ViewVC Help
View File | Revision Log | Show Annotations | View Changeset | Root Listing
root/group/interfacial/interfacial.tex
Revision: 3750
Committed: Tue Jul 26 18:10:49 2011 UTC (12 years, 11 months ago) by skuang
Content type: application/x-tex
File size: 50571 byte(s)
Log Message:
add more references.

File Contents

# User Rev Content
1 gezelter 3717 \documentclass[11pt]{article}
2     \usepackage{amsmath}
3     \usepackage{amssymb}
4     \usepackage{setspace}
5     \usepackage{endfloat}
6     \usepackage{caption}
7     %\usepackage{tabularx}
8     \usepackage{graphicx}
9     \usepackage{multirow}
10     %\usepackage{booktabs}
11     %\usepackage{bibentry}
12     %\usepackage{mathrsfs}
13     %\usepackage[ref]{overcite}
14     \usepackage[square, comma, sort&compress]{natbib}
15     \usepackage{url}
16     \pagestyle{plain} \pagenumbering{arabic} \oddsidemargin 0.0cm
17     \evensidemargin 0.0cm \topmargin -21pt \headsep 10pt \textheight
18     9.0in \textwidth 6.5in \brokenpenalty=10000
19    
20     % double space list of tables and figures
21     \AtBeginDelayedFloats{\renewcommand{\baselinestretch}{1.66}}
22     \setlength{\abovecaptionskip}{20 pt}
23     \setlength{\belowcaptionskip}{30 pt}
24    
25 skuang 3727 %\renewcommand\citemid{\ } % no comma in optional reference note
26 gezelter 3740 \bibpunct{[}{]}{,}{n}{}{;}
27     \bibliographystyle{achemso}
28 gezelter 3717
29     \begin{document}
30    
31     \title{Simulating interfacial thermal conductance at metal-solvent
32     interfaces: the role of chemical capping agents}
33    
34     \author{Shenyu Kuang and J. Daniel
35     Gezelter\footnote{Corresponding author. \ Electronic mail: gezelter@nd.edu} \\
36     Department of Chemistry and Biochemistry,\\
37     University of Notre Dame\\
38     Notre Dame, Indiana 46556}
39    
40     \date{\today}
41    
42     \maketitle
43    
44     \begin{doublespace}
45    
46     \begin{abstract}
47 skuang 3725
48 skuang 3732 With the Non-Isotropic Velocity Scaling algorithm (NIVS) we have
49     developed, an unphysical thermal flux can be effectively set up even
50     for non-homogeneous systems like interfaces in non-equilibrium
51     molecular dynamics simulations. In this work, this algorithm is
52     applied for simulating thermal conductance at metal / organic solvent
53     interfaces with various coverages of butanethiol capping
54     agents. Different solvents and force field models were tested. Our
55     results suggest that the United-Atom models are able to provide an
56     estimate of the interfacial thermal conductivity comparable to
57     experiments in our simulations with satisfactory computational
58     efficiency. From our results, the acoustic impedance mismatch between
59     metal and liquid phase is effectively reduced by the capping
60     agents, and thus leads to interfacial thermal conductance
61     enhancement. Furthermore, this effect is closely related to the
62     capping agent coverage on the metal surfaces and the type of solvent
63     molecules, and is affected by the models used in the simulations.
64 skuang 3725
65 gezelter 3717 \end{abstract}
66    
67     \newpage
68    
69     %\narrowtext
70    
71     %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
72     % BODY OF TEXT
73     %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
74    
75     \section{Introduction}
76 skuang 3725 Interfacial thermal conductance is extensively studied both
77 skuang 3737 experimentally and computationally\cite{cahill:793}, due to its
78     importance in nanoscale science and technology. Reliability of
79     nanoscale devices depends on their thermal transport
80     properties. Unlike bulk homogeneous materials, nanoscale materials
81     features significant presence of interfaces, and these interfaces
82     could dominate the heat transfer behavior of these
83 skuang 3733 materials. Furthermore, these materials are generally heterogeneous,
84 skuang 3737 which challenges traditional research methods for homogeneous
85     systems.
86 gezelter 3717
87 skuang 3733 Heat conductance of molecular and nano-scale interfaces will be
88     affected by the chemical details of the surface. Experimentally,
89     various interfaces have been investigated for their thermal
90     conductance properties. Wang {\it et al.} studied heat transport
91     through long-chain hydrocarbon monolayers on gold substrate at
92     individual molecular level\cite{Wang10082007}; Schmidt {\it et al.}
93     studied the role of CTAB on thermal transport between gold nanorods
94     and solvent\cite{doi:10.1021/jp8051888}; Juv\'e {\it et al.} studied
95     the cooling dynamics, which is controlled by thermal interface
96     resistence of glass-embedded metal
97     nanoparticles\cite{PhysRevB.80.195406}. Although interfaces are
98     commonly barriers for heat transport, Alper {\it et al.} suggested
99     that specific ligands (capping agents) could completely eliminate this
100     barrier ($G\rightarrow\infty$)\cite{doi:10.1021/la904855s}.
101    
102 skuang 3737 Theoretical and computational models have also been used to study the
103     interfacial thermal transport in order to gain an understanding of
104     this phenomena at the molecular level. Recently, Hase and coworkers
105     employed Non-Equilibrium Molecular Dynamics (NEMD) simulations to
106     study thermal transport from hot Au(111) substrate to a self-assembled
107 skuang 3738 monolayer of alkylthiol with relatively long chain (8-20 carbon
108 skuang 3737 atoms)\cite{hase:2010,hase:2011}. However, ensemble averaged
109     measurements for heat conductance of interfaces between the capping
110     monolayer on Au and a solvent phase has yet to be studied.
111 skuang 3738 The comparatively low thermal flux through interfaces is
112 skuang 3736 difficult to measure with Equilibrium MD or forward NEMD simulation
113 skuang 3750 methods. Therefore, the Reverse NEMD (RNEMD)
114     methods\cite{MullerPlathe:1997xw} would have the advantage of having
115     this difficult to measure flux known when studying the thermal
116     transport across interfaces, given that the simulation methods being
117     able to effectively apply an unphysical flux in non-homogeneous
118     systems. Garde and coworkers\cite{garde:nl2005,garde:PhysRevLett2009}
119     applied this approach to various liquid interfaces and studied how
120     thermal conductance (or resistance) is dependent on chemistry details
121     of interfaces, e.g. hydrophobic and hydrophilic aqueous interfaces.
122 skuang 3734
123 skuang 3725 Recently, we have developed the Non-Isotropic Velocity Scaling (NIVS)
124     algorithm for RNEMD simulations\cite{kuang:164101}. This algorithm
125     retains the desirable features of RNEMD (conservation of linear
126     momentum and total energy, compatibility with periodic boundary
127     conditions) while establishing true thermal distributions in each of
128 skuang 3737 the two slabs. Furthermore, it allows effective thermal exchange
129     between particles of different identities, and thus makes the study of
130     interfacial conductance much simpler.
131 skuang 3725
132 skuang 3737 The work presented here deals with the Au(111) surface covered to
133     varying degrees by butanethiol, a capping agent with short carbon
134     chain, and solvated with organic solvents of different molecular
135     properties. Different models were used for both the capping agent and
136     the solvent force field parameters. Using the NIVS algorithm, the
137     thermal transport across these interfaces was studied and the
138 skuang 3747 underlying mechanism for the phenomena was investigated.
