1: %\documentclass[twocolumn,showpacs,preprintnumbers,amsmath,amssymb,epsf]{revtex4}
2: %\documentclass[floatfix,showkeys,showpacs,twocolumn,amsmath,amssymb,epsf,aps]{revtex4}
3: \documentclass[floatfix,showkeys,twocolumn,citeautoscript,showpacs,amsmath,amssymb,epsf,aps]{revtex4}
4: \usepackage{dcolumn}
5: \usepackage{epsfig}
6: \renewcommand{\baselinestretch}{1.4}
7:
8: \usepackage{graphicx}% Include figure files
9: \usepackage{longtable}% Include figure files
10: \usepackage{dcolumn}% Align table columns on decimal point
11: \usepackage{bm}% bold math
12:
13: \def\etal{{\em et al. }}
14: %\def\BN #1{B${_#1}$N${_#1}$}
15: \def\BNF{B$_{24}$N$_{24}$ }
16: \def\ocm{cm$^{-1}$ }
17:
18: \def\BNE{B$_{28}$N$_{28}$ }
19: \def\BNT{B$_{32}$N$_{32}$ }
20: \def\BNS{B$_{36}$N$_{36}$ }
21: \def\BNR{B$_{96}$N$_{96}$ }
22:
23:
24: \begin{document}
25: \preprint{USC/002}
26:
27: \title{ Theoretical infra-red, Raman, and Optical spectra of the \BNS cage}
28:
29: \author{Rajendra R. Zope$^{1}$, Tunna Baruah$^{2,3}$, Mark R. Pederson${^2}$, and Brett I. Dunlap$^4$}
30:
31: \affiliation{$^1$Department of Chemistry, George Washington University, Washington DC, 20052, USA}
32:
33:
34: \affiliation{$^2$Code 6392, Center for Computational Materials Science, US Naval Research Laboratory,
35: Washington DC 20375, USA}
36:
37: \affiliation{$^3$Department of Electrical Engineering and Materials Science Research Center,
38: Howard University, Washington DC 200059, USA}
39:
40:
41: \affiliation{$^4$Code 6189, Theoretical Chemistry Section, US Naval Research Laboratory,
42: Washington, DC 20375}
43:
44: \date{\today}
45:
46: \begin{abstract}
47: The \BNS fullerene-like cage structure was proposed as candidate structure for the single-shell
48: boron-nitride cages observed in electron-beam irradiation experiment. We have performed all
49: electron density functional calculations, with large polarized Gaussian basis sets, on the \BNS
50: cage. We show that the cage is energetically and vibrationally stable. The infra-red,
51: Raman and optical spectra are calculated. The predicted spectra, in combination with experimentally
52: measured spectra, will be useful in conclusive assignment of the proposed \BNS cage.
53: The vertical and adiabatic ionization potentials as well as static dipole polarizability
54: are also reported.
55: \end{abstract}
56:
57:
58: \pacs{ }
59:
60: %Use showkeys class option if keyword display desired
61: \keywords{ boron nitride, nanotubes, fullerene}
62:
63: \maketitle
64: The discovery of carbon fullerenes\cite{KHOCS85}, the synthesis of
65: boron nitride (BN) nanotubes\cite{Zettl}, and the fact that the BN pair
66: is isoelectronic with a
67: pair carbon atoms, led to the search for fullerene analogues of BN. The carbon
68: fullerenes are made close by introducing 12 pentagons in the hexagonal network
69: of carbon atoms. The exact BN analogues of fullerenes are not preferred as the presence of
70: pentagonal ring does not permit alternate sequence of B and N atoms. However, by
71: Euler's theorem, it can be shown that fullerene-like alternate BN cages can
72: be formed using six isolated squares\cite{SSL95,ZSK97}. The square contains
73: four BN bonds with alternate boron and nitrogen atoms.
