ViewVC Help
View File | Revision Log | Show Annotations | View Changeset | Root Listing
root/group/trunk/multipole/multipole1.tex
Revision: 4175
Committed: Fri Jun 6 16:01:36 2014 UTC (10 years ago) by gezelter
Content type: application/x-tex
File size: 69235 byte(s)
Log Message:
Latest edits

File Contents

# User Rev Content
1 gezelter 3906 % ****** Start of file aipsamp.tex ******
2     %
3     % This file is part of the AIP files in the AIP distribution for REVTeX 4.
4     % Version 4.1 of REVTeX, October 2009
5     %
6     % Copyright (c) 2009 American Institute of Physics.
7     %
8     % See the AIP README file for restrictions and more information.
9     %
10     % TeX'ing this file requires that you have AMS-LaTeX 2.0 installed
11     % as well as the rest of the prerequisites for REVTeX 4.1
12     %
13     % It also requires running BibTeX. The commands are as follows:
14     %
15     % 1) latex aipsamp
16     % 2) bibtex aipsamp
17     % 3) latex aipsamp
18     % 4) latex aipsamp
19     %
20     % Use this file as a source of example code for your aip document.
21     % Use the file aiptemplate.tex as a template for your document.
22     \documentclass[%
23 gezelter 3980 aip,jcp,
24 gezelter 3906 amsmath,amssymb,
25     preprint,%
26 gezelter 4172 % reprint,%
27 gezelter 3906 %author-year,%
28     %author-numerical,%
29 gezelter 3984 jcp]{revtex4-1}
30 gezelter 3906
31     \usepackage{graphicx}% Include figure files
32     \usepackage{dcolumn}% Align table columns on decimal point
33 gezelter 3980 %\usepackage{bm}% bold math
34 gezelter 3982 \usepackage{times}
35 gezelter 3980 \usepackage[version=3]{mhchem} % this is a great package for formatting chemical reactions
36     \usepackage{url}
37 gezelter 3985 \usepackage{rotating}
38 gezelter 4172
39 gezelter 3906 %\usepackage[mathlines]{lineno}% Enable numbering of text and display math
40     %\linenumbers\relax % Commence numbering lines
41    
42     \begin{document}
43    
44 gezelter 3988 %\preprint{AIP/123-QED}
45 gezelter 3906
46 gezelter 3988 \title{Real space alternatives to the Ewald
47 gezelter 4174 Sum. I. Shifted electrostatics for multipoles}
48 gezelter 3906
49     \author{Madan Lamichhane}
50     \affiliation{Department of Physics, University
51     of Notre Dame, Notre Dame, IN 46556}
52    
53     \author{J. Daniel Gezelter}
54     \email{gezelter@nd.edu.}
55     \affiliation{Department of Chemistry and Biochemistry, University
56     of Notre Dame, Notre Dame, IN 46556}
57    
58     \author{Kathie E. Newman}
59     \affiliation{Department of Physics, University
60     of Notre Dame, Notre Dame, IN 46556}
61    
62    
63     \date{\today}% It is always \today, today,
64     % but any date may be explicitly specified
65    
66     \begin{abstract}
67 gezelter 3980 We have extended the original damped-shifted force (DSF)
68     electrostatic kernel and have been able to derive two new
69     electrostatic potentials for higher-order multipoles that are based
70     on truncated Taylor expansions around the cutoff radius. For
71     multipole-multipole interactions, we find that each of the distinct
72     orientational contributions has a separate radial function to ensure
73     that the overall forces and torques vanish at the cutoff radius. In
74     this paper, we present energy, force, and torque expressions for the
75     new models, and compare these real-space interaction models to exact
76     results for ordered arrays of multipoles.
77 gezelter 3906 \end{abstract}
78    
79 gezelter 3988 %\pacs{Valid PACS appear here}% PACS, the Physics and Astronomy
80 gezelter 3906 % Classification Scheme.
81 gezelter 3988 %\keywords{Suggested keywords}%Use showkeys class option if keyword
82 gezelter 3906 %display desired
83     \maketitle
84    
85     \section{Introduction}
86 gezelter 3982 There has been increasing interest in real-space methods for
87     calculating electrostatic interactions in computer simulations of
88     condensed molecular
89 gezelter 3980 systems.\cite{Wolf99,Zahn02,Kast03,BeckD.A.C._bi0486381,Ma05,Fennell:2006zl,Chen:2004du,Chen:2006ii,Rodgers:2006nw,Denesyuk:2008ez,Izvekov:2008wo}
90     The simplest of these techniques was developed by Wolf {\it et al.}
91     in their work towards an $\mathcal{O}(N)$ Coulombic sum.\cite{Wolf99}
92 gezelter 3982 For systems of point charges, Fennell and Gezelter showed that a
93     simple damped shifted force (DSF) modification to Wolf's method could
94     give nearly quantitative agreement with smooth particle mesh Ewald
95     (SPME)\cite{Essmann95} configurational energy differences as well as
96     atomic force and molecular torque vectors.\cite{Fennell:2006zl}
97 gezelter 3906
98 gezelter 3980 The computational efficiency and the accuracy of the DSF method are
99     surprisingly good, particularly for systems with uniform charge
100     density. Additionally, dielectric constants obtained using DSF and
101 gezelter 3986 similar methods where the force vanishes at $r_{c}$ are
102 gezelter 3980 essentially quantitative.\cite{Izvekov:2008wo} The DSF and other
103     related methods have now been widely investigated,\cite{Hansen:2012uq}
104 gezelter 3985 and DSF is now used routinely in a diverse set of chemical
105     environments.\cite{doi:10.1021/la400226g,McCann:2013fk,kannam:094701,Forrest:2012ly,English:2008kx,Louden:2013ve,Tokumasu:2013zr}
106     DSF electrostatics provides a compromise between the computational
107     speed of real-space cutoffs and the accuracy of fully-periodic Ewald
108     treatments.
109 gezelter 3906
110 gezelter 3980 One common feature of many coarse-graining approaches, which treat
111     entire molecular subsystems as a single rigid body, is simplification
112     of the electrostatic interactions between these bodies so that fewer
113     site-site interactions are required to compute configurational
114 gezelter 3986 energies. To do this, the interactions between coarse-grained sites
115     are typically taken to be point
116     multipoles.\cite{Golubkov06,ISI:000276097500009,ISI:000298664400012}
117 gezelter 3906
118 gezelter 3986 Water, in particular, has been modeled recently with point multipoles
119     up to octupolar
120     order.\cite{Chowdhuri:2006lr,Te:2010rt,Te:2010ys,Te:2010vn} For
121     maximum efficiency, these models require the use of an approximate
122     multipole expansion as the exact multipole expansion can become quite
123     expensive (particularly when handled via the Ewald
124     sum).\cite{Ichiye:2006qy} Point multipoles and multipole
125     polarizability have also been utilized in the AMOEBA water model and
126 gezelter 3980 related force fields.\cite{Ponder:2010fk,schnieders:124114,Ren:2011uq}
127 gezelter 3906
128 gezelter 3980 Higher-order multipoles present a peculiar issue for molecular
129     dynamics. Multipolar interactions are inherently short-ranged, and
130     should not need the relatively expensive Ewald treatment. However,
131     real-space cutoff methods are normally applied in an orientation-blind
132     fashion so multipoles which leave and then re-enter a cutoff sphere in
133     a different orientation can cause energy discontinuities.
134 gezelter 3906
135 gezelter 3980 This paper outlines an extension of the original DSF electrostatic
136 gezelter 3985 kernel to point multipoles. We describe two distinct real-space
137 gezelter 3982 interaction models for higher-order multipoles based on two truncated
138     Taylor expansions that are carried out at the cutoff radius. We are
139     calling these models {\bf Taylor-shifted} and {\bf Gradient-shifted}
140     electrostatics. Because of differences in the initial assumptions,
141 gezelter 3986 the two methods yield related, but somewhat different expressions for
142     energies, forces, and torques.
143 gezelter 3906
144 gezelter 3982 In this paper we outline the new methodology and give functional forms
145     for the energies, forces, and torques up to quadrupole-quadrupole
146     order. We also compare the new methods to analytic energy constants
147 gezelter 3986 for periodic arrays of point multipoles. In the following paper, we
148 gezelter 3985 provide numerical comparisons to Ewald-based electrostatics in common
149     simulation enviornments.
150 gezelter 3982
151 gezelter 3980 \section{Methodology}
152 gezelter 3986 An efficient real-space electrostatic method involves the use of a
153     pair-wise functional form,
154     \begin{equation}
155 gezelter 4175 U = \sum_i \sum_{j>i} U_\mathrm{pair}(\mathbf{r}_{ij}, \Omega_i, \Omega_j) +
156     \sum_i U_i^\mathrm{self}
157 gezelter 3986 \end{equation}
158     that is short-ranged and easily truncated at a cutoff radius,
159     \begin{equation}
160 gezelter 4175 U_\mathrm{pair}(\mathbf{r}_{ij},\Omega_i, \Omega_j) = \left\{
161 gezelter 3986 \begin{array}{ll}
162 gezelter 4175 U_\mathrm{approx} (\mathbf{r}_{ij}, \Omega_i, \Omega_j) & \quad \left| \mathbf{r}_{ij} \right| \le r_c \\
163 gezelter 4098 0 & \quad \left| \mathbf{r}_{ij} \right| > r_c ,
164 gezelter 3986 \end{array}
165     \right.
166     \end{equation}
167 gezelter 3989 along with an easily computed self-interaction term ($\sum_i
168 gezelter 4175 U_i^\mathrm{self}$) which scales linearly with the number of
169 gezelter 3986 particles. Here $\Omega_i$ and $\Omega_j$ represent orientational
170 gezelter 4098 coordinates of the two sites, and $\mathbf{r}_{ij}$ is the vector
171     between the two sites. The computational efficiency, energy
172 gezelter 3986 conservation, and even some physical properties of a simulation can
173 gezelter 4175 depend dramatically on how the $U_\mathrm{approx}$ function behaves at
174 gezelter 3986 the cutoff radius. The goal of any approximation method should be to
175     mimic the real behavior of the electrostatic interactions as closely
176     as possible without sacrificing the near-linear scaling of a cutoff
177     method.
178 gezelter 3906
179 gezelter 3980 \subsection{Self-neutralization, damping, and force-shifting}
180     The DSF and Wolf methods operate by neutralizing the total charge
181     contained within the cutoff sphere surrounding each particle. This is
182     accomplished by shifting the potential functions to generate image
183     charges on the surface of the cutoff sphere for each pair interaction
184 gezelter 4098 computed within $r_c$. Damping using a complementary error function is
185     applied to the potential to accelerate convergence. The interaction
186     for a pair of charges ($C_i$ and $C_j$) in the DSF method,
187 gezelter 3980 \begin{equation*}
188 gezelter 4175 U_\mathrm{DSF}(r_{ij}) = C_i C_j \Biggr{[}
189 gezelter 3980 \frac{\mathrm{erfc}\left(\alpha r_{ij}\right)}{r_{ij}}
190 gezelter 3986 - \frac{\mathrm{erfc}\left(\alpha r_c\right)}{r_c} + \left(\frac{\mathrm{erfc}\left(\alpha r_c\right)}{r_c^2}
191 gezelter 3980 + \frac{2\alpha}{\pi^{1/2}}
192 gezelter 3986 \frac{\exp\left(-\alpha^2r_c^2\right)}{r_c}
193     \right)\left(r_{ij}-r_c\right)\ \Biggr{]}
194 gezelter 3980 \label{eq:DSFPot}
195     \end{equation*}
196 gezelter 4098 where $\alpha$ is the adjustable damping parameter. Note that in this
197     potential and in all electrostatic quantities that follow, the
198     standard $1/4 \pi \epsilon_{0}$ has been omitted for clarity.
199 gezelter 3980
200     To insure net charge neutrality within each cutoff sphere, an
201     additional ``self'' term is added to the potential. This term is
202     constant (as long as the charges and cutoff radius do not change), and
203     exists outside the normal pair-loop for molecular simulations. It can
204     be thought of as a contribution from a charge opposite in sign, but
205     equal in magnitude, to the central charge, which has been spread out
206     over the surface of the cutoff sphere. A portion of the self term is
207     identical to the self term in the Ewald summation, and comes from the
208     utilization of the complimentary error function for electrostatic
209 gezelter 3986 damping.\cite{deLeeuw80,Wolf99} There have also been recent efforts to
210     extend the Wolf self-neutralization method to zero out the dipole and
211     higher order multipoles contained within the cutoff
212 gezelter 3985 sphere.\cite{Fukuda:2011jk,Fukuda:2012yu,Fukuda:2013qv}
213 gezelter 3982
214 gezelter 3985 In this work, we extend the idea of self-neutralization for the point
215     multipoles by insuring net charge-neutrality and net-zero moments
216     within each cutoff sphere. In Figure \ref{fig:shiftedMultipoles}, the
217     central dipolar site $\mathbf{D}_i$ is interacting with point dipole
218     $\mathbf{D}_j$ and point quadrupole, $\mathbf{Q}_k$. The
219     self-neutralization scheme for point multipoles involves projecting
220     opposing multipoles for sites $j$ and $k$ on the surface of the cutoff
221     sphere. There are also significant modifications made to make the
222     forces and torques go smoothly to zero at the cutoff distance.
223 gezelter 3982
224 gezelter 4172 \begin{figure}
225     \includegraphics[width=3in]{SM}
226     \caption{Reversed multipoles are projected onto the surface of the
227     cutoff sphere. The forces, torques, and potential are then smoothly
228     shifted to zero as the sites leave the cutoff region.}
229     \label{fig:shiftedMultipoles}
230     \end{figure}
231 gezelter 3980
232 gezelter 3986 As in the point-charge approach, there is an additional contribution
233     from self-neutralization of site $i$. The self term for multipoles is
234 gezelter 3982 described in section \ref{sec:selfTerm}.
