cond-mat0601470/HF.tex
1: \nonstopmode
2: 
3: \newcommand{\ie}{i.e.{}}
4: \newcommand{\etal}{\textit{et al.}}
5: \newcommand{\U}[1]{\,{\rm{#1}}}
6: \newcommand{\I}[1]{_{\textrm{\scriptsize #1}}}
7: \newcommand{\Sum}{\sum\limits}
8: \newcommand{\angstrom}{\U{\hbox{\AA}}}
9: \newcommand{\degree}{{}^{\circ}}
10: \newcommand{\oHF}[1]{(HF)$_{#1}$}
11: \newcommand{\oHCl}[1]{(HCl)$_{#1}$}
12: \newcommand{\HF}{(HF)$_{\infty}$}
13: \newcommand{\HCl}{(HCl)$_{\infty}$}
14: \newcommand{\DCl}{(DCl)$_{\infty}$}
15: \newcommand{\Cal}[1]{{\cal #1}}
16: \newcommand{\mHartree}{\U{mE_{\mathit h}}}
17: \newcommand{\muHartree}{\U{\mu E_{\mathit h}}}
18: \newcommand{\bra}[1]{\left<\right.\!#1\!\left.\right|}
19: \newcommand{\ket}[1]{\left|\right.\!#1\!\left.\right>}
20: \newcommand{\euler}{\textrm{e}}
21: 
22: 
23: \documentclass[prb,showpacs,nopreprintnumbers,twocolumn]{revtex4}
24: \usepackage{graphicx,dcolumn,subeqnarray}
25: 
26: \begin{document}
27: \title{Hydrogen bonding in infinite hydrogen fluoride
28: and hydrogen chloride chains}
29: \date{October 14, 2006}
30: \author{Christian Buth}
31: \email[Corresponding author. Electronic address: ]{Christian.Buth@web.de}
32: \author{Beate Paulus}
33: \affiliation{Max-Planck-Institut f\"ur Physik komplexer Systeme, N\"othnitzer
34: Stra\ss{}e~38, 01187~Dresden, Germany}
35: 
36: 
37: \begin{abstract}
38: Hydrogen bonding in infinite~\HF{} and \HCl{} bent (zigzag) chains
39: is studied using the \emph{ab initio}
40: coupled-cluster singles and doubles~(CCSD) correlation method.
41: The correlation contribution to the binding energy is
42: decomposed in terms of nonadditive many-body interactions between
43: the monomers in the chains, the so-called energy increments.
44: Van der Waals constants for the two-body dispersion interaction
45: between distant monomers in the infinite chains
46: are extracted from this decomposition.
47: They allow a partitioning of the correlation contribution to
48: the binding energy into short- and long-range terms.
49: This finding affords a significant reduction in the computational
50: effort of \emph{ab initio} calculations for solids as only
51: the short-range part requires a sophisticated treatment
52: whereas the long-range part can be summed immediately
53: to infinite distances.
54: \end{abstract}
55: 
56: 
57: %
58: % The 2006 PACS classification scheme
59: %
60: % 31.15.Ar Ab initio calculations
61: % 31.25.Qm Electron correlation calculations for polyatomic molecules
62: % 71.15.Nc Total energy and cohesive energy calculations
63: % 71.20.Ps Other inorganic compounds
64: %
65: 
66: \pacs{71.15.Nc, 71.20.Ps, 31.15.Ar, 31.25.Qm}
67: \preprint{arXiv:cond-mat/0601470}
68: \maketitle
69: 
70: 
71: \section{Introduction}
72: 
73: Hydrogen fluoride~\cite{Atoji:CS-54,Habuda:NM-71,Johnson:CS-75,%
74: Otto:BE-86,Panas:HF-93,Berski:DHF-98}
75: and hydrogen chloride~\cite{Sandor:CS-67,Sandor:CC-67,Sandor:ND-69}
76: are representatives of molecular crystals; the electronic structure
77: of the constituent HF or HCl~monomers is
78: essentially preserved upon crystallization.
79: The monomers in both crystals are hydrogen bonded;~\cite{Hamilton:HB-68,%
80: Pauling:NC-93,Scheiner:HB-97,Hadzi:TT-97,Karpfen:CE-02}
81: it is a directional and anisotropic bonding
82: of the hydrogen in a HF and HCl~monomer to the fluorine
83: or chlorine atom, respectively, of a neighboring monomer.
84: The bonding is caused by a partial withdrawal of charge from
85: the hydrogen atom due to the high electronegativity of
86: the fluorine or chlorine atoms.~\cite{Hamilton:HB-68,%
87: Pauling:NC-93,Scheiner:HB-97,Hadzi:TT-97,Karpfen:CE-02}
88: Hydrogen bonds are intermediates between ionic
89: bonding and van der Waals bonding and are of great
90: importance for the physical and chemical properties of many organic and
91: inorganic crystals. Moreover, they turn out to be crucial for the structure of
92: many biopolymers such as proteins and nucleic acids.~\cite{Hamilton:HB-68,%
93: Pauling:NC-93,Scheiner:HB-97,Hadzi:TT-97,Karpfen:CE-02}
94: 
95: At low temperature, HF and HCl crystals are structurally
96: very similar.
97: While HF forms strong hydrogen bonds, HCl forms
98: weak hydrogen bonds. Therefore, HF and HCl
99: represent good candidates for
100: a thorough analysis of this special type of bonding in crystals.
101: The monomers in both compounds are found to be arranged in terms of
102: parallel zigzag chains with a large interchain distance and, hence,
103: a weak interchain interaction [see Sec.~\ref{sec:comp} for details].
104: Frequently, a single infinite chain is considered as a simple but
105: realistic model of the crystals.
106: The isolated \HF{}~chain has fascinated many theoreticians
107: and much work has been carried out dealing with it;
108: namely, model studies,~\cite{Springborg:ES-87,Springborg:ES-88}
109: semiempirical~(intermediate neglect of differential overlap)
110: examinations of~\HF{} by Zunger,~\cite{Zunger:BS-75}
111: density functional theory calculations~[local density
112: approximation] of Springborg,~\cite{Springborg:ES-87,Springborg:ES-88}
113: and \emph{ab initio}
114: investigations.~\cite{Kertesz:AI-75,Karpfen:AI-76,%
115: Blumen:CC-76,Blumen:EB-77,Kertesz:AI-78,Karpfen:AI-80,Karpfen:HB-81,%
116: Karpfen:AI-82,Ihaya:AI-84,Liegener:AI-87,%
117: Mayer:CH-97,Berski:PHF-97,Hirata:AI-98,Jacquemin:LR-99,Buth:BS-04,Buth:MC-05}
118: In contrast, the \HCl{} chain has not been investigated very extensively;
119: there are only a few \emph{ab initio} examinations.~\cite{Blumen:CC-76,Blumen:EB-77,Blumen:EBS-77,Berski:PHF-98}
120: For both isolated chains the zigzag geometry of Fig.~\ref{fig:polygeom}
121: is the energetically favored arrangement in comparison with
122: the linear geometry. For \HF{} chains this has been shown in
123: early studies of Karpfen~\etal~\cite{Karpfen:AI-76,Karpfen:AI-80} and
124: Beyer and Karpfen.~\cite{Karpfen:AI-82}
125: The structure of isolated~\HCl{} chains is investigated
126: in Ref.~\onlinecite{Berski:PHF-98}.
127: 
128: 
129: \begin{figure}
130:   \includegraphics[width=\hsize,clip]{HF_chain.eps}
131:   \caption{(Color online) Structure of infinite~\HF{} zigzag chains.
132:            The structure of \HCl{}~chains is analogous.
133:            The grey (red) numbers below the monomers indicate the relative
134:            position of a monomer in the chain with respect to the
135:            monomer~``0'' at the origin.}
136:   \label{fig:polygeom}
137: \end{figure}
138: 
139: 
140: The large number of \emph{ab initio} studies that have
141: been carried out for the chains also exhibits how
142: challenging hydrogen-bonded systems are.
143: An accurate treatment of hydrogen bonds requires both a
144: good electron correlation method and a large one-particle
145: basis set.~\cite{Halkier:BS-99,Buth:BS-04,Buth:MC-05}
146: Hence, in Refs.~\onlinecite{Buth:BS-04,Buth:MC-05}, we applied
147: basis set extrapolation schemes---which
148: are powerful methods to improve the accuracy
149: of both Hartree-Fock and correlation
150: energies for small molecules---to extended systems,
151: showing their validity and usefulness also in
152: this situation.
153: Thereby, a periodic Hartree-Fock treatment
154: was accompanied by correlation calculations.
155: 
156: We use the coupled-cluster singles and doubles~(CCSD) correlation
157: method~\cite{Purvis:CC-82} in conjunction with the
158: incremental scheme~\cite{Stoll:CD-92,%
159: Stoll:CS-92,Stoll:CG-92,Fulde:EC-95,Fulde:WF-02}
160: which has been used in many applications and yields
161: a physically meaningful many-body decomposition
162: of the correlation energy.
163: It has been applied to a few one-dimensional systems before: the
164: \textit{trans}-polyacetylene polymer,~\cite{Yu:IA-97}
165: the infinite lithium hydride chain,~\cite{Abdurahman:AI-00}
166: the beryllium hydride polymer,~\cite{Abdurahman:AI-00} and
167: poly(\textit{para}-phenylene).~\cite{Willnauer:QC-04}
168: In the context of the present paper, the contributions of
169: Doll~\etal,~\cite{Doll:CE-95,Doll:CE-96,Doll:CP-97,Doll:GS-98}
170: who studied ionic solids, as well as the investigations of
171: rare-gas crystals by Ro\'sciszewski~\etal~\cite{Rosciszewski:AI-99,%
172: Rosciszewski:AI-00} are worth mentioning because
173: hydrogen-bonded crystals fall in between these two
174: types of bonding.
175: 
176: This paper is structured as follows.
177: The CCSD electron correlation method and the incremental scheme
178: are introduced in Sec.~\ref{sec:theory} whereas Sec.~\ref{sec:comp}
179: describes geometries, basis sets, and the computer programs employed.
180: In Sec.~\ref{sec:elcorr}, we discuss electron
181: correlation effects and their implications.
182: The contributions to the binding energy of
183: the~\HF{} and \HCl{} chains are evaluated in
184: Sec.~\ref{sec:bind} and conclusions are drawn in
185: Sec.~\ref{sec:conclusion}.
186: 
187: 
188: \section{Theory}
189: \label{sec:theory}
190: 
191: We choose the Hartree-Fock approximation as a starting point to
192: study electron correlations, employing the full non-relativistic
193: Hamiltonian of the chains within the fixed-nuclei
194: approximation.~\cite{Szabo:MQC-89,Fulde:EC-95,Ladik:PS-99,%
195: Fulde:WF-02,Buth:MC-05}
196: The resulting Hartree-Fock Bloch orbitals are transformed
197: to Wannier orbitals because the latter provide a more
198: appropriate, local representation of the Hamiltonian for
199: a subsequent treatment of electron
200: correlations.~\cite{Fulde:EC-95,Fulde:WF-02,Forner:LC-92}
201: The ground-state wave function is modeled by the coupled-cluster
202: ansatz~\cite{Szabo:MQC-89}
203: %
204: \begin{equation}
205:   \ket{\Psi_0^N} = \euler^{\hat T} \ket{\Phi_0^N}
206: \end{equation}
207: %
208: which relates the $N$-electron Hartree-Fock ground-state wave
209: function~$\Phi_0^N$ to the correlated ground-state wave
210: function~$\Psi_0^N$ employing the cluster operator~$\hat T$.
211: For the calculations in this study,
212: it is sufficient to restrict $\hat T$~to single and double
213: excitations, resulting in the coupled-cluster singles and
214: doubles~(CCSD) scheme~\cite{Purvis:CC-82,Hirata:CC-03}
215: %
216: \begin{eqnarray}
217:   \hat T &=& \Sum_{\vec R_1^{\prime} \, \alpha_1 \atop \vec R_1 \, \kappa_1}
218:     t^{\vec R_1^{\prime} \, \alpha_1}_{\vec R_1 \, \kappa_1}
219:     \hat c^{\dagger}_{\vec R_1^{\prime} \, \alpha_1}
220:     \hat c_{\vec R_1 \, \kappa_1} \\
221: %
222:   &&{} + \Sum_{\vec R_1^{\prime} \, \alpha_1 < \vec R_2^{\prime} \, \alpha_2
223:     \atop \vec R_1 \, \kappa_1 < \vec R_2 \, \kappa_2}
224:     t^{\vec R_1^{\prime} \, \alpha_1, \vec R_2^{\prime} \, \alpha_2}_{\vec
225:        R_1 \, \kappa_1, \vec R_2 \, \kappa_2}
226:     \hat c^{\dagger}_{\vec R_1^{\prime} \, \alpha_1}
227:     \hat c^{\dagger}_{\vec R_2^{\prime} \, \alpha_2} \hat c_{\vec R_1 \, \kappa_1}
228:     \hat c_{\vec R_2 \, \kappa_2}
229:     \; . \nonumber
230: \end{eqnarray}
231: %
232: The coefficients~$t^{\vec R_1^{\prime} \, \alpha_1}_{\vec R_1 \, \kappa_1}$
233: and $t^{\vec R_1^{\prime} \, \alpha_1, \vec R_2^{\prime} \, \alpha_2}_{\vec
234: R_1 \, \kappa_1, \vec R_2 \, \kappa_2}$~are referred to as excitation
235: amplitudes.
236: Let~$\vec r$ and $s$ denote spatial and spin coordinates;
237: then $\hat c^{\dagger}_{\vec R_1^{\prime} \, \alpha_1}$ creates electrons
238: in virtual spin Wannier orbitals~$w_{\vec R_1^{\prime} \, \alpha_1}(\vec
239: r \, s)$ whereas $\hat c_{\vec R_1 \, \kappa_1}$ annihilates
240: electrons from occupied spin Wannier orbitals~$w_{\vec R_1 \,
241: \kappa_1}(\vec r \, s)$.
242: Here $\vec R_1$ and $\vec R_1^{\prime}$ denote the unit cell in
243: which the Wannier orbital is located and $\alpha_1$ and
244: $\kappa_1$ refer to Wannier orbital indices.
245: Assuming Born-von K\'arm\'an boundary conditions,
246: the correlation energy per unit
247: cell~$\Cal E_{\rm corr}$ is given by
248: %
249: \begin{equation}
250:   \label{eq:ccsdenergy}
251:   N_0 \, \Cal E_{\rm corr} = \bra{\Psi_0^N} \hat H \ket{\Psi_0^N}
252:     - \bra{\Phi_0^N} \hat H \ket{\Phi_0^N} \; ,
253: \end{equation}
254: %
255: where $N_0$~is the number of unit cells in the chains.
256: 
257: Occupied Wannier orbitals are grouped in terms of $n_{\rm one}$
258: pairwise disjunct one-body orbital sets which are defined
259: by
260: %
261: \begin{equation}
262:   \label{eq:onebset}
263:   \vec R \, I_l = \{w_{\vec R \, \alpha}(\vec r \, s) \> |
264:     \> \alpha \in I_l\}
265: \end{equation}
266: %
267: for~$l = 1, \ldots, n_{\rm one}$.
268: 
269: The expression for the correlation energy of a
270: crystal~(\ref{eq:ccsdenergy}) is rearranged and decomposed in terms
271: of correlation energies~$\varepsilon_{\vec R_1 \, I_1 \cdots
272: \vec R_K \, I_K}$ of the electrons from
273: the one-body orbital sets~$\vec R_1 \, I_1 \cdots \vec R_K \, I_K$.
274: The resulting formula for the correlation energy of the chains
275: per unit cell~$\Cal E_{\rm corr}$ reads~\cite{Stoll:CD-92,%
276: Stoll:CS-92,Stoll:CG-92,Fulde:EC-95,Yu:IA-97,Abdurahman:AI-00,%
277: Fulde:WF-02,Willnauer:QC-04,Buth:MC-05}
278: %
279: \begin{widetext}
280: \begin{equation}
281:   \label{eq:inccryst}
282:   \begin{array}{rcl}
283:     N_0 \, \Cal E_{\rm corr} &=& \frac{1}{1!} \, \Sum_{\vec R_1 \, I_1}
284:       \Delta \varepsilon_{\vec R_1 \, I_1} + \frac{1}{2!} \, \Sum_{\vec R_1 \,
285:       I_1 \neq \vec R_2 \, I_2} \Delta \varepsilon_{\vec R_1 \, I_1 \; \vec R_2
286:       \, I_2} \\
287:     &&{} + \cdots + \frac{1}{K!} \, \Sum_{\vec R_1 \, I_1 \cdots \vec R_K \, I_K
288:       \atop \textrm{\scriptsize pairwise disjunct}}
289:       \Delta \varepsilon_{\vec R_1 \, I_1 \cdots \vec R_K \, I_K}
290:       + \cdots \; . \nonumber \\
291: \end{array}
292: \end{equation}
293: %
294: The one-body~$\Delta \varepsilon_{\vec R_1 \, I_1}$, the
295: two-body~$\Delta \varepsilon_{\vec R_1 \, I_1 \; \vec R_2 \, I_2}$,
296: \dots, up to the $K$-body~$\Delta \varepsilon_{\vec R_1 \, I_1 \cdots
297: \vec R_K \, I_K}$,~\ldots{} energy increments are defined recursively
298: by~\cite{Stoll:CD-92,Stoll:CS-92,Stoll:CG-92,Fulde:EC-95,Yu:IA-97,%
299: Abdurahman:AI-00,Fulde:WF-02,Willnauer:QC-04,Buth:MC-05}
300: %
301: \begin{subeqnarray}
302:   \label{eq:incdefs}
303:   \Delta \varepsilon_{\vec R_1 \, I_1} &=& \varepsilon_{\vec R_1 \, I_1} \; , \\
304: %
305:   \Delta \varepsilon_{\vec R_1 \, I_1 \; \vec R_2 \, I_2} &=&
306:   \varepsilon_{\vec R_1 \, I_1 \; \vec R_2 \, I_2} - \Delta
307:     \varepsilon_{\vec R_1 \, I_1} - \Delta \varepsilon_{\vec R_2 \, I_2} \; , \\
308: %
309:   & \vdots & \nonumber \\
310: %
311:   \Delta \varepsilon_{\vec R_1 \, I_1 \cdots \vec R_K \, I_K} &=&
312:     \varepsilon_{\vec R_1 \, I_1 \cdots \vec R_K \, I_K}
313:     - \Sum_{n = 1}^{K - 1}
314:     \frac{1}{n!} \Sum_{{{
315:     \{ \vec R_1^{\prime} \, I_1^{\prime} \cdots \vec R_n^{\prime}
316:     \, I_n^{\prime} \}
317:     \atop \subset \{ \vec R_1 \, I_1 \cdots \vec R_K \, I_K \} }
318:     \atop \vec R_1^{\prime} \, I_1^{\prime} \cdots \vec R_n^{\prime}
319:     \, I_n^{\prime} }
320:     \atop \textrm{\scriptsize pairwise disjunct}}
321:     \Delta \varepsilon_{\vec R_1^{\prime} \, I_1^{\prime} \cdots \vec
322:     R_n^{\prime} \, I_n^{\prime}} \; . \\
323: %
324:   & \vdots & \nonumber
325: \end{subeqnarray}
326: \end{widetext}
327: %
328: The factors~$\frac{1}{1!}$, $\frac{1}{2!}$, \ldots, $\frac{1}{K!}$,
329: \ldots{} in front of the sums on the right-hand side of
330: Eq.~(\ref{eq:inccryst}) account for permutations among the one-body
331: orbital sets of a certain $K$-body energy increment~($K \geq 2$).
332: One can eliminate all permutations that lead to the
333: same energy increment by letting the sums in Eq.~(\ref{eq:inccryst})
334: run only over distinct sets of one-body orbital sets.
335: 
336: The translational relation of the Wannier orbitals
337: leads to the translational symmetry of the energy
338: increments, \ie, $\Delta \varepsilon_{\vec 0 \, I_1 \cdots \vec R_K
339: - \vec R_1 \, I_K} = \Delta \varepsilon_{\vec R_1 \, I_1 \cdots \vec R_K
340: \, I_K}$. This can be exploited in Eq.~(\ref{eq:inccryst})
341: to make the right-hand side of Eq.~(\ref{eq:inccryst})
342: independent of the first lattice sum~$\Sum_{\vec R_1}$ which, for this
343: reason, is $N_0$~times the sum of the translational
344: symmetry-adapted energy increments.
345: It allows us to eliminate the factor~$N_0$ in front of the
346: left-hand side of Eq.~(\ref{eq:inccryst})
347: and enormously reduces the number of energy increments
348: that need to be calculated to describe the chains with a
349: given accuracy.
350: 
351: The localized occupied molecular orbitals of oligomers, comprising a few
352: monomers arranged in the geometry of the infinite chains, are
353: found to be a good approximation to the Wannier orbitals of infinite
354: chains and crystals.~\cite{Stoll:CD-92,Stoll:CS-92,Stoll:CG-92,%
355: Fulde:EC-95,Yu:IA-97,Abdurahman:AI-00,Fulde:WF-02,Willnauer:QC-04}
356: This facilitates the determination of the correlation contribution
357: to the binding energy of infinite chains from
358: the energy increments in oligomers using Eq.~(\ref{eq:inccryst}).
359: Yet the virtual Wannier orbitals involved in the determination
360: of the energy increments, and implicitly contained in
361: Eq.~(\ref{eq:incdefs}), are replaced by all the virtual canonical
362: molecular orbitals of the oligomers.
363: The procedure outlined in this paragraph is termed the incremental
364: scheme.~\cite{Stoll:CD-92,Stoll:CS-92,Stoll:CG-92,%
365: Fulde:EC-95,Fulde:WF-02}
366: 
367: 
368: \section{Computational details}
369: \label{sec:comp}
370: 
371: Hydrogen fluoride and hydrogen chloride crystallize in an orthorhombic
372: low-temperature phase described by the space groups~$Bm2_1b$ for
373: HF,~\cite{Atoji:CS-54,Habuda:NM-71,Johnson:CS-75} and
374: $Bb2_1m$ for HCl.~\cite{Sandor:CS-67,Sandor:CC-67,Sandor:ND-69}
375: The unit cells of both crystals contain four monomers which
376: are arranged in terms of weakly interacting parallel zigzag
377: chains~[Fig.~\ref{fig:polygeom}].
378: The chains are described by a unit cell which comprises two monomers
379: and are considered as an excellent one-dimensional model for
380: HF and HCl crystals.~\cite{Zunger:BS-75,%
381: Kertesz:AI-75,Karpfen:AI-76,Blumen:CC-76,%
382: Blumen:EB-77,Blumen:EBS-77,Kertesz:AI-78,%
383: Karpfen:AI-80,Karpfen:HB-81,Karpfen:AI-82,%
384: Ihaya:AI-84,Liegener:AI-87,%
385: Springborg:ES-87,Springborg:ES-88,Mayer:CH-97,%
386: Berski:PHF-97,Berski:PHF-98,Hirata:AI-98,%
387: Jacquemin:LR-99,Buth:BS-04,Buth:MC-05}
388: The structure of a single chain is determined by
389: three parameters, the H---$X$~distance~$r$,
390: the $X \cdots X$~distance~$R$, and the
391: angle~$\alpha = \angle({\rm H}X{\rm H})$, $X ={}$F$,$Cl.
392: Experimental values for the parameters are~$r = 0.92 \angstrom$,
393: $R = 2.50 \angstrom$, and $\alpha = 120 \degree$ for
394: \HF{},~\cite{Atoji:CS-54,Holleman:IC-01} and
395: $r = 1.25 \angstrom$, $R = 3.688 \angstrom$, and
396: $\alpha = 93.3 \degree$ for
397: \DCl{}.~\cite{Sandor:CS-67} HCl and deuterated DCl crystals
398: have very similar lattice
399: constants and are considered to be isomorphous.~\cite{Sandor:CS-67}
400: Unfortunately, further structural information for HCl crystals
401: is unavailable.
402: 
403: To calculate the energy increments in Eq.~(\ref{eq:inccryst}), we
404: perform molecular calculations on oligomers~\oHF{n} and
405: \oHCl{n}, respectively, \ie, short fragments of the chains.
406: We utilize the program package \textsc{molpro},~\cite{MOLPRO2002.6}
407: employing Foster-Boys localization~\cite{Boys:MO-60,Foster:MO-60}
408: of the molecular orbitals and the
409: coupled-cluster singles and doubles~(CCSD) correlation
410: method.~\cite{Purvis:CC-82,Hampel:CE-92,MOLPRO2002.6}
411: Thereby, we are consistent with Refs.~\onlinecite{Buth:BS-04,Buth:MC-05}
412: where also the CCSD~method was employed.
413: We use the aug-cc-pVDZ basis set~\cite{Dunning:GBS-89,%
414: Kendall:EA-92,Woon:GBS-93,basislib-04} which
415: yields somewhat less accurate total
416: binding energies than those obtained in
417: Refs.~\onlinecite{Buth:BS-04,Buth:MC-05}.
418: Yet we are mainly interested in the long-range behavior of
419: electron correlations in the hydrogen-bonded chains;
420: it is well described by the aug-cc-pVDZ basis set due
421: to vanishingly small geometrical orbital
422: overlaps among the one-body orbital sets
423: in the $K$-body energy increments considered.
424: In fact, the modulus of the two-body energy increment
425: involving the third-nearest-neighbor monomer~$|\varepsilon_{03}|$,
426: as obtained with the aug-cc-pVDZ basis set,~\cite{Dunning:GBS-89,%
427: Kendall:EA-92,Woon:GBS-93,basislib-04} is smaller
428: %
429: % HF: EINC47 aug-cc-pVDZ - aug-cc-pVTZ
430: % (-0.00000737 - (-0.00000752)) / 0.00000737 = 2.0%
431: %
432: by~$2 \%$ for~\HF{} and
433: %
434: % HCl: EINC36 aug-cc-pVDZ - aug-cc-pVTZ
435: % (-0.00001059 - (-0.00001106)) / -0.00001059 = 4.4%
436: %
437: by~$4 \%$ for~\HCl{} than~$|\varepsilon_{03}|$
438: obtained with the larger aug-cc-pVTZ basis set~\cite{Dunning:GBS-89,%
439: Kendall:EA-92,Woon:GBS-93,basislib-04}
440: [Sec.~\ref{sec:elcorr}].
441: Similarly, $K$-body energy increments for $K \geq 3$ agree
442: well as soon as the geometrical orbital overlaps between the
443: one-body orbital sets become negligible.
444: 
445: In Tab.~\ref{tab:bindtotal} below, reference is made to periodic
446: Hartree-Fock calculations of~\HF{} and \HCl{}
447: chains, described in Refs.~\onlinecite{Buth:BS-04,Buth:MC-05},
448: which were carried out with the \textsc{crystal}
449: program.~\cite{Pisani:HF-88,Pisani:QM-96,crystal03}
450: For these calculations, the $f, g, h, i$ functions
451: were removed from the basis sets.
452: To estimate the influence of the neglected basis functions on the
453: total binding energy of the chains, we use
454: \textsc{molpro},~\cite{MOLPRO2002.6} to calculate the Hartree-Fock
455: binding energy per monomer for~\oHF{9} and \oHCl{9}
456: both with and without $f, g, h$~functions using the
457: cc-pV5Z basis set;~\cite{Dunning:GBS-89,Woon:GBS-93,basislib-04}
458: at the Hartree-Fock level it is close to completeness with
459: respect to basis functions with the angular momenta~$s, p, d$.
460: The binding energies of both chains
461: obtained with $f, g, h$~functions are smaller
462: %
463: % Estimate Hartree-Fock binding in (HF)_9 energy of cc-pV5Z minus
464: %       cc-pV5Z-without-f,g,h-functions
465: %
466: % HF
467: % cc-pV5Z        : -900.70115997 / 9 - (-100.07035200) = -0.00755466
468: % cc-pV5Z-crystal: -900.69710425 / 9 - (-100.06988760) = -0.00756843
469: % cc-pV5Z minus cc-pV5Z-crystal = -0.00755466 - (-0.00756843) = 0.00001377
470: %
471: %
472: % HCl
473: % cc-pV5Z        : -4141.01380657 / 9 - (-460.11238499) = -0.0002602
474: % cc-pV5Z-crystal: -4141.00858712 / 9 - (-460.11179208) = -0.0002732
475: % cc-pV5Z minus cc-pV5Z-crystal = -0.00026020 - (-0.00027320) = 0.00001300
476: %
477: by~$\approx 13 \U{\mu E_h}$ than those determined without them.
478: This deviation turns out to be sufficiently small to be neglected
479: in the subsequent calculation of total binding energies
480: in Sec.~\ref{sec:bind}.
481: However, note that for the correlation calculations
482: the $f, g, h$~functions are fully considered.
483: 
484: The calculated binding energies reported here
485: account for the basis set superposition
486: error~\cite{Duijneveldt:CP-94}~(BSSE)
487: by the counterpoise correction~(CP);
488: it is determined by surrounding an isolated HF or HCl
489: monomer by sufficiently many ghost atoms positioned in
490: the geometry of the infinite
491: chains.~\cite{Boys:CP-70,Duijneveldt:CP-94}
492: 
493: 
494: \section{Electron correlations}
495: \label{sec:elcorr}
496: \subsection{Transferability}
497: \label{sec:transfer}
498: 
499: We have to ensure that the Foster-Boys-localized
500: molecular orbitals~\cite{Boys:MO-60,Foster:MO-60}
501: which are extracted from oligomers
502: are a good approximation to the Wannier orbitals of the
503: infinite \HF{} and \HCl{}~chains.
504: They have to be approximately translationally
505: related within a certain oligomer and
506: must not differ substantially
507: among oligomers of varying length.
508: These two properties of the localized
509: occupied molecular orbitals are termed transferability.
510: In order to achieve transferability for
511: moderately sized molecular clusters, one is
512: frequently obliged to account for the omitted
513: monomers in terms of an appropriate substitute.
514: In the case of~\HF{} and \HCl{}, two options
515: offer themselves. On the one hand, both
516: chains are cut out from a molecular crystal, implying
517: no substitutes as in rare-gas
518: crystals;~\cite{Rosciszewski:AI-99,Rosciszewski:AI-00}
519: on the other hand, particularly~\HF{} is rather
520: ionic and a surrounding by point charges
521: can be envisaged to create the embedding
522: which adequately models the infinite
523: chains.~\cite{Evjen:OS-32,Doll:CE-95,%
524: Doll:CE-96,Doll:CP-97,Doll:GS-98}
525: I explore both possibilities to elucidate what kind of procedure is
526: adequate for hydrogen-bonded solids.
527: The virtual orbitals of the oligomers are not altered
528: and remain canonical molecular orbitals
529: in the calculation of the individual energy increments.
530: The oligomer approximation thus corresponds
531: to a sort of domain decomposition of the
532: virtual space which also is applied in other
533: local correlation methods.~\cite{Pisani:LT-03}
534: 
535: 
536: \begin{figure}
537:   \includegraphics[width=\hsize,clip]{conv_one-body.eps}
538:    \caption{(Color online) One-body energy increments~$\Delta\varepsilon_0$
539:             of~\HF{} and \HCl{} chains as determined in oligomers of varying
540:             length reduced by the CP-corrected correlation energy~$\Cal
541:             E\I{mon}^{\rm corr}$ of the corresponding isolated monomer.
542:             The mean value of the two innermost energy increments
543:             of the oligomers is taken.
544:             Circles stand for~\HF{} and squares
545:             for~\HCl{} data. Open symbols refer to
546:             oligomers surrounded by point charges
547:             whereas closed symbols denote isolated
548:             oligomers.}
549:   \label{fig:one-body}
550: \end{figure}
551: 
552: 
553: \begin{figure}
554:   \includegraphics[width=\hsize,clip]{conv_two-body.eps}
555:    \caption{(Color online) Two-body energy increments~$\Delta\varepsilon_{01}$
556:             of~\HF{} and \HCl{} chains as determined in oligomers of varying length.
557:             The two innermost one-body orbital sets of the respective
558:             oligomer are taken to determine~$\Delta\varepsilon_{01}$.
559:             Symbols are chosen as in Fig.~\ref{fig:one-body}.}
560:   \label{fig:two-body}
561: \end{figure}
562: 
563: 
564: In the case of~\HF{} and \HCl{}, the outer and inner valence
565: orbitals, \ie, eight electrons, are assigned to the one-body
566: orbital set representing a particular HF or HCl~monomer~(\ref{eq:onebset}).
567: To test the transferability, we
568: calculate all one-body energy increments
569: and all two-body energy increments between
570: two adjacent monomers, the so-called
571: connected two-body energy increments,
572: in isolated oligomers~\oHF{n} and
573: \oHCl{n} for $n = 2, 4, 6, 10$.
574: They are compared with the energy increments from the
575: oligomers which were surrounded by point charges
576: up to twentieth-nearest neighbors.
577: The hydrogen atoms are represented by~$q = +1$;
578: fluorine and chlorine atoms are described
579: by~$q = -1$ with the exception of the terminal charges
580: which are set to~$q = -\frac{1}{2}$.~\cite{Evjen:OS-32,%
581: Doll:CE-95,Doll:CE-96,Doll:CP-97,Doll:GS-98}
582: 
583: Comparing the one-body and the connected two-body energy
584: increments obtained in oligomers of varying length in
585: Figs.~\ref{fig:one-body} and \ref{fig:two-body}, we
586: observe a rapid convergence of their values toward the
587: limit of the infinite chains
588: where the values determined for the isolated
589: oligomers and the values resulting from
590: the point-charge-embedded oligomers
591: approach each other quickly.
592: The energy increments taken from the isolated oligomers
593: increase (decrease) in Fig.~\ref{fig:one-body}~(Fig.~\ref{fig:two-body})
594: monotonically with the oligomer length
595: whereas the embedding with point charges causes an unsystematic
596: but quicker convergence.
597: The convergence behavior of the individual curves
598: in Figs.~\ref{fig:one-body} and \ref{fig:two-body}
599: is governed by a small variation in the localized orbitals
600: and, predominantly, by the improvement of the electronic
601: structure which approaches rapidly
602: the electronic structure of the infinite chains with
603: increasing length of the oligomers.
604: Obviously, the decision whether to employ point
605: charges or not has only a small impact on the values of the
606: energy increments.
607: The one-body (two-body) energy increments differ by
608: %
609: % CCSD, aug-cc-pVDZ, (HF)_10
610: % EINC4 - EINC4pointcharges
611: % -0.22338000 - (-0.22331345) = -0.00006655 Eh = -67 muEh
612: %
613: $67 \muHartree$
614: %
615: % CCSD, aug-cc-pVDZ, (HF)_10
616: % EINC45 - EINC45pointcharges
617: % -0.00408785 - (-0.00410198) =  0.00001413 Eh =  14 muEh
618: %
619: ($14 \muHartree$) in~\HF{} and by
620: %
621: % CCSD, aug-cc-pVDZ, (HCl)_10
622: % EINC4 - EINC4pointcharges
623: % -0.175099 - (-0.17507709) = -0.00002191 Eh = -21.91 muEh
624: %
625: $22 \muHartree$
626: %
627: % CCSD, aug-cc-pVDZ, (HCl)_10
628: % EINC45 - EINC45pointcharges
629: % -0.00282125 - (-0.00285217) = 0.00003092 Eh = 30.92 muEh
630: %
631: ($31 \muHartree$) in~\HCl{}.
632: In what follows only energy increments taken from
633: isolated oligomers are considered, as in
634: Refs.~\onlinecite{Buth:BS-04,Buth:MC-05}.
635: 
636: The error on the binding
637: energy due to the approximate translational
638: relation of the Wannier orbitals
639: taken from the finite-length oligomers can be estimated
640: by investigating the numerical
641: differences between energy increments whose values should
642: be identical due to (translational) symmetry. The
643: connected two-body increments, for example,
644: vary at most by~$40 \U{\mu E_h}$ in~\oHF{10} and at most
645: by~$11 \U{\mu E_h}$ in~\oHCl{10}.%
646: %
647: ~\footnote{The two first and the two last monomers at the
648: ends of the oligomers are not regarded as they are
649: appreciably influenced by finite-size effects.}
650: %
651: We conclude that the transferability prerequisite
652: of the energy increments is satisfied within good accuracy.
653: The agreement between the results obtained in the two
654: different ways described before corroborates
655: the applicability of oligomers to represent the
656: electronic structure of infinite~\HF{}
657: and \HCl{} chains.
658: 
659: 
660: \begin{table}
661:   \centering
662:   \begin{ruledtabular}
663:     \begin{tabular}{ldd}
664:       Increment     & $\HF$ & $\HCl$ \\
665:       \hline
666:       \quad $\Delta\varepsilon_{0}$          & -223.3632 & -175.0985 \\
667:       \quad $\Delta\varepsilon_{0\,1}$       &   -4.0879 &   -2.8213 \\
668:       \quad $\Delta\varepsilon_{0\,2}$       &   -0.0861 &   -0.1347 \\
669:       \quad $\Delta\varepsilon_{0\,1\,2}$    &   -0.0163 &    0.0014 \\
670:       \quad $\Delta\varepsilon_{0\,3}$       &   -0.0063 &   -0.0092 \\
671:       \quad $\Delta\varepsilon_{0\,1\,3}$    &   -0.0015 &   -0.0016 \\
672:       \quad $\Delta\varepsilon_{0\,1\,2\,3}$ &    0.0003 &    0.0006 \\
673:     \end{tabular}
674:   \end{ruledtabular}
675:   \caption{Exemplary energy increments of~\HF{} and \HCl{} chains
676:            taken from~\oHF{10} and \oHCl{10} oligomers.
677:            The energy increments are computed using the one-body
678:            orbital sets of the innermost monomers in the oligomers.
679:            All data are given in millihartrees.}
680:   \label{tab:increments}
681: \end{table}
682: 
683: 
684: \subsection{Short-range correlations}
685: \label{sec:increments}
686: 
687: Let us discuss the short-range correlation contributions first.
688: The most relevant energy increments of~\HF{}
689: and \HCl{} are summarized in Tab.~\ref{tab:increments}.
690: The CP-corrected correlation energy of the
691: %
692: % aug-cc-pVDZ, CP in (HF)_9
693: % -100.261311923055 - (-100.033971301475) = -0.227340621580
694: %
695: HF monomer, $-227.3406 \mHartree$,
696: %
697: % aug-cc-pVDZ, CP in (HCl)_9
698: % -460.268686021020 - (-460.092470961000) = -0.176215060020
699: %
700: and of the HCl monomer, $-176.2151 \mHartree$, are
701: %
702: % aug-cc-pVDZ, HF
703: % (-0.227340621580 - (-0.223363216268)) / (-0.223363216268)
704: %       = 0.01780689487935986815
705: %
706: only larger in magnitude by~$1.8 \%$
707: %
708: % aug-cc-pVDZ, HCl
709: % (-0.176215060020 - (-0.175098524764)) / (-0.175098524764)
710: %       = -0.00637661
711: %
712: and $0.6 \%$, respectively, than the
713: corresponding one-body energy
714: increments~$\Delta\varepsilon_0$.
715: They are a bit larger because the electrons of a monomer
716: in the chains experience Pauli repulsion exerted by the
717: neighboring monomers.
718: However, these effects are small in our case and the electronic
719: structure of the~HF and HCl monomers is essentially
720: preserved in the \HF{} and \HCl{}~chains.
721: 
722: The $K$-body energy increments, $K \geq 2$, describe the mutual
723: correlation of the valence electrons of several monomers.
724: They give rise to a pronounced nonlinear increase of the binding energy
725: of small clusters and short oligomers which is termed bond
726: cooperativity (Ref.~\onlinecite{Karpfen:CE-02}
727: and references therein).
