1 \documentclass[10pt]{article}
4 \usepackage{amsmath} % needed for math align
5 \usepackage{bm} % needed for maths bold face
6 \usepackage{graphicx} % needed for including graphics e.g. EPS, PS
7 \usepackage{fancyhdr} % needed for header
11 \topmargin -1.5cm % read Lamport p.163
12 \oddsidemargin -0.04cm % read Lamport p.163
13 \evensidemargin -0.04cm % same as oddsidemargin but for left-hand pages
16 %\pagestyle{empty} % Uncomment if don't want page numbers
17 \parskip 7.2pt % sets spacing between paragraphs
18 %\renewcommand{\baselinestretch}{1.5} % Uncomment for 1.5 spacing between lines
19 \parindent 0pt % sets leading space for paragraphs
22 \newcommand{\vect}[1]{\boldsymbol{#1}} % Draw a vector
23 \newcommand{\divg}[1]{\nabla \cdot #1} % divergence
24 \newcommand{\curl}[1]{\nabla \times #1} % curl
25 \newcommand{\grad}[1]{\nabla #1} %gradient
26 \newcommand{\pd}[3][ ]{\frac{\partial^{#1} #2}{\partial #3^{#1}}} %partial derivative
27 %\newcommand{\d}[3][ ]{\frac{d^{#1} #2}{d #3^{#1}}} %full derivative
28 \newcommand{\phasor}[1]{\tilde{#1}} % make a phasor
29 \newcommand{\laplacian}[1]{\nabla^2 {#1}} % The laplacian operator
34 \definecolor{darkgray}{rgb}{0.95,0.95,0.95}
35 \definecolor{darkred}{rgb}{0.75,0,0}
36 \definecolor{darkblue}{rgb}{0,0,0.75}
37 \definecolor{pink}{rgb}{1,0.5,0.5}
38 \lstset{language=Java}
39 \lstset{backgroundcolor=\color{darkgray}}
40 \lstset{numbers=left, numberstyle=\tiny, stepnumber=1, numbersep=5pt}
41 \lstset{keywordstyle=\color{darkred}\bfseries}
42 \lstset{commentstyle=\color{darkblue}}
43 %\lstset{stringsyle=\color{red}}
44 \lstset{showstringspaces=false}
45 \lstset{basicstyle=\small}
55 \fancyfoot[CO, C]{\thepage}
66 \section*{Overview of VASE: Basic Theory and Applications}
68 \item Reference: J.A Woolam\cite{woolam1}
70 \subsection*{Introduction}
72 \item Mathematical theory is based on Fresnel reflection/transmission equations
74 \item These come from solutions to Maxwell's equations
75 \item p polarised: $\vect{E}$ field parallel to plane of incidence
76 \item s polarised: $\vect{E}$ field perpendicular to plane of incidence (parallel to surface)
78 \item Ellipsometric measurement:
80 \tan(\Psi) e^{i \Delta} = \rho = \frac{r_p}{r_s}
82 \item Spectroscopic Ellipsometry (SE): Measure $\rho$ as a function of $\lambda$
83 \item Variable Angle SE (VASE): Measure $\rho$ as a function of $\lambda$ and angle of incidence.
84 \item Measure \emph{ratio} of 2 values $\implies$ accurate and reproducable
85 \item Measure $\Delta$ (phase quantity), sensitive to presence of thin films
86 \item VASE gathers many data points off a single sample, and is well suited to modelling and fitting
88 \subsection*{What can be Determined by VASE?}
91 \item Surface/interfacial roughness
92 \item Optical constants $\implies$ any parameter that depends on these
93 \item Gradients in properties vs depth in film
94 \item Optical anisotropy
96 \langle\epsilon\rangle &= \langle\epsilon_1\rangle + i \langle\epsilon_2\rangle \\
97 &= \langle\phasor{n}\rangle^2 = \left( \langle n\rangle + i \langle k \rangle \right)^2 \\
98 &= \sin(\phi)^2 \left[ 1 + \tan(\phi)^2 \left(\frac{1 - \rho}{1 + \rho}\right)\right]
102 \subsection*{Data Analysis}
104 \item Ellipsometry doesn't directly measure film parameters; measures $\Psi$ and $\Delta$
105 \item Necessary to perform a model dependent analysis of $\Psi$ and $\Delta$ data
106 \item VASE increases data points recorded $\implies$ good for fitting model to data
110 \item Use assumed model to predict expected data
111 \item Compare generated to experimental data, and adjust model parameters to fit
113 \item Fitting algorithm - Marquardt-Levenberg
114 \item Objective - minimise (root) Mean Squared Error (MSE)
116 \text{MSE} &= \sqrt{\frac{1}{2N - M} \displaystyle\sum_{i=1}^N \left[ \left(\frac{\Psi_i^{\text{Mod}} - \Psi_i^{\text{Exp}}}{\sigma_{\Psi_i}^{\text{Exp}}} \right)^2 + \left(\frac{\Delta_i^{\text{Mod}} - \Delta_i^{\text{Exp}}}{\sigma_{\Delta_i}^{\text{Exp}}} \right)^2\right]}
118 \item Iterative process; start with simple model and refine
119 \item ``...the danger in making the model more complex is that paramters become correlated, in which case multiple sets of paramters will give the same good MSE fit.''
122 \subsection*{Considerations for VASE Analysis}
124 \item Initial guesses must be close to the actual value
125 \item This way the final MSE is the actual best fit, not a local minima as shown in \ref{MSE-woolam1}
129 \includegraphics[scale=0.80]{MSE-woolam1.png}
130 \captionof{figure}{From \cite{woolam1}}
134 \subsection*{Building Optical Model}
136 \item Must have enough flexibility to accurately fit the experimental data
137 \item Example: Surface roughness added to model increases accuracy of fit
140 \section*{Harris 1953}
142 \item Reference \cite{harris53}
143 \item Measure conductivity of gold-blacks from reflection & transmission measurements in far IR
144 \item Conductivity depends on wavelength
146 \item ``Condenser Effect''
148 \item Structure of material; yarn like metal strands
149 \item Gaps in metal strands act as condensers
150 \item ``optical conductivity'' is actually an admittivity
151 \item As $f$ of radiation increases, more strands capable of conducting current
152 \item $\implies$ conductivity decreases as $\lambda$ increases
154 \item ``Relaxation Effect''
156 \item Finite relaxation time of Electrons (Reference Drude and Zener)
157 \item Electrons lag behind imposed EMF. Lag increases as $f$ increases.
158 \item $\implies$ conductivity increases as $\lambda$ increases
160 \item At Resonance frequencies, optical absorbtivity (and hence conductivity) passes through a maximum.
162 \item For metal blacks, the effects of the 3 factors dominate in different regions of $\lambda$
164 \item Resonance $f$ lie in visible and near IR regions
165 \item $\lambda > 100$microns
167 \item Relaxation effect is the dominant effect considered in this paper.
171 \bibliographystyle{plain}