ViewVC Help
View File | Revision Log | Show Annotations | View Changeset | Root Listing
root/group/trunk/chainLength/GoldThiolsPaper.tex
Revision: 3857
Committed: Mon Feb 11 23:12:08 2013 UTC (12 years, 5 months ago) by kstocke1
Content type: application/x-tex
File size: 41203 byte(s)
Log Message:

File Contents

# Content
1 \documentclass[11pt]{article}
2 \usepackage{amsmath}
3 \usepackage{amssymb}
4 \usepackage{times}
5 \usepackage{mathptm}
6 \usepackage{setspace}
7 \usepackage{endfloat}
8 \usepackage{caption}
9 \usepackage{tabularx}
10 \usepackage{longtable}
11 \usepackage{graphicx}
12 \usepackage{multirow}
13 \usepackage{multicol}
14 \usepackage[version=3]{mhchem} % this is a great package for formatting chemical reactions
15 \usepackage[square, comma, sort&compress]{natbib}
16 \usepackage{url}
17 \pagestyle{plain} \pagenumbering{arabic} \oddsidemargin 0.0cm
18 \evensidemargin 0.0cm \topmargin -21pt \headsep 10pt \textheight
19 9.0in \textwidth 6.5in \brokenpenalty=10000
20
21 % double space list of tables and figures
22 %\AtBeginDelayedFloats{\renewcomand{\baselinestretch}{1.66}}
23 \setlength{\abovecaptionskip}{20 pt}
24 \setlength{\belowcaptionskip}{30 pt}
25
26 \bibpunct{}{}{,}{s}{}{;}
27
28 \citestyle{nature}
29 \bibliographystyle{achemso}
30
31 \begin{document}
32
33 \newcolumntype{A}{p{1.5in}}
34 \newcolumntype{B}{p{0.75in}}
35
36 \title{Simulations of Heat Conduction at Thiolate-Capped Gold
37 Surfaces: The Role of Chain Length and Solvent Penetration}
38
39 \author{Kelsey M. Stocker and J. Daniel
40 Gezelter\footnote{Corresponding author. \ Electronic mail:
41 gezelter@nd.edu} \\
42 251 Nieuwland Science Hall, \\
43 Department of Chemistry and Biochemistry,\\
44 University of Notre Dame\\
45 Notre Dame, Indiana 46556}
46
47 \date{\today}
48
49 \maketitle
50
51 \begin{doublespace}
52
53 \begin{abstract}
54
55 We report on simulations of heat conduction through Au(111) / hexane
56 interfaces in which the surface has been protected by a mixture of
57 short and long chain alkanethiolate ligands. Reverse
58 non-equilibrium molecular dynamics (RNEMD) was used to create a
59 thermal flux between the metal and solvent, and thermal conductance
60 was computed using the resulting thermal profiles near the
61 interface. We find a non-monotonic dependence of the interfacial
62 thermal conductance on the fraction of long chains present at the
63 interface, and correlate this behavior to both solvent ordering and
64 the rate of solvent escape from the thiolate layer immediately in
65 contact with the metal surface. Our results suggest that a mixed
66 vibrational transfer / convection model is necessary to explain the
67 features of heat transfer at this interface. The alignment of the
68 solvent chains with the ordered ligand allows rapid transfer of
69 energy to the trapped solvent and is the dominant feature for
70 ordered ligand layers. Diffusion of the vibrationally excited
71 solvent into the bulk also plays a significant role when the ligands
72 are less tightly packed.
73
74 \end{abstract}
75
76 \newpage
77
78 %\narrowtext
79
80 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
81 % **INTRODUCTION**
82 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
83 \section{Introduction}
84
85 The structural details of the interfaces of metal nanoparticles
86 determine how energy flows between these particles and their
87 surroundings. Understanding this energy flow is essential in designing
88 and functionalizing metallic nanoparticles for use in plasmonic photothermal
89 therapies,\cite{Jain:2007ux,Petrova:2007ad,Gnyawali:2008lp,Mazzaglia:2008to,Huff:2007ye,Larson:2007hw}
90 which rely on the ability of metallic nanoparticles to absorb light in
91 the near-IR, a portion of the spectrum in which living tissue is very
92 nearly transparent. The relevant physical property controlling the
93 transfer of this energy as heat into the surrounding tissue is the
94 interfacial thermal conductance, $G$, which can be somewhat difficult
95 to determine experimentally.\cite{Wilson:2002uq,Plech:2005kx}
96
97 Metallic particles have also been proposed for use in efficient
98 thermal-transfer fluids, although careful experiments by Eapen {\it et
99 al.} have shown that metal-particle-based nanofluids have thermal
100 conductivities that match Maxwell predictions.\cite{Eapen:2007th} The
101 likely cause of previously reported non-Maxwell
102 behavior\cite{Eastman:2001wb,Keblinski:2002bx,Lee:1999ct,Xue:2003ya,Xue:2004oa}
103 is percolation networks of nanoparticles exchanging energy via the
104 solvent,\cite{Eapen:2007mw} so it is important to get a detailed
105 molecular picture of particle-ligand and ligand-solvent interactions
106 in order to understand the thermal behavior of complex fluids. To
107 date, there have been few reported values (either from theory or
108 experiment) of $G$ for ligand-protected nanoparticles embedded in
109 liquids, and there is a significant gap in knowledge about how
110 chemically distinct ligands or protecting groups will affect heat
111 transport from the particles.
