Session 4A: Advanced Topics

James N. Hilfiker
University of Toledo
August 20, 2009
 Advanced Grading

ƒ Optical properties can vary with depth into a film.

ƒ Model this behavior by dividing film into ‘slices’
where optical constants can vary between each.
14 270 14 270
Model Fit Model Fit
Exp Ψ -E 65° Exp Ψ -E 65°

Ψin degrees
Exp Δ -E 65° 180 Exp Δ -E 65° 180
Ψin degrees

12 12

Δin degrees

Δin degrees
90 90

10 10
0 0
(A) (B)
8 -90 8 -90
400 600 800 1000 1200
Wavelength (nm)

TiO 2 Graded TiO2 2

BK7 BK7

2 © 2009 J.A. Woollam Co., All Rights Reserved

 Function-based Grade

ƒ Any “fit” parameter can be graded.

– Examples: Cauchy An parameter, Oscillator
amplitude, Drude resistivity term

Tauc-Lorentz model for TiO2 Grade oscillator amplitude

3 © 2009 J.A. Woollam Co., All Rights Reserved

 Function-based Grade

ƒ Grading oscillator amplitude varies index

over all wavelengths and UV absorption.
Graded Optical Constants
2.8 1.2

2.6 1.0

Extinction Coefficient ' k'

Index of refraction ' n'

n top
2.4 n bottom
k top
0.8
2.2 k bottom
0.6
2.0
0.4
1.8

1.6 0.2

1.4 0.0
0 300 600 900 1200 1500 1800
Wavelength (nm)

4 © 2009 J.A. Woollam Co., All Rights Reserved

 Function-based Grade

ƒ More complex
profiles available
1. “Nodes” (next slide)
2. Parameter Grading

– Example of a
Gaussian
distribution for free
carrier density
(Sb implanted Ge)

5 © 2009 J.A. Woollam Co., All Rights Reserved

 Function-based Grade

ƒ “Nodes” to describe grading profile

Depth Profile of Optical Constants at 500nm
2.6 0.10

2.4 n

Extinction Coefficient ' k'

k 0.08

Index of refraction ' n'

2.2
0.06
2.0
0.04
1.8

0.02
1.6

1.4 0.00
-200 0 200 400 600 800
Distance from Substrate in Å

Node 1 Node 2 Node 3

Interface (61.5%) surface
(0%) (100%)

6 © 2009 J.A. Woollam Co., All Rights Reserved

 ITO Example: No Grading

•ITO: No Grading
Generated and Experimental
40
Model Fit
Exp E 45°
Exp E 50°
30
2 srough 5.3383 nm Exp E 55°
Exp E 60°

Ψ in degrees
Exp E 65°
1 ito genosc 61.85 nm 20
Exp E 70°

0 sl glass genosc 1.1 mm

10

0
0 300 600 900 1200 1500 1800

ito genosc Optical Constants

MSE=9.6 Wavelength (nm)

Generated and Experimental

300
2.5 1.0 Model Fit
Exp E 45°
n
Exp E 50°
k
Extinction Coefficient ' k'
2.0 0.8 Exp E 55°
200
Index of refraction ' n'

Exp E 60°
Exp E 65°

Δ in degrees
Exp E 70°
1.5 0.6
100

1.0 0.4
0
0.5 0.2

0.0 0.0 -100

0 300 600 900 1200 1500 1800 0 300 600 900 1200 1500 1800
Wavelength (nm) Wavelength (nm)

7 © 2009 J.A. Woollam Co., All Rights Reserved

 2-node Function-Based Grading

•ITO: Function-based Generated and Experimental

Grading 2-nodes
40
Model Fit
Exp E 45°
Exp E 50°
30 Exp E 55°
Exp E 60°

Ψ in degrees
Exp E 65°
Exp E 70°
3 srough 5.6959 nm 20

2 graded2 (ito genosc) 62.018 nm

10
1 ito genosc 0 nm
0 sl glass genosc 1.1 mm 0
0 300 600 900 1200 1500 1800
Wavelength (nm)
•MSE=4.97
2x improvement
graded2 Optical Constants Generated and Experimental
2.5 1.2 300
Model Fit
Exp E 45°
n top Exp E 50°
1.0
Extinction Coefficient ' k'