235 gezelter 3906
236 gezelter 3982 \subsection{The multipole expansion}
237    
238 gezelter 3980 Consider two discrete rigid collections of point charges, denoted as
239 gezelter 3982 $\bf a$ and $\bf b$. In the following, we assume that the two objects
240     interact via electrostatics only and describe those interactions in
241     terms of a standard multipole expansion. Putting the origin of the
242     coordinate system at the center of mass of $\bf a$, we use vectors
243 gezelter 3980 $\mathbf{r}_k$ to denote the positions of all charges $q_k$ in $\bf
244     a$. Then the electrostatic potential of object $\bf a$ at
245     $\mathbf{r}$ is given by
246 gezelter 3906 \begin{equation}
247 gezelter 4175 \phi_a(\mathbf r) =
248 gezelter 3906 \sum_{k \, \text{in \bf a}} \frac{q_k}{\lvert \mathbf{r} - \mathbf{r}_k \rvert}.
249     \end{equation}
250 gezelter 3982 The Taylor expansion in $r$ can be written using an implied summation
251     notation. Here Greek indices are used to indicate space coordinates
252     ($x$, $y$, $z$) and the subscripts $k$ and $j$ are reserved for
253 gezelter 4098 labeling specific charges in $\bf a$ and $\bf b$ respectively. The
254 gezelter 3982 Taylor expansion,
255 gezelter 3906 \begin{equation}
256     \frac{1}{\lvert \mathbf{r} - \mathbf{r}_k \rvert} =
257     \left( 1
258     - r_{k\alpha} \frac{\partial}{\partial r_{\alpha}}
259     + \frac{1}{2} r_{k\alpha} r_{k\beta} \frac{\partial^2}{\partial r_{\alpha} \partial r_{\beta}} +\dots
260     \right)
261 gezelter 3982 \frac{1}{r} ,
262 gezelter 3906 \end{equation}
263 gezelter 3982 can then be used to express the electrostatic potential on $\bf a$ in
264     terms of multipole operators,
265 gezelter 3906 \begin{equation}
266 gezelter 4175 \phi_{\bf a}(\mathbf{r}) =\hat{M}_{\bf a} \frac{1}{r}
267 gezelter 3906 \end{equation}
268     where
269     \begin{equation}
270     \hat{M}_{\bf a} = C_{\bf a} - D_{{\bf a}\alpha} \frac{\partial}{\partial r_{\alpha}}
271     + Q_{{\bf a}\alpha\beta}
272     \frac{\partial^2}{\partial r_{\alpha} \partial r_{\beta}} + \dots
273     \end{equation}
274 gezelter 3980 Here, the point charge, dipole, and quadrupole for object $\bf a$ are
275     given by $C_{\bf a}$, $D_{{\bf a}\alpha}$, and $Q_{{\bf
276 gezelter 3982 a}\alpha\beta}$, respectively. These are the primitive multipoles
277     which can be expressed as a distribution of charges,
278     \begin{align}
279 gezelter 3996 C_{\bf a} =&\sum_{k \, \text{in \bf a}} q_k , \label{eq:charge} \\
280     D_{{\bf a}\alpha} =&\sum_{k \, \text{in \bf a}} q_k r_{k\alpha}, \label{eq:dipole}\\
281     Q_{{\bf a}\alpha\beta} =& \frac{1}{2} \sum_{k \, \text{in \bf a}} q_k
282     r_{k\alpha} r_{k\beta} . \label{eq:quadrupole}
283 gezelter 3982 \end{align}
284     Note that the definition of the primitive quadrupole here differs from
285     the standard traceless form, and contains an additional Taylor-series
286 gezelter 3996 based factor of $1/2$. We are essentially treating the mass
287     distribution with higher priority; the moment of inertia tensor,
288     $\overleftrightarrow{\mathsf I}$, is diagonalized to obtain body-fixed
289     axes, and the charge distribution may result in a quadrupole tensor
290     that is not necessarily diagonal in the body frame. Additional
291     reasons for utilizing the primitive quadrupole are discussed in
292     section \ref{sec:damped}.
293 gezelter 3906
294     It is convenient to locate charges $q_j$ relative to the center of mass of $\bf b$. Then with $\bf{r}$ pointing from
295     $\bf a$ to $\bf b$ ($\mathbf{r}=\mathbf{r}_b - \mathbf{r}_b $), the interaction energy is given by
296     \begin{equation}
297 gezelter 3982 U_{\bf{ab}}(r)
298 gezelter 3985 = \hat{M}_a \sum_{j \, \text{in \bf b}} \frac {q_j}{\vert \bf{r}+\bf{r}_j \vert} .
299 gezelter 3982 \end{equation}
300     This can also be expanded as a Taylor series in $r$. Using a notation
301     similar to before to define the multipoles on object {\bf b},
302     \begin{equation}
303 gezelter 3906 \hat{M}_{\bf b} = C_{\bf b} + D_{{\bf b}\alpha} \frac{\partial}{\partial r_{\alpha}}
304     + Q_{{\bf b}\alpha\beta}
305     \frac{\partial^2}{\partial r_{\alpha} \partial r_{\beta}} + \dots
306     \end{equation}
307 gezelter 3982 we arrive at the multipole expression for the total interaction energy.
308 gezelter 3906 \begin{equation}
309 gezelter 3985 U_{\bf{ab}}(r)=\hat{M}_{\bf a} \hat{M}_{\bf b} \frac{1}{r} \label{kernel}.
310 gezelter 3906 \end{equation}
311 gezelter 3982 This form has the benefit of separating out the energies of
312     interaction into contributions from the charge, dipole, and quadrupole
313 gezelter 3986 of $\bf a$ interacting with the same multipoles on $\bf b$.
314 gezelter 3906
315 gezelter 3982 \subsection{Damped Coulomb interactions}
316 gezelter 3996 \label{sec:damped}
317 gezelter 3982 In the standard multipole expansion, one typically uses the bare
318     Coulomb potential, with radial dependence $1/r$, as shown in
319     Eq.~(\ref{kernel}). It is also quite common to use a damped Coulomb
320     interaction, which results from replacing point charges with Gaussian
321     distributions of charge with width $\alpha$. In damped multipole
322     electrostatics, the kernel ($1/r$) of the expansion is replaced with
323     the function:
324 gezelter 3906 \begin{equation}
325     B_0(r)=\frac{\text{erfc}(\alpha r)}{r} = \frac{2}{\sqrt{\pi}r}
326     \int_{\alpha r}^{\infty} \text{e}^{-s^2} ds .
327     \end{equation}
328 gezelter 3982 We develop equations below using the function $f(r)$ to represent
329 gezelter 3986 either $1/r$ or $B_0(r)$, and all of the techniques can be applied to
330     bare or damped Coulomb kernels (or any other function) as long as
331     derivatives of these functions are known. Smith's convenient
332 gezelter 3996 functions $B_l(r)$ are summarized in Appendix A. (N.B. there is one
333     important distinction between the two kernels, which is the behavior
334     of $\nabla^2 \frac{1}{r}$ compared with $\nabla^2 B_0(r)$. The former
335     is zero everywhere except for a delta function evaluated at the
336     origin. The latter also has delta function behavior, but is non-zero
337     for $r \neq 0$. Thus the standard justification for using a traceless
338     quadrupole tensor fails for the damped case.)
339 gezelter 3906
340 gezelter 3982 The main goal of this work is to smoothly cut off the interaction
341     energy as well as forces and torques as $r\rightarrow r_c$. To
342     describe how this goal may be met, we use two examples, charge-charge
343 gezelter 3986 and charge-dipole, using the bare Coulomb kernel, $f(r)=1/r$, to
344     explain the idea.
345 gezelter 3906
346 gezelter 3984 \subsection{Shifted-force methods}
347 gezelter 3982 In the shifted-force approximation, the interaction energy for two
348     charges $C_{\bf a}$ and $C_{\bf b}$ separated by a distance $r$ is
349     written:
350 gezelter 3906 \begin{equation}
351 gezelter 3985 U_{C_{\bf a}C_{\bf b}}(r)= C_{\bf a} C_{\bf b}
352 gezelter 3906 \left({ \frac{1}{r} - \frac{1}{r_c} + (r - r_c) \frac{1}{r_c^2} }
353     \right) .
354     \end{equation}
355 gezelter 3982 Two shifting terms appear in this equations, one from the
356 gezelter 3984 neutralization procedure ($-1/r_c$), and one that causes the first
357     derivative to vanish at the cutoff radius.
358 gezelter 3982
359     Since one derivative of the interaction energy is needed for the
360     force, the minimal perturbation is a term linear in $(r-r_c)$ in the
361     interaction energy, that is:
362 gezelter 3906 \begin{equation}
363     \frac{d\,}{dr}
364     \left( {\frac{1}{r} - \frac{1}{r_c} + (r - r_c) \frac{1}{r_c^2} }
365     \right) = \left(- \frac{1}{r^2} + \frac{1}{r_c^2}
366     \right) .
367     \end{equation}
368 gezelter 3985 which clearly vanishes as the $r$ approaches the cutoff radius. There
369     are a number of ways to generalize this derivative shift for
370 gezelter 3984 higher-order multipoles. Below, we present two methods, one based on
371     higher-order Taylor series for $r$ near $r_c$, and the other based on
372     linear shift of the kernel gradients at the cutoff itself.
373 gezelter 3906
374 gezelter 3984 \subsection{Taylor-shifted force (TSF) electrostatics}
375 gezelter 3982 In the Taylor-shifted force (TSF) method, the procedure that we follow
376     is based on a Taylor expansion containing the same number of
377     derivatives required for each force term to vanish at the cutoff. For
378     example, the quadrupole-quadrupole interaction energy requires four
379     derivatives of the kernel, and the force requires one additional
380 gezelter 3986 derivative. For quadrupole-quadrupole interactions, we therefore
381     require shifted energy expressions that include up to $(r-r_c)^5$ so
382     that all energies, forces, and torques are zero as $r \rightarrow
383     r_c$. In each case, we subtract off a function $f_n^{\text{shift}}(r)$
384     from the kernel $f(r)=1/r$. The subscript $n$ indicates the number of
385     derivatives to be taken when deriving a given multipole energy. We
386     choose a function with guaranteed smooth derivatives -- a truncated
387     Taylor series of the function $f(r)$, e.g.,
388 gezelter 3906 %
389     \begin{equation}
390 gezelter 3984 f_n^{\text{shift}}(r)=\sum_{m=0}^{n+1} \frac {(r-r_c)^m}{m!} f^{(m)}(r_c) .
391 gezelter 3906 \end{equation}
392     %
393     The combination of $f(r)$ with the shifted function is denoted $f_n(r)=f(r)-f_n^{\text{shift}}(r)$.
394     Thus, for $f(r)=1/r$, we find
395     %
396     \begin{equation}
397     f_1(r)=\frac{1}{r}- \frac{1}{r_c} + (r - r_c) \frac{1}{r_c^2} - \frac{(r-r_c)^2}{r_c^3} .
398     \end{equation}
399     %
400 gezelter 3982 Continuing with the example of a charge $\bf a$ interacting with a
401     dipole $\bf b$, we write
402 gezelter 3906 %
403     \begin{equation}
404     U_{C_{\bf a}D_{\bf b}}(r)=
405 gezelter 3985 C_{\bf a} D_{{\bf b}\alpha} \frac {\partial f_1(r) }{\partial r_\alpha}
406     = C_{\bf a} D_{{\bf b}\alpha}
407 gezelter 3906 \frac {r_\alpha}{r} \frac {\partial f_1(r)}{\partial r} .
408     \end{equation}
409     %
410 gezelter 3984 The force that dipole $\bf b$ exerts on charge $\bf a$ is
411 gezelter 3906 %
412     \begin{equation}
413 gezelter 3985 F_{C_{\bf a}D_{\bf b}\beta} = C_{\bf a} D_{{\bf b}\alpha}
414 gezelter 3906 \left[ \frac{\delta_{\alpha\beta}}{r} \frac {\partial}{\partial r} +
415     \frac{r_\alpha r_\beta}{r^2}
416     \left( -\frac{1}{r} \frac {\partial} {\partial r}
417     + \frac {\partial ^2} {\partial r^2} \right) \right] f_1(r) .
418     \end{equation}
419     %
420 gezelter 3984 For undamped coulombic interactions, $f(r)=1/r$, we find
421 gezelter 3906 %
422     \begin{equation}
423     F_{C_{\bf a}D_{\bf b}\beta} =
424 gezelter 3985 \frac{C_{\bf a} D_{{\bf b}\beta}}{r}
425 gezelter 3906 \left[ -\frac{1}{r^2}+\frac{1}{r_c^2}-\frac{2(r-r_c)}{r_c^3} \right]
426 gezelter 3985 +C_{\bf a} D_{{\bf b}\alpha}r_\alpha r_\beta
427 gezelter 3906 \left[ \frac{3}{r^5}-\frac{3}{r^3r_c^2} \right] .
428     \end{equation}
429     %
430     This expansion shows the expected $1/r^3$ dependence of the force.
431    
432 gezelter 3984 In general, we can write
433 gezelter 3906 %
434     \begin{equation}
435 gezelter 4098 U^{\text{TSF}}= (\text{prefactor}) (\text{derivatives}) f_n(r)
436 gezelter 3906 \label{generic}
437     \end{equation}
438     %
439 gezelter 3985 with $n=0$ for charge-charge, $n=1$ for charge-dipole, $n=2$ for
440     charge-quadrupole and dipole-dipole, $n=3$ for dipole-quadrupole, and
441     $n=4$ for quadrupole-quadrupole. For example, in
442     quadrupole-quadrupole interactions for which the $\text{prefactor}$ is
443     $Q_{{\bf a}\alpha\beta}Q_{{\bf b}\gamma\delta}$, the derivatives are
444     $\partial^4/\partial r_\alpha \partial r_\beta \partial
445     r_\gamma \partial r_\delta$, with implied summation combining the
446     space indices.