112
113 Experimentally, the thermal properties of various nanostructured
114 interfaces have been investigated by a number of groups. Cahill and
115 coworkers studied thermal transport from metal nanoparticle/fluid
116 interfaces, epitaxial TiN/single crystal oxide interfaces, and
117 hydrophilic and hydrophobic interfaces between water and solids with
118 different self-assembled
119 monolayers.\cite{cahill:793,Wilson:2002uq,PhysRevB.67.054302,doi:10.1021/jp048375k,PhysRevLett.96.186101} Schmidt {\it et al.} studied the role of
120 cetyltrimethylammonium bromide (CTAB) on the thermal transport between
121 gold nanorods and solvent.\cite{doi:10.1021/jp8051888}
122 Wang {\it et al.} studied heat transport through long-chain
123 hydrocarbon monolayers on unsolvated gold substrate at the individual molecular
124 level.\cite{Wang10082007} The introduction of solvent adds yet another interface and potential barrier for heat transfer. Juv\'e {\it
125 et al.} studied the cooling dynamics of glass-embedded nanoparticles, which is controlled by thermal
126 interfacial resistance and heat diffusion in the matrix.\cite{PhysRevB.80.195406} Hartland also notes the importance of heat transfer through diffusive solvent behavior.\cite{hartland2011} Although interfaces are
127 normally considered barriers for heat transport, Alper {\it et al.}
128 have suggested that specific ligands (capping agents) could completely
129 eliminate this barrier
130 ($G\rightarrow\infty$).\cite{doi:10.1021/la904855s}
131
132 Recently, Hase and coworkers employed Non-Equilibrium Molecular
133 Dynamics (NEMD) simulations to study thermal transport from hot
134 Au(111) substrate to a self-assembled monolayer of alkylthiol with
135 relatively long chain (8-20 carbon atoms).\cite{hase:2010,hase:2011}
136 These simulations explained many of the features of the experiments of
137 Wang {\it et al.} However, ensemble averaged measurements for heat
138 conductance of interfaces between the capping monolayer on Au and a
139 solvent phase have yet to be studied with their approach. In previous
140 simulations, our group applied a variant of reverse non-equilibrium
141 molecular dynamics (RNEMD) to calculate the interfacial thermal
142 conductance at a metal / organic solvent interface that had been
143 chemically protected by butanethiolate groups.\cite{kuang:AuThl} Our
144 calculations suggested an explanation for the very large thermal
145 conductivity at alkanethiol-capped metal surfaces when compared with
146 bare metal/solvent interfaces. Specifically, the chemical bond
147 between the metal and the ligand introduces a vibrational overlap that
148 is not present without the protecting group, and the overlap between
149 the vibrational spectra (metal to ligand, ligand to solvent) provides
150 a mechanism for rapid thermal transport across the interface.
151
152 A notable result of the previous simulations was the non-monotonic
153 dependence of $G$ on the fractional coverage of the metal surface by
154 the chemical protecting group. Gaps in surface coverage allow the
155 solvent molecules to come into direct contact with ligands that have
156 been heated by the metal surface, absorb thermal energy from the
157 ligands, and then diffuse away. Quantifying the role of overall
158 surface coverage was difficult because the ligands have lateral
159 mobility on the surface and can aggregate to form domains on the
160 timescale of the simulation.
161
162 To isolate the effect of ligand/solvent coupling while avoiding
163 lateral mobility of the surface ligands, the current work utilizes
164 monolayers of mixed chain-lengths in which the length mismatch between
165 long and short chains is sufficient to accommodate solvent
166 molecules. These completely covered (but mixed-chain) surfaces are
167 significantly less prone to surface domain formation, and allow us to
168 further investigate the mechanism of heat transport to the solvent. A
169 thermal flux is created using velocity shearing and scaling reverse
170 non-equilibrium molecular dynamics (VSS-RNEMD), and the resulting
171 temperature profiles are analyzed to yield information about the
172 interfacial thermal conductance.
173
174
175 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
176 % **METHODOLOGY**
177 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
178 \section{Methodology}
179
180 There are many ways to compute bulk transport properties from
181 classical molecular dynamics simulations. Equilibrium Molecular
182 Dynamics (EMD) simulations can be used to compute the relevant time
183 correlation functions and transport coefficients can be calculated
184 assuming that linear response theory
185 holds.\cite{PhysRevB.37.5677,MASSOBRIO:1984bl,PhysRev.119.1,Viscardy:2007rp,che:6888,kinaci:014106}
186 For some transport properties (notably thermal conductivity), EMD
187 approaches are subject to noise and poor convergence of the relevant
188 correlation functions. Traditional Non-equilibrium Molecular Dynamics
189 (NEMD) methods impose a gradient (e.g. thermal or momentum) on a
190 simulation.\cite{ASHURST:1975tg,Evans:1982zk,ERPENBECK:1984sp,MAGINN:1993hc,Berthier:2002ij,Evans:2002ai,Schelling:2002dp,PhysRevA.34.1449,JiangHao_jp802942v}
191 However, the resulting flux is often difficult to
192 measure. Furthermore, problems arise for NEMD simulations of
193 heterogeneous systems, such as phase-phase boundaries or interfaces,
194 where the type of gradient to enforce at the boundary between
195 materials is unclear.