74: Several theoretical works have reported the possible candidate
75: structures for the BN cages\cite{SSL95, ZSK97, Strout,P2000,WJ04,ZD04,ZBPD04,Oku00}.
76:
77: On the experimental side, there are also a few reports of synthesis of hollow BN
78: structures\cite{Stephan98,Goldberg98,Oku00,Oku03,Oku04}. In a series of experiments
79: by Oku and coorkwers, boron nitride clusters of different sizes were
80: reported\cite{Oku00,Oku03,Oku04}. Perhaps the most interesting result of these works was
81: the production of \BNF in abundance. We performed theoretical calculations\cite{ZD04,ZBPD04}
82: on \BNF, and think that the roundness of the octahedral \BNF cluster is the most likely
83: explanation of its unexpected abundance.
84:
85: In another set of experiments,
86: St\'ephan \etal irradiated BN-powder and BN material\cite{Stephan98} and
87: observed small BN cage-like molecules. The irradiated derivatives were either
88: close-packed cages or nested cages. The most observed cages in the experiment
89: had diameters in the size range 4 to 7 \AA; the octahedral B$_{12}$N$_{12}$, B$_{16}$N$_{16}$,
90: and \BNE cages were proposed as possible structures.
91: Subsequent tilting experiment in an electron microscope
92: by Goldberg and coworkers permitted viewing of the irradiation-induced BN cages in
93: different directions and corroborated the proposal of octahedral structures for the observed
94: BN cages\cite{Goldberg98}. From the high resolution transmission electron
95: microscopy images (HRTEM), they observed single-shelled cages of size 9-10 \AA,
96: with rectangle-like outlines. The \BNS cage has roughly similar dimension and was proposed
97: as candidate structure by Alexandre \etal \cite{AMC99} and Oku \etal \cite{Oku00}.
98: Alexander \etal also performed a finite-range pseudopotential density functional calculation
99: on the \BNS and showed that the \BNS cage is energetically stable
100: and has large energy gap between its highest molecular orbital and lowest molecular orbital
101: (HOMO-LUMO). These conclusions were subsequently confirmed by all electron calculations using
102: the Slater-Roothan (SR) method\cite{ZD04}. These calculations show that the proposed \BNS
103: cage has overall size of about 8 \AA, smaller than single-shelled BN cages observed
104: in the experiment.
105: So far the proposal of \BNS as the observed structure has not yet been
106: verified in alternative experiments. It is the purpose of the present work to ascertain
107: the stability of proposed cage structure of \BNS with respect to small distortions and to
108: provide its spectroscopic data : the infra-red (IR) spectra, Raman and optical spectra, for
109: comparison with experimental spectra. The comparison of the calculated and experimentally
110: measured spectra has played an important role in structure
111: determination in past. We hope that the present work will stimulate further experimental works,
112: which combined with the predicted spectra, would unambiguously confirm the
113: proposal of \BNS fullerene-like cage.
114:
115:
116: %\input{b36n36_xyz}
117:
118: \begin{table}
119: \caption{The optimized coordinates of ineqivalent atoms in \BNS.}
120: \begin{tabular}{lccc}
121: \hline
122: Atom & X & Y & Z \\
123: \hline
124: B & -1.340 & 4.713 & 4.713 \\
125: B & -7.771 & -1.224 & -1.224 \\
126: B & 6.099 & 2.739 & 2.739 \\
127: N & 1.457 & 4.876 & 4.876 \\
128: N & 8.189 & 1.504 & 1.504 \\
129: N & -6.057 & -2.702 & -2.702 \\
130: \hline
131: \end{tabular}
132: \label{tab:xyz}
133: \end{table}
134:
135:
136:
137: Our calculations were carried out at the all electron level using the { NRLMOL} suite of
138: codes \cite{NRLMOL}. The exchange-correlation effects were treated at the level of the
139: generalized gradient approximation using the Perdew-Burke-Ernzerhof scheme(PBE-GGA)\cite{PBE}.