447 gezelter 3906
448 gezelter 3984 In the formulas presented in the tables below, the placeholder
449     function $f(r)$ is used to represent the electrostatic kernel (either
450     damped or undamped). The main functions that go into the force and
451 gezelter 3985 torque terms, $g_n(r), h_n(r), s_n(r), \mathrm{~and~} t_n(r)$ are
452     successive derivatives of the shifted electrostatic kernel, $f_n(r)$
453     of the same index $n$. The algebra required to evaluate energies,
454     forces and torques is somewhat tedious, so only the final forms are
455 gezelter 3986 presented in tables \ref{tab:tableenergy} and \ref{tab:tableFORCE}.
456 gezelter 3996 One of the principal findings of our work is that the individual
457     orientational contributions to the various multipole-multipole
458     interactions must be treated with distinct radial functions, but each
459     of these contributions is independently force shifted at the cutoff
460     radius.
461 gezelter 3906
462 gezelter 3982 \subsection{Gradient-shifted force (GSF) electrostatics}
463 gezelter 3985 The second, and conceptually simpler approach to force-shifting
464     maintains only the linear $(r-r_c)$ term in the truncated Taylor
465     expansion, and has a similar interaction energy for all multipole
466     orders:
467 gezelter 3906 \begin{equation}
468 gezelter 3996 U^{\text{GSF}} = \sum \left[ U(\mathbf{r}, \hat{\mathbf{a}}, \hat{\mathbf{b}}) -
469 gezelter 3990 U(\mathbf{r}_c,\hat{\mathbf{a}}, \hat{\mathbf{b}}) - (r-r_c) \hat{r}
470 gezelter 3996 \cdot \nabla U(\mathbf{r},\hat{\mathbf{a}}, \hat{\mathbf{b}}) \Big \lvert _{r_c} \right]
471 gezelter 3985 \label{generic2}
472 gezelter 3906 \end{equation}
473 gezelter 3996 where the sum describes a separate force-shifting that is applied to
474     each orientational contribution to the energy. Both the potential and
475     the gradient for force shifting are evaluated for an image multipole
476     projected onto the surface of the cutoff sphere (see fig
477     \ref{fig:shiftedMultipoles}). The image multipole retains the
478     orientation ($\hat{\mathbf{b}}$) of the interacting multipole. No
479 gezelter 3990 higher order terms $(r-r_c)^n$ appear. The primary difference between
480     the TSF and GSF methods is the stage at which the Taylor Series is
481     applied; in the Taylor-shifted approach, it is applied to the kernel
482     itself. In the Gradient-shifted approach, it is applied to individual
483     radial interactions terms in the multipole expansion. Energies from
484     this method thus have the general form:
485 gezelter 3906 \begin{equation}
486 gezelter 3985 U= \sum (\text{angular factor}) (\text{radial factor}).
487     \label{generic3}
488 gezelter 3906 \end{equation}
489    
490 gezelter 3986 Functional forms for both methods (TSF and GSF) can both be summarized
491 gezelter 3985 using the form of Eq.~(\ref{generic3}). The basic forms for the
492     energy, force, and torque expressions are tabulated for both shifting
493 gezelter 3986 approaches below -- for each separate orientational contribution, only
494 gezelter 3985 the radial factors differ between the two methods.
495 gezelter 3906
496     \subsection{\label{sec:level2}Body and space axes}
497 gezelter 3989 Although objects $\bf a$ and $\bf b$ rotate during a molecular
498     dynamics (MD) simulation, their multipole tensors remain fixed in
499     body-frame coordinates. While deriving force and torque expressions,
500     it is therefore convenient to write the energies, forces, and torques
501     in intermediate forms involving the vectors of the rotation matrices.
502     We denote body axes for objects $\bf a$ and $\bf b$ using unit vectors
503     $\hat{a}_m$ and $\hat{b}_m$, respectively, with the index $m=(123)$.
504     In a typical simulation , the initial axes are obtained by
505     diagonalizing the moment of inertia tensors for the objects. (N.B.,
506     the body axes are generally {\it not} the same as those for which the
507     quadrupole moment is diagonal.) The rotation matrices are then
508     propagated during the simulation.
509 gezelter 3906
510 gezelter 3989 The rotation matrices $\hat{\mathbf {a}}$ and $\hat{\mathbf {b}}$ can be
511 gezelter 3985 expressed using these unit vectors:
512 gezelter 3906 \begin{eqnarray}
513     \hat{\mathbf {a}} =
514     \begin{pmatrix}
515     \hat{a}_1 \\
516     \hat{a}_2 \\
517     \hat{a}_3
518 gezelter 3989 \end{pmatrix}, \qquad
519 gezelter 3906 \hat{\mathbf {b}} =
520     \begin{pmatrix}
521     \hat{b}_1 \\
522     \hat{b}_2 \\
523     \hat{b}_3
524     \end{pmatrix}
525     \end{eqnarray}
526     %
527 gezelter 3985 These matrices convert from space-fixed $(xyz)$ to body-fixed $(123)$
528 gezelter 3989 coordinates.
529    
530     Allen and Germano,\cite{Allen:2006fk} following earlier work by Price
531     {\em et al.},\cite{Price:1984fk} showed that if the interaction
532     energies are written explicitly in terms of $\hat{r}$ and the body
533     axes ($\hat{a}_m$, $\hat{b}_n$) :
534 gezelter 3906 %
535 gezelter 3985 \begin{equation}
536 gezelter 3989 U(r, \{\hat{a}_m \cdot \hat{r} \},
537     \{\hat{b}_n\cdot \hat{r} \},
538     \{\hat{a}_m \cdot \hat{b}_n \}) .
539     \label{ugeneral}
540     \end{equation}
541     %
542     the forces come out relatively cleanly,
543     %
544     \begin{equation}
545     \mathbf{F}_{\bf a}=-\mathbf{F}_{\bf b} = \frac{\partial U}{\partial \mathbf{r}}
546     = \frac{\partial U}{\partial r} \hat{r}
547     + \sum_m \left[
548     \frac{\partial U}{\partial (\hat{a}_m \cdot \hat{r})}
549     \frac { \partial (\hat{a}_m \cdot \hat{r})}{\partial \mathbf{r}}
550     + \frac{\partial U}{\partial (\hat{b}_m \cdot \hat{r})}
551     \frac { \partial (\hat{b}_m \cdot \hat{r})}{\partial \mathbf{r}}
552     \right] \label{forceequation}.
553     \end{equation}
554    
555     The torques can also be found in a relatively similar
556     manner,
557     %
558     \begin{eqnarray}
559     \mathbf{\tau}_{\bf a} =
560     \sum_m
561     \frac{\partial U}{\partial (\hat{a}_m \cdot \hat{r})}
562     ( \hat{r} \times \hat{a}_m )
563     -\sum_{mn}
564     \frac{\partial U}{\partial (\hat{a}_m \cdot \hat{b}_n)}
565     (\hat{a}_m \times \hat{b}_n) \\
566     %
567     \mathbf{\tau}_{\bf b} =
568     \sum_m
569     \frac{\partial U}{\partial (\hat{b}_m \cdot \hat{r})}
570     ( \hat{r} \times \hat{b}_m)
571     +\sum_{mn}
572     \frac{\partial U}{\partial (\hat{a}_m \cdot \hat{b}_n)}
573     (\hat{a}_m \times \hat{b}_n) .
574     \end{eqnarray}
575    
576     Note that our definition of $\mathbf{r}=\mathbf{r}_b - \mathbf{r}_b $
577     is opposite in sign to that of Allen and Germano.\cite{Allen:2006fk}
578     We also made use of the identities,
579     %
580     \begin{align}
581     \frac { \partial (\hat{a}_m \cdot \hat{r})}{\partial \mathbf{r}}
582     =& \frac{1}{r} \left( \hat{a}_m - (\hat{a}_m \cdot \hat{r})\hat{r}
583     \right) \\
584     \frac { \partial (\hat{b}_m \cdot \hat{r})}{\partial \mathbf{r}}
585     =& \frac{1}{r} \left( \hat{b}_m - (\hat{b}_m \cdot \hat{r})\hat{r}
586     \right) .
587     \end{align}
588    
589     Many of the multipole contractions required can be written in one of
590     three equivalent forms using the unit vectors $\hat{r}$, $\hat{a}_m$,
591     and $\hat{b}_n$. In the torque expressions, it is useful to have the
592     angular-dependent terms available in all three fashions, e.g. for the
593     dipole-dipole contraction:
594     %
595     \begin{equation}
596 gezelter 3906 \mathbf{D}_{\mathbf {a}} \cdot \mathbf{D}_{\mathbf{b}}
597 gezelter 3985 = D_{\bf {a}\alpha} D_{\bf {b}\alpha} =
598     \sum_{mn} {D_{\mathbf{a}m} \hat{a}_m \cdot \hat{b}_n D_{\mathbf{b}n}}
599     \end{equation}
600 gezelter 3906 %
601 gezelter 3985 The first two forms are written using space coordinates. The first
602     form is standard in the chemistry literature, while the second is
603     expressed using implied summation notation. The third form shows
604     explicit sums over body indices and the dot products now indicate
605     contractions using space indices.
606 gezelter 3906
607 gezelter 3989 In computing our force and torque expressions, we carried out most of
608     the work in body coordinates, and have transformed the expressions
609     back to space-frame coordinates, which are reported below. Interested
610     readers may consult the supplemental information for this paper for
611     the intermediate body-frame expressions.
612 gezelter 3906
613 gezelter 3982 \subsection{The Self-Interaction \label{sec:selfTerm}}
614    
615 gezelter 3985 In addition to cutoff-sphere neutralization, the Wolf
616     summation~\cite{Wolf99} and the damped shifted force (DSF)
617 gezelter 4098 extension~\cite{Fennell:2006zl} also include self-interactions that
618 gezelter 3985 are handled separately from the pairwise interactions between
619     sites. The self-term is normally calculated via a single loop over all
620     sites in the system, and is relatively cheap to evaluate. The
621     self-interaction has contributions from two sources.
622    
623     First, the neutralization procedure within the cutoff radius requires
624     a contribution from a charge opposite in sign, but equal in magnitude,
625     to the central charge, which has been spread out over the surface of
626     the cutoff sphere. For a system of undamped charges, the total
627     self-term is
628 gezelter 3980 \begin{equation}
629 gezelter 4175 U_\textrm{self} = - \frac{1}{r_c} \sum_{{\bf a}=1}^N C_{\bf a}^{2}
630 gezelter 3980 \label{eq:selfTerm}
631     \end{equation}
632 gezelter 3985
633     Second, charge damping with the complementary error function is a
634     partial analogy to the Ewald procedure which splits the interaction
635     into real and reciprocal space sums. The real space sum is retained
636     in the Wolf and DSF methods. The reciprocal space sum is first
637     minimized by folding the largest contribution (the self-interaction)
638     into the self-interaction from charge neutralization of the damped
639     potential. The remainder of the reciprocal space portion is then
640     discarded (as this contributes the largest computational cost and
641     complexity to the Ewald sum). For a system containing only damped
642     charges, the complete self-interaction can be written as
643 gezelter 3980 \begin{equation}
644 gezelter 4175 U_\textrm{self} = - \left(\frac{\textrm{erfc}(\alpha r_c)}{r_c} +
645 gezelter 3980 \frac{\alpha}{\sqrt{\pi}} \right) \sum_{{\bf a}=1}^N
646     C_{\bf a}^{2}.
647     \label{eq:dampSelfTerm}
648     \end{equation}
649    
650     The extension of DSF electrostatics to point multipoles requires
651     treatment of {\it both} the self-neutralization and reciprocal
652     contributions to the self-interaction for higher order multipoles. In
653     this section we give formulae for these interactions up to quadrupolar
654     order.
655    
656     The self-neutralization term is computed by taking the {\it
657     non-shifted} kernel for each interaction, placing a multipole of
658     equal magnitude (but opposite in polarization) on the surface of the
659     cutoff sphere, and averaging over the surface of the cutoff sphere.
660     Because the self term is carried out as a single sum over sites, the
661     reciprocal-space portion is identical to half of the self-term
662     obtained by Smith and Aguado and Madden for the application of the
663     Ewald sum to multipoles.\cite{Smith82,Smith98,Aguado03} For a given
664 gezelter 4098 site which posesses a charge, dipole, and quadrupole, both types of
665 gezelter 3980 contribution are given in table \ref{tab:tableSelf}.
666    
667     \begin{table*}
668     \caption{\label{tab:tableSelf} Self-interaction contributions for
669     site ({\bf a}) that has a charge $(C_{\bf a})$, dipole
670     $(\mathbf{D}_{\bf a})$, and quadrupole $(\mathbf{Q}_{\bf a})$}
671     \begin{ruledtabular}
672     \begin{tabular}{lccc}
673     Multipole order & Summed Quantity & Self-neutralization & Reciprocal \\ \hline
674     Charge & $C_{\bf a}^2$ & $-f(r_c)$ & $-\frac{\alpha}{\sqrt{\pi}}$ \\
675     Dipole & $|\mathbf{D}_{\bf a}|^2$ & $\frac{1}{3} \left( h(r_c) +
676     \frac{2 g(r_c)}{r_c} \right)$ & $-\frac{2 \alpha^3}{3 \sqrt{\pi}}$\\
677 gezelter 3989 Quadrupole & $2 \mathbf{Q}_{\bf a}:\mathbf{Q}_{\bf a} + \text{Tr}(\mathbf{Q}_{\bf a})^2$ &
678 gezelter 3980 $- \frac{1}{15} \left( t(r_c)+ \frac{4 s(r_c)}{r_c} \right)$ &
679     $-\frac{4 \alpha^5}{5 \sqrt{\pi}}$ \\
680     Charge-Quadrupole & $-2 C_{\bf a} \text{Tr}(\mathbf{Q}_{\bf a})$ & $\frac{1}{3} \left(
681     h(r_c) + \frac{2 g(r_c)}{r_c} \right)$& $-\frac{2 \alpha^3}{3 \sqrt{\pi}}$ \\
682     \end{tabular}
683     \end{ruledtabular}
684     \end{table*}
685    
686     For sites which simultaneously contain charges and quadrupoles, the
687     self-interaction includes a cross-interaction between these two
688     multipole orders. Symmetry prevents the charge-dipole and
689     dipole-quadrupole interactions from contributing to the
690     self-interaction. The functions that go into the self-neutralization
691 gezelter 3985 terms, $g(r), h(r), s(r), \mathrm{~and~} t(r)$ are successive
692     derivatives of the electrostatic kernel, $f(r)$ (either the undamped
693     $1/r$ or the damped $B_0(r)=\mathrm{erfc}(\alpha r)/r$ function) that
694     have been evaluated at the cutoff distance. For undamped
695     interactions, $f(r_c) = 1/r_c$, $g(r_c) = -1/r_c^{2}$, and so on. For
696     damped interactions, $f(r_c) = B_0(r_c)$, $g(r_c) = B_0'(r_c)$, and so
697     on. Appendix \ref{SmithFunc} contains recursion relations that allow
698     rapid evaluation of these derivatives.