196
197 {\it Reverse} Non-Equilibrium Molecular Dynamics (RNEMD) methods adopt
198 a different approach in that an unphysical {\it flux} is imposed
199 between different regions or ``slabs'' of the simulation
200 box.\cite{MullerPlathe:1997xw,ISI:000080382700030,Kuang2010} The
201 system responds by developing a temperature or momentum {\it gradient}
202 between the two regions. Since the amount of the applied flux is known
203 exactly, and the measurement of a gradient is generally less
204 complicated, imposed-flux methods typically take shorter simulation
205 times to obtain converged results for transport properties. The
206 corresponding temperature or velocity gradients which develop in
207 response to the applied flux are then related (via linear response
208 theory) to the transport coefficient of interest. These methods are
209 quite efficient, in that they do not need many trajectories to provide
210 information about transport properties. To date, they have been
211 utilized in computing thermal and mechanical transfer of both
212 homogeneous
213 liquids~\cite{MullerPlathe:1997xw,ISI:000080382700030,Maginn:2010} as
214 well as heterogeneous
215 systems.\cite{garde:nl2005,garde:PhysRevLett2009,kuang:AuThl}
216
217 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
218 % VSS-RNEMD
219 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
220 \subsection{VSS-RNEMD}
221 The original ``swapping'' approaches by M\"{u}ller-Plathe {\it et
222 al.}\cite{ISI:000080382700030,MullerPlathe:1997xw} can be understood
223 as a sequence of imaginary elastic collisions between particles in
224 regions separated by half of the simulation cell. In each collision,
225 the entire momentum vectors of both particles may be exchanged to
226 generate a thermal flux. Alternatively, a single component of the
227 momentum vectors may be exchanged to generate a shear flux. This
228 algorithm turns out to be quite useful in many simulations of bulk
229 liquids. However, the M\"{u}ller-Plathe swapping approach perturbs the
230 system away from ideal Maxwell-Boltzmann distributions, which has
231 undesirable side-effects when the applied flux becomes
232 large.\cite{Maginn:2010}
233
234 The most useful alternative RNEMD approach developed so far utilizes a
235 series of simultaneous velocity shearing and scaling (VSS) exchanges between
236 the two slabs.\cite{Kuang2012} This method provides a set of
237 conservation constraints while simultaneously creating a desired flux
238 between the two slabs. Satisfying the constraint equations ensures
239 that the new configurations are sampled from the same NVE ensemble.
240
241 The VSS moves are applied periodically to scale and shift the particle
242 velocities ($\mathbf{v}_i$ and $\mathbf{v}_j$) in two slabs ($H$ and
243 $C$) which are separated by half of the simulation box,
244 \begin{displaymath}
245 \begin{array}{rclcl}
246
247 & \underline{\mathrm{shearing}} & &
248 \underline{~~~~~~~~~~~~\mathrm{scaling}~~~~~~~~~~~~} \\ \\
249 \mathbf{v}_i \leftarrow &
250 \mathbf{a}_c\ & + & c\cdot\left(\mathbf{v}_i - \langle\mathbf{v}_c
251 \rangle\right) + \langle\mathbf{v}_c\rangle \\
252 \mathbf{v}_j \leftarrow &
253 \mathbf{a}_h & + & h\cdot\left(\mathbf{v}_j - \langle\mathbf{v}_h
254 \rangle\right) + \langle\mathbf{v}_h\rangle .
255
256 \end{array}
257 \end{displaymath}
258 Here $\langle\mathbf{v}_c\rangle$ and $\langle\mathbf{v}_h\rangle$ are
259 the center of mass velocities in the $C$ and $H$ slabs, respectively.
260 Within the two slabs, particles receive incremental changes or a
261 ``shear'' to their velocities. The amount of shear is governed by the
262 imposed momentum flux, $\mathbf{j}_z(\mathbf{p})$
263 \begin{eqnarray}
264 \mathbf{a}_c & = & - \mathbf{j}_z(\mathbf{p}) \Delta t / M_c \label{vss1}\\
265 \mathbf{a}_h & = & + \mathbf{j}_z(\mathbf{p}) \Delta t / M_h \label{vss2}
266 \end{eqnarray}
267 where $M_{\{c,h\}}$ is the total mass of particles within each of the
268 slabs and $\Delta t$ is the interval between two separate operations.
269
270 To simultaneously impose a thermal flux ($J_z$) between the slabs we
271 use energy conservation constraints,
272 \begin{eqnarray}
273 K_c - J_z\Delta t & = & c^2 (K_c - \frac{1}{2}M_c \langle\mathbf{v}_c
274 \rangle^2) + \frac{1}{2}M_c (\langle \mathbf{v}_c \rangle + \mathbf{a}_c)^2 \label{vss3}\\
275 K_h + J_z\Delta t & = & h^2 (K_h - \frac{1}{2}M_h \langle\mathbf{v}_h
276 \rangle^2) + \frac{1}{2}M_h (\langle \mathbf{v}_h \rangle +
277 \mathbf{a}_h)^2 \label{vss4}.