140: Large polarized Gaussian basis sets\cite{Porezag99} optimized for density functional
141: calculations (PBE-GGA) are employed. The B-atom basis set consists of 5 $s-$type,
142: 4 $p-$type, and 3 $d-$type orbitals each of which were constructed from a linear combination of
143: 12 primitive Gaussians. For N atoms, 5 $s-$type, 4 $p-$type, and 3 $d-$type functions comprised
144: of 13 Gaussian are used. The geometry optimization was performed using the limited-memory
145: Broyden-Fletcher-Goldfarb-Shanno scheme until the forces on each atom were less
146: than 0.001 a.u. The self-consistent calculation was iterated until the energy
147: difference of successive iteration was smaller 0.0000001 a.u. The vibrational
148: frequencies were calculated by diagonalizing a dynamical matrix constructed by displacing
149: the atoms by small amount\cite{Porezag96}. The IR and Raman frequencies were determined
150: from the derivative of the dipole moment and the polarizability tensor.
151:
152:
153: \begin{figure}
154: \epsfig{file=B36N36_AB.eps,width=8.5cm,clip=true}
155: \caption{\label{fig1} (Color online) Two different views of optimized
156: \BNS cages. In the second view the \BNS cage is rotated by 45$^o$ around a vertical
157: axis passing through center of square.}
158: \end{figure}
159:
160: \begin{figure}
161: \epsfig{file=ir_raman.eps,width=8.5cm,clip=true}
162: \caption{\label{fig:ir} (Color online) The infra-red (top) and Raman (lower panel) spectrum of \BNS cage.}
163: \end{figure}
164:
165: %\begin{figure}
166: %\epsfig{file=raman.eps,width=8.5cm,clip=true}
167: %\caption{\label{fig:raman} (Color online) Raman spectrum of \BNS cage.}
168: %\end{figure}
169:
170: The optimization of \BNS cage (See Fig. \ref{fig1}) was started using its optimized geometry obtained by
171: the SR method\cite{ZD04}.
172: % The optimized \BNS cage is shown in Figure \ref{fig1}.
173: The cage has
174: $T_d$ symmetry and can be generated from six inequivalent atoms, three for boron, and
175: three for nitrogen, using symmetry operations. The optimized coordinates of the
176: inequivalent atoms are given in Table \ref{tab:xyz}. These coordinates are with respect
177: to the origin at the center of mass. The average spherical radius of
178: the cage, obtained as the mean of distance of each atom from the center of mass, is
179: 3.94 \AA. The SR method predicts a cage of 3.76 \AA \, radius\cite{ZD04}.
180: Thus, the cage has roughly dimension of 8 \AA, somewhat smaller than the
181: observed dimensions of cage structure in the experiment. Part of this discrepancy could
182: be due to the thermal effects as our calculations are at $T=0K$.
183: The cage is energetically
184: stable with respect to isolated atoms and has a binding energy (BE) of 8.43 eV per
185: atom.
186: The SR method gives a binding energy of 7.53 eV/atom using the 6-311G* basis.
187: Th present BE is higher than that of \BNF cages, which have energies of about
188: 8.3 eV\cite{ZBPD04}. This result agrees with our predictions by
189: SR method\cite{ZD04}.
190: The trend of larger BN cages being energetically more stable
191: than smaller ones is similar to that observed in case of carbon fullerenes\cite{D91}.
192: The present calculations, consistent with earlier predictions by
193: lower theories, show that the \BNS cage is characterized by a large HOMO-LUMO gap of 5.0 eV.
194: The HOMO level has t$_2$ symmetry.
195: The vertical ionization potential (VIP) is calculated as the self-consistent energy
196: differences of the cage and its positive ion with the same ionic configurations.
197: The calculated VIP is 8.2 eV. This agrees well with 8.4 eV, obtained by the SR method\cite{ZD04}.