2.0 n bottom Exp E 55°

Index of refraction ' n'

k top 200 Exp E 60°

k bottom 0.8 Exp E 65°

Δ in degrees
1.5 Exp E 70°
0.6 100
1.0
0.4
0
0.5
0.2

0.0 0.0 -100

0 300 600 900 1200 1500 1800 0 300 600 900 1200 1500 1800
Wavelength (nm) Wavelength (nm)

8 © 2009 J.A. Woollam Co., All Rights Reserved

 3-node Function-Based Grading

ITO: Function-based Grading Generated and Experimental

40
3-nodes Model Fit
Exp E 45°
Exp E 50°
30 Exp E 55°
Exp E 60°

Ψ in degrees
Exp E 65°
3 srough 5.8724 nm Exp E 70°
20
2 graded2 (ito genosc) 62.075 nm
1 ito genosc 0 nm 10

0 sl glass genosc 1.1 mm

0
0 300 600 900 1200 1500 1800
Wavelength (nm)

•MSE=3.27
& improved near-IR
graded2 Optical Constants Generated and Experimental
2.5 1.2 300 Model Fit
Exp E 45°
n top Exp E 50°
1.0
Extinction Coefficient ' k'

2.0 n bottom Exp E 55°

Index of refraction ' n'

k top 200 Exp E 60°

k bottom Exp E 65°
0.8

Δ in degrees
1.5 Exp E 70°

0.6 100
1.0
0.4
0
0.5
0.2

0.0 0.0 -100

0 300 600 900 1200 1500 1800 0 300 600 900 1200 1500 1800
Wavelength (nm) Wavelength (nm)

9 © 2009 J.A. Woollam Co., All Rights Reserved

 3-node Function-Based Grading

• ITO: Function-based Grading

•resistivity vs. depth (Ω-cm) •n & k vs. depth

•Node 1
•Node 2 •Node 3

10 © 2009 J.A. Woollam Co., All Rights Reserved

 DEMONSTRATION

ƒ ITO on Glass (#5).

– Use Oscillator Model for ITO.
– Add Function-based Grading.
– Add extra nodes.
(when does MSE improve)

•11 © 2009 J.A. Woollam Co., All Rights Reserved

 ITO Example

ƒ “ITO-6.dat”

ITO conductivity changes with depth due to

annealing process. Try to grade the Drude
absorption in Near Infrared.

•12 © 2009 J.A. Woollam Co., All Rights Reserved

 ITO #5, Result
•MSE=3.2 Graded Optical Constants
2.5 2.5
2 Graded (ito_l2) 100.487 nm
n top
1 ito_l2 0.000 nm n bottom

Extinction Coefficient ' k'

2.0 k top 2.0

Index of refraction ' n'

0 7059_u 1 mm k bottom

Generated and Experimental 1.5 1.5

50
1.0 1.0
40
Ψ in degrees

30
0.5 0.5
Model Fit
Exp E 45°
Exp E 50°
20 Exp E 55° 0.0 0.0
Exp E 60° 300 600 900 1200 1500 1800
Exp E 65°
10 Exp E 70° Wavelength (nm)
Exp E 75°
Exp E 80°
0 Depth Profile of Optical Constants at 1500nm
300 600 900 1200 1500 1800
1.8 1.8
Wavelength (nm)
Generated and Experimental n
1.5 k 1.5

Extinction Coefficient ' k'

300

Index of refraction ' n'

Model Fit
Exp E 45°
Exp E 50°
1.2 1.2
200 Exp E 55°
Exp E 60°
0.9 0.9
Δ in degrees

Exp E 65°
Exp E 70°
100 Exp E 75°
Exp E 80° 0.6 0.6

0 0.3 0.3

0.0 0.0
-100 -20 0 20 40 60 80 100 120
300 600 900 1200 1500 1800
Wavelength (nm) Distance from Substrate in nm

13 © 2009 J.A. Woollam Co., All Rights Reserved

 ITO #6, Result
•MSE=7 Graded Optical Constants
2.5 2.5
2 Graded (ito_l2) 196.191 nm

Extinction Coefficient ' k'

1 ito_l2 0.000 nm 2.0 2.0

Index of refraction ' n'

0 7059_u 1 mm
1.5 1.5
Generated and Experimental
50 n top
1.0 n bottom 1.0
k top
40 k bottom
0.5 0.5
Ψ in degrees

30
Model Fit
Exp Eb 45.2° 0.0 0.0
20 Exp Eb 60.3° 300 600 900 1200 1500 1800
Exp Eb 75.5°
Wavelength (nm)
10 Depth Profile of Optical Constants at 1500nm
2.1 1.8
0
300 600 900 1200 1500 1800 1.8 1.5

Extinction Coefficient ' k'

Wavelength (nm)

Index of refraction ' n'

Generated and Experimental 1.5
300
1.2
1.2
0.9
200 0.9 n
k
0.6
Δ in degrees

0.6
100
0.3 0.3

0 Model Fit 0.0 0.0

Exp Eb 45.2° -50 0 50 100 150 200
Exp Eb 60.3°
Exp Eb 75.5° Distance from Substrate in nm
-100
300 600 900 1200 1500 1800
Wavelength (nm)

14 © 2009 J.A. Woollam Co., All Rights Reserved

 Function-Based Grading

ƒ Use when model parameter is varying

from bottom to top of film
ƒ Two most common applications:
– Conductive Oxides (ITO)
Grade the Drude absorption
– Liquid Crystals
Grade the Orientation (next section)

© 2009 J.A. Woollam Co., All Rights Reserved

 ANISOTROPY - 2

ƒ When to use g-SE?

ƒ Euler Angles
ƒ Anisotropic Substrates
ƒ Liquid Crystals

© 2009 J.A. Woollam Co., All Rights Reserved

 When is g-SE needed?
ƒ Anisotropic materials
“may” cause cross-
polarization between p-
and s-.
ƒ Off-diagonal elements
only non-zero when
optical axis misaligned
with ellipsometer.

Figure from:
Spectroscopic Ellipsometry:
Principles and Applications
by Hiroyuki Fujiwara (2003)

17 © 2009 J.A. Woollam Co., All Rights Reserved

 Traditional Ellipsometry measures...

⎛ rpp ⎞ i (δ −δ ) E prefl E inp rpp

( )
ρ = tan Ψ e = tan⎜⎜ ⎟⎟ e p s = refl in =
iΔ

⎝ rss ⎠ Es Es rss

Assumes no cross-coupling of p- and s- polarized light in the sample,

that is, sample is a diagonal Jones Matrix:

⎛ E pout ⎞ ⎛ E inp ⎞ ⎛ rpp 0 ⎞ ⎛ E inp ⎞ Standard

⎜ out ⎟ = J⎜ in ⎟ = ⎜ ⎟ ⎜ in ⎟ Ellipsometry
assumes
⎝ Es ⎠ ⎝ Es ⎠ ⎝ 0 rss ⎠ ⎝ Es ⎠ cross-terms = 0

This holds for specular isotropic samples, and certain classes of

anisotropic samples, when the optical axis is oriented along
axes of symmetry

18 © 2009 J.A. Woollam Co., All Rights Reserved

 Generalized Ellipsometry ...

ƒ Does not assume zero off-diagonal components

for the Jones matrix. Uses the following
relationship…
⎡ pout ⎤ ⎡rpp rsp ⎤ ⎡ pin ⎤ Generalized
⎢ s ⎥ = ⎢r ⎥⋅⎢ ⎥ Ellipsometry allows
⎣ out ⎦ ⎢⎣ ps ss ⎥⎦ ⎣ sin ⎦
r cross-terms ≠ 0
ƒ Thus, a generalized sample is described by…

iΔ r pp
AnE = tan(Ψ) ⋅ e = rss
iΔ ps r ps
Aps = tan(Ψ ps ) ⋅ e =
r pp

i Δ sp rsp
Asp = tan( Ψ sp ) ⋅ e =
rss
19 © 2009 J.A. Woollam Co., All Rights Reserved
 Euler Angles

ƒ The Euler Angles (φ, θ, ψ ) define the orientation of

SAMPLE DIELECTRIC TENSOR relative to the
LABORATORY Frame of reference.

Laboratory
Laboratoryframe
frameof
ofreference
reference
isisdefined
definedby:
by:
Plane of
••sample
samplesurface
surface(x-y
(x-yplane)
plane) incidence
••plane
planeofofincidence
incidence(x-z
(x-zplane)
plane)
x

y z
Shown: sample

(φ, θ, ψ ) = (0°, 0°, 0°) surface ε11 direction

ε22 direction
ε33 direction

20 © 2009 J.A. Woollam Co., All Rights Reserved

 Euler Angles (φ, θ, ψ )
2) θ : rotate y’,z’ about x’
ƒ 3 Euler Angle steps:
sample
surface
x ε22 direction ε33 direction

y ε11 direction
1) φ : rotate x,y about z
z

θ z”
sample y’ x’
surface
x ε22
y” z’
direction
y z
φ ε11 direction
3) ψ : rotate x”,y” about z”
ε33
direction
y’ x’ sample ε33
surface direction
x ε11 direction
z ε22 direction
y z

ψ x”
x’’’
y’’’ z’’’
y”

21 © 2009 J.A. Woollam Co., All Rights Reserved

 Anisotropic Substrates

ƒ Anisotropy can produce two refracted beams

that remain ‘coherent’ with each other, but travel
different path lengths.
incoherent

coherent

p- & s- E-field
components

Refracted beam splits

into two beams

© 2009 J.A. Woollam Co., All Rights Reserved

 Birefringent Substrate

ƒ Interference between p- and s- waves from

backside of substrate are coherent and cause
“oscillations” in data, dependent on path length
and index-difference (Δn)
30.0

28.0 Birefringent SubstrateΔ( n=0.005)

Isotropic Substrate
Ψ in degrees

26.0

24.0

22.0

20.0
400 500 600 700 800 900 1000
Wavelength (nm)

© 2009 J.A. Woollam Co., All Rights Reserved

 Handling Substrates

1. Roughen backside (minimizes out-of-

plane)

2. Align with “specific” direction.

3. Measure Pseudo-N
4. Save Optical Constants as reference for
future measurements with same
alignment.

OR
Complete Characterization of substrate.
Come to our Advanced Course
24 © 2009 J.A. Woollam Co., All Rights Reserved
 Data Analysis Review

ƒ Transparent: Cauchy
ƒ Absorbing: Point-by-Point Fit
GENOSC
SE + Transmission
ƒ Microstructure: EMA
Roughness
Grading

25 © 2009 J.A. Woollam Co., All Rights Reserved

 Substrate Modeling Summary
Bare Substrate

Dielectric
Semiconductor Metal
Crystalline values Opacity
Surface Oxide Direct Fit
Isotropic Anisotropic Ignore oxide/
•Cauchy/ Sellmeier Generalized SE roughness
•Backside BIAXIAL layer
•Roughness
•Transmission

© 2009 J.A. Woollam Co., All Rights Reserved

 Thin Film Interference

ƒ Each reflected wave has phase and

amplitude.

n~0

n~1

n~2

•27 © 2009 J.A. Woollam Co., All Rights Reserved

 Thin Film Modeling
Thin Film

Transparent Semi-Absorbing Absorbing

Cauchy 3 Step Approach 9 Methods
Add Roughness, 1. Thickness (transparent λs)
Grading, or
2. Pt-by-Pt “n,k”
Anisotropy
3. GENOSC

© 2009 J.A. Woollam Co., All Rights Reserved

 Course Overview

ƒ Theory and Introduction

ƒ Substrates
ƒ Transparent Films
– Cauchy and Sellmeier
– Microstructure (Roughness, Grading)
– Simple Anisotropy
ƒ Multilayers
ƒ UV Absorption
ƒ Absorbing Films
ƒ Genosc Layer
ƒ Depolarization Data
ƒ Advanced Topics
– Function-based Grading
– Anisotropy
ƒ Review

•29 © 2009 J.A. Woollam Co., All Rights Reserved