278 \label{constraint}
279 \end{eqnarray}
280 Simultaneous solution of these quadratic formulae for the scaling
281 coefficients, $c$ and $h$, will ensure that the simulation samples from
282 the original microcanonical (NVE) ensemble. Here $K_{\{c,h\}}$ is the
283 instantaneous translational kinetic energy of each slab. At each time
284 interval, it is a simple matter to solve for $c$, $h$, $\mathbf{a}_c$,
285 and $\mathbf{a}_h$, subject to the imposed momentum flux,
286 $j_z(\mathbf{p})$, and thermal flux, $J_z$, values. Since the VSS
287 operations do not change the kinetic energy due to orientational
288 degrees of freedom or the potential energy of a system, configurations
289 after the VSS move have exactly the same energy (and linear
290 momentum) as before the move.
291
292 As the simulation progresses, the VSS moves are performed on a regular
293 basis, and the system develops a thermal or velocity gradient in
294 response to the applied flux. Using the slope of the temperature or
295 velocity gradient, it is quite simple to obtain the thermal
296 conductivity ($\lambda$),
297 \begin{equation}
298 J_z = -\lambda \frac{\partial T}{\partial z},
299 \end{equation}
300 and shear viscosity ($\eta$),
301 \begin{equation}
302 j_z(p_x) = -\eta \frac{\partial \langle v_x \rangle}{\partial z}.
303 \end{equation}
304 Here, the quantities on the left hand side are the imposed flux
305 values, while the slopes are obtained from linear fits to the
306 gradients that develop in the simulated system.
307
308 The VSS-RNEMD approach is versatile in that it may be used to
309 implement both thermal and shear transport either separately or
310 simultaneously. Perturbations of velocities away from the ideal
311 Maxwell-Boltzmann distributions are minimal, as is thermal anisotropy.
312 This ability to generate simultaneous thermal and shear fluxes has
313 been previously utilized to map out the shear viscosity of SPC/E water
314 over a wide range of temperatures (90~K) with a single 1 ns
315 simulation.\cite{Kuang2012}
316
317 \begin{figure}
318 \includegraphics[width=\linewidth]{figures/rnemd}
319 \caption{The VSS-RNEMD approach imposes unphysical transfer of
320 linear momentum or kinetic energy between a ``hot'' slab and a
321 ``cold'' slab in the simulation box. The system responds to this
322 imposed flux by generating velocity or temperature gradients. The
323 slope of the gradients can then be used to compute transport
324 properties (e.g. shear viscosity or thermal conductivity).}
325 \label{fig:rnemd}
326 \end{figure}
327
328 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
329 % INTERFACIAL CONDUCTANCE
330 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
331 \subsection{Reverse Non-Equilibrium Molecular Dynamics approaches
332 to interfacial transport}
333
334 Interfaces between dissimilar materials have transport properties
335 which can be defined as derivatives of the standard transport
336 coefficients in a direction normal to the interface. For example, the
337 {\it interfacial} thermal conductance ($G$) can be thought of as the
338 change in the thermal conductivity ($\lambda$) across the boundary
339 between materials:
340 \begin{align}
341 G &= \left(\nabla\lambda \cdot \mathbf{\hat{n}}\right)_{z_0} \\
342 &= J_z\left(\frac{\partial^2 T}{\partial z^2}\right)_{z_0} \Big/
343 \left(\frac{\partial T}{\partial z}\right)_{z_0}^2.
344 \label{derivativeG}
345 \end{align}
346 where $z_0$ is the location of the interface between two materials and
347 $\mathbf{\hat{n}}$ is a unit vector normal to the interface (assumed
348 to be the $z$ direction from here on). RNEMD simulations, and
349 particularly the VSS-RNEMD approach, function by applying a momentum
350 or thermal flux and watching the gradient response of the
351 material. This means that the {\it interfacial} conductance is a
352 second derivative property which is subject to significant noise and
353 therefore requires extensive sampling. Previous work has demonstrated
354 the use of the second derivative approach to compute interfacial
355 conductance at chemically-modified metal / solvent interfaces.
356 However, a definition of the interfacial conductance in terms of a
357 temperature difference ($\Delta T$) measured at $z_0$,
358 \begin{displaymath}
359 G = \frac{J_z}{\Delta T_{z_0}},
360 \end{displaymath}
361 is useful once the RNEMD approach has generated a stable temperature
362 gap across the interface.
363
364 \begin{figure}
365 \includegraphics[width=\linewidth]{figures/resistor_series}
366 \caption{The inverse of the interfacial thermal conductance, $G$, is
367 the Kapitza resistance, $R_K$. Because the gold / thiolate /
368 solvent interface extends a significant distance from the metal
369 surface, the interfacial resistance $R_K$ can be computed by
370 summing a series of temperature drops between adjacent temperature
371 bins along the $z$ axis. The depicted temperature profile is from a RNEMD simulation of 100\% butanethiolate (C$_4$) coverage.}
372 \label{fig:resistor_series}
373 \end{figure}
374
375 In the particular case we are studying here, there are two interfaces
376 involved in the transfer of heat from the gold slab to the solvent:
377 the metal / thiolate interface and the thiolate / solvent interface. We
378 can treat the temperature on each side of an interface as discrete,
379 making the interfacial conductance the inverse of the Kaptiza
380 resistance, or $G = \frac{1}{R_k}$. To model the total conductance
381 across multiple interfaces, it is useful to think of the interfaces as
382 a set of resistors wired in series. The total resistance is then
383 additive, $R_{total} = \sum_i R_{i}$, and the interfacial conductance
384 is the inverse of the total resistance, or $G = \frac{1}{\sum_i
385 R_i}$). In the interfacial region, we treat each bin in the
386 VSS-RNEMD temperature profile as a resistor with resistance
387 $\frac{T_{2}-T_{1}}{J_z}$, $\frac{T_{3}-T_{2}}{J_z}$, etc. The sum of
388 the set of resistors which spans the gold / thiolate interface, thiolate
389 chains, and thiolate / solvent interface simplifies to
390 \begin{equation}
391 R_{K} = \frac{T_{n}-T_{1}}{J_z},
392 \label{eq:finalG}
393 \end{equation}
394 or the temperature difference between the gold side of the
395 gold / thiolate interface and the solvent side of the thiolate / solvent
396 interface over the applied flux.