728: The modulus of the connected two-body energy increment
729: of~\HCl{}, $|\Delta\varepsilon_{01}^{\rm HCl}|$,
730: %
731: % aug-cc-pVDZ
732: % (-4.0879 - (-2.8213)) / (-4.0879) = 0.309841
733: %
734: is~$31 \%$ smaller than the modulus of the corresponding
735: energy increment of~\HF{}, $|\Delta\varepsilon_{01}^{\rm HF}|$.
736: The connected three-body energy
737: increment~$\Delta\varepsilon_{012}^{\rm HCl}$
738: is repulsive, \ie, greater than zero, but~$\Delta\varepsilon_{012}^{\rm HF}$
739: is again attractive, \ie, smaller than zero.
740: The reverse trend is observed for the remaining energy increments.
741: The modulus of the energy increment~$\Delta\varepsilon_{02}^{\rm HCl}$
742: %
743: % aug-cc-pVDZ
744: % (-0.0861 - (-0.1347)) / (-0.0861) = -0.56446
745: %
746: is~$57 \%$ larger than~$|\Delta\varepsilon_{02}^{\rm HF}|$.
747: The connected four-body increment~$\Delta\varepsilon_{0123}^{\rm HCl}$
748: %
749: % aug-cc-pVDZ
750: % (0.0006 - (0.0003)) / (0.0003) = 1.0
751: %
752: is~$100 \%$ larger than~$\Delta\varepsilon_{0123}^{\rm HF}$ and both
753: are repulsive.
754: These two trends of the energy increments can be explained
755: by two effects.
756: First, short-range correlation is effective for nearest
757: neighbors [see also the ensuing Sec.~\ref{sec:long}];
758: it is apparently stronger in~\HF{} than in \HCl{} due to
759: the tighter packing of the monomers in~\HF{}
760: and the higher compactness of the HF~monomer itself.
761: Second, chlorine atoms have a higher polarizability than
762: fluorine atoms because their valence electrons
763: are more diffuse than those of the latter atom.
764: This causes the van der Waals interaction
765: to be stronger in~\HCl{} than in~\HF{},~\cite{Craig:MQ-84}
766: leading to the more distant energy increments being larger in~\HCl{}
767: when compared with~\HF{};
768: all this despite the larger intermonomer distances
769: in~\HCl{} than in~\HF{}.
770: 
771: However, note that one should not associate too much physics with the
772: particular values of energy increments;
773: namely, they depend on the specific unitary transformation
774: used to localize the occupied orbitals
775: (here the one of Foster and Boys~\cite{Boys:MO-60,Foster:MO-60}).
776: Only the binding energy, which involves the sum of all the energy
777: increments [Eqs.~(\ref{eq:RHF_bind}) and (\ref{eq:corrpoly})],
778: is a physical observable and thus invariant under
779: orbital transformations.
780: When an analysis of the individual energy increments
781: is to be meaningful, their value should be
782: fairly independent of the localization procedure
783: which is well satisfied in our case.
784: 
785: 
786: \begin{table}
787:   \centering
788:   \begin{ruledtabular}
789:     \begin{tabular}{rddd}
790:       Compound & C_6^{\rm odd} & C_6^{\rm even} & $Experiment$%
791: %
792: \footnote{The experimental data are taken from Ref.~\onlinecite{Kumar:PS-85};
793: it is a rotational average of the van der Waals constants for the
794: dispersion interaction between two isolated monomers.} \\
795: %
796:       \hline
797:       % Cardinal number fit
798:       \HF      & 23.4         & 20.2          & 19.00     \\
799:       % Cardinal number fit
800:       \HCl     & 121          & 120           & 130.4     \\
801:     \end{tabular}
802:   \end{ruledtabular}
803:   \caption{Van der Waals constants $C_6^{\rm odd}$
804:            and $C_6^{\rm even}$ of the two-body dispersion interaction
805:            for the~\HF{} and \HCl{} chains. They are obtained by fitting
806:            the two-body energy increments of Fig.~\ref{fig:two-body-short}
807:            with either odd or even cardinal numbers~$n$, respectively,
808:            to Eq.~(\ref{eq:vdwdisp}).
809:            All data are given in~hartree~bohr$^6$.}
810:   \label{tab:van-der-Waals}
811: \end{table}
812: 
813: 
814: \subsection{Long-range correlations}
815: \label{sec:long}
816: 
817: At a separation of two isolated monomers where the geometrical
818: overlap between the orbitals from distinct one-body orbital sets of the
819: monomers becomes negligible, only van der Waals dispersion
820: interactions remain.
821: This is also the case for the interaction between two
822: monomers in the infinite chains which is
823: correspondingly described by the two-body energy increments.
824: Their absolute values in~\HF{} and \HCl{}
825: for two monomers at a distance up to
826: seventh nearest neighbors are displayed in
827: Fig.~\ref{fig:two-body-short};
828: the curves drop off rapidly with distance.
829: 
830: The monomers in Fig.~\ref{fig:polygeom} that are labeled by
831: odd cardinal numbers are tilted with respect
832: to monomer~$0$. Likewise, monomers with even cardinal
833: numbers are arranged parallel to the monomer~$0$.
834: We will refer to the two types of two-body energy increments
835: that result from a parallel or tilted setting of monomers
836: as even or odd energy increments,
837: respectively.
838: The long-range interaction between two monomers
839: is approximated by the leading term of the two-body
840: van der Waals dispersion interaction~\cite{Klein:RG-76,Craig:MQ-84}
841: %
842: \begin{equation}
843:   \label{eq:vdwdisp}
844:   \varepsilon_{0 \, n}^{\textrm{\scriptsize vdW}} = \cases{
845:     \textstyle - \frac{C_6^{\rm odd }}{(\frac{n}{2} \, a)^6} \; , & $n$~odd       \cr
846:     \textstyle - \frac{C_6^{\rm even}}{(\frac{n}{2} \, a)^6} \; , & $n$~even$\;,$ \cr }
847: \end{equation}
848: %
849: where individual van der Waals constants~$C_6^{\rm odd}$ and
850: $C_6^{\rm even}$, respectively, are affixed for odd and even
851: energy increments.
852: Here $a$~represents the lattice constant of the chains.
853: %
854: % xmgrace 5.1.14 uses the Levenberg-Marquardt algorithm
855: % for non-linear curve fitting which is based on
856: % LMDIF from MINPACK, with some modifications.
857: %
858: % Literature (LMDIF):
859: % K. Levenberg (1944) A method for the solution of certain non-linear problems
860: % in least squares.  Quart. Appl. Math. 2:164-168.
861: %
862: % D.W. Marquardt (1963) An algorithm for least squares estimation of nonlinear
863: % parameters.  J. Soc. Industr. Appl. Math. 11:431-441  (also: SIAM J. Appl. Math.).
864: %
865: % Jorge J. More, The Levenberg-Marquardt Algorithm: Implementation
866: % and Theory. Numerical Analysis, G. A. Watson, editor.
867: % Lecture Notes in Mathematics 630, Springer-Verlag, 1977.
868: %
869: The van der Waals constants are obtained by a weighted
870: nonlinear curve fit~\cite{More:LM-78}
871: of the data in Fig.~\ref{fig:two-body-short} for either the odd
872: or the even energy increments.
873: The weights are chosen such that the fit reproduces the energy
874: increments with large~$n$ best, as for them the geometrical
875: orbital overlaps between the orbitals from two different
876: one-body orbital sets are negligible.
877: The van der Waals constants for the fits of the two sets
878: of translationally equivalent monomers are given in
879: Tab.~\ref{tab:van-der-Waals}.
880: Our theoretical data compare satisfactorily with the
881: experimental data for the dispersion interaction between
882: two isolated monomers also given in the table.
883: 
884: Having identified the van der Waals contribution to the
885: two-body energy increments in the chains, we can
886: subtract it
887: from the two-body energy increments to obtain
888: van der Waals-reduced energy increments
889: %
890: \begin{equation}
891:   \label{eq:vdw-reduced}
892:   \Delta \varepsilon^{\textrm{\scriptsize vdW}}_{0 n}
893:     = \Delta \varepsilon_{0 n} - \varepsilon_{0 n}^{
894:     \textrm{\scriptsize vdW}} \; .
895: \end{equation}
896: %
897: Their absolute value~$|\Delta \varepsilon_{0 n}^{
898: \textrm{\scriptsize vdW}}|$ is shown next to the absolute value of
899: the two-body energy increments~$|\Delta \varepsilon_{0 n}|$ in
900: Fig.~\ref{fig:two-body-short}.
901: 
902: Two regions can be identified in Fig.~\ref{fig:two-body-short}.
903: First, from the nearest to the third-nearest neighbors, there
904: is the local correlation zone, where, due to geometrical orbital
905: overlaps, short-range electron correlations are effective.
906: There, the decay of two-body energy increments
907: with the distance between the two monomers
908: is faster than the one that would result from a pure van
909: der Waals interaction, following Eq.~(\ref{eq:vdwdisp}).
910: We observe that the van der Waals contribution to~$|\Delta \varepsilon^{
911: \textrm{\scriptsize vdW}}_{0 1}|$ of~\HCl{} is appreciably
912: larger than in~\HF{} as the curves for~$|\Delta \varepsilon_{0 1}|$
913: and $|\Delta \varepsilon^{\textrm{\scriptsize vdW}}_{0 1}|$ are
914: much closer for~\HCl{} than for~\HF{}.
915: Second, there is the van der Waals zone reaching from the
916: third- up to the seventh-nearest neighbor.
917: Here a typical $r^{-6}$~decay is observed, leading to van
918: der Waals-reduced energy increments that are essentially zero.
919: A slight deviation of the two-body energy increments
920: from an $r^{-6}$~behavior is perceived beyond
921: fifth-nearest neighbors which likely can be
922: attributed to inaccuracies caused by the oligomer
923: approximation.
924: The translational relation of the Wannier orbitals is better
925: satisfied in~\HCl{} compared with~\HF{} [Sec.~\ref{sec:transfer}]
926: which leads to a lower absolute value of the fourth to seventh
927: van der Waals-reduced energy increments.
928: The low absolute value of~$\Delta \varepsilon^{\textrm{\scriptsize
929: vdW}}_{0 6}$ for both chains is an artifact of the fitting
930: process.
931: Figure~\ref{fig:two-body-short} also reveals that the two-body
932: energy increments
933: are of satisfactory accuracy even beyond the estimates
934: given in~Sec.~\ref{sec:transfer} of~$40 \U{\mu E_h}$ for~\oHF{10}
935: and $11 \U{\mu E_h}$ for~\oHCl{10}
936: due to a considerable error cancellation.
937: 
938: 
939: \begin{figure}
940:   \includegraphics[width=\hsize,clip]{decay_two-body.eps}
941:   \caption{(Color online) Absolute values of two-body energy
942:            increments~$|\Delta \varepsilon_{0 \, n}|$ (closed symbols)
943:            and van der Waals-reduced two-body energy
944:            increments~$|\Delta \varepsilon_{0 \, n}^{\textrm{\scriptsize
945:            vdW}}|$ (open symbols)
946:            using Eq.~(\ref{eq:vdwdisp}) and the
947:            constants in Tab.~\ref{tab:van-der-Waals} for
948:            both the \HF{} chain (circles) and the \HCl{} chain (squares).}
949:   \label{fig:two-body-short}
950: \end{figure}
951: 
952: 
953: Given the van der Waals constants in Tab.~\ref{tab:van-der-Waals}, we
954: can use Eq.~(\ref{eq:vdwdisp}) to sum up the van der Waals
955: contribution of the two-body energy increments to the binding energy
956: to infinite distance~$\Sum_{n=3}^{\infty} \varepsilon_{0 n}^{
957: \textrm{\scriptsize vdW}}$ which yields
958: $-8 \U{\mu E_h}$~for~\HF{} and $-12 \U{\mu E_h}$~for~\HCl{}.
959: The bigger polarizability of chlorine atoms compared with
960: fluorine atoms leads to the larger contribution
961: in~\HCl{} when compared with~\HF{} as can be expected from the
962: analysis of the energy increments in Sec.~\ref{sec:increments}.
963: The contribution of two-body energy increments beyond
964: %
965: % HF: Amount of third-neighrest neighbor increment
966: % (-0.008 - (-0.0063)) / (-0.0063) = 0.2698
967: %
968: %
969: % HCl: Amount of third-neighrest neighbor increment
970: % (-0.012 - (-0.0092)) / (-0.0092) = 0.3044
971: %
972: third-nearest neighbors is in both cases~$\approx 30 \%$
973: of~$\Delta \varepsilon_{0 \, 3}$.
974: 
975: Generally, once a good estimate of the van der Waals constants of a crystal is
976: available, the decomposition of the energy increments
977: into a short- and a long-range van der Waals
978: contribution can be made. This allows one to
979: focus on the short-range part.
980: Thereby, the required number of energy increments
981: to be calculated to reach a certain accuracy of the correlation
982: energy is reduced considerably.
983: Therefore, the computational demand of its determination is
984: also significantly decreased.
985: 
986: 
987: \section{Binding energy}
988: \label{sec:bind}
989: 
990: Having understood the impact of electron correlations
991: in~\HF{} and \HCl{} chains, we are now in a position to examine
992: the total binding energy per monomer~$\Delta E$ of the chains;
993: it is given by
994: %
995: \begin{equation}
996:   \label{eq:RHF_bind}
997:   \begin{array}{rcl}
998:     \Delta E &=& \Delta E_{\rm SCF} + \Delta E_{\rm corr} \\
999:              &=& \frac{1}{2} \, (\Cal E^{\rm SCF}_{\rm chain}
1000:                  + \Cal E^{\rm corr}_{\rm chain})
1001:                  - \Cal E^{\rm SCF}_{\rm monomer}
1002:                  - \Cal E^{\rm corr}_{\rm monomer} \; .