198: We have also determined the adiabatic ionization potential
199: from the energy differences of the neutral \BNS cage and its optimized singly charged cation.
200: The adiabatic ionization potential is 8.1 eV, which is an overestimate as
201: the cation was optimized under symmetry constraint.
202: The mean polarizability of \BNS calculated from the finite field method is 78 \AA$^3$.
203: The nuclear frame of \BNS was assumed to be frozen during the polarizability calculation.
204:
205:
206:
207:
208:
209: The vibrational analysis within the harmonic approximation indicate that all
210: frequencies are real. Thus, the \BNS cage is vibrationally stable and corresponds
211: to a minimum on the potential surface. The calculated IR spectrum is presented
212: in Fig. ~\ref{fig:ir}. The IR spectrum is broadened by 6 \ocm \!\!. The IR spectra
213: shows two conspicuous peaks at 1432 \ocm and 1456 \ocm ; the intensity of the
214: latter peak (42 Debye$^2$/amu/\AA$^2$) is half of the former peak. A weak
215: but noticeable absorption (intensity 6 Debye$^2$/amu/\AA$^2$) is found at 797 \ocm.
216: All the peaks have $T_2$ symmetry. The Raman spectrum
217: is also shown in Fig. ~\ref{fig:ir}. The Raman spectrum is characterized by multiple
218: peaks in the frequency range $150-900$ \ocm. Of the two prominent peaks in the low frequency
219: range, the first one is at 165 \ocm and is of $E$ type. The second one at 188 \ocm
220: is due to the triply degenerate mode of $T_2$ symmetry. All other Raman active frequencies:
221: at 354, 439, 796, 847, and 869 \ocm are of $A_1$ symmetry. The absorption at 354 \ocm
222: corresponds to symmetric stretching or breathing mode. All vibrational frequencies
223: along with their symmetry type are listed in Table \ref{vibs}.
224:
225:
226: \begin{table*}[h]
227: \caption{The frequencies, symmetry, IR and Raman activeness of the vibrational modes.}
228: \label{vibs}
229: \begin{ruledtabular}
230: \begin{tabular}{lccccccc}
231: Frequency & Symmetry & IR & Raman & \hskip 0.7in Frequency & Symmetry & IR & Raman \\
232: (\ocm) & &active & active & \hskip 0.7in (\ocm) & & active & active \\
233: \hline
234: 164 & E & & $\surd$ & \hskip 0.7in 796 & T$_2$ &$\surd$ & $\surd$ \\
235: 187 & T$_2$ &$\surd$ & $\surd$ & \hskip 0.7in 806 & A$_2$ & & \\
236: 208 & A$_1$ & & $\surd$ & \hskip 0.7in 829 & A$_1$ & & $\surd$ \\
237: 222 & T$_1$ & & & \hskip 0.7in 847 & T$_2$ &$\surd$ & $\surd$ \\
238: 245 & T$_2$ &$\surd$ & $\surd$ & \hskip 0.7in 860 & A$_1$ & & $\surd$ \\
239: 269 & T$_1$ & & & \hskip 0.7in 905 & T$_1$ & & \\
240: 282 & E & & $\surd$ & \hskip 0.7in 933 & E & & $\surd$ \\
241: 293 & T$_1$ & & & \hskip 0.7in 953 & T$_2$ &$\surd$ & $\surd$ \\