139 skuang 3733
140 skuang 3737 [MAY ADD WHY STUDY AU-THIOL SURFACE; CITE SHAOYI JIANG]
141 skuang 3734
142 skuang 3721 \section{Methodology}
143 skuang 3737 \subsection{Imposd-Flux Methods in MD Simulations}
144 skuang 3749 Steady state MD simulations has the advantage that not many
145     trajectories are needed to study the relationship between thermal flux
146     and thermal gradients. For systems including low conductance
147     interfaces one must have a method capable of generating or measuring
148     relatively small fluxes, compared to those required for bulk
149     conductivity. This requirement makes the calculation even more
150     difficult for those slowly-converging equilibrium
151     methods\cite{Viscardy:2007lq}. Forward methods may impose gradient,
152     but in interfacial conditions it is not clear what behavior to impose
153     at the interfacial boundaries. Imposed-flux reverse non-equilibrium
154 skuang 3721 methods\cite{MullerPlathe:1997xw} have the flux set {\it a priori} and
155 skuang 3749 the thermal response becomes easier to measure than the flux. Although
156     M\"{u}ller-Plathe's original momentum swapping approach can be used
157     for exchanging energy between particles of different identity, the
158     kinetic energy transfer efficiency is affected by the mass difference
159     between the particles, which limits its application on heterogeneous
160     interfacial systems.
161 skuang 3721
162 skuang 3737 The non-isotropic velocity scaling (NIVS)\cite{kuang:164101} approach to
163     non-equilibrium MD simulations is able to impose a wide range of
164     kinetic energy fluxes without obvious perturbation to the velocity
165     distributions of the simulated systems. Furthermore, this approach has
166 skuang 3721 the advantage in heterogeneous interfaces in that kinetic energy flux
167     can be applied between regions of particles of arbitary identity, and
168 skuang 3737 the flux will not be restricted by difference in particle mass.
169 skuang 3721
170     The NIVS algorithm scales the velocity vectors in two separate regions
171     of a simulation system with respective diagonal scaling matricies. To
172     determine these scaling factors in the matricies, a set of equations
173     including linear momentum conservation and kinetic energy conservation
174 skuang 3737 constraints and target energy flux satisfaction is solved. With the
175     scaling operation applied to the system in a set frequency, bulk
176     temperature gradients can be easily established, and these can be used
177     for computing thermal conductivities. The NIVS algorithm conserves
178     momenta and energy and does not depend on an external thermostat.
179 skuang 3721
180 skuang 3727 \subsection{Defining Interfacial Thermal Conductivity $G$}
181 skuang 3747 Given a system with thermal gradients and the corresponding thermal
182     flux, for interfaces with a relatively low interfacial conductance,
183     the bulk regions on either side of an interface rapidly come to a
184     state in which the two phases have relatively homogeneous (but
185     distinct) temperatures. The interfacial thermal conductivity $G$ can
186     therefore be approximated as:
187 skuang 3727 \begin{equation}
188     G = \frac{E_{total}}{2 t L_x L_y \left( \langle T_\mathrm{hot}\rangle -
189     \langle T_\mathrm{cold}\rangle \right)}
190     \label{lowG}
191     \end{equation}
192     where ${E_{total}}$ is the imposed non-physical kinetic energy
193     transfer and ${\langle T_\mathrm{hot}\rangle}$ and ${\langle
194     T_\mathrm{cold}\rangle}$ are the average observed temperature of the
195     two separated phases.
196 skuang 3721
197 skuang 3737 When the interfacial conductance is {\it not} small, there are two
198     ways to define $G$.
199 skuang 3727
200 skuang 3737 One way is to assume the temperature is discrete on the two sides of
201     the interface. $G$ can be calculated using the applied thermal flux
202     $J$ and the maximum temperature difference measured along the thermal
203 skuang 3745 gradient max($\Delta T$), which occurs at the Gibbs deviding surface
204     (Figure \ref{demoPic}):
205 skuang 3727 \begin{equation}
206     G=\frac{J}{\Delta T}
207     \label{discreteG}
208     \end{equation}
209    
210 skuang 3745 \begin{figure}
211     \includegraphics[width=\linewidth]{method}
212     \caption{Interfacial conductance can be calculated by applying an
213     (unphysical) kinetic energy flux between two slabs, one located
214     within the metal and another on the edge of the periodic box. The
215     system responds by forming a thermal response or a gradient. In
216     bulk liquids, this gradient typically has a single slope, but in
217     interfacial systems, there are distinct thermal conductivity
218     domains. The interfacial conductance, $G$ is found by measuring the
219     temperature gap at the Gibbs dividing surface, or by using second
220     derivatives of the thermal profile.}
221     \label{demoPic}
222     \end{figure}
223    
224 skuang 3727 The other approach is to assume a continuous temperature profile along
225     the thermal gradient axis (e.g. $z$) and define $G$ at the point where
226     the magnitude of thermal conductivity $\lambda$ change reach its
227     maximum, given that $\lambda$ is well-defined throughout the space:
228     \begin{equation}
229     G^\prime = \Big|\frac{\partial\lambda}{\partial z}\Big|
230     = \Big|\frac{\partial}{\partial z}\left(-J_z\Big/
231     \left(\frac{\partial T}{\partial z}\right)\right)\Big|
232     = |J_z|\Big|\frac{\partial^2 T}{\partial z^2}\Big|
233     \Big/\left(\frac{\partial T}{\partial z}\right)^2
234     \label{derivativeG}
235     \end{equation}
236    
237     With the temperature profile obtained from simulations, one is able to
238     approximate the first and second derivatives of $T$ with finite
239 skuang 3737 difference methods and thus calculate $G^\prime$.
240 skuang 3727
241 skuang 3737 In what follows, both definitions have been used for calculation and
242     are compared in the results.
243 skuang 3727
244 skuang 3737 To compare the above definitions ($G$ and $G^\prime$), we have modeled
245     a metal slab with its (111) surfaces perpendicular to the $z$-axis of
246 skuang 3744 our simulation cells. Both with and without capping agents on the
247 skuang 3737 surfaces, the metal slab is solvated with simple organic solvents, as
248 skuang 3746 illustrated in Figure \ref{gradT}.
249 skuang 3727
250 skuang 3737 With the simulation cell described above, we are able to equilibrate
251     the system and impose an unphysical thermal flux between the liquid
252     and the metal phase using the NIVS algorithm. By periodically applying
253     the unphysical flux, we are able to obtain a temperature profile and
254     its spatial derivatives. These quantities enable the evaluation of the
255     interfacial thermal conductance of a surface. Figure \ref{gradT} is an
256 skuang 3747 example of how an applied thermal flux can be used to obtain the 1st
257 skuang 3737 and 2nd derivatives of the temperature profile.
