| 1 |
%!TEX root = /Users/charles/Desktop/nanoglass/nanoglass.tex |
| 2 |
|
| 3 |
\section{Computational Methodology} |
| 4 |
\label{sec:details} |
| 5 |
|
| 6 |
\subsection{Initial Geometries and Heating} |
| 7 |
|
| 8 |
Cu-core / Ag-shell and random alloy structures were constructed on an |
| 9 |
underlying FCC lattice (4.09 {\AA}) at the bulk eutectic composition |
| 10 |
$\mathrm{Ag}_6\mathrm{Cu}_4$. Both initial geometries were considered |
| 11 |
although experimental results suggest that the random structure is the |
| 12 |
most likely structure to be found following |
| 13 |
synthesis.\cite{Jiang:2005lr,gonzalo:5163} Three different sizes of |
| 14 |
nanoparticles corresponding to a 20 \AA radius (1961 atoms), 30 {\AA} |
| 15 |
radius (6603 atoms) and 40 {\AA} radius (15683 atoms) were |
| 16 |
constructed. These initial structures were relaxed to their |
| 17 |
equilibrium structures at 20 K for 20 ps and again at 300 K for 100 ps |
| 18 |
sampling from a Maxwell-Boltzmann distribution at each temperature. |
| 19 |
|
| 20 |
To mimic the effects of the heating due to laser irradiation, the |
| 21 |
particles were allowed to melt by sampling velocities from the Maxwell |
| 22 |
Boltzmann distribution at a temperature of 900 K. The particles were |
| 23 |
run under microcanonical simulation conditions for 1 ns of simualtion |
| 24 |
time prior to studying the effects of heat transfer to the solvent. |
| 25 |
In all cases, center of mass translational and rotational motion of |
| 26 |
the particles were set to zero before any data collection was |
| 27 |
undertaken. Structural features (pair distribution functions) were |
| 28 |
used to verify that the particles were indeed liquid droplets before |
| 29 |
cooling simulations took place. |
| 30 |
|
| 31 |
\subsection{Modeling random alloy and core shell particles in solution |
| 32 |
phase environments} |
| 33 |
|
| 34 |
To approximate the effects of rapid heat transfer to the solvent |
| 35 |
following a heating at the plasmon resonance, we utilized a |
| 36 |
methodology in which atoms contained in the outer $4$ {\AA} radius of |
| 37 |
the nanoparticle evolved under Langevin Dynamics, |
| 38 |
\begin{equation} |
| 39 |
m \frac{\partial^2 \vec{x}}{\partial t^2} = F_\textrm{sys}(\vec{x}(t)) |
| 40 |
- 6 \pi a \eta \vec{v}(t) + F_\textrm{ran} |
| 41 |
\label{eq:langevin} |
| 42 |
\end{equation} |
| 43 |
with a solvent friction ($\eta$) approximating the contribution from |
| 44 |
the solvent and capping agent. Atoms located in the interior of the |
| 45 |
nanoparticle evolved under Newtonian dynamics. The set-up of our |
| 46 |
simulations is nearly identical with the ``stochastic boundary |
| 47 |
molecular dynamics'' ({\sc sbmd}) method that has seen wide use in the |
| 48 |
protein simulation |
| 49 |
community.\cite{BROOKS:1985kx,BROOKS:1983uq,BRUNGER:1984fj} A sketch |
| 50 |
of this setup can be found in Fig. \ref{fig:langevinSketch}. In |
| 51 |
equation \ref{eq:langevin} the frictional forces of a spherical atom |
| 52 |
of radius $a$ depend on the solvent viscosity. The random forces are |
| 53 |
usually taken as gaussian random variables with zero mean and a |
| 54 |
variance tied to the solvent viscosity and temperature, |