397
398
399 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
400 % **COMPUTATIONAL DETAILS**
401 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
402 \section{Computational Details}
403
404 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
405 % FORCE-FIELD PARAMETERS
406 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
407 \subsection{Force-Field Parameters}
408
409 Our simulations include a number of chemically distinct components.
410 Figure \ref{fig:structures} demonstrates the sites defined for both
411 the {\it n}-hexane and alkanethiolate ligands present in our
412 simulations.
413
414 \begin{figure}
415 \includegraphics[width=\linewidth]{figures/structures}
416 \caption{Topologies of the thiolate capping agents and solvent
417 utilized in the simulations. The chemically-distinct sites (S,
418 \ce{CH2}, and \ce{CH3}) are treated as united atoms. Most
419 parameters are taken from references \bibpunct{}{}{,}{n}{}{,}
420 \protect\cite{TraPPE-UA.alkanes} and
421 \protect\cite{TraPPE-UA.thiols}. Cross-interactions with the Au
422 atoms were adapted from references
423 \protect\cite{landman:1998},~\protect\cite{vlugt:cpc2007154},~and
424 \protect\cite{hautman:4994}.}
425 \label{fig:structures}
426 \end{figure}
427
428 The Au-Au interactions in the metal lattice slab were described by the
429 quantum Sutton-Chen (QSC) formulation.\cite{PhysRevB.59.3527} The QSC
430 potentials include zero-point quantum corrections and are
431 reparametrized for accurate surface energies compared to the
432 Sutton-Chen potentials.\cite{Chen90}
433
434 For the {\it n}-hexane solvent molecules, the TraPPE-UA
435 parameters\cite{TraPPE-UA.alkanes} were utilized. In this model,
436 sites are located at the carbon centers for alkyl groups. Bonding
437 interactions, including bond stretches and bends and torsions, were
438 used for intra-molecular sites closer than 3 bonds. For non-bonded
439 interactions, Lennard-Jones potentials were used. We have previously
440 utilized both united atom (UA) and all-atom (AA) force fields for
441 thermal conductivity,\cite{kuang:AuThl,Kuang2012} and since the united
442 atom force fields cannot populate the high-frequency modes that are
443 present in AA force fields, they appear to work better for modeling
444 thermal conductivity. The TraPPE-UA model for alkanes is known to
445 predict a slightly lower boiling point than experimental values. This
446 is one of the reasons we used a lower average temperature (220 K) for
447 our simulations.
448
449 The TraPPE-UA force field includes parameters for thiol
450 molecules\cite{TraPPE-UA.thiols} which were used for the
451 alkanethiolate molecules in our simulations. To derive suitable
452 parameters for butanethiolate adsorbed on Au(111) surfaces, we adopted
453 the S parameters from Luedtke and Landman\cite{landman:1998} and
454 modified the parameters for the CTS atom to maintain charge neutrality
455 in the molecule.
456
457 To describe the interactions between metal (Au) and non-metal atoms,
458 we refer to an adsorption study of alkyl thiols on gold surfaces by
459 Vlugt {\it et al.}\cite{vlugt:cpc2007154} They fitted an effective
460 Lennard-Jones form of potential parameters for the interaction between
461 Au and pseudo-atoms CH$_x$ and S based on a well-established and
462 widely-used effective potential of Hautman and Klein for the Au(111)
463 surface.\cite{hautman:4994} As our simulations require the gold slab
464 to be flexible to accommodate thermal excitation, the pair-wise form
465 of potentials they developed was used for our study.
466
467 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
468 % SIMULATION PROTOCOL
469 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
470 \subsection{Simulation Protocol}
471
472 We have implemented the VSS-RNEMD algorithm in OpenMD, our group
473 molecular dynamics code.\cite{openmd} An 1188 atom gold slab was
474 prepared and equilibrated at 1 atm and 200 K. The periodic box was
475 then expanded in the $z$ direction to expose two Au(111) faces on
476 either side of the 11-layer slab.
477
478 A full monolayer of thiolates (1/3 the number of surface gold atoms)
479 was placed on three-fold hollow sites on the gold interfaces. The
480 effect of thiolate binding sites on the thermal conductance was tested
481 by placing thiolates at both fcc and hcp hollow sites. No appreciable
482 difference in the temperature profile due to the location of
483 thiolate binding was noted.