699 gezelter 3980
700 gezelter 3985 \section{Interaction energies, forces, and torques}
701     The main result of this paper is a set of expressions for the
702     energies, forces and torques (up to quadrupole-quadrupole order) that
703     work for both the Taylor-shifted and Gradient-shifted approximations.
704     These expressions were derived using a set of generic radial
705     functions. Without using the shifting approximations mentioned above,
706     some of these radial functions would be identical, and the expressions
707     coalesce into the familiar forms for unmodified multipole-multipole
708     interactions. Table \ref{tab:tableenergy} maps between the generic
709     functions and the radial functions derived for both the Taylor-shifted
710     and Gradient-shifted methods. The energy equations are written in
711     terms of lab-frame representations of the dipoles, quadrupoles, and
712     the unit vector connecting the two objects,
713 gezelter 3906
714     % Energy in space coordinate form ----------------------------------------------------------------------------------------------
715     %
716     %
717     % u ca cb
718     %
719 gezelter 3983 \begin{align}
720     U_{C_{\bf a}C_{\bf b}}(r)=&
721 gezelter 3985 C_{\bf a} C_{\bf b} v_{01}(r) \label{uchch}
722 gezelter 3983 \\
723 gezelter 3906 %
724     % u ca db
725     %
726 gezelter 3983 U_{C_{\bf a}D_{\bf b}}(r)=&
727 gezelter 3985 C_{\bf a} \left( \mathbf{D}_{\mathbf{b}} \cdot \hat{r} \right) v_{11}(r)
728 gezelter 3906 \label{uchdip}
729 gezelter 3983 \\
730 gezelter 3906 %
731     % u ca qb
732     %
733 gezelter 3985 U_{C_{\bf a}Q_{\bf b}}(r)=& C_{\bf a } \Bigl[ \text{Tr}Q_{\bf b}
734     v_{21}(r) + \left( \hat{r} \cdot \mathbf{Q}_{{\mathbf b}} \cdot
735     \hat{r} \right) v_{22}(r) \Bigr]
736 gezelter 3906 \label{uchquad}
737 gezelter 3983 \\
738 gezelter 3906 %
739     % u da cb
740     %
741 gezelter 3983 %U_{D_{\bf a}C_{\bf b}}(r)=&
742     %-\frac{C_{\bf b}}{4\pi \epsilon_0}
743     %\left( \mathbf{D}_{\mathbf{a}} \cdot \hat{r} \right) v_{11}(r) \label{udipch}
744     %\\
745 gezelter 3906 %
746     % u da db
747     %
748 gezelter 3983 U_{D_{\bf a}D_{\bf b}}(r)=&
749 gezelter 3985 -\Bigr[ \left( \mathbf{D}_{\mathbf {a}} \cdot
750 gezelter 3906 \mathbf{D}_{\mathbf{b}} \right) v_{21}(r)
751     +\left( \mathbf{D}_{\mathbf {a}} \cdot \hat{r} \right)
752     \left( \mathbf{D}_{\mathbf {b}} \cdot \hat{r} \right)
753     v_{22}(r) \Bigr]
754     \label{udipdip}
755 gezelter 3983 \\
756 gezelter 3906 %
757     % u da qb
758     %
759     \begin{split}
760     % 1
761 gezelter 3983 U_{D_{\bf a}Q_{\bf b}}(r) =&
762 gezelter 3985 -\Bigl[
763 gezelter 3906 \text{Tr}\mathbf{Q}_{\mathbf{b}}
764     \left( \mathbf{D}_{\mathbf{a}} \cdot \hat{r} \right)
765     +2 ( \mathbf{D}_{\mathbf{a}} \cdot
766     \mathbf{Q}_{\mathbf{b}} \cdot \hat{r} ) \Bigr] v_{31}(r) \\
767     % 2
768 gezelter 3985 &- \left( \mathbf{D}_{\mathbf{a}} \cdot \hat{r} \right)
769 gezelter 3906 \left( \hat{r} \cdot \mathbf{Q}_{{\mathbf b}} \cdot \hat{r} \right) v_{32}(r)
770     \label{udipquad}
771     \end{split}
772 gezelter 3983 \\
773 gezelter 3906 %
774     % u qa cb
775     %
776 gezelter 3983 %U_{Q_{\bf a}C_{\bf b}}(r)=&
777     %\frac{C_{\bf b }}{4\pi \epsilon_0} \Bigl[ \text{Tr}\mathbf{Q}_{\bf a} v_{21}(r)
778     %\left( \hat{r} \cdot \mathbf{Q}_{{\mathbf a}} \cdot \hat{r} \right) v_{22}(r) \Bigr]
779     %\label{uquadch}
780     %\\
781 gezelter 3906 %
782     % u qa db
783     %
784 gezelter 3983 %\begin{split}
785 gezelter 3906 %1
786 gezelter 3983 %U_{Q_{\bf a}D_{\bf b}}(r)=&
787     %\frac{1}{4\pi \epsilon_0} \Bigl[
788     %\text{Tr}\mathbf{Q}_{\mathbf{a}}
789     %\left( \mathbf{D}_{\mathbf{b}} \cdot \hat{r} \right)
790     %+2 ( \mathbf{D}_{\mathbf{b}} \cdot
791     %\mathbf{Q}_{\mathbf{a}} \cdot \hat{r}) \Bigr] v_{31}(r)\\
792 gezelter 3906 % 2
793 gezelter 3983 %&+\frac{1}{4\pi \epsilon_0}
794     %\left( \mathbf{D}_{\mathbf{b}} \cdot \hat{r} \right)
795     %\left( \hat{r} \cdot \mathbf{Q}_{{\mathbf a}} \cdot \hat{r} \right) v_{32}(r)
796     %\label{uquaddip}
797     %\end{split}
798     %\\
799 gezelter 3906 %
800     % u qa qb
801     %
802     \begin{split}
803     %1
804 gezelter 3983 U_{Q_{\bf a}Q_{\bf b}}(r)=&
805 gezelter 3985 \Bigl[
806 gezelter 3906 \text{Tr} \mathbf{Q}_{\mathbf{a}} \text{Tr} \mathbf{Q}_{\mathbf{b}}
807 gezelter 3989 +2
808     \mathbf{Q}_{\mathbf{a}} : \mathbf{Q}_{\mathbf{b}} \Bigr] v_{41}(r)
809 gezelter 3906 \\
810     % 2
811 gezelter 3985 &+\Bigl[ \text{Tr}\mathbf{Q}_{\mathbf{a}}
812 gezelter 3906 \left( \hat{r} \cdot
813     \mathbf{Q}_{{\mathbf b}} \cdot \hat{r} \right)
814     +\text{Tr}\mathbf{Q}_{\mathbf{b}}
815     \left( \hat{r} \cdot \mathbf{Q}_{{\mathbf a}}
816     \cdot \hat{r} \right) +4 (\hat{r} \cdot
817     \mathbf{Q}_{{\mathbf a}}\cdot \mathbf{Q}_{{\mathbf b}} \cdot \hat{r})
818     \Bigr] v_{42}(r)
819     \\
820     % 4
821 gezelter 3985 &+
822 gezelter 3906 \left( \hat{r} \cdot \mathbf{Q}_{{\mathbf a}} \cdot \hat{r} \right)
823     \left( \hat{r} \cdot \mathbf{Q}_{{\mathbf b}} \cdot \hat{r} \right) v_{43}(r).
824     \label{uquadquad}
825     \end{split}
826 gezelter 3983 \end{align}
827 gezelter 3985 %
828 gezelter 3983 Note that the energies of multipoles on site $\mathbf{b}$ interacting
829     with those on site $\mathbf{a}$ can be obtained by swapping indices
830     along with the sign of the intersite vector, $\hat{r}$.
831 gezelter 3906
832     %
833     %
834     % TABLE of radial functions ----------------------------------------------------------------------------------------------------------------
835     %
836    
837 gezelter 3985 \begin{sidewaystable}
838     \caption{\label{tab:tableenergy}Radial functions used in the energy
839     and torque equations. The $f, g, h, s, t, \mathrm{and} u$
840     functions used in this table are defined in Appendices B and C.}
841     \begin{tabular}{|c|c|l|l|} \hline
842     Generic&Bare Coulomb&Taylor-Shifted&Gradient-Shifted
843 gezelter 3906 \\ \hline
844     %
845     %
846     %
847     %Ch-Ch&
848     $v_{01}(r)$ &
849     $\frac{1}{r}$ &
850     $f_0(r)$ &
851     $f(r)-f(r_c)-(r-r_c)g(r_c)$
852     \\
853     %
854     %
855     %
856     %Ch-Di&
857     $v_{11}(r)$ &
858     $-\frac{1}{r^2}$ &
859     $g_1(r)$ &
860     $g(r)-g(r_c)-(r-r_c)h(r_c)$ \\
861     %
862     %
863     %
864     %Ch-Qu/Di-Di&
865     $v_{21}(r)$ &
866     $-\frac{1}{r^3} $ &
867     $\frac{g_2(r)}{r} $ &
868     $\frac{g(r)}{r}-\frac{g(r_c)}{r_c} -(r-r_c)
869     \left( -\frac{g(r_c)}{r_c^2} + \frac{h(r_c)}{r_c} \right)$ \\
870     $v_{22}(r)$ &
871     $\frac{3}{r^3} $ &
872     $\left(-\frac{g_2(r)}{r} + h_2(r) \right)$ &
873     $\left(-\frac{g(r)}{r}+h(r) \right)
874 gezelter 3985 -\left(-\frac{g(r_c)}{r_c}+h(r_c) \right)$ \\
875     &&& $ ~~~-(r-r_c) \left( \frac{g(r_c)}{r_c^2}-\frac{h(r_c)}{r_c}+s(r_c) \right)$
876 gezelter 3906 \\
877     %
878     %
879     %
880     %Di-Qu &
881     $v_{31}(r)$ &
882     $\frac{3}{r^4} $ &
883     $\left(-\frac{g_3(r)}{r^2} + \frac{h_3(r)}{r} \right)$ &
884     $\left( -\frac{g(r)}{r^2}+\frac{h(r)}{r} \right)
885     -\left(-\frac{g(r_c)}{r_c^2}+\frac{h(r_c)}{r_c} \right) $\\
886 gezelter 3985 &&&$ ~~~ -(r-r_c) \left(\frac{2g(r_c)}{r_c^3}-\frac{2h(r_c)}{r_c^2}+\frac{s(r_c)}{r_c} \right)$
887 gezelter 3906 \\
888     %
889     $v_{32}(r)$ &
890     $-\frac{15}{r^4} $ &