| 55 |
\begin{equation} |
| 56 |
\langle F_\textrm{ran}(t) \cdot F_\textrm{ran} (t') |
| 57 |
\rangle = 2 k_B T (6 \pi \eta a) \delta(t - t') |
| 58 |
\label{eq:stochastic} |
| 59 |
\end{equation} |
| 60 |
Due to the presence of the capping agent and the lack of details about |
| 61 |
the atomic-scale interactions between the metallic atoms and the |
| 62 |
solvent, the effective viscosity is a essentially a free parameter |
| 63 |
that must be tuned to give experimentally relevant simulations. |
| 64 |
\begin{figure}[htbp] |
| 65 |
\centering |
| 66 |
\includegraphics[width=5in]{images/stochbound.pdf} |
| 67 |
\caption{Methodology used to mimic the experimental cooling conditions |
| 68 |
of a hot nanoparticle surrounded by a solvent. Atoms in the core of |
| 69 |
the particle evolved under Newtonian dynamics, while atoms that were |
| 70 |
in the outer skin of the particle evolved under Langevin dynamics. |
| 71 |
The radius of the spherical region operating under Newtonian dynamics, |
| 72 |
$r_\textrm{Newton}$ was set to be 4 {\AA} smaller than the original |
| 73 |
radius ($R$) of the liquid droplet.} |
| 74 |
\label{fig:langevinSketch} |
| 75 |
\end{figure} |
| 76 |
|
| 77 |
The viscosity ($\eta$) can be tuned by comparing the cooling rate that |
| 78 |
a set of nanoparticles experience with the known cooling rates for |
| 79 |
similar particles obtained via the laser heating experiments. |
| 80 |
Essentially, we tune the solvent viscosity until the thermal decay |
| 81 |
profile matches a heat-transfer model using reasonable values for the |
| 82 |
interfacial conductance and the thermal conductivity of the solvent. |
| 83 |
|
| 84 |
Cooling rates for the experimentally-observed nanoparticles were |
| 85 |
calculated from the heat transfer equations for a spherical particle |
| 86 |
embedded in a ambient medium that allows for diffusive heat transport. |
| 87 |
Following Plech {\it et al.},\cite{plech:195423} we use a heat |
| 88 |
transfer model that consists of two coupled differential equations |
| 89 |
in the Laplace domain, |
| 90 |
\begin{eqnarray} |
| 91 |
Mc_{P}\cdot(s\cdot T_{p}(s)-T_{0})+4\pi R^{2} G\cdot(T_{p}(s)-T_{f}(r=R,s)=0\\ |
| 92 |
\left(\frac{\partial}{\partial r} T_{f}(r,s)\right)_{r=R} + |
| 93 |
\frac{G}{K}(T_{p}(s)-T_{f}(r,s) = 0 |
| 94 |
\label{eq:heateqn} |
| 95 |
\end{eqnarray} |
| 96 |
where $s$ is the time-conjugate variable in Laplace space. The |
| 97 |
variables in these equations describe a nanoparticle of radius $R$, |
| 98 |
mass $M$, and specific heat $c_{p}$ at an initial temperature |
| 99 |
$T_0$. The surrounding solvent has a thermal profile $T_f(r,t)$, |
| 100 |
thermal conductivity $K$, density $\rho$, and specific heat $c$. $G$ |
| 101 |
is the interfacial conductance between the nanoparticle and the |
| 102 |
surrounding solvent, and contains information about heat transfer to |
| 103 |
the capping agent as well as the direct metal-to-solvent heat loss. |
| 104 |
The temperature of the nanoparticle as a function of time can then |
| 105 |
obtained by the inverse Laplace transform, |
| 106 |
\begin{equation} |
| 107 |
T_{p}(t)=\frac{2 k R^2 g^2 |
| 108 |
T_0}{\pi}\int_{0}^{\infty}\frac{\exp(-\kappa u^2 |
| 109 |
