ViewVC Help
View File | Revision Log | Show Annotations | View Changeset | Root Listing
root/group/stokes/stokes.tex
Revision: 3778
Committed: Sat Dec 10 02:42:55 2011 UTC (12 years, 9 months ago) by skuang
Content type: application/x-tex
File size: 36115 byte(s)
Log Message:
finish much of SPC/E results.

File Contents

# Content
1 \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 %\renewcommand\citemid{\ } % no comma in optional reference note
26 \bibpunct{[}{]}{,}{n}{}{;}
27 \bibliographystyle{achemso}
28
29 \begin{document}
30
31 \title{ENTER TITLE HERE}
32
33 \author{Shenyu Kuang and J. Daniel
34 Gezelter\footnote{Corresponding author. \ Electronic mail: gezelter@nd.edu} \\
35 Department of Chemistry and Biochemistry,\\
36 University of Notre Dame\\
37 Notre Dame, Indiana 46556}
38
39 \date{\today}
40
41 \maketitle
42
43 \begin{doublespace}
44
45 \begin{abstract}
46 REPLACE ABSTRACT HERE
47 With the Non-Isotropic Velocity Scaling (NIVS) approach to Reverse
48 Non-Equilibrium Molecular Dynamics (RNEMD) it is possible to impose
49 an unphysical thermal flux between different regions of
50 inhomogeneous systems such as solid / liquid interfaces. We have
51 applied NIVS to compute the interfacial thermal conductance at a
52 metal / organic solvent interface that has been chemically capped by
53 butanethiol molecules. Our calculations suggest that coupling
54 between the metal and liquid phases is enhanced by the capping
55 agents, leading to a greatly enhanced conductivity at the interface.
56 Specifically, the chemical bond between the metal and the capping
57 agent introduces a vibrational overlap that is not present without
58 the capping agent, and the overlap between the vibrational spectra
59 (metal to cap, cap to solvent) provides a mechanism for rapid
60 thermal transport across the interface. Our calculations also
61 suggest that this is a non-monotonic function of the fractional
62 coverage of the surface, as moderate coverages allow diffusive heat
63 transport of solvent molecules that have been in close contact with
64 the capping agent.
65
66 \end{abstract}
67
68 \newpage
69
70 %\narrowtext
71
72 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
73 % BODY OF TEXT
74 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
75
76 \section{Introduction}
77 [REFINE LATER, ADD MORE REF.S]
78 Imposed-flux methods in Molecular Dynamics (MD)
79 simulations\cite{MullerPlathe:1997xw} can establish steady state
80 systems with a set applied flux vs a corresponding gradient that can
81 be measured. These methods does not need many trajectories to provide
82 information of transport properties of a given system. Thus, they are
83 utilized in computing thermal and mechanical transfer of homogeneous
84 or bulk systems as well as heterogeneous systems such as liquid-solid
85 interfaces.\cite{kuang:AuThl}
86
87 The Reverse Non-Equilibrium MD (RNEMD) methods adopt constraints that
88 satisfy linear momentum and total energy conservation of a system when
89 imposing fluxes in a simulation. Thus they are compatible with various
90 ensembles, including the micro-canonical (NVE) ensemble, without the
91 need of an external thermostat. The original approaches by
92 M\"{u}ller-Plathe {\it et
93 al.}\cite{MullerPlathe:1997xw,ISI:000080382700030} utilize simple
94 momentum swapping for generating energy/momentum fluxes, which is also
95 compatible with particles of different identities. Although simple to
96 implement in a simulation, this approach can create nonthermal
97 velocity distributions, as discovered by Tenney and
98 Maginn\cite{Maginn:2010}. Furthermore, this approach to kinetic energy
99 transfer between particles of different identities is less efficient
100 when the mass difference between the particles becomes significant,
101 which also limits its application on heterogeneous interfacial
102 systems.
103
104 Recently, we developed a different approach, using Non-Isotropic
105 Velocity Scaling (NIVS) \cite{kuang:164101} algorithm to impose
106 fluxes. Compared to the momentum swapping move, it scales the velocity
107 vectors in two separate regions of a simulated system with respective
108 diagonal scaling matrices. These matrices are determined by solving a
109 set of equations including linear momentum and kinetic energy
110 conservation constraints and target flux satisfaction. This method is
111 able to effectively impose a wide range of kinetic energy fluxes
112 without obvious perturbation to the velocity distributions of the
113 simulated systems, regardless of the presence of heterogeneous
114 interfaces. We have successfully applied this approach in studying the
115 interfacial thermal conductance at metal-solvent
116 interfaces.\cite{kuang:AuThl}
117
118 However, the NIVS approach limits its application in imposing momentum
119 fluxes. Temperature anisotropy can happen under high momentum fluxes,
120 due to the nature of the algorithm. Thus, combining thermal and
121 momentum flux is also difficult to implement with this
122 approach. However, such combination may provide a means to simulate
123 thermal/momentum gradient coupled processes such as freeze
124 desalination. Therefore, developing novel approaches to extend the
125 application of imposed-flux method is desired.