891     $\left( \frac{3g_3(r)}{r^2} - \frac{3h_3(r)}{r} + s_3(r) \right)$ &
892     $\left( \frac{3g(r)}{r^2} - \frac{3h(r)}{r} + s(r) \right)
893     - \left( \frac{3g(r_c)}{r_c^2} - \frac{3h(r_c)}{r_c} + s(r_c) \right)$ \\
894 gezelter 3985 &&&$ ~~~ -(r-r_c) \left( \frac{-6g(r_c)}{r_c^3}+\frac{6h(r_c)}{r_c^2}-\frac{3s(r_c)}{r_c}+t(r_c) \right)$
895 gezelter 3906 \\
896     %
897     %
898     %
899     %Qu-Qu&
900     $v_{41}(r)$ &
901     $\frac{3}{r^5} $ &
902     $\left(-\frac{g_4(r)}{r^3} +\frac{h_4(r)}{r^2} \right) $ &
903     $\left( -\frac{g(r)}{r^3} + \frac{h(r)}{r^2} \right)
904     - \left( -\frac{g(r_c)}{r_c^3} + \frac{h(r_c)}{r_c^2} \right)$ \\
905 gezelter 3985 &&&$ ~~~ -(r-r_c) \left( \frac{3g(r_c)}{r_c^4}-\frac{3h(r_c)}{r_c^3}+\frac{s(r_c)}{r_c^2} \right)$
906 gezelter 3906 \\
907     % 2
908     $v_{42}(r)$ &
909     $- \frac{15}{r^5} $ &
910     $\left( \frac{3g_4(r)}{r^3} - \frac{3h_4(r)}{r^2}+\frac{s_4(r)}{r} \right)$ &
911     $\left( \frac{3g(r)}{r^3} - \frac{3h(r)}{r^2}+\frac{s(r)}{r} \right)
912     -\left( \frac{3g(r_c)}{r_c^3} - \frac{3h(r_c)}{r_c^2}+\frac{s(r_c)}{r_c} \right)$ \\
913 gezelter 3985 &&&$ ~~~ -(r-r_c) \left(- \frac{9g(r_c)}{r_c^4}+\frac{9h(r_c)}{r_c^3}
914 gezelter 3906 -\frac{4s(r_c)}{r_c^2} + \frac{t(r_c)}{r_c}\right)$
915     \\
916     % 3
917     $v_{43}(r)$ &
918     $ \frac{105}{r^5} $ &
919     $\left(-\frac{15g_4(r)}{r^3}+\frac{15h_4(r)}{r^2}-\frac{6s_4(r)}{r} + t_4(r)\right) $ &
920     $\left(-\frac{15g(r)}{r^3}+\frac{15h(r)}{r^2}-\frac{6s(r)}{r} + t(r)\right)$ \\
921 gezelter 3985 &&&$~~~ -\left(-\frac{15g(r_c)}{r_c^3}+\frac{15h(r_c)}{r_c^2}-\frac{6s(r_c)}{r_c} + t(r_c)\right)$ \\
922     &&&$~~~ -(r-r_c)\left(\frac{45g(r_c)}{r_c^4}-\frac{45h(r_c)}{r_c^3}+\frac{21s(r_c)}{r_c^2}
923     -\frac{6t(r_c)}{r_c}+u(r_c) \right)$ \\ \hline
924 gezelter 3906 \end{tabular}
925 gezelter 3985 \end{sidewaystable}
926 gezelter 3906 %
927     %
928     % FORCE TABLE of radial functions ----------------------------------------------------------------------------------------------------------------
929     %
930    
931 gezelter 3985 \begin{sidewaystable}
932 gezelter 3906 \caption{\label{tab:tableFORCE}Radial functions used in the force equations.}
933 gezelter 3985 \begin{tabular}{|c|c|l|l|} \hline
934     Function&Definition&Taylor-Shifted&Gradient-Shifted
935 gezelter 3906 \\ \hline
936     %
937     %
938     %
939     $w_a(r)$&
940 gezelter 3985 $\frac{d v_{01}}{dr}$&
941     $g_0(r)$&
942     $g(r)-g(r_c)$ \\
943 gezelter 3906 %
944     %
945     $w_b(r)$ &
946 gezelter 3985 $\frac{d v_{11}}{dr} - \frac{v_{11}(r)}{r} $&
947     $\left( -\frac{g_1(r)}{r}+h_1(r) \right)$ &
948     $h(r)- h(r_c) - \frac{v_{11}(r)}{r} $ \\
949 gezelter 3906 %
950     $w_c(r)$ &
951 gezelter 3985 $\frac{v_{11}(r)}{r}$ &
952     $\frac{g_1(r)}{r} $ &
953     $\frac{v_{11}(r)}{r}$\\
954 gezelter 3906 %
955     %
956     $w_d(r)$&
957 gezelter 3985 $\frac{d v_{21}}{dr}$&
958     $\left( -\frac{g_2(r)}{r^2} + \frac{h_2(r)}{r} \right) $ &
959     $\left( -\frac{g(r)}{r^2} + \frac{h(r)}{r} \right)
960     -\left( -\frac{g(r_c)}{r_c^2} + \frac{h(r_c)}{r_c} \right) $ \\
961 gezelter 3906 %
962     $w_e(r)$ &
963 gezelter 3985 $\left(-\frac{g_2(r)}{r^2} + \frac{h_2(r)}{r} \right)$ &
964     $\frac{v_{22}(r)}{r}$ &
965 gezelter 3906 $\frac{v_{22}(r)}{r}$ \\
966     %
967     %
968     $w_f(r)$&
969 gezelter 3985 $\frac{d v_{22}}{dr} - \frac{2v_{22}(r)}{r}$&
970     $\left( \frac{3g_2(r)}{r^2}-\frac{3h_2(r)}{r}+s_2(r) \right)$ &
971     $ \left( \frac{g(r)}{r^2}-\frac{h(r)}{r}+s(r) \right) $ \\
972     &&& $ ~~~- \left( \frac{g(r_c)}{r_c^2}-\frac{h(r_c)}{r_c}+s(r_c)
973     \right)-\frac{2v_{22}(r)}{r}$\\
974 gezelter 3906 %
975     $w_g(r)$&
976 gezelter 3985 $\frac{v_{31}(r)}{r}$&
977     $ \left( -\frac{g_3(r)}{r^3}+\frac{h_3(r)}{r^2} \right)$&
978 gezelter 3906 $\frac{v_{31}(r)}{r}$\\
979     %
980     $w_h(r)$ &
981 gezelter 3985 $\frac{d v_{31}}{dr} -\frac{v_{31}(r)}{r}$&
982     $\left(\frac{3g_3(r)}{r^3} -\frac{3h_3(r)}{r^2} +\frac{s_3(r)}{r} \right) $ &
983     $ \left(\frac{2g(r)}{r^3} -\frac{2h(r)}{r^2} +\frac{s(r)}{r} \right) - \left(\frac{2g(r_c)}{r_c^3} -\frac{2h(r_c)}{r_c^2} +\frac{s(r_c)}{r_c} \right) $ \\
984     &&& $ ~~~ -\frac{v_{31}(r)}{r}$ \\
985 gezelter 3906 % 2
986     $w_i(r)$ &
987 gezelter 3985 $\frac{v_{32}(r)}{r}$ &
988     $\left(\frac{3g_3(r)}{r^3} -\frac{3h_3(r)}{r^2} +\frac{s_3(r)}{r} \right) $ &
989     $\frac{v_{32}(r)}{r}$\\
990 gezelter 3906 %
991     $w_j(r)$ &
992 gezelter 3985 $\frac{d v_{32}}{dr} - \frac{3v_{32}}{r}$&
993     $\left(\frac{-15g_3(r)}{r^3} + \frac{15h_3(r)}{r^2} - \frac{6s_3(r)}{r} + t_3(r) \right) $ &
994     $\left(\frac{-6g(r)}{r^3} +\frac{6h(r)}{r^2} -\frac{3s(r)}{r} +t(r) \right)$ \\
995     &&& $~~~-\left(\frac{-6g(_cr)}{r_c^3} +\frac{6h(r_c)}{r_c^2}
996     -\frac{3s(r_c)}{r_c} +t(r_c) \right) -\frac{3v_{32}}{r}$ \\
997 gezelter 3906 %
998     $w_k(r)$ &
999 gezelter 3985 $\frac{d v_{41}}{dr} $ &
1000     $\left(\frac{3g_4(r)}{r^4} -\frac{3h_4(r)}{r^3} +\frac{s_4(r)}{r^2} \right)$ &
1001     $\left(\frac{3g(r)}{r^4} -\frac{3h(r)}{r^3} +\frac{s(r)}{r^2} \right)
1002     -\left(\frac{3g(r_c)}{r_c^4} -\frac{3h(r_c)}{r_c^3} +\frac{s(r_c)}{r_c^2} \right)$ \\
1003 gezelter 3906 %
1004     $w_l(r)$ &
1005 gezelter 3985 $\frac{d v_{42}}{dr} -\frac{2v_{42}(r)}{r}$ &
1006     $\left(-\frac{15g_4(r)}{r^4} +\frac{15h_4(r)}{r^3} -\frac{6s_4(r)}{r^2} +\frac{t_4(r)}{r} \right)$ &
1007     $\left(-\frac{9g(r)}{r^4} +\frac{9h(r)}{r^3} -\frac{4s(r)}{r^2} +\frac{t(r)}{r} \right)$ \\
1008     &&& $~~~ -\left(-\frac{9g(r_c)}{r_c^4} +\frac{9h(r_c)}{r_c^3} -\frac{4s(r_c)}{r_c^2} +\frac{t(r_c)}{r_c} \right)
1009     -\frac{2v_{42}(r)}{r}$\\
1010 gezelter 3906 %
1011     $w_m(r)$ &
1012 gezelter 3985 $\frac{d v_{43}}{dr} -\frac{4v_{43}(r)}{r}$&
1013     $\left(\frac{105g_4(r)}{r^4} - \frac{105h_4(r)}{r^3} + \frac{45s_4(r)}{r^2} - \frac{10t_4(r)}{r} +u_4(r) \right)$ &
1014     $\left(\frac{45g(r)}{r^4} -\frac{45h(r)}{r^3} +\frac{21s(r)}{r^2} -\frac{6t(r)}{r} +u(r) \right)$\\
1015     &&& $~~~- \left(\frac{45g(r_c)}{r_c^4} -\frac{45h(r_c)}{r_c^3}
1016     +\frac{21s(r_c)}{r_c^2} -\frac{6t(r_c)}{r_c} +u(r_c) \right) $\\
1017     &&& $~~~-\frac{4v_{43}(r)}{r}$ \\
1018 gezelter 3906 %
1019     $w_n(r)$ &
1020 gezelter 3985 $\frac{v_{42}(r)}{r}$ &
1021     $\left(\frac{3g_4(r)}{r^4} -\frac{3h_4(r)}{r^3} +\frac{s_4(r)}{r^2} \right)$ &
1022     $\frac{v_{42}(r)}{r}$\\
1023 gezelter 3906 %
1024     $w_o(r)$ &
1025 gezelter 3985 $\frac{v_{43}(r)}{r}$&
1026     $\left(-\frac{15g_4(r)}{r^4} +\frac{15h_4(r)}{r^3} -\frac{6s_4(r)}{r^2} +\frac{t_4(r)}{r} \right)$ &
1027     $\frac{v_{43}(r)}{r}$ \\ \hline
1028 gezelter 3906 %
1029    
1030     \end{tabular}
1031 gezelter 3985 \end{sidewaystable}
1032 gezelter 3906 %
1033     %
1034     %
1035    
1036     \subsection{Forces}
1037 gezelter 3985 The force on object $\bf{a}$, $\mathbf{F}_{\bf a}$, due to object
1038     $\bf{b}$ is the negative of the force on $\bf{b}$ due to $\bf{a}$. For
1039     a simple charge-charge interaction, these forces will point along the
1040     $\pm \hat{r}$ directions, where $\mathbf{r}=\mathbf{r}_b -
1041     \mathbf{r}_a $. Thus
1042 gezelter 3906 %
1043     \begin{equation}
1044     F_{\bf a \alpha} = \hat{r}_\alpha \frac{\partial U_{C_{\bf a}C_{\bf b}}}{\partial r}
1045     \quad \text{and} \quad F_{\bf b \alpha}
1046     = - \hat{r}_\alpha \frac{\partial U_{C_{\bf a}C_{\bf b}}} {\partial r} .
1047     \end{equation}
1048     %
1049 gezelter 3985 We list below the force equations written in terms of lab-frame
1050     coordinates. The radial functions used in the two methods are listed
1051     in Table \ref{tab:tableFORCE}
1052 gezelter 3906 %
1053 gezelter 3985 %SPACE COORDINATES FORCE EQUATIONS
1054 gezelter 3906 %
1055     % **************************************************************************
1056     % f ca cb
1057     %
1058 gezelter 3985 \begin{align}
1059     \mathbf{F}_{{\bf a}C_{\bf a}C_{\bf b}} =&
1060     C_{\bf a} C_{\bf b} w_a(r) \hat{r} \\
1061 gezelter 3906 %
1062     %
1063     %
1064 gezelter 3985 \mathbf{F}_{{\bf a}C_{\bf a}D_{\bf b}} =&
1065     C_{\bf a} \Bigl[
1066 gezelter 3906 \left( \hat{r} \cdot \mathbf{D}_{\mathbf{b}} \right)
1067     w_b(r) \hat{r}
1068 gezelter 3985 + \mathbf{D}_{\mathbf{b}} w_c(r) \Bigr] \\
1069 gezelter 3906 %
1070     %
1071     %
1072 gezelter 3985 \mathbf{F}_{{\bf a}C_{\bf a}Q_{\bf b}} =&
1073     C_{\bf a } \Bigr[
1074 gezelter 3906 \text{Tr}\mathbf{Q}_{\bf b} w_d(r) \hat{r}
1075     + 2 \mathbf{Q}_{{\mathbf b}} \cdot \hat{r} w_e(r)
1076 gezelter 3985 + \left( \hat{r} \cdot \mathbf{Q}_{{\mathbf b}} \cdot \hat{r}
1077     \right) w_f(r) \hat{r} \Bigr] \\
1078 gezelter 3906 %
1079     %
1080     %
1081 gezelter 3985 % \begin{equation}
1082     % \mathbf{F}_{{\bf a}D_{\bf a}C_{\bf b}} =
1083     % -C_{\bf{b}} \Bigl[
1084     % \left( \hat{r} \cdot \mathbf{D}_{\mathbf{a}} \right) w_b(r) \hat{r}
1085     % + \mathbf{D}_{\mathbf{a}} w_c(r) \Bigr]
1086     % \end{equation}
1087 gezelter 3906 %
1088     %
1089     %
1090 gezelter 3985 \begin{split}
1091     \mathbf{F}_{{\bf a}D_{\bf a}D_{\bf b}} =&
1092 gezelter 3906 - \mathbf{D}_{\mathbf {a}} \cdot \mathbf{D}_{\mathbf{b}} w_d(r) \hat{r}
1093     + \left( \mathbf{D}_{\mathbf {a}}
1094     \left( \mathbf{D}_{\mathbf{b}} \cdot \hat{r} \right)
1095 gezelter 3985 + \mathbf{D}_{\mathbf {b}} \left( \mathbf{D}_{\mathbf{a}} \cdot \hat{r} \right) \right) w_e(r)\\
1096 gezelter 3906 % 2
1097 gezelter 3985 & - \left( \hat{r} \cdot \mathbf{D}_{\mathbf {a}} \right)
1098     \left( \hat{r} \cdot \mathbf{D}_{\mathbf {b}} \right) w_f(r) \hat{r}
1099     \end{split}\\
1100 gezelter 3906 %
1101     %
1102     %
1103     \begin{split}
1104 gezelter 3985 \mathbf{F}_{{\bf a}D_{\bf a}Q_{\bf b}} =& - \Bigl[
1105 gezelter 3906 \text{Tr}\mathbf{Q}_{\mathbf{b}} \mathbf{ D}_{\mathbf{a}}
1106     +2 \mathbf{D}_{\mathbf{a}} \cdot
1107     \mathbf{Q}_{\mathbf{b}} \Bigr] w_g(r)
1108 gezelter 3985 - \Bigl[
1109 gezelter 3906 \text{Tr}\mathbf{Q}_{\mathbf{b}}
1110     \left( \hat{r} \cdot \mathbf{D}_{\mathbf{a}} \right)
1111     +2 ( \mathbf{D}_{\mathbf{a}} \cdot
1112     \mathbf{Q}_{\mathbf{b}} \cdot \hat{r}) \Bigr] w_h(r) \hat{r} \\
1113     % 3
1114 gezelter 3985 & - \Bigl[\mathbf{ D}_{\mathbf{a}} (\hat{r} \cdot \mathbf{Q}_{{\mathbf b}} \cdot \hat{r})
1115 gezelter 3906 +2 (\hat{r} \cdot \mathbf{D}_{\mathbf{a}} ) (\hat{r} \cdot \mathbf{Q}_{{\mathbf b}} ) \Bigr]
1116     w_i(r)
1117     % 4
1118 gezelter 3985 -
1119 gezelter 3906 (\hat{r} \cdot \mathbf{D}_{\mathbf{a}} )
1120 gezelter 3985 (\hat{r} \cdot \mathbf{Q}_{{\mathbf b}} \cdot \hat{r}) w_j(r) \hat{r} \end{split} \\
1121 gezelter 3906 %
1122     %
1123 gezelter 3985 % \begin{equation}
1124     % \mathbf{F}_{{\bf a}Q_{\bf a}C_{\bf b}} =
1125     % \frac{C_{\bf b }}{4\pi \epsilon_0} \Bigr[
1126     % \text{Tr}\mathbf{Q}_{\bf a} w_d(r) \hat{r}
1127     % + 2 \mathbf{Q}_{{\mathbf a}} \cdot \hat{r} w_e(r)
1128     % + \left( \hat{r} \cdot \mathbf{Q}_{{\mathbf a}} \cdot \hat{r} \right) w_f(r) \hat{r} \Bigr]
1129     % \end{equation}
1130     % %
1131     % \begin{equation}
1132     % \begin{split}
1133     % \mathbf{F}_{{\bf a}Q_{\bf a}D_{\bf b}} =
1134     % &\frac{1}{4\pi \epsilon_0} \Bigl[
1135     % \text{Tr}\mathbf{Q}_{\mathbf{a}} \mathbf{D}_{\mathbf{b}}
1136     % +2 \mathbf{D}_{\mathbf{b}} \cdot \mathbf{Q}_{\mathbf{a}} \Bigr] w_g(r)
1137     % % 2
1138     % + \frac{1}{4\pi \epsilon_0} \Bigl[ \text{Tr}\mathbf{Q}_{\mathbf{a}}
1139     % (\hat{r} \cdot \mathbf{D}_{\mathbf{b}})
1140     % +2 (\mathbf{D}_{\mathbf{b}} \cdot
1141     % \mathbf{Q}_{\mathbf{a}} \cdot \hat{r}) \Bigr] w_h(r) \hat{r} \\
1142     % % 3
1143     % &+ \frac{1}{4\pi \epsilon_0} \Bigl[ \mathbf{D}_{\mathbf{b}}
1144     % (\hat{r} \cdot \mathbf{Q}_{{\mathbf a}} \cdot \hat{r})
1145     % +2 (\hat{r} \cdot \mathbf{D}_{\mathbf{b}})
1146     % (\hat{r} \cdot \mathbf{Q}_{{\mathbf a}} ) \Bigr] w_i(r)
1147     % % 4
1148     % +\frac{1}{4\pi \epsilon_0}
1149     % (\hat{r} \cdot \mathbf{D}_{\mathbf{b}})
1150     % (\hat{r} \cdot \mathbf{Q}_{{\mathbf a}} \cdot \hat{r}) w_j(r) \hat{r}
1151     % \end{split}
1152     % \end{equation}
1153 gezelter 3906 %
1154     %
1155     %
1156     \begin{split}
1157 gezelter 3985 \mathbf{F}_{{\bf a}Q_{\bf a}Q_{\bf b}} =&
1158     \Bigl[
1159 gezelter 3989 \text{Tr}\mathbf{Q}_{\mathbf{a}} \text{Tr}\mathbf{Q}_{\mathbf{b}}
1160     + 2 \mathbf{Q}_{\mathbf{a}} : \mathbf{Q}_{\mathbf{b}} \Bigr] w_k(r) \hat{r} \\
1161 gezelter 3906 % 2
1162 gezelter 3985 &+ \Bigl[
1163 gezelter 3906 2\text{Tr}\mathbf{Q}_{\mathbf{b}} (\hat{r} \cdot \mathbf{Q}_{\mathbf{a}} )
1164     + 2\text{Tr}\mathbf{Q}_{\mathbf{a}} (\hat{r} \cdot \mathbf{Q}_{\mathbf{b}} )
1165     % 3
1166     +4 (\mathbf{Q}_{\mathbf{a}} \cdot \mathbf{Q}_{\mathbf{b}} \cdot \hat{r})
1167     + 4(\hat{r} \cdot \mathbf{Q}_{\mathbf{a}} \cdot \mathbf{Q}_{\mathbf{b}}) \Bigr] w_n(r) \\
1168     % 4
1169 gezelter 3985 &+ \Bigl[
1170 gezelter 3906 \text{Tr}\mathbf{Q}_{\mathbf{a}} (\hat{r} \cdot \mathbf{Q}_{\mathbf{b}} \cdot \hat{r})
1171     + \text{Tr}\mathbf{Q}_{\mathbf{b}}
1172     (\hat{r} \cdot \mathbf{Q}_{\mathbf{a}} \cdot \hat{r})
1173     % 5
1174     +4 (\hat{r} \cdot \mathbf{Q}_{\mathbf{a}} \cdot
1175     \mathbf{Q}_{\mathbf{b}} \cdot \hat{r}) \Bigr] w_l(r) \hat{r} \\
1176     %
1177 gezelter 3985 &+ \Bigl[
1178 gezelter 3906 + 2 (\hat{r} \cdot \mathbf{Q}_{\mathbf{a}} )
1179     (\hat{r} \cdot \mathbf{Q}_{\mathbf{b}} \cdot \hat{r})
1180     %6
1181     +2 (\hat{r} \cdot \mathbf{Q}_{\mathbf{a}} \cdot \hat{r})
1182     (\hat{r} \cdot \mathbf{Q}_{\mathbf{b}} ) \Bigr] w_o(r) \\
1183     % 7
1184 gezelter 3985 &+
1185 gezelter 3906 (\hat{r} \cdot \mathbf{Q}_{\mathbf{a}} \cdot \hat{r})
1186 gezelter 3985 (\hat{r} \cdot \mathbf{Q}_{\mathbf{b}} \cdot \hat{r}) w_m(r) \hat{r} \end{split}
1187     \end{align}
1188     Note that the forces for higher multipoles on site $\mathbf{a}$
1189     interacting with those of lower order on site $\mathbf{b}$ can be
1190     obtained by swapping indices in the expressions above.
1191    
1192 gezelter 3906 %
1193 gezelter 3985 % Torques SECTION -----------------------------------------------------------------------------------------
1194 gezelter 3906 %
1195     \subsection{Torques}
1196 gezelter 3989
1197 gezelter 3906 %
1198 gezelter 3985 The torques for both the Taylor-Shifted as well as Gradient-Shifted
1199     methods are given in space-frame coordinates:
1200 gezelter 3906 %
1201     %
1202 gezelter 3985 \begin{align}
1203     \mathbf{\tau}_{{\bf b}C_{\bf a}D_{\bf b}} =&
1204     C_{\bf a} (\hat{r} \times \mathbf{D}_{\mathbf{b}}) v_{11}(r) \\
1205 gezelter 3906 %
1206     %
1207     %
1208 gezelter 3985 \mathbf{\tau}_{{\bf b}C_{\bf a}Q_{\bf b}} =&
1209     2C_{\bf a}
1210     \hat{r} \times ( \mathbf{Q}_{{\mathbf b}} \cdot \hat{r}) v_{22}(r) \\
1211 gezelter 3906 %
1212     %
1213     %
1214 gezelter 3985 % \begin{equation}
1215     % \mathbf{\tau}_{{\bf a}D_{\bf a}C_{\bf b}} =
1216     % -\frac{C_{\bf b}}{4\pi \epsilon_0}
1217     % (\hat{r} \times \mathbf{D}_{\mathbf{a}}) v_{11}(r)
1218     % \end{equation}
1219 gezelter 3906 %
1220     %
1221     %
1222 gezelter 3985 \mathbf{\tau}_{{\bf a}D_{\bf a}D_{\bf b}} =&
1223     \mathbf{D}_{\mathbf {a}} \times \mathbf{D}_{\mathbf{b}} v_{21}(r)
1224 gezelter 3906 % 2
1225 gezelter 3985 -
1226 gezelter 3906 (\hat{r} \times \mathbf{D}_{\mathbf {a}} )
1227 gezelter 3985 (\hat{r} \cdot \mathbf{D}_{\mathbf {b}} ) v_{22}(r)\\
1228 gezelter 3906 %
1229     %
1230     %
1231 gezelter 3985 % \begin{equation}
1232     % \mathbf{\tau}_{{\bf b}D_{\bf a}D_{\bf b}} =
1233     % -\frac{1}{4\pi \epsilon_0} \mathbf{D}_{\mathbf {a}} \times \mathbf{D}_{\mathbf{b}} v_{21}(r)
1234     % % 2
1235     % +\frac{1}{4\pi \epsilon_0}
1236     % (\hat{r} \cdot \mathbf{D}_{\mathbf {a}} )
1237     % (\hat{r} \times \mathbf{D}_{\mathbf {b}} ) v_{22}(r)
1238     % \end{equation}
1239 gezelter 3906 %
1240     %
1241     %
1242 gezelter 3985 \mathbf{\tau}_{{\bf a}D_{\bf a}Q_{\bf b}} =&
1243     \Bigl[
1244 gezelter 3906 -\text{Tr}\mathbf{Q}_{\mathbf{b}}
1245     (\hat{r} \times \mathbf{D}_{\mathbf{a}} )
1246     +2 \mathbf{D}_{\mathbf{a}} \times
1247     (\mathbf{Q}_{\mathbf{b}} \cdot \hat{r})
1248     \Bigr] v_{31}(r)
1249     % 3
1250 gezelter 3985 - (\hat{r} \times \mathbf{D}_{\mathbf{a}} )
1251     (\hat{r} \cdot \mathbf{Q}_{{\mathbf b}} \cdot \hat{r}) v_{32}(r)\\
1252 gezelter 3906 %
1253     %
1254     %
1255 gezelter 3985 \mathbf{\tau}_{{\bf b}D_{\bf a}Q_{\bf b}} =&
1256     \Bigl[
1257 gezelter 3906 +2 ( \mathbf{D}_{\mathbf{a}} \cdot \mathbf{Q}_{\mathbf{b}} ) \times
1258     \hat{r}
1259     -2 \mathbf{D}_{\mathbf{a}} \times
1260     (\mathbf{Q}_{\mathbf{b}} \cdot \hat{r})
1261     \Bigr] v_{31}(r)
1262     % 2
1263 gezelter 3985 +
1264 gezelter 3906 (\hat{r} \cdot \mathbf{D}_{\mathbf{a}})
1265 gezelter 3985 (\hat{r} \cdot \mathbf{Q}_{\mathbf{b}}) \times \hat{r} v_{32}(r)\\
1266 gezelter 3906 %
1267     %
1268     %
1269 gezelter 3985 % \begin{equation}
1270     % \mathbf{\tau}_{{\bf a}Q_{\bf a}D_{\bf b}} =
1271     % \frac{1}{4\pi \epsilon_0} \Bigl[
1272     % -2 (\mathbf{D}_{\mathbf{b}} \cdot \mathbf{Q}_{\mathbf{a}} ) \times \hat{r}
1273     % +2 \mathbf{D}_{\mathbf{b}} \times
1274     % (\mathbf{Q}_{\mathbf{a}} \cdot \hat{r})
1275     % \Bigr] v_{31}(r)
1276     % % 3
1277     % - \frac{2}{4\pi \epsilon_0}
1278     % (\hat{r} \cdot \mathbf{D}_{\mathbf{b}} )
1279     % (\hat{r} \cdot \mathbf
1280     % {Q}_{{\mathbf a}}) \times \hat{r} v_{32}(r)
1281     % \end{equation}
1282 gezelter 3906 %
1283     %
1284     %
1285 gezelter 3985 % \begin{equation}
1286     % \mathbf{\tau}_{{\bf b}Q_{\bf a}D_{\bf b}} =
1287     % \frac{1}{4\pi \epsilon_0} \Bigl[
1288     % \text{Tr}\mathbf{Q}_{\mathbf{a}}
1289     % (\hat{r} \times \mathbf{D}_{\mathbf{b}} )
1290     % +2 \mathbf{D}_{\mathbf{b}} \times
1291     % ( \mathbf{Q}_{\mathbf{a}} \cdot \hat{r}) \Bigr] v_{31}(r)
1292     % % 2
1293     % +\frac{1}{4\pi \epsilon_0}
1294     % (\hat{r} \times \mathbf{D}_{\mathbf{b}} )
1295     % (\hat{r} \cdot \mathbf{Q}_{{\mathbf a}} \cdot \hat{r}) v_{32}(r)
1296     % \end{equation}
1297 gezelter 3906 %
1298     %
1299     %
1300     \begin{split}
1301 gezelter 3985 \mathbf{\tau}_{{\bf a}Q_{\bf a}Q_{\bf b}} =&
1302     -4
1303 gezelter 3906 \mathbf{Q}_{{\mathbf a}} \times \mathbf{Q}_{{\mathbf b}}
1304     v_{41}(r) \\
1305     % 2
1306 gezelter 3985 &+
1307 gezelter 3906 \Bigl[-2\text{Tr}\mathbf{Q}_{\mathbf{b}}
1308     (\hat{r} \cdot \mathbf{Q}_{{\mathbf a}} ) \times \hat{r}
1309     +4 \hat{r} \times
1310     ( \mathbf{Q}_{{\mathbf a}} \cdot \mathbf{Q}_{{\mathbf b}} \cdot \hat{r})
1311     % 3
1312     -4 (\hat{r} \cdot \mathbf{Q}_{{\mathbf a}} )\times
1313     ( \mathbf{Q}_{{\mathbf b}} \cdot \hat{r} ) \Bigr] v_{42}(r) \\
1314     % 4
1315 gezelter 3985 &+ 2
1316 gezelter 3906 \hat{r} \times ( \mathbf{Q}_{{\mathbf a}} \cdot \hat{r})
1317 gezelter 3985 (\hat{r} \cdot \mathbf{Q}_{{\mathbf b}} \cdot \hat{r}) v_{43}(r) \end{split}\\
1318 gezelter 3906 %
1319     %
1320     %
1321     \begin{split}
1322     \mathbf{\tau}_{{\bf b}Q_{\bf a}Q_{\bf b}} =
1323 gezelter 3985 &4
1324 gezelter 3906 \mathbf{Q}_{{\mathbf a}} \times \mathbf{Q}_{{\mathbf b}} v_{41}(r) \\
1325     % 2
1326 gezelter 3985 &+ \Bigl[- 2\text{Tr}\mathbf{Q}_{\mathbf{a}}
1327 gezelter 3906 (\hat{r} \cdot \mathbf{Q}_{{\mathbf b}} ) \times \hat{r}
1328     -4 (\hat{r} \cdot \mathbf{Q}_{{\mathbf a}} \cdot
1329     \mathbf{Q}_{{\mathbf b}} ) \times
1330     \hat{r}
1331     +4 ( \hat{r} \cdot \mathbf{Q}_{{\mathbf a}} ) \times
1332     ( \mathbf{Q}_{{\mathbf b}} \cdot \hat{r})
1333     \Bigr] v_{42}(r) \\
1334     % 4
1335 gezelter 3985 &+2
1336 gezelter 3906 (\hat{r} \cdot \mathbf{Q}_{{\mathbf a}} \cdot \hat{r})
1337 gezelter 3985 \hat{r} \times ( \mathbf{Q}_{{\mathbf b}} \cdot \hat{r}) v_{43}(r)\end{split}
1338     \end{align}
1339     %
1340     Here, we have defined the matrix cross product in an identical form
1341     as in Ref. \onlinecite{Smith98}:
1342     \begin{equation}
1343     \left[\mathbf{A} \times \mathbf{B}\right]_\alpha = \sum_\beta
1344     \left[\mathbf{A}_{\alpha+1,\beta} \mathbf{B}_{\alpha+2,\beta}
1345     -\mathbf{A}_{\alpha+2,\beta} \mathbf{B}_{\alpha+2,\beta}
1346     \right]
1347 gezelter 3906 \end{equation}
1348 gezelter 3985 where $\alpha+1$ and $\alpha+2$ are regarded as cyclic
1349     permuations of the matrix indices.