484
485 To test the role of thiolate chain length on interfacial thermal
486 conductance, full coverages of each of five chain lengths were tested:
487 butanethiolate (C$_4$), hexanethiolate (C$_6$), octanethiolate
488 (C$_8$), decanethiolate (C$_{10}$), and dodecanethiolate
489 (C$_{12}$). To test the effect of mixed chain lengths, full coverages
490 of C$_4$ / C$_{10}$ and C$_4$ / C$_{12}$ mixtures were created in
491 short/long chain percentages of 25/75, 50/50, 62.5/37.5, 75/25, and
492 87.5/12.5. The short and long chains were placed on the surface hollow
493 sites in a random configuration.
494
495 The simulation box $z$-dimension was set to roughly double the length
496 of the gold / thiolate block. Hexane solvent molecules were placed in
497 the vacant portion of the box using the packmol
498 algorithm.\cite{packmol} Figure \ref{fig:timelapse} shows two of the
499 mixed chain length interfaces both before and after the RNEMD simulation.
500
501 \begin{figure}
502 \includegraphics[width=\linewidth]{figures/timelapse}
503 \caption{Images of 75\%~C$_4$~/~25\%~C$_{12}$ (top panel) and 25\%~C$_4$~/~75\%~C$_{12}$ (bottom panel) interfaces at the beginning and end of 3 ns simulations. Solvent molecules that were initially present in the thiolate layer are colored light blue. Diffusion of the initially-trapped solvent into the bulk is apparent in the interface with fewer long chains. Trapped solvent is orientationally locked to the ordered ligands (and is less able to diffuse into the bulk) when the fraction of long chains increases.}
504 \label{fig:timelapse}
505 \end{figure}
506
507 The system was equilibrated to 220 K in the canonical (NVT) ensemble,
508 allowing the thiolates and solvent to warm gradually. Pressure
509 correction to 1 atm was done using an isobaric-isothermal (NPT)
510 integrator that allowed expansion or contraction only in the $z$
511 direction, maintaining the crystalline structure of the gold as close
512 to the bulk result as possible. The diagonal elements of the pressure
513 tensor were monitored during the pressure equilibration stage. If the
514 $xx$ and/or $yy$ elements had a mean above zero throughout the
515 simulation -- indicating residual surface tension in the plane of the
516 gold slab -- an additional short NPT equilibration step was performed
517 allowing all box dimensions to change. Once the pressure was stable
518 at 1 atm, a final equilibration stage was performed at constant
519 temperature. All systems were equilibrated in the microcanonical (NVE)
520 ensemble before proceeding with the VSS-RNEMD and data collection
521 stages.
522
523 A kinetic energy flux was applied using VSS-RNEMD during a data
524 collection period of 3 nanoseconds, with velocity scaling moves
525 occurring every 10 femtoseconds. The ``hot'' slab was centered in the
526 gold and the ``cold'' slab was placed in the center of the solvent
527 region. The entire system had a (time-averaged) temperature of 220 K,
528 with a temperature difference between the hot and cold slabs of
529 approximately 30 K. The average temperature and kinetic energy flux
530 were selected to avoid solvent freezing (or glass formation) and to
531 prevent the thiolates from burying in the gold slab. The Au-S
532 interaction has a deep potential energy well, which allows sulfur
533 atoms to embed into the gold slab at temperatures above 250 K.
534 Simulation conditions were chosen to avoid both of these
535 situations.
536
537 Temperature profiles of the system were created by dividing the box
538 into $\sim$ 3 \AA \, bins along the $z$ axis and recording the average
539 temperature of each bin.
540
541
542
543 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
544 % **RESULTS**
545 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
546 \section{Results}
547
548 The solvent, hexane, is a straight chain flexible alkane that is structurally
549 similar to the thiolate alkane tails. Previous work has shown that UA models
550 of hexane and butanethiolate have a high degree of vibrational
551 overlap.\cite{kuang:AuThl} This overlap provides a mechanism for thermal
552 energy conduction from the thiolates to the solvent. Indeed, we observe that
553 the interfacial conductance is twice as large with the thiolate monolayers (of
554 all chain lengths) than with the bare metal surface.
555
556 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
557 % CHAIN LENGTH
558 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
559 \subsection{Effect of Chain Length}
560
561 We examined full coverages of five alkyl chain lengths, C$_{4}$, C$_{6}$, C$_{8}$, C$_{10}$, and C$_{12}$. As shown in table \ref{table:chainlengthG}, the interfacial conductance is roughly constant as a function of chain length, indicating that the length of the thiolate alkyl chains does not play a significant role in the transport of heat across the gold/thiolate and thiolate/solvent interfaces. In all cases, the hexane solvent was unable to penetrate into the thiolate layer, leading to a persistent 2-4 \AA \, gap between the solvent region and the thiolates. However, while the identity of the alkyl thiolate capping agent has little effect on the interfacial thermal conductance, the presence of a full monolayer of capping agent provides a two-fold increase in the G value relative to a bare gold surface.