t/R^2)u^2}{(u^2(1 + R g) - k R g)^2+(u^3 - k R g u)^2}\mathrm{d}u. |
| 110 |
\label{eq:laplacetransform} |
| 111 |
\end{equation} |
| 112 |
For simplicity, we have introduced the thermal diffusivity $\kappa = |
| 113 |
K/(\rho c)$, and defined $k=4\pi R^3 \rho c /(M c_p)$ and $g = G/K$ in |
| 114 |
Eq. \ref{eq:laplacetransform}. |
| 115 |
|
| 116 |
Eq. \ref{eq:laplacetransform} was solved numerically for the Ag-Cu |
| 117 |
system using mole-fraction weighted values for $c_p$ and $\rho_p$ of |
| 118 |
0.295 $(\mathrm{J g^{-1} K^{-1}})$ and $9.826\times 10^6$ $(\mathrm{g |
| 119 |
m^{-3}})$ respectively. Since most of the laser excitation experiments |
| 120 |
have been done in aqueous solutions, parameters used for the fluid are |
| 121 |
$K$ of $0.6$ $(\mathrm{Wm^{-1}K^{-1}})$, $\rho$ of $1.0\times10^6$ |
| 122 |
$(\mathrm{g m^{-3}})$ and $c$ of $4.184$ $(\mathrm{J g^{-1} K^{-1}})$. |
| 123 |
|
| 124 |
Values for the interfacial conductance have been determined by a |
| 125 |
number of groups for similar nanoparticles and range from a low |
| 126 |
$87.5\times 10^{6}$ $(\mathrm{Wm^{-2}K^{-1}})$ to $120\times 10^{6}$ |
| 127 |
$(\mathrm{Wm^{-2}K^{-1}})$.\cite{hartlandPrv2007} Similarly, Plech |
| 128 |
{\it et al.} reported a value for the interfacial conductance of |
| 129 |
$G=105\pm 15$ $(\mathrm{Wm^{-2}K^{-1}})$ and $G=130\pm 15$ |
| 130 |
$(\mathrm{Wm^{-2}K^{-1}})$ for Pt |
| 131 |
nanoparticles.\cite{plech:195423,PhysRevB.66.224301} |
| 132 |
|
| 133 |
We conducted our simulations at both ends of the range of |
| 134 |
experimentally-determined values for the interfacial conductance. |
| 135 |
This allows us to observe both the slowest and fastest heat transfers |
| 136 |
from the nanoparticle to the solvent that are consistent with |
| 137 |
experimental observations. For the slowest heat transfer, a value for |
| 138 |
G of $87.5\times 10^{6}$ $(\mathrm{Wm^{-2}K^{-1}})$ was used and for |
| 139 |
the fastest heat transfer, a value of $117\times 10^{6}$ |
| 140 |
$(\mathrm{Wm^{-2}K^{-1}})$ was used. Based on calculations we have |
| 141 |
done using raw data from the Hartland group's thermal half-time |
| 142 |
experiments on Au nanospheres, the true G values are probably in the |
| 143 |
faster regime: $117\times 10^{6}$ $(\mathrm{Wm^{-2}K^{-1}})$. |
| 144 |
|
| 145 |
The rate of cooling for the nanoparticles in a molecular dynamics |
| 146 |
simulation can then be tuned by changing the effective solvent |
| 147 |
viscosity ($\eta$) until the nanoparticle cooling rate matches the |
| 148 |
cooling rate described by the heat-transfer equations |
| 149 |
(\ref{eq:heateqn}). The effective solvent viscosity (in poise) for a G |
| 150 |
of $87.5\times 10^{6}$ $(\mathrm{Wm^{-2}K^{-1}})$ is 0.17, 0.20, and |
| 151 |
0.22 for 20 {\AA}, 30 {\AA}, and 40 {\AA} particles, respectively. The |
| 152 |
effective solvent viscosity (again in poise) for an interfacial |
| 153 |
conductance of $117\times 10^{6}$ $(\mathrm{Wm^{-2}K^{-1}})$ is 0.23, |
| 154 |
0.29, and 0.30 for 20 {\AA}, 30 {\AA} and 40 {\AA} particles. Cooling |
| 155 |
traces for each particle size are presented in |
| 156 |
Fig. \ref{fig:images_cooling_plot}. It should be noted that the |
| 157 |
Langevin thermostat produces cooling curves that are consistent with |
| 158 |
Newtonian (single-exponential) cooling, which cannot match the cooling |
| 159 |
profiles from Eq. \ref{eq:laplacetransform} exactly. Fitting the |
| 160 |
Langevin cooling profiles to a single-exponential produces |
| 161 |
$\tau=25.576$ ps, $\tau=43.786$ ps, and $\tau=56.621$ ps for the 20, |
| 162 |
30 and 40 {\AA} nanoparticles and a G of $87.5\times 10^{6}$ |
| 163 |
$(\mathrm{Wm^{-2}K^{-1}})$. For comparison's sake, similar |
| 164 |
single-exponential fits with an interfacial conductance of G of |
| 165 |
$117\times 10^{6}$ $(\mathrm{Wm^{-2}K^{-1}})$ produced a $\tau=13.391$ |
| 166 |
ps, $\tau=30.426$ ps, $\tau=43.857$ ps for the 20, 30 and 40 {\AA} |
| 167 |
nanoparticles. |
| 168 |
|
| 169 |
\begin{figure}[htbp] |
| 170 |
\centering |
| 171 |
\includegraphics[width=5in]{images/cooling_plot.pdf} |
| 172 |
\caption{Thermal cooling curves obtained from the inverse Laplace |
| 173 |
transform heat model in Eq. \ref{eq:laplacetransform} (solid line) as |
| 174 |
well as from molecular dynamics simulations (circles). Effective |
| 175 |
solvent viscosities of 0.23-0.30 poise (depending on the radius of the |
| 176 |
particle) give the best fit to the experimental cooling curves. |
| 177 |
%Since |
| 178 |
%this viscosity is substantially in excess of the viscosity of liquid |
| 179 |
%water, much of the thermal transfer to the surroundings is probably |
| 180 |
%due to the capping agent. |
| 181 |
} |
| 182 |
\label{fig:images_cooling_plot} |
| 183 |
\end{figure} |
| 184 |
|
| 185 |
\subsection{Potentials for classical simulations of bimetallic |
| 186 |
nanoparticles} |
| 187 |
|
| 188 |
Several different potential models have been developed that reasonably |
| 189 |
describe interactions in transition metals. In particular, the |
| 190 |
Embedded Atom Model ({\sc eam})~\cite{PhysRevB.33.7983} and |
| 191 |
Sutton-Chen ({\sc sc})~\cite{Chen90} potential have been used to study |
| 192 |
a wide range of phenomena in both bulk materials and |
| 193 |
nanoparticles.\cite{Vardeman-II:2001jn,ShibataT._ja026764r} Both |
| 194 |
potentials are based on a model of a metal which treats the nuclei and |
| 195 |
core electrons as pseudo-atoms embedded in the electron density due to |
| 196 |
the valence electrons on all of the other atoms in the system. The |
| 197 |
{\sc sc} potential has a simple form that closely resembles that of |
| 198 |
the ubiquitous Lennard Jones potential, |
| 199 |
\begin{equation} |
| 200 |
\label{eq:SCP1} |
| 201 |
U_{tot}=\sum _{i}\left[ \frac{1}{2}\sum _{j\neq i}D_{ij}V^{pair}_{ij}(r_{ij})-c_{i}D_{ii}\sqrt{\rho_{i}}\right] , |
| 202 |
\end{equation} |
| 203 |
where $V^{pair}_{ij}$ and $\rho_{i}$ are given by |
| 204 |
\begin{equation} |
| 205 |
\label{eq:SCP2} |
| 206 |
V^{pair}_{ij}(r)=\left( \frac{\alpha_{ij}}{r_{ij}}\right)^{n_{ij}}, \rho_{i}=\sum_{j\neq i}\left( \frac{\alpha_{ij}}{r_{ij}}\right) ^{m_{ij}}. |
| 207 |
\end{equation} |
| 208 |
$V^{pair}_{ij}$ is a repulsive pairwise potential that accounts for |
| 209 |
interactions between the pseudoatom cores. The $\sqrt{\rho_i}$ term in |
| 210 |
Eq. (\ref{eq:SCP1}) is an attractive many-body potential that models |
| 211 |
the interactions between the valence electrons and the cores of the |
| 212 |
pseudo-atoms. $D_{ij}$, $D_{ii}$ set the appropriate overall energy |
| 213 |
scale, $c_i$ scales the attractive portion of the potential relative |
| 214 |
to the repulsive interaction and $\alpha_{ij}$ is a length parameter |
| 215 |
that assures a dimensionless form for $\rho$. These parameters are |
| 216 |
tuned to various experimental properties such as the density, cohesive |
| 217 |
energy, and elastic moduli for FCC transition metals. The quantum |
| 218 |
Sutton-Chen ({\sc q-sc}) formulation matches these properties while |
| 219 |
including zero-point quantum corrections for different transition |
| 220 |
metals.\cite{PhysRevB.59.3527} This particular parametarization has |
| 221 |
been shown to reproduce the experimentally available heat of mixing |
| 222 |
data for both FCC solid solutions of Ag-Cu and the high-temperature |
| 223 |
liquid.\cite{sheng:184203} In contrast, the {\sc eam} potential does |
| 224 |
not reproduce the experimentally observed heat of mixing for the |
| 225 |
liquid alloy.\cite{MURRAY:1984lr} Combination rules for the alloy were |
| 226 |
taken to be the arithmatic average of the atomic parameters with the |
| 227 |
exception of $c_i$ since its values is only dependent on the identity |
| 228 |
of the atom where the density is evaluated. For the {\sc q-sc} |
| 229 |
potential, cutoff distances are traditionally taken to be |
| 230 |
$2\alpha_{ij}$ and include up to the sixth coordination shell in FCC |
| 231 |
metals. |
| 232 |
|
| 233 |
%\subsection{Sampling single-temperature configurations from a cooling |
| 234 |
%trajectory} |
| 235 |
|
| 236 |
To better understand the structural changes occurring in the |
| 237 |
nanoparticles throughout the cooling trajectory, configurations were |
| 238 |
sampled at regular intervals during the cooling trajectory. These |
| 239 |
configurations were then allowed to evolve under NVE dynamics to |
| 240 |
sample from the proper distribution in phase space. Figure |
| 241 |
\ref{fig:images_cooling_time_traces} illustrates this sampling. |
| 242 |
|
| 243 |
|
| 244 |
\begin{figure}[htbp] |
| 245 |
\centering |
| 246 |
\includegraphics[height=3in]{images/cooling_time_traces.pdf} |
| 247 |
\caption{Illustrative cooling profile for the 40 {\AA} |
| 248 |
nanoparticle evolving under stochastic boundary conditions |
| 249 |
corresponding to $G=$$87.5\times 10^{6}$ |
| 250 |
$(\mathrm{Wm^{-2}K^{-1}})$. At temperatures along the cooling |
| 251 |
trajectory, configurations were sampled and allowed to evolve in the |
| 252 |
NVE ensemble. These subsequent trajectories were analyzed for |
| 253 |
structural features associated with bulk glass formation.} |
| 254 |
\label{fig:images_cooling_time_traces} |
| 255 |
\end{figure} |
| 256 |
|
| 257 |
|
| 258 |
\begin{figure}[htbp] |
| 259 |
\centering |
| 260 |
\includegraphics[width=5in]{images/cross_section_array.jpg} |
| 261 |
\caption{Cutaway views of 30 \AA\ Ag-Cu nanoparticle structures for |
| 262 |
random alloy (top) and Cu (core) / Ag (shell) initial conditions |
| 263 |
(bottom). Shown from left to right are the crystalline, liquid |
| 264 |
droplet, and final glassy bead configurations.} |
| 265 |
\label{fig:cross_sections} |
| 266 |
\end{figure} |