1350 gezelter 3980
1351 gezelter 3985 All of the radial functions required for torques are identical with
1352     the radial functions previously computed for the interaction energies.
1353     These are tabulated for both shifted force methods in table
1354     \ref{tab:tableenergy}. The torques for higher multipoles on site
1355     $\mathbf{a}$ interacting with those of lower order on site
1356     $\mathbf{b}$ can be obtained by swapping indices in the expressions
1357     above.
1358    
1359 gezelter 3990 \section{Related real-space methods}
1360 gezelter 4098 One can also formulate an extension of the Wolf approach for point
1361     multipoles by simply projecting the image multipole onto the surface
1362     of the cutoff sphere, and including the interactions with the central
1363     multipole and the image. This effectively shifts the total potential
1364     to zero at the cutoff radius,
1365 gezelter 3990 \begin{equation}
1366 gezelter 4098 U^{\text{SP}} = \sum \left[ U(\mathbf{r}, \hat{\mathbf{a}}, \hat{\mathbf{b}}) -
1367     U(\mathbf{r}_c,\hat{\mathbf{a}}, \hat{\mathbf{b}}) \right]
1368 gezelter 3990 \label{eq:SP}
1369     \end{equation}
1370 gezelter 4098 where the sum describes separate potential shifting that is applied to
1371     each orientational contribution to the energy.
1372    
1373     The energies and torques for the shifted potential (SP) can be easily
1374     obtained by zeroing out the $(r-r_c)$ terms in the final column of
1375     table \ref{tab:tableenergy}. Forces for the SP method retain
1376     discontinuities at the cutoff sphere, and can be obtained by
1377 gezelter 3990 eliminating all functions that depend on $r_c$ in the last column of
1378 gezelter 4098 table \ref{tab:tableFORCE}. The self-energy contributions for the SP
1379 gezelter 3990 potential are identical to both the GSF and TSF methods.
1380    
1381 gezelter 3980 \section{Comparison to known multipolar energies}
1382    
1383     To understand how these new real-space multipole methods behave in
1384     computer simulations, it is vital to test against established methods
1385     for computing electrostatic interactions in periodic systems, and to
1386     evaluate the size and sources of any errors that arise from the
1387 gezelter 3990 real-space cutoffs. In this paper we test both TSF and GSF
1388     electrostatics against analytical methods for computing the energies
1389     of ordered multipolar arrays. In the following paper, we test the new
1390     methods against the multipolar Ewald sum for computing the energies,
1391     forces and torques for a wide range of typical condensed-phase
1392     (disordered) systems.
1393 gezelter 3980
1394     Because long-range electrostatic effects can be significant in
1395     crystalline materials, ordered multipolar arrays present one of the
1396     biggest challenges for real-space cutoff methods. The dipolar
1397     analogues to the Madelung constants were first worked out by Sauer,
1398     who computed the energies of ordered dipole arrays of zero
1399     magnetization and obtained a number of these constants.\cite{Sauer}
1400     This theory was developed more completely by Luttinger and
1401 gezelter 3986 Tisza\cite{LT,LT2} who tabulated energy constants for the Sauer arrays
1402 gezelter 3990 and other periodic structures.
1403 gezelter 3986
1404 gezelter 3990 To test the new electrostatic methods, we have constructed very large,
1405     $N=$ 16,000~(bcc) arrays of dipoles in the orientations described in
1406     Ref. \onlinecite{LT}. These structures include ``A'' lattices with
1407     nearest neighbor chains of antiparallel dipoles, as well as ``B''
1408     lattices with nearest neighbor strings of antiparallel dipoles if the
1409     dipoles are contained in a plane perpendicular to the dipole direction
1410     that passes through the dipole. We have also studied the minimum
1411 gezelter 3980 energy structure for the BCC lattice that was found by Luttinger \&
1412 gezelter 3986 Tisza. The total electrostatic energy for any of the arrays is given
1413     by:
1414 gezelter 3980 \begin{equation}
1415     E = C N^2 \mu^2
1416     \end{equation}
1417 gezelter 3990 where $C$ is the energy constant (equivalent to the Madelung
1418     constant), $N$ is the number of dipoles per unit volume, and $\mu$ is
1419     the strength of the dipole. Energy constants (converged to 1 part in
1420     $10^9$) are given in the supplemental information.
1421 gezelter 3980
1422 gezelter 4172 \begin{figure}
1423     \includegraphics[width=\linewidth]{Dipoles_rCutNew.pdf}
1424     \caption{Convergence of the lattice energy constants as a function of
1425     cutoff radius (normalized by the lattice constant, $a$) for the new
1426     real-space methods. Three dipolar crystal structures were sampled,
1427     and the analytic energy constants for the three lattices are
1428     indicated with grey dashed lines. The left panel shows results for
1429     the undamped kernel ($1/r$), while the damped error function kernel,
1430     $B_0(r)$ was used in the right panel.}
1431     \label{fig:Dipoles_rCut}
1432     \end{figure}
1433 mlamichh 4163
1434 gezelter 4172 \begin{figure}
1435     \includegraphics[width=\linewidth]{Dipoles_alphaNew.pdf}
1436     \caption{Convergence to the lattice energy constants as a function of
1437     the reduced damping parameter ($\alpha^* = \alpha a$) for the
1438     different real-space methods in the same three dipolar crystals in
1439     Figure \ref{fig:Dipoles_rCut}. The left panel shows results for a
1440     relatively small cutoff radius ($r_c = 4.5 a$) while a larger cutoff
1441     radius ($r_c = 6 a$) was used in the right panel. }
1442     \label{fig:Dipoles_alpha}
1443 mlamichh 4163 \end{figure}
1444 gezelter 4172
1445 gezelter 3990 For the purposes of testing the energy expressions and the
1446     self-neutralization schemes, the primary quantity of interest is the
1447     analytic energy constant for the perfect arrays. Convergence to these
1448     constants are shown as a function of both the cutoff radius, $r_c$,
1449 gezelter 4174 and the damping parameter, $\alpha$ in Figs.\ref{fig:Dipoles_rCut}
1450 gezelter 4172 and \ref{fig:Dipoles_alpha}. We have simultaneously tested a hard
1451     cutoff (where the kernel is simply truncated at the cutoff radius), as
1452     well as a shifted potential (SP) form which includes a
1453 gezelter 3990 potential-shifting and self-interaction term, but does not shift the
1454     forces and torques smoothly at the cutoff radius. The SP method is
1455     essentially an extension of the original Wolf method for multipoles.
1456 gezelter 3986
1457 gezelter 4174 The hard cutoff exhibits oscillations around the analytic energy
1458 gezelter 3988 constants, and converges to incorrect energies when the complementary
1459     error function damping kernel is used. The shifted potential (SP) and
1460     gradient-shifted force (GSF) approximations converge to the correct
1461 gezelter 4174 energy smoothly by $r_c = 6 a$ even for the undamped case. This
1462 gezelter 3988 indicates that the correction provided by the self term is required
1463     for obtaining accurate energies. The Taylor-shifted force (TSF)
1464     approximation appears to perturb the potential too much inside the
1465     cutoff region to provide accurate measures of the energy constants.
1466 gezelter 4174
1467 gezelter 3986 {\it Quadrupolar} analogues to the Madelung constants were first
1468     worked out by Nagai and Nakamura who computed the energies of selected
1469     quadrupole arrays based on extensions to the Luttinger and Tisza
1470 gezelter 4174 approach.\cite{Nagai01081960,Nagai01091963}
1471 gezelter 3980
1472     In analogy to the dipolar arrays, the total electrostatic energy for
1473     the quadrupolar arrays is:
1474     \begin{equation}
1475 gezelter 3996 E = C N \frac{3\bar{Q}^2}{4a^5}
1476 gezelter 3980 \end{equation}
1477 gezelter 3996 where $a$ is the lattice parameter, and $\bar{Q}$ is the effective
1478     quadrupole moment,
1479     \begin{equation}
1480 gezelter 4098 \bar{Q}^2 = 2 \left(3 Q : Q - (\text{Tr} Q)^2 \right)
1481 gezelter 3996 \end{equation}
1482     for the primitive quadrupole as defined in Eq. \ref{eq:quadrupole}.
1483     (For the traceless quadrupole tensor, $\Theta = 3 Q - \text{Tr} Q$,
1484 gezelter 4098 the effective moment, $\bar{Q}^2 = \frac{2}{3} \Theta : \Theta$.)
1485 gezelter 3980
1486 gezelter 4174 To test the new electrostatic methods for quadrupoles, we have
1487     constructed very large, $N=$ 8,000~(sc), 16,000~(bcc), and
1488     32,000~(fcc) arrays of linear quadrupoles in the orientations
1489     described in Ref. \onlinecite{Nagai01081960}. We have compared the
1490     energy constants for the lowest energy configurations for these linear
1491     quadrupoles. Convergence to these constants are shown as a function
1492     of both the cutoff radius, $r_c$, and the damping parameter, $\alpha$
1493     in Figs.~\ref{fig:Quadrupoles_rCut} and \ref{fig:Quadrupoles_alpha}.
1494    
1495 gezelter 4172 \begin{figure}
1496 gezelter 4173 \includegraphics[width=\linewidth]{Quadrupoles_rcutConvergence.pdf}
1497     \caption{Convergence of the lattice energy constants as a function of
1498     cutoff radius (normalized by the lattice constant, $a$) for the new
1499     real-space methods. Three quadrupolar crystal structures were
1500     sampled, and the analytic energy constants for the three lattices
1501     are indicated with grey dashed lines. The left panel shows results
1502     for the undamped kernel ($1/r$), while the damped error function
1503     kernel, $B_0(r)$ was used in the right panel.}
1504 gezelter 4174 \label{fig:Quadrupoles_rCut}
1505 gezelter 4172 \end{figure}
1506    
1507    
1508     \begin{figure}[!htbp]
1509 gezelter 4174 \includegraphics[width=\linewidth]{Quadrupoles_newAlpha.pdf}
1510     \caption{Convergence to the lattice energy constants as a function of
1511     the reduced damping parameter ($\alpha^* = \alpha a$) for the
1512     different real-space methods in the same three quadrupolar crystals
1513     in Figure \ref{fig:Quadrupoles_rCut}. The left panel shows
1514     results for a relatively small cutoff radius ($r_c = 4.5 a$) while a
1515     larger cutoff radius ($r_c = 6 a$) was used in the right panel. }
1516     \label{fig:Quadrupoles_alpha}
1517 gezelter 4172 \end{figure}
1518    
1519 gezelter 4174 Again, we find that the hard cutoff exhibits oscillations around the
1520     analytic energy constants. The shifted potential (SP) and
1521     gradient-shifted force (GSF) approximations converge to the correct
1522     energy smoothly by $r_c = 4 a$ even for the undamped case. The
1523     Taylor-shifted force (TSF) approximation again appears to perturb the
1524     potential too much inside the cutoff region to provide accurate
1525     measures of the energy constants.
1526 gezelter 4172
1527 gezelter 4174
1528 gezelter 3985 \section{Conclusion}
1529 gezelter 4174 We have presented three efficient real-space methods for computing the
1530 gezelter 3985 interactions between point multipoles. These methods have the benefit
1531     of smoothly truncating the energies, forces, and torques at the cutoff
1532     radius, making them attractive for both molecular dynamics (MD) and
1533     Monte Carlo (MC) simulations. We find that the Gradient-Shifted Force
1534     (GSF) and the Shifted-Potential (SP) methods converge rapidly to the
1535     correct lattice energies for ordered dipolar and quadrupolar arrays,
1536     while the Taylor-Shifted Force (TSF) is too severe an approximation to
1537 gezelter 4174 provide accurate convergence to lattice energies.
1538 gezelter 3980
1539 gezelter 3985 In most cases, GSF can obtain nearly quantitative agreement with the
1540     lattice energy constants with reasonably small cutoff radii. The only
1541     exception we have observed is for crystals which exhibit a bulk
1542     macroscopic dipole moment (e.g. Luttinger \& Tisza's $Z_1$ lattice).
1543     In this particular case, the multipole neutralization scheme can
1544     interfere with the correct computation of the energies. We note that
1545     the energies for these arrangements are typically much larger than for
1546     crystals with net-zero moments, so this is not expected to be an issue
1547     in most simulations.
1548 gezelter 3980
1549 gezelter 3985 In large systems, these new methods can be made to scale approximately
1550     linearly with system size, and detailed comparisons with the Ewald sum
1551     for a wide range of chemical environments follows in the second paper.
1552 gezelter 3980
1553 gezelter 3906 \begin{acknowledgments}
1554 gezelter 3985 JDG acknowledges helpful discussions with Christopher
1555     Fennell. Support for this project was provided by the National
1556 gezelter 4172 Science Foundation under grant CHE-1362211. Computational time was
1557 gezelter 3985 provided by the Center for Research Computing (CRC) at the
1558     University of Notre Dame.
1559 gezelter 3906 \end{acknowledgments}
1560    
1561 gezelter 3984 \newpage
1562 gezelter 3906 \appendix
1563    
1564 gezelter 3984 \section{Smith's $B_l(r)$ functions for damped-charge distributions}
1565 gezelter 3985 \label{SmithFunc}
1566 gezelter 3984 The following summarizes Smith's $B_l(r)$ functions and includes
1567     formulas given in his appendix.\cite{Smith98} The first function
1568     $B_0(r)$ is defined by
1569 gezelter 3906 %
1570     \begin{equation}
1571     B_0(r)=\frac{\text{erfc}(\alpha r)}{r} = \frac{2}{\sqrt{\pi}r}=
1572     \int_{\alpha r}^{\infty} \text{e}^{-s^2} ds .
1573     \end{equation}
1574     %
1575     The first derivative of this function is
1576     %
1577     \begin{equation}
1578     \frac{dB_0(r)}{dr}=-\frac{1}{r^2}\text{erfc}(\alpha r)
1579     -\frac{2\alpha}{r\sqrt{\pi}}\text{e}^{-{\alpha}^2r^2}
1580     \end{equation}
1581     %
1582 gezelter 3984 which can be used to define a function $B_1(r)$:
1583 gezelter 3906 %
1584     \begin{equation}
1585     B_1(r)=-\frac{1}{r}\frac{dB_0(r)}{dr}
1586     \end{equation}
1587     %
1588 gezelter 3984 In general, the recurrence relation,
1589 gezelter 3906 \begin{equation}
1590     B_l(r)=-\frac{1}{r}\frac{dB_{l-1}(r)}{dr}
1591     = \frac{1}{r^2} \left[ (2l-1)B_{l-1}(r) + \frac {(2\alpha^2)^l}{\alpha \sqrt{\pi}}
1592     \text{e}^{-{\alpha}^2r^2}
1593 gezelter 3984 \right] ,
1594 gezelter 3906 \end{equation}
1595 gezelter 3984 is very useful for building up higher derivatives. Using these
1596     formulas, we find:
1597 gezelter 3906 %
1598 gezelter 3984 \begin{align}
1599     \frac{dB_0}{dr}=&-rB_1(r) \\
1600     \frac{d^2B_0}{dr^2}=& - B_1(r) + r^2 B_2(r) \\
1601     \frac{d^3B_0}{dr^3}=& 3 r B_2(r) - r^3 B_3(r) \\
1602     \frac{d^4B_0}{dr^4}=& 3 B_2(r) - 6 r^2 B_3(r) + r^4 B_4(r) \\
1603     \frac{d^5B_0}{dr^5}=& - 15 r B_3(r) + 10 r^3 B_4(r) - r^5 B_5(r) .
1604     \end{align}
1605 gezelter 3906 %
1606 gezelter 3984 As noted by Smith, it is possible to approximate the $B_l(r)$
1607     functions,
1608 gezelter 3906 %
1609     \begin{equation}
1610     B_l(r)=\frac{(2l)!}{l!2^lr^{2l+1}} - \frac {(2\alpha^2)^{l+1}}{(2l+1)\alpha \sqrt{\pi}}
1611     +\text{O}(r) .
1612     \end{equation}
1613 gezelter 3984 \newpage
1614     \section{The $r$-dependent factors for TSF electrostatics}
1615 gezelter 3906
1616     Using the shifted damped functions $f_n(r)$ defined by:
1617     %
1618     \begin{equation}
1619 gezelter 3984 f_n(r)= B_0(r) -\sum_{m=0}^{n+1} \frac {(r-r_c)^m}{m!} B_0^{(m)}(r_c) ,
1620 gezelter 3906 \end{equation}
1621     %
1622 gezelter 3984 where the superscript $(m)$ denotes the $m^\mathrm{th}$ derivative. In
1623     this Appendix, we provide formulas for successive derivatives of this
1624     function. (If there is no damping, then $B_0(r)$ is replaced by
1625     $1/r$.) First, we find:
1626 gezelter 3906 %
1627     \begin{equation}
1628     \frac{\partial f_n}{\partial r_\alpha}=\hat{r}_\alpha \frac{d f_n}{d r} .
1629     \end{equation}
1630     %
1631 gezelter 3984 This formula clearly brings in derivatives of Smith's $B_0(r)$
1632     function, and we define higher-order derivatives as follows:
1633 gezelter 3906 %
1634 gezelter 3984 \begin{align}
1635     g_n(r)=& \frac{d f_n}{d r} =
1636     B_0^{(1)}(r) -\sum_{m=0}^{n} \frac {(r-r_c)^m}{m!} B_0^{(m+1)}(r_c) \\
1637     h_n(r)=& \frac{d^2f_n}{d r^2} =
1638     B_0^{(2)}(r) -\sum_{m=0}^{n-1} \frac {(r-r_c)^m}{m!} B_0^{(m+2)}(r_c) \\
1639     s_n(r)=& \frac{d^3f_n}{d r^3} =
1640     B_0^{(3)}(r) -\sum_{m=0}^{n-2} \frac {(r-r_c)^m}{m!} B_0^{(m+3)}(r_c) \\
1641     t_n(r)=& \frac{d^4f_n}{d r^4} =
1642     B_0^{(4)}(r) -\sum_{m=0}^{n-3} \frac {(r-r_c)^m}{m!} B_0^{(m+4)}(r_c) \\
1643     u_n(r)=& \frac{d^5f_n}{d r^5} =
1644     B_0^{(5)}(r) -\sum_{m=0}^{n-4} \frac {(r-r_c)^m}{m!} B_0^{(m+5)}(r_c) .
1645     \end{align}
1646 gezelter 3906 %
1647 gezelter 3984 We note that the last function needed (for quadrupole-quadrupole interactions) is
1648 gezelter 3906 %
1649     \begin{equation}
1650 gezelter 3984 u_4(r)=B_0^{(5)}(r) - B_0^{(5)}(r_c) .
1651 gezelter 3906 \end{equation}
1652 gezelter 3989 % The functions
1653     % needed are listed schematically below:
1654     % %
1655     % \begin{eqnarray}
1656     % f_0 \quad f_1 \qquad \qquad \quad & \nonumber \\
1657     % g_0 \quad g_1 \quad g_2 \quad g_3 \quad &g_4 \nonumber \\
1658     % h_1 \quad h_2 \quad h_3 \quad &h_4 \nonumber \\
1659     % s_2 \quad s_3 \quad &s_4 \nonumber \\
1660     % t_3 \quad &t_4 \nonumber \\
1661     % &u_4 \nonumber .
1662     % \end{eqnarray}
1663 gezelter 3984 The functions $f_n(r)$ to $u_n(r)$ can be computed recursively and
1664 gezelter 3989 stored on a grid for values of $r$ from $0$ to $r_c$. Using these
1665     functions, we find
1666 gezelter 3906 %
1667 gezelter 3984 \begin{align}
1668     \frac{\partial f_n}{\partial r_\alpha} =&r_\alpha \frac {g_n}{r} \label{eq:b9}\\
1669     \frac{\partial^2 f_n}{\partial r_\alpha \partial r_\beta} =&\delta_{\alpha \beta}\frac {g_n}{r}
1670     +r_\alpha r_\beta \left( -\frac{g_n}{r^3} +\frac{h_n}{r^2}\right) \\
1671 gezelter 3989 \frac{\partial^3 f_n}{\partial r_\alpha \partial r_\beta \partial r_\gamma} =&
1672 gezelter 3906 \left( \delta_{\alpha \beta} r_\gamma + \delta_{\alpha \gamma} r_\beta +
1673     \delta_{ \beta \gamma} r_\alpha \right)
1674 gezelter 3989 \left( -\frac{g_n}{r^3} +\frac{h_n}{r^2} \right) \nonumber \\
1675     & + r_\alpha r_\beta r_\gamma
1676 gezelter 3984 \left( \frac{3g_n}{r^5}-\frac{3h_n}{r^4} +\frac{s_n}{r^3} \right) \\
1677 gezelter 3989 \frac{\partial^4 f_n}{\partial r_\alpha \partial r_\beta \partial
1678     r_\gamma \partial r_\delta} =&
1679 gezelter 3906 \left( \delta_{\alpha \beta} \delta_{\gamma \delta}
1680     + \delta_{\alpha \gamma} \delta_{\beta \delta}
1681     +\delta_{ \beta \gamma} \delta_{\alpha \delta} \right)
1682     \left( - \frac{g_n}{r^3} + \frac{h_n}{r^2} \right) \nonumber \\
1683 gezelter 3984 &+ \left( \delta_{\alpha \beta} r_\gamma r_\delta
1684     + \text{5 permutations}
1685 gezelter 3906 \right) \left( \frac{3 g_n}{r^5} - \frac{3h_n}{r^4} + \frac{s_n}{r^3}
1686     \right) \nonumber \\
1687 gezelter 3984 &+ r_\alpha r_\beta r_\gamma r_\delta
1688 gezelter 3906 \left( -\frac{15g_n}{r^7} + \frac{15h_n}{r^6} - \frac{6s_n}{r^5}
1689 gezelter 3984 + \frac{t_n}{r^4} \right)\\
1690 gezelter 3906 \frac{\partial^5 f_n}
1691 gezelter 3989 {\partial r_\alpha \partial r_\beta \partial r_\gamma \partial
1692     r_\delta \partial r_\epsilon} =&
1693 gezelter 3906 \left( \delta_{\alpha \beta} \delta_{\gamma \delta} r_\epsilon
1694 gezelter 3984 + \text{14 permutations} \right)
1695 gezelter 3906 \left( \frac{3g_n}{r^5}-\frac{3h_n}{r^4} +\frac{s_n}{r^3} \right) \nonumber \\
1696 gezelter 3984 &+ \left( \delta_{\alpha \beta} r_\gamma r_\delta r_\epsilon
1697     + \text{9 permutations}
1698 gezelter 3906 \right) \left(- \frac{15g_n}{r^7}+\frac{15h_n}{r^7} -\frac{6s_n}{r^5} +\frac{t_n}{r^4}
1699     \right) \nonumber \\
1700 gezelter 3984 &+ r_\alpha r_\beta r_\gamma r_\delta r_\epsilon
1701 gezelter 3906 \left( \frac{105g_n}{r^9} - \frac{105h_n}{r^8} + \frac{45s_n}{r^7}
1702 gezelter 3984 - \frac{10t_n}{r^6} +\frac{u_n}{r^5} \right) \label{eq:b13}
1703     \end{align}
1704 gezelter 3906 %
1705     %
1706     %
1707 gezelter 3984 \newpage
1708     \section{The $r$-dependent factors for GSF electrostatics}
1709 gezelter 3906
1710 gezelter 3984 In Gradient-shifted force electrostatics, the kernel is not expanded,
1711     rather the individual terms in the multipole interaction energies.
1712     For damped charges , this still brings into the algebra multiple
1713     derivatives of the Smith's $B_0(r)$ function. To denote these terms,
1714 gezelter 3989 we generalize the notation of the previous appendix. For either
1715     $f(r)=1/r$ (undamped) or $f(r)=B_0(r)$ (damped),
1716 gezelter 3906 %
1717 gezelter 3984 \begin{align}
1718     g(r)=& \frac{df}{d r}\\
1719     h(r)=& \frac{dg}{d r} = \frac{d^2f}{d r^2} \\
1720     s(r)=& \frac{dh}{d r} = \frac{d^3f}{d r^3} \\
1721     t(r)=& \frac{ds}{d r} = \frac{d^4f}{d r^4} \\
1722     u(r)=& \frac{dt}{d r} = \frac{d^5f}{d r^5} .
1723     \end{align}
1724 gezelter 3906 %
1725 gezelter 3989 For undamped charges Table I lists these derivatives under the column
1726     ``Bare Coulomb.'' Equations \ref{eq:b9} to \ref{eq:b13} are still
1727     correct for GSF electrostatics if the subscript $n$ is eliminated.
1728 gezelter 3906
1729 gezelter 4172 \newpage
1730    
1731 gezelter 3980 \bibliography{multipole}
1732    
1733 gezelter 3906 \end{document}
1734     %
1735     % ****** End of file multipole.tex ******