1003:   \end{array}
1004: \end{equation}
1005: %
1006: It consists of the Hartree-Fock and the correlation
1007: contributions~$\Delta E_{\rm SCF}$ and $\Delta
1008: E_{\rm corr}$, respectively, where $\Cal E^{\rm
1009: SCF}_{\rm chain}$ and $\Cal E^{\rm corr}_{\rm chain}$
1010: denote the Hartree-Fock and correlation energies per
1011: unit cell of the chains.
1012: Furthermore, $\Cal E^{\rm SCF}_{\rm monomer}$
1013: and $\Cal E^{\rm corr}_{\rm monomer}$ are the corresponding
1014: energies for the monomers.
1015: The Hartree-Fock energies are readily available from molecular and
1016: periodic calculations~\cite{Pisani:HF-88,Pisani:QM-96,crystal03}
1017: but electron correlations are more involved and were discussed
1018: in the previous Sec.~\ref{sec:elcorr}.
1019: 
1020: In Tab.~\ref{tab:bindtotal}, we communicate basis-set-extrapolated
1021: Hartree-Fock binding energies~$\Delta E_{\rm SCF}(\infty)$ for the
1022: chains taken from Refs.~\onlinecite{Buth:BS-04,Buth:MC-05}.
1023: At the Hartree-Fock equilibrium geometry of~\HF{},
1024: Bayer and Karpfen give~$-10.360 \mHartree$ for basis set~3 in
1025: Ref.~\onlinecite{Karpfen:AI-82} while Hirata and Iwata
1026: find~$-10.855 \mHartree$ for the 6-311++G(d,p) basis
1027: set.~\cite{Hirata:AI-98}
1028: Both numbers are in good agreement with our result which
1029: is~$-10.202 \mHartree$.
1030: Yet Berski and Latajka report~$-9.696 \mHartree$ for~\HF{},
1031: when using the 6-311G(d,p) basis set,~\cite{Berski:PHF-97} and
1032: $-2.073 \mHartree$ for~\HCl{} harnessing the DZ(d,p)
1033: basis set.~\cite{Berski:PHF-98}
1034: The latter number in particular deviates considerably from our
1035: result, $-0.586 \mHartree$,
1036: which is most likely due to the fact that
1037: they did not remove the BSSE.~\cite{Berski:PHF-97,Berski:PHF-98}
1038: 
1039: To determine the correlation energy per monomer of the
1040: infinite chains~$\Cal E_{\rm corr}$, we inspect
1041: the energy increments in Tab.~\ref{tab:increments}.
1042: They converge rapidly both with increasing distance among the
1043: monomers involved and with increasing order of the many-body
1044: expansion, \ie, with the maximum number of monomers correlated
1045: in a specific energy increment.
1046: In order to obtain an accuracy in~$\Cal E_{\rm corr}$
1047: of $\approx 50 \U{\mu E_h}$ for~\HF{} and
1048: of $\approx 11 \U{\mu E_h}$ for~\HCl{},
1049: we see from Tab.~\ref{tab:increments} that it is
1050: enough to include the following energy increments:
1051: %
1052: \begin{equation}
1053:   \label{eq:corrpoly}
1054:   \Cal E_{\rm corr} \approx \Delta \varepsilon_0 + \Delta \varepsilon_{01}
1055:   + \Delta \varepsilon_{02} \; .
1056: \end{equation}
1057: %
1058: Hence, it is sufficient to correlate
1059: the electrons in the Wannier orbitals of only two
1060: monomers at a time for \HF{} and \HCl{}~chains
1061: in which more distant two-body energy increments contribute
1062: only $-8 \U{\mu E_h}$~for~\HF{} and $-12 \U{\mu E_h}$~for~\HCl{}
1063: as shown in Sec.~\ref{sec:long}.
1064: In three-dimensional crystals, three-body terms
1065: become more important according to the analyses in
1066: Refs.~\onlinecite{Axilrod:IV-43,Fink:AI-95,Rosciszewski:AI-00}
1067: and the much larger number of three-body energy increments
1068: compared with the number of two-body energy increments.
1069: 
1070: We report~$\Delta E$ based on the approximation
1071: of Eq.~(\ref{eq:corrpoly}) as obtained by basis-set
1072: extrapolation~\cite{Buth:BS-04,Buth:MC-05} in Tab.~\ref{tab:bindtotal}.
1073: Using density functional theory, Hirata and Iwata~\cite{Hirata:AI-98}
1074: find at optimized geometries for the total binding energy
1075: of~\HF{} $-13.521 \mHartree$ (BLYP) and $-13.864 \mHartree$ (B3LYP)
1076: [utilizing the 6-311++G(d,p) basis set] which are larger in magnitude
1077: by~$14\%$ and $17\%$, respectively, than our result
1078: for the experimental geometry.
1079: The observed differences can be partially ascribed
1080: to the different geometries used;
1081: our experimental geometry leads to a binding energy that is
1082: somewhat smaller in magnitude than the binding energy in
1083: the optimized geometry.
1084: The 6-311++G(d,p) basis set of Iwata and Hirata~\cite{Hirata:AI-98} also
1085: differs substantially from our choices
1086: in Refs.~\onlinecite{Buth:BS-04,Buth:MC-05}.
1087: Above all, we do not consider the impact of triple excitations in our study;
1088: they increase the absolute value of the short-range correlation
1089: contributions and, consequently, also the absolute value of
1090: the binding energies.
1091: 
1092: 
1093: \section{Conclusion}
1094: \label{sec:conclusion}
1095: 
1096: \begin{table}
1097: %
1098: % HF Hartree-Fock binding energy per monomer: -11.82525 mEh
1099: % Mean of all least squares fitted Hartree-Fock binding energies:
1100: % ((-10.199) + (-10.194) + (-10.194) + (-10.222)) / 4 = -10.20225 mEh = -0.27761735 eV
1101: %
1102: % Mean of the best extrapolated CCSD binding energies:
1103: % ((-1.655) + (-1.606) + (-1.646) + (-1.585)) / 4 = -1.623 mEh = -0.044164 eV
1104: %
1105: %
1106: %
1107: % HCl Hartree-Fock binding energy per monomer: -3.34975 mEh
1108: % Mean of all least squares fitted Hartree-Fock binding energies:
1109: % ((-0.596) + (-0.586) + (-0.584) + (-0.577)) / 4 = -0.58575 mEh = -0.0159391 eV
1110: %
1111: % Mean of the best extrapolated CCSD binding energies:
1112: % ((-2.773) + (-2.747) + (-2.774) + (-2.762)) / 4 = -2.764 mEh = -0.075212 eV
1113: %
1114: %
1115:   \centering
1116:   \begin{ruledtabular}
1117:     \begin{tabular}{rdd}
1118:                    & $\HF{}$ & $\HCl{}$ \\
1119:       \hline
1120:       $\Delta E_{\rm SCF}(\infty)$  & -10.202 & -0.586 \\
1121:       $\Delta E_{\rm corr}(\infty)$ &  -1.623 & -2.764 \\
1122:       $\Delta E(\infty)$            & -11.826 & -3.350 \\
1123:     \end{tabular}
1124:   \end{ruledtabular}
1125:   \caption{Basis-set-extrapolated total binding energies per
1126:            monomer~$\Delta E(\infty)$ of~\HF{} and \HCl{}~chains and their
1127:            decomposition into basis-set-extrapolated
1128:            Hartree-Fock~$\Delta E_{\rm SCF}(\infty)$ and
1129:            electron correlation~$\Delta E_{\rm corr}(\infty)$
1130:            contributions.
1131:            All data are taken from Refs.~\onlinecite{Buth:BS-04,Buth:MC-05}
1132:            and are given in millihartrees.}
1133:   \label{tab:bindtotal}
1134: \end{table}
1135: 
1136: 
1137: We study hydrogen bonding in infinite \HF{} and \HCl{} chains.
1138: Whereas in~\HF{} the Hartree-Fock contribution dominates the total
1139: binding energy
1140: %
1141: % -10.202 / -11.826 = 86.27%
1142: %
1143: by~$86\%$ due to the electrostatic contribution of the rather ionic
1144: HF~monomers, in the much more weakly bound~\HCl{}, the Hartree-Fock
1145: calculation yields
1146: %
1147: % -0.586 /  -3.350 = 17.49%
1148: %
1149: only~$18\%$ of the binding energy [Tab.~\ref{tab:bindtotal}].
1150: The transitional character of hydrogen-bonded crystals between the
1151: ionic and van der Waals regimes is reflected in these numbers.
1152: The very weak hydrogen bonds in the \HCl{} chains bear a
1153: close resemblance to purely van der Waals-bonded systems,
1154: like rare-gas solids, where
1155: bonding is entirely caused by electron
1156: correlations.~\cite{Fink:AI-95,Rosciszewski:AI-99,%
1157: Rosciszewski:AI-00}
1158: In fact only the inclusion of electron correlations
1159: puts the binding energy per monomer of the \HCl{} chain
1160: into the range conventionally ascribed to hydrogen
1161: bonding,~\cite{Pauling:NC-93}
1162: %
1163: % Pauling gives for hydrogen bonding the range 2 to 10 kJ / mol
1164: % per bond.
1165: %
1166: %  2 kJ / mol =  2 * 4.184 / 96.485 / 27.2113845 * 1000 mE_h
1167: %            =  3.187 mE_h
1168: %
1169: % 10 kJ / mol = 10 * 4.184 / 96.485 / 27.2113845 * 1000 mE_h
1170: %            = 15.936 mE_h
1171: %
1172: namely~\hbox{$3$--$16 \mHartree$}.
1173: 
1174: The incremental scheme, which provides a decomposition of the
1175: contribution of electron correlations to the binding energy of the
1176: chains in terms of nonadditive many-body energy increments,
1177: is shown to converge rapidly with respect to the number
1178: of monomers, \ie, bodies, involved in the energy increments
1179: and the distance among them, thus providing a
1180: good tool to study hydrogen bonding in crystals.
1181: In contrast to three-dimensional crystals, the dominant
1182: contribution of electron correlations to the binding energies of
1183: the infinite chains is, in the present investigation,
1184: already given by pair interactions
1185: between the valence electrons of a monomer and its nearest
1186: and next-nearest neighbors.
1187: The $K$-body energy increments represent a general framework
1188: to describe van der Waals dispersion
1189: interaction among the electrons from $K$~one-body orbital sets.
1190: The decay of the two-body energy increments,
1191: \ie, the electron correlations between two monomers in the
1192: chains, with the intermonomer distance is investigated and
1193: van der Waals constants are determined,
1194: which affords a partition of the energy increments
1195: into a short-range, van der Waals-reduced part
1196: and a long-range part that can be summed analytically
1197: to infinite distances.
1198: 
1199: This partitioning affords a methodological advancement;
1200: if the van der Waals constants of a crystal
1201: are available from other sources, then the number of energy increments
1202: in the incremental expansion can be reduced by considering only
1203: van der Waals-reduced energy increments;
1204: they typically need to comprise only the nearest- and
1205: next-nearest-neighbor energy increments.
1206: The small number of van der Waals-reduced energy increments
1207: permits one to significantly decrease the number of energy
1208: increments which have actually to be calculated
1209: to reach a certain accuracy of the correlation energy
1210: compared with the conventional incremental series.
1211: Furthermore, the long-range part turns out to be
1212: well described using small (double-$\zeta$) basis
1213: sets---in contrast to the short-range part where basis-set
1214: extrapolation is required for an accurate
1215: description~\cite{Buth:BS-04,Buth:MC-05}---thus facilitating
1216: further computational savings.
1217: 
1218: 
1219: \begin{acknowledgments}
1220: We are highly indebted to Uwe Birkenheuer, Krzysztof Ro\'sciszewski,
1221: Hermann Stoll, and Peter Fulde for helpful discussions and
1222: a critical reading of the manuscript.
1223: \end{acknowledgments}
1224: 
1225: 
1226: \begin{thebibliography}{77}
1227: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
1228: \expandafter\ifx\csname bibnamefont\endcsname\relax
1229:   \def\bibnamefont#1{#1}\fi
1230: \expandafter\ifx\csname bibfnamefont\endcsname\relax
1231:   \def\bibfnamefont#1{#1}\fi
1232: \expandafter\ifx\csname citenamefont\endcsname\relax
1233:   \def\citenamefont#1{#1}\fi
1234: \expandafter\ifx\csname url\endcsname\relax
1235:   \def\url#1{\texttt{#1}}\fi
1236: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL }\fi
1237: \providecommand{\bibinfo}[2]{#2}
1238: \providecommand{\eprint}[2][]{\url{#2}}
1239: 
1240: \bibitem[{\citenamefont{Atoji and Lipscomb}(1954)}]{Atoji:CS-54}
1241: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Atoji}} \bibnamefont{and}
1242:   \bibinfo{author}{\bibfnamefont{W.~N.} \bibnamefont{Lipscomb}},
1243:   \bibinfo{journal}{Acta. Cryst.} \textbf{\bibinfo{volume}{7}},
1244:   \bibinfo{pages}{173} (\bibinfo{year}{1954}).
1245: 
1246: \bibitem[{\citenamefont{Habuda and Gagarinsky}(1971)}]{Habuda:NM-71}
1247: \bibinfo{author}{\bibfnamefont{S.~P.} \bibnamefont{Habuda}} \bibnamefont{and}
1248:   \bibinfo{author}{\bibfnamefont{Y.~V.} \bibnamefont{Gagarinsky}},
1249:   \bibinfo{journal}{Acta. Cryst.} \textbf{\bibinfo{volume}{B27}},
1250:   \bibinfo{pages}{1677} (\bibinfo{year}{1971}).
1251: 
1252: \bibitem[{\citenamefont{Johnson et~al.}(1975)\citenamefont{Johnson, S\'andor,
1253:   and Arzi}}]{Johnson:CS-75}
1254: \bibinfo{author}{\bibfnamefont{M.~W.} \bibnamefont{Johnson}},
1255:   \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{S\'andor}}, \bibnamefont{and}
1256:   \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Arzi}},
1257:   \bibinfo{journal}{Acta. Cryst.} \textbf{\bibinfo{volume}{B31}},
1258:   \bibinfo{pages}{1998} (\bibinfo{year}{1975}).
1259: 
1260: \bibitem[{\citenamefont{Otto and Steinborn}(1986)}]{Otto:BE-86}
1261: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Otto}} \bibnamefont{and}
1262:   \bibinfo{author}{\bibfnamefont{E.~O.} \bibnamefont{Steinborn}},
1263:   \bibinfo{journal}{Solid State Communications} \textbf{\bibinfo{volume}{58}},
1264:   \bibinfo{pages}{281} (\bibinfo{year}{1986}).
1265: 
1266: \bibitem[{\citenamefont{Panas}(1993)}]{Panas:HF-93}
1267: \bibinfo{author}{\bibfnamefont{I.}~\bibnamefont{Panas}}, \bibinfo{journal}{Int.
1268:   J. Quantum Chem.} \textbf{\bibinfo{volume}{46}}, \bibinfo{pages}{109}
1269:   (\bibinfo{year}{1993}).
1270: 
1271: \bibitem[{\citenamefont{Berski and
1272:   Latajka}(1998{\natexlab{a}})}]{Berski:DHF-98}
1273: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Berski}} \bibnamefont{and}
1274:   \bibinfo{author}{\bibfnamefont{Z.}~\bibnamefont{Latajka}},
1275:   \bibinfo{journal}{J. Mol. Struct.} \textbf{\bibinfo{volume}{450}},
1276:   \bibinfo{pages}{259} (\bibinfo{year}{1998}{\natexlab{a}}).
1277: 
1278: \bibitem[{\citenamefont{S\'andor and
1279:   Farrow}(1967{\natexlab{a}})}]{Sandor:CS-67}
1280: \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{S\'andor}} \bibnamefont{and}
1281:   \bibinfo{author}{\bibfnamefont{R.~F.~C.} \bibnamefont{Farrow}},
1282:   \bibinfo{journal}{Nature} \textbf{\bibinfo{volume}{213}},
1283:   \bibinfo{pages}{171} (\bibinfo{year}{1967}{\natexlab{a}}).
1284: 
1285: \bibitem[{\citenamefont{S\'andor and
1286:   Farrow}(1967{\natexlab{b}})}]{Sandor:CC-67}
1287: \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{S\'andor}} \bibnamefont{and}
1288:   \bibinfo{author}{\bibfnamefont{R.~F.~C.} \bibnamefont{Farrow}},
1289:   \bibinfo{journal}{Nature} \textbf{\bibinfo{volume}{215}},
1290:   \bibinfo{pages}{1265} (\bibinfo{year}{1967}{\natexlab{b}}).
1291: 
1292: \bibitem[{\citenamefont{S\'andor and Farrow}(1969)}]{Sandor:ND-69}
1293: \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{S\'andor}} \bibnamefont{and}
1294:   \bibinfo{author}{\bibfnamefont{R.~F.~C.} \bibnamefont{Farrow}},
1295:   \bibinfo{journal}{Disc. Faraday Soc.} \textbf{\bibinfo{volume}{48}},
1296:   \bibinfo{pages}{78} (\bibinfo{year}{1969}).
1297: 
1298: \bibitem[{\citenamefont{Hamilton and Ibers}(1968)}]{Hamilton:HB-68}
1299: \bibinfo{author}{\bibfnamefont{W.~C.} \bibnamefont{Hamilton}} \bibnamefont{and}
1300:   \bibinfo{author}{\bibfnamefont{J.~A.} \bibnamefont{Ibers}},
1301:   \emph{\bibinfo{title}{Hydrogen bonding in solids}}
1302:   (\bibinfo{publisher}{Benjamin}, \bibinfo{address}{New York, Amsterdam},
1303:   \bibinfo{year}{1968}).
1304: 
1305: \bibitem[{\citenamefont{Pauling}(1993)}]{Pauling:NC-93}
1306: \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Pauling}},
1307:   \emph{\bibinfo{title}{The nature of the chemical bond and the structure of
1308:   molecules and crystals}} (\bibinfo{publisher}{Cornell University Press},
1309:   \bibinfo{address}{Ithaca (New York)}, \bibinfo{year}{1993}),
1310:   \bibinfo{edition}{3rd} ed., ISBN \bibinfo{isbn}{0-8014-0333-2}.
1311: 
1312: \bibitem[{\citenamefont{Scheiner}(1997)}]{Scheiner:HB-97}
1313: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Scheiner}},
1314:   \emph{\bibinfo{title}{Hydrogen bonding---A theoretical perspective}}
1315:   (\bibinfo{publisher}{Oxford University Press}, \bibinfo{address}{New York,
1316:   Oxford}, \bibinfo{year}{1997}), ISBN \bibinfo{isbn}{0-19-509011-X}.
1317: 
1318: \bibitem[{\citenamefont{Had\v{z}i}(1997)}]{Hadzi:TT-97}
1319: \bibinfo{editor}{\bibfnamefont{D.}~\bibnamefont{Had\v{z}i}}, ed.,
1320:   \emph{\bibinfo{title}{Theoretical treatments of hydrogen bonding}}
1321:   (\bibinfo{publisher}{John Wiley \& Sons}, \bibinfo{address}{Chichester, New
1322:   York}, \bibinfo{year}{1997}), ISBN \bibinfo{isbn}{0-471-97395-5}.
1323: 
1324: \bibitem[{\citenamefont{Karpfen}(2002)}]{Karpfen:CE-02}
1325: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Karpfen}}, in
1326:   \emph{\bibinfo{booktitle}{Adv. Chem. Phys.}}, edited by
1327:   \bibinfo{editor}{\bibfnamefont{I.}~\bibnamefont{Prigogine}} \bibnamefont{and}
1328:   \bibinfo{editor}{\bibfnamefont{S.~A.} \bibnamefont{Rice}}
1329:   (\bibinfo{publisher}{John Wiley~\& Sons}, \bibinfo{address}{New York},
1330:   \bibinfo{year}{2002}), vol. \bibinfo{volume}{123}, pp.
1331:   \bibinfo{pages}{469--510}.
1332: 
1333: \bibitem[{\citenamefont{Springborg}(1987)}]{Springborg:ES-87}
1334: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Springborg}},
1335:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{59}},
1336:   \bibinfo{pages}{2287} (\bibinfo{year}{1987}).
1337: 
1338: \bibitem[{\citenamefont{Springborg}(1988)}]{Springborg:ES-88}
1339: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Springborg}},
1340:   \bibinfo{journal}{Phys. Rev.~B} \textbf{\bibinfo{volume}{38}},
1341:   \bibinfo{pages}{1483} (\bibinfo{year}{1988}).
1342: 
1343: \bibitem[{\citenamefont{Zunger}(1975)}]{Zunger:BS-75}
1344: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Zunger}}, \bibinfo{journal}{J.
1345:   Chem. Phys.} \textbf{\bibinfo{volume}{63}}, \bibinfo{pages}{1713}
1346:   (\bibinfo{year}{1975}).
1347: 
1348: \bibitem[{\citenamefont{Kert\'esz et~al.}(1975)\citenamefont{Kert\'esz, Koller,
1349:   and A$\rm\check{z}$man}}]{Kertesz:AI-75}
1350: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Kert\'esz}},
1351:   \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Koller}}, \bibnamefont{and}
1352:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{A$\rm\check{z}$man}},
1353:   \bibinfo{journal}{Chem. Phys. Lett.} \textbf{\bibinfo{volume}{36}},
1354:   \bibinfo{pages}{576} (\bibinfo{year}{1975}).
1355: 
1356: \bibitem[{\citenamefont{Karpfen and Schuster}(1976)}]{Karpfen:AI-76}
1357: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Karpfen}} \bibnamefont{and}
1358:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Schuster}},
1359:   \bibinfo{journal}{Chem. Phys. Lett.} \textbf{\bibinfo{volume}{44}},
1360:   \bibinfo{pages}{459} (\bibinfo{year}{1976}).
1361: 
1362: \bibitem[{\citenamefont{Blumen and Merkel}(1976)}]{Blumen:CC-76}
1363: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Blumen}} \bibnamefont{and}
1364:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Merkel}},
1365:   \bibinfo{journal}{Solid State Commun.} \textbf{\bibinfo{volume}{20}},
1366:   \bibinfo{pages}{755} (\bibinfo{year}{1976}).
1367: 
1368: \bibitem[{\citenamefont{Blumen and Merkel}(1977{\natexlab{a}})}]{Blumen:EB-77}
1369: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Blumen}} \bibnamefont{and}
1370:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Merkel}},
1371:   \bibinfo{journal}{phys. stat. sol. (b)} \textbf{\bibinfo{volume}{83}},
1372:   \bibinfo{pages}{425} (\bibinfo{year}{1977}{\natexlab{a}}).
1373: 
1374: \bibitem[{\citenamefont{Kert\'esz et~al.}(1978)\citenamefont{Kert\'esz, Koller,
1375:   and A$\rm\check{z}$man}}]{Kertesz:AI-78}
1376: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Kert\'esz}},
1377:   \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Koller}}, \bibnamefont{and}
1378:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{A$\rm\check{z}$man}},
1379:   \bibinfo{journal}{Z. Naturforsch. A} \textbf{\bibinfo{volume}{33}},
1380:   \bibinfo{pages}{249} (\bibinfo{year}{1978}).
1381: 
1382: \bibitem[{\citenamefont{Karpfen}(1980)}]{Karpfen:AI-80}
1383: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Karpfen}},
1384:   \bibinfo{journal}{Chem. Phys.} \textbf{\bibinfo{volume}{47}},
1385:   \bibinfo{pages}{401} (\bibinfo{year}{1980}).
1386: 
1387: \bibitem[{\citenamefont{Karpfen et~al.}(1981)\citenamefont{Karpfen, Beyer, and
1388:   Schuster}}]{Karpfen:HB-81}
1389: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Karpfen}},
1390:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Beyer}}, \bibnamefont{and}
1391:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Schuster}},
1392:   \bibinfo{journal}{Int. J. Quantum Chem.} \textbf{\bibinfo{volume}{19}},
1393:   \bibinfo{pages}{1113} (\bibinfo{year}{1981}).
1394: 
1395: \bibitem[{\citenamefont{Beyer and Karpfen}(1982)}]{Karpfen:AI-82}
1396: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Beyer}} \bibnamefont{and}
1397:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Karpfen}},
1398:   \bibinfo{journal}{Chem. Phys.} \textbf{\bibinfo{volume}{64}},
1399:   \bibinfo{pages}{343} (\bibinfo{year}{1982}).
1400: 
1401: \bibitem[{\citenamefont{I'haya et~al.}(1984)\citenamefont{I'haya, Narita,
1402:   Fujita, and Ujino}}]{Ihaya:AI-84}
1403: \bibinfo{author}{\bibfnamefont{Y.~J.} \bibnamefont{I'haya}},
1404:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Narita}},
1405:   \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Fujita}}, \bibnamefont{and}
1406:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Ujino}},
1407:   \bibinfo{journal}{Int. J. Quantum Chem. Symp.} \textbf{\bibinfo{volume}{18}},
1408:   \bibinfo{pages}{153} (\bibinfo{year}{1984}).
1409: 
1410: \bibitem[{\citenamefont{Liegener and Ladik}(1987)}]{Liegener:AI-87}
1411: \bibinfo{author}{\bibfnamefont{C.-M.} \bibnamefont{Liegener}} \bibnamefont{and}
1412:   \bibinfo{author}{\bibfnamefont{J.~J.} \bibnamefont{Ladik}},
1413:   \bibinfo{journal}{Phys. Rev.~B} \textbf{\bibinfo{volume}{35}},
1414:   \bibinfo{pages}{6403} (\bibinfo{year}{1987}).
1415: 
1416: \bibitem[{\citenamefont{Mayer et~al.}(1997)\citenamefont{Mayer, R\"ather, and
1417:   Suhai}}]{Mayer:CH-97}
1418: \bibinfo{author}{\bibfnamefont{I.}~\bibnamefont{Mayer}},
1419:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{R\"ather}}, \bibnamefont{and}
1420:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Suhai}},
1421:   \bibinfo{journal}{Chem. Phys. Lett.} \textbf{\bibinfo{volume}{270}},
1422:   \bibinfo{pages}{211} (\bibinfo{year}{1997}).
1423: 
1424: \bibitem[{\citenamefont{Berski and Latajka}(1997)}]{Berski:PHF-97}
1425: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Berski}} \bibnamefont{and}
1426:   \bibinfo{author}{\bibfnamefont{Z.}~\bibnamefont{Latajka}},
1427:   \bibinfo{journal}{J. Mol. Struct. (Theochem)} \textbf{\bibinfo{volume}{389}},
1428:   \bibinfo{pages}{147} (\bibinfo{year}{1997}).
1429: 
1430: \bibitem[{\citenamefont{Hirata and Iwata}(1998)}]{Hirata:AI-98}
1431: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Hirata}} \bibnamefont{and}
1432:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Iwata}}, \bibinfo{journal}{J.
1433:   Phys. Chem. A} \textbf{\bibinfo{volume}{102}}, \bibinfo{pages}{8426}
1434:   (\bibinfo{year}{1998}).
1435: 
1436: \bibitem[{\citenamefont{Jacquemin et~al.}(1999)\citenamefont{Jacquemin,
1437:   Andr\'e, and Champagne}}]{Jacquemin:LR-99}
1438: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Jacquemin}},
1439:   \bibinfo{author}{\bibfnamefont{J.-M.} \bibnamefont{Andr\'e}},
1440:   \bibnamefont{and}
1441:   \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Champagne}},
1442:   \bibinfo{journal}{J. Chem. Phys.} \textbf{\bibinfo{volume}{111}},
1443:   \bibinfo{pages}{5324} (\bibinfo{year}{1999}).
1444: 
1445: \bibitem[{\citenamefont{Buth and Paulus}(2004)}]{Buth:BS-04}
1446: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Buth}} \bibnamefont{and}
1447:   \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Paulus}},
1448:   \bibinfo{journal}{Chem. Phys. Lett.} \textbf{\bibinfo{volume}{398}},
1449:   \bibinfo{pages}{44} (\bibinfo{year}{2004}),
1450:   \bibinfo{note}{\href{http://de.arxiv.org/abs/cond-mat/0408243}
1451:   {arXiv:cond-mat/0408243}}.
1452: 
1453: \bibitem[{\citenamefont{Buth}(2005)}]{Buth:MC-05}
1454: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Buth}},
1455:   \bibinfo{type}{Dissertation}, \bibinfo{school}{Technische Universit\"at
1456:   Dresden}, \bibinfo{address}{01062~Dresden, Germany} (\bibinfo{year}{2005}),
1457:   \bibinfo{note}{\href{http://nbn-resolving.de/urn:nbn:de:swb:14-1132580113554%
1458: -34509} {nbn-resolving.de/urn:nbn:de:swb:14-1132580113554-34509}}.
1459: 
1460: \bibitem[{\citenamefont{Blumen and Merkel}(1977{\natexlab{b}})}]{Blumen:EBS-77}
1461: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Blumen}} \bibnamefont{and}
1462:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Merkel}},
1463:   \bibinfo{journal}{Chem. Phys. Lett.} \textbf{\bibinfo{volume}{45}},
1464:   \bibinfo{pages}{47} (\bibinfo{year}{1977}{\natexlab{b}}).
1465: 
1466: \bibitem[{\citenamefont{Berski and
1467:   Latajka}(1998{\natexlab{b}})}]{Berski:PHF-98}
1468: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Berski}} \bibnamefont{and}
1469:   \bibinfo{author}{\bibfnamefont{Z.}~\bibnamefont{Latajka}},
1470:   \bibinfo{journal}{Polish J. Chem.} \textbf{\bibinfo{volume}{72}},
1471:   \bibinfo{pages}{1540} (\bibinfo{year}{1998}{\natexlab{b}}).
1472: 
1473: \bibitem[{\citenamefont{Halkier et~al.}(1999)\citenamefont{Halkier, Klopper,
1474:   Helgaker, J\o{}rgensen, and Taylor}}]{Halkier:BS-99}
1475: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Halkier}},
1476:   \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Klopper}},
1477:   \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Helgaker}},
1478:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{J\o{}rgensen}},
1479:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{P.~R.}
1480:   \bibnamefont{Taylor}}, \bibinfo{journal}{J. Chem. Phys.}
1481:   \textbf{\bibinfo{volume}{111}}, \bibinfo{pages}{9157} (\bibinfo{year}{1999}).
1482: 
1483: \bibitem[{\citenamefont{Purvis~III and Bartlett}(1982)}]{Purvis:CC-82}
1484: \bibinfo{author}{\bibfnamefont{G.~D.} \bibnamefont{Purvis~III}}
1485:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{R.~J.}
1486:   \bibnamefont{Bartlett}}, \bibinfo{journal}{J. Chem. Phys.}
1487:   \textbf{\bibinfo{volume}{76}}, \bibinfo{pages}{1910} (\bibinfo{year}{1982}).
1488: 
1489: \bibitem[{\citenamefont{Stoll}(1992{\natexlab{a}})}]{Stoll:CD-92}
1490: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Stoll}},
1491:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{46}},
1492:   \bibinfo{pages}{6700} (\bibinfo{year}{1992}{\natexlab{a}}).
1493: 
1494: \bibitem[{\citenamefont{Stoll}(1992{\natexlab{b}})}]{Stoll:CS-92}
1495: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Stoll}},
1496:   \bibinfo{journal}{Chem. Phys. Lett.} \textbf{\bibinfo{volume}{191}},
1497:   \bibinfo{pages}{548} (\bibinfo{year}{1992}{\natexlab{b}}).
1498: 
1499: \bibitem[{\citenamefont{Stoll}(1992{\natexlab{c}})}]{Stoll:CG-92}
1500: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Stoll}}, \bibinfo{journal}{J.
1501:   Chem. Phys.} \textbf{\bibinfo{volume}{97}}, \bibinfo{pages}{8449}
1502:   (\bibinfo{year}{1992}{\natexlab{c}}).
1503: 
1504: \bibitem[{\citenamefont{Fulde}(1995)}]{Fulde:EC-95}
1505: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Fulde}},
1506:   \emph{\bibinfo{title}{Electron correlations in molecules and solids}}, vol.
1507:   \bibinfo{volume}{100} of \emph{\bibinfo{series}{Springer series in
1508:   solid-state sciences}} (\bibinfo{publisher}{Springer},
1509:   \bibinfo{address}{Berlin}, \bibinfo{year}{1995}), \bibinfo{edition}{3rd} ed.,
1510:   ISBN \bibinfo{isbn}{3-540-59364-0}.
1511: 
1512: \bibitem[{\citenamefont{Fulde}(2002)}]{Fulde:WF-02}
1513: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Fulde}}, \bibinfo{journal}{Adv.
1514:   Phys.} \textbf{\bibinfo{volume}{51}}, \bibinfo{pages}{909}
1515:   (\bibinfo{year}{2002}).
1516: 
1517: \bibitem[{\citenamefont{Yu et~al.}(1997)\citenamefont{Yu, Kalvoda, and
1518:   Dolg}}]{Yu:IA-97}
1519: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Yu}},
1520:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Kalvoda}}, \bibnamefont{and}
1521:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Dolg}},
1522:   \bibinfo{journal}{Chem. Phys.} \textbf{\bibinfo{volume}{224}},
1523:   \bibinfo{pages}{121} (\bibinfo{year}{1997}).
1524: 
1525: \bibitem[{\citenamefont{Abdurahman et~al.}(2000)\citenamefont{Abdurahman,
1526:   Shukla, and Dolg}}]{Abdurahman:AI-00}
1527: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Abdurahman}},
1528:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Shukla}}, \bibnamefont{and}
1529:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Dolg}}, \bibinfo{journal}{J.
1530:   Chem. Phys.} \textbf{\bibinfo{volume}{112}}, \bibinfo{pages}{4801}
1531:   (\bibinfo{year}{2000}).
1532: 
1533: \bibitem[{\citenamefont{Willnauer and Birkenheuer}(2004)}]{Willnauer:QC-04}
1534: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Willnauer}} \bibnamefont{and}
1535:   \bibinfo{author}{\bibfnamefont{U.}~\bibnamefont{Birkenheuer}},
1536:   \bibinfo{journal}{J. Chem. Phys.} \textbf{\bibinfo{volume}{120}},
1537:   \bibinfo{pages}{11910} (\bibinfo{year}{2004}).
1538: 
1539: \bibitem[{\citenamefont{Doll et~al.}(1995)\citenamefont{Doll, Dolg, Fulde, and
1540:   Stoll}}]{Doll:CE-95}
1541: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Doll}},
1542:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Dolg}},
1543:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Fulde}}, \bibnamefont{and}
1544:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Stoll}},
1545:   \bibinfo{journal}{Phys. Rev.~B} \textbf{\bibinfo{volume}{52}},
1546:   \bibinfo{pages}{4842} (\bibinfo{year}{1995}).
1547: 
1548: \bibitem[{\citenamefont{Doll et~al.}(1996)\citenamefont{Doll, Dolg, and
1549:   Stoll}}]{Doll:CE-96}
1550: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Doll}},
1551:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Dolg}}, \bibnamefont{and}
1552:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Stoll}},
1553:   \bibinfo{journal}{Phys. Rev.~B} \textbf{\bibinfo{volume}{54}},
1554:   \bibinfo{pages}{13529} (\bibinfo{year}{1996}).
1555: 
1556: \bibitem[{\citenamefont{Doll and Stoll}(1997)}]{Doll:CP-97}
1557: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Doll}} \bibnamefont{and}
1558:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Stoll}},
1559:   \bibinfo{journal}{Phys. Rev.~B} \textbf{\bibinfo{volume}{56}},
1560:   \bibinfo{pages}{10121} (\bibinfo{year}{1997}).
1561: 
1562: \bibitem[{\citenamefont{Doll and Stoll}(1998)}]{Doll:GS-98}
1563: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Doll}} \bibnamefont{and}
1564:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Stoll}},
1565:   \bibinfo{journal}{Phys. Rev.~B} \textbf{\bibinfo{volume}{57}},
1566:   \bibinfo{pages}{4327} (\bibinfo{year}{1998}).
1567: 
1568: \bibitem[{\citenamefont{Ro\'sciszewski
1569:   et~al.}(1999)\citenamefont{Ro\'sciszewski, Paulus, Fulde, and
1570:   Stoll}}]{Rosciszewski:AI-99}
1571: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Ro\'sciszewski}},
1572:   \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Paulus}},
1573:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Fulde}}, \bibnamefont{and}
1574:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Stoll}},
1575:   \bibinfo{journal}{Phys. Rev.~B} \textbf{\bibinfo{volume}{60}},
1576:   \bibinfo{pages}{7905} (\bibinfo{year}{1999}).
1577: 
1578: \bibitem[{\citenamefont{Ro\'sciszewski
1579:   et~al.}(2000)\citenamefont{Ro\'sciszewski, Paulus, Fulde, and
1580:   Stoll}}]{Rosciszewski:AI-00}
1581: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Ro\'sciszewski}},
1582:   \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Paulus}},
1583:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Fulde}}, \bibnamefont{and}
1584:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Stoll}},
1585:   \bibinfo{journal}{Phys. Rev.~B} \textbf{\bibinfo{volume}{62}},
1586:   \bibinfo{pages}{5482} (\bibinfo{year}{2000}).
1587: 
1588: \bibitem[{\citenamefont{Szabo and Ostlund}(1989)}]{Szabo:MQC-89}
1589: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Szabo}} \bibnamefont{and}
1590:   \bibinfo{author}{\bibfnamefont{N.~S.} \bibnamefont{Ostlund}},
1591:   \emph{\bibinfo{title}{Modern quantum chemistry: Introduction to advanced
1592:   electronic structure theory}} (\bibinfo{publisher}{McGraw-Hill},
1593:   \bibinfo{address}{New York}, \bibinfo{year}{1989}), \bibinfo{edition}{{1st,
1594:   revised}} ed., ISBN \bibinfo{isbn}{0-486-69186-1}.
1595: 
1596: \bibitem[{\citenamefont{Ladik}(1999)}]{Ladik:PS-99}
1597: \bibinfo{author}{\bibfnamefont{J.~J.} \bibnamefont{Ladik}},
1598:   \bibinfo{journal}{Phys. Rep.} \textbf{\bibinfo{volume}{313}},
1599:   \bibinfo{pages}{171} (\bibinfo{year}{1999}).
1600: 
1601: \bibitem[{\citenamefont{F\"orner}(1992)}]{Forner:LC-92}
1602: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{F\"orner}},
1603:   \bibinfo{journal}{Int. J. Quantum Chem.} \textbf{\bibinfo{volume}{43}},
1604:   \bibinfo{pages}{221} (\bibinfo{year}{1992}).
1605: 
1606: \bibitem[{\citenamefont{Hirata et~al.}(2004)\citenamefont{Hirata, Podeszwa,
1607:   Tobita, and Bartlett}}]{Hirata:CC-03}
1608: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Hirata}},
1609:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Podeszwa}},
1610:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Tobita}}, \bibnamefont{and}
1611:   \bibinfo{author}{\bibfnamefont{R.~J.} \bibnamefont{Bartlett}},
1612:   \bibinfo{journal}{J. Chem. Phys.} \textbf{\bibinfo{volume}{120}},
1613:   \bibinfo{pages}{2581} (\bibinfo{year}{2004}).
1614: 
1615: \bibitem[{\citenamefont{Wiberg et~al.}(2001)\citenamefont{Wiberg, Holleman, and
1616:   Wiberg}}]{Holleman:IC-01}
1617: \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Wiberg}},
1618:   \bibinfo{author}{\bibfnamefont{A.~F.} \bibnamefont{Holleman}},
1619:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Wiberg}},
1620:   \emph{\bibinfo{title}{Inorganic chemistry}} (\bibinfo{publisher}{Academic
1621:   Press}, \bibinfo{address}{New York}, \bibinfo{year}{2001}), ISBN
1622:   \bibinfo{isbn}{0-123-52651-5}.
1623: 
1624: \bibitem[{\citenamefont{Amos et~al.}(2002)\citenamefont{Amos, Bernhardsson,
1625:   Berning, Celani, Cooper, Deegan, Dobbyn, Eckert, Hampel, Hetzer
1626:   et~al.}}]{MOLPRO2002.6}
1627: \bibinfo{author}{\bibfnamefont{R.~D.} \bibnamefont{Amos}},
1628:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Bernhardsson}},
1629:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Berning}},
1630:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Celani}},
1631:   \bibinfo{author}{\bibfnamefont{D.~L.} \bibnamefont{Cooper}},
1632:   \bibinfo{author}{\bibfnamefont{M.~J.~O.} \bibnamefont{Deegan}},
1633:   \bibinfo{author}{\bibfnamefont{A.~J.} \bibnamefont{Dobbyn}},
1634:   \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Eckert}},
1635:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Hampel}},
1636:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Hetzer}},
1637:   \bibnamefont{et~al.}, \emph{\bibinfo{title}{\textsc{molpro}, a package of
1638:   \emph{ab initio} programs designed by {H.-J.~Werner} and {P.~J. Knowles},
1639:   version 2002.6}} (\bibinfo{year}{2002}).
1640: 
1641: \bibitem[{\citenamefont{Boys}(1960)}]{Boys:MO-60}
1642: \bibinfo{author}{\bibfnamefont{S.~F.} \bibnamefont{Boys}},
1643:   \bibinfo{journal}{Rev. Mod. Phys.} \textbf{\bibinfo{volume}{32}},
1644:   \bibinfo{pages}{296} (\bibinfo{year}{1960}).
1645: 
1646: \bibitem[{\citenamefont{Foster and Boys}(1960)}]{Foster:MO-60}
1647: \bibinfo{author}{\bibfnamefont{J.~M.} \bibnamefont{Foster}} \bibnamefont{and}
1648:   \bibinfo{author}{\bibfnamefont{S.~F.} \bibnamefont{Boys}},
1649:   \bibinfo{journal}{Rev. Mod. Phys.} \textbf{\bibinfo{volume}{32}},
1650:   \bibinfo{pages}{300} (\bibinfo{year}{1960}).
1651: 
1652: \bibitem[{\citenamefont{Hampel et~al.}(1992)\citenamefont{Hampel, Peterson, and
1653:   Werner}}]{Hampel:CE-92}
1654: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Hampel}},
1655:   \bibinfo{author}{\bibfnamefont{K.~A.} \bibnamefont{Peterson}},
1656:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{H.-J.}
1657:   \bibnamefont{Werner}}, \bibinfo{journal}{Chem. Phys. Lett.}
1658:   \textbf{\bibinfo{volume}{190}}, \bibinfo{pages}{1} (\bibinfo{year}{1992}).
1659: 
1660: \bibitem[{\citenamefont{{Dunning, Jr.}}(1989)}]{Dunning:GBS-89}
1661: \bibinfo{author}{\bibfnamefont{T.~H.} \bibnamefont{{Dunning, Jr.}}},
1662:   \bibinfo{journal}{J. Chem. Phys.} \textbf{\bibinfo{volume}{90}},
1663:   \bibinfo{pages}{1007} (\bibinfo{year}{1989}).
1664: 
1665: \bibitem[{\citenamefont{Kendall et~al.}(1992)\citenamefont{Kendall, {Dunning,
1666:   Jr.}, and Harrison}}]{Kendall:EA-92}
1667: \bibinfo{author}{\bibfnamefont{R.~A.} \bibnamefont{Kendall}},
1668:   \bibinfo{author}{\bibfnamefont{T.~H.} \bibnamefont{{Dunning, Jr.}}},
1669:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{R.~J.}
1670:   \bibnamefont{Harrison}}, \bibinfo{journal}{J. Chem. Phys.}
1671:   \textbf{\bibinfo{volume}{96}}, \bibinfo{pages}{6796} (\bibinfo{year}{1992}).
1672: 
1673: \bibitem[{\citenamefont{Woon and {Dunning, Jr.}}(1993)}]{Woon:GBS-93}
1674: \bibinfo{author}{\bibfnamefont{D.~E.} \bibnamefont{Woon}} \bibnamefont{and}
1675:   \bibinfo{author}{\bibfnamefont{T.~H.} \bibnamefont{{Dunning, Jr.}}},
1676:   \bibinfo{journal}{J. Chem. Phys.} \textbf{\bibinfo{volume}{98}},
1677:   \bibinfo{pages}{1358} (\bibinfo{year}{1993}).
1678: 
1679: \bibitem[{bas()}]{basislib-04}
1680: \bibinfo{note}{Basis sets were obtained from the Extensible Computational
1681:   Chemistry Environment Basis Set Database, Version 02/25/04, as developed and
1682:   distributed by the Molecular Science Computing Facility, Environmental and
1683:   Molecular Sciences Laboratory which is part of the Pacific Northwest
1684:   Laboratory, P.O.~Box~999, Richland, Washington~99352, USA, and funded by the
1685:   U.S.~Department of Energy. The Pacific Northwest Laboratory is a
1686:   multi-program laboratory operated by Battelle Memorial Institute for the
1687:   U.S.~Department of Energy under contract DE-AC06-76RLO 1830. Contact David
1688:   Feller or Karen Schuchardt for further information.}
1689: 
1690: \bibitem[{\citenamefont{Pisani et~al.}(1988)\citenamefont{Pisani, Dosevi, and
1691:   Roetti}}]{Pisani:HF-88}
1692: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Pisani}},
1693:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Dosevi}}, \bibnamefont{and}
1694:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Roetti}},
1695:   \emph{\bibinfo{title}{Hartree-Fock ab initio treatment of crystalline
1696:   systems}}, vol.~\bibinfo{volume}{48} of \emph{\bibinfo{series}{Lecture notes
1697:   in chemistry}} (\bibinfo{publisher}{Springer}, \bibinfo{address}{Berlin,
1698:   Heidelberg}, \bibinfo{year}{1988}), ISBN \bibinfo{isbn}{3-540-19317-0}.
1699: 
1700: \bibitem[{\citenamefont{Pisani}(1996)}]{Pisani:QM-96}
1701: \bibinfo{editor}{\bibfnamefont{C.}~\bibnamefont{Pisani}}, ed.,
1702:   \emph{\bibinfo{title}{Quantum-mechanical ab initio calculation of the
1703:   properties of crystalline materials}}, vol.~\bibinfo{volume}{67} of
1704:   \emph{\bibinfo{series}{Lecture notes in chemistry}}
1705:   (\bibinfo{publisher}{Springer}, \bibinfo{address}{Berlin, Heidelberg},
1706:   \bibinfo{year}{1996}), ISBN \bibinfo{isbn}{3-540-61645-4}.
1707: 
1708: \bibitem[{\citenamefont{Saunders et~al.}(2003)\citenamefont{Saunders, Dovesi,
1709:   Roetti, Orlando, Zicovich-Wilson, Harrison, Doll, Civalleri, Bush, D'Arco
1710:   et~al.}}]{crystal03}
1711: \bibinfo{author}{\bibfnamefont{V.~R.} \bibnamefont{Saunders}},
1712:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Dovesi}},
1713:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Roetti}},
1714:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Orlando}},
1715:   \bibinfo{author}{\bibfnamefont{C.~M.} \bibnamefont{Zicovich-Wilson}},
1716:   \bibinfo{author}{\bibfnamefont{N.~M.} \bibnamefont{Harrison}},
1717:   \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Doll}},
1718:   \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Civalleri}},
1719:   \bibinfo{author}{\bibfnamefont{I.~J.} \bibnamefont{Bush}},
1720:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{D'Arco}},
1721:   \bibnamefont{et~al.}, \emph{\bibinfo{title}{\textsc{crystal03} User's
1722:   Manual}}, \bibinfo{organization}{University of Torino},
1723:   \bibinfo{address}{Torino, Italy} (\bibinfo{year}{2003}),
1724:   \bibinfo{note}{\href{http://www.crystal.unito.it} {www.crystal.unito.it}}.
1725: 
1726: \bibitem[{\citenamefont{{van Duijneveldt} et~al.}(1994)\citenamefont{{van
1727:   Duijneveldt}, {van Duijneveldt-van de Rijdt}, and {van
1728:   Lenthe}}}]{Duijneveldt:CP-94}
1729: \bibinfo{author}{\bibfnamefont{F.~B.} \bibnamefont{{van Duijneveldt}}},
1730:   \bibinfo{author}{\bibfnamefont{J.~G. C.~M.} \bibnamefont{{van Duijneveldt-van
1731:   de Rijdt}}}, \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.~H.}
1732:   \bibnamefont{{van Lenthe}}}, \bibinfo{journal}{Chem. Rev.}
1733:   \textbf{\bibinfo{volume}{94}}, \bibinfo{pages}{1873} (\bibinfo{year}{1994}).
1734: 
1735: \bibitem[{\citenamefont{Boys and Bernardi}(1970)}]{Boys:CP-70}
1736: \bibinfo{author}{\bibfnamefont{S.~F.} \bibnamefont{Boys}} \bibnamefont{and}
1737:   \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Bernardi}},
1738:   \bibinfo{journal}{Mol. Phys.} \textbf{\bibinfo{volume}{19}},
1739:   \bibinfo{pages}{553} (\bibinfo{year}{1970}).
1740: 
1741: \bibitem[{\citenamefont{Evjen}(1932)}]{Evjen:OS-32}
1742: \bibinfo{author}{\bibfnamefont{H.~M.} \bibnamefont{Evjen}},
1743:   \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{39}},
1744:   \bibinfo{pages}{675} (\bibinfo{year}{1932}).
1745: 
1746: \bibitem[{\citenamefont{Pisani}(2003)}]{Pisani:LT-03}
1747: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Pisani}}, \bibinfo{journal}{J.
1748:   Mol. Struct. (THEOCHEM)} \textbf{\bibinfo{volume}{621}}, \bibinfo{pages}{141}
1749:   (\bibinfo{year}{2003}).
1750: 
1751: \bibitem[{\citenamefont{Craig and Thirunamachandran}(1984)}]{Craig:MQ-84}
1752: \bibinfo{author}{\bibfnamefont{D.~P.} \bibnamefont{Craig}} \bibnamefont{and}
1753:   \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Thirunamachandran}},
1754:   \emph{\bibinfo{title}{Molecular quantum electrodynamics}}
1755:   (\bibinfo{publisher}{Academic Press}, \bibinfo{address}{London},
1756:   \bibinfo{year}{1984}), ISBN \bibinfo{isbn}{0-486-40214-2}.
1757: 
1758: \bibitem[{\citenamefont{Kumar and Meath}(1985)}]{Kumar:PS-85}
1759: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Kumar}} \bibnamefont{and}
1760:   \bibinfo{author}{\bibfnamefont{W.~J.} \bibnamefont{Meath}},
1761:   \bibinfo{journal}{Mol. Phys.} \textbf{\bibinfo{volume}{54}},
1762:   \bibinfo{pages}{823} (\bibinfo{year}{1985}).
1763: 
1764: \bibitem[{\citenamefont{Klein and Venables}(1976)}]{Klein:RG-76}
1765: \bibinfo{editor}{\bibfnamefont{M.~L.} \bibnamefont{Klein}} \bibnamefont{and}
1766:   \bibinfo{editor}{\bibfnamefont{J.~A.} \bibnamefont{Venables}}, eds.,
1767:   \emph{\bibinfo{title}{Rare gas solids}}, vol.~\bibinfo{volume}{1}
1768:   (\bibinfo{publisher}{Academic Press}, \bibinfo{address}{London},
1769:   \bibinfo{year}{1976}), ISBN \bibinfo{isbn}{0-12-413501-3}.
1770: 
1771: \bibitem[{\citenamefont{Mor\'e}(1977)}]{More:LM-78}
1772: \bibinfo{author}{\bibfnamefont{J.~J.} \bibnamefont{Mor\'e}}, in
1773:   \emph{\bibinfo{booktitle}{Numerical analysis}}, edited by
1774:   \bibinfo{editor}{\bibfnamefont{G.~A.} \bibnamefont{Watson}}
1775:   (\bibinfo{publisher}{Springer-Verlag}, \bibinfo{address}{Berlin, New York},
1776:   \bibinfo{year}{1977}), vol. \bibinfo{volume}{630} of
1777:   \emph{\bibinfo{series}{Lecture notes in mathematics}}, pp.
1778:   \bibinfo{pages}{105--116}.
1779: 
1780: \bibitem[{\citenamefont{Axilrod and Teller}(1943)}]{Axilrod:IV-43}
1781: \bibinfo{author}{\bibfnamefont{B.~M.} \bibnamefont{Axilrod}} \bibnamefont{and}
1782:   \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Teller}},
1783:   \bibinfo{journal}{J. Chem. Phys.} \textbf{\bibinfo{volume}{11}},
1784:   \bibinfo{pages}{299} (\bibinfo{year}{1943}).
1785: 
1786: \bibitem[{\citenamefont{Fink and Staemmler}(1995)}]{Fink:AI-95}
1787: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Fink}} \bibnamefont{and}
1788:   \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{Staemmler}},
1789:   \bibinfo{journal}{J. Chem. Phys.} \textbf{\bibinfo{volume}{103}},
1790:   \bibinfo{pages}{2603} (\bibinfo{year}{1995}).
1791: \end{thebibliography}
1792: \end{document}
1793: