
tials of OER and rds. If the oxygen of the intermediate binds with the catalyst surface
too strongly, the rds is the formation of HOO•, if too weakly – the rds is the oxidation
of adsorbed OH– ions. As in the hydrogen case, the best catalysts for OER have moder-
ate values of binding energy. At present, the best catalysts applied for OER are IrO2,
RuO2, or mixed oxides; however, even the benchmark IrO2 catalyst has an overpoten-
tial of about 0.3–0.4 V, at a current density of 10 mA/cm2. Plenty of efforts are made to
improve the efficiency of electrocatalysts in OER. Much more about the development
and progress in the field of electrocatalysts for HER and OER can be found in reviews
[4–7] and references therein.
Theoretically, in standard conditions, the minimal potential that should be used
for water splitting is 1.229 V (or 1.481 V), but as a result of energy losses, the potentials
demanded are higher. In practice, most water electrolyzers operate in the potential
range of 1.6–2.0 V depending on the type of electrolyzer, its construction, and operat-
ing condition.
If we want to estimate the usefulness of the water electrolytic cells for hydrogen
production we have to compare their efficiency. Many formulas are applied, and
there is controversy about what value HHV or LHV should be used in calculations.
For electrochemical processes, the Faradaic efficiency is a basic dependence. In
the case of hydrogen production, it can be defined as the ratio of the amount of hydro-
gen (in mole, volume) evolved during electrolysis and the theoretical amount of hy-
drogen that should be produced in the same electrolytic condition. The theoretical
amount of H2 is calculated by using Faraday’s law, assuming 100% of Faradaic effi-
ciency; the amount of evolved H2 during electrolysis may be determined by using gas
chromatography or other methods:

amount of H2 produced
Faradaic efficiency =
amount of H2 calculated

It is accepted in Europe and recommended that the water electrolysis efficiency

should be calculated by using HHV of hydrogen. The standard potential correspond-
 14.2 Water electrolysis 321

ing to the HHV, termed thermoneutral potential, is equal to 1.481 V (see above). The
formulas for the calculation of the efficiency of an electrolytic cell are as follows:

Thermoneutral potential
Potential efficiency ðHHVÞ =
Operating potential

The operating potential means the external applied potential for the cell:

HHV of H2 production
Electrical efficiency ðHHVÞ = ,
Electrical input

The electrical input means the total electrical energy applied for electrolysis.
The same formulas are used when the LHV value is applied instead of HHV.
Therefore, it should be pointed out which value is used for calculation. The formulas
described above apply to the electrolytic cells (electrolyzers) operating at tempera-
tures below the boiling point of water, termed low-temperature electrolysis (LTE)
cells . At present, the high-temperature steam electrolysis (HTE) cells are developed,
operating in the range of temperature 700–1,000 °C. At such temperatures, the Gibbs
free energy of water splitting (ΔrG✶) and enthalpy of reaction (ΔrH✶) are much lower
than at standard temperature, resulting in a significant decrease of thermodynamic
potentials ΔE✶ and ΔEtn✶ calculated on their basis. In the case of HTE, the following
formulas are used:

Thermodynamic potential ðΔE✶ Þ

Potential efficiency =
Operating potential
HHV of H2 production
Electrical efficiency =
Electrical input + heat supplied

14.2.2 Methods and technologies

Depending on the operating temperature, two types of electrolysis method and elec-
trolytic cells (electrolyzers) can be distinguished: low-temperature electrolysis (LTE)
and high-temperature electrolysis (HTE). To the first group belongs the alkaline water
electrolysis (AWE) and polymer membrane electrolysis (PEM), to the second the solid
oxide electrolysis (SOE). At present, low-temperature technologies such as AWE is
used for hydrogen production on a large industry scale, while PEM electrolyzers are
applied on a lower scale, in small factories. SOE technology is still under development.
In the last years, the other low-temperature method of electrolysis, microbial electrol-
ysis (MEC), is investigated and developed at the laboratory level.
322 14 Green fuel: hydrogen production

14.2.2.1 Alkaline water electrolysis (AWE)

The technology of AWE was developed for many years (about a hundred) and is very
well established. The principles of AWE are very simple. The electrolyzer contains an
aqueous solution of 25–30% KOH or NaOH and two electrodes – cathode and anode –
separated by the porous diaphragm. On the cathode, the reduction of water takes
place with the formation of hydrogen atoms that recombine forming gaseous H2. In
solution, the hydroxide ions are charge carriers. They move across the diaphragm in
the electric field to the anode where they are oxidized to water and O2 (see reactions
(14.2a) and (14.2b)). Both H2 and O2 are removed from the vicinity of electrodes by col-
lectors. Figure 14.1 illustrates the principle of alkaline electrolyzer operation.

C A
H2 O2

H2 O2

OH-

H2O D OH-

Fig. 14.1: Schematic illustration of alkaline water

30% KOH inlet inlet 30% KOH
electrolyzer. A, anode; C, cathode; D, diaphragm.

As pointed out in Section 14.2.1, the HER and OER are complicated multistep reactions
with the formation of adsorbed, intermediate products, which determine the rate of
evolution of H2 and O2.
This rate can be increased by using a proper catalyst as an electrode material. For
HER, the best catalysts are Pt and the metal of its group, and for OER, iridium oxide
and ruthenium oxide are often used. However, as the cost of these materials is high,
the other materials are applied in the commercial electrolyzers. Since in alkali media
the steel and Ni are stable and resistant against corrosion, steel plates plated with Ni are
commonly used as electrodes. However, at high operating temperatures, Ni can react
with hydroxide ions forming nickel hydroxides. To eliminate such reactions, some metals
(Fe, V, and Co) are added to electrode materials. The other Ni materials with high surface
area were also tested such as Ni–Zn and Ni–Al alloys, or hot galvanized Ni meshes. A
very important part of the electrolyzer is the diaphragm in the form of porous foil with a
thickness between 0.05 and 0.5 mm. The diaphragm should avoid mixing of produced
gases and be permeable for hydroxide ions. In alkaline electrolyzers, asbestos porous di-
aphragms are commonly used. However, as asbestos is toxic and undergoes corrosion at
higher temperatures, other materials such as Zirfon, a composite material of zirconia
and polysulfone, or the composite of potassium titanate (K2TiO3) fibers with polytetra-
fluoroethylene are tested. The new approach in AWE is the application of anion-
 14.2 Water electrolysis 323

exchange membranes (AEM) made up of polymer with anionic (hydroxide ions) conduc-
tivity. Conventionally working alkaline electrolyzers operate typically at temperatures
and pressures in the range between 60 and 90 °C, 10 and 30 bars, and in the cell potential
and current density range of 1.8–2.4 V and 0.2–0.6 A/cm2, respectively. Energy efficiency
is 70–80%. Hydrogen produced in this way has a purity of 99% [2, 8, 9].

14.2.2.2 Polymer electrolyte membrane electrolysis (PEM)

PEM water electrolysis technologies are similar to those applied in PEM fuel cells (Sec-
tion 8.1.3.4). The PEM electrolyzer contains a membrane electrode assembly (MEA) to-
gether with current collector and plate separators. MEA consists of a polymer-exchange
proton membrane, covered on both sides by proper catalysts for anode and cathode re-
actions. The ionomer solution is added to the vicinity of catalytic layers increasing the
proton transport from catalysts to the membrane. Figure 14.2 illustrates the operating
principle of the PEM electrolyzer.

gas diffusion
layer C A

H2 O2 O2
H2 H2 O2

H+

unused H+ H2O
H2O H2O

Cathodic Anodic
catalyst layer PEM catalyst layer

Fig. 14.2: Schematic illustration of PEM electrolyzer. A, anode; C, cathode; PEM, proton-exchange
membrane.

The water is pumped to the anode reaching the catalyst layer, where it is oxidized to
protons and oxygen. Protons go from the anode through the proton conducting mem-
brane to the catalyst layer of the cathode, where hydrogen is generated (see reactions
(14.1a) and (14.1b)). The gases left the cell via anodic and cathodic current collectors
and separator plates made from materials with porous structure providing facilities
for water transport to electrodes and elimination of gases from the cell. At present,
porous titanium materials are used as collectors and separators. They are character-
ized with good electric and thermal conductivity and mechanical stability. Titanium
materials are resistive against corrosion in acidic media; however, they corrode in O2,
so anode current collectors and separators should be protected by anticorrosion
layers. The heart of PEM electrolyzer is a membrane with catalyst layers. Membranes
324 14 Green fuel: hydrogen production

that are commonly used in PEM electrolyzers are perfluorosulfonic polymers such as
Nafion (CH2)18–28SO3H, referred to as Nafion 115, 117, 212.

CF2 CF2
CF2 x CF y
O
O
O CF2
S H2O
FC CF2
z O
HO
CF2

Fig. 14.3: Scheme of Nafion. See: https://en.wikipedia.org/wiki/Nafion.

The membranes are made from perfluorinated and sulfonated polyalkene chains.
Their sulfonic groups dissociate in the presence of water causing high proton conduc-
tivity (0.1 ± 0.02 S/cm). Nafion membranes work as solid polymer electrolytes perme-
able only for proton ions. They are characterized with good mechanical stability and
low resistance, and can work at higher current densities (2 A/cm). After the dissocia-
tion of membrane sulfonic groups, the concentration of proton is so high that it is
equivalent to approximately 0.5–1 M H2SO4 solution. Hence, the materials used as
electrocatalysts of electrode reactions (HER and OER) are limited to a relatively small
group. Typically Pt, Pt-black, and Pt dispersed on black carbon support (Vulcan) have
been used as cathode electrocatalysts, having good stability in acidic media and show-
ing good electrocatalytic properties to HER. Pt-based catalysts are very expensive;
therefore, the investigations are carried out in several directions such as (i) reduction
of the Pt loading in the support materials, (ii) increase in the surface of catalyst par-
ticles by using highly dispersed carbon materials and carbon nanostructures (graphite
nanofibers and carbon nanotubes doped with N, P, or S) as catalyst support, and (iii)
investigations of materials being alternative to Pt. Materials such as NiC, Mo2C, Ni2P,
WO2, and Mo2S on different carbon substrates are extensively studied.
Metal oxides, mainly IrO2, RuO2, or their mixture, are used as electrocatalysts for
OER at the anode. Among these oxides, RuO2 has the highest activity for OER. However,
in oxygen, ruthenium oxide is not stable and undergoes further oxidation to RuO4. IrO2
is stable, but its activity toward OER is lower; therefore, mixed oxides of Ir and Ru are
used. To minimize the catalyst cost, the research is focused on decreasing Ir and Ru
amount by using transition metal oxides such as TiO2, Nb2O5, and Ta2O5.
The technology of PEM electrolysis is partially established, and PEM electrolyzers
are commercially used in small factories, but still plenty of work should be done to
minimize the cost of electrocatalysts and hydrogen production. PEM electrolyzers op-
erate typically in the range of temperature and pressure from 50 to 80 °C and from 20
to 50 bars, in the potential range of 1.6–1.8 V and current density of 1.0–2.0 A/cm2. The
energy efficiency of PEM electrolyzer is 80–90%. The obtained hydrogen has a purity
 14.2 Water electrolysis 325

of 99.99%. Much more about PEM electrolysis and catalysts applied can be found in [2,
9, 10] and references therein.

14.2.2.3 Solid oxide electrolysis (SOE)

Solid oxide electrolysis (SOE) of water is the high-temperature process carried out typ-
ically in the range between 700 and 1,000 °C, at a pressure of up to 15 bars, and cur-
rent density ranging from 0.3 to 1.0 A/cm2. The important part of the device is the
electrolyte–SO ion conductor. In such electrolyte, the current flow occurs by the
movement of thermally activated hopping of oxide ions through the crystal lattice di-
rected by the electric field. As the movement of oxide ions is thermally activated, the
conductivity of SO is highly temperature-dependent. Much more about solid electrolytes,
their structure, properties, and applications can be found in Section 1.3.3, and in [11].
The principles of operating SO electrolyzers are shown in Fig. 14.4.

C A

H2 H2 O2 O2
H2 O2

gas diffusion
O2–
layer

H2O
H2O

Cathodic Anodic
catalyst layer solid catalyst layer
oxide
electrolyte

Fig. 14.4: Schematic illustration of solid oxide electrolyzers. A, anode; C, cathode.

The water steam is pumped in a porous cathode, where under an electric field it dif-
fuses to the electrode surface. At the cathode interface, the water is reduced to oxygen
ions (O2–) and molecular hydrogen (H2). Oxygen ions migrate across the SO electrolyte
to the anode, where they are oxidized to molecular oxygen (O2). Both gases diffuse
back through the cathode and anode, respectively, and are collected. The reactions
that take place during electrolysis are as follows:

Cathode: H2 O + 2e ! H2 + O2−

Anode: O2 − ! 1=2 O2 + 2e
326 14 Green fuel: hydrogen production

The most common material used as an electrolyte in SO electrolyzer is an oxygen ion

O2– conductor ZrO2 doped with 8 mol% of Y2O3 known as YSZ. YSZ is chemically stable
up to 1,000 °C and has a high ionic conductivity of about 0.02 S/cm at 900 °C. Other ma-
terials were also tested as electrolytes, such as scandium-stabilized zirconia (ScSZ),
ceria oxides doped with Gd2O3 or Sm2O3, and LaGaO3 doped with Sr and Mg. In recent
years, some of the ceramic proton conducting materials have been investigated and de-
veloped, such as yttrium-doped barium cerate, yttrium-doped barium zirconate, and yt-
trium-doped strontium zirconate. For the cathode, the commonly used material is
porous cermet composed of Ni and YSZ, but also the perovskite-type lanthanum stron-
tium manganese (LSM) and LSM doped with scandium were tested. For the anode, the
composite containing LSM and YSZ is applied. The predicted electrical efficiency of hy-
drogen production in SOE is the highest among other water electrolyzers (90–100%),
and also the purity of generated hydrogen is high. However, the main problem that
should be solved before the technology is commercially available relates to the degrada-
tion of Ni/YSZ cathodes decreasing the lifetime of SO electrolyzers [2, 9, 12].
All these briefly described technologies have some advantages and disadvantages.
For instance, the advantages of AWE are well-established technology, low-cost cata-
lysts, and high durability. As disadvantages, we can consider the low current densities
and low purity of gases which need further purification. The advantages of PME cells
are compact system design, high current densities, and high purity of produced gases,
but their technology is not fully established, the cost of catalysts is very high, and du-
rability is low. SOE cells operate with low-cost catalysts, and their efficiency is the
best in comparison with other water electrolyzers, but durability is low, design is
large, and they are mostly at the laboratory stage.

14.2.2.4 Seawater electrolysis

In the water electrolyzers, briefly described above, fresh water is applied and enriched
with proper electrolytes. However, such water constitutes about 1% of Earth’s water
and is unevenly distributed. Fresh water is a scarce resource in many parts of the
Earth, and about 80% of the human population faces a risk of water shortage. On the
other hand, about 71% of the Earth is covered by oceans and seas, so seawater is the
richest source of water. Plenty of efforts are made to develop the technology for seawa-
ter electrolysis but researches are still at the laboratory level. Unfortunately, seawaters
are salty waters, containing plenty of ions such as Na+, K+, Ca2+, Mg2+, Cl–, Br–, and SO42–.
Moreover, they contain not only different salts but also microorganisms and pollution,
the products of human activity. Therefore, to remove sediments, plastics, and microor-
ganisms, the seawater needs initial purification. After that, two different routes can be
considered: (i) two-step indirect electrolysis with desalination of seawater by using re-
verse osmosis, and application of such purified water in commercially working electro-
lyzers (mainly AWE); and (ii) one-step electrolysis carried out by applying seawater
directly to electrolytic cells, without initial desalination (direct seawater electrolysis). In
 14.2 Water electrolysis 327

electrolyzers filled with fresh water, H2 and O2 are evolved at the cathode and anode,
respectively. If seawater is applied, the H2 is still evolved at the cathode, but at the
anode, instead of O2, Cl2 is formed:

Cathode 2H2 O ðlÞ + 2e ! H2 ðgÞ + 2OH−ðaqÞ, E0 = − 0.828 V vs. SHE

Anode 2Cl−ðaqÞ ! Cl2 ðgÞ + 2e, E0 = 1.358 V vs. SHE

Overall reaction: 2H2 O ðlÞ + 2Cl−ðaqÞ ! H2 ðgÞ + Cl2 ðgÞ + 2OH−ðaqÞ

ΔE0 = EA0 − EC0 = 2.186 V ∼ 2.19 V

ΔE0 = 2.19 V means the standard minimal potential required for seawater splitting and
ΔEeq is the equilibrium potential of the cell which depends on partial pressures of Cl2
and H2, concentration of Cl–, and pH of the solution.
The overall reaction is very well known in the chloralkali industry, based on the
electrolysis of brine (a concentrated solution of sodium chloride), and is applied for
the production of chlorine. The primary products of chloralkali processes, besides
chlorine gas, are hydrogen gas and sodium hydroxide solution (caustic soda). The con-
centration of sodium chloride in such type of electrolysis is about 7 M. Because of the
very high corrosive properties of chlorine toward metals, some other materials for
the anode are applied. For many years, graphite electrodes were applied in industry;
at present, titanium–ruthenium oxide is commonly used. The applied potential for
brine electrolysis is in the range of 2.9–4 V depending on the type of the electrolyzer.
You can find much more on brine electrolysis in the Electrochemistry Encyclopedia
online. It is beyond the scope of this book.
The concentration of chloride ions in seawater is much lower than in brine
(about 0.5 M), also pH is near neutral, so electrolysis of seawater can be carried out in
more friendly conditions, but still chlorine evolved at the anode is aggressive and
should be eliminated. Therefore, the main task for researchers is to carry out seawa-
ter electrolysis in such a way that at the anode instead of Cl2, the O2 is evolved, as
during electrolysis of fresh water.
Let us look more closely at the problems of anodic reactions. Let us suppose that
both reactions, Cl2 and O2 evolutions, are competitive and depend on the electrolysis
conditions. If we compare the standard potentials of the below reactions, we find out
that they do not differ too much; however, only the potential of O2 formation is pH-
dependent:

2 H2 O ðlÞ ! O2 ðgÞ + 4H+ ðaqÞ + 4e , E0 = 1.229 V vs. SHE

2 Cl−ðaqÞ ! Cl2 ðgÞ + 2e, E0 = 1.358 V vs. SHE

Evolved chlorine is not stable in water solution and undergoes disproportionation.

This reaction creates hypochlorous acid (HClO) in acidic solutions and hypochlorite
ions (ClO–) in alkaline media (basic solutions):
328 14 Green fuel: hydrogen production

 2
Cl2 ðgÞ + H2 O ðlÞ $ HClO + H+ + Cl−, Kac = 4.2 × 10 − 4 mol2 = dm3
 
Cl2 ðgÞ + 2OH−ðaqÞ $ OCl− + H2 O ðlÞ + Cl−ðaqÞ, Kb = 7.5 × 1015 mol= dm3

Kac and Kb are equilibrium constants in acidic and basic solutions, respectively.
Hypochlorous acid and hypochlorite ions can also be generated at the anode as a
result of the side reaction involving Cl– ions:

Cl− ðaqÞ + 2H2 O ðlÞ ! HClO ðaqÞ + H + ðaqÞ + 2e, E0 = 1.484 V vs. SHE, aH+ = 1

Cl− + 2OH− ! ClO− + H2 O + 2e, E0 = 0.89 vs. SHE , aOH− = 1

In solutions with pH > 7.5, the equilibrium between HClO/ClO– is fully shifted to ClO–
ion formation. These ions can also be oxidized at the anode to ClO2– ions (chlorite ions)
and further to ClO3– ions (chlorate ions) but as the amount of ClO– ions is small in com-
parison with Cl– ions, reactions that involve chloride ions are dominant.
Comparing values of potentials of O2 evolution with values of potentials obtained
for side reactions at different pH, we can find that in all pH range the potential of O2
evolution is lower (see Pourbaix diagram, Fig. 3.5). This means that the OER is more
favorable from the thermodynamic point of view than chlorine or hypochlorite ion
formation. The factor that inhibits the evolution of O2 is the high activation overpotential
of this reaction described in Section 14.2.1. To illustrate the differences between both
evolution reactions (Cl2 and O2), we can compare their exchange currents. The ratio of
exchange current densities for chlorine and oxygen evolution on most anodes tested
during seawater electrolysis is in the range of 103–107 [3]. It showed differences in the
kinetics of both reactions and pointed out how easy chlorine evolution is. Therefore, the
development of suitable OER electrocatalysts is crucial for the industrial application of
direct seawater electrolysis. These electrocatalysts should not only provide high selectiv-
ity for OER but also should exhibit anticorrosive properties against aggressive chlorine
and chloride ions. More about seawater electrolysis can be found in [2, 13].
The excess availability of seawater and wastewater, as well as the reduction in en-
ergy resources, has initiated a new thought of green process “waste to energy.” Apart
from the classic green technology like solar and wind systems, the utilization of seawa-
ter and wastewater as a source of energy depends on the conversion of the chemical
energy trapped in waste to green energy.
For this purpose, a new bioelectrochemical system has been identified for sus-
tainable recovery/production of valuable resources such as hydrogen and oxygen that
utilizes wastewater and seawater electrolysis. Here, we will focus on a general de-
scription of such a system involved in the production of hydrogen on the cathode and
a model reaction for microbial anode.
 14.2 Water electrolysis 329

14.2.2.5 Microbial electrolysis (MEC)

At present, a new method of water electrolysis called MEC is being investigated and
developed. This method is related to the microbial fuel cell (MFC, see Section 9.4)
where the chemical energy is converted into electrical energy as a result of microor-
ganism action. On the contrary, in microbial electrolysis cells (MECs), electrochemi-
cally active microorganisms (bacteria and microbes) participate in the production of
chemical energy by oxidizing the organic compounds at the anode, producing CO2,
protons, and electrons. At the cathode, protons are reduced and H2 is evolved. In mi-
crobial cells, hydrogen can be produced from pure organic compounds such as acetic
acid, lactic acid, and glucose as well as from biomass and wastewater. Moreover, the
demand for electrical energy is much lower than for other electrolytic cells, as a part
of energy is derived by the microorganism activities. Such cells can operate in the po-
tential range from 0.25 to 0.8 V – much lower, compared with typical water electro-
lytic cells operating at potentials higher than 1.6 V. The efficiency of H2 production
depends on the kind of organic substances used. In the case of lactic or acetic acids,
the efficiency achieved is 87%; for not pretreated cellulose, or glucose it is about 63%.
A microbial electrolytic cell is a one- or two-compartment cell. In a two-compartment
cell (Fig. 14.5), the cathodic and anodic compartments are separated by proton-conducting
membranes.