126
127 In this paper, we improve the NIVS method and propose a novel approach
128 to impose fluxes. This approach separate the means of applying
129 momentum and thermal flux with operations in one time step and thus is
130 able to simutaneously impose thermal and momentum flux. Furthermore,
131 the approach retains desirable features of previous RNEMD approaches
132 and is simpler to implement compared to the NIVS method. In what
133 follows, we first present the method to implement the method in a
134 simulation. Then we compare the method on bulk fluids to previous
135 methods. Also, interfacial frictions are computed for a series of
136 interfaces.
137
138 \section{Methodology}
139 Similar to the NIVS methodology,\cite{kuang:164101} we consider a
140 periodic system divided into a series of slabs along a certain axis
141 (e.g. $z$). The unphysical thermal and/or momentum flux is designated
142 from the center slab to one of the end slabs, and thus the center slab
143 would have a lower temperature than the end slab (unless the thermal
144 flux is negative). Therefore, the center slab is denoted as ``$c$''
145 while the end slab as ``$h$''.
146
147 To impose these fluxes, we periodically apply separate operations to
148 velocities of particles {$i$} within the center slab and of particles
149 {$j$} within the end slab:
150 \begin{eqnarray}
151 \vec{v}_i & \leftarrow & c\cdot\left(\vec{v}_i - \langle\vec{v}_c
152 \rangle\right) + \left(\langle\vec{v}_c\rangle + \vec{a}_c\right) \\
153 \vec{v}_j & \leftarrow & h\cdot\left(\vec{v}_j - \langle\vec{v}_h
154 \rangle\right) + \left(\langle\vec{v}_h\rangle + \vec{a}_h\right)
155 \end{eqnarray}
156 where $\langle\vec{v}_c\rangle$ and $\langle\vec{v}_h\rangle$ denotes
157 the instantaneous bulk velocity of slabs $c$ and $h$ respectively
158 before an operation occurs. When a momentum flux $\vec{j}_z(\vec{p})$
159 presents, these bulk velocities would have a corresponding change
160 ($\vec{a}_c$ and $\vec{a}_h$ respectively) according to Newton's
161 second law:
162 \begin{eqnarray}
163 M_c \vec{a}_c & = & -\vec{j}_z(\vec{p}) \Delta t \\
164 M_h \vec{a}_h & = & \vec{j}_z(\vec{p}) \Delta t
165 \end{eqnarray}
166 where
167 \begin{eqnarray}
168 M_c & = & \sum_{i = 1}^{N_c} m_i \\
169 M_h & = & \sum_{j = 1}^{N_h} m_j
170 \end{eqnarray}
171 and $\Delta t$ is the interval between two operations.
172
173 The above operations conserve the linear momentum of a periodic
174 system. To satisfy total energy conservation as well as to impose a
175 thermal flux $J_z$, one would have
176 [SUPPORT INFO MIGHT BE NECESSARY TO PUT EXTRA MATH IN]
177 \begin{eqnarray}
178 K_c - J_z\Delta t & = & c^2 (K_c - \frac{1}{2}M_c \langle\vec{v}_c
179 \rangle^2) + \frac{1}{2}M_c (\langle \vec{v}_c \rangle + \vec{a}_c)^2 \\
180 K_h + J_z\Delta t & = & h^2 (K_h - \frac{1}{2}M_h \langle\vec{v}_h
181 \rangle^2) + \frac{1}{2}M_h (\langle \vec{v}_h \rangle + \vec{a}_h)^2
182 \end{eqnarray}
183 where $K_c$ and $K_h$ denotes translational kinetic energy of slabs
184 $c$ and $h$ respectively before an operation occurs. These
185 translational kinetic energy conservation equations are sufficient to
186 ensure total energy conservation, as the operations applied do not
187 change the potential energy of a system, given that the potential
188 energy does not depend on particle velocity.
189
190 The above sets of equations are sufficient to determine the velocity
191 scaling coefficients ($c$ and $h$) as well as $\vec{a}_c$ and
192 $\vec{a}_h$. Note that two roots of $c$ and $h$ exist
193 respectively. However, to avoid dramatic perturbations to a system,
194 the positive roots (which are closer to 1) are chosen. Figure
195 \ref{method} illustrates the implementation of this algorithm in an
196 individual step.