258 skuang 3727
259     \begin{figure}
260     \includegraphics[width=\linewidth]{gradT}
261 skuang 3745 \caption{A sample of Au-butanethiol/hexane interfacial system and the
262     temperature profile after a kinetic energy flux is imposed to
263     it. The 1st and 2nd derivatives of the temperature profile can be
264     obtained with finite difference approximation (lower panel).}
265 skuang 3727 \label{gradT}
266     \end{figure}
267    
268     \section{Computational Details}
269 skuang 3730 \subsection{Simulation Protocol}
270 skuang 3737 The NIVS algorithm has been implemented in our MD simulation code,
271     OpenMD\cite{Meineke:2005gd,openmd}, and was used for our
272 skuang 3747 simulations. Different metal slab thickness (layer numbers of Au) was
273 skuang 3737 simulated. Metal slabs were first equilibrated under atmospheric
274     pressure (1 atm) and a desired temperature (e.g. 200K). After
275     equilibration, butanethiol capping agents were placed at three-fold
276 skuang 3747 hollow sites on the Au(111) surfaces. These sites could be either a
277     {\it fcc} or {\it hcp} site. However, Hase {\it et al.} found that
278     they are equivalent in a heat transfer process\cite{hase:2010}, so
279     they are not distinguished in our study. The maximum butanethiol
280     capacity on Au surface is $1/3$ of the total number of surface Au
281     atoms, and the packing forms a $(\sqrt{3}\times\sqrt{3})R30^\circ$
282 skuang 3749 structure\cite{doi:10.1021/ja00008a001,doi:10.1021/cr9801317}. A
283     series of different coverages was derived by evenly eliminating
284     butanethiols on the surfaces, and was investigated in order to study
285     the relation between coverage and interfacial conductance.
286 skuang 3727
287 skuang 3737 The capping agent molecules were allowed to migrate during the
288     simulations. They distributed themselves uniformly and sampled a
289     number of three-fold sites throughout out study. Therefore, the
290     initial configuration would not noticeably affect the sampling of a
291     variety of configurations of the same coverage, and the final
292     conductance measurement would be an average effect of these
293 skuang 3746 configurations explored in the simulations. [MAY NEED SNAPSHOTS]
294 skuang 3727
295 skuang 3737 After the modified Au-butanethiol surface systems were equilibrated
296     under canonical ensemble, organic solvent molecules were packed in the
297     previously empty part of the simulation cells\cite{packmol}. Two
298     solvents were investigated, one which has little vibrational overlap
299     with the alkanethiol and a planar shape (toluene), and one which has
300     similar vibrational frequencies and chain-like shape ({\it n}-hexane).
301 skuang 3727
302 skuang 3737 The space filled by solvent molecules, i.e. the gap between
303 skuang 3730 periodically repeated Au-butanethiol surfaces should be carefully
304     chosen. A very long length scale for the thermal gradient axis ($z$)
305     may cause excessively hot or cold temperatures in the middle of the
306     solvent region and lead to undesired phenomena such as solvent boiling
307     or freezing when a thermal flux is applied. Conversely, too few
308     solvent molecules would change the normal behavior of the liquid
309     phase. Therefore, our $N_{solvent}$ values were chosen to ensure that
310     these extreme cases did not happen to our simulations. And the
311 skuang 3749 corresponding spacing is usually $35 \sim 75$\AA.
312 skuang 3730
313 skuang 3746 The initial configurations generated are further equilibrated with the
314     $x$ and $y$ dimensions fixed, only allowing length scale change in $z$
315     dimension. This is to ensure that the equilibration of liquid phase
316     does not affect the metal crystal structure in $x$ and $y$ dimensions.
317     To investigate this effect, comparisons were made with simulations
318 skuang 3747 that allow changes of $L_x$ and $L_y$ during NPT equilibration, and
319     the results are shown in later sections. After ensuring the liquid
320     phase reaches equilibrium at atmospheric pressure (1 atm), further
321 skuang 3746 equilibration are followed under NVT and then NVE ensembles.
322 skuang 3728
323 skuang 3727 After the systems reach equilibrium, NIVS is implemented to impose a
324     periodic unphysical thermal flux between the metal and the liquid
325 skuang 3728 phase. Most of our simulations are under an average temperature of
326     $\sim$200K. Therefore, this flux usually comes from the metal to the
327 skuang 3727 liquid so that the liquid has a higher temperature and would not
328 skuang 3747 freeze due to excessively low temperature. After this induced
329     temperature gradient is stablized, the temperature profile of the
330     simulation cell is recorded. To do this, the simulation cell is
331     devided evenly into $N$ slabs along the $z$-axis and $N$ is maximized
332     for highest possible spatial resolution but not too many to have some
333     slabs empty most of the time. The average temperatures of each slab
334     are recorded for 1$\sim$2 ns. When the slab width $d$ of each slab is
335     the same, the derivatives of $T$ with respect to slab number $n$ can
336     be directly used for $G^\prime$ calculations:
337 skuang 3727 \begin{equation}
338     G^\prime = |J_z|\Big|\frac{\partial^2 T}{\partial z^2}\Big|
339     \Big/\left(\frac{\partial T}{\partial z}\right)^2
340     = |J_z|\Big|\frac{1}{d^2}\frac{\partial^2 T}{\partial n^2}\Big|
341     \Big/\left(\frac{1}{d}\frac{\partial T}{\partial n}\right)^2
342     = |J_z|\Big|\frac{\partial^2 T}{\partial n^2}\Big|
343     \Big/\left(\frac{\partial T}{\partial n}\right)^2
344     \label{derivativeG2}
345     \end{equation}
346    
347 skuang 3747 All of the above simulation procedures use a time step of 1 fs. And
348     each equilibration / stabilization step usually takes 100 ps, or
349     longer, if necessary.
350    
351 skuang 3725 \subsection{Force Field Parameters}
352 skuang 3744 Our simulations include various components. Figure \ref{demoMol}
353     demonstrates the sites defined for both United-Atom and All-Atom
354     models of the organic solvent and capping agent molecules in our
355     simulations. Force field parameter descriptions are needed for
356     interactions both between the same type of particles and between
357     particles of different species.
358 skuang 3721
359 skuang 3736 \begin{figure}
360 gezelter 3740 \includegraphics[width=\linewidth]{structures}
361     \caption{Structures of the capping agent and solvents utilized in
362     these simulations. The chemically-distinct sites (a-e) are expanded
363     in terms of constituent atoms for both United Atom (UA) and All Atom
364     (AA) force fields. Most parameters are from
365     Refs. \protect\cite{TraPPE-UA.alkanes,TraPPE-UA.alkylbenzenes} (UA) and
366     \protect\cite{OPLSAA} (AA). Cross-interactions with the Au atoms are given
367     in Table \ref{MnM}.}
368 skuang 3736 \label{demoMol}
369     \end{figure}
370    
371 skuang 3744 The Au-Au interactions in metal lattice slab is described by the
372     quantum Sutton-Chen (QSC) formulation\cite{PhysRevB.59.3527}. The QSC
373     potentials include zero-point quantum corrections and are
374     reparametrized for accurate surface energies compared to the
375     Sutton-Chen potentials\cite{Chen90}.
376    
377 skuang 3728 For both solvent molecules, straight chain {\it n}-hexane and aromatic
378     toluene, United-Atom (UA) and All-Atom (AA) models are used
379     respectively. The TraPPE-UA
380     parameters\cite{TraPPE-UA.alkanes,TraPPE-UA.alkylbenzenes} are used
381 skuang 3744 for our UA solvent molecules. In these models, sites are located at
382     the carbon centers for alkyl groups. Bonding interactions, including
383     bond stretches and bends and torsions, were used for intra-molecular
384     sites not separated by more than 3 bonds. Otherwise, for non-bonded
385 skuang 3747 interactions, Lennard-Jones potentials are used. [CHECK CITATION]
386 skuang 3721
387 skuang 3744 By eliminating explicit hydrogen atoms, these models are simple and
388     computationally efficient, while maintains good accuracy. However, the
389     TraPPE-UA for alkanes is known to predict a lower boiling point than
390     experimental values. Considering that after an unphysical thermal flux
391     is applied to a system, the temperature of ``hot'' area in the liquid
392 skuang 3747 phase would be significantly higher than the average of the system, to
393     prevent over heating and boiling of the liquid phase, the average
394     temperature in our simulations should be much lower than the liquid
395     boiling point.