562 \begin{longtable}{p{4cm} p{3cm}}
563 \caption{Computed interfacial thermal conductance ($G$) values for bare gold and 100\% coverages of various thiolate alkyl chain lengths.}
564 \\
565 \centering {\bf Chain Length} & \centering\arraybackslash {\bf G (MW/m$^2$/K)} \\ \hline
566 \endhead
567 \hline
568 \endfoot
569 \centering bare metal & \centering\arraybackslash 30.2 \\
570 $~~~~~~~~~~~~~~~~~~$ C$_{4}$ & \centering\arraybackslash 59.4 \\
571 $~~~~~~~~~~~~~~~~~~$ C$_{6}$ & \centering\arraybackslash 60.2 \\
572 $~~~~~~~~~~~~~~~~~~$ C$_{8}$ & \centering\arraybackslash 61.0 \\
573 $~~~~~~~~~~~~~~~~~~$ C$_{10}$ & \centering\arraybackslash 58.2 \\
574 $~~~~~~~~~~~~~~~~~~$ C$_{12}$ & \centering\arraybackslash 58.8
575 \label{table:chainlengthG}
576 \end{longtable}
577
578 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
579 % MIXED CHAINS
580 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
581 \subsection{Effect of Mixed Chain Lengths}
582
583 Previous simulations have demonstrated non-monotonic behavior for $G$ as a function of the surface coverage. One difficulty with the previous study was the ability of butanethiolate ligands to migrate on the Au(111) surface and to form segregated domains. To simulate the effect of low coverages while preventing thiolate domain formation, we maintain 100\% thiolate coverage while varying the proportions of short (butanethiolate, C$_4$) and long (decanethiolate, C$_{10}$, or dodecanethiolate, C$_{12}$) alkyl chains. Data on the conductance trend as the fraction of long chains was varied is shown in figure \ref{fig:Gstacks}. Note that as in the previous study, $G$ is dependent upon solvent accessibility to thermally excited ligands. Our simulations indicate a similar (but less dramatic) non-monotonic dependence on the fraction of long chains.
584 \begin{figure}
585 \includegraphics[width=\linewidth]{figures/Gstacks}
586 \caption{Interfacial thermal conductivity of mixed-chains has a non-monotonic dependence on the fraction of long chains (lower panels). At low fractions of long chains, the solvent escape rate ($k_{escape}$) dominates the heat transfer process, while the solvent-thiolate orientational ordering ($<d>$) dominates in systems with higher fractions of long chains (upper panels).}
587 \label{fig:Gstacks}
588 \end{figure}
589
590 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
591 % **DISCUSSION**
592 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
593 \section{Discussion}
594
595 In the mixed chain-length simulations, solvent molecules
596 can become temporarily trapped or entangled with the thiolate chains. Their
597 residence in close proximity to the higher temperature environment close to
598 the surface allows them to carry heat away from the surface quite efficiently.
599 There are two aspects of this behavior that are relevant to thermal
600 conductance of the interface: the residence time of solvent molecules in the
601 thiolate layer, and the alignment of solvent molecules with the ligand alkyl
602 chains as a mechanism for transferring vibrational energy to these entrapped
603 solvent molecules. To quantify these competing effects, we have computed
604 solvent escape rates from the thiolate layer as well as a joint orientational
605 order parameter between the trapped solvent and the thiolate ligands.
606
607 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
608 % RESIDENCE TIME
609 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
610 \subsection{Mobility of solvent in the interfacial layer}
611
612 We use a simple survival correlation function, $C(t)$, to measure the
613 residence time of a solvent molecule in the thiolate
614 layer. This function correlates the identity of all hexane molecules
615 within the $z$-coordinate range of the thiolate layer at two separate
616 times. If the solvent molecule is present at both times, the
617 configuration contributes a $1$, while the absence of the molecule at
618 the later time indicates that the solvent molecule has migrated into
619 the bulk, and this configuration contributes a $0$. A steep decay in
620 $C(t)$ indicates a high turnover rate of solvent molecules from the
621 chain region to the bulk. We define the escape rate for trapped solvent molecules at the interface as
622 \begin{equation}
623 k_{escape} = \left( \int_0^T C(t) dt \right)^{-1}
624 \label{eq:mobility}
625 \end{equation}
626 where T is the length of the simulation. This is a direct measure of
627 the rate at which solvent molecules initially entangled in the thiolate layer
628 can escape into the bulk. As $k_{escape} \rightarrow 0$, the
629 solvent becomes permanently trapped in the thiolate layer. In
630 figure \ref{fig:Gstacks} we show that interfacial solvent mobility
631 decreases as the fraction of long thiolate chains increases.
632
633 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
634 % ORDER PARAMETER
635 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
636
637 \subsection{Vibrational coupling via orientational ordering}
638
639 As the fraction of long-chain thiolates increases, the entrapped
640 solvent molecules must find specific orientations relative to the mean
641 orientation of the thiolate chains. This alignment allows for
642 efficient thermal energy exchange between the thiolate alkyl chain and
643 the solvent molecules.
644
645 Once the interfacial solvent molecules have picked up thermal energy from the
646 thiolates, they carry heat away from the gold as they diffuse back
647 into the bulk solvent. When the percentage of long chains decreases,
648 the tails of the long chains are much more disordered and do not
649 provide structured channels for the solvent to fill.