242: 301 & T$_2$ &$\surd$ & $\surd$ & \hskip 0.7in 966 & E & & $\surd$ \\
243: 335 & E & & $\surd$ & \hskip 0.7in 967 & T$_1$ & & \\
244: 346 & T$_2$ &$\surd$ & $\surd$ & \hskip 0.7in 976 & A$_2$ & & \\
245: 354 & A$_1$ & & $\surd$ & \hskip 0.7in 979 & T$_1$ & & \\
246: 360 & E & & $\surd$ & \hskip 0.7in 1006 & T$_2$ &$\surd$ & $\surd$ \\
247: 369 & T$_1$ & & & \hskip 0.7in 1029 & A$_1$ & & $\surd$ \\
248: 374 & T$_2$ &$\surd$ & $\surd$ & \hskip 0.7in 1061 & T$_2$ &$\surd$ & $\surd$ \\
249: 419 & T$_1$ & & & \hskip 0.7in 1078 & E & & $\surd$ \\
250: 437 & T$_2$ &$\surd$ & $\surd$ & \hskip 0.7in 1140 & T$_1$ & & \\
251: 438 & E & & $\surd$ & \hskip 0.7in 1143 & T$_2$ &$\surd$ & \\
252: 462 & A$_2$ & & & \hskip 0.7in 1154 & E & & $\surd$ \\
253: 480 & T$_2$ &$\surd$ & $\surd$ & \hskip 0.7in 1157 & T$_2$ &$\surd$ & $\surd$ \\
254: 513 & T$_1$ & & & \hskip 0.7in 1159 & T$_1$ & & \\
255: 528 & T$_2$ &$\surd$ & $\surd$ & \hskip 0.7in 1163 & A$_2$ & & \\
256: 558 & T$_1$ & & & \hskip 0.7in 1166 & T$_1$ & & \\
257: 573 & E & & $\surd$ & \hskip 0.7in 1175 & A$_2$ & & \\
258: 589 & T$_2$ &$\surd$ & $\surd$ & \hskip 0.7in 1189 & A$_1$ & & $\surd$ \\
259: 592 & T$_1$ & & & \hskip 0.7in 1198 & T$_1$ & & \\
260: 605 & T$_2$ &$\surd$ & $\surd$ & \hskip 0.7in 1206 & E & &$\surd$ \\
261: 612 & E & & $\surd$ & \hskip 0.7in 1207 & T$_2$ &$\surd$ & $\surd$ \\
262: 636 & T$_2$ &$\surd$ &$\surd$ & \hskip 0.7in 1232 & T$_2$ &$\surd$ & $\surd$ \\
263: 637 & T$_1$ & & & \hskip 0.7in 1237 & T$_1$ & & \\
264: 640 & E & &$\surd$ & \hskip 0.7in 1271 & T$_1$ & & \\
265: 653 & T$_2$ &$\surd$ &$\surd$ & \hskip 0.7in 1288 & T$_2$ &$\surd$ & $\surd$ \\
266: 664 & E & &$\surd$ & \hskip 0.7in 1299 & T$_1$ & & \\
267: 675 & A$_1$ & &$\surd$ & \hskip 0.7in 1304 & E & & $\surd$ \\
268: 677 & T$_1$ & & & \hskip 0.7in 1356 & T$_1$ & & \\
269: 710 & T$_2$ &$\surd$ &$\surd$ & \hskip 0.7in 1356 & T$_2$ &$\surd$ & $\surd$ \\
270: 714 & E & &$\surd$ & \hskip 0.7in 1365 & E & & $\surd$ \\
271: 715 & A$_1$ & & $\surd$ & \hskip 0.7in 1395 & A$_1$ & & $\surd$ \\
272: 723 & T$_1$ & & & \hskip 0.7in 1420 & T$_2$ &$\surd$ & $\surd$ \\
273: 726 & T$_2$ &$\surd$ &$\surd$ & \hskip 0.7in 1432 & T$_2$ &$\surd$ & $\surd$ \\
274: 740 & T$_2$ &$\surd$ &$\surd$ & \hskip 0.7in 1455 & T$_2$ &$\surd$ & $\surd$ \\
275: 787 & T$_1$ & & & \hskip 0.7in 1456 & E & & $\surd$ \\
276: 792 & E & &$\surd$ & \hskip 0.7in 1464 & A$_1$ & & $\surd$ \\
277: 795 & A$_1$ & &$\surd$ & \hskip 0.7in & & & \\
278: \end{tabular}
279: \end{ruledtabular}
280: \end{table*}
281:
282: The density of states (DOS) and optical spectra of the \BNS cluster
283: calculated within DFT are shown in Fig. \ref{optical}.