396 skuang 3744
397     For UA-toluene model, the non-bonded potentials between
398     inter-molecular sites have a similar Lennard-Jones formulation. For
399     intra-molecular interactions, considering the stiffness of the benzene
400     ring, rigid body constraints are applied for further computational
401     efficiency. All bonds in the benzene ring and between the ring and the
402     methyl group remain rigid during the progress of simulations.
403    
404 skuang 3729 Besides the TraPPE-UA models, AA models for both organic solvents are
405 skuang 3730 included in our studies as well. For hexane, the OPLS-AA\cite{OPLSAA}
406 skuang 3744 force field is used. Additional explicit hydrogen sites were
407     included. Besides bonding and non-bonded site-site interactions,
408     partial charges and the electrostatic interactions were added to each
409     CT and HC site. For toluene, the United Force Field developed by
410     Rapp\'{e} {\it et al.}\cite{doi:10.1021/ja00051a040} is
411     adopted. Without the rigid body constraints, bonding interactions were
412     included. For the aromatic ring, improper torsions (inversions) were
413     added as an extra potential for maintaining the planar shape.
414 skuang 3747 [CHECK CITATION]
415 skuang 3728
416 skuang 3729 The capping agent in our simulations, the butanethiol molecules can
417     either use UA or AA model. The TraPPE-UA force fields includes
418 skuang 3730 parameters for thiol molecules\cite{TraPPE-UA.thiols} and are used for
419     UA butanethiol model in our simulations. The OPLS-AA also provides
420     parameters for alkyl thiols. However, alkyl thiols adsorbed on Au(111)
421     surfaces do not have the hydrogen atom bonded to sulfur. To adapt this
422     change and derive suitable parameters for butanethiol adsorbed on
423 skuang 3736 Au(111) surfaces, we adopt the S parameters from Luedtke and
424 skuang 3747 Landman\cite{landman:1998}[CHECK CITATION]
425     and modify parameters for its neighbor C
426 skuang 3736 atom for charge balance in the molecule. Note that the model choice
427     (UA or AA) of capping agent can be different from the
428     solvent. Regardless of model choice, the force field parameters for
429     interactions between capping agent and solvent can be derived using
430 skuang 3738 Lorentz-Berthelot Mixing Rule:
431     \begin{eqnarray}
432 skuang 3742 \sigma_{ij} & = & \frac{1}{2} \left(\sigma_{ii} + \sigma_{jj}\right) \\
433     \epsilon_{ij} & = & \sqrt{\epsilon_{ii}\epsilon_{jj}}
434 skuang 3738 \end{eqnarray}
435 skuang 3721
436     To describe the interactions between metal Au and non-metal capping
437 skuang 3730 agent and solvent particles, we refer to an adsorption study of alkyl
438     thiols on gold surfaces by Vlugt {\it et
439     al.}\cite{vlugt:cpc2007154} They fitted an effective Lennard-Jones
440     form of potential parameters for the interaction between Au and
441     pseudo-atoms CH$_x$ and S based on a well-established and widely-used
442 skuang 3736 effective potential of Hautman and Klein\cite{hautman:4994} for the
443     Au(111) surface. As our simulations require the gold lattice slab to
444     be non-rigid so that it could accommodate kinetic energy for thermal
445 skuang 3730 transport study purpose, the pair-wise form of potentials is
446     preferred.
447 skuang 3721
448 skuang 3730 Besides, the potentials developed from {\it ab initio} calculations by
449     Leng {\it et al.}\cite{doi:10.1021/jp034405s} are adopted for the
450 skuang 3744 interactions between Au and aromatic C/H atoms in toluene. A set of
451     pseudo Lennard-Jones parameters were provided for Au in their force
452     fields. By using the Mixing Rule, this can be used to derive pair-wise
453     potentials for non-bonded interactions between Au and non-metal sites.
454 skuang 3725
455 skuang 3730 However, the Lennard-Jones parameters between Au and other types of
456 skuang 3744 particles, such as All-Atom normal alkanes in our simulations are not
457     yet well-established. For these interactions, we attempt to derive
458     their parameters using the Mixing Rule. To do this, Au pseudo
459     Lennard-Jones parameters for ``Metal-non-Metal'' (MnM) interactions
460     were first extracted from the Au-CH$_x$ parameters by applying the
461     Mixing Rule reversely. Table \ref{MnM} summarizes these ``MnM''
462 skuang 3730 parameters in our simulations.
463 skuang 3729
464 skuang 3730 \begin{table*}
465     \begin{minipage}{\linewidth}
466     \begin{center}
467 gezelter 3741 \caption{Non-bonded interaction parameters (including cross
468     interactions with Au atoms) for both force fields used in this
469     work.}
470     \begin{tabular}{lllllll}
471 skuang 3730 \hline\hline
472 gezelter 3741 & Site & $\sigma_{ii}$ & $\epsilon_{ii}$ & $q_i$ &
473     $\sigma_{Au-i}$ & $\epsilon_{Au-i}$ \\
474     & & (\AA) & (kcal/mol) & ($e$) & (\AA) & (kcal/mol) \\
475 skuang 3730 \hline
476 gezelter 3741 United Atom (UA)
477     &CH3 & 3.75 & 0.1947 & - & 3.54 & 0.2146 \\
478     &CH2 & 3.95 & 0.0914 & - & 3.54 & 0.1749 \\
479     &CHar & 3.695 & 0.1003 & - & 3.4625 & 0.1680 \\
480     &CRar & 3.88 & 0.04173 & - & 3.555 & 0.1604 \\
481     \hline
482     All Atom (AA)
483     &CT3 & 3.50 & 0.066 & -0.18 & 3.365 & 0.1373 \\
484     &CT2 & 3.50 & 0.066 & -0.12 & 3.365 & 0.1373 \\
485     &CTT & 3.50 & 0.066 & -0.065 & 3.365 & 0.1373 \\
486     &HC & 2.50 & 0.030 & 0.06 & 2.865 & 0.09256 \\
487     &CA & 3.55 & 0.070 & -0.115 & 3.173 & 0.0640 \\
488     &HA & 2.42 & 0.030 & 0.115 & 2.746 & 0.0414 \\
489     \hline
490 skuang 3744 Both UA and AA
491     & S & 4.45 & 0.25 & - & 2.40 & 8.465 \\
492 skuang 3730 \hline\hline
493     \end{tabular}
494     \label{MnM}
495     \end{center}
496     \end{minipage}
497     \end{table*}
498 skuang 3729
499 skuang 3746 \subsection{Vibrational Spectrum}
500 skuang 3747 To investigate the mechanism of interfacial thermal conductance, the
501     vibrational spectrum is utilized as a complementary tool. Vibrational
502     spectra were taken for individual components in different
503     simulations. To obtain these spectra, simulations were run after
504     equilibration, in the NVE ensemble. Snapshots of configurations were
505     collected at a frequency that is higher than that of the fastest
506     vibrations occuring in the simulations. With these configurations, the
507     velocity auto-correlation functions can be computed:
508     \begin{equation}
509     C_A (t) = \langle\vec{v}_A (t)\cdot\vec{v}_A (0)\rangle
510     \label{vCorr}
511     \end{equation}
512     Followed by Fourier transforms, the power spectrum can be constructed:
513     \begin{equation}
514     \hat{f}(\omega) = \int_{-\infty}^{\infty} C_A (t) e^{-2\pi it\omega}\,dt
515     \label{fourier}
516     \end{equation}
517 skuang 3746
518 skuang 3730 \section{Results and Discussions}
519 skuang 3747 In what follows, how the parameters and protocol of simulations would
520     affect the measurement of $G$'s is first discussed. With a reliable
521     protocol and set of parameters, the influence of capping agent
522     coverage on thermal conductance is investigated. Besides, different
523     force field models for both solvents and selected deuterated models
524     were tested and compared. Finally, a summary of the role of capping
525     agent in the interfacial thermal transport process is given.
526    
527 skuang 3730 \subsection{How Simulation Parameters Affects $G$}
528     We have varied our protocol or other parameters of the simulations in
529     order to investigate how these factors would affect the measurement of
530     $G$'s. It turned out that while some of these parameters would not
531     affect the results substantially, some other changes to the
532     simulations would have a significant impact on the measurement
533     results.
534 skuang 3725
535 skuang 3730 In some of our simulations, we allowed $L_x$ and $L_y$ to change
536 skuang 3744 during equilibrating the liquid phase. Due to the stiffness of the
537     crystalline Au structure, $L_x$ and $L_y$ would not change noticeably
538     after equilibration. Although $L_z$ could fluctuate $\sim$1\% after a
539     system is fully equilibrated in the NPT ensemble, this fluctuation, as
540     well as those of $L_x$ and $L_y$ (which is significantly smaller),
541     would not be magnified on the calculated $G$'s, as shown in Table
542     \ref{AuThiolHexaneUA}. This insensivity to $L_i$ fluctuations allows
543     reliable measurement of $G$'s without the necessity of extremely
544     cautious equilibration process.
545 skuang 3725
546 skuang 3730 As stated in our computational details, the spacing filled with
547     solvent molecules can be chosen within a range. This allows some
548     change of solvent molecule numbers for the same Au-butanethiol
549     surfaces. We did this study on our Au-butanethiol/hexane
550     simulations. Nevertheless, the results obtained from systems of
551     different $N_{hexane}$ did not indicate that the measurement of $G$ is
552     susceptible to this parameter. For computational efficiency concern,
553     smaller system size would be preferable, given that the liquid phase
554     structure is not affected.
555    
556     Our NIVS algorithm allows change of unphysical thermal flux both in
557     direction and in quantity. This feature extends our investigation of
558     interfacial thermal conductance. However, the magnitude of this
559     thermal flux is not arbitary if one aims to obtain a stable and
560     reliable thermal gradient. A temperature profile would be
561     substantially affected by noise when $|J_z|$ has a much too low
562     magnitude; while an excessively large $|J_z|$ that overwhelms the
563     conductance capacity of the interface would prevent a thermal gradient
564     to reach a stablized steady state. NIVS has the advantage of allowing
565     $J$ to vary in a wide range such that the optimal flux range for $G$
566     measurement can generally be simulated by the algorithm. Within the
567     optimal range, we were able to study how $G$ would change according to
568     the thermal flux across the interface. For our simulations, we denote
569     $J_z$ to be positive when the physical thermal flux is from the liquid
570     to metal, and negative vice versa. The $G$'s measured under different
571 skuang 3744 $J_z$ is listed in Table \ref{AuThiolHexaneUA} and
572     \ref{AuThiolToluene}. These results do not suggest that $G$ is
573     dependent on $J_z$ within this flux range. The linear response of flux
574     to thermal gradient simplifies our investigations in that we can rely
575     on $G$ measurement with only a couple $J_z$'s and do not need to test
576     a large series of fluxes.
577 skuang 3730
578 skuang 3725 \begin{table*}
579     \begin{minipage}{\linewidth}
580     \begin{center}
581     \caption{Computed interfacial thermal conductivity ($G$ and
582 skuang 3731 $G^\prime$) values for the 100\% covered Au-butanethiol/hexane
583     interfaces with UA model and different hexane molecule numbers
584 skuang 3745 at different temperatures using a range of energy
585     fluxes. Error estimates indicated in parenthesis.}
586 skuang 3730
587 skuang 3738 \begin{tabular}{ccccccc}
588 skuang 3730 \hline\hline
589 skuang 3738 $\langle T\rangle$ & $N_{hexane}$ & Fixed & $\rho_{hexane}$ &
590     $J_z$ & $G$ & $G^\prime$ \\
591     (K) & & $L_x$ \& $L_y$? & (g/cm$^3$) & (GW/m$^2$) &
592 skuang 3730 \multicolumn{2}{c}{(MW/m$^2$/K)} \\
593     \hline
594 skuang 3745 200 & 266 & No & 0.672 & -0.96 & 102(3) & 80.0(0.8) \\
595 skuang 3743 & 200 & Yes & 0.694 & 1.92 & 129(11) & 87.3(0.3) \\
596     & & Yes & 0.672 & 1.93 & 131(16) & 78(13) \\
597 skuang 3745 & & No & 0.688 & 0.96 & 125(16) & 90.2(15) \\
598 skuang 3743 & & & & 1.91 & 139(10) & 101(10) \\
599     & & & & 2.83 & 141(6) & 89.9(9.8) \\
600     & 166 & Yes & 0.679 & 0.97 & 115(19) & 69(18) \\
601     & & & & 1.94 & 125(9) & 87.1(0.2) \\
602     & & No & 0.681 & 0.97 & 141(30) & 78(22) \\
603     & & & & 1.92 & 138(4) & 98.9(9.5) \\
604 skuang 3739 \hline
605 skuang 3743 250 & 200 & No & 0.560 & 0.96 & 75(10) & 61.8(7.3) \\
606     & & & & -0.95 & 49.4(0.3) & 45.7(2.1) \\
607     & 166 & Yes & 0.570 & 0.98 & 79.0(3.5) & 62.9(3.0) \\
608     & & No & 0.569 & 0.97 & 80.3(0.6) & 67(11) \\
609     & & & & 1.44 & 76.2(5.0) & 64.8(3.8) \\
610     & & & & -0.95 & 56.4(2.5) & 54.4(1.1) \\
611     & & & & -1.85 & 47.8(1.1) & 53.5(1.5) \\
612 skuang 3730 \hline\hline
613     \end{tabular}
614     \label{AuThiolHexaneUA}
615     \end{center}
616     \end{minipage}
617     \end{table*}
618    
619     Furthermore, we also attempted to increase system average temperatures
620     to above 200K. These simulations are first equilibrated in the NPT
621     ensemble under normal pressure. As stated above, the TraPPE-UA model
622     for hexane tends to predict a lower boiling point. In our simulations,
623     hexane had diffculty to remain in liquid phase when NPT equilibration
624     temperature is higher than 250K. Additionally, the equilibrated liquid
625     hexane density under 250K becomes lower than experimental value. This
626     expanded liquid phase leads to lower contact between hexane and
627 skuang 3744 butanethiol as well.[MAY NEED SLAB DENSITY FIGURE]
628     And this reduced contact would
629 skuang 3730 probably be accountable for a lower interfacial thermal conductance,
630     as shown in Table \ref{AuThiolHexaneUA}.