BA - bioanode

C H2 CO2, H+ BA

H+

H+

organics

Fig. 14.5: Schematic illustration of microbial cell.

PME BA, bioanode; C, cathode.

The membrane is an important part of a cell. It is not only the conductor for proton
ions released at the anode but also the membrane reduces the crossover of organic
compounds from anode to cathode, improving purity of the produced hydrogen and
preventing microbial consumption of hydrogen. The most common membrane used in
MECs is the proton-exchange membrane, but other membranes have also been tested
such as the AEMs, and the bipolar membranes. Much more information about mem-
branes and membrane potentials can be found in Section 1.4. The cathodic and anodic
parts of the cell contain electrodes: conventional cathode and microbial anode. Anode
330 14 Green fuel: hydrogen production

is placed in the electrolyte containing microorganisms and the medium proper for their
existence and growth. Some microorganisms can spontaneously adsorb on an anode
forming electroactive biofilms, and catalyzing the oxidation of organic compounds. In
this case, the anode works as a support for biofilm formation and electron collector.
The oxidation of acetic acid is considered a model reaction for microbial anode:

C2 H4 O2 + 2 H2 O ! 2CO2 + 8H+ + 8e, E0 = 0.071 V ðvs. SHEÞ

The equilibrium potential of the anode depends on pH:

2.3 RT 2.3RT p2CO2

EA = E0 − pH + log (14:10a)
F 8F aC2 H4 O2

2.3 RT
if Cacid − 1 M, pCO2 − 1 bar, aH2 O − 1, formal potential EA f = E0 − pH (14:10b)
F
at pH = 7, it is equal to –0.291 V (vs. SHE).

Electrons released in this reaction participate in water reduction at the cathode and
in the production of hydrogen:

8 H+ + 8e ! 4H2
2.3 RT RT
EC = E0 − pH − log p4H2 , E0 = 0 V ðvs. SHEÞ (14:11a)
F 8F
2.3 RT
Formal potential EC f = − pH (14:11b)
F
and at pH 7, it is equal to –0.413 V:

Overall reaction: C2 H4 O2 + 2H2 O ! 2CO2 + 4H2

At pH 7, the equilibrium potential ΔE = EAf – ECf of MEC containing acetic acid as sub-
strate is equal to 0.122 V. This is the minimal thermodynamic potential required for
hydrogen generation in MECs in such conditions. Comparing this value with the mini-
mal thermodynamic value 1.229 V, required for conventional water electrolysis, we
see the benefits and attraction of MEC. In practice, the potential that has to be used in
MECs is higher (about 0.25 V and more up to 0.8 V), but still much lower than in the
typical water electrolytic cells. In the case of conventional cells, the values of applied
potentials are dependent on the energy losses. They result not only from activation
overpotentials of reaction but also from ohmic losses. Due to the potential applica-
tions of MECs in “green chemistry,” polluted electrolytes such as sewage media or
wastewater are tested as a fuel. Their ionic conductivity is low, in the order of 0.002 S/
cm; therefore, some salts have to be added to improve the conductivity. However, in
microbial cells, the important anodic reactions are enzyme-catalyzed reactions, very
sensitive to pH, salt concentrations, and temperature. Therefore, the microbial anode
 14.3 Solar energy in photo water splitting 331

can operate efficiently in the limited temperature range of 25–40 °C, and in neutral,
not too concentrated solutions.
Plenty of materials were tested as anodes or biofilm supports in MECs, such as
stainless steel, carbon papers, graphite brushes, and graphite granules. As cathodes,
stainless steel, nickel, and cobalt, and some alloys were applied. It seems that stainless
steel after special treatment is a good material for both electrodes. The alternative to
the abiotic cathode is the application of a biocathode. Microorganisms that are in-
volved in the formation of such bioelectrode contain the enzyme hydrogenase, cata-
lyzing the reaction 2H+ + 2e– ! H2.
Microbial cells, like other water electrolytic cells, have some advantages such as
(i) low operating potential, usually in the range of 0.25–0.8 V; (ii) low cost of electrode
materials, since special treatment of anode is not required as biofilm is formed spon-
taneously; and (iii) application of polluted media as a source of hydrogen. The main
disadvantages of MECs are (i) low hydrogen production rate, (ii) low purity of hydro-
gen, and (iii) microbial anode is very sensitive to pH, T, and concentration of sub-
strates. MECs are in the phase of investigations, and many efforts are ahead as their
operations depend upon the type of microorganisms and organic compound applied.
Much more about MEC can be found in [14, 15] and references therein.

14.3 Solar energy in photo water splitting

Hydrogen is the fuel of the future. The best way to produce hydrogen without pollu-
tion is the electrolysis of water. However, if fossil fuels are used as a source of electri-
cal energy, the problem of CO2 emission and other environmental pollution remains
unsolved. Therefore, many efforts are made for the application of renewable energy
sources. The Sun is the cleanest and inexhaustible energy source and its energy pro-
duces heat and can be converted to electrical energy.
Two approaches to the application of solar energy can be distinguished in water
splitting. The first can be realized by the installation of complex devices containing
PV (photovoltaic) panels and water electrolysis units. PV panels can convert solar en-
ergy into electrical one and this energy can be further used for feeding electrolyzers.
The other way is the direct application of solar energy for the photoelectrolysis of
water by using photoactive materials – semiconductors (SCs). In the future, photo
one-step technology of water splitting may be economically more profitable than PV
production, followed by water electrolysis. The advantages lay in the application of a
single plant and the mitigation of efficiency losses inevitable in the case of two plants
that are required in a two-step technology. There is a huge amount of research car-
ried out on the improvement of efficiency of water photoelectrolysis and hydrogen
production; however, we are still far away from commercial application of this tech-
nology, and it is song of future. The current approaches are focused on the evaluation
332 14 Green fuel: hydrogen production

of new light-sensitive electrode materials and multijunction configurations of the

photocells.
Two terms should be distinguished for a better understanding of this section: pho-
toelectrochemical electrolysis of water and photocatalytic water splitting. In both
methods, solar energy can be applied. The term “photoelectrochemical electrolysis” is
correctly used when photoelectrochemical reactions take place in electrochemical
cells at SC electrode/electrodes (SCE). The form of SC photoelectrodes may be differ-
ent. The solid plates, thin films, and SC particles (SCPs) distributed on different sub-
strates are applied. The photocatalytic water splitting occurs on suspensions of SCPs.
There are no essential differences between both methods, in the description of the
photo mechanism and the applied SC materials. The crucial difference between SCEs
and SCPs is the place of reactions. In a photoelectrochemical cell (PEC), the oxidation
of water occurs at the n-SC anode (O2 evolution), while the evolution of H2 takes place
at the cathode (Pt, or p-SC). Hence, the separation of O2 and H2 is much easier than in
the case of direct photocatalytic splitting of water, where both reactions take place at
the same particles. The process of separation increases the cost of production of pho-
tocatalytic H2; therefore, some approaches described further are developed. The re-
search of photoelectrolysis of water is carried out in the typical PECs, while for
photocatalytic water splitting, chemical reactors are used. The SCPs where photoreac-
tions occur are often regarded as short-circuited microelectrochemical cells with a
cathode and anode placed together. It is worth mentioning that hydrogen may be ob-
tained by the half-reaction of water splitting in the presence of sacrificial agents – do-
nors. Then, oxygen is not evolved at the SCEs or SCPs. Such reactions cannot be
named “water splitting.” The term is reserved strictly for situation when water split-
ting and photo water splitting of H2 and O2 is produced in stoichiometric amount. In
other cases, we suggest using the term photo-sacrificial water splitting or photo water
splitting in the presence of a sacrificial agent.
Part III of this book was devoted to the electrochemistry and photoelectrochemis-
try of SC materials. In Chapter 10, we described the behavior of the SC/electrolyte in-
terface in dark and under illumination. In Chapter 11, we discussed the types of
photocells, their operation, and the application of SC materials. Special attention was
devoted to the photosynthetic cell and photoelectrolysis of water, to the energetic dia-
grams of direct photoelectrolysis and assisted water photoelectrolysis. Some brief,
basic information about SCPs, their photoreactions, and their applications can be
found in Chapter 12. We encourage all readers interested in a deeper understanding
of photoelectrochemical and photocatalytic processes, and their applications in the
resolution of environmental problems to read carefully these chapters. Below we con-
centrate on comparison between photoelectrochemical and photocatalytic water split-
ting, looking for some similarities and differences and on the new developments and
trends.
 14.3 Solar energy in photo water splitting 333

14.3.1 Thermodynamics, energy losses, and efficiency

The overall reaction of water photo splitting may be expressed, independently from
the reaction location (SCE and SCP), in the form:

4 hν + 2 H2 O ðlÞ ! 2H2 ðgÞ + O2 ðgÞ

where h is Planck’s constant and ν is the frequency.

This reaction takes place if the energy of photons Eph is equal to or larger than

Δ r G0
Eph = hν ≥ (14:12)
2NA
Taking into account the values of Δr G0 – the standard Gibbs free energy of reaction
equal to 237.13 kJ/mol, and NA – Avogadro’s number equal to 6.022 × 1023 particles/mol,
1 eV = 1.602 × 10−19 J, we obtained the minimal energy of photons required for photo
water splitting equal to 1.229 eV. This value is equivalent to the value of ΔE0 = 1.229 V,
the standard minimal potential required for water splitting when electrical energy is
used. As in the case of water splitting, also in photo water splitting, there are losses of
energy, and some of them are common and some are different as SC materials and
photons are applied. Taking into account all losses, we can roughly estimate the opti-
mal value of bandgap energy of SC, and optimal photon energy as 2.0–2.4 eV. This
range corresponds to the wavelength of light 620–517 nm, respectively, and lies in the
visible range of the solar spectrum (see Fig. 11.2). Before going further, we should
briefly consider the energy losses during photo water splitting at SCEs and SCPs. We
can distinguish three main steps in photo water splitting.

A. The first step that takes place during illumination (irradiation) of SC materials
(SCEs and SCPs) is the absorption of photons. Only photons with energy Eph = hν ≥ Eg
(Eg – bandgap energy of SC) are absorbed and lead to a generation of excitons: bonded
pairs electron–hole, electrons (e) at the bottom of conduction band (CB), and holes (h+)
at the top of the valence band (VB) of SC:

hν + SCE ðor SCPÞ ! e ðCBÞ + h+ ðVBÞ

Note the differences between the direct and indirect SC bandgap energy (Section 10.1.1).
In the first case, photons with energy hν = Eg are absorbed; in the second case, photons
should have energy higher than Eg. In both cases, the excess of energy is dissipated as
thermal energy. What about photons with energy lower than Eg? The SC materials are
transparent to them and their energy is lost. The other factor, decreasing available
solar energy, is the reflectivity of applied materials. All these energy losses connected
with sunlight reflection and threshold absorption are named optical losses, and they
determine the threshold efficiency of solar energy conversion.
The efficiency of solar energy conversion is substantially decreased by recombina-
tion processes. The generation of electron–hole pairs during illumination perturbs the
334 14 Green fuel: hydrogen production

equilibrium of the charge carriers’ concentration. The processes of recombination re-

store the equilibrium. The excited electrons in CB might fall back to VB and then recom-
bine with holes. It is a so-called direct recombination with a one-step transition.
Sometimes, the recombination occurs via impurities, defects acting as the bulk, or sur-
face traps. Recombination annihilates the photogenerated electrons and holes, so they
cannot participate in any redox processes. In consequence, the recombination decreases
the efficiency of reactions that occur at SCEs and SCPs, and the efficiency of solar energy
conversion. Therefore, the excited electron–hole pairs should be effectively separated
before annihilation. Much more about the processes of direct and indirect recombina-
tion, surface recombination, and trapping can be found in Sections 10.1.2 and 10.5.

B. The second step is the separation of excited pairs e–h+ and the motion of free
charge carriers to the reaction sites. There are differences in separation processes oc-
curring in solid SCEs and SCPs. In both cases, separation is caused by an electric field,
but the origin of this field is different. Figure 14.6(a, b) illustrates the differences and
the pathways of recombination.

a) b)
hν
e e e e
CB CB H+ /H2
3 3
O2 /H2O
5 2 1
2 1 O2 /H2O
4
4 VB
VB h+ h+ h+
h+
Ldiff LSC
LD
n-type SC SCP

Fig. 14.6: Schematic illustration of generation and recombination of charge carriers (electrons and holes)
in (a) n-type of SC and (b) SC particles. LSC, thickness of space charge layer; Ldiff, diffusion length; LD = LSC + Ldiff.
Arrows 1, 2 – generation, direct recombination; 3,4 – bulk recombination; 5 – surface recombination.

In the beginning, let us consider the formation of an electric field in a SCE. The con-
centration of electrons in SC materials is much lower than in metals; therefore, the
compensation of the charge of electrolyte ions takes place in the thin layer of the SC.
This charge region of SC is named the space charge layer, and its existence is crucial
for SCE behavior in dark and under illumination. Much more about the space charge
layer and its significance in electrochemistry and photoelectrochemistry of SCs can be
found in Chapters 10 and 11. This space charge layer is characterized by the charge
QSC, potential drop φSC , and thickness LSC, and can exist in three states: depletion, in-
version, and accumulation (see Fig. 10.7). The best charge separation is obtainable
when the space charge layer exists in a depletion state. In this state, in n-type SC, the
Qsc is positive, created by ionized donor atoms. On the contrary, in p-type SC, the QSC
 14.3 Solar energy in photo water splitting 335

is created by ionized acceptor atoms and is negative. In Fig. 14.6(a), the energy dia-
gram for n-type SC in a depletion state is shown. There, besides the space charge layer
thickness LSC, the diffusion length Ldiff is marked. Note that the sum of LSC and Ldiff
means the Debye’s length LD. What processes occur during illumination depends on
the depth of photon penetration characterized by the reciprocal of the absorption co-
efficient (x = α − 1 ). If photons are absorbed at x > LD, the photogenerated e–h+ pairs re-
combine (arrows 1 and 2 marked generation and recombination processes). If photons
are absorbed in the vicinity of LSC, the photogenerated e–h+ pairs dissociate in the
electric field and free minority carriers – holes migrate to the n-SC/electrolyte inter-
face where they can participate in the oxidation reaction, or are trapped by surface
states (arrow 5). The e–h+ pairs photogenerated in the range of Ldiff diffuse to the
space charge region, where they are separated by the electric field. They can also be
trapped by defects of the crystal lattice (deep energy levels, arrows 3 and 4) and re-
combine. The free majority carriers – electrons, obtained during dissociation, diffuse
in the bulk of SC to the back contact or to the other electrode where they participate
in reduction. They can also be trapped and recombined on their way.
A different situation exists when SCPs are illuminated, and then separation starts
to depend on the particle size. The large particles (dimension higher than LD) behave
similarly to the bulk SCs. The separation of electrons and holes occurs in the electric
field of space charge layers, and free charge carriers migrate to the interface of SCP/
electrolyte taking part in redox reactions. The situation is different when small par-
ticles are illuminated (Fig. 14.6(b)). Then photogenerated e–h+ pairs diffuse to the SCP/
electrolyte interface, where they dissociate in the electric field of an electric double
layer (Section 1.1.1), and there they participate in redox reactions. They can also be
trapped by surface states created by particle inhomogeneity or species adsorbed from
solution. Earlier, other processes can also occur in SCPs such as recombination and
bulk trapping. All these processes have different timescales. It is worth mentioning
that trapping processes at particle surface can also play a positive function, as they
compete with surface recombination of charges. Trapped electrons or holes partici-
pate in a redox reaction at the particle surface after release. Size effect, other pro-
cesses, and their timescale are briefly described in Section 12.2. In the case of SCPs,
the main problem is the improvement of the separation of photogenerated e–h+ pairs
and the decrease of recombination processes.

C. In the third step, the photogenerated, separated free charge carriers (electrons and
holes) participate in the redox reaction with substrates in the solution. The evolution of
O2 occurs at the n-SC photoanode, while H2 is generated at the cathode (Pt or p-SC). In the
SCP case, both H2 and O2 are evolved at the same particles. The half-reactions are as
follows:
Anodic side: 2 H2 O ðlÞ + 4h+ ðSCÞ ! 4 H+ ðaqÞ + O2 ðgÞ

Cathodic side:: 4H+ ðaqÞ + 4e ðSCÞ ! 2H2 ðgÞ

336 14 Green fuel: hydrogen production

These reactions will take place only when the relation between energy levels of CB
(EC) and VB (EV), and energy levels of reactions H+/H2 and O2/H2O will be appropriate.
Figure 14.7(a, b) presents the simplified energy diagrams for n-SCE photoanode and
SCPs, where proper positions of energy levels for SC materials and redox systems (H+/H2
and O2/H2O) are marked.

a) (V) (eV) b)

e e e –1 –3.5 e e
EC H+/H2 –0.41 EC

0 –4.5
hν O2/H2O hν
+0.82
1 –5.5
EV EV
h + +
h h + h+ h+
2 –6.5
vs.SHE
n-type SC SCP

Fig. 14.7: Schematic energy diagram showing requirements for direct photo water splitting: (a) n-type SC
electrode and (b) SC particles. pH is 7.

In electrochemistry, the potential scale is mainly applied, so two scales are used in this
figure. The correlation of both scales is simple and is based on the value of the Fermi
energy level (electrochemical potential of electrons) for the SHE electrode equal to −4.5
± 0.01 eV. It means that 0 V vs. SHE is equivalent to −4.5 eV. The photogenerated elec-
trons and holes should have sufficient energy to reduce protons and to oxidize water.
Generally, the energy of the CB edge (Ec) for n-SCE and SCP has to be higher than the
energy level of the H+/H2 redox couple, that is, it is located at a more negative potential
relative to the H+/H2 redox potential in the solution. In the case of water oxidation, the
energy of the VB edge (EV) should be lower than the energy level of the O2/H2O redox
couple, that is, it is located at more positive potential than the O2/H2O redox potential in
solution. In the energy diagram of the SC electrode (Fig. 14.7(a)), the energy bands are
flat. It means that the charge of the space charge layer, QSC, was compensated by an
external source, for instance, the external potential applied to the SC electrode. This po-
tential at which the energy bands are flat is referred to as the flat band potential, Efb,
and has to be more negative than the potential of the H+/H2 redox couple. Note that the
redox potential of H+/H2 and O2/H2O couples depends on the pH of the solution; there-
fore, the position of their energy levels may be changed. In Fig. 14.7(a, b), they are
marked at values corresponding to the redox potential of H+/H2 and O2/H2O couples at
pH 7. If we applied in PEC the p-type SC photocathode, the photogenerated minority
charge carriers, electrons, participate in the reduction of hydrogen ions. Also then, the
flat band potential has to be more negative than the redox potential of the H+/H2 couple.
If these conditions are not fulfilled for n-SC and p-SC electrodes, we have to use the ad-
 14.3 Solar energy in photo water splitting 337

ditional external potential (bias voltage) to overcome the energy demand for photo
water splitting. This type of photoelectrolysis is referred to as assisted water photoelec-
trolysis, contrary to direct photoelectrolysis. The detailed requirements, which have to
be fulfilled by the photoanodes, photocathodes, and more advanced schematic energy
diagrams for water photo splitting, can be found in Section 11.3.4. Therein, on the energy
scheme, the anodic and cathodic overpotentials (ηA and ηC) are also marked (Fig. 11.5). As
in the case of water splitting, they cause energy losses influencing the efficiency of solar
energy conversion. Their formulas differ from those applied for the reaction at metal
electrodes (see Section 10.5). The energy diagram for SC particles is shown in Fig. 14.7b.
The requirements for the relation between EC, EV, and redox potentials of H+/H2 and
O2/H2O are the same as for SC electrodes.
There are many definitions and equations describing the efficiency of solar en-
ergy conversion. They depend on whether the solar energy is converted into electrical
energy (regenerative photocells), or is stored as chemical energy in a fuel (photosyn-
thetic cells). Some of these definitions are common, as the basic processes in SC mate-
rials are the same. We will concentrate here on the efficiency of conversion of solar
energy during photoelectrolysis of water and on the efficiency of photocatalytic water
splitting. Some general information about solar energy, conversion, and efficiency of
regenerative photocells can be found in Sections 11.1 and 11.2.
Generally, the efficiency of solar energy conversion in photodevices can be deter-
mined as the ratio of power output (Pout) of any devices and systems, and power
input (Pin) of solar irradiation ε = Pout =Pin . The solar-to-hydrogen (STH) conversion ef-
ficiency is described for direct water photoelectrolysis by equation (14.13a). Sometimes,
instead of Gibbs free energy of reaction, the enthalpy of reaction is used (14.13b). In this
case, it is assumed that the hydrogen will be burnt and the stored energy of this reac-
tion will be recovered as heat:

Δr G0 R
εSTH = (14:13a)
Pin A
Δr H 0 R
εSTH = (14:13b)
Pin A
ΔrG0 is the standard Gibbs free energy of water splitting (kJ/mol), ΔrH0 is the standard
enthalpy of water splitting (kJ/mol), R is the rate of hydrogen generation (mol/s), Pin is
solar power (W/m2), A is the illuminated (irradiated) area (m2).
If we determined the solar conversion efficiency to hydrogen during the assisted
water photoelectrolysis, we should also take into account the additional electrical energy,
which is applied to fulfill energy demand for photo water splitting equal to UI (see
Fig. 11.5). In this case, equations (14.13) are modified. In these equations, the standard po-
tential ΔE0 = 1.229 V ≈ 1.23 V or the standard thermoneutral potential ΔEtn 0
= 1.481 V ≈
1.48 V of water splitting are used:
338 14 Green fuel: hydrogen production

Δr G0 R − UI
εSTH = (14:14a)
Pin A
ð1.23 − U ÞI
εSTH = (14:14b)
Pin A
Δr H 0 R − UI
εSTH = (14:15a)
Pin A
ð1.48 − U ÞI
εSTH = (14:15b)
Pin A

where U is the applied potential and I is the current flowing through the cell.
The efficiency of solar energy conversion is decreased by the processes described
above. Their influence is taken into account in the equation:

ε = εthresh ∅ εch
stor εohm (14:16)

(i) εthresh – the threshold efficiency pointing out that only part of a photon is ab-
sorbed by SCE or SCP (see A above, Section 14.3.1)
ð∞
Eg N ðEÞð1 − RÞdE
Eg
εthresh = ð∞ (14:17)
EN ðEÞdE
0

N(E) is the number of photons striking the SC surface in unit time and R is the
reflection coefficient
(ii) ∅ – the quantum yield is defined as a ratio of effective photon incidents Neff ,
leading to the generation of photo e–h+ pairs, to the total number of absorbed
photons:
Neff
∅= (14:18)
Ntotal

∅ is connected with the efficiency of separation of photogenerated electron–hole

ph
pairs (see B above, Section 14.3.1) and manifests itself in the value of jsc,c short
circuit photocurrent (see Section 11.2).

If the light energy is converted into chemical energy in direct photoelectrolysis, the
efficiency of energy storage is determined by the ratio of free Gibbs energy of the
overall reaction occurring in the photocell to the energy of absorbed photons hν = Eg.
In the case of assisted photoelectrolysis, the power input (UI) should be taken into
account:
 14.3 Solar energy in photo water splitting 339

Δr G
εch
stor = (14:19a)
Eg
Δr G − UI
stor =
εch (14:19b)
Eg

The ohmic losses in the photocell are manifested by changes in the fill factor, ff (see
Section 11.2, Fig. 11.5). Note that the fill factor determines the quality of PE cells. The
ohmic losses are mostly caused by the lower conductivity of SC materials, the prob-
lems with the transport of charge carriers to the reaction place, and the transport of
electrons from photoanode to cathode where H2 is evolved. The efficiency of PE cells
may also be evaluated by IPCE, the incident photon to current conversion efficiency.
IPCE is defined as the ratio of electron numbers generated by light to the number of
incident photons:

1, 240 × photocurrent density ½A=cm2 

IPCE = × 100%
wavelength ½nm × photon flux ½W=cm2 

where 1,240 is the unit conversion coefficient.

All these equations may be used when solar energy is converted to chemical en-
ergy in a photoelectrochemical cell. To determine the efficiency of solar energy con-
version at SCPs, we may use equations (14.13a) and (14.13b). In practice, the activities
of SCPs applied in photo water splitting are usually assessed by the ratio of evolved
gases H2 and O2 (mol/h) per SCP amount (g) under specified illumination conditions.
Having the amount of H2 or O2, the apparent ∅ may be determined:

2 × number of evolved hydrogen molecules

Quantum yield ðH2 Þ = × 100%
number of incident photons
4 × number of evolved hydrogen molecules
Quantum yield ðO2 Þ = × 100%
number of incident photons

Here, numbers 2 and 4 point out that two or four photons are required to generate one
H2 and O2 molecule, respectively.
To compare the efficiencies of solar energy conversion between different PECs
and also at different SCPs, we should carry out the experiments under the same photo
conditions. It is accepted to apply the photodevices generating irradiance of power
100 mW/cm2, which also mimics the air mass 1.5 spectra (AM 1.5). The STH conversion
efficiency in PECs depends on their configuration. Theoretically predicted STH con-
version efficiencies for ideal photosystems containing single photoabsorber (S) or
dual photoabsorber (D) are about 30% and 41%, respectively. In reality, taking into
account all energy losses (including electrochemical ones), they do not exceed about
10% and 16% [16]. Note that a single photoabsorber in PEC means one SCE (photoa-
node or photocathode), while a dual photoabsorber means two n-SC and p-SC individ-
ual electrodes operating in PEC, or a special design of two SC materials coupled
340 14 Green fuel: hydrogen production

together. Such photocells are referred to as one-bandgap or two-bandgap devices. In

the literature concerning PEC and STH, you can find notations: S2 and D4. These nota-
tions mean that two or four photons are required for the generation of one H2 mole-
cule. Data concerning activities and ∅ (quantum yield) of many SCPs applied in the
photocatalytic water splitting can be found in review [17] and references therein. The
obtained values depend on the modification of particles, applied solutions, and illumi-
nation conditions, but in comparison with the efficiency of SCE, they are much lower.
To date, the efficiencies of conversion of solar energy to chemical energy are too low
for practical, commercial application of photo water splitting. At present, to improve
the efficiency of conversion, the investigations are carried out in a few directions: the
evolution of materials applied for photoelectrodes and photocatalysis, the more so-
phisticated design of the PECs, and special arrays of SCPs in photocatalytic splitting of
water. We very briefly describe these trends below.

14.3.2 Photoelectrolysis of water: trends

In photo water electrolysis, the typically studied cells are single photoelectrode and two
photoelectrode cells. Single photoelectrode cells contain n-SC photoanode and metal cath-
ode which is not light sensitive. In dual photocell, two electrodes are light sensitive, n-
type SC photoanode, and p-type SC photocathode. The “hearts” of these cells are photo-
electrodes. To be efficient in the photo water splitting, they have to fulfill several specific
requirements described above. From a practical point of view, the photoelectrodes
should be stable and the cost of their manufacturing should be low. The stability of pho-
toelectrodes depends on corrosion and photocorrosion of SC materials. The problems of
corrosion and photocorrosion of SC materials are described in Section 10.6. Let us concen-
trate firstly on the photoanodes applied in the photoelectrolysis of water. Most of them
are fabricated from metal oxides or materials with oxide structure; they are n-type SCs.
We can distinguish two main groups: the binary oxides and the perovskite-structured
oxide materials (ABO3). The first group is represented by TiO2, ZnO, WO3, Fe2O3, and
In2O3. Perovskite-structured SCs have a general formula ABO3. They are represented by
titanates (SrTiO3, BaTiO3, and FeTiO3), tantalates (NaTO3), and niobates (KNbO3). The ener-
getic diagram for selected oxides is shown in Fig. 14.8(a). The bandgap energy (Eg) and
positions of the energy band edges EC and EV are marked there in correlation to (E0) stan-
dard potential of H+/H2 O2/H2O couples in the energy and potential scale.
The family of the perovskite-structured oxides is very large, but they are mostly
applied in photocatalytic water splitting. The group of oxides studied for possible appli-
cation as photoanodes in photoelectrolysis of water is much higher. The interesting ma-
terials tested lately are vanadates BiVO4 (2.5 eV), InVO4 (2.0 eV), and oxynitrides such as
TaON (2.5 eV). Let us divide the studied oxides into two groups: one having the bandgap
energy Eg > 3 eV and the other with Eg < 3 eV. Generally, oxides with Eg > 3 eV are chemi-
cally stable even in acidic solutions and avoid photocorrosion. Unfortunately, they absorb
 14.3 Solar energy in photo water splitting 341

a) (V) (eV)

–2 –2.5
SrTio3 NaTaO3
–1 –3.5 In2O3
ZnO WO3
0 –4.5 H+/H2
Fe2O3
2.8 eV 3.2 eV 3.9 eV
1 –5.5 O2/H2O
3.2 eV 2.8 eV 2.2 eV
2 –6.5

3 –7.5

4 –8.5
vs. SHE n-type SC

Fig. 14.8: Schematic energy diagram showing correlation between the energy of conduction and valence
band edges (EC, EV) and redox potentials of H+/H2 and O2/H2O couples: (a) for n-type SC and (b) p-type SC
materials (adapted from [17, 19]).

solar energy in the UV range of the solar spectrum. The better absorption ability in the
visible range of the spectrum has oxides with Eg < 3 eV. However, they are not stable and
undergo photocorrosion under illumination. Moreover, for WO3 and hematite (Fe2O3),
the CB edges lay at too positive potentials with respect to hydrogen evolution potential.
Therefore, they cannot be used in direct photoelectrolysis of water. Others such as In2O3
are not direct SCs and absorb photons with energy higher than Eg.
In the photoelectrolysis of water, except photoanodes, the light-sensitive photo-
cathodes, p-type SCs are used. In earlier studies, III–V and II–VI semiconductors such
as GaAs (1.42 eV), GaP (2.25 eV), and CdTe (1.56 eV) were mostly tested. Great popular-
ity enjoyed the cupric oxides: CuO (1.57 eV) and CuO2 (2.47 eV). All these compounds
can absorb photons from the visible range of the solar spectrum; however, they are
unstable and undergo photocorrosion in the electrolyte solutions. In recent years, the
attention of researchers has been devoted to other materials, called ternary oxides.
They are represented by oxides with spinel structure (AB2O4) or the layered delafossite-
type oxides (ABO2). Among them are CaFe2O4, CuBi2O4, CuFe2O4, and CuRhO2 (Fig. 14.8(b)).
Note that they have the appropriate Eg values and can absorb photons in the visible
range of solar spectrum. They have also the potentials of band edges more negative than
the redox potential of the H+/H2 couple and more positive than the redox potential of
water oxidation O2/H2O except CaFe2O4. Therefore, they may simultaneously evolve H2
and O2 and can be applied in the photocatalysis of water splitting.
Plenty of studies have been carried out to improve the photon absorption effi-
ciency of SCEs possessing broad forbidden bands. Among them are: (i) modifications
of electronic structure through doping of SCs with metals (Au, Pt, and Ag), nonmetals
(N, C, S, and P), and ions (Ru4+, Pt4+, and Cr3+); (ii) nanostructuring and formation of
ordered arrays of SC (nanotubes, nanowires, nanorods, and nanosheets); (iii) modifi-
342 14 Green fuel: hydrogen production

b) (V) (eV)

–2 –2.5
CaFe2O4 CuRhO2
CuFe2O4
–1 –3.5 CuBi2O4
1.9 eV H+/H2
0 –4.5
2.1 eV 2.9 eV
1.75 eV
1 –5.5
O2/H2O

2 –6.5

3 –7.5
p-type SC
vs. SHE

Fig. 14.8 (continued )

cations of photoelectrode surface with sensitizers such as organic dyes, SC quantum

dots, or metal nanoparticles (Ag, Au, and Pt). Dye and quantum dot-sensitized solar
cells are described in Sections 11.3.2 and 11.3.3. All these approaches have advantages
and disadvantages. For instance, doping can create additional energetic levels in the
vicinity of the energy gap, which can change the photon absorption threshold, moving
it to the visible range of light. However, these additional energy levels can also facili-
tate the recombination of charge carriers decreasing their amount. Nanostructuring
of photoelectrode materials increases the active surface area of light absorption and
decreases the path of majority carriers limiting their recombination. The nanostruc-
tures also improved the light absorption by photon scattering. On the other hand, it
was shown that nanostructuring reduces the open-circuit photopotential when the
surface area of the SC/electrolyte interface increases. For in-depth information about
oxides applied as photoanodes, photocathodes, their modification, and their influence
on the efficiency of water photoelectrolysis, an interested reader is directed to the re-
views [16, 18, 19] and references therein.
Many SC materials suitable energetically for the photoelectrolysis of water are
unstable. To cope with this problem, the surface of such photoelectrodes was coated
by a thin layer of metals (nanometer thickness layers of Au, Pt, and Pd) or a thin layer
of stable semiconductor (TiO2). The compact metal films protect the SCEs against pho-
tocorrosion. Additionally, they create the photosensitive metal/SC junction, and the
photopotential of which can help to drive the reactions at the interface. From an elec-
trochemical point of view, such electrodes work as metal electrodes, and the kinetics
of reactions depends on the metal used. When the metal layers are porous, they do
not protect the photoelectrodes; however, they can catalyze the reactions. The applica-
tion of the thin films of a more stable semiconductor (SCouter), with much higher Eg
than SC substrate (SCinter), has another advantage: this bilayer photoelectrode absorbs
 14.3 Solar energy in photo water splitting 343

much more solar energy. The UV (short wavelength) part of the spectrum is absorbed
in the outer layer, while the photons with lower energy (longer waves) penetrate
across the outer layer and are absorbed in the substrate. The future of minority car-
riers generated in the interlayer depends on their energy, whether they have enough
energy to overcome the barrier at the junction SCinter/SCouter and cross the outer
layer. The studies of bilayer electrodes containing a thin layer of n-TiO2 and p-type
substrates GaP, GaAs, or Si confirmed that TiO2 effectively protects those SCs against
photocorrosion, but photoelectrochemical behavior is entirely determined by TiO2.
The new approaches in protections against photocorrosion focus on the application of
catalytically active coatings: oxides (NiOx and CoOx), nitrides (GaN and TaNx), carbides
(TiC and ZrC), for photoanodes, and the application of oxides such as Al2O3, TiO2, AZO
(Al-doped ZnO/TiO2) for photocathodes. Up to date, plenty of different coatings for
photoelectrode protection have been tested, and much more can be found in review
[20] and references therein.
The efficiency of H2 and O2 production is strongly dependent on the surface prop-
erties of photoelectrodes. The surface recombination, the low density of active sites,
and the sluggish kinetics of charge transfer decrease the efficiency of photo water
splitting. These problems can be overcome by the deposition of catalysts (cocatalysts)
on the electrode surface. They may be deposited directly on the SC surface in the
form of particles, nanoparticles, or embedded in the protective layer of the electrode,
if such is used. The typical catalysts enhancing HER are noble metals such as Pt, Pd,
Ru, and Au, as in the case of water electrolysis. The nonnoble metal catalysts NiS, CuS,
and CoP, and binary systems Ni–Mo and Ni–Co have been explored as alternatives to
them. To improve the efficiency of OERs, the oxides such as IrO2, RuO2, ZrO2. NiO2,
and SnO2 were tested. More detailed information about catalysts studied in the photo-
electrolysis of water can be found in reviews [16, 21]. All these approaches described
above improve the efficiency of STH at the individual electrodes. STH efficiency can
also be increased by the appropriate design of photoelectrolysis cells.
The configurations of the most studied and developed photocells applied in water
splitting are shown in Fig. 14.9(a–d). For many years, the investigations of photo
water splitting have been concentrated on the photoelectrolysis cells containing a sin-
gle photoanode or photocathode. Cells such as n-type photoanode PEC and p-type pho-
tocathode PEC are referred to as single bandgap devices (S2, Fig. 14.9(a)). In the last
years, the huge development of dual-bandgap devices (D4) has been observed. The
simplest dual bandgap photocells studied earlier contained two photoelectrodes: pho-
toanode and photocathode immersed in electrolytic solution independently. The other
configuration of PEC, referred to as “photochemical diode,” also contains two photo-
electrodes, and they are pressed together and connected in series by ohmic contact
(Fig. 14.9(b)). In both types of cells, named p/n-PECs, the minority carriers (h+) photo-
generated in n-SC (photoanode) and (e) photogenerated in p-SC (photocathode) partic-
ipate in the water splitting. Such dual bandgap configuration has some advantages in
comparison with S2 photocells. They absorb a higher amount of solar spectra, as the
344 14 Green fuel: hydrogen production

Eg of the photocathode is usually much lower than the Eg of the photoanode. The reac-
tion of water splitting is divided between two SC/liquid electrolyte interfaces and is
harvested by their properties. The open-circuit photopotential of the cell is higher
than the photopotential of individual photoelectrodes and, in some cases, is higher
than the potential required for direct water splitting. The efficiency of STH predicted
theoretically for p/n-PECs, including the losses associated with reactions (activation
and concentration overpotentials), is 16%.

a) b) e e
(V) (eV)
e EC hν H+ /H2
e
EC
H+ /H2
O2 /H2O
hν
hν
O2 /H2O
h+ EV p-type SC
h+ + photocathode
n-type SC
EV Ohmic
h+ phoanode
contact
n-type SC cathode n/p - PEC
(Me)
n-type SC photoanode PEC

c) d)
(V) (eV)
e
EC hν H+ /H2 H+ /H2
p-n PV
O2 /H2O
hν
O2 /H2O
EV p-n PV p-n PV
h+
h+ +
n-type SC
Ohmic cathode
photoanode
contact (Me)
n-type photoanode PV-PEC anode cathode
(Me) Ohmic contact (Me)
dual PV-PEC

Fig. 14.9: Schematic illustration of different configurations of photoelectrochemical cell (PEC): (a) n-type
photoanode PEC; (b) n/p PEC; (c) n-type photoanode PV-PEC; and (d) dual PV-PEC. PV, photovoltaic cell
(adapted from [16]).

In practice, many PEC systems containing individual photoelectrodes were tested

such as p-GaP/n-TiO2, p-CdTe/n-TiO2, and p-GaP/n-SrTiO3. They show the STH values
lower than 1%. Much better results were obtained when the photoelectrodes were
pressed together and modified. For the p/n-PEC containing p-InP/n-GaAs system with
the surface of p-InP covered with Pt and n-GaAs with MnO2 layer, the efficiency of
STH equal to 8.2% was reported. It should be mentioned that configurations applied
in PECs “two in one” are referred to as “back to back” wireless configurations and are
widely studied and applied. The back-to-back configurations are realized in the n-type
photoanode-PV–PV and p-type photocathode-PV–PV cells. In these PECs, the photoelec-
 14.3 Solar energy in photo water splitting 345

trode is connected directly with the p–n PV layer cell (Fig. 14.9(c)). The photogener-
ated minority carriers participate in reactions of water oxidation (photoanode) or re-
duction of H+ ions (photocathode), while the majority charge carriers photogenerated
in the PV cell are provided to the counter metal electrodes reducing protons at the
cathode or oxidizing water at the anode. The system containing p–n Si cell (PV) cov-
ered on the top by an n-type TiO2 layer (photoanode) is an example. In this case, only
the n-TiO2 layer is exposed to the electrolyte solution. Such a design is more effective
in the conversion of solar energy. TiO2 absorbed the photon of Eg ≥ 3.2 eV, and the
remaining part of the solar spectrum is absorbed by the p–n Si part of the system (Eg
of Si = 1.2 eV). The additional potential generated by the p–n Si cell is about 0.7 V and
may be helpful in water splitting. Other PV cells can also be used such as dye-
sensitized solar cells. The efficiency of STH obtained in such a dual system was 3.1%.
Much higher efficiency of STH, equal to 12.4%, was reported for the dual system con-
taining p-type photocathode GaInP2 coupled together with n–p GaAs PV cell.
Another interesting approach is the application of devices containing two n–p PV
cells connected together back to back. Their outer surfaces are covered with metal,
forming a cathode and anode (Fig. 14.9(d)). In this case, the splitting of water takes
place at the metal/electrolyte solution interface, such as in the direct water electroly-
sis (Section 14.2.2). The electrical energy required for water electrolysis is delivered by
SC junctions. The energetics of these junctions are not directly related to the redox
potentials of the H+/H2 and O2/H2O couples, and the relation between these potentials
and the position of EC and EV edges is not important. Such designs are referred to as
PV-PECs. They are similar to the system containing a water electrolyzer fed by PV pan-
els, where electrical energy is delivered by wires, but PV-PEC systems are wireless.
Note that ohmic contact should be formed between p–n PV junctions (Fig. 14.9(b–d)).
Efficiency of STH, equal to 16%, was reported for such PV-PEC systems containing n/p
GaInP2/GaAs covered on both sides with Pt. Much more detailed information about
the configuration of photodevices applied in water splitting and their efficiency can
be found in reviews [16, 22] and references therein.

14.3.3 Photocatalysis of water: trends

In the photocatalysis of water, powdered SC materials are applied in the form of suspen-
sions, SCPs embedded in a matrix, or spatially organized nanostructures. All these SC
materials operate as photoabsorber of solar energy catalyzing photosplitting of water.
As in photoelectrolysis, the water is split into H2 and O2. However, during photocatalysis,
H2 and O2 are not separated between electrodes and evolved simultaneously at the same
particles. The requirements that should be fulfilled by SC materials applied in these pro-
cesses are common and described in Section 14.3.1. Here also the equations for efficiency
of STH and apparent quantum yields are given. In practice, the efficiency of H2 forma-
tion is determined by the amount of H2 evolved (mol/h) in unit time (h) per unit mass (g)
346 14 Green fuel: hydrogen production

of SCPs. It was pointed out that STH conversion efficiency of 5–10% obtained in photo-
catalysis of water would render this method economically acceptable for H2 production.
However, the STH values published in articles are much lower. On the laboratory levels,
they can reach values of 1–3%, but only when photocatalysts with special designs are
used. The other factor that should be considered before the practical application of SCPs
is their selectivity to the reaction path. Selectivity is determined by equation (14.20) as
the ratio of the quantum yield of i-product formation to the quantum yield of r-reactant
decomposition:

∅i
S= (14:20)
∅r

Changes in selectivity depend on the correlation between the surface concentration

(number) of electrons (es) and the surface concentration of holes (h+s). If the concen-
tration of es is larger than the concentration of holes h+s, the reaction of reduction on
SCPs is dominant (reduction of H+); in the reverse case, the oxidation reaction is pre-
ferred. This factor is important when photo water splitting is carried out in solutions
containing other redox systems.
More than a hundred SC materials, mostly oxides, were tested as potential photo-
catalysts in water splitting. Various syntheses, doping, cocatalysts, and specially orga-
nized nanostructures were applied, but still the efficiency of photocatalytic water
splitting and H2 formation is not satisfactory. Very rich data about applied photocata-
lysts, their efficiency, and special treatments can be found in reviews [17, 21, 23] and
references therein. Note that some of these SC materials and strategies are also used
in the photoelectrolysis of water (Section 14.3.2). Here, we briefly describe some ap-
proaches more characteristic for SCPs.
As in the case of SCEs, it is convenient to divide the photocatalysts into two
groups: the photocatalyst absorbing the photons from the UV range of the solar spec-
trum and photocatalysts absorbing photons in the visible light range. The first group
involves oxides such as TiO2, Nb2O3, Ta2O3, Ga2O3, titanates, niobates, tantalates, a ma-
terial with perovskite structure, and others. These SC materials are stable during illu-
mination; however, their bandgap energy Eg is higher than 3 eV and they can only
absorb photons in the UV range of the spectrum, which accounts for a small fraction
of solar spectra (ca. 4%). The other group includes the photocatalysts possessing
bandgap energy Eg lower than 3 eV, such as metal oxides (Fe2O3 and WO3), sulfides,
selenides (CdS and CdSe), perovskites (BiFeO3 and LaFeO3), and oxynitrides (TaON
and T3N5). Unfortunately, some of them catalyze only half of the reaction of photoca-
talytic water splitting, mainly O2 evolution (WO3, In2O3, SnNb2O3, and Bi2WO3). The
photocatalysts with Eg < 3 eV are often applied in the architecture of the Z-scheme sys-
tem (see further). All these photocatalysts contain metal ions in their structures. Re-
cently, the applications of metal-free photocatalysts are in progress. Plenty of such
materials were studied, including elemental catalysts such as P, Si, and Se, and binary
compounds such as carbon nitrides CxNy. The intensively studied material is g-C3N4, a
 14.3 Solar energy in photo water splitting 347

graphite carbon-nitride-based polymer. It is a stable, visible light-active material with

a bandgap energy of 2.7 eV. It was shown that this photocatalyst causes the evolution
of H2 from water at wavelength >420 nm, even without the cocatalyst used. Much
more about metal-free photocatalysts and their application in hydrogen production
can be found in the review [24].
Generally, three main steps can be distinguished during photocatalysis of water: the
photon absorption by SCPs, separation of photogenerated charge carriers e–h+, and reac-
tions of H2 and O2 evolution on SCPs. The efficiency of water photocatalysis is deter-
mined by all of them and is low. Different strategies are developed to improve the
efficiency of water splitting and the effectiveness of these steps. As in the case of photo-
electrodes (Section 14.3.2), to improve photon absorption range, the particles of photoca-
talyst were doped with metals (Ag, Au, and Pt), nonmetals (N, P, and S), and metal ions
(Cr3+, Ni2+, and Rh3+) or were sensitized by using visible light sensitizer – noble metals
(Ag, Au, and Pt) or organic dyes. Doping of SC materials changes their electronic struc-
ture causing a decrease in bandgap energy, or the formation of new energy levels in a
bandgap. These effects shift the edge of photon absorption to the Vis range of the solar
spectrum. However, the additional energy levels created in the bandgap can work as
recombination centers decreasing the amount of photogenerated charge carriers. Metal
nanoparticles deposited on photocatalyst surface can implement several functions. They
are not only sensitizers but also can work as cocatalysts, and as surface traps protecting
the charge carriers (electrons and holes) against recombination. Particularly interesting
is the application of Ag and Au nanoparticles loaded on SCPs. These nano-MePs can ab-
sorb visible light via surface plasmon resonance, enhancing photocatalytic activity. The
application of dyes is limited because during illumination they are oxidized by O2
evolved at particles and decompose.
The other problems that should be solved are connected with the separation of
electron–hole pairs generated during the illumination of SCPs and the separation of
H2 and O2 evolved, before the backward reaction of water formation takes place. Few
approaches are developed to cope with these problems. One of them is the application
of cocatalysts, the other is the design of the systems containing two-photon absorbers
(SCPs) referred to as Z-scheme and cascade system.
Many different cocatalysts were tested in the photocatalysis of water. Noble met-
als (Pt, Ru, Rh, Au, and Ag) were widely applied to improve the efficiency of hydrogen
evolution. However, they are too expensive. Therefore, other nonnoble metal catalysts
have been developed such as MoS2, NiS, NiO, and CoO. For improvement of the OER,
oxides IrO2 and RhO2 were applied, but also other oxides were tested such as NiOx,
CoOx, and MnOx. The cocatalysts can be loaded on SCPs in the form of nanoparticles
by various methods. These methods are mainly divided into two groups. In one of
them, the cocatalysts are grown directly on the surface of photocatalyst particles
using a precursor. In other, the cocatalysts are synthesized and after that loaded on
the surface of photocatalysts. The precipitation and photodeposition are used for the
348 14 Green fuel: hydrogen production

direct formation of cocatalysts, while for loading of cocatalysts impregnation, grind-

ing, and calcination are applied.
How does this system cocatalyst/photocatalyst work? Figure 14.10 presents the
processes between metal cocatalysts (MeC) and host photocatalysts (SCP). Let us sup-
pose that the cocatalyst has Fermi energy EF,Me lower than the SC photocatalyst EF,SC.
After the deposition of Me nanoparticles on the photocatalyst, the equilibrium is es-
tablished between electrons of both materials. This state is characterized by equality
EF,Me = EF,SC and is reached by the flow of electrons from the CB of SCP to the metal. As
a result, the metal part of the interface is charged negatively, and the SC part of the
interface is charged positively. Note that it means the formation at the SC surface of a
thin space charge layer in a depletion state. The barrier formed at the interface (junc-
tion) is referred to as the Schottky barrier. During illumination and absorption of UV
photons, this barrier decreases as the energy of SC electrons and EF,SC increase (to
✶
EF,SC Þ. These photoexcited electrons flow from CB to the metal, also increasing EF,Me.
and participate in the reduction of H+ ions at the metal/solution interface. At the same
time, the photogenerated holes oxidized water at the SCP/solution interface. To improve
the efficiency of O2 evolution, the additional cocatalyst can be deposited on SCPs. These
cocatalysts are mainly oxides, and plenty of them have SC properties; therefore, the
type of junctions that are formed between cocatalysts and photocatalysts depends on
the type of SC and band energy matching.

cocatalyst (Me)

charge layer
e
EC EC EF,Me
EF,SC EF,SC E✶F,SC
EF,Me EF,Me

hν

EV EV

SCP SCP

Fig. 14.10: Schematic illustration of cocatalyst/photocatalyst system. Principle of work. EF,SC, E✶F,SC, EF,Me –
Fermi energies of an electron in semiconductor (SC) and metal (Me); ✶ means illumination of SC.

The electric field of the junctions, formed at the interface of cocatalyst/photocatalyst,

enhances significantly the separation of photogenerated electron–hole pairs, acceler-
ates the motion of free charge carriers to the surface reaction sides, and prevents the
backward reactions of water formation. The applied cocatalysts work also as ordinary
catalysts, decreasing the activation energy of HER and OER. Much more about cocata-
lysts, their synthesis, and their application can be found in the review [25]. This cocat-
 14.3 Solar energy in photo water splitting 349

alyst/photocatalyst design can be classified as one photon absorber (SCP), one-stage

system with four photons taking part in overall water splitting.
The design that is widely tested and applied is the so-called Z-scheme, a two-
photon absorber two-stage system (Fig. 14.11). In the Z-scheme design, the reaction of
photocatalytic water splitting is divided between two different photocatalysts – SCPs;
hence, the evolved H2 and O2 are separated. These two stages are combined by a
mediator–redox couple in the solution, or by a solid system. During illumination, in
both particles (SCP1 and SCP2), charge carriers are photogenerated. At SCP1, electrons
reduced H+ ions to H2 and holes oxidized the Red form of redox couple in the solution,
while at SCP2 holes oxidized H2O and electrons reduced the Ox form of redox couple
(see also Fig. 11.7). If the solid mediator is used, the holes from SCP1 should recombine
with electrons from SCP2.

hν

e e e H+ /H2
H+ /H2
EC

hν e e EC
Ox Ox/Red
Red
EV hν
O2/H2O
h+ h+ h+ O2/H2O
EV
h+ h+
SCP 1 h+
SCP 2

Fig. 14.11: Schematic energy diagram of photocatalytic water splitting Z-system (Z-scheme).

Plenty of SCPs were tested in the design of Z-scheme for water splitting. Among them
SrTiO3, TaON, BaTaO2N, C3N4 and WO3, TiO2, and Ta3N5 are applied for H2 and O2 evo-
lution, respectively. It is worth mentioning that the Z-scheme system containing
SrTiO3 (doped with Cr and Ta) and WO3 coupled by mediator iodate/iodide ions can
split water under visible light illumination. Z-designs applied for water splitting have
some advantages:
– They enhance the separation of photogenerated charge carriers.
– They eliminate the backward reaction of water formation.
– They can utilize the visible part of the solar spectrum, as two particles with differ-
ent Eg are used.

The other two absorber photon systems, which can utilize visible light and facilitate
the separation of photogenerated electron–hole pairs, are cascade systems. Cascade is
designed in the form of two different SCPs coupled together or SCP particles deposited
350 14 Green fuel: hydrogen production

on the nanosheet of SC. Systems such as SrTiO3/TiO2, WO3/TiO2, and CdS/TiO2 were
tested. They can be used for hydrogen production and degradation of organic com-
pounds. However, there is no information about the ability for water splitting.
Plenty of efforts have been made to improve the efficiency of water photocataly-
sis. For many years, the rates of H2 and O2 evolutions were on the level μmol/h g not
having practical meaning. The situation changed with the huge development of nano-
technologies. New strategies are developed, systems with nanoarchitecture are de-
signed, and visible-light-driven photocatalysts are synthesized. It is not possible to
describe all of them, but we should mention the one-step water splitting photocata-
lysts such as the plasmonic structure composites (Au–TiO2 systems), solid solutions
[(Ga1–xZnx)(N1–xOx)], and a new generation of cocatalyst design containing noble metal
(core)/Cr2O3 (shell) nanoparticles. At present, the best rates of H2 and O2 evolutions were
obtained on CoO one-step photocatalyst, under visible light, as 71.7 and 34.5 mmol/h g
for H2 and O2 evolved, referred to as 5% STH [17, 23]. However, for most of photocata-
lysts, the rates of H2 and O2 evolutions are contained in the range from μmol/h g to few
mmol/h g [17].
If we consider only H2 evolution, the best efficiencies are obtained if the reaction of
hydrogen generation is carried out in the aqueous solution containing sacrificial re-
agents. In these cases, the photocatalysts participate only in half-reaction of water split-
ting. The rates of H2 evolution depend not only on the type of photocatalysts, and
incident light used, but also on the type and concentration of sacrificial reagents. The
rates of H2 evolution are in the range of 10–150 mmol/h g. If the photocatalytic reaction is
carried out in the presence of electron donors (hole scavengers), photogenerated holes
oxidize the reducing reagent (electron donor) instead of water, while photogenerated
electrons participate in H2 evolution, enhancing the rate of reaction. If we want to en-
hance the rate of O2 evolution, we utilize, as the sacrificial reagents, acceptors of elec-
trons, which are reduced by photogenerated electrons, while photogenerated holes
oxidize water. The reactions with sacrificial reagents are often used at the laboratory
level for the evaluation of photocatalysts applied in water splitting. Alcohols, sulfide, sul-
fite ions, Ag+, and Fe3+ are often used as electron and donor acceptors for such purposes.
This half-reaction of H2 evolution may be important for practical hydrogen production if,
as reducing reagents, biomass or other abundant compounds will be used.

References

[1] Tashie Lewis BCh, Nnabuife SG. Hydrogen production, distribution, storage and power conversion
in a hydrogen economy – A technology review. Chem. Eng. J. Adv. 2021, 8, 100172–100191.
[2] Smolinka T, Garche J. Electrochemical power sources. Fundamentals, systems and applications.
Hydrogen production by water electrolysis. Amsterdam, Netherland, Elsevier, 2021.
[3] Bockris JO`M, Reddy AKN. Modern electrochemistry 2B. NY, USA, Kluwer Academic/Plenum
Publisher, 2000.
 References 351

[4] Wang S, Lu A, Zong CJ. Hydrogen production from water electrolysis: Role of catalysts. Nano
Converg. 2021, 8, 4. doi: 10.1186/s40580-021-00254-x.
[5] Durovic M, Hnat J, Bouzek K. Electrocatalysts for the hydrogen evolution reaction in alkaline and
neutral media. A comparative review. J. Power Sources 2021, 493, 229708.
[6] Song J, Wei C, Huang ZF, Liu C, Zeng I, Wang X, Hu ZJ. A review of fundamentals for designing
oxygen evolution electrocatalysts. Chem. Soc. Rev. 2020, 49, 2196–2214.
[7] Li J. Oxygen evolution reaction in energy conversion and storage: Design strategies under and
beyond the energy scaling relationship. Nano-Micro Lett. 2022, 14, 112.
[8] Zeng K, Zhang D. Recent progress in alkaline water electrolysis for hydrogen production and
application. Prog. Energy Combust. Sci. 2010, 36, 307–326.
[9] Kecebas A, Kayfeci M, Bayat M. Electrochemical hydrogen generation. In: Calise F, D`Accadio M,
Ferrevo D, eds. Solar hydrogen production. Processes, systems and technologies. Amsterdam,
Netherland, Elsevier, 2019, 299–317.
[10] Kumar SS, Himabindu V. Hydrogen production by PEM water electrolysis – A review. Mater. Sci.
Energy Technol. 2019, 2, 442–454.
[11] Tu H. Solid electrolytes. In: Bagotsky VS, ed. Fundamentals of electrochemistry. Hoboken, USA,
Wiley & Sons Inc, 2006, 419–436.
[12] Laguna-Bercero MA. Recent advances in high temperature electrolysis using solid oxide fuel cells: A
review. J. Power Sources 2012, 203, 4–16.
[13] Mohammed- Ibrahim J, Mousab H. Recent advances on hydrogen production through seawater
electrolysis. Mater. Sci. Energy Technol. 2020, 3, 780–807.
[14] Rousseau R, Etcheverry L, Roumbaud E, Besseguy R, Delia M-L, Bergel A. Microbial electrolysis cell
(MEC): Strengths, weaknesses and research need from electrochemical engineering standpoint.
Appl. Energy 2020, 257, 113938.
[15] Hasany M, Mardanpour MM, Yaghmaei S. Biocatalyst in microbial electrolysis cells: A review. Int.
J. Hydrog. Energy 2016, 41, 1477–1493.
[16] Walter MG, Warren EL, McKone JR, Boetcher SW, Mi Q, Santori EA, Lewis NS. Solar water splitting
cells. Chem. Rev. 2010, 110, 6446–6473.
[17] Cao S, Piao L, Chen X. Emerging photocatalysts for hydrogen evolution. Trends Chem. 2020, 2,
57–70.
[18] Rajeshwar K. Hydrogen generation at irradiated oxide semiconductor – Solution interface. J. Appl.
Electrochem. 2007, 37, 765–787.
[19] Diez-Garcia MI, Gomez R. Progress in ternary metal oxides as photocathodes for water splitting
cells: Optimization strategies. Solar RRR 2022, 6, 2100871.
[20] Hu S, Levis NS, Ager JW, Yang J, McKone JR, Strandwitz NC. Thin film materials for protection of
semiconducting photoelectrodes in solar fuel generators. J. Phys. Chem. C 2015, 119, 24201–24228.
[21] Wang Z, Li C, Domen K. Recent developments in heterogenous photocatalysts for solar –driven
overall water splitting. Chem. Soc. Rev. 2019, 48, 2109–2125.
[22] Chowdhury F. Recent advances and demonstrated potentials for clean hydrogen via overall solar
water splitting. MRS Adv. 2019, 4, 2771–2785.
[23] Ismail AA, Bahneman DW. Photochemical splitting of water for hydrogen production. Sol. Energy
Mater. Sol. Cells 2014, 128, 85–101.
[24] Rahman MZ, Kibria G, Mullins Ch B. Metal-free photocatalysts for hydrogen evolution. Chem. Soc.
Rev. 2020, 49, 1887–1931.
[25] Tian L, Guan X, Zong S, Dai A, Qu J. Cocatalysts for photocatalytic overall water splitting: A mini
review. Catalyst 2023, 13, 355.
15 Electrochemical and photocatalytic methods
in pollutant removal

15.1 General remarks

All living creatures need air and water for life. For centuries, the equilibrium has
been established between the natural environment and human beings. From the nine-
teenth century with the Industrial Revolution, everything changed as a result of
human activities, developments of technologies, and the increasing demand for goods
and energy. We feel the climate changes, and our health is deteriorating as the
amount of pollutants due to our activity increases in nature. The problems of detec-
tion and removal of pollutants are crucial in our days. The detection of some pollu-
tants by using sensors based on electrochemical reactions is described in Chapter 13.
The pollutants are everywhere. The main air pollutants are gases such as CO, SO2, NO,
NO2, and NH3, volatile organic compounds (VOCs), and particulates. They can be re-
moved by applying physicochemical processes such as absorption, adsorption, and
combustion. For instance, SO2 is removed by a process called flue gas desulfurization,
in which gas dissolves in/or reacts with absorbent, being trapped. Gas adsorption is
used for the control and removal of VOCs. For such purposes also combustion is ap-
plied, converting organic pollutants into water and CO2. Carbon dioxide for many
years was treated as a natural air compound. At present, CO2 is referred to as a green-
house gas responsible for global warming and climate change. Among other green-
house gases (e.g., H2O, N2O, CH4, and O3), CO2 is not as effective in heat absorption and
re-radiation as N2O or CH4; however, its emission is very high (82%) in comparison
with CH4 (10%) and N2O (8%) and should be significantly reduced. The main source of
CO2 is human activity and increasing demand for energy, which is mostly produced
by burning fossil fuels (coal, natural gas, petroleum, and gasoline). The best way to
reduce the CO2 level in the air is the application of alternative energy sources: solar
energy (see Chapters 11 and 14), nuclear energy, and wind energy. Currently, the tech-
nology for the capture and storage of CO2 is being developed. In this two-step technol-
ogy, CO2 is firstly separated from other gases and after that stored in geological beds
In the presence of oxygen and in the humid environment, gases NO, NO2, and SO2
form nitrous, nitric, and sulfuric acids (Chapter 13) contributing to acidic rain precip-
itations. They fall to the ground, on surface waters, and at soil causing a pH decrease
and other deteriorating effects. In the last decades, the acidification of oceans has
been observed as a result of the huge dissolution of CO2 and carbonic acid formation
(H2CO3). Most of the gaseous pollutants have anthropogenic origins, but it is worth
mentioning that NH3 and H2S can be produced in bacterial processes of biomass
decomposition.
Water, an excellent solvent, can dissolve plenty of inorganic and organic com-
pounds, and is therefore very susceptible to pollution. The ground and surface waters

https://doi.org/10.1515/9783111160986-016
 15.2 Removal and recovery of heavy metals: electrochemical treatment 353

(e.g., oceans, lakes, and rivers) are contaminated by air and soil pollutants, which can
have natural or anthropogenic origins. The most polluted waters are wastewater which
comes from domestic, industrial, and agricultural activities. These waters contain
plenty of inorganic and organic contaminants. Among them are heavy metals, sulfides,
nitrates, phosphates, organohalogens, organophosphorus, pharmaceutical compounds,
and others. They exist in the form of dissolved species, colloids, and suspensions. Some
of them are toxic and hazardous. The heavy metals are introduced into waters natu-
rally, by erosion of minerals and soil, through volcanic eruptions and biogenic pro-
cesses. They came also as a result of human activity from the metallurgy industry,
factories of batteries, textiles, electroplating processes, and many others. In waste-
waters, depending on the composition and solution pH, the heavy metals can exist as
free ions, hydroxides, and stable complexes, formed with compounds such as chloride,
citrate, cyanide, and ethylenediaminetetraacetic acid (EDTA). They are highly toxic, po-
tentially carcinogenic, nonbiodegradable, and can accumulate in human beings and
other biosystems. A variety of techniques have been developed and applied to remove
heavy metals and other contaminants from water, such as adsorption, coagulation, re-
verse osmosis, and membrane filtration technology. Among them are electrochemical
methods: electrodeposition (ED), electrodialysis (EDl), electrocoagulation (EC), and elec-
troflotation (EF) [1]. The principles, some advantages, and disadvantages of these meth-
ods are briefly described in the next section.
In recent years, plenty of research has been carried out on the application of ad-
vanced oxidation processes (AOPs) for the purification of different wastewater. AOPs
can be described as oxidative reactions in which input energy (chemical, light, and
electrical) is used for the production of very reactive oxidizing species – hydroxyl rad-
icals, enabling the in situ oxidation and decomposition of organic pollutants. It was
shown that AOPs are very effective in the degradation of organic compounds and or-
ganic heavy metal complexes. AOPs can be realized by a variety of methods. Among
them are Fenton oxidation, electro-Fenton oxidation, photocatalytic oxidation, and
others. The application of photocatalytic oxidation of organic pollutants raised signifi-
cant attention, as solar energy can be used for such purposes. Some principles of the
application of electro-Fenton oxidation and photocatalysis for the removal of organic
compounds are also described in this chapter.

15.2 Removal and recovery of heavy metals: electrochemical

treatment

In the last few years, the fastest growing waste in the world is electronic waste (e-
waste). Their production is estimated at about 50–60 million tons a year. Only a small
amount of them (about 17%) is properly treated and recycled. Electronic waste (e-
waste) means discarded electrical and electronic devices. All of them contain plenty
of metals, divided mostly into three categories: (i) base metals which constitute about
354 15 Electrochemical and photocatalytic methods in pollutant removal

30 wt% (e.g., Cu, Cd, Zn, Cr, Ni, and Co); (ii) precious metals constitute 0.1–1 wt% (e.g.,
Pd, Pt, Au, Ag, and Ir); and (iii) rare earth elements such as Gd, La, Sc, and Pr, and the
amount of them is lower than 0.1 wt%. Since some of these metals are toxic, some are
valuable, and the supply of some of them is not sufficient for further high-tech devel-
opment, they have to be recovered. Different strategies are developed, but the first
step is common – the extraction of metals by leaching the waste with acid, alkali, or
even with complex compounds and conversion of solid metal-containing species into
soluble ionic forms. An example is acid leaching of metal sulfides or hydroxides:

ZnSðsÞ + H2 SO4 ðaqÞ + 1=2O2 ðgÞ ! ZnSO4 ðaqÞ + H2 O ðlÞ + SðsÞ

AgOH ðsÞ + HNO3 ðaqÞ ! AgNO3 ðaqÞ + H2 O ðlÞ

Leachate solutions from acidic treatment (or others) of e-waste contain multiple
metal ions, which differ from each other, posing a challenge to the selective recovery
of pure metals. For such purposes, electrochemical methods can be used such as elec-
trowinning, often referred to as electroextraction and electrorefining. In the electro-
extraction method, the metal ions dissolved in a leach solution are reduced onto the
cathode and extracted in their metallic form. In the electrorefining method, an anode
is made from impure metals, which have to be refined. During electrooxidation, the
anode is partially dissolved in electrolytic media and the formed free metal ions mi-
grate to the cathode, where they are reduced. In both methods, the main processes
are ED processes. As pointed out above, metal contaminants exist in the ground and
surface waters and in larger amounts in industrial wastewater. ED is applied for the
removal of free metal ions from all kinds of water.

15.2.1 Electrodeposition and electrodialysis

Electrodeposition (ED) means electrolytic processes carried out in two-electrode cells

at constant potential or current. They are governed by Faraday’s law. During ED, new
solid phases are formed. Much more about the principles, phase formations, and prac-
tical aspects can be found in Chapter 5. In the ED method, the free metal ions dis-
solved in wastewater or leaching solutions are reduced at the cathode. Their redox
potential should be more positive than the redox potential of the H+/H2 redox couple.
If not, the hydrogen is evolved at the electrode, disturbing the reduction of ions and
decreasing the efficiency of the reaction:
RT
Men+ ðaqÞ + ne ! Me ðsÞ, Eeq = E0 + ln aMen+
nF
2H+ðaqÞ + 2e ! H2 ðgÞ, E0 = 0 V
 15.2 Removal and recovery of heavy metals: electrochemical treatment 355

The redox potential of H+ ions depends on pH of the solution and is described by the
equation:

E = − 0.059 pH, and at pH 7 it is equal to − 0.413 V ðvs. SHEÞ

If we compare only the values of standard potentials of Men+/Me with this value, it is
clearly seen that by ED, we can recover metals such as Ag, Cu, Pb, Ni, and Co, if their ions
do not exist in the form of hydroxides or stable complexes. At the anode, the oxidation of
some metallic contaminants takes place. This process is utilized in electrorefining:

Me ðsÞ ! Men+ ðaqÞ + ne

If the potential of the anode is sufficient, the oxidation of water or hydroxide ions
occurs, depending on the pH of the solution (see Section 14.2.1)

2H2 O ðlÞ ! O2 ðgÞ + 4H+ ðaqÞ + 4e, E0 = 1.229 V ðvs. SHEÞ

4OH−ðaqÞ ! O2 ðgÞ + 2H2 O ðlÞ + 4e, E0 = 0.401 V ðvs. SHEÞ

In alkali media, most of the transition ions form insoluble hydroxides, which can be
adsorbed on anode undergoing further oxidation to metal oxides:

Men+ ðaqÞ + nOH−ðaqÞ ! MeðOHÞn,ads

MeðOHÞn,ads ! MeOn ðsÞ + nH+ ðaqÞ + ne

Effluents originating from electroplating baths or leaching solutions can contain a mix-
ture of free heavy metal ions. In theory, they can be separated and selectively reduced by
careful control of potential. However, in practice, the deposition reactions depend on the
pH of the solution and impurities. It was shown that it is possible to deposit Cu from a
solution containing Cu2+, Cd2+, or Zn2+ ions, but not Pb in the presence of In3+ ions, despite
differences in the standard potentials (about 0.2 V). Selective recovery of Cu (95%) from e-
waste was reached in an ammonia-based electrolyte, while selective recovery of Fe from
the leaching of magnets (Nd–Fe–B) was obtained when ammonium sulfate and sodium
citrate were used. The processes of selective recovery of metals from wastewater and
leaching solutions are very complicated and require detailed analysis of solution compo-
sition. In EDs, applied for removal and recovery of metals from leaching solutions, aque-
ous media are mostly used. However, some metal ions have a very high redox potential,
for example, Al3+/Al − 1.67 V, U3+/U − 1.66 V, Pt2+/Pt + 1.19 V, and some react in aqueous
solutions and cannot be removed in such a way. Therefore, often nonaqueous electro-
lytes are used. They include conventional organic solvents such as dimethylsulfoxide
and dimethylformamide, but also ionic liquids (see Section 1.3.1). More details about the
electrochemical recovery of metals from e-waste and wastewater can be found in re-
views [2, 3] and references therein. The other problem that should be solved during the
ED of metals from wastewaters is a low concentration of metal ions (not more than a
few thousand ppm), resulting in the decrease of the ion transportation rate and the high
356 15 Electrochemical and photocatalytic methods in pollutant removal

values of concentration overpotential (see Section 1.1.3). A few methods are used to im-
prove the transport of ions to the electrode and increase their concentration, such as
extensive mixing of solution by using rotating electrodes, gas bubbling, concentrator
electrochemical technique, or electrodialysis [1]. The selectivity and efficiency of metal
recovery from wastewater may be improved when the EDl and ED are coupled together.
To date, EDl has been widely used in the desalinization of water and the brine
industry. At present, the EDl method is also applied for the removal and recovery of
heavy metal ions from the electroplating industry residues. The operating principles
of EDl are simple. The scheme of the three-compartment EDl cell is shown in Fig. 15.1.

CEM AEM

C A
Men+

Xn–
Catholyte

Anolyte

influent

Fig. 15.1: Schematic illustration of an electrodialysis cell. A, anode; C, cathode; CEM, cation-exchange
membrane; AEM, anion-exchange membrane.

Cell contains two electrodes (cathode and anode) and two ion-exchange membranes
(IEM) (cation-exchange membrane (CEM) and anion-exchange membrane (AEM)).
CEM is negatively charged and can attract positively charged metal ions, while AEM is
positively charged and can attract negatively charged anions. The electric field, gener-
ated between two electrodes, works as a driving force causing the migration of ions
toward electrodes. Migration of ions across the IEM allows an increase of cationic con-
centration in the catholyte part of the cell and anionic concentration in the anolyte
part, respectively. The cell can be fed with industry effluents continuously and catho-
lyte solutions may be removed for further treatments. In this case (Fig. 15.1), the cell
contains two IEMs, but cells with only one or a stack of membranes are also used.
More about membranes and membrane potential can be found in Section 1.4. The EDl
may be carried out in two ways, one without separation of metal ions but with an
increase of their concentrations, and the second with an increase of both separation
and concentration. The second way is referred to as selective EDl and is very impor-
tant for the recovery of metal from the industry effluents containing a mixture of
metal ions. The selective separation can be achieved by pH matching, application of
membranes selective to the charge of ions, and application of complexing compounds.
 15.2 Removal and recovery of heavy metals: electrochemical treatment 357

The specially prepared selective membranes are used for the separation of differently
charged cations. For instance, the Nafion membrane covered with a thin layer of poly-
ethyleneimine shows better selectivity in the separation of Cr3+ from a solution con-
taining Na+ and Cr3+ ions. Lately, the heterogeneous and chelating membranes have
been tested in the recovery of metals. The heterogeneous CEM consisting of acrylam-
ide-2-methyl propane sulfonic acid was successfully used for the selective separation
of Ni2+, Pb2+, and K+ ions. The other approach is the application of complexing agents
such as chloride, citric acid, or widely used EDTA. They form complexes with metal ions
changing their charge from positive to negative. For instance, Ni2+ and Co2+ ions can be
separated in the presence of EDTA, since nickel ions form negatively charged complexes
Ni(EDTA)2– transferred to the anolyte part of the chamber. The EDl technology is simple,
and what is important is the removal of ions can be carried out in a continuous flow of
industry effluent; therefore, the application of this method for recovery of ions is grow-
ing. However, some drawbacks have to be overcome such as the high demand for elec-
trical energy caused by the resistance of EDl cells and the high cost of membranes. More
detailed information about EDl and its application in the separation of metal ions can be
found in review [4].

15.2.2 Electrocoagulation and electroflotation

The classical methods of the removal of organic pollutants and heavy metals from
wastewater are coagulation/flocculation and flotation. Both methods have the same
origin and are based on the precipitation of coagulated species. In both methods, co-
agulants such as Fe2+ and Al3+ are used. The sludge formed during the precipitation of
species in coagulation/flocculation processes falls to the bottom of the reservoir. In
the flotation, the precipitates or ions with surfactants are driven to the surface by gas
bubbles. Coagulation processes are well-known in colloid science [5].
The wastewater contains plenty of colloid pollutants. They include organic species,
metal oxides, sulfides, and biotic materials. Colloid systems consist of two phases: the
dispersed phase containing particles with diameters ranging from 1 nm to 1 μm, and
the continuous phase (dispersing phase). If this phase is liquid, we use the term “sol.”
Sols can be divided into two groups: lyophilic sols and lyophobic sols. In aqueous sus-
pensions, they are referred to as hydrophilic and hydrophobic sols. Hydrophilic sols are
formed mainly by large organic molecules (e.g., peptides) and their charge originates
from ionized function groups. They are highly hydrated and stabilized by solvent
(water) molecules. The hydrophobic sols contain insoluble particles of metal oxides, sul-
fides, chlorides, and others. Their charge comes from particles’ ionized surface groups
or chemically adsorbed ions (or both). Their stability is caused by the interaction be-
tween particles’ charge and oppositely charged other ions existing in water. As a result,
the electrical double layer (e.d.l.) is formed at the interface dispersed particle/solution.
This e.d.l. consists of two parts: (i) compact layer referred to as the Helmholtz layer,
358 15 Electrochemical and photocatalytic methods in pollutant removal

with a linear potential drop in its space, due to the particle surface charge, water mole-
cules, and solvated counter ions from solution, and (ii) diffusion layer, referred to as
the Gouy–Chapman layer, where the distribution of ions is described by Maxwell’s rela-
tion, and its potential changes exponentially with distance from the compact layer (see
Section 1.1.1, Figs. 1.3–1.5, setting for the charge on a metal surface the charge on a parti-
cle surface). In an external electric field, the hydrophobic sol particles, together with
their compact layer, move together as one species to the respective electrode, whereas
the rest of their e.d.l. moves in the opposite direction. The “plane” that divides the two
regions of the e.d.l. is called the plane (surface) of shear. The potential that exists on
this surface is called the zeta potential (ζ) or electrokinetic potential and is an indirect
measure of the charge of the colloid particles (Section 1.1.1., Fig. 1.9). It is worth stressing
that this potential can be measured only when the colloidal particles are set in motion
with respect to the dispersing phase by the presence of an external electric field. The
formation of e.d.l. in the hydrophobic sols is responsible for all electrokinetic phenom-
ena such as electroosmosis, electrophoresis, streaming potential, and sedimentation po-
tential. The addition of ions that can adsorb at particle/solution interface, decreasing
charge of e.d.l., and zeta potential destabilizes the colloid system. The particles become
thermodynamically unstable and at zeta potential near zero begin to coagulate. The co-
agulation ability is dependent on the valence of coagulating ions: the greater the va-
lence of ions, the better the ability for coagulation. The coagulation of hydrophilic
colloids can be achieved by the application of other solvents than water, or ions with
higher energy of hydration than energy of interaction of water molecules with colloid
particles. Coagulation is connected to flocculation. Coagulation means the destabiliza-
tion of colloids by neutralization of forces (mostly charges) which keep them separated,
and results in the agglomeration on a molecular level and formation of microparticles.
Flocculation means the agglomeration of destabilized colloid particles in flocks (large
aggregates) by flocculants such as polyacrylamide (PAM), polyaluminum chloride, or
polyferric sulfate. Note that very often both terms are used interchangeably. In coagula-
tion, the coagulant agents Fe2+ and Al3+ come from ionized salts, mostly chlorides and
sulfates. In electrocoagulation, Fe2+ and Al3+ ions are produced electrochemically,
in situ in wastewaters, by anodic electrooxidation of Fe, steel, or Al. At the cathode, the
reduction of water takes place with H2 evolution. The reactions that occur at the anode
and further in the solution are complicated and depend on the pH of the solution
(wastewater). The Pourbaix diagrams of Fe/H2O, Al/H2O, equilibriums, and reactions in
such systems are described elsewhere [6, 7]. Here we consider some of them. Let us
start with the Fe anode. The reactions that occur in acidic media are

Anode: Fe ðsÞ ! Fe2+ ðaqÞ + 2e, E0 = − 0.441V ðvs. SHEÞ

Cathode 2H+ ðaqÞ + 2e ! H2 ðgÞ, E0 = 0 V ðvs. SHEÞ

The reactions at the anode are multistep reactions with the formation of adsorbed
species (FeOH and FeOH+). Fe2+ ions are formed in the oxidation reaction of FeOH+ads:
 15.2 Removal and recovery of heavy metals: electrochemical treatment 359

FeOH+ ads + H+ ðaqÞ ! Fe2+ ðaqÞ + H2 O ðlÞ

In the presence of O2 in acidic media, Fe2+ ions are oxidized to Fe3+ ions and insoluble
Fe(OH)3 is formed:

4Fe2+ ðaqÞ + 10 H2 OðlÞ + O2 ðgÞ ! 4FeðOHÞ3 ðsÞ + 8H+ ðaqÞ

In alkaline (basic) media, the ferrous ions produced in the anode oxidation react with
hydroxide ions forming insoluble Fe(OH)2. The reactions are

Anode: FeðsÞ ! Fe2+ ðaqÞ + 2e, Fe2+ ðaqÞ + 2OH−ðaqÞ ! FeðOHÞ2 ðsÞ

Cathode: 2H2 O ðlÞ + 2e ! H2 ðgÞ + 2OH−ðaqÞ, E0 = − 0.828V ðvs. SHEÞ

Insoluble Fe(OH)2 and Fe(OH)3 hydroxides remain in wastewaters as suspensions,

which can remove pollutants by the formation of surface complexes or by electro-
static interactions followed by coagulation. In complexes, the pollutants work as li-
gands to hydrous ions:

Pollutant-H + ðHOÞOFeðsÞ ! pollutant-OFe ðsÞ + H2 OðlÞ

If in EC, Al anode is used, electrooxidation leads to the formation of Al3+ ions:

Al ðsÞ ! Al3+ ðaqÞ + 3e, E0 = − 1.676 V ðvs. SHEÞ

Al3+ ions are not stable in aqueous solutions and undergo spontaneous hydrolysis. In
these reactions, different species are generated depending on the pH of the solutions. In
acidic media Al(OH)2+, Al(OH)2+ soluble species dominate and insoluble aluminum hy-
droxide Al(OH)3 is formed in reactions:

Al3+ ðaqÞ + H2 O ðlÞ ! AlðOHÞ2+ ðaqÞ + H+ ðaqÞ

AlðOHÞ2+ ðaqÞ + H2 O ðlÞ ! AlðOHÞ2 + ðaqÞ + H+ ðaqÞ

AlðOHÞ2 + ðaqÞ + H2 O ðlÞ ! AlðOHÞ3 ðsÞ + H+ ðaqÞ.

In alkaline media, at pH < 9, insoluble Al(OH)3 species are formed, while at pH > 9 sol-
uble Al(OH)4– ions dominate in solutions. At the same time, at the cathode, H2 is
evolved (see Fe reactions). All the species generated during and after the dissolution
of the anode are charged. As in the case of Fe, they can remove the pollutants in two
ways: by the neutralization of the negative charge of colloid particles followed by co-
agulation and by the formation of surface complexes with pollutants. The H2 evolved
during the reduction of hydrogen ions or water molecules at the cathode can also be
applied for the removal of pollutants by the flotation process. EC technology has been
widely used for the removal of suspended particles of many pollutants; among them
are organic compounds, heavy metals (e.g., Zn, Ni, Co, Mn, and Mo), and also anions
360 15 Electrochemical and photocatalytic methods in pollutant removal

such as CN–, F–, and PO43–. Note, that by using EC we can only remove pollutants from
water, but we cannot separate them. Much more about EC applications and parameters
influencing the efficiency of EC, an interested reader can find in references [1, 8]. Flota-
tion and electroflotation are other methods widely used for the removal of suspended
species: metal particles, colloids, flocks, and oil contaminants from wastewater. Re-
moval is achieved by the bubbling gases, which adhere to the species and flow them to
the surface. In EF [9], the gases are generated electrochemically, and these gases are
used for the removal of pollutants. As in EC, in EF, the hydrogen is evolved at the cath-
ode and the pollutants can flow to the surface. Additionally at proper potential or cur-
rent density, at the anode not only Al is oxidized to Al3+ ions but also water is
decomposed into O2:

Anode 2H2 O ðlÞ ! O2 ðgÞ + 4H+ ðaqÞ + 4e, E0 = 1.229 V ðvs. SHEÞ, acidic media

Anode 4OH−ðaqÞ ! O2 ðgÞ + 2H2 O ðlÞ + 4e, E0 = 0.401 V ðvs. SHEÞ, basic media

Both gases H2 and O2 are used in flotation processes. EF can be applied separately, but
very often the combination of EC and EF is used. In Fig. 15.2, the processes of coagula-
tion, flotation, and sedimentation, which proceed during EC and EF, are schematically
presented.

A C
(Fe)
H2
Fe2+

H2 Fig. 15.2: Schematic illustration of electrocoagulation,

electroflotation, and sedimentation in an
electrochemical reactor.

All these methods, described above, can be used for metal removal and recovery.
However, the combination of ED and EDl has a major advantage, over EC and EF pro-
cesses: the recovery is selective. Therefore, these methods are often used for recover-
ing valuable metals needed for high-tech devices. During the application of EC and EF
methods, the heavy metals and other metals precipitate together with organic pollu-
tants forming flocks, sludge, or surface suspensions. Such a mixture needs further
treatment, so the cost of recovery of metals increases. By using EC and EF, we can
remove from wastewater organic pollutants, but many of them are toxic and should
be degraded. For such purposes, other methods are used.
 15.3 Degradation and removal of organic pollutants: electrochemical treatment 361

15.3 Degradation and removal of organic pollutants:

electrochemical treatment
For the removal and degradation of organic pollutants, oxidizing processes are used.
Among them, we can distinguish the special group referred to as AOPs. These pro-
cesses are widely used for the purification of wastewater from biologically toxic, non-
degradable organic materials such as pesticides, volatile organics, or detergents.
During AOPs, all these pollutants and others can be converted to CO2 and H2O, and
inorganic compounds (the mineralization of organics). It was shown that AOPs are
very efficient and decrease the concentration of some pollutants to very low levels
from ppm to ppb. Generally, AOPs mean the oxidation reactions of organic com-
pounds by very strong hydroxyl radicals •OH generated chemically, electrochemically,
or by light, mainly from H2O2. It is worth mentioning that •OH radicals have one of the
highest redox standard potential E0 = 2.8 V (vs. SHE) for the •OH/H2O system [10]. Pro-
cesses where hydroxyl radicals are generated electrochemically are referred to as elec-
trochemical AOPs (EAOPs). Alternative to AOPs, EAOPs, and technologies connected
with them, other oxidative electrochemical technologies are developed. They are based
on common anodic oxidation processes. Below, we briefly describe Fenton and elec-
tro-Fenton reactions, where the source of hydroxyl radicals •OH is hydrogen perox-
ide (H2O2) and some of the anodic processes utilized in the degradation of organic
pollutants.

15.3.1 Fenton and electro-Fenton oxidation

Fenton processes (FPs) refer to the reactions where reagents Fe2+ and H2O2, often
called Fenton reagents, are involved in the generation of hydroxyl radicals •OH. Plenty
of studies were carried out to describe the mechanism of these reactions. Generally,
we can distinguish a few crucial reactions in which radicals (•OH and HO2•) are
formed and Fe2+ ions are recovered:

Fe2+ ðaqÞ + H2 O2 ðaqÞ ! Fe3+ ðaqÞ +  OH. + OH−ðaqÞ


OH + H2 O2 ðaqÞ ! HO2  + H2 O ðlÞ

Fe3+ ðaqÞ + HO2  ! Fe2+ ðaqÞ + H+ ðaqÞ + O2 ðgÞ

Fe3+ ðaqÞ + H2 O2 ðaqÞ ! Fe2+ ðaqÞ + HO2  + H+ ðaqÞ

FPs are much more complicated. The interested reader is referred to reviews [10, 11]
and references therein. It was mentioned above that Fenton reactions are widely ap-
362 15 Electrochemical and photocatalytic methods in pollutant removal

plied in the degradation of organic species. The degradation is caused by •OH and or-
ganic radicals generated during oxidation and fragmentation of organic compounds:

RH +  OH ! R + H2 O

Organic radicals are very reactive and in further reactions may produce various radi-
cal species (RO• and ROO•). All these generated radicals, ferrous and ferric ions, react
further in chain reactions resulting in the mineralization of organic compounds. The
crucial parameter in the Fenton reactions is the pH of the wastewater. The reactions
should be carried out at pH 3, which means that the pH of neutral or slightly basic
wastewater should be adjusted. The cost of application of the Fenton reaction to puri-
fication is fairly high, as expensive chemical reagents are required. The cost may de-
crease when electro-FPs (EFPs) are applied.
In EFPs, the Fenton reagents (Fe2+ and H2O2) are formed electrochemically in situ.
The wastewater and other waters flow continuously through an electrochemical reac-
tor, where Fenton reagents are produced at electrodes. The reagents react further
generating the radicals and initiating the chain of reactions shown partially above.
The EFPs can also be realized in other ways. In one way, only one reagent is produced
electrochemically (mostly H2O2) and iron (II) salts are externally added, in another
way both reagents are supplied externally. Let us concentrate on the first case, where
Fenton reagents are produced electrochemically. At a proper current density (voltage)
in an electrochemical reactor, an anode (Fe or steel) dissolves with the formation of
ferrous ions:

Fe ðsÞ ! Fe2+ ðaqÞ + 2e, E0 = − 0.441 V ðvs. SHEÞ

while at the cathode in an acidic solution, peroxide (H2O2) is formed in the reduction
of O2 dissolved in solution, or coming from the air. Generally, in acidic solutions, O2
can be reduced in two pathways: direct four-electron reduction or two-electron reduc-
tion which involves the formation of H2O2:

O2 ðgÞ + 4H+ ðaqÞ + 4e ! 2H2 O ðlÞ, E0 = 1.229 V ðvs. SHEÞ

or

O2 ðgÞ + 2H+ ðaqÞ + 2e ! H2 O2 ðaqÞ, E0 = 0.695 V ðvs. SHEÞ

Both reactions are sluggish, and the exchange current of these reactions ranges from
10−10 to 10−11 A/cm2. The catalysts have to be applied in both reactions. Plenty of stud-
ies were carried out to determine the mechanism and to find the proper catalysts. It
was shown that direct reaction is preferable (water formation) when, in the initial
step of the reaction, O2 is strongly adsorbed and O−O bonds are broken (see Sec-
tion 4.5.1). The catalysts that are used in the reaction of H2O2 formation should weakly
 15.3 Degradation and removal of organic pollutants: electrochemical treatment 363

adsorb O2 molecules, have a low value of activation overpotential for O2 reduction,

and also release H2O2 before further reduction or decomposition occurs:

H2 O2 ðaqÞ + 2H+ ðaqÞ + 2e ! 2 H2 O ðlÞ, E0 = 1.76 V ðvs. SHEÞ

2H2 O2 ðaqÞ ! O2 ðgÞ + 2H2 O ðlÞ

It was found that carbon materials have such properties toward H2O2 formation.
Plenty of them are used and tested in the electrochemical reactors applied in the H2O2
industry and the treatment of wastewater. Among them are graphite, carbon foam,
mesoporous carbon, graphene composites, carbon nanotubes with metal particles,
and others (see Section 4.3.2). When H2O2 is produced at the cathode, but Fe2+ ions
come from added Fe(II) salts, the inert anode is applied, made mainly from oxides
such as IrO2/RuO2, TiO2, and Nb2O5. In this case at the anode, water is oxidized with O2
evolution. Much more information about this reaction and the catalyst applied, an in-
terested reader can find in Section 14.2.1. The organic pollutants from wastewater can
be removed by EFPs in electrochemical reactors. Their design may be different, but a
few parameters are important and common:
– The pH of effluents should be kept at 3. In other cases, Fe3+ ions created in Fenton
reactions precipitate as Fe(OH)3; moreover, H2O2 is not stable in neutral and alka-
line solutions.
– In many cases, the conductivity of wastewater is low; therefore, some electrolytes
should be added.
– An essential parameter is current density determined by operating conditions.
– It was established that the optimal temperature for Fenton reactions is 30 °C.

If we compare EFPs with classical ones (FPs), we clearly see the advantages of EFPs:
– The EFPs are less expensive. H2O2 can be produced in situ in wastewater; there-
fore, the transport of reagent is eliminated.
– By using proper current density or potential (voltage), the processes are more
controllable.
– By using the proper pH of reaction media, the production of sludge is limited or
fully eliminated.

Hydroxyl radicals generated during both Fenton methods are very strong oxidants able
to nonselectively react with organic pollutants yielding their complete mineralization.

15.3.2 Anodic oxidation

Electrochemical oxidation processes (anodic oxidation processes – AOx) are widely

applied in the degradation and removal of organic pollutants. They are friendly to the
environment and able to mineralize nonbiodegradable organic compounds. These
364 15 Electrochemical and photocatalytic methods in pollutant removal

AOx processes can undergo two routes: (i) direct oxidation occurring at the anode sur-
face and (ii) indirect oxidation where electrochemically generated mediators (oxi-
dant) oxidize the pollutant in solution. In direct AOx, the organic molecules adsorbed
on the anode surface are oxidized in direct charge (electron) transfer without involve-
ment of any other substances:

Rads ðMÞ ! R+ + e, M is the active site of the anode surface and R is organic compounds

This process may be carried out at potentials lower than the potential of water oxida-
tion and oxygen evolution. However, very often, direct transfer reaction resulted in
the formation of insoluble, not fully oxidized product fouling and poisoning electrode
surface. This effect can be avoided by leading the oxidation at potentials higher than
the potential of water decomposition, or by using indirect oxidation. If the oxidation
is carried out according to the first way, the processes can be catalyzed by intermedi-
ate, oxygen-containing radicals (•OH, O•) generated during water oxidation and oxy-
gen evolution (see Section 14.2.1). They adsorb on the anode active side M and may
participate in the oxidation of organic molecules:

M + H2 O ðlÞ ! M½ OHads + H+ ðaqÞ + e

M½ OHads ! M½O ads + H+ ðaqÞ + e

R + M½ OHads ! RO + H+ ðaqÞ + M + e

R + M½O ads ! RO + M, RO is partially oxidized organic compounds

The most important is the hydroxyl radical •OH, which can be chemisorbed or
weakly adsorbed on the anode surface depending on the anode nature. If hydroxyl
radicals are weakly adsorbed (physical adsorption), the radicals are freer to react
with pollutant molecules at electrodes and assist in the complete mineralization of
organics:

R + M½ OHads ! CO2 ðgÞ + H2 O ðlÞ + H+ ðaqÞ + M + e

Chemisorbed radicals can cause partial oxidation of organic compounds (RO).

A different mechanism is proposed for the oxidation of organics at some metal
oxide anode (MOx). In the first step, the •OH radicals are also formed during water
decomposition and are chemisorbed on/in active sites of anode, followed by oxidation
and formation of so-called higher oxide (MO(x +1)). This higher oxide acts as a mediator
in the selective oxidation of organic pollutants:

MOx ðsÞ + H2 O ðlÞ ! MOx ð OHÞads + H+ ðaqÞ + e

MOx ð OHÞads ! MOðx+1Þ ðsÞ + H+ ðaqÞ + e

MOðx+1Þ + R ! MOx ðsÞ + RO

 15.3 Degradation and removal of organic pollutants: electrochemical treatment 365

If the adsorption of •OH radicals on metal oxide anode is weak, the hydroxyl radical
may assist in the nonselective complete oxidation of organic pollutants:

MOx ð OHÞads + R ! MOx + CO2 + H2 O

Note the possibility of competitive reaction of O2 evolution; therefore, O2 can also par-
ticipate in the oxidation of pollutants. Since in the case of direct anodic oxidation pro-
cesses, the •OH radicals also participate, and the direct AOx reactions belong to EAO
processes.
In indirect anodic oxidation, the pollutants are oxidized by strong oxidizing re-
agents generated electrochemically in situ in wastewater at anode or cathode (H2O2).
These reagents work as mediators in charge transfer between the electrode and organ-
ics in solution. They may be divided into two groups: the metallic redox couple such as
Ag+/Ag2+ (E0 = 1.98 V) and Co2+/Co3+ (E0 = 1.82 V) and strong oxidizing reagents such as
chlorine (Cl–/Cl2, E0 = 1.358), ozone (O2/O3, E0 = 2.075 V), peroxide (H2O/H2O2, E0 = 1.763 V),
and persulfate (SO42–/S2O82–, E0 = 2.01 V) [10]. All these compounds possess the high val-
ues of standard redox potentials pointing to their oxidation power. The processes
where metallic redox couples take place are referred to as mediated electrochemical
oxidation (MEO). They are preferably used for the treatment of solid waste and concen-
trated solutions containing more than 20% of waste. The metal ions in an acidic solution
are anodically oxidized to the higher oxidation state, in which they react with organic
pollutants inducing their degradation:

Men+ ! Meðn+1Þ + + e, Meðn+1Þ + + R ! Men+ + CO2 + H2 O

Such reactions were successfully applied for the degradation of cellulose, oil, resins,
urea, organic acid, and others. When wastewaters contain low concentrations of or-
ganic pollutants (lower than 20%), the effluents are treated by other oxidizing reac-
tants electrogenerated in situ. They originate from the oxidation of water and ions
present in wastewater effluents such as chlorides, sulfates, and carbonates. Among
them, chlorine is most widely applied in wastewater treatment. Chlorides are oxidized
in the reaction:

2Cl−ðaqÞ ! Cl2 ðgÞ + 2e, E0 = 1.358 V ðvs. SHEÞ

Chlorine is not stable undergoing disproportionation in water solution:

Cl2 ðgÞ + H2 O ðlÞ ! HClO ðaqÞ + H+ ðaqÞ + Cl−ðaqÞ

Cl2 ðgÞ + 2 OH−ðaqÞ ! ClO−ðaqÞ + H2 O ðlÞ + Cl−ðaqÞ

Hypochlorous acid and hypochlorite ions can undergo further oxidation to ClO2– chlo-
rite ions and ClO3– chlorate ions in an alkaline medium. All these chlorine-containing
oxidants are strongly reactive and can cause the mineralization of pollutant mole-
cules, but they also can produce some harmful, unwanted products. Chlorine was
366 15 Electrochemical and photocatalytic methods in pollutant removal

used for the degradation of phenols, glucose, textile dyes, pesticides, and others. Much
higher oxidant power than chlorine has other reagents: ozone, persulfate, and perox-
ide (H2O2/H2O, E0 = 1.765 V) generated at the cathode (see Section 15.3.1):

3H2 O ðlÞ ! O3 ðgÞ + 6H+ ðaqÞ + 6e, E0 + 1.51 V ðvs. SHEÞ

2SO4 2 − ðaqÞ ! S2 O8 2 − ðaqÞ + 2e, E0 = 2.01 V ðvs. SHEÞ

All these oxidants described above and others were tested in wastewater treatment
for removal of very hard degradable toxic organic pollutants.
The anodic oxidation of pollutants is carried out in a typical electrochemical reac-
tor equipped or not with a system for the continuous flow of wastewater. The main
parts of reactors are electrodes. The cathode is conventionally made from steel or
graphite. The heart of the reactor is the anode, its material, catalytic properties, and
adsorption ability to decide on the reaction pathway and the selectivity and efficiency
of mineralization of pollutants and their removal. Anodes applied in the oxidation of
organic pollutants differ in their efficiency and selectivity. On some of them, the reac-
tion of oxidation of organics is completed with the formation of CO2 and H2O, and at
others, partial oxidation occurs. It was found that anodes such as SnO2, PbO2, and
BDD (boron-doped diamond) at which the potential of O2 evolution is higher (1.9 V,
1.9 V, 2.3 V in H2SO4 solution) favor the complete oxidation of organic pollutants.
Others such as RuO2, IrO2, and Pt, characterized by lower potential evolution of O2
(1.47 V, 1.52 V, 1.6 V in H2SO4 solution) participate in partial oxidation of organic
pollutants.
In practice, most anodes can participate in both AOx pathways. Applying high
enough potential to the anode, all reactions with direct charge transfer and mediator
generation may occur simultaneously. Depending on the pollutants that have to be re-
moved, different anodes are applied in AOx processes such as Pt, metal oxides, mixed
metal oxides, or graphite. Pt is considered a good anode for many electrooxidation reac-
tions, for example, phenol, dyes, and herbicides. However, as the cost of Pt is too high,
it should be replaced with other materials. Metal oxides such as IrO2, RuO2, Nb2O5, Ta2O5,
SnO2, and mixed oxides are very effective in promoting Cl2 and hypochlorite production.
Some of them, IrO2 and RuO2, are widely used as catalysts in water oxidation and O2
evolution. It was shown that at high potential, the strong O3 oxidant may be produced at
PbO2. In recent years, the application of BDD (boron doped diamond) electrodes in AOx
processes has aroused interest. BDD electrodes are chemically stable, are resistive
against corrosion even in harmful media, and have a very large potential for oxygen
evolution (2.3 V vs. SHE). Therefore, BDD anodes are widely used for near complete re-
moval of very hazardous pollutants from wastewater such as textile dyes, pharmaceuti-
cal contaminants, or surfactants. The other materials widely used for AOx processes are
carbon-based and graphite electrodes, for example, graphite rods, activated carbon, car-
bon papers, carbon brushes, and carbon nanostructures. Carbon electrodes cannot work
at high potentials; however, they are often used, since they are cheap and may be fabri-
 15.4 Solar energy: pollutant degradation and removal 367

cated as large porous assemblies with large surfaces. The interested readers can find
much more about the application of AOx processes in the degradation of pollutants in
reviews [11–13] and references therein.

15.4 Solar energy: pollutant degradation and removal

Sunlight is a renewable and sustainable source of natural energy that can be utilized
for the degradation and removal of pollutants from our environment. For such pur-
poses, solar energy may be used in two ways: indirect and direct. We have above-
described processes and electrochemical methods applied in the degradation and re-
moval of inorganic and organic pollutants from industrial and domestic wastewater.
Some of these methods such as electrodeposition (ED), electrocoagulation (EC), and
anodic oxidation electrolysis require a large amount of electrical energy delivered by
conventional energy sources. These electrochemical reactors can also be fed with elec-
trical energy coming from photovoltaic panels converting solar energy into electricity.
This can be treated as an indirect application of solar energy in pollutant removal.
The direct way of the utilization of solar energy in the purification of wastewater and
air is the application of photocatalytic processes. Photocatalytic processes convert
solar energy into chemical energy, making it possible to mineralize organic pollutants.
They can be divided into homogeneous and heterogeneous processes. Typical exam-
ples of homogeneous photoprocesses are photo-Fenton reactions, where reagents and
catalysts are in the same phase. The heterogeneous photocatalysis is carried out on
the solid phase: suspension of powdered catalyst or catalyst particles, nanoparticles
embedded in membranes, or other materials. The compounds used in heterophotoca-
talysis are semiconductors (SC). Much more information about SC materials, their
electrochemical, and photoelectrochemical properties, and their application in photo-
electrochemical cells, an interested reader can find in Chapters 10 and 11. In Chapter 12
and Section 14.3.3, the basis of photocatalytic processes on SC particles (SCPs), their
application in water photocatalysis, and the efforts to improve the light absorption
and quantum efficiency of SCPs are described. Here, we briefly described the applica-
tion of photocatalytic processes in the degradation and removal of some pollutants
pointing also to the possibility of using pollutants as electron donors or acceptors in
photocatalytic reactions, yielding the formation of valuable products, for example, H2,
CH3OH, and HOOH. Before going further, we have to bring to mind how these systems
work. Three steps can be distinguished during irradiation of SCPs by photons with
energy hν ≥ Eg: (i) photogeneration of charge carriers: electrons in the conduction
band (CB) and holes (h+) in the valence band (VB); (ii) separation of charge carriers:
electrons and holes; and (iii) participation of charge carriers in reactions at SCP/solu-
tion interface. Electrons generated in CB reduce electron acceptors (A), whereas holes
from VB oxidize the electron donors (D). Processes are illustrated in Fig. 15.3.
368 15 Electrochemical and photocatalytic methods in pollutant removal

hν
(V) (eV)

e e e
EC A A– A/A– (P)

hν
D D+ D+/D (P)
EV
h+ h+ h+

SPC

Fig. 15.3: Schematic illustration of photocatalytic reaction. A, acceptor; D, donor; P, pollutant.

Reactions occur when the thermodynamics requirements are fulfilled. Generally, the
energy of the CB edge, EC, has to be higher than the energy level of the A/A– couple,
that is, it is located at a more negative potential relative to the redox potential of A/A–
in solution. The energy of the VB edge, EV, should be lower than the energy level of
the D/D+ couple, that is, located at a more positive potential than the redox potential
of the D/D+ couple in solution (see Fig. 15.3). The roles of electron acceptors or electron
donors (known also as sacrificial agents) can play pollutant molecules. Contaminant
molecules adsorbed on SCP surfaces are irreversibly reduced or oxidized. During
their degradation, sometimes, valuable products are formed. Let us consider some ex-
amples of pollutant degradation.

15.4.1 Inorganic pollutants

For illustrative purposes, we choose carbon dioxide CO2 (greenhouse gas) and H2S pol-
lutants existing in natural gas, petroleum, and wastewaters, being also products of
biomass fermentation. Their degradation differs; carbon dioxide adsorbed on the SCP
surface is reduced by electrons from CB during illumination of SCPs, while H2S is oxi-
dized as a gas, or sulfide ions in solution by photogenerated holes from VB. The pro-
cess of reduction of CO2 is difficult as the energy of the C = O bond is high (850 kJ/mol).
During the reduction of CO2 in aqueous solutions, different products can be formed,
including CO, HCOOH, HCHO, CH3OH, CH4, and C2H5OH. These reactions are multielec-
tron reactions involving from 2 to 12 electrons. For instance, 2 electrons are involved
in the formation of carbon monoxide and formic acid, 4 electrons participate in form-
aldehyde formation, 6 in methanol formation, and 8 and 12 in methane and ethanol
formation. Some of these reactions are quoted in Section 11.3.4. The values of redox
potentials of these reactions are close to each other, so during CO2 reduction, a mix-
ture of products can be formed. For instance, the redox potentials of CO2/HCOOH, CO2/CO,
CO2/HCOH, and CO2/CH3OH, in an aqueous solution, at pH 7 are as follows: −0.58 V,
 15.4 Solar energy: pollutant degradation and removal 369

−0.52 V, −0.46 V, −0.37 V vs. NHE, respectively. They are not only close to themselves but
also close to the equilibrium potential of H+/H2 at pH 7 equal to −0.413 V; therefore, during
CO2 reduction, hydrogen may also be evolved. Plenty of researches were carried out on
the photocatalytic reduction of CO2 on different kinds of SCPs. In Fig. 12.4 (Section 12.3),
you find the energetic diagram pointing to the correlation between some popular oxide
photocatalysts and redox potentials of CO2 reduction. Oxides such as TiO2, ZnO, and NiO
may be used for such purposes.
However, these oxides absorb the UV part of the solar spectrum; therefore, plenty
of efforts are made for their modification, especially doping with metals (Pt), nonme-
tals (N), structural changes by using materials in the form of nanorods, nanotubes,
and so on. Other photocatalysts were also tested for CO2 reduction such as sulfides
(CdS and Bi2S3), metal-organic compounds, and photocatalysts modified with enzymes.
Much more about the photocatalytic reduction of CO2 and applied photocatalysts, one
can find in reviews [14, 15] and references therein. However, the main problem is still
not solved: how to selectively reduce CO2, for instance, to CH3OH, or more generally,
how to carry out the synthetic reactions

nCO2 + nH2 O + hν ! ðCH2 OÞn + nO2

in which the absorbed greenhouse gas will be utilized and converted into valuable
products.
During the reduction of CO2 by electrons from CB, the photogenerated holes in VB
react with electron donors (D) specially added to the solution or oxidize water mole-
cules producing a radical intermediate or O2. Other processes occur during the illumi-
nation of SCPs in the presence of H2S. H2S is a very interesting compound, in which
the O molecule atom is replaced by S, and the molecule may be split with H2 evolu-
tion, according to the decomposition reaction of H2S (g), or H2S (aq):

H2 S ! H2 ðgÞ + SðsÞ

This reaction requires much less energy than in the case of water decomposition.
Standard Gibbs free energy of reactions ΔrG0 are 33.56 kJ/mol and 27.89 kJ/mol, respec-
tively, while ΔrG0 for H2O (l) is equal to 237.13 kJ/mol. In consequence, there is a large
difference in the standard redox potential of reactions of H2O (l) and H2S (aq) splitting
(1.229 V and 0.144 V vs. SHE, respectively). In solution, H2S is a weak aprotic acid that
ionizes in two steps:

H2 S $ H+ + HS−, pK = 7, and further HS− $ H+ + S2 − , pK = 11.96

During photocatalytic decomposition of H2S in solution, photogenerated electrons and
holes reduce proton ions and oxidize sulfide ions, respectively:

SC + hν ! e ðCBÞ + h+ ðVBÞ

2H+ ðaqÞ + 2e ðCBÞ ! H2 ðgÞ

370 15 Electrochemical and photocatalytic methods in pollutant removal

2HS−ðaqÞ + 2h+ ðVBÞ ! 2H+ ðaqÞ + S2 2 − ðaqÞ

2S2− ðaqÞ + 2h+ ðVBÞ ! S2 2−

Photocatalytic decomposition of H2S can be also carried out in the gas phase. In this
case, H2 and S are produced. The decomposition potential of H2S is much lower than
the potential of water splitting; therefore, SC materials possessing low bandgap en-
ergy Eg are widely used. They can absorb the photons from the visible part of the
solar spectrum by increasing the efficiency of decomposition of H2S and hydrogen
evolution. Plenty of SC materials were tested; among them are metal sulfides CdS and
ZnS. To enhance their activity, some efforts have been made to combine them with
other SC, forming heterojunctions, for example, CdS/TiO2, CdS/ZnO, and ZnS/CdS, to
dope them with ions such as Cu2+ and Ni2+ (ZnS), or to use them in nanostructures.
Much more about photocatalytic H2S decomposition and applied photocatalysts, you
can find in review [16]. Note that most of the results are obtained by researchers
using different sources (UV–Vis lamps) for illumination; however, you can also find
some results where systems with simulated sunlight were used. The photocatalytic
processes can be used for the removal of heavy metal ions in wastewater [17]. They
are reduced and the electrons photogenerated in CB of SC photocatalyst participate in
the reduction. It was shown that such photocatalysts as TiO2–ZrO2 can easily reduce
Cu2+ and Pb2+ ions to Cu, Pb, and Cr(VI) ions to Cr(III) in reaction:

Cr2 O7 2− + 14H+ + 6e ðCBÞ ! 2Cr3+ + 7H2 O

The photocatalysts and photocatalytic processes also have substantial significance in

the degradation and removal of organic pollutants.

15.4.2 Organic pollutants

In Section 15.3, we described the processes (AOPs) that cause the oxidation of organic
pollutants to a high degree, even to mineralization. A strong oxidant – hydroxyl radi-
cal •OH – participates in these reactions; however, other radicals such as HO2• can
also be involved. As was pointed out, the radicals are produced during catalytic pro-
cesses in Fenton reactions between Fe2+ and H2O2, or during anodic oxidation of
water, where radicals are intermediates in oxygen evolution. Such radicals can also
be formed during photocatalytic oxidation of water or photolysis of solution contain-
ing Fe2+ ions and H2O2 (Fenton system). It was shown that illumination of the Fenton
system with UV/Vis light in the range of 400–200 nm (approximately) accelerated the
degradation rate of pollutants. It results from the additional formation of •OH radi-
cals, and they are not only created in elementary Fenton reaction:
 15.4 Solar energy: pollutant degradation and removal 371

Fe2+ ðaqÞ + H2 O2 ðaqÞ ! Fe3+ ðaqÞ +  OH + OH−ðaqÞ

but also in reactions:

Fe3+ ðaqÞ + H2 O ðlÞ + hν ! Fe2+ ðaqÞ +  OH + H+ ðaqÞ

H2 O2 ðaqÞ + hν ! 2 OH

Thus, illumination not only regenerates Fe2+ ions crucial in catalytic reactions but
also increases the amount of •OH radicals participating in the reactions. Such Fenton
reactions during illumination are referred to as photo-FPs and belong to the group of
homogeneous photocatalysis as Fe2+ ions, H2O2, and pollutants are in the same liquid
phase. Recent works have shown that instead of the application of expensive illumina-
tion with a UV–Vis lamp, the sunlight energy may be used in FPs and remediation of
wastewaters. Such reactions are referred to as solar-FP. Much more about different
FPs and their applications in the degradation of the pollutants, the interested readers
can find in review [10] and references therein.
The FPs may be realized also in another way, using solid photocatalysts – iron-
based compounds such as iron oxides, for example, α -Fe2O3 (hematite), γ -Fe2O3 (maghe-
mite), Fe3O4 (magnetite), iron (hydroxyl) oxides FeOOH, or spinel ferrites MeFe2O4
(Me – Mn, Fe, Co, and Ni). They can be used in the form of particle suspensions, par-
ticles embedded in membranes, or different nanostructures. These compounds have
large adsorption ability toward pollutant molecules. Some of them possess the bandgap
energy Eg lower than 2.8 eV (α -Fe2O3 and γ -Fe2O3 are 2.2 eV and 1.83 eV, respectively),
so they can absorb the photons from the visible part of the solar spectrum. They are
also partially dissolved in dark and under light, being the source of Fe2+ and Fe3+ ions at
the particle surface. Ferric ions can react further with H2O2 produced during photore-
duction of O2:

SC + hν ! e ðCBÞ + h+ ðVBÞ

e ðCBÞ + O2 ðgÞ + 2H+ ðaqÞ ! H2 O2

Fe2+ + H2 O2 ! Fe3+ + 2 OHads

The hydroxyl radicals oxidize adsorbed pollutant molecules causing their degradation:

Rads +  OHads ! ROads ðpartially oxidized intermediateÞ ! nCO2 + nH2 O

Plenty of iron-based materials were tested in photocatalytic degradation of different

pollutants, among them were pharmaceutical drugs. For instance, CoFe2O4 was used
for the degradation of doxycycline, Fe2O3-modified graphene nanoplatelets for the re-
moval of amoxicillin, and the maghemite nanoparticles for the degradation of tetracy-
cline. More detailed information, interested readers can find in the review [18] and
references therein.
372 15 Electrochemical and photocatalytic methods in pollutant removal

The degradation of pollutants is caused not only by •OH and HO2• radicals gener-
ated during different Fenton systems or anodic oxidation, these radicals are also
formed during photocatalytic reactions of oxygen reduction and water oxidation. Dur-
ing illumination of SCPs, photogenerated CB and VB electrons and holes can partici-
pate in reactions:

SC + hν ! e ðCBÞ + h+ ðVBÞ

h+ ðVBÞ + H2 Oads ! H+ +  OH

h+ ðVBÞ + OH− ! 
OH

e ðCBÞ + O2 ! O2 −

O2  − + H+ ! HO2  , HO2  notation is equal to HOO

Radicals HO2• and O2•− undergo further reactions

The interested reader can find the sequence of reactions and diagram presenting
the correlation between energy of CB, VB edges, and redox potentials of oxygen radi-
cals in Section 12.3 (Fig. 12.3). There are plenty of organic pollutants in wastewater
such as phenolic compounds, pesticides, hydrocarbons, dyes, and drugs. For their deg-
radation and removal, mostly binary metal oxides such as TiO2, ZnO, SnO2, and Fe2O3,
and materials with oxide structures such as perovskites (ABO3), that is, SrTiO3 and
NaTaO3 are used, since they are stable in a hazardous environment and can partici-
pate in production of highly oxidant radicals. They degrade a wide range of organic
pollutants in biodegradable nontoxic molecules, which can be further mineralized to
CO2 and H2O. Detailed information about the application of photocatalysts in the deg-
radation of different classes of pollutants can be found in reviews [15, 17–19]. Photo-
catalytic degradations and removal of pollutants are carried out in photocatalytic
reactors. They differ in configurations and design, and their construction depends on
the kind of pollutants. Their description is beyond the scope of this book. Interested
readers can find such information in [20].

References

[1] Yasri NG, Gunasekaran S. Electrochemical technologies for environmental remediation. In: Anjum
NA, Gill SS, Tuleja Z, eds. Enhancing cleanup of environmental pollutants. Non biological
approaches. Dordrecht, Germany, Springer, 2017, vol. 2, 5–73.
[2] Rai V, Liu D, Xia D, Jayaraman Y, Gabriel JCP. Electrochemical approaches for the recovery of metals
from electronic waste: A critical review. Recycling 2021, 6, 53.
[3] Qasem NAA, Mohammed RH, Lawal DU. Removal of heavy metal ions from wastewater: A
comprehensive and critical review. Npj Clean Water 2021, 4, 36. doi: 10.1038/s41545-021-00127-0.
[4] Juve J-MA, Christensen FMS, Wang Y, Wei Z. Electrodialysis for metal removal and recovery. A
review. Chem. Eng. 2022, 435, 134857.
 References 373

[5] Hiemenz PC, Rajagoplan R. Principles of colloid and surface chemistry. NY, USA, Marcel Dekker
Inc, 1997.
[6] Bard AJ. Encyclopedia of electrochemistry of elements. V9, Part A: Hg, Fe, H. NY, USA, Marcel Dekker
Inc, 1979.
[7] Bard AJ. Encyclopedia of electrochemistry of elements. NY, USA, Marcel Dekker Inc, 1976.
[8] Mickova IL. Advanced electrochemical technologies in wastewater treatment. Part I
Electrocoagulation. Am. Sci. Res. J. Eng. Technol. Sci. 2015, 14, 233–257.
[9] Mickova IL. Advanced electrochemical technologies in wastewater treatment. Part II Electro –
Flocculation and Electro-Flotation. Am. Sci. Res. J. Eng. Technol. Sci. 2015, 14, 273–294.
[10] Brillas E, Sires I, Oturan MA. Electro-Fenton processes and related electrochemical technologies
based on Fenton`s reaction chemistry. Chem. Rev. 2009, 109, 6570–6631.
[11] Vasudevan S, Oturan MA. Electrochemistry and water pollution. In: Lichtfous E, Schwarzbauer J,
Robert D, eds. Green materials for energy, products and depollution. Environmental Chemistry for a
sustainable world. Dordrecht, Germany, Springer, 2013, vol. 3, 27–68.
[12] Panizza M, Cerisola G. Direct and mediated anodic oxidation of organic pollutants. Chem. Rev. 2009,
109, 6541–6569.
[13] Sires I, Brillas E, Oturan MA, Rodrigo MA, Panizza M. Electrochemical advanced oxidation processes:
Today and tomorrow. A review. Environ. Sci. Pollut. Res. 2014, 21, 8336–8367.
[14] Wang W-N, Soulis J, Yang JY, Biswas P. Comparison of CO2. Photoreduction systems. A review.
Aerosol Air Qual. Res. 2014, 14, 533–549.
[15] Guo W, Guo T, Zhang Y, Yin L, Dai Y. Progress on simultaneous photocatalytic degradation of
pollutants and production of clean energy: A review. Chemosphere 2023, 333, 139486.
[16] Preethi Y, Kanmani S. Photocatalytic hydrogen production. Mater. Sci. Semicond. Process 2013, 16,
561–575.
[17] Ren G, Han H, Wang Y, Liu S, Zhan J, Meng X, Li Z. Recent advances of photocatalytic application in
water treatment: A review. Nanomater 2021, 11, 1804.
[18] Olusegun SJ, Souza TGF, Souza GO, Osial M, Mohallen NDS, Ciminelli VST, Krysinski P. Iron-based
materials for the adsorption and photocatalytical degradation of pharmaceutical drugs. A
comprehensive review of the mechanism pathway. J. Water Process Eng. 2023, 51, 103457.
[19] Al-Nuaim MA, Alvasiti AA, Shnain ZY. The photocatalytic process in the treatment of polluted water.
Chem. Pap. 2023, 77, 677–701.
[20] Shanghagi M, Sargazi H, Bazargan A, Bellardita M. Photocatalytic reactor types and configuration.
In: Bazagran A, ed. Photocatalytic water and wastewater treatment. London, United Kingdom, Iwa
Publishing, 2022, 73–110.
List of abbreviations
A Anode, acceptor
AC Acetonitrile
AC Alternating current
AEM Anion-exchange membrane
AFC Alkaline fuel cell
AFM Atomic force microscopy
AOx Anodic oxidation
AOP Advanced oxidation process
AWE Alkaline water electrolysis
BET Adsorption isotherm, Brunauer S, Emmet PH, and Teller EJ
BFC Biofuel cell
C Cathode
CB Conduction band
CCS Carbon capture and storage
CE Counter electrode
CEM Cation-exchange membrane
CNFs Carbon nanofibers
CNTs Carbon nanotubes
CP Conducting polymer
CPE Constant phase element; composite polymer electrode
CV Cyclic voltammetry, cyclic voltammogram
D Donor; diffusivity of charges; diffusion coefficient
xD Dimensional (x – 0, 1, 2, 3)
DC Direct current
DET Direct electron transport
DMF Dimethylformamide
DMFC Direct methanol fuel cell
DMSO Dimethylsulfoxide
DSSC Dye-sensitized solar cell
E Enzyme
Ea Activation energy
EAOP Electrochemical advanced oxidation process
EC Electrocoagulation, ethylene carbonate
ECALE Electrochemical atomic layer epitaxy
e.d.l Electrical double layer
EDTA Ethylenediaminetetraacetic acid
ED Electrodeposition
EDl Electrodialysis
EDLC Electrical double layer capacitor
EF Electroflotation
EFP Electro-Fenton process
EIS Electrochemical impedance spectroscopy
EMF Electromotive force
EPS Electrochemical power sources
EQE Quantum conversion efficiency
EQCM Electrochemical quartz-crystal microbalance
ES Enzyme–substrate complex

https://doi.org/10.1515/9783111160986-017
376 List of abbreviations

FA Formic acid
FP Fenton process
G Graphene
GC Glassy carbon
GDH Glucose dehydrogenase
GO Graphene oxide
(GO)X Glucose oxidase
HER Hydrogen evolution reaction
HHV High heating value of hydrogen
HOR Hydrogen oxidation reaction
HOPG Highly oriented pyrolytic graphite
HRP Horseradish peroxidase
HTE High-temperature electrolysis
IPCE Incident photon to current efficiency
IL Ionic liquid
ISE Inorganic solid electrolyte; ion-selective electrode
ITIES Interface between two immiscible electrolyte solutions
Lac Laccase
LHE Light-harvesting efficiency
LHV Low heating value of hydrogen
LTE Low-temperature electrolysis
M Active side of electrode/catalyst
MCFC Molten carbonate fuel cell
MCO Multicopper oxidase
MEA Membrane electrode assembly
MEC Microbial electrolysis
MEO Mediated electrochemical oxidation
MET Mediated electron transport
MFC Membrane fuel cell
MHLD Marcus, Hush, Levich, Dogonadze theory
NHE Normal hydrogen electrode
OCP Open-circuit potential
OER Oxygen evolution reaction
OHP Outer Helmholtz plane
OPD Overpotential deposition
ORR Oxygen reduction reaction
Ox Oxidized form of Redox system
PAA Porous anodic alumina
PAFC Phosphoric acid fuel cell
PAM Polyacrylamide
PC Propylene carbonate
PEC Photoelectrochemical cell
PEM Proton membrane electrolysis
PEMFC Proton-exchange membrane fuel cell
PV Photovoltaic cell
QC Quartz crystal
QCM Quartz-crystal microbalance
QDSSC Quantum dot-sensitized solar cell
QS Quantum size
 List of abbreviations 377

RE, Ref Reference electrode

rds Rate-determining step
Red Reduced form of Redox system
RGO Reduced graphene oxide
RHE Reference hydrogen electrode
S Substrate
SC Semiconductor
SCC Stress corrosion cracking
SCE Saturated calomel electrode, semiconductor electrode
SCP Semiconductor particle
SERR Surface-limited replacement reaction
SHE Standard hydrogen electrode
SMR Steam methane reforming
SOE Solid oxide electrolysis
SOFC Solid oxide fuel cell
SPE Solid polymer electrolyte
SSE Solid-state electrolyte
STE Energy conversion of solar to electricity
STF Energy conversion of solar to fuel
STH Energy conversion of solar to hydrogen
STM Scanning tunneling microscopy
SRH Shockley–Read–Hall (recombination)
ti Ionic transference number of “i”
Ui Electrolytic mobility
UHV Ultra-high vacuum
UPCD Underpotential codeposition
UPD Underpotential deposition
WE Working electrode
VB Valence band
VOC Volatile organic compound
List of symbols
A Surface area cm
ai Activity of species “i” mol/dm
Ce.d.l Differential capacitance of the electrical double layer μ F/cm2
CG-Ch, CH Differential capacitance of Gouy–Chapman and Helmholtz layers, μ F/cm2
respectively
CF Faradaic capacitance μ F/cm2
D Diffusion coefficient cm/s
E Electrical potential, electrode potential V
Ef Formal potential of an electrode V
E0 Standard potential of an electrode V
Eeq Equilibrium (Nernst) potential of an electrode V
ΔEeq Difference of equilibrium potentials of electrodes in an electrochemical V
cell, electromotive force (emf), and zero-current cell potential
ΔEeq
0 Difference of standard potentials of electrodes in an electrochemical cell V
ΔE Potential (voltage) applied to the electrolysis cell V
ΔEtn
0 Thermoneutral potential of electrolysis cell V
EML Potential of UPD monolayer formation V
ΔEUPD Difference between the reversible and the peak potentials during UPD V
formation
fi Activity coefficient Dimensionless
f0 Resonance frequency of a quartz crystal microbalance Hz, s−
Δf Change of resonance frequency of a quartz crystal microbalance s−
G Thermodynamic potential, Gibbs free energy J/mol
ΔG0# Standard Gibbs free energy of activation J/mol
ΔG0ads Standard Gibbs free energy of adsorption J/mol
ΔGc Gibbs free energy of formation of critical size nuclei J/mol
ΔG0mix Standard Gibbs free energy of the mixing (alloy) J/mol
ΔrG Gibbs free energy of reaction J/mol
Δr G0 Standard Gibbs free energy of reaction J/mol
H Enthalpy J/mol
ΔrH Enthalpy of reaction J/mol
Δr H0 Standard enthalpy of reaction J/mol
ΔHa0# Standard activation enthalpy J/mol
ΔHads 0 Standard enthalpy of adsorption J/mol
Ich Charging current of the cell A
Id Discharge current of the cell A
i, I Current A
j Current density A/cm
j0 Exchange current density A/cm
j0, A , j0, C Exchange current density of reaction at anode (A) and cathode (C) A/cm
jL Limiting current density A/cm
jL, A , jL, C Limiting current density of reaction at anode (A) and cathode (C) A/cm
K Equilibrium constant Dimensionless
Kads Adsorption equilibrium constant Dimensionless
k Rate constant for homogeneous reaction Depend on order
kf,b Rate constant for heterogeneous forward and backward reactions cm/s

https://doi.org/10.1515/9783111160986-018
380 List of symbols

m, M Mass and molecular weight g, g/mol

n Number of electrons transferred per molecule Dimensionless
ni Number of moles of species “i”
nF Charge transferred per mol C/mol
N0 Initial number of active nucleation sites Dimensionless
P Power W
p Pressure Pa
Q, q Charge C
R Resistance Ω
r, h Radius and height of disk shape nuclei nm
Rin , Rex Internal and external resistance of the cell Ω
S Entropy J/mol∙K
ΔrS Entropy of reaction J/mol∙K
Δr S0 Standard entropy of reaction J/mol∙K
T Temperature K, C
U Internal energy J/mol
Uch Charge potential of the cell V
Ud Discharge potential of the cell V
V Volume m−
W Warburg impedance Ohm
Wa Wagner number Dimensionless
wrev , wmax Reversible work and maximal work J
xA , xB Molar fraction of components Dimensionless
zi Valency of ion “i” Dimensionless

Greek symbols

α Cathodic transfer coefficient Dimensionless

Β Anodic transfer coefficient Dimensionless
γ Molar surface free energy, interfacial tension, N/m; activity coefficient Dimensionless
δ Thickness of diffuse layer m
ε Dielectric permittivity Dimensionless
ε Dielectric permittivity of vacuum F/m
ζ Electrokinetic potential and zeta potential V
η Overpotential V
ηa Activation overpotential V
ηc Concentration overpotential V
ηOhm Ohmic overpotential V
ηa, A , ηa, C Activation overpotential of reaction at anode (A) and cathode (C) V
ηc, A , ηc, C Concentration overpotential of reaction at anode (A) and cathode (C) V
ηF Faradaic efficiency Dimensionless
Θ Degree of surface coverage Dimensionless
γ Activity coefficient of an ion Dimensionless
κ Ionic conductivity Ohm− m−
Λ Molar conductivity Ohm− m mol−
 Symbols applied in photoelectrochemistry of SC (Chapters 10–12, 14, and 15) 381

λ Debye screening length, m; molar conductivity of an ion at infinite (Ohm− m mol−)

dilution
μi Chemical potential of species “i” J/mol
μei Electrochemical potential of species “i” J/mol
νi Stoichiometric coefficient of species “i” Dimensionless
ν Rate of potential sweep mV/s
ρ Density, g/cm, charge density, C/cm; resistivity Ohm m
σ Surface charge density, C/cm; conductivity (Ohm− m−)
τ Time constant s
Φ Electron work function eV
φ Internal potential V
χ Surface potential V
ψ External potential V
ω Angular frequency rad/s
ν Reaction rate, mol/dm s mol/m s

Symbols applied in photoelectrochemistry of SC

(Chapters 10–12, 14, and 15)

CH Capacity of Helmholtz layer μF/cm

CSC Capacity of space charge layer in SC μF/cm
CSS Capacity of surface states μF/cm
c Bulk concentration mol/dm
cOx , cRed Concentration of oxidized and reduced forms of the redox couple in mol/dm
solution
Dn , Dp Diffusion coefficient of electrons and holes cm/s
DRedox Density of electron states in redox system J− cm−
Dox Density of unoccupied electron states in redox system J− cm−
DRed Density of occupied electron states in redox system J− cm−
E Energy, potential eV, V
EF Fermi energy level of electron in SC eV
EF, Redox Fermi energy level of electron in redox system, in equilibrium eV
EF, dec, n , Fermi energy (electrochemical potential) for the decomposition reaction eV
EF, dec, p of SC electrode with participation of electrons and holes
EC , EV Energy of the edges of conduction and valence bands (CB and VB) in the eV
SC bulk
EC, S , EV, S Energy of the edges of conduction and valence bands at the surface of SC eV
Eg Energy of bandgap eV
ESS Energy level of the surface state eV
* *
n EF , p EF Energy of quasi-Fermi levels for electrons and holes eV
E (V) Electrode potential V
E0 Standard electrode potential V
Efb Flat band potential V
OC Open-circuit photopotential V
Eph
Edoc Open-circuit potential, in dark V
Eph Photopotential, energy of photon V, eV
ERedox , Eeq Equilibrium potential of redox system V
382 List of symbols

0 Standard potential of redox system V

ERedox
0 Standard potential of SC decomposition with participation of electrons V
Edec, n
0 Standard potential of SC decomposition with participation of holes V
Edec, p
ff Fill factor Dimensionless
Δr G Gibbs free energy of reaction J/mol
Δr G0 Standard Gibbs free energy of reaction J/mol.
Gn , Gp Rate of generation of electrons and holes under external stimulus cm− s−
I0 Light incident intensity W/cm
Iab Absorbed light (photon) flux W/cm
jn , jp Current density of electrons and holes A/cm
jn, diff Diffusion current density of electrons A/cm
jp, diff Diffusion current density of holes A/cm
ph
jsc, c
Short-circuit photocurrent density A/cm
jCB , j0, CB Current density, exchange current density, and currents that flow across A/cm
conduction band
jVB , j0, VB Current density, exchange current density, and currents that flow across A/cm
valence band
jcorr Corrosion current density A/cm
jph Photocurrent density (SC electrode) A/cm
k Rate constant of chemical reaction (first order) s−
Kads Equilibrium adsorption constant Dimensionless
LD Debye length μm
Ldiff Diffusion length μm
Ln , Lp Diffusion length of electrons and holes μm
LH Thickness of Helmholtz layer nm
LSC Thickness of space charge layer in SC nm, μm
Nc , NV Density of electron states in conduction and valence bands of SC J− cm−
ND , NA Concentration of donors and acceptors in the SC bulk cm−
NSS Density of surface states J/cm
n, p Concentration of electrons in conduction band and holes in valence band cm−
n0 , p0 Concentration of electrons in conduction band and holes in valence band cm−
in the equilibrium state
Pin Input power of solar irradiation W/m
Pout Output power of irradiated system W/m
QSC Charge of space charge layer in SC C
QSS Charge of surface states C
R Rate of hydrogen generation mol/s
Rn , Rp Recombination rates of electrons and holes cm− s−
S Selectivity Dimensionless
t Time s
U External potential applied to the SC electrode V.
 Constants 383

Greek symbols

ε Power efficiency Dimensionless

εSTH Solar to hydrogen conversion efficiency Dimensionless
εthresh Threshold efficiency Dimensionless
ϵch
stor
Energy storage efficiency Dimensionless
ϵSC Relative dielectric permittivity of SC Dimensionless
α Absorption coefficient m−
λ Wavelength m
τn , τp Lifetime of electrons and holes s
μn , μp Mobility of electrons and holes m/V s
~μe ðSCÞ Electrochemical potential of electrons in SC eV
~μe Electrochemical potential of electrons in redox system eV
(Redox)
φG Galvani potential V
φG − Ch Potential drop in Gouy–Chapman layer V
φH Potential drop in the Helmholtz layer V
φs , φb Potential at the surface, in the bulk of SC V
φSC Potential drop in space charge layer in SC V
φ0SC Potential drop in space charge layer; the SC is in equilibrium with V
redox system
∅ Quantum yield Dimensionless
∅i Quantum yield of i-product formation Dimensionless
∅r Quantum yield of r-reactant decomposition Dimensionless
νr Rate of chemical reaction mol/dm s
ν Frequency s−
ηA , ηC Overpotential of reactions at anode (A) and cathode (C) V

Constants

e Elementary charge and absolute value of electrical charge . × − C

ϵ0 Dielectric permittivity of vacuum . × − J− C m−, F/m
kB Boltzmann’s constant . × − J/K
h Planck’s constant . × − Js
R Gas constant . J/K mol
F Faraday’s constant . ×  C/mol
Index
The index lists only those page numbers when a given term is used or explained for the first time, or
where another aspect of the same term is described

Acceptor – charging, discharging 181, 182

– in semiconductor 229, 230 – classification 183
– in solution 287 Bioanode 218
Accumulation layer in SC 240, 244 Biocathode 214
Activation energy 32, 33 Bioelectrocatalysis 205
Activation polarization 35 see Activation (kinetic) Biofuel cell, BFC 203, 214
overpotential Biosensors 208, 215
Activity 5 Bipolar membranes (BPMs) 63
– catalytic 102, 125, 205 Boltzman distribution function 231
– coefficient of 19 Breathable membrane 301
Advanced oxidation processes (AOPs) 353, 361 Butler-Volmer equation 35, 36
Adsorption 5 see Gibbs’ adsorption isotherm Bode plots 75
Alkaline fuel cells (AFC) 188
Alkaline water electrolysis (AWE) 322 Cadmium chalcogenides electrodes
Alloys 122, 152 (CdX, X-Se, S, Te) 229, 238, 261
– catalyst 122 Capacitance, differential 8
– electrodeposition 152 Capacitive reactance 75
Ammonia 296 Capacitors 8, 11
Amperometric sensors 293, 302 Carbon monoxide 297
– ammonia 296 Catalysts 118, 121
– carbon monoxide 297 Catalysts in oxygen reduction reaction 132
– nitric oxide 294 Cation-exchange membrane (CEM) 356
– oxygen 297 Charging current 65
– sulfur dioxide 296 Chemical power sources see Batteries
Amperometry 65 see Chronoamperometry Colloid systems 357
Anion-exchange membrane (AEM) 356 Composite polymer electrolyte (CPE) 300
Anode 92 Conducting polymer (CP) materials 305
Anodic oxidation 363 Conductivity 40
Aqueous electrolytes 40 – of electrolytes 44
– ion–ion interactions 41 Conductometric sensors 307
– ion–solvent interaction 40 Constant phase element (CPE) 80
Atomic scale 51 Conversion efficiency (SC)
Auxiliary electrodes (AE), also counter electrode – power 265
(CE) 65 – quantum 265
Avrami theorem 145 – solar energy to fuel (STF) 267
– solar energy to hydrogen (STH) 267, 337
Band theory of solid state 229 – solar energy storage 338
– application in semiconductors 230, 232 Chronoamperometry 65
Batteries 175 Corrosion of metals 91
– characteristics 179 – corrosion current 94

https://doi.org/10.1515/9783111160986-019
386 Index

– corrosion macrocell 95 Electrocapillary curve 8

– corrosion microcell 96, 97 Electrocatalysis 118
– corrosion potential 93, 97 – comparison with catalysis 119
– electrochemical 92 Electrocatalytic activity 120
Corrosion of semiconductor 258 – metals 122
– corrosion in dark 258 – alloys 123
– criteria for stability 259, 260 – oxides 124
– effect of light see Photocorrosion – carbons 125
Critical size of nuclei 141 Electrochemical cell 175, 306, 307
– in electrolytic mode 178
Debye‒Hückel limiting equation 42, 43 – in equilibrium 176
Debye‒Hückel‒Onsager (D‒H‒O) theory 47 – in galvanic mode 178
Debye length 14, 42, 241 Electrochemical potential 15
Depletion layer, state (SC) 239 – electrons in metal 16
– capacity 241 – ions in solution 15
– energy diagram 240 Electrochemical power sources see Batteries
– thickness 241 Electrochemical quartz crystal microbalance
Deposit morphology 148 (EQCM) 82
Depth of light penetration (SC) 251 – operating principle 85
Differential capacitance of electrodes – quartz piezoelectric effect 85
– diffuse ionic layer, electrolyte 9 Electrochemical scale of potentials, relation to
– Helmholtz layer 9 energetic scale 237
– space charge layer (SC) 242 Electrochemical sensors 292
– surface states (SC) 255 – amperometric 293
Diffusion – conductometric 307
– coefficient of electrons, holes (SC) 237 – potentiometric 302
– of ions in solution, coefficient of 25 Electrochemical treatment
– layer (region) 28 – heavy metals 353
– length of electrons, holes (SC) 251 – organic pollutants 361
Direct methanol fuel cell 191 Electrocoagulation 357, 360
Dissociation coefficient (constant) 40 Electrocrystallization 139
Donnan equilibrium potential 55, 56, 58 Electrodeposition 138, 354
Donors 229, 287 Electrodialysis 354, 356
– in semiconductor 229, 232 Electro-Fenton oxidation 361
– in solution 287 Electroflotation 357, 360
Double layer capacitors see Supercapacitors Electro-FPs (EFPs) 362, 363
Doped semiconductor (n, p) 232 Electrokinetic (zeta) potential 15, 358
Dye-sensitized solar cells (DSSC) 269 Electrolytes 40, 299
– Grätzel cell 269 – aqueous 40
– light harvesting efficiency 271 – materials, catalysts 300
– operating principle 270 – nonaqueous 48
Dynamic viscosity 47 Electrolytic mobility 57
Electromotive force (EMF) 307
Efficiency of solar energy conversion Electronic nose (e-nose) 311
see Conversion efficiency Electronic tongue (e-tongue) 311
Electrical double layer (EDL) 7 Electronic waste (e-waste) 353
– differential capacitance 8 Electron work function see Work function
– potential distribution 11 Electrophoretic effect 47
– in semiconductors 237, 239 Electrowinning 354
 Index 387

Environmental sensing 291 – standard 237

Enzymes 39 Hydrogen evolution reaction (HER) 135, 319
– artificial 172 Hydrogen oxidation reaction 135
– electroactive 202 Hydrogen-oxygen fuel cell 188
Equilibrium potentials 53 Hydrogen production 313
Exchange current 120, 153, 246 – solar energy, photo water splitting 331
Exciton 233 – water electrolysis 314
External potential 16
Evans diagrams 103 Ideally polarizable electrodes 7
Impedance 70, 73
Faradaic efficiency 139, 320 Incident photon to current efficiency (IPCE) 339
Faradaic resistance 151 Inorganic pollutants 368
Faraday’s law 32, 138 Inorganic solid electrolyte (ISE) 299
Fenton processes (FPs) 361 Interband transitions (SC)
Fermi-Dirac distribution function 231 – direct 233
Fermi energy 231, 232, 244 – indirect 233
Fick’s first law 26 Intrinsic semiconductor 229, 232
Fick’s second law 67 Inversion state, layer 239
Fill factor 265 Ion exchange membranes 58
Fischer‒Tropsch catalytic process 297 Ionic liquids 48
Flat band potential 240 Ionic transference number 57
Flue gas desulfurization 352 Ion–ion interactions 41
Flux density 22 Ion–solvent interaction 40
Formal electrode potential 29, 31
Frumkin adsorption isotherm 128, 129 Kinetic parameters 32, 205
Fuel cells 188 – rate constant 32, 33
– rate of reaction 32
Galvani potential 17 Kohlrausch coefficient 47
Gibbs’ adsorption isotherm 5 Kohlrausch law 44, 45
Gibbs’ energy of reaction 27
Gibbs’ free energy 3, 41 Lambda sensors 306, 307
Gibbs’ free energy of activation 33 Langmuir isotherm 127, 286
Gibbs’ model of interface 4 Lead-acid battery 184
Gibbs surface excess see Gibbs adsorption Leclanché cell 183
isotherm Lewis‒Sargent equation 58
Gouy-Chapman EDL model 10, 42 Lifetime of electrons, holes 235
Grahame EDL model 13 Light absorption (SC) 252
Green fuel see Hydrogen production Light energy conversion see Solar energy
Growth of crystals 139 conversion
Limiting diffusion current 29
Heavy metals 353 Linear scan voltammetry 66, 106
Helmholtz EDL model 9 Lineweaver-Burk plot 207
Helmholtz plane 9 Liquid junction potential 57
Henderson approach 58 Liquid membrane electrodes 304
Heyrovsky reaction 112, 135 Liquid photovoltaic cells 267
Hydroformylation 297 Lithium batteries 186
Hydrogen electrodes – primary 187
– normal 135, 237 – secondary 187
388 Index

Macroscopic scale 53 Overpotential 29, 316

Mass-transport 25, 67 see Fick’s law – activation 35, 318
Majority carriers (SC) 230 – concentration 29, 317
Mediated electrochemical oxidation (MEO) 365 ohmic 317
Membrane electrode assemblies 189 Overpotential deposition (OPD) 157
Membrane fuel cells 189 Overvoltage see Overpotential
Membrane potentials 53 Oxide semiconductors 264, 271, 276, 286
– equilibrium 53 Oxygen concentration 297
– nonequilibrium 56 Oxygen evolution reaction (OER) 319
Michaelis-Menten kinetics 205 Oxygen reduction reaction 132
Microbial cells 329, 331
Microbial electrolysis (MEC) 329 Particles in photocatalysis 280
Minority carriers (SC) 230 – size effect 281
Molar conductivity 44, 45 – processes at 283
Molten salts 50 Photoanodes (SC) 259, 277, 283
Mott-Schottky equation, plot (SC) 241 – in water splitting 273
Photoassisted electrolysis (SC) 273
Nanoparticles 166 Photocatalysis of water 345
Nanostructures 162, 164 Photocatalytic cells 267, 278
Nanozymes 171, 172 see Artificial enzymes Photocatalytic degradation of pollutans 284
Nernst potential 19 Photocatalytic reactions 367, 368
Nernst–Planck equation 24, 56, 60 Photocathodes (SC) 267
Nickel–cadmium battery 185 – in water splitting 274, 276
Nikolsky‒Eisenman equation 61 Photochemical diode 343
Nitric oxide 294 Photocorrosion (SC) 258, 261, 269
Nonaqueous electrolytes 48 Photocurrent (SC) 261
– ionic liquids 48 – Gärtner model 252
– molten salts 50 – surface states influence 257
– solid electrolytes 51 Photodecomposition (SC) see Photocorrosin
Nonequilibrium potentials 56 Photodegradation of pollutants (SC) 284, 367
Nucleation 139 Photoelectrochemical cell (PEC) 332, 344
– analytical approach 144 Photoelectrochemical cells, classification (SC) 267
– instantaneous 143 Photoelectrolysis of water 340
– progressive 143 Photoelectrosynthetic cells (SC) 264, 272
Nucleus growth 139, 148 Photon absorption efficiency 341, 347
Nyquist plot 74 Photopotential (SC) 251, 265
Photoreduction of CO2 277, 287
Ohm’s law 26, 72 Pinning of energy levels (SC) 242
Onsager coefficients 23 Poisson’s equation 10
– coupling coefficients 23 Polarization see Electrode polarization
– direct coefficients 23 Polymer electrolyte membrane electrolysis
– symmetry postulate 24 (PEM) 323
Open-circuit potential (OCP) 91, 251, 303 Polymer membrane electrodes 304
Open-circuit voltage (OPV) decay 257 Potentials
Open-circuit photopotential 265, 267 – chemical 15
Organic pollutants 361, 370 – electrochemical 15
Organic radicals 362 – flat band 240
Outer Helmholtz plane 9, 26 – zero charge 8
 Index 389

Potentiometric sensors 302 Solid oxide fuel cells (SOFC) 191

Pourbaix diagram 98 Solid polymer electrolyte (SPE) 299
Primary batteries 183 Solid polymer fuel cell 189
Protection against photocorrosion (SC) 259, 261 Space charge layer (SC) 238
Pseudocapacitance 160, 195 – accumulation 240
– depletion, inversion 240
Quantum dot-sensitized solar cells 271 – potential drop 240
Quantum yield 254, 265, 271, 285, 338 Stabilization against photocorrosion (SC) 259,
Quasi Fermi levels (SC) 234 260, 261
Standard electrode potential 20, 237
Ragone diagram, plot 182, 197 State functions 3
Randles’ equivalent circuit 78, 81 State parameters 3
Recombination (SC) 235 Stern edl model 11
– radiative 233, 235 Stockholm convention 27
– rate 235 Stoichiometric coefficient, number 178, 196
– surface 256 Sulfur dioxide 296
– trap-assisted 235, 236 Supercapacitors
Regenerative photoelectrochemical cells see Liquid – edl capacitors 195
photovoltaic cells – hybrid capacitors 197
Redox reactions 19 – pseudocapacitors 195
Relaxation time 46 Surface charge 6
Reorganization energy 38 Surface diffusion 140
Resistance Surface excesses see Gibbs’ model of interface
– electrode reaction 36 Surface phase see Gibbs’ model of interface
– polarization 36 Surface potential 17
Reversibility of reactions 27 Surface states (SC) 243, 254
Room-temeprature ionic liquids (RTILs) 50 Surface tension 4

Schottky barrier 348 Tafel equation 318

Seawater electrolysis 326 – plot, slope 36, 108, 121
Secondary batteries 184, 186 Tafel reaction 112, 135
Semiconductor electrodes Temkin adsorption isotherm 128, 129
– electrode reactions 243 Template
– photo effects 250 – active 165
Semiconductor electrodepiosition 154 – restricted 166
Semipermeable membrane 54 Ternary oxides 341
Sensors 208, 215 see Biosensors Thermodynamic activity see Activity
Size effect in catalysis 129 Thermodynamic force 22
Solar energy, photo water splitting 331 Thionyl chloride lithium batteries 187
– photocatalysis of water 345 Throwing power 151
– photoelectrolysis of water 340 Titanium dioxide (SC) in DSS cells 269
– thermodynamics, energy losses and – in photocatalysis 284
efficiency 333 – in water photoelectrolysis 276
Solar energy conversion 262, 264 Transfer coefficient
– efficiency 264, 265 – charge 33
Solid electrolytes 51 – mass 31
Solid oxide electrolysis (SOE) 325 Transient methods 64
390 Index

Transition state 33 see Activation energy, Eyring- – polymer electrolyte membrane electrolysis
Polanyi theory (PEM) 323
– seawater electrolysis 326
Underpotential deposition (UPD) – solid oxide electrolysis (SOE) 325
– catalytic properties 161 – thermodynamics, energy losses and
– definition 157 efficiency 314
– features 158 – thermoneutral potential 316
Water-gas shift reaction 313
Volmer-Butler equation see Butler-Volmer equation Water photoelectrolysis 273
Volmer reaction 112, 136 – electrodes 274
Voltammetry 66, 68 – principle of photo cell operations 273, 274
Volta potential 18 – thermodynamics 273
Water splitting see Water photoelectrolysis
Wagner number 150 Work function 20, 126
Warburg impedance 80 Working electrode (WE) 65
Water electrolysis 314
– alkaline water electrolysis (AWE) 322 Zeta potential see Electrokinetic potential
– microbial electrolysis (MEC) 329 Z scheme in photocatalysis 276, 349