197
198 \begin{figure}
199 \includegraphics[width=\linewidth]{method}
200 \caption{Illustration of the implementation of the algorithm in a
201 single step. Starting from an ideal velocity distribution, the
202 transformation is used to apply both thermal and momentum flux from
203 the ``c'' slab to the ``h'' slab. As the figure shows, the thermal
204 distributions preserve after this operation.}
205 \label{method}
206 \end{figure}
207
208 By implementing these operations at a certain frequency, a steady
209 thermal and/or momentum flux can be applied and the corresponding
210 temperature and/or momentum gradients can be established.
211
212 This approach is more computationaly efficient compared to the
213 previous NIVS method, in that only quadratic equations are involved,
214 while the NIVS method needs to solve a quartic equations. Furthermore,
215 the method implements isotropic scaling of velocities in respective
216 slabs, unlike the NIVS, where an extra criteria function is necessary
217 to choose a set of coefficients that performs the most isotropic
218 scaling. More importantly, separating the momentum flux imposing from
219 velocity scaling avoids the underlying cause that NIVS produced
220 thermal anisotropy when applying a momentum flux.
221
222 The advantages of the approach over the original momentum swapping
223 approach lies in its nature to preserve a Gaussian
224 distribution. Because the momentum swapping tends to render a
225 nonthermal distribution, when the imposed flux is relatively large,
226 diffusion of the neighboring slabs could no longer remedy this effect,
227 and nonthermal distributions would be observed. Results in later
228 section will illustrate this effect.
229
230 \section{Computational Details}
231 The algorithm has been implemented in our MD simulation code,
232 OpenMD\cite{Meineke:2005gd,openmd}. We compare the method with
233 previous RNEMD methods or equilibrium MD methods in homogeneous fluids
234 (Lennard-Jones and SPC/E water). And taking advantage of the method,
235 we simulate the interfacial friction of different heterogeneous
236 interfaces (gold-organic solvent and gold-SPC/E water and ice-liquid
237 water).
238
239 \subsection{Simulation Protocols}
240 The systems to be investigated are set up in a orthorhombic simulation
241 cell with periodic boundary conditions in all three dimensions. The
242 $z$ axis of these cells were longer and was set as the gradient axis
243 of temperature and/or momentum. Thus the cells were divided into $N$
244 slabs along this axis, with various $N$ depending on individual
245 system. The $x$ and $y$ axis were usually of the same length in
246 homogeneous systems or close to each other where interfaces
247 presents. In all cases, before introducing a nonequilibrium method to
248 establish steady thermal and/or momentum gradients for further
249 measurements and calculations, canonical ensemble with a Nos\'e-Hoover
250 thermostat\cite{hoover85} and microcanonical ensemble equilibrations
251 were used to prepare systems ready for data
252 collections. Isobaric-isothermal equilibrations are performed before
253 this for SPC/E water systems to reach normal pressure (1 bar), while
254 similar equilibrations are used for interfacial systems to relax the
255 surface tensions.
256
257 While homogeneous fluid systems can be set up with random
258 configurations, our interfacial systems needs extra steps to ensure
259 the interfaces be established properly for computations. The
260 preparation and equilibration of butanethiol covered gold (111)
261 surface and further solvation and equilibration process is described
262 as in reference \cite{kuang:AuThl}.
263
264 As for the ice/liquid water interfaces, the basal surface of ice
265 lattice was first constructed. Hirsch {\it et
266 al.}\cite{doi:10.1021/jp048434u} explored the energetics of ice
267 lattices with different proton orders. We refer to their results and
268 choose the configuration of the lowest energy after geometry
269 optimization as the unit cells of our ice lattices. Although
270 experimental solid/liquid coexistant temperature near normal pressure
271 is 273K, Bryk and Haymet's simulations of ice/liquid water interfaces
272 with different models suggest that for SPC/E, the most stable
273 interface is observed at 225$\pm$5K. Therefore, all our ice/liquid
274 water simulations were carried out under 225K. To have extra
275 protection of the ice lattice during initial equilibration (when the
276 randomly generated liquid phase configuration could release large
277 amount of energy in relaxation), a constraint method (REF?) was
278 adopted until the high energy configuration was relaxed.
279 [MAY ADD A FIGURE HERE FOR BASAL PLANE, MAY INCLUDE PRISM IF POSSIBLE]
280
281 \subsection{Force Field Parameters}
282 For comparison of our new method with previous work, we retain our
283 force field parameters consistent with the results we will compare
284 with. The Lennard-Jones fluid used here for argon , and reduced unit
285 results are reported for direct comparison purpose.
286
287 As for our water simulations, SPC/E model is used throughout this work
288 for consistency. Previous work for transport properties of SPC/E water
289 model is available\cite{Bedrov:2000,10.1063/1.3330544,Medina2011} so
290 that unnecessary repetition of previous methods can be avoided.
