--- /dev/null
+\chapter{Overview} \label{chapter_overview}
+
+\section{Black Metal Films}
+
+So called black metal films\footnote{Naming conventions vary in the literature} are the result of deposition of metal elements at a relatively high pressure\footnote{of the order of $10^{-2}$ mbar} or ``bad vacuum''. The films are named due to their high absorbance at visible wavelengths; a sufficiently thick film will appear black to the naked eye. There is a remarkable contrast between such films and metal films deposited under low pressure\footnote{less than $10^{-5}$mbar}, which are typically highly reflective and brightly coloured at comparable thicknesses. It has been established that Black metal films may be prepared in any gas, but when oxygen is present the resulting films contain tungsten oxides due to the use of tungsten heating filaments for the deposition of films\cite{harris1952} \cite{mckenzie2006}.
+
+% First mentions and early research; Pfund
+The formation of black-metal films at high pressure has been known since the early 20th century, with the first papers on the subject published by Pfund in the 1930s, motivated by the potential application of black-metal films to radiometetric devices \cite{pfund1930}, \cite{pfund1933}. Pfund established the conditions for formation of black-metals and showed that the transmission spectrum of metallic black films is almost zero in visible wavelengths, but increases to a plateau in the far infrared. Subsequent researchers have also focused on measuring the properties of black-metal films as a function of preparation conditions, with the aim of producing selective filters for infra-red detectors. \cite{harris1948} \cite{harris1952}. More recently, it was shown that black-Au coatings can be used to increase the efficiency of thin film solar cells \cite{panjwani2011}.
+
+There have been several attempts to relate the structure of black-metal films to measured optical and electrical properties. Metallic nanostructured films deposited at low pressures generally consist of a series of metallic islets; for such films, Mie theories of scattering are often used. However, it is clear that black-metal films usually consist of a highly irregular strandlike structure; Mie theories fail to accurately predict the optical properties of such films \cite{mckenzie2006}.
+
+Harris et al. have produced experimental results of the transmission of black-metal films from visible wavelengths to the far-infrared. By modelling black-metal films as consisting of layers of ``yarn like'' metal strands, Harris et al. have arrived at an expression for the electron relaxation time of Au-black, leading to a calculated transmission spectrum in good agreement with experimental results \cite{harris1953}.
+
+
+% Artificially ``blackened'' thin films
+The optical properties of black-metal films have been found to vary due to changes in structure when the film is exposed to atmosphere, or is heated \cite{advena1993}. This ``degradation'' of the black-metal films is inconvenient for maintaining consistent calibration of devices. Recently there has been interest in artificial ``blackening'' of metal surfaces for optical applications. These ``meta-materials'' offer a promising alternative to the traditional black-metal films, due to the ability to more precisely control properties of the film. In particular, artifially blackened metal films have been produced which suppress reflection of light via surface plasmon resonances \cite{sondergaard2012}.
+
+In light of the wealth of previous research, the aims of this project were first to reproduce black-metal films using equipment available in the CAMSP research group, and then employ several techniques for the study of these samples in comparison with other metal films. Most of the existing research has been conducted using optical transmission or reflection spectroscopy techniques. A total current secondary electron spectroscopy experiment was therefore integrated into the sample preparation vacuum chamber, to allow for almost immediate study of prepared samples using this technique. A Variable Angle Spectroscopic Ellipsometer (VASE) has been used in an attempt to relate the optical and structural properties of the samples. We have also used optical transmission spectroscopy at visible wavelengths to characterise the transmission of black-metal films.
+
+
+\section{Plasmonics}
+
+Since optical properties of nanostructured films are often determined by plasmon-photon interaction, a brief description will be given of plasmonic behaviour in such films. Generally the optical properties of a metal are modelled using a combination of classical and quantum theories. For noble metals (eg: Au and Ag), conduction band electrons are well described by free electron gas models \cite{oates2011}.
+
+\subsection{Bulk Plasmons}
+
+In the Lorentz model, electrons responding to incident light are modelled as non-interacting, damped and forced harmonic oscillators. For each electron:
+\begin{align}
+ m \frac{d^2 \vect{x}}{dt^2} &= -m_e \Gamma \der{\vect{x}}{t} - m_e \omega_0^2 \vect{x} - e \vect{E_0} e^{-i\omega t} \label{lorentz}
+\end{align}
+where $m$ is the electron mass, $\vect{x}$ is the displacement of the electron from equilibrium, $\Gamma$ is the damping constant, $-m \omega_0^2 \vect{x}$ is the restoring force, and $-e \vect{E_0} e^{-i\omega t}$ is the forcing term due to an incident plane wave of frequency $\omega$.
+
+The polarisation density $\vect{P}$\footnote{Not to be confused with the polarisation state of light} of the electron gas describes the material's response to the incident light. It can be writen as:
+\begin{align*}
+ \vect{P} &= \phasor{\epsilon} \vect{E} = -e N \vect{x}(t)
+\end{align*}
+Where $\phasor{\epsilon}$ is the psuedo-dielectric constant and $N$ is the number of electrons per unit volume.
+
+Solving \label{lorentz} for $\vect{x}(t)$ gives the Lorentz Oscillator expression:
+\begin{align*}
+ \phasor{\epsilon}(\omega) &= 1 + \frac{e^2 N}{\epsilon_0 m} \frac{1}{\omega_0^2 - \omega^2 - i \Gamma \omega}
+\end{align*}
+
+In the noble metals (Au, Ag, etc), conduction electrons are extremely loosely bound. The Drude model, which neglects the restoring force, is a good model for electron behaviour in these metals. In this case, $\omega_0 = 0$, and:
+\begin{align*}
+ \phasor{\epsilon}(\omega) &= 1 - \frac{\omega_p^2}{\omega^2 + \Gamma^2} + i \frac{\omega_p^2 \Gamma}{\omega(\omega^2 + \Gamma^2)}
+\end{align*}
+where the Drude Plasma frequency $\omega_p$ is:
+\begin{align*}
+ \omega_p &= \sqrt{\frac{e^2 N}{\epsilon_0 m}}
+\end{align*}
+
+If $\Gamma$ is small, (as for the noble metals), then as $\omega$ approaches $\omega_{p}$, the real part of $\phasor{\epsilon}$ approaches zero, and longitudinal oscillating waves may propagate through the material \cite{oates2011}. These oscillations are called ``Bulk'' or ``Volume'' plasmons. They were first excited in Electron Energy Loss Spectroscopy (EELS) experiments by Ferrel in 1958 \cite{ferrel1958}.
+
+In real metals, interband transitions contribute to $\phasor{\epsilon}$. The ``screened'' plasmon frequency is given by:
+\begin{align}
+ \omega_{sp}^2 &= \frac{\omega_p^2}{\epsilon_\infty} - \Gamma^2 \label{screened_plasmon}
+\end{align}
+where $\epsilon_\infty$ is a constant representing the real part of $\phasor{\epsilon}$ with the contribution of interband transitions.
+The accuracy of this formula decreases as $\hbar \omega$ approaches the energy for interband transitions. \cite{oates2011}.
+
+From \eqref{screened_plasmon}, it can be seen that interband transitions effectively reduce the frequency at which bulk plasmons are excited. For example, in Ag $\epsilon_\infty \approx 4$ and $\hbar \omega_{sp} = 2.8$ eV. Observed values for the bulk plasmon frequency in Ag are typically $\hbar \omega = 3.8\text{ eV}$ \cite{kittel}. Ag's relatively low plasmon threshold energy makes it a material of choice for plasmonic meta-material applications \cite{oates2011}.
+
+
+
+\subsection{Surface Plasmons}
+
+One of the most important developments leading to the modern field of plasmonics has been the prediction of so called Surface Plasmons (SP) in 1957 by Ritchie \cite{ritchie1957}. In his paper, Ritchie predicted surface plasmons as the mechanism for energy loss of electrons passing through a thin film, building on previous work by Pines and Bohm \cite{bohm1951} \cite{bohm1952} \cite{bohm1953}. Ritchie's prediction was experimentally verified here at UWA by Cedric Powel and John Swan just three years later \cite{powel1959}.
+
+A coupled Surface Plasmon and Photon are referred to as a Surface Plasmon Polariton (SPP). The behaviour of SPPs can be derived by solving Maxwell's equations for $p$ (parallel to the plane of incidence) and $s$ (parallel to the surface) polarised light at the interface between a metal and a dielectric \cite{pitark2007}.
+
+The Surface plasmon frequency is given by:
+\begin{align*}
+ \omega_{s} &= \frac{\omega_p}{\sqrt{1 + \epsilon_a}}
+\end{align*}
+where $\epsilon_a$ is the dielectric constant for the dielectric. For air, $\epsilon_a = 1$ and $\omega_{s} = \frac{\omega_p}{\sqrt{2}}$
+
+Theoretically, SPPs can only couple to $p$ polarised light. This makes Ellipsometry, which measures the change in polarisation of reflected light, a useful technique for detecting excitation of SPPs \cite{oates2011}.
+
+In order to excite an SPP in a flat surface, some form of ``third body'' is required to match the wavevector of the photon with the plasmon. The Kretchsman configuration is one possible method, in which light is totally internally reflected at the interface between a glass prism and one side of the metal film. The evanescent fields created at the interface between a thin film and the glass may lead to the excitation of a SPP on the other side of the thin film.
+
+\begin{figure}[H]
+ \centering
+ \includegraphics[width=0.6\textwidth]{figures/plasmon/spp.png}
+ \caption{Illustration of the charge distribution and electric fields of a surface plasmon polariton taken from Oates et al. \cite{oates2011}}
+\end{figure}
+
+\subsection{Surface Plasmon Resonances}
+
+Due to different boundary conditions, a nanostructured surface exhibits very different plasmonic effects to a smooth surface.
+Although the term ``plasmon'' was not coined until the 1950s, Mie's solution for scattering of light from spherical nanoparticles does incorporate the Drude model, and predicts resonance of the free electron gas inside the metal spheres. These are now refered to as Mie plasmons, or surface plasmon resonances (SPR) \cite{pitark2007,oates2011}. The frequency of these resonances is:
+\begin{align}
+ \omega_{pp}^2 &= \frac{\omega_p^2}{\epsilon_\infty + 2 \epsilon_a} - \Gamma^2
+\end{align}
+For a pure Drude metal, this reduces to $\omega_{pp} = \frac{\omega_p}{\sqrt{3}}$. Light of frequency $\omega_{pp}$ is preferentially scattered.
git.ucc.asn.au / matches / honours.git / blobdiff
UCC git Repository :: git.ucc.asn.au