| 1 |
+ |
%!TEX root = /Users/charles/Desktop/nanoglass/nanoglass.tex |
| 2 |
+ |
|
| 3 |
|
\section{Computational Methodology} |
| 4 |
|
\label{sec:details} |
| 5 |
|
|
| 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 |
< |
$\mathrm{Ag}_6\mathrm{Cu}_4$. Three different sizes of nanoparticles |
| 11 |
< |
corresponding to a 20 \AA radius (1961 atoms), 30 {\AA} radius (6603 |
| 12 |
< |
atoms) and 40 {\AA} radius (15683 atoms) were constructed. These |
| 13 |
< |
initial structures were relaxed to their equilibrium structures at 20 |
| 14 |
< |
K for 20 ps and again at 300 K for 100 ps sampling from a |
| 15 |
< |
Maxwell-Boltzmann distribution at each temperature. |
| 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 |
| 19 |
> |
temperature. All simulations were conducted using the {\sc oopse} |
| 20 |
> |
molecular dynamics package.\cite{Meineke:2004uq} |
| 21 |
|
|
| 22 |
|
To mimic the effects of the heating due to laser irradiation, the |
| 23 |
|
particles were allowed to melt by sampling velocities from the Maxwell |
| 36 |
|
To approximate the effects of rapid heat transfer to the solvent |
| 37 |
|
following a heating at the plasmon resonance, we utilized a |
| 38 |
|
methodology in which atoms contained in the outer $4$ {\AA} radius of |
| 39 |
< |
the nanoparticle evolved under Langevin Dynamics with a solvent |
| 40 |
< |
friction approximating the contribution from the solvent and capping |
| 41 |
< |
agent. Atoms located in the interior of the nanoparticle evolved |
| 42 |
< |
under Newtonian dynamics. The set-up of our simulations is nearly |
| 43 |
< |
identical with the ``stochastic boundary molecular dynamics'' ({\sc |
| 44 |
< |
sbmd}) method that has seen wide use in the protein simulation |
| 39 |
> |
the nanoparticle evolved under Langevin Dynamics, |
| 40 |
> |
\begin{equation} |
| 41 |
> |
m \frac{\partial^2 \vec{x}}{\partial t^2} = F_\textrm{sys}(\vec{x}(t)) |
| 42 |
> |
- 6 \pi a \eta \vec{v}(t) + F_\textrm{ran} |
| 43 |
> |
\label{eq:langevin} |
| 44 |
> |
\end{equation} |
| 45 |
> |
with a solvent friction ($\eta$) approximating the contribution from |
| 46 |
> |
the solvent and capping agent. Atoms located in the interior of the |
| 47 |
> |
nanoparticle evolved under Newtonian dynamics. The set-up of our |
| 48 |
> |
simulations is nearly identical with the ``stochastic boundary |
| 49 |
> |
molecular dynamics'' ({\sc sbmd}) method that has seen wide use in the |
| 50 |
> |
protein simulation |
| 51 |
|
community.\cite{BROOKS:1985kx,BROOKS:1983uq,BRUNGER:1984fj} A sketch |
| 52 |
< |
of this setup can be found in Fig. \ref{fig:langevinSketch}. For a |
| 53 |
< |
spherical atom of radius $a$, the Langevin frictional forces can be |
| 54 |
< |
determined by Stokes' law |
| 55 |
< |
\begin{equation} |
| 56 |
< |
\mathbf{F}_{\mathrm{frictional}}=6\pi a \eta \mathbf{v} |
| 52 |
> |
of this setup can be found in Fig. \ref{fig:langevinSketch}. In |
| 53 |
> |
Eq. (\ref{eq:langevin}) the frictional forces of a spherical atom |
| 54 |
> |
of radius $a$ depend on the solvent viscosity. The random forces are |
| 55 |
> |
usually taken as gaussian random variables with zero mean and a |
| 56 |
> |
variance tied to the solvent viscosity and temperature, |
| 57 |
> |
\begin{equation} |
| 58 |
> |
\langle F_\textrm{ran}(t) \cdot F_\textrm{ran} (t') |
| 59 |
> |
\rangle = 2 k_B T (6 \pi \eta a) \delta(t - t') |
| 60 |
> |
\label{eq:stochastic} |
| 61 |
|
\end{equation} |
| 62 |
< |
where $\eta$ is the effective viscosity of the solvent in which the |
| 63 |
< |
particle is embedded. Due to the presence of the capping agent and |
| 64 |
< |
the lack of details about the atomic-scale interactions between the |
| 65 |
< |
metallic atoms and the solvent, the effective viscosity is a |
| 49 |
< |
essentially a free parameter that must be tuned to give experimentally |
| 50 |
< |
relevant simulations. |
| 62 |
> |
Due to the presence of the capping agent and the lack of details about |
| 63 |
> |
the atomic-scale interactions between the metallic atoms and the |
| 64 |
> |
solvent, the effective viscosity is a essentially a free parameter |
| 65 |
> |
that must be tuned to give experimentally relevant simulations. |
| 66 |
|
\begin{figure}[htbp] |
| 67 |
|
\centering |
| 68 |
< |
\includegraphics[width=\linewidth]{images/stochbound.pdf} |
| 69 |
< |
\caption{Methodology for nanoparticle cooling. Equations of motion for metal atoms contained in the outer 4 {\AA} were determined by Langevins' Equations of motion. Metal atoms outside this region were allowed to evolve under Newtonian dynamics.} |
| 68 |
> |
\includegraphics[width=5in]{images/stochbound.pdf} |
| 69 |
> |
\caption{Methodology used to mimic the experimental cooling conditions |
| 70 |
> |
of a hot nanoparticle surrounded by a solvent. Atoms in the core of |
| 71 |
> |
the particle evolved under Newtonian dynamics, while atoms that were |
| 72 |
> |
in the outer skin of the particle evolved under Langevin dynamics. |
| 73 |
> |
The radius of the spherical region operating under Newtonian dynamics, |
| 74 |
> |
$r_\textrm{Newton}$ was set to be 4 {\AA} smaller than the original |
| 75 |
> |
radius ($R$) of the liquid droplet.} |
| 76 |
|
\label{fig:langevinSketch} |
| 77 |
|
\end{figure} |
| 78 |
+ |
|
| 79 |
|
The viscosity ($\eta$) can be tuned by comparing the cooling rate that |
| 80 |
|
a set of nanoparticles experience with the known cooling rates for |
| 81 |
< |
those particles obtained via the laser heating experiments. |
| 81 |
> |
similar particles obtained via the laser heating experiments. |
| 82 |
|
Essentially, we tune the solvent viscosity until the thermal decay |
| 83 |
|
profile matches a heat-transfer model using reasonable values for the |
| 84 |
|
interfacial conductance and the thermal conductivity of the solvent. |
| 85 |
|
|
| 86 |
|
Cooling rates for the experimentally-observed nanoparticles were |
| 87 |
|
calculated from the heat transfer equations for a spherical particle |
| 88 |
< |
embedded in a ambient medium that allows for diffusive heat |
| 89 |
< |
transport. The heat transfer model is a set of two coupled |
| 90 |
< |
differential equations in the Laplace domain, |
| 88 |
> |
embedded in a ambient medium that allows for diffusive heat transport. |
| 89 |
> |
Following Plech {\it et al.},\cite{plech:195423} we use a heat |
| 90 |
> |
transfer model that consists of two coupled differential equations |
| 91 |
> |
in the Laplace domain, |
| 92 |
|
\begin{eqnarray} |
| 93 |
|
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\\ |
| 94 |
|
\left(\frac{\partial}{\partial r} T_{f}(r,s)\right)_{r=R} + |
| 113 |
|
\end{equation} |
| 114 |
|
For simplicity, we have introduced the thermal diffusivity $\kappa = |
| 115 |
|
K/(\rho c)$, and defined $k=4\pi R^3 \rho c /(M c_p)$ and $g = G/K$ in |
| 116 |
< |
Eq. \ref{eq:laplacetransform}. |
| 116 |
> |
Eq. (\ref{eq:laplacetransform}). |
| 117 |
|
|
| 118 |
< |
Eq. \ref{eq:laplacetransform} was solved numerically for the Ag-Cu |
| 118 |
> |
Eq. (\ref{eq:laplacetransform}) was solved numerically for the Ag-Cu |
| 119 |
|
system using mole-fraction weighted values for $c_p$ and $\rho_p$ of |
| 120 |
|
0.295 $(\mathrm{J g^{-1} K^{-1}})$ and $9.826\times 10^6$ $(\mathrm{g |
| 121 |
|
m^{-3}})$ respectively. Since most of the laser excitation experiments |
| 122 |
|
have been done in aqueous solutions, parameters used for the fluid are |
| 123 |
< |
$K$ of $0.6$ $(\mathrm{Wm^{-1}K^{-1}})$, $\rho$ of $1.0\times10^6$ $(\mathrm{g |
| 124 |
< |
m^{-3}})$ and $c$ of $4.184$ $(\mathrm{J g^{-1} K^{-1}})$. |
| 123 |
> |
$K$ of $0.6$ $(\mathrm{Wm^{-1}K^{-1}})$, $\rho$ of $1.0\times10^6$ |
| 124 |
> |
$(\mathrm{g m^{-3}})$ and $c$ of $4.184$ $(\mathrm{J g^{-1} K^{-1}})$. |
| 125 |
|
|
| 126 |
|
Values for the interfacial conductance have been determined by a |
| 127 |
|
number of groups for similar nanoparticles and range from a low |
| 128 |
< |
$87.5\times 10^{6}$ $(\mathrm{Wm^{-2}K^{-1}})$ to $120\times 10^{6}$ |
| 129 |
< |
$(\mathrm{Wm^{-2}K^{-1}})$.\cite{hartlandPrv2007} Plech {\it et al.} reported a value for the interfacial conductance of $G=105\pm 15$ $(\mathrm{Wm^{-2}K^{-1}})$ and |
| 130 |
< |
$G=130\pm 15$ $(\mathrm{Wm^{-2}K^{-1}})$ for Pt nanoparticles.\cite{plech:195423,PhysRevB.66.224301} |
| 128 |
> |
$87.5\times 10^{6} (\mathrm{Wm^{-2}K^{-1}})$ to $130\times 10^{6} |
| 129 |
> |
(\mathrm{Wm^{-2}K^{-1}})$.\cite{Wilson:2002uq} Wilson {\it |
| 130 |
> |
et al.} worked with Au, Pt, and AuPd nanoparticles and obtained an |
| 131 |
> |
estimate for the interfacial conductance of $G=130 |
| 132 |
> |
(\mathrm{Wm^{-2}K^{-1}})$.\cite{Wilson:2002uq} Similarly, Plech {\it |
| 133 |
> |
et al.} reported a value for the interfacial conductance of $G=105\pm |
| 134 |
> |
15 (\mathrm{Wm^{-2}K^{-1}})$ for Au nanoparticles.\cite{plech:195423} |
| 135 |
|
|
| 136 |
|
We conducted our simulations at both ends of the range of |
| 137 |
|
experimentally-determined values for the interfacial conductance. |
| 142 |
|
the fastest heat transfer, a value of $117\times 10^{6}$ |
| 143 |
|
$(\mathrm{Wm^{-2}K^{-1}})$ was used. Based on calculations we have |
| 144 |
|
done using raw data from the Hartland group's thermal half-time |
| 145 |
< |
experiments on Au nanospheres, we believe that the true G values are |
| 146 |
< |
closer to the faster regime: $117\times 10^{6}$ $(\mathrm{Wm^{-2}K^{-1}})$. |
| 145 |
> |
experiments on Au nanospheres,\cite{HuM._jp020581+} the true G values |
| 146 |
> |
are probably in the faster regime: $117\times 10^{6}$ |
| 147 |
> |
$(\mathrm{Wm^{-2}K^{-1}})$. |
| 148 |
|
|
| 121 |
– |
|
| 149 |
|
The rate of cooling for the nanoparticles in a molecular dynamics |
| 150 |
|
simulation can then be tuned by changing the effective solvent |
| 151 |
|
viscosity ($\eta$) until the nanoparticle cooling rate matches the |
| 152 |
< |
cooling rate described by the heat-transfer equations |
| 153 |
< |
(\ref{eq:heateqn}). The effective solvent viscosity (in poise) for a G |
| 154 |
< |
of $87.5\times 10^{6}$ $(\mathrm{Wm^{-2}K^{-1}})$ is 0.17, 0.20, and |
| 155 |
< |
0.22 for 20 {\AA}, 30 {\AA}, and 40 {\AA} particles, respectively. The |
| 156 |
< |
effective solvent viscosity (again in poise) for an interfacial |
| 157 |
< |
conductance of $117\times 10^{6}$ $(\mathrm{Wm^{-2}K^{-1}})$ is 0.23, |
| 158 |
< |
0.29, and 0.30 for 20 {\AA}, 30 {\AA} and 40 {\AA} particles. Cooling |
| 159 |
< |
traces for each particle size are presented in |
| 152 |
> |
cooling rate described by the heat-transfer Eq. |
| 153 |
> |
(\ref{eq:heateqn}). The effective solvent viscosity (in Pa s) for a G |
| 154 |
> |
of $87.5\times 10^{6}$ $(\mathrm{Wm^{-2}K^{-1}})$ is $4.2 \times |
| 155 |
> |
10^{-6}$, $5.0 \times 10^{-6}$, and |
| 156 |
> |
$5.5 \times 10^{-6}$ for 20 {\AA}, 30 {\AA}, and 40 {\AA} particles, respectively. The |
| 157 |
> |
effective solvent viscosity (again in Pa s) for an interfacial |
| 158 |
> |
conductance of $117\times 10^{6}$ $(\mathrm{Wm^{-2}K^{-1}})$ is $5.7 |
| 159 |
> |
\times 10^{-6}$, $7.2 \times 10^{-6}$, and $7.5 \times 10^{-6}$ |
| 160 |
> |
for 20 {\AA}, 30 {\AA} and 40 {\AA} particles. These viscosities are |
| 161 |
> |
essentially gas-phase values, a fact which is consistent with the |
| 162 |
> |
initial temperatures of the particles being well into the |
| 163 |
> |
super-critical region for the aqueous environment. Gas bubble |
| 164 |
> |
generation has also been seen experimentally around gold nanoparticles |
| 165 |
> |
in water.\cite{kotaidis:184702} Instead of a single value for the |
| 166 |
> |
effective viscosity, a time-dependent parameter might be a better |
| 167 |
> |
mimic of the cooling vapor layer that surrounds the hot particles. |
| 168 |
> |
This may also be a contributing factor to the size-dependence of the |
| 169 |
> |
effective viscosities in our simulations. |
| 170 |
> |
|
| 171 |
> |
Cooling traces for each particle size are presented in |
| 172 |
|
Fig. \ref{fig:images_cooling_plot}. It should be noted that the |
| 173 |
|
Langevin thermostat produces cooling curves that are consistent with |
| 174 |
|
Newtonian (single-exponential) cooling, which cannot match the cooling |
| 175 |
< |
profiles from Eq. \ref{eq:laplacetransform} exactly. |
| 175 |
> |
profiles from Eq. (\ref{eq:laplacetransform}) exactly. Fitting the |
| 176 |
> |
Langevin cooling profiles to a single-exponential produces |
| 177 |
> |
$\tau=25.576$ ps, $\tau=43.786$ ps, and $\tau=56.621$ ps for the 20, |
| 178 |
> |
30 and 40 {\AA} nanoparticles and a G of $87.5\times 10^{6}$ |
| 179 |
> |
$(\mathrm{Wm^{-2}K^{-1}})$. For comparison's sake, similar |
| 180 |
> |
single-exponential fits with an interfacial conductance of G of |
| 181 |
> |
$117\times 10^{6}$ $(\mathrm{Wm^{-2}K^{-1}})$ produced a $\tau=13.391$ |
| 182 |
> |
ps, $\tau=30.426$ ps, $\tau=43.857$ ps for the 20, 30 and 40 {\AA} |
| 183 |
> |
nanoparticles. |
| 184 |
|
|
| 185 |
|
\begin{figure}[htbp] |
| 186 |
|
\centering |
| 187 |
< |
\includegraphics[width=\linewidth]{images/cooling_plot.pdf} |
| 187 |
> |
\includegraphics[width=5in]{images/cooling_plot.pdf} |
| 188 |
|
\caption{Thermal cooling curves obtained from the inverse Laplace |
| 189 |
< |
transform heat model in Eq. \ref{eq:laplacetransform} (solid line) as |
| 189 |
> |
transform heat model in Eq. (\ref{eq:laplacetransform}) (solid line) as |
| 190 |
|
well as from molecular dynamics simulations (circles). Effective |
| 191 |
< |
solvent viscosities of 0.23-0.30 poise (depending on the radius of the |
| 192 |
< |
particle) give the best fit to the experimental cooling curves. Since |
| 193 |
< |
this viscosity is substantially in excess of the viscosity of liquid |
| 194 |
< |
water, much of the thermal transfer to the surroundings is probably |
| 195 |
< |
due to the capping agent.} |
| 191 |
> |
solvent viscosities of 4.2-7.5 $\times 10^{-6}$ Pa s (depending on the |
| 192 |
> |
radius of the particle) give the best fit to the experimental cooling |
| 193 |
> |
curves. This viscosity suggests that the nanoparticles in these |
| 194 |
> |
experiments are surrounded by a vapor layer (which is a reasonable |
| 195 |
> |
assumptions given the initial temperatures of the particles). } |
| 196 |
|
\label{fig:images_cooling_plot} |
| 197 |
|
\end{figure} |
| 198 |
|
|
| 204 |
|
Embedded Atom Model ({\sc eam})~\cite{PhysRevB.33.7983} and |
| 205 |
|
Sutton-Chen ({\sc sc})~\cite{Chen90} potential have been used to study |
| 206 |
|
a wide range of phenomena in both bulk materials and |
| 207 |
< |
nanoparticles.\cite{Vardeman-II:2001jn,ShibataT._ja026764r} Both |
| 207 |
> |
nanoparticles.\cite{Vardeman-II:2001jn,ShibataT._ja026764r,Sankaranarayanan:2005lr,Chui:2003fk,Wang:2005qy,Medasani:2007uq} Both |
| 208 |
|
potentials are based on a model of a metal which treats the nuclei and |
| 209 |
|
core electrons as pseudo-atoms embedded in the electron density due to |
| 210 |
|
the valence electrons on all of the other atoms in the system. The |
| 236 |
|
data for both FCC solid solutions of Ag-Cu and the high-temperature |
| 237 |
|
liquid.\cite{sheng:184203} In contrast, the {\sc eam} potential does |
| 238 |
|
not reproduce the experimentally observed heat of mixing for the |
| 239 |
< |
liquid alloy.\cite{MURRAY:1984lr} Combination rules for the alloy were |
| 240 |
< |
taken to be the arithmatic average of the atomic parameters with the |
| 241 |
< |
exception of $c_i$ since its values is only dependent on the identity |
| 242 |
< |
of the atom where the density is evaluated. For the {\sc q-sc} |
| 243 |
< |
potential, cutoff distances are traditionally taken to be |
| 244 |
< |
$2\alpha_{ij}$ and include up to the sixth coordination shell in FCC |
| 245 |
< |
metals. |
| 239 |
> |
liquid alloy.\cite{MURRAY:1984lr} In this work, we have utilized the |
| 240 |
> |
{\sc q-sc} formulation for our potential energies and forces. |
| 241 |
> |
Combination rules for the alloy were taken to be the arithmetic |
| 242 |
> |
average of the atomic parameters with the exception of $c_i$ since its |
| 243 |
> |
values is only dependent on the identity of the atom where the density |
| 244 |
> |
is evaluated. For the {\sc q-sc} potential, cutoff distances are |
| 245 |
> |
traditionally taken to be $2\alpha_{ij}$ and include up to the sixth |
| 246 |
> |
coordination shell in FCC metals. |
| 247 |
|
|
| 248 |
< |
\subsection{Sampling single-temperature configurations from a cooling |
| 249 |
< |
trajectory} |
| 248 |
> |
%\subsection{Sampling single-temperature configurations from a cooling |
| 249 |
> |
%trajectory} |
| 250 |
|
|
| 251 |
< |
ffdsafjdksalfdsa |
| 251 |
> |
To better understand the structural changes occurring in the |
| 252 |
> |
nanoparticles throughout the cooling trajectory, configurations were |
| 253 |
> |
sampled at regular intervals during the cooling trajectory. These |
| 254 |
> |
configurations were then allowed to evolve under NVE dynamics to |
| 255 |
> |
sample from the proper distribution in phase space. Fig. |
| 256 |
> |
\ref{fig:images_cooling_time_traces} illustrates this sampling. |
| 257 |
> |
|
| 258 |
> |
|
| 259 |
> |
\begin{figure}[htbp] |
| 260 |
> |
\centering |
| 261 |
> |
\includegraphics[height=3in]{images/cooling_time_traces.pdf} |
| 262 |
> |
\caption{Illustrative cooling profile for the 40 {\AA} |
| 263 |
> |
nanoparticle evolving under stochastic boundary conditions |
| 264 |
> |
corresponding to $G=$$87.5\times 10^{6}$ |
| 265 |
> |
$(\mathrm{Wm^{-2}K^{-1}})$. At temperatures along the cooling |
| 266 |
> |
trajectory, configurations were sampled and allowed to evolve in the |
| 267 |
> |
NVE ensemble. These subsequent trajectories were analyzed for |
| 268 |
> |
structural features associated with bulk glass formation.} |
| 269 |
> |
\label{fig:images_cooling_time_traces} |
| 270 |
> |
\end{figure} |
| 271 |
> |
|
| 272 |
> |
|
| 273 |
> |
\begin{figure}[htbp] |
| 274 |
> |
\centering |
| 275 |
> |
\includegraphics[width=5in]{images/cross_section_array.jpg} |
| 276 |
> |
\caption{Cutaway views of 30 \AA\ Ag-Cu nanoparticle structures for |
| 277 |
> |
random alloy (top) and Cu (core) / Ag (shell) initial conditions |
| 278 |
> |
(bottom). Shown from left to right are the crystalline, liquid |
| 279 |
> |
droplet, and final glassy bead configurations.} |
| 280 |
> |
\label{fig:cross_sections} |
| 281 |
> |
\end{figure} |