291
292 The Au-Au interaction parameters in all simulations are described by
293 the quantum Sutton-Chen (QSC) formulation.\cite{PhysRevB.59.3527} The
294 QSC potentials include zero-point quantum corrections and are
295 reparametrized for accurate surface energies compared to the
296 Sutton-Chen potentials.\cite{Chen90} For gold/water interfaces, the
297 Spohr potential was adopted\cite{ISI:000167766600035} to depict
298 Au-H$_2$O interactions.
299
300 The small organic molecules included in our simulations are the Au
301 surface capping agent butanethiol and liquid hexane and toluene. The
302 United-Atom
303 models\cite{TraPPE-UA.thiols,TraPPE-UA.alkanes,TraPPE-UA.alkylbenzenes}
304 for these components were used in this work for better computational
305 efficiency, while maintaining good accuracy. We refer readers to our
306 previous work\cite{kuang:AuThl} for further details of these models,
307 as well as the interactions between Au and the above organic molecule
308 components.
309
310 \subsection{Thermal conductivities}
311 When $\vec{j}_z(\vec{p})$ is set to zero and a target $J_z$ is set to
312 impose kinetic energy transfer, the method can be used for thermal
313 conductivity computations. Similar to previous RNEMD methods, we
314 assume linear response of the temperature gradient with respect to the
315 thermal flux in general case. And the thermal conductivity ($\lambda$)
316 can be obtained with the imposed kinetic energy flux and the measured
317 thermal gradient:
318 \begin{equation}
319 J_z = -\lambda \frac{\partial T}{\partial z}
320 \end{equation}
321 Like other imposed-flux methods, the energy flux was calculated using
322 the total non-physical energy transferred (${E_{total}}$) from slab
323 ``c'' to slab ``h'', which is recorded throughout a simulation, and
324 the time for data collection $t$:
325 \begin{equation}
326 J_z = \frac{E_{total}}{2 t L_x L_y}
327 \end{equation}
328 where $L_x$ and $L_y$ denotes the dimensions of the plane in a
329 simulation cell perpendicular to the thermal gradient, and a factor of
330 two in the denominator is present for the heat transport occurs in
331 both $+z$ and $-z$ directions. The temperature gradient
332 ${\langle\partial T/\partial z\rangle}$ can be obtained by a linear
333 regression of the temperature profile, which is recorded during a
334 simulation for each slab in a cell. For Lennard-Jones simulations,
335 thermal conductivities are reported in reduced units
336 (${\lambda^*=\lambda \sigma^2 m^{1/2} k_B^{-1}\varepsilon^{-1/2}}$).
337
338 \subsection{Shear viscosities}
339 Alternatively, the method can carry out shear viscosity calculations
340 by switching off $J_z$. One can specify the vector
341 $\vec{j}_z(\vec{p})$ by choosing the three components
342 respectively. For shear viscosity simulations, $j_z(p_z)$ is usually
343 set to zero. Although for isotropic systems, the direction of
344 $\vec{j}_z(\vec{p})$ on the $xy$ plane should not matter, the ability
345 of arbitarily specifying the vector direction in our method provides
346 convenience in anisotropic simulations.
347
348 Similar to thermal conductivity computations, linear response of the
349 momentum gradient with respect to the shear stress is assumed, and the
350 shear viscosity ($\eta$) can be obtained with the imposed momentum
351 flux (e.g. in $x$ direction) and the measured gradient:
352 \begin{equation}
353 j_z(p_x) = -\eta \frac{\partial v_x}{\partial z}
354 \end{equation}
355 where the flux is similarly defined:
356 \begin{equation}
357 j_z(p_x) = \frac{P_x}{2 t L_x L_y}
358 \end{equation}
359 with $P_x$ being the total non-physical momentum transferred within
360 the data collection time. Also, the velocity gradient
361 ${\langle\partial v_x/\partial z\rangle}$ can be obtained using linear
362 regression of the $x$ component of the mean velocity, $\langle
363 v_x\rangle$, in each of the bins. For Lennard-Jones simulations, shear
364 viscosities are reported in reduced units
365 (${\eta^*=\eta\sigma^2(\varepsilon m)^{-1/2}}$).
366
367 \subsection{Interfacial friction and Slip length}
368 While the shear stress results in a velocity gradient within bulk
369 fluid phase, its effect at a solid-liquid interface could vary due to
370 the interaction strength between the two phases. The interfacial
371 friction coefficient $\kappa$ is defined to relate the shear stress
372 (e.g. along $x$-axis) and the relative fluid velocity tangent to the
373 interface:
374 \begin{equation}
375 j_z(p_x)|_{interface} = \kappa\Delta v_x|_{interface}
376 \end{equation}
377 Under ``stick'' boundary condition, $\Delta v_x|_{interface}
378 \rightarrow 0$, which leads to $\kappa\rightarrow\infty$. However, for
379 ``slip'' boundary condition at the solid-liquid interface, $\kappa$
380 becomes finite. To characterize the interfacial boundary conditions,
381 slip length ($\delta$) is defined using $\kappa$ and the shear
382 viscocity of liquid phase ($\eta$):
383 \begin{equation}
384 \delta = \frac{\eta}{\kappa}
385 \end{equation}
386 so that $\delta\rightarrow 0$ in the ``no-slip'' boundary condition,
387 and depicts how ``slippery'' an interface is. Figure \ref{slipLength}
388 illustrates how this quantity is defined and computed for a
389 solid-liquid interface. [MAY INCLUDE A SNAPSHOT IN FIGURE]
390
391 \begin{figure}
392 \includegraphics[width=\linewidth]{defDelta}
393 \caption{The slip length $\delta$ can be obtained from a velocity
394 profile of a solid-liquid interface. An example of Au/hexane
395 interfaces is shown.}
396 \label{slipLength}
397 \end{figure}
398
399 In our method, a shear stress can be applied similar to shear
400 viscosity computations by applying an unphysical momentum flux
401 (e.g. $j_z(p_x)$). A corresponding velocity profile can be obtained as
402 shown in Figure \ref{slipLength}, in which the velocity gradients
403 within liquid phase and velocity difference at the liquid-solid
404 interface can be measured respectively. Further calculations and
405 characterizations of the interface can be carried out using these
406 data.
407
408 \section{Results and Discussions}
409 \subsection{Lennard-Jones fluid}
410 Our orthorhombic simulation cell of Lennard-Jones fluid has identical
411 parameters to our previous work\cite{kuang:164101} to facilitate
412 comparison. Thermal conductivitis and shear viscosities were computed
413 with the algorithm applied to the simulations. The results of thermal
414 conductivity are compared with our previous NIVS algorithm. However,
415 since the NIVS algorithm could produce temperature anisotropy for
416 shear viscocity computations, these results are instead compared to
417 the momentum swapping approaches. Table \ref{LJ} lists these
418 calculations with various fluxes in reduced units.
419
420 \begin{table*}
421 \begin{minipage}{\linewidth}
422 \begin{center}
423
424 \caption{Thermal conductivity ($\lambda^*$) and shear viscosity
425 ($\eta^*$) (in reduced units) of a Lennard-Jones fluid at
426 ${\langle T^* \rangle = 0.72}$ and ${\rho^* = 0.85}$ computed
427 at various momentum fluxes. The new method yields similar
428 results to previous RNEMD methods. All results are reported in
429 reduced unit. Uncertainties are indicated in parentheses.}
430
431 \begin{tabular}{cccccc}
432 \hline\hline
433 \multicolumn{2}{c}{Momentum Exchange} &
434 \multicolumn{2}{c}{$\lambda^*$} &
435 \multicolumn{2}{c}{$\eta^*$} \\
436 \hline
437 Swap Interval & Equivalent $J_z^*$ or $j_z^*(p_x)$ &
438 NIVS & This work & Swapping & This work \\
439 \hline
440 0.116 & 0.16 & 7.30(0.10) & 6.25(0.23) & 3.57(0.06) & 3.53(0.16)\\
441 0.232 & 0.09 & 6.95(0.09) & 8.02(0.56) & 3.64(0.05) & 3.43(0.17)\\
442 0.463 & 0.047 & 7.19(0.07) & 6.48(0.15) & 3.52(0.16) & 3.51(0.08)\\
443 0.926 & 0.024 & 7.19(0.28) & 7.02(0.13) & 3.72(0.05) & 3.79(0.11)\\
444 1.158 & 0.019 & 7.98(0.33) & 7.37(0.23) & 3.42(0.06) & 3.53(0.06)\\
445 \hline\hline
446 \end{tabular}
447 \label{LJ}
448 \end{center}
449 \end{minipage}
450 \end{table*}
451
452 \subsubsection{Thermal conductivity}
453 Our thermal conductivity calculations with this method yields
454 comparable results to the previous NIVS algorithm. This indicates that
455 the thermal gradients rendered using this method are also close to
456 previous RNEMD methods. Simulations with moderately higher thermal
457 fluxes tend to yield more reliable thermal gradients and thus avoid
458 large errors, while overly high thermal fluxes could introduce side
459 effects such as non-linear temperature gradient response or
460 inadvertent phase transitions.
461
462 Since the scaling operation is isotropic in this method, one does not
463 need extra care to ensure temperature isotropy between the $x$, $y$
464 and $z$ axes, while thermal anisotropy might happen if the criteria
465 function for choosing scaling coefficients does not perform as
466 expected. Furthermore, this method avoids inadvertent concomitant
467 momentum flux when only thermal flux is imposed, which could not be
468 achieved with swapping or NIVS approaches. The thermal energy exchange
469 in swapping ($\vec{p}_i$ in slab ``c'' with $\vec{p}_j$ in slab ``h'')
470 or NIVS (total slab momentum conponemts $P^\alpha$ scaled to $\alpha
471 P^\alpha$) would not obtain this result unless thermal flux vanishes
472 (i.e. $\vec{p}_i=\vec{p}_j$ or $\alpha=1$ which are trivial to apply a
473 thermal flux). In this sense, this method contributes to having
474 minimal perturbation to a simulation while imposing thermal flux.
475
476 \subsubsection{Shear viscosity}
477 Table \ref{LJ} also compares our shear viscosity results with momentum
478 swapping approach. Our calculations show that our method predicted
479 similar values for shear viscosities to the momentum swapping
480 approach, as well as the velocity gradient profiles. Moderately larger
481 momentum fluxes are helpful to reduce the errors of measured velocity
482 gradients and thus the final result. However, it is pointed out that
483 the momentum swapping approach tends to produce nonthermal velocity
484 distributions.\cite{Maginn:2010}
485
486 To examine that temperature isotropy holds in simulations using our
487 method, we measured the three one-dimensional temperatures in each of
488 the slabs (Figure \ref{tempXyz}). Note that the $x$-dimensional
489 temperatures were calculated after subtracting the effects from bulk
490 velocities of the slabs. The one-dimensional temperature profiles
491 showed no observable difference between the three dimensions. This
492 ensures that isotropic scaling automatically preserves temperature
493 isotropy and that our method is useful in shear viscosity
494 computations.
495
496 \begin{figure}
497 \includegraphics[width=\linewidth]{tempXyz}
498 \caption{Unlike the previous NIVS algorithm, the new method does not
499 produce a thermal anisotropy. No temperature difference between
500 different dimensions were observed beyond the magnitude of the error
501 bars. Note that the two ``hotter'' regions are caused by the shear
502 stress, as reported by Tenney and Maginn\cite{Maginn:2010}, but not
503 an effect that only observed in our methods.}
504 \label{tempXyz}
505 \end{figure}
506
507 Furthermore, the velocity distribution profiles are tested by imposing
508 a large shear stress into the simulations. Figure \ref{vDist}
509 demonstrates how our method is able to maintain thermal velocity
510 distributions against the momentum swapping approach even under large
511 imposed fluxes. Previous swapping methods tend to deplete particles of
512 positive velocities in the negative velocity slab (``c'') and vice
513 versa in slab ``h'', where the distributions leave a notch. This
514 problematic profiles become significant when the imposed-flux becomes
515 larger and diffusions from neighboring slabs could not offset the
516 depletion. Simutaneously, abnormal peaks appear corresponding to
517 excessive velocity swapped from the other slab. This nonthermal
518 distributions limit applications of the swapping approach in shear
519 stress simulations. Our method avoids the above problematic
520 distributions by altering the means of applying momentum
521 fluxes. Comparatively, velocity distributions recorded from
522 simulations with our method is so close to the ideal thermal
523 prediction that no observable difference is shown in Figure
524 \ref{vDist}. Conclusively, our method avoids problems happened in
525 previous RNEMD methods and provides a useful means for shear viscosity
526 computations.
527
528 \begin{figure}
529 \includegraphics[width=\linewidth]{velDist}
530 \caption{Velocity distributions that develop under the swapping and
531 our methods at high flux. These distributions were obtained from
532 Lennard-Jones simulations with $j_z(p_x)\sim 0.4$ (equivalent to a
533 swapping interval of 20 time steps). This is a relatively large flux
534 to demonstrate the nonthermal distributions that develop under the
535 swapping method. Distributions produced by our method are very close
536 to the ideal thermal situations.}
537 \label{vDist}
538 \end{figure}
539
540 \subsection{Bulk SPC/E water}
541 Since our method was in good performance of thermal conductivity and
542 shear viscosity computations for simple Lennard-Jones fluid, we extend
543 our applications of these simulations to complex fluid like SPC/E
544 water model. A simulation cell with 1000 molecules was set up in the
545 same manner as in \cite{kuang:164101}. For thermal conductivity
546 simulations, measurements were taken to compare with previous RNEMD
547 methods; for shear viscosity computations, simulations were run under
548 a series of temperatures (with corresponding pressure relaxation using
549 the isobaric-isothermal ensemble[CITE NIVS REF 32]), and results were
550 compared to available data from Equilibrium MD methods[CITATIONS].
551
552 \subsubsection{Thermal conductivity}
553 Table \ref{spceThermal} summarizes our thermal conductivity
554 computations under different temperatures and thermal gradients, in
555 comparison to the previous NIVS results\cite{kuang:164101} and
556 experimental measurements\cite{WagnerKruse}. Note that no appreciable
557 drift of total system energy or temperature was observed when our
558 method is applied, which indicates that our algorithm conserves total
559 energy even for systems involving electrostatic interactions.
560
561 Measurements using our method established similar temperature
562 gradients to the previous NIVS method. Our simulation results are in
563 good agreement with those from previous simulations. And both methods
564 yield values in reasonable agreement with experimental
565 values. Simulations using moderately higher thermal gradient or those
566 with longer gradient axis ($z$) for measurement seem to have better
567 accuracy, from our results.
568
569 \begin{table*}
570 \begin{minipage}{\linewidth}
571 \begin{center}
572
573 \caption{Thermal conductivity of SPC/E water under various
574 imposed thermal gradients. Uncertainties are indicated in
575 parentheses.}
576
577 \begin{tabular}{ccccc}
578 \hline\hline
579 $\langle T\rangle$ & $\langle dT/dz\rangle$ & \multicolumn{3}{c}
580 {$\lambda (\mathrm{W m}^{-1} \mathrm{K}^{-1})$} \\
581 (K) & (K/\AA) & This work & Previous NIVS\cite{kuang:164101} &
582 Experiment\cite{WagnerKruse} \\
583 \hline
584 300 & 0.8 & 0.815(0.027) & 0.770(0.008) & 0.61 \\
585 318 & 0.8 & 0.801(0.024) & 0.750(0.032) & 0.64 \\
586 & 1.6 & 0.766(0.007) & 0.778(0.019) & \\
587 & 0.8 & 0.786(0.009)$^a$ & & \\
588 \hline\hline
589 \end{tabular}
590 $^a$Simulation with $L_z$ twice as long.
591 \label{spceThermal}
592 \end{center}
593 \end{minipage}
594 \end{table*}
595
596 \subsubsection{Shear viscosity}
597 The improvement our method achieves for shear viscosity computations
598 enables us to apply it on SPC/E water models. The series of
599 temperatures under which our shear viscosity calculations were carried
600 out covers the liquid range under normal pressure. Our simulations
601 predict a similar trend of $\eta$ vs. $T$ to EMD results we refer to
602 (Table \ref{spceShear}).
603
604 \begin{table*}
605 \begin{minipage}{\linewidth}
606 \begin{center}
607
608 \caption{Computed shear viscosity of SPC/E water under different
609 temperatures. Results are compared to those obtained with EMD
610 method[CITATION]. Uncertainties are indicated in parentheses.}
611
612 \begin{tabular}{cccc}
613 \hline\hline
614 $T$ & $\partial v_x/\partial z$ & \multicolumn{2}{c}
615 {$\eta (\mathrm{mPa}\cdot\mathrm{s})$} \\
616 (K) & (10$^{10}$s$^{-1}$) & This work & Previous simulations[CITATION]\\
617 \hline
618 273 & & 1.218(0.004) & \\
619 & & 1.140(0.012) & \\
620 303 & & 0.646(0.008) & \\
621 318 & & 0.536(0.007) & \\
622 & & 0.510(0.007) & \\
623 & & & \\
624 333 & & 0.428(0.002) & \\
625 363 & & 0.279(0.014) & \\
626 & & 0.306(0.001) & \\
627 \hline\hline
628 \end{tabular}
629 \label{spceShear}
630 \end{center}
631 \end{minipage}
632 \end{table*}
633
634 [MAY COMBINE JZPX AND JZKE TO RUN ONE JOB BUT GET ETA=F(T)]
635 [PUT RESULTS AND FIGURE HERE IF IT WORKS]
636 \subsection{Interfacial frictions}
637 [SLIP BOUNDARY VS STICK BOUNDARY]
638
639 qualitative agreement w interfacial thermal conductance
640
641 [ETA OBTAINED HERE DOES NOT NECESSARILY EQUAL TO BULK VALUES]
642
643
644 [ATTEMPT TO CONSTRUCT BASAL PLANE ICE-WATER INTERFACE]
645
646 [FUTURE WORK HERE OR IN CONCLUSIONS]
647
648
649 \begin{table*}
650 \begin{minipage}{\linewidth}
651 \begin{center}
652
653 \caption{Computed interfacial thermal conductance ($G$ and
654 $G^\prime$) values for interfaces using various models for
655 solvent and capping agent (or without capping agent) at
656 $\langle T\rangle\sim$200K. Here ``D'' stands for deuterated
657 solvent or capping agent molecules. Error estimates are
658 indicated in parentheses.}
659
660 \begin{tabular}{llccc}
661 \hline\hline
662 Butanethiol model & Solvent & $G$ & $G^\prime$ \\
663 (or bare surface) & model & \multicolumn{2}{c}{(MW/m$^2$/K)} \\
664 \hline
665 UA & UA hexane & 131(9) & 87(10) \\
666 & UA hexane(D) & 153(5) & 136(13) \\
667 & AA hexane & 131(6) & 122(10) \\
668 & UA toluene & 187(16) & 151(11) \\
669 & AA toluene & 200(36) & 149(53) \\
670 \hline
671 bare & UA hexane & 46.5(3.2) & 49.4(4.5) \\
672 & UA hexane(D) & 43.9(4.6) & 43.0(2.0) \\
673 & AA hexane & 31.0(1.4) & 29.4(1.3) \\
674 & UA toluene & 70.1(1.3) & 65.8(0.5) \\
675 \hline\hline
676 \end{tabular}
677 \label{modelTest}
678 \end{center}
679 \end{minipage}
680 \end{table*}
681
682 On bare metal / solvent surfaces, different force field models for
683 hexane yield similar results for both $G$ and $G^\prime$, and these
684 two definitions agree with each other very well. This is primarily an
685 indicator of weak interactions between the metal and the solvent.
686
687 For the fully-covered surfaces, the choice of force field for the
688 capping agent and solvent has a large impact on the calculated values
689 of $G$ and $G^\prime$. The AA thiol to AA solvent conductivities are
690 much larger than their UA to UA counterparts, and these values exceed
691 the experimental estimates by a large measure. The AA force field
692 allows significant energy to go into C-H (or C-D) stretching modes,
693 and since these modes are high frequency, this non-quantum behavior is
694 likely responsible for the overestimate of the conductivity. Compared
695 to the AA model, the UA model yields more reasonable conductivity
696 values with much higher computational efficiency.
697
698 \subsubsection{Effects due to average temperature}
699
700 We also studied the effect of average system temperature on the
701 interfacial conductance. The simulations are first equilibrated in
702 the NPT ensemble at 1 atm. The TraPPE-UA model for hexane tends to
703 predict a lower boiling point (and liquid state density) than
704 experiments. This lower-density liquid phase leads to reduced contact
705 between the hexane and butanethiol, and this accounts for our
706 observation of lower conductance at higher temperatures. In raising
707 the average temperature from 200K to 250K, the density drop of
708 $\sim$20\% in the solvent phase leads to a $\sim$40\% drop in the
709 conductance.
710
711 Similar behavior is observed in the TraPPE-UA model for toluene,
712 although this model has better agreement with the experimental
713 densities of toluene. The expansion of the toluene liquid phase is
714 not as significant as that of the hexane (8.3\% over 100K), and this
715 limits the effect to $\sim$20\% drop in thermal conductivity.
716
717 Although we have not mapped out the behavior at a large number of
718 temperatures, is clear that there will be a strong temperature
719 dependence in the interfacial conductance when the physical properties
720 of one side of the interface (notably the density) change rapidly as a
721 function of temperature.
722
723 \section{Conclusions}
724 [VSIS WORKS! COMBINES NICE FEATURES OF PREVIOUS RNEMD METHODS AND
725 IMPROVEMENTS TO THEIR PROBLEMS! ROBUST AND VERSATILE!]
726
727 The NIVS algorithm has been applied to simulations of
728 butanethiol-capped Au(111) surfaces in the presence of organic
729 solvents. This algorithm allows the application of unphysical thermal
730 flux to transfer heat between the metal and the liquid phase. With the
731 flux applied, we were able to measure the corresponding thermal
732 gradients and to obtain interfacial thermal conductivities. Under
733 steady states, 2-3 ns trajectory simulations are sufficient for
734 computation of this quantity.
735
736 Our simulations have seen significant conductance enhancement in the
737 presence of capping agent, compared with the bare gold / liquid
738 interfaces. The vibrational coupling between the metal and the liquid
739 phase is enhanced by a chemically-bonded capping agent. Furthermore,
740 the coverage percentage of the capping agent plays an important role
741 in the interfacial thermal transport process. Moderately low coverages
742 allow higher contact between capping agent and solvent, and thus could
743 further enhance the heat transfer process, giving a non-monotonic
744 behavior of conductance with increasing coverage.
745
746 Our results, particularly using the UA models, agree well with
747 available experimental data. The AA models tend to overestimate the
748 interfacial thermal conductance in that the classically treated C-H
749 vibrations become too easily populated. Compared to the AA models, the
750 UA models have higher computational efficiency with satisfactory
751 accuracy, and thus are preferable in modeling interfacial thermal
752 transport.
753
754 \section{Acknowledgments}
755 Support for this project was provided by the National Science
756 Foundation under grant CHE-0848243. Computational time was provided by
757 the Center for Research Computing (CRC) at the University of Notre
758 Dame.
759
760 \newpage
761
762 \bibliography{stokes}
763
764 \end{doublespace}
765 \end{document}