650
651 To measure this cooperative ordering, we compute the orientational order
652 parameters and director axes for both the thiolate chains and for the
653 entrapped solvent. The director axis can be easily obtained by diagonalization
654 of the order parameter tensor,
655 \begin{equation}
656 \mathsf{Q}_{\alpha \beta} = \frac{1}{2 N} \sum_{i=1}^{N} \left( 3 \mathbf{e}_{i
657 \alpha} \mathbf{e}_{i \beta} - \delta_{\alpha \beta} \right)
658 \end{equation}
659 where $\mathbf{e}_{i \alpha}$ is the $\alpha = x,y,z$ component of
660 the unit vector $\mathbf{e}_{i}$ along the long axis of molecule $i$.
661 For both the solvent and the ligand, the $\mathbf{e}$ vector is defined using
662 the terminal atoms of the molecule.
663
664 The largest eigenvalue of $\overleftrightarrow{\mathsf{Q}}$ is
665 traditionally used to obtain the orientational order parameter, while the
666 eigenvector corresponding to the order parameter yields the director
667 axis ($\mathbf{d}(t)$), which defines the average direction of
668 molecular alignment at any time $t$. The overlap between the director
669 axes of the thiolates and the entrapped solvent is time-averaged,
670 \begin{equation}
671 \langle d \rangle = \langle \mathbf{d}_{thiolates} \left( t \right) \cdot
672 \mathbf{d}_{solvent} \left( t \right) \rangle_t
673 \label{eq:orientation}
674 \end{equation}
675 and reported in figure \ref{fig:Gstacks}. Values of $\langle d \rangle$ range from $0$ (solvent molecules in the ligand layer are perpendicular to the thiolate chains) to $1$ (solvent and ligand chains are aligned parallel to each other).
676
677 C$_4$ / C$_{10}$ mixed monolayers have a peak interfacial conductance with 75\% long chains. At this fraction of long chains, the cooperative orientational ordering of the solvent molecules and chains becomes the dominant effect while the solvent escape rate is quite slow. C$_4$ / C$_{12}$ mixtures have a peak interfacial conductance for 87.5\% long chains. The solvent-thiolate orientational ordering reaches its maximum value at this long chain fraction. Long chain fractions of over $0.5$ for the C$_4$ / C$_{12}$ system are well ordered, but this effect is tempered by the exceptionally slow solvent escape rate ($\sim$ 1 molecule / 2 ns).
678
679 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
680 % **CONCLUSIONS**
681 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
682 \section{Conclusions}
683 Our results suggest that a mixed vibrational transfer / convection
684 model may be necessary to explain the features of heat transfer at
685 this interface. The alignment of the solvent chains with the ordered
686 ligand allows rapid transfer of energy to the trapped solvent and
687 becomes the dominant feature for ordered ligand layers. Diffusion of
688 the vibrationally excited solvent into the bulk also plays a
689 significant role when the ligands are less tightly packed.
690
691 In the language of earlier continuum approaches to interfacial
692 conductance,\cite{RevModPhys.61.605} the alignment of the chains is an
693 important factor in the transfer of phonons from the thiolate layer to the
694 trapped solvent. The aligned solvent and thiolate chains have nearly identical
695 acoustic impedances and the phonons can scatter directly into a solvent
696 molecule that has been forced into alignment. When the entrapped solvent has
697 more configurations available, the likelihood of an impedance mismatch is
698 higher, and the phonon scatters into the solvent with lower
699 efficiency. The fractional coverage of the long chains is therefore a simple
700 way of tuning the acoustic mismatch between the thiolate layer and the hexane
701 solvent.
702
703 Efficient heat transfer also can be accomplished via convective or diffusive motion of vibrationally excited solvent back into the bulk. Once the entrapped solvent becomes too tightly aligned with the ligands, however, the convective avenue of heat transfer is cut off.
704
705 Our simulations suggest a number of routes to make interfaces with high thermal conductance. If it is possible to create an interface which forces the solvent into alignment with a ligand that shares many of the solvent's vibrational modes, while simultaneously preserving the ability of the solvent to diffuse back into the bulk, we would expect a significant jump in the interfacial conductance. One possible way to accomplish this is to use polyene ligands with alternating unsaturated bonds. These are significantly more rigid than long alkanes, and could force solvent alignment (even at low relative coverages) while preserving mobility of solvent molecules within the ligand layer.
706
707 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
708 % **ACKNOWLEDGMENTS**
709 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
710 \section*{Acknowledgments}
711
712 We gratefully acknowledge conversations with Dr. Shenyu Kuang. Support for
713 this project was provided by the National Science Foundation under grant
714 CHE-0848243. Computational time was provided by the Center for Research
715 Computing (CRC) at the University of Notre Dame.
716
717 \newpage
718
719 \bibliography{thiolsRNEMD}
720
721 \newpage
722
723 \begin{figure}
724 \centering{\includegraphics{figures/toc2}}
725 \caption{Images of 75\%~C$_4$~/~25\%~C$_{12}$ (top panel) and 25\%~C$_4$~/~75\%~C$_{12}$ (bottom panel) interfaces at the beginning and end of 3 ns simulations. Solvent molecules that were initially present in the thiolate layer are colored light blue. Diffusion of the initially-trapped solvent into the bulk is apparent in the interface with fewer long chains. Trapped solvent is orientationally locked to the ordered ligands (and is less able to diffuse into the bulk) when the fraction of long chains increases.}
726 \label{fig:toc}
727 \end{figure}
728
729 \end{doublespace}
730 \end{document}
731
732
733
734
735
736
737
738

Properties

Name Value
svn:executable *