284: The HOMO-LUMO gap of this cluster is 5.0 eV. The HOMO has a predominant N character whereas the
285: LUMO is delocalized over the whole cluster. The states near the HOMO and the LUMO have more N character
286: than B character. The HOMO has $t_2$ symmetry and the LUMO has $a_1$ symmetry.
287: The HOMO-LUMO transition is weak though symmetry allowed.
288: The optical spectra shows a sharp peak at 5.4 eV which occurs from transitions from occupied states
289: within 0.59 eV of HOMO to the unoccupied states within 0.43 eV of LUMO. This peak has contributions
290: from the $t_2 \rightarrow t_2$ and $e \rightarrow t_2$ transitions. It also has weak contributions from
291: $t_2 \rightarrow e$ transitions.
292: There are other higher energy peaks seen at 9.7 eV and 13.8 eV. The small peak at 9.7 eV arises
293: mainly from a $t_2 \rightarrow t_2$ transition involving occupied states lying 3.4 eV below
294: HOMO and unoccupied
295: states 1.3 eV above the LUMO. The broad peak at 13.8 eV is due to a large number of weak transitions
296: involving a large number of low lying states rather than any sharp strong transition.
297: It should be noted that the present calculations are performed within the PBE-GGA at zero temperature
298: and hence the predicted optical spectra will be useful only for qualitative comparison. More accurate
299: spectra can be obtained by the TDDFT or GW approximation.
300:
301:
302:
303: \begin{figure}
304: \epsfig{file=DOS.eps,width=0.9\linewidth,clip=true}
305: \caption{ The density of states of \BNS cluster.}
306: \label{spn}
307: \end{figure}
308:
309: \begin{figure}
310: \epsfig{file=j.eps,width=0.9\linewidth,clip=true}
311: \caption{ The optical absorption spectra of \BNS cluster.}
312: \label{optical}
313: \end{figure}
314:
315:
316: In summary, density functional theory at the level of PBE-GGA is used to examine
317: the vibrational stability of the proposed fullerene-like cage structure of the \BNS
318: observed in a recent electron-radiation measurement. The cage structure was proposed on the criteria
319: that it satisfied the isolated six square rule for alternate BN cages and has dimension
320: similar to the single-shell BN cages observed in the HRTEM images. The calculations
321: indicate somewhat smaller dimension ($\sim 8 $ \AA) of the \BNS cage than that those
322: observed in experiments ($9-10$ \AA). The cage is energetically and vibrationally stable.
323: It has vertical and adiatbatic ionization potential of 8.2 and 8 eV, respectively.
324: Its mean static dipole polarizability is 79 \AA$^3$. We have predicted IR, Raman and
325: optical spectra at the DFT GGA level.
326: We hope that the predicted spectra in this work will stimulate experimental
327: measurements of these spectra. The measured spectra, in combination with
328: the predicated spectra in this work, will play a decisive role in confirmation of the
329: proposed candidate structure.
330:
331:
332: The Office of Naval Research, directly and through the Naval Research Laboratory, and and the
333: Department of Defense's High Performance Computing Modernization Program, through the Common High
334: Performance Computing Software Support Initiative Project MBD-5, supported this work.
335:
336: \begin{thebibliography}{99}
337: \bibitem{KHOCS85} H. W. Kroto, J. R. Heath, S. C. O'Brien, R. F. Curl, and R. E. Smalley, Nature {\bf 318}, 162, (1985).
338:
339: \bibitem{Zettl} N. G. Chopra, R. J. Luyken, K. Cherrey, V. H. Crespi, M. L. Cohen, S. G. Louie,
340: and A. Zettle, Science {\bf 269}, 966 (1995).