631    
632     A similar study for TraPPE-UA toluene agrees with the above result as
633     well. Having a higher boiling point, toluene tends to remain liquid in
634     our simulations even equilibrated under 300K in NPT
635     ensembles. Furthermore, the expansion of the toluene liquid phase is
636     not as significant as that of the hexane. This prevents severe
637     decrease of liquid-capping agent contact and the results (Table
638     \ref{AuThiolToluene}) show only a slightly decreased interface
639     conductance. Therefore, solvent-capping agent contact should play an
640     important role in the thermal transport process across the interface
641     in that higher degree of contact could yield increased conductance.
642    
643     \begin{table*}
644     \begin{minipage}{\linewidth}
645     \begin{center}
646     \caption{Computed interfacial thermal conductivity ($G$ and
647 skuang 3731 $G^\prime$) values for a 90\% coverage Au-butanethiol/toluene
648     interface at different temperatures using a range of energy
649 skuang 3745 fluxes. Error estimates indicated in parenthesis.}
650 skuang 3725
651 skuang 3738 \begin{tabular}{ccccc}
652 skuang 3725 \hline\hline
653 skuang 3738 $\langle T\rangle$ & $\rho_{toluene}$ & $J_z$ & $G$ & $G^\prime$ \\
654     (K) & (g/cm$^3$) & (GW/m$^2$) & \multicolumn{2}{c}{(MW/m$^2$/K)} \\
655 skuang 3725 \hline
656 skuang 3745 200 & 0.933 & 2.15 & 204(12) & 113(12) \\
657     & & -1.86 & 180(3) & 135(21) \\
658     & & -3.93 & 176(5) & 113(12) \\
659 skuang 3738 \hline
660 skuang 3745 300 & 0.855 & -1.91 & 143(5) & 125(2) \\
661     & & -4.19 & 135(9) & 113(12) \\
662 skuang 3725 \hline\hline
663     \end{tabular}
664     \label{AuThiolToluene}
665     \end{center}
666     \end{minipage}
667     \end{table*}
668    
669 skuang 3730 Besides lower interfacial thermal conductance, surfaces in relatively
670     high temperatures are susceptible to reconstructions, when
671     butanethiols have a full coverage on the Au(111) surface. These
672     reconstructions include surface Au atoms migrated outward to the S
673     atom layer, and butanethiol molecules embedded into the original
674     surface Au layer. The driving force for this behavior is the strong
675     Au-S interactions in our simulations. And these reconstructions lead
676     to higher ratio of Au-S attraction and thus is energetically
677     favorable. Furthermore, this phenomenon agrees with experimental
678     results\cite{doi:10.1021/j100035a033,doi:10.1021/la026493y}. Vlugt
679     {\it et al.} had kept their Au(111) slab rigid so that their
680     simulations can reach 300K without surface reconstructions. Without
681     this practice, simulating 100\% thiol covered interfaces under higher
682     temperatures could hardly avoid surface reconstructions. However, our
683     measurement is based on assuming homogeneity on $x$ and $y$ dimensions
684     so that measurement of $T$ at particular $z$ would be an effective
685     average of the particles of the same type. Since surface
686     reconstructions could eliminate the original $x$ and $y$ dimensional
687     homogeneity, measurement of $G$ is more difficult to conduct under
688     higher temperatures. Therefore, most of our measurements are
689 skuang 3732 undertaken at $\langle T\rangle\sim$200K.
690 skuang 3725
691 skuang 3730 However, when the surface is not completely covered by butanethiols,
692     the simulated system is more resistent to the reconstruction
693 skuang 3744 above. Our Au-butanethiol/toluene system had the Au(111) surfaces 90\%
694     covered by butanethiols, but did not see this above phenomena even at
695     $\langle T\rangle\sim$300K. The empty three-fold sites not occupied by
696     capping agents could help prevent surface reconstruction in that they
697     provide other means of capping agent relaxation. It is observed that
698 skuang 3738 butanethiols can migrate to their neighbor empty sites during a
699     simulation. Therefore, we were able to obtain $G$'s for these
700     interfaces even at a relatively high temperature without being
701     affected by surface reconstructions.
702 skuang 3725
703 skuang 3730 \subsection{Influence of Capping Agent Coverage on $G$}
704     To investigate the influence of butanethiol coverage on interfacial
705     thermal conductance, a series of different coverage Au-butanethiol
706     surfaces is prepared and solvated with various organic
707     molecules. These systems are then equilibrated and their interfacial
708 skuang 3744 thermal conductivity are measured with our NIVS algorithm. Figure
709     \ref{coverage} demonstrates the trend of conductance change with
710     respect to different coverages of butanethiol. To study the isotope
711     effect in interfacial thermal conductance, deuterated UA-hexane is
712     included as well.
713 skuang 3730
714 skuang 3748 \begin{figure}
715     \includegraphics[width=\linewidth]{coverage}
716     \caption{Comparison of interfacial thermal conductivity ($G$) values
717     for the Au-butanethiol/solvent interface with various UA models and
718     different capping agent coverages at $\langle T\rangle\sim$200K
719     using certain energy flux respectively.}
720     \label{coverage}
721     \end{figure}
722    
723 skuang 3731 It turned out that with partial covered butanethiol on the Au(111)
724 skuang 3744 surface, the derivative definition for $G^\prime$
725     (Eq. \ref{derivativeG}) was difficult to apply, due to the difficulty
726     in locating the maximum of change of $\lambda$. Instead, the discrete
727     definition (Eq. \ref{discreteG}) is easier to apply, as the Gibbs
728     deviding surface can still be well-defined. Therefore, $G$ (not
729     $G^\prime$) was used for this section.
730 skuang 3725
731 skuang 3744 From Figure \ref{coverage}, one can see the significance of the
732 skuang 3731 presence of capping agents. Even when a fraction of the Au(111)
733     surface sites are covered with butanethiols, the conductivity would
734     see an enhancement by at least a factor of 3. This indicates the
735     important role cappping agent is playing for thermal transport
736 skuang 3744 phenomena on metal / organic solvent surfaces.
737 skuang 3725
738 skuang 3731 Interestingly, as one could observe from our results, the maximum
739     conductance enhancement (largest $G$) happens while the surfaces are
740     about 75\% covered with butanethiols. This again indicates that
741     solvent-capping agent contact has an important role of the thermal
742     transport process. Slightly lower butanethiol coverage allows small
743     gaps between butanethiols to form. And these gaps could be filled with
744     solvent molecules, which acts like ``heat conductors'' on the
745     surface. The higher degree of interaction between these solvent
746     molecules and capping agents increases the enhancement effect and thus
747     produces a higher $G$ than densely packed butanethiol arrays. However,
748     once this maximum conductance enhancement is reached, $G$ decreases
749     when butanethiol coverage continues to decrease. Each capping agent
750     molecule reaches its maximum capacity for thermal
751     conductance. Therefore, even higher solvent-capping agent contact
752     would not offset this effect. Eventually, when butanethiol coverage
753     continues to decrease, solvent-capping agent contact actually
754     decreases with the disappearing of butanethiol molecules. In this
755 skuang 3744 case, $G$ decrease could not be offset but instead accelerated. [NEED
756 skuang 3746 SNAPSHOT SHOWING THE PHENOMENA / SLAB DENSITY ANALYSIS]
757 skuang 3725
758 skuang 3731 A comparison of the results obtained from differenet organic solvents
759     can also provide useful information of the interfacial thermal
760     transport process. The deuterated hexane (UA) results do not appear to
761     be much different from those of normal hexane (UA), given that
762     butanethiol (UA) is non-deuterated for both solvents. These UA model
763     studies, even though eliminating C-H vibration samplings, still have
764     C-C vibrational frequencies different from each other. However, these
765 skuang 3732 differences in the infrared range do not seem to produce an observable
766 skuang 3748 difference for the results of $G$ (Figure \ref{uahxnua}).
767 skuang 3730
768 skuang 3748 \begin{figure}
769     \includegraphics[width=\linewidth]{uahxnua}
770     \caption{Vibrational spectra obtained for normal (upper) and
771     deuterated (lower) hexane in Au-butanethiol/hexane
772     systems. Butanethiol spectra are shown as reference. Both hexane and
773     butanethiol were using United-Atom models.}
774     \label{uahxnua}
775     \end{figure}
776    
777 skuang 3731 Furthermore, results for rigid body toluene solvent, as well as other
778     UA-hexane solvents, are reasonable within the general experimental
779 skuang 3749 ranges\cite{Wilson:2002uq,cahill:793,PhysRevB.80.195406}. This
780     suggests that explicit hydrogen might not be a required factor for
781     modeling thermal transport phenomena of systems such as
782     Au-thiol/organic solvent.
783 skuang 3731
784     However, results for Au-butanethiol/toluene do not show an identical
785 skuang 3744 trend with those for Au-butanethiol/hexane in that $G$ remains at
786 skuang 3731 approximately the same magnitue when butanethiol coverage differs from
787     25\% to 75\%. This might be rooted in the molecule shape difference
788 skuang 3744 for planar toluene and chain-like {\it n}-hexane. Due to this
789 skuang 3731 difference, toluene molecules have more difficulty in occupying
790     relatively small gaps among capping agents when their coverage is not
791     too low. Therefore, the solvent-capping agent contact may keep
792     increasing until the capping agent coverage reaches a relatively low
793     level. This becomes an offset for decreasing butanethiol molecules on
794     its effect to the process of interfacial thermal transport. Thus, one
795     can see a plateau of $G$ vs. butanethiol coverage in our results.
796    
797 skuang 3730 \subsection{Influence of Chosen Molecule Model on $G$}
798 skuang 3732 In addition to UA solvent/capping agent models, AA models are included
799     in our simulations as well. Besides simulations of the same (UA or AA)
800     model for solvent and capping agent, different models can be applied
801     to different components. Furthermore, regardless of models chosen,
802     either the solvent or the capping agent can be deuterated, similar to
803     the previous section. Table \ref{modelTest} summarizes the results of
804     these studies.
805 skuang 3725
806     \begin{table*}
807     \begin{minipage}{\linewidth}
808     \begin{center}
809    
810     \caption{Computed interfacial thermal conductivity ($G$ and
811 skuang 3732 $G^\prime$) values for interfaces using various models for
812     solvent and capping agent (or without capping agent) at
813 skuang 3739 $\langle T\rangle\sim$200K. (D stands for deuterated solvent
814     or capping agent molecules; ``Avg.'' denotes results that are
815 skuang 3742 averages of simulations under different $J_z$'s. Error
816     estimates indicated in parenthesis.)}
817 skuang 3725
818 skuang 3742 \begin{tabular}{llccc}
819 skuang 3725 \hline\hline
820 skuang 3732 Butanethiol model & Solvent & $J_z$ & $G$ & $G^\prime$ \\
821     (or bare surface) & model & (GW/m$^2$) &
822     \multicolumn{2}{c}{(MW/m$^2$/K)} \\
823 skuang 3725 \hline
824 skuang 3742 UA & UA hexane & Avg. & 131(9) & 87(10) \\
825     & UA hexane(D) & 1.95 & 153(5) & 136(13) \\
826     & AA hexane & Avg. & 131(6) & 122(10) \\
827     & UA toluene & 1.96 & 187(16) & 151(11) \\
828     & AA toluene & 1.89 & 200(36) & 149(53) \\
829 skuang 3739 \hline
830 skuang 3742 AA & UA hexane & 1.94 & 116(9) & 129(8) \\
831     & AA hexane & Avg. & 442(14) & 356(31) \\
832     & AA hexane(D) & 1.93 & 222(12) & 234(54) \\
833     & UA toluene & 1.98 & 125(25) & 97(60) \\
834     & AA toluene & 3.79 & 487(56) & 290(42) \\
835 skuang 3739 \hline
836 skuang 3742 AA(D) & UA hexane & 1.94 & 158(25) & 172(4) \\
837     & AA hexane & 1.92 & 243(29) & 191(11) \\
838     & AA toluene & 1.93 & 364(36) & 322(67) \\
839 skuang 3739 \hline
840 skuang 3742 bare & UA hexane & Avg. & 46.5(3.2) & 49.4(4.5) \\
841     & UA hexane(D) & 0.98 & 43.9(4.6) & 43.0(2.0) \\
842     & AA hexane & 0.96 & 31.0(1.4) & 29.4(1.3) \\
843     & UA toluene & 1.99 & 70.1(1.3) & 65.8(0.5) \\
844 skuang 3725 \hline\hline
845     \end{tabular}
846 skuang 3732 \label{modelTest}
847 skuang 3725 \end{center}
848     \end{minipage}
849     \end{table*}
850    
851 skuang 3732 To facilitate direct comparison, the same system with differnt models
852     for different components uses the same length scale for their
853     simulation cells. Without the presence of capping agent, using
854     different models for hexane yields similar results for both $G$ and
855     $G^\prime$, and these two definitions agree with eath other very
856     well. This indicates very weak interaction between the metal and the
857     solvent, and is a typical case for acoustic impedance mismatch between
858     these two phases.
859 skuang 3730
860 skuang 3732 As for Au(111) surfaces completely covered by butanethiols, the choice
861     of models for capping agent and solvent could impact the measurement
862     of $G$ and $G^\prime$ quite significantly. For Au-butanethiol/hexane
863     interfaces, using AA model for both butanethiol and hexane yields
864     substantially higher conductivity values than using UA model for at
865     least one component of the solvent and capping agent, which exceeds
866 skuang 3744 the general range of experimental measurement results. This is
867     probably due to the classically treated C-H vibrations in the AA
868     model, which should not be appreciably populated at normal
869     temperatures. In comparison, once either the hexanes or the
870     butanethiols are deuterated, one can see a significantly lower $G$ and
871     $G^\prime$. In either of these cases, the C-H(D) vibrational overlap
872 skuang 3748 between the solvent and the capping agent is removed (Figure
873     \ref{aahxntln}). Conclusively, the improperly treated C-H vibration in
874     the AA model produced over-predicted results accordingly. Compared to
875     the AA model, the UA model yields more reasonable results with higher
876     computational efficiency.
877 skuang 3731
878 skuang 3748 \begin{figure}
879     \includegraphics[width=\linewidth]{aahxntln}
880     \caption{Spectra obtained for All-Atom model Au-butanethil/solvent
881     systems. When butanethiol is deuterated (lower left), its
882     vibrational overlap with hexane would decrease significantly,
883     compared with normal butanethiol (upper left). However, this
884     dramatic change does not apply to toluene as much (right).}
885     \label{aahxntln}
886     \end{figure}
887    
888 skuang 3732 However, for Au-butanethiol/toluene interfaces, having the AA
889     butanethiol deuterated did not yield a significant change in the
890 skuang 3739 measurement results. Compared to the C-H vibrational overlap between
891     hexane and butanethiol, both of which have alkyl chains, that overlap
892     between toluene and butanethiol is not so significant and thus does
893 skuang 3749 not have as much contribution to the heat exchange
894     process. Conversely, extra degrees of freedom such as the C-H
895     vibrations could yield higher heat exchange rate between these two
896     phases and result in a much higher conductivity.
897 skuang 3731
898 skuang 3732 Although the QSC model for Au is known to predict an overly low value
899 skuang 3738 for bulk metal gold conductivity\cite{kuang:164101}, our computational
900 skuang 3732 results for $G$ and $G^\prime$ do not seem to be affected by this
901 skuang 3739 drawback of the model for metal. Instead, our results suggest that the
902     modeling of interfacial thermal transport behavior relies mainly on
903     the accuracy of the interaction descriptions between components
904     occupying the interfaces.
905 skuang 3732
906 skuang 3746 \subsection{Role of Capping Agent in Interfacial Thermal Conductance}
907 skuang 3747 The vibrational spectra for gold slabs in different environments are
908     shown as in Figure \ref{specAu}. Regardless of the presence of
909     solvent, the gold surfaces covered by butanethiol molecules, compared
910     to bare gold surfaces, exhibit an additional peak observed at the
911     frequency of $\sim$170cm$^{-1}$, which is attributed to the S-Au
912     bonding vibration. This vibration enables efficient thermal transport
913     from surface Au layer to the capping agents. Therefore, in our
914     simulations, the Au/S interfaces do not appear major heat barriers
915     compared to the butanethiol / solvent interfaces.
916 skuang 3732
917 skuang 3747 Simultaneously, the vibrational overlap between butanethiol and
918     organic solvents suggests higher thermal exchange efficiency between
919     these two components. Even exessively high heat transport was observed
920     when All-Atom models were used and C-H vibrations were treated
921     classically. Compared to metal and organic liquid phase, the heat
922     transfer efficiency between butanethiol and organic solvents is closer
923     to that within bulk liquid phase.
924    
925 skuang 3749 Furthermore, our observation validated previous
926     results\cite{hase:2010} that the intramolecular heat transport of
927     alkylthiols is highly effecient. As a combinational effects of these
928     phenomena, butanethiol acts as a channel to expedite thermal transport
929     process. The acoustic impedance mismatch between the metal and the
930     liquid phase can be effectively reduced with the presence of suitable
931     capping agents.
932 skuang 3747
933 skuang 3725 \begin{figure}
934     \includegraphics[width=\linewidth]{vibration}
935     \caption{Vibrational spectra obtained for gold in different
936 skuang 3745 environments.}
937 skuang 3747 \label{specAu}
938 skuang 3725 \end{figure}
939    
940 skuang 3747 [MAY ADD COMPARISON OF AU SLAB WIDTHS]
941 skuang 3732
942 skuang 3730 \section{Conclusions}
943 skuang 3732 The NIVS algorithm we developed has been applied to simulations of
944     Au-butanethiol surfaces with organic solvents. This algorithm allows
945     effective unphysical thermal flux transferred between the metal and
946     the liquid phase. With the flux applied, we were able to measure the
947     corresponding thermal gradient and to obtain interfacial thermal
948 skuang 3747 conductivities. Under steady states, single trajectory simulation
949     would be enough for accurate measurement. This would be advantageous
950     compared to transient state simulations, which need multiple
951     trajectories to produce reliable average results.
952    
953     Our simulations have seen significant conductance enhancement with the
954     presence of capping agent, compared to the bare gold / liquid
955     interfaces. The acoustic impedance mismatch between the metal and the
956     liquid phase is effectively eliminated by proper capping
957 skuang 3732 agent. Furthermore, the coverage precentage of the capping agent plays
958 skuang 3747 an important role in the interfacial thermal transport
959     process. Moderately lower coverages allow higher contact between
960     capping agent and solvent, and thus could further enhance the heat
961     transfer process.
962 skuang 3725
963 skuang 3732 Our measurement results, particularly of the UA models, agree with
964     available experimental data. This indicates that our force field
965     parameters have a nice description of the interactions between the
966     particles at the interfaces. AA models tend to overestimate the
967     interfacial thermal conductance in that the classically treated C-H
968     vibration would be overly sampled. Compared to the AA models, the UA
969     models have higher computational efficiency with satisfactory
970     accuracy, and thus are preferable in interfacial thermal transport
971 skuang 3747 modelings. Of the two definitions for $G$, the discrete form
972     (Eq. \ref{discreteG}) was easier to use and gives out relatively
973     consistent results, while the derivative form (Eq. \ref{derivativeG})
974     is not as versatile. Although $G^\prime$ gives out comparable results
975     and follows similar trend with $G$ when measuring close to fully
976     covered or bare surfaces, the spatial resolution of $T$ profile is
977     limited for accurate computation of derivatives data.
978 skuang 3730
979 skuang 3732 Vlugt {\it et al.} has investigated the surface thiol structures for
980     nanocrystal gold and pointed out that they differs from those of the
981     Au(111) surface\cite{vlugt:cpc2007154}. This difference might lead to
982     change of interfacial thermal transport behavior as well. To
983     investigate this problem, an effective means to introduce thermal flux
984     and measure the corresponding thermal gradient is desirable for
985     simulating structures with spherical symmetry.
986 skuang 3730
987 gezelter 3717 \section{Acknowledgments}
988     Support for this project was provided by the National Science
989     Foundation under grant CHE-0848243. Computational time was provided by
990     the Center for Research Computing (CRC) at the University of Notre
991 skuang 3730 Dame. \newpage
992 gezelter 3717
993     \bibliography{interfacial}
994    
995     \end{doublespace}
996     \end{document}
997