341:
342: \bibitem{SSL95} M.-L. Sun, Z. Slanina, and S.-L. Lee, Chem. Phys. Lett. {\bf 233}, 279 (1995).
343:
344: \bibitem{ZSK97} H.-Y. Zhu, T. G. Schmalz, and D. J. Klein, Int. J. Quantum Chem. {\bf 63}, 393 (1997).
345:
346: \bibitem{Strout} D. L. Strout, Chem. Phys. Lett. {\bf 95}, 383 (2004) and refereces therein.
347:
348: \bibitem{P2000} V. V. Pokropivny, V. V. Skorokhod, G. S. Oleinik, A. V. Kurdyumov, T. S. Bartnitskaya,
349: A. V. Pokropivny, A. G. Sisonyuk, D. M. Sheichenko, J. Solid State Chem. {\bf 154} (2000) 214.
350:
351: \bibitem{WJ04} H. Wu, H. Jiao, Chem. Phys. Lett. {\bf 386}, 369 (2004).
352:
353: \bibitem{ZD04} R. R. Zope and B. I. Dunlap, Chem. Phys. Lett. {\bf 386}, 403 (2004).
354:
355: \bibitem{ZBPD04} R. R. Zope, T. Baruah, M. R. Pederson, and B. I. Dunlap,
356: Chem. Phys. Lett. {\bf 393}, 300 (2004).
357:
358: \bibitem{Oku00} T. Oku, T. Hirano, M. Kuno, T. Kusunose, K. Nihara,and K. Suanuma,
359: Mater. Sci. and Engg. {\bf B74}, 206 (2000).
360:
361: \bibitem{Oku03} T. Oku, A. Nishiwaki, I. Narita, and M. Gonda, Chem. Phys. Lett. {\bf 380}, 620 (2003).
362:
363: \bibitem{Oku04} T. Oku, T. Hirano, M. Kuno, T. Kusnose, K. Nihara, and K. Suganuma,
364: Mater. Sci. and Engg. {\bf B74}, 206 (2004).
365:
366: \bibitem{Stephan98} O. Stephan, Y. Bando, A. Loiseau, F. Willamie, N. Shramchenko, T. Tamiya, and
367: T. Sato, Appl. Phys. A {\bf 67}, 107 (1998).
368:
369: \bibitem{Goldberg98} D. Goldberg, Y. Bando, O. Ste'pahan, and K. Kurashima,
370: Appl. Phys. Lett. {\bf 73}, 2441 (1998).
371:
372: \bibitem{AMC99} S. S. Alexandre, M. S. C. Mazzoni, and H. Chacham,
373: Appl. Phys. Lett. {\bf 75}, 61 (1999).
374:
375: \bibitem{KS65} W. Kohn, L. J. Sham, Phys. Rev. {\bf 140} (1965) A1133.
376:
377: \bibitem{NRLMOL}
378: M. R. Pederson, K. A. Jackson, Phys. Rev. B. {\bf 41}, 7453 (1990); {\it ibid} {\bf 43}, 7312 (1991);
379: K. A. Jackson, M. R. Pederson, Phys. Rev. B. {\bf 42}, 3276 (1990);
380:
381: \bibitem{PBE}
382: J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. {\bf 77 }, 3865(1996).
383:
384: \bibitem{Porezag99}
385: D. V. Porezag, M. R. Pederson, Phys. Rev. A {\bf 60}, 2840 (1999).
386:
387: \bibitem{Porezag96}
388: D. V. Porezag, M. R. Pederson, Phys. Rev. A {\bf 54}, 7830 (1996).
389:
390:
391: \bibitem{D91}
392: B. I. Dunlap, D. W. Brenner, J. W. Mintmire, R. C. Mowery, and C. T. White,
393: J. Phys. Chem. {\bf 95}, 8737 (1991).
394:
395:
396:
397: \end{thebibliography}
398: \end{document}
399:
400:
401: