cond-mat0104346/efg.tex
1: \documentstyle[osa,manuscript]{revtex}
2: %
3: \newcommand{\MF}{{\large{\manual META}\-{\manual FONT}}}
4: \newcommand{\manual}{rm}        % Substitute rm (Roman) font.
5: \newcommand\bs{\char '134 }     % add backslash char to \tt font
6: %
7: \begin{document}
8: %
9: 
10: \title{Electric field gradients in
11: $s$-, $p$- and $d$-metal diborides and the effect of pressure 
12: on the band structure and T$_c$ in MgB$_2$ }
13: 
14: \author{
15: N.I. Medvedeva$^{1,2}$, A.L. Ivanovskii$^1$,
16: J.E. Medvedeva$^{2,3}$, A.J.Freeman$^2$, D.L. Novikov$^4$}
17: 
18: \address{
19: $^1$Institute of Solid State Chemistry, Ekaterinburg, Russia}
20: 
21: \address{
22: $^2$Department of Physics and Astronomy, Northwestern University,
23: Evanston, Illinois}
24: 
25: \address{
26: $^3$Institute of Metal Physics,  Ekaterinburg, Russia} 
27: 
28: \address{
29: $^4$Arthur D. Little, Inc. Cambridge, Massachusetts, 02140}
30: 
31: 
32: \maketitle
33: \begin{abstract}
34: Results of FLMTO-GGA (full-potential linear muffin-tin orbital --
35: generalized gradient approximation) calculations of the band
36: structure and boron electric field gradients (EFG) for the new
37: medium-T$_c$ superconductor (MTSC), MgB$_2$, and related diborides
38: MB$_2$, M=Be, Al, Sc, Ti, V, Cr, Mo and Ta are reported. The
39: boron EFG variations  are found to be related to
40: specific features of their band structure and particularly to the
41: M-B hybridization. 
42: The strong charge anisotropy at the B site 
43: in MgB$_2$ is completely defined by the 
44: valence electrons - a property which sets MgB$_2$ apart from other diborides.
45: The boron EFG in MgB$_2$ is weakly dependent of applied pressure:
46: the  B p electron anisotropy increases with pressure, 
47: but it is partly compensated by the increase of  core charge assymetry. 
48: The concentration of holes in bonding $\sigma$ bands
49: is found to decrease slightly  from 0.067 to 0.062 holes/B 
50: under a  pressure of 10 GPa.
51: Despite a small decrease of N(E$_F$), 
52: the Hopfield parameter increases with pressure 
53: and we believe that  the main reason for the
54: reduction under pressure of the superconducting 
55: transition temperature, T$_c$,
56: is the strong pressure
57: dependence of  phonon frequencies, which is sufficient to
58: compensate the electronic effects. 
59: \end{abstract}
60: 
61: %PASC numbers: 74.72.-h
62: 
63: \newpage
64: 
65: Recently, Akimitsu et al \cite{c1} reported the discovery of
66: medium-T$_c$ superconductivity (MTSC) with T$_c$ of about 39 K in
67: magnesium diboride (MgB$_2$) with a simple composition and crystal
68: structure (AlB$_2$-type, space group P6/mmm, Z=1). 
69: The band structure
70: calculations showed \cite{c3,c4,c5,c6,c7,c8} that the MTSC in
71: MgB$_2$ can be attributed to a strong electron-phonon coupling, a
72: rather high density of states from 2D (in-plane) metallic boron
73: $\sigma (p_{x,y})$ bands  at E$_F$ and the existence of
74: $p_{x,y}$-band holes.
75: By now, a number of studies have been performed with NMR
76: \cite{c9,c10,c11} which is a very powerful technique for
77: investigating  the properties of  MgB$_2$. The 
78: measured quadrupole interaction is determined by  the value of
79: the electric field gradient (EFG), which is directly related to
80: the  charge distribution  around the  nucleus. 
81: Thus, theoretical EFG studies are  important in order
82: to give a reliable interpretation of the experimental data
83: based on the electronic structure.
84: 
85: The  EFG at the boron site in MgB$_2$ was  found 
86:  experimentally to be  much larger
87: than those  for $d$- diborides (Table I).
88: As seen from Table I, the variation in EFG for diborides 
89: covers two  orders of  magnitude 
90: and shows  trends
91: which cannot be explained by the crystal structure changes.     
92: For example, these EFG's in 
93: AlB$_2$ and TiB$_2$ 
94: differ  by  almost three times, 
95: but their lattice parameters  $a$ and $c$ are approximately the same;
96: and, vice versa, equal EFG's were obtained in BeB$_2$ and MgB$_2$,
97: for which the lattice parameters have  
98: the largest differences among diborides
99: under consideration. 
100: No attempts was made previosuly to relate the EFG changes 
101: to the peculiarities of electronic structure.
102: 
103: In this paper, we present  results of first-principles
104: full-potential LMTO-GGA \cite{c12} calculations (within the
105: generalized gradient approximation (GGA) for the exchange correlation
106: potential) of the electronic structure and  EFG's at the
107: boron site for  MgB$_2$ and other $s$-,$p$-,$d$-diborides.
108: We compare the calculated EFG's  with the experimental 
109: and other theoretical data and 
110: explain the EFG's variation  based on the anisotropy of 
111: boron 2p occupancies.
112: The pressure dependence of the EFG, 
113: which is very sensitive to the charge distribution, represents a good test
114: for the anisotropy study.    
115: The experimental and theoretical data on the
116: electronic and elastic behavior of MgB$_2$ under pressure 
117: are contradictory: it was found to be nearly isotropic \cite{Loa,c24},
118: anisotropic \cite{Islam} or strong anisotropic \cite{Rabin,c23}. 
119: We simulated the pressure effect  on the band structure of MgB$_2$  
120: and estimated  the changes in 
121: the EFG, boron p-occupancies, hole concentration 
122: and Hopfield parameter under pressure. 
123: These investigations are of interest, 
124: since the pressure dependence of the T$_c$ is the key difference 
125: between
126: the conventional BCS \cite{c4} and hole \cite{Hirsh} 
127: superconductivity mechanisms.
128:   
129: The band structures of MgB$_2$ and some other diborides are shown
130: in Fig.1.
131: There are two distinct sets of B 2$p$-bands: $\sigma$ (2$p_{x,y}$)
132: and $\pi$ ($p_z$)-types with considerably different dispersions.
133: The  B $2p_{x,y}$ bands are quasi-two dimensional (2D) along the
134: $\Gamma$-A line in the Brillouin zone (BZ) 
135: and  make a considerable contribution to the density of states 
136: at the Fermi level, N(E$_F$),
137: for MgB$_2$. Now was shown \cite{c4,c5,c6,c7,c8}, that the existence
138: of degenerate $p_{x,y}$-states above E$_F$ at the $\Gamma$ point
139: in the BZ is crucial for the MTSC in diborides. 
140: The high Tc is explained by the strong coupling of these
141: holes to the in-plane E$_{2g}$ phonon modes \cite{c7,kong}. 
142: The B
143: $2p_z$-bands are responsible for the weaker $pp_\pi$-interactions
144: and these 3D-like bands have maximum dispersion along $\Gamma-A$.
145: The bonding and antibonding B $p_z$ bands cross E$_F$ at the K
146: point
147: and their location and dispersion depend on the M-B hybridization. For
148: BeB$_2$ (Fig.1b) and AlB$_2$ (Fig.1c), the $p_{x,y}$ bands  are,
149: respectively, partly and completely filled, the Fermi surface
150: topology changes and medium -T$_c$ superconductivity is absent
151: \cite{c13,c14}.
152: The $p_z$-bands progressively move down in going from BeB$_2$ to
153: MgB$_2$ and AlB$_2$, demonstrating the strengthening of M-B
154: bonding.
155: 
156: The band structure and chemical bonding of all
157: 3$d$, 4$d$ and 5$d$-metal diborides were previously investigated in
158: detail \cite{c16,c17,c17a,c17b}. These studies showed that the cohesive
159: properties of AlB$_2$-type diborides are explained in terms of the
160: band filling. The Fermi level for TiB$_2$ (ZrB$_2$ and HfB$_2$)
161: falls in the pseudogap where bonding states are occupied and
162: antibonding states are empty (Fig.1d). The shift of  E$_F$
163: results in the partial emptying of the bonding states (ScB$_2$,
164: Fig.1e) or the occupation of antibonding states (VB$_2$, CrB$_2$,
165: MoB$_2$, TaB$_2$). Both cases correspond to the lowering of the
166: cohesive properties (melting temperature, enthalpies of formation,
167: etc.). From first-principles estimates of M-M, M-B and B-B
168: bonding strengths, we found that the cohesive energy of
169: $d$-diborides with filled bonding states 
170: decreases when the atomic number increases in the row
171: due to the weakening of M-B hybridization
172: and that the AlB$_2$-type diborides of the group VII and VIII elements are
173: unstable\cite{c16,c17,c17a,c17b}. Thus, the main feature of the
174: band structure of TM diborides is the progressive filling of B
175: $p_{x,y}$, $p_{z}$-bands and the dominant role of $d$-states near
176: E$_F$.
177: 
178: A systematic search for superconductivity in the $d$-diborides
179: (with M = Ti, Zr, Hf, V, Ta, Cr, Mo)
180: showed that T$_c$  is
181: below $\sim 0.4$ K \cite{c18}. Only NbB$_2$ and ScB$_2$ were found
182: to be superconductors with a T$_c$ of about 0.6 K \cite{c18} and
183: 1.5 K \cite{c15}, respectively. Recently, a relatively high critical
184: temperature, T$_c$ $\sim$ 9 K, was found in TaB$_2$ \cite{kacz}.
185: As seen in Fig.1f, the bonding states in TaB$_2$ are fully
186: occupied, the Fermi level is shifted away from the pseudogap to the
187: region of antibonding states and  Ta 5$d$ states define  
188: N(E$_F$) (0.9 states/eV). Among the 3$d$-diborides, only
189: in ScB$_2$ are the 2$p_{x,y}$ bands  not completely filled (there
190: is a small hole concentration of these states at $A$), but they
191: lie below E$_F$ at $\Gamma$ and the largest contribution to N(E$_F$)
192: arises from Sc 3$d$ states. Based on the band structure results and
193: calculations of the electron-phonon interaction \cite{kong}, one
194: may conclude that   superconductivity with medium T$_c$ is
195: unlikely in undoped diborides, except for MgB$_2$; the absence of
196: holes in the two-dimensional $\sigma$-bands at $\Gamma$ results in 
197: hardening of zone-phonon modes and weak electron-phonon coupling.
198: 
199: The electric field gradient tensor, defined as 
200: the second derivative of the electrostatic potential at the nucleus,
201: was calculated directly from 
202: the  FLMTO charge  density.
203: The calculated principal components of the boron  EFG tensor, 
204: $V_{zz}^B$, 
205: are shown together with other theoretical and
206: experimental data in Table 1. Note that the asymmetry parameter, 
207: $(|V_{xx}|-|V_{yy}|)/|V_{zz}|$, is equal to 0 for the AlB$_2$-type
208: structure. 
209: 
210: The largest boron EFG's in MgB$_2$ and BeB$_2$
211: demonstrate the strongest assymetry of the charge distribution
212: as compared with other diborides
213: (note, that here and below we consider the absolute value of EFG's). 
214: For the $3d$-diborides, the
215: EFG decreases from ScB$_2$ to TiB$_2$ and increases when going
216: from TiB$_2$ to VB$_2$ and CrB$_2$. 
217: The boron EFG's for 4$d$ (MoB$_2$) and 5$d$ (TaB$_2$) diborides are
218: much smaller than the EFG's for isoelectronic 3$d$ diborides. 
219: Note that all calculated EFG's are in
220: very good agreement with available experimental data and with FLAPW
221: \cite{Wimmer} theoretical results \cite{c19} (Table 1).
222: 
223: To analyze the variation  of the boron EFG's in diborides, 
224: we consider it as a sum of
225: electron ($V_{zz}^{el}$) and lattice (ion) ($V_{zz}^{lat}$)
226: contributions (Table 1).
227: For  the $s,p$ diborides,
228: the  ion contribution is small,
229: and the boron EFG is mainly determined by
230: the anisotropy of the valence electrons. 
231: For the $d-$diborides, $V_{zz}^{lat}$
232: is larger  and  depends somewhat on the  metal,
233: (except ScB$_2$ where the lattice parameters are the largest).
234: For TaB$_2$, the electron and ion contributions 
235: are almost equal,  the EFG is positive and         
236: smallest among all diborides.
237: The electronic contributions explain  the boron EFG variation,
238: although they overestimate the calculated and observed EFG's;
239: clearly, the lattice contributions must be taken into account 
240: in order to obtain  good agreement with experiment for $d$ diborides.
241: Among the diborides, the lattice contribution is the smallest for
242: MgB$_2$ and  
243: the  strong charge anisotropy  in MgB$_2$
244: is completely defined by the valence B $p$ electrons --
245: a property which sets MgB$_2$ apart from other diborides.  
246: 
247: A qualitative explanation of EFG behavior 
248: may be given based on the  anisotropy of B 2$p$ partial 
249: occupancies, $\Delta n_p$=$p_z-(p_x+p_y)/2$,
250: since 
251: $V_{zz}^{el}$ $\sim$  $<1/r^3>$ $\Delta n_p$
252: and one may consider the boron $p$ 
253: $<1/r^3>$ expectation value  to be constant for all diborides
254: discussed. 
255: Thus, the variation of the electronic EFG's is determined by the
256: interplay of $p_z$ and $p_x, p_y$ occupations.
257: As seen from the partial DOS (PDOS)
258: obtained by means of a Mulliken population analysis
259: , the $p_{x,y}$ orbitals are
260: more occupied than are  $p_z$ orbitals (Fig.2) and 
261: $V_{zz}^{el}$ is negative for all diborides considered. For MgB$_2$ and
262: BeB$_2$, the high  $p_{x,y}$  peaks lying below -2 eV
263: lead to large negative $\Delta n_p$ values and, therefore, to  large boron
264: EFG's. The small increase of $p_z$ occupancy explains the EFG lowering
265: for MgB$_2$ (and AlB$_2$) as compared to BeB$_2$. Thus, the weaker
266: M-B bonds for $s$-, $p$-diborides correspond to  larger 
267: boron EFG's.
268: 
269: For $d$-diborides, the $p_z$ PDOS is more localized due to 
270: strong covalent M 3$d$-B 2$p$ bonding and 
271: the  high peak  at 3-5 eV below E$_F$ 
272: decreases $\Delta n_p$ and $V_{zz}^{el}$ compared with $s$-, $p$-diborides. 
273: Among the $3d$-diborides, the EFG is  smallest for
274: TiB$_2$, which has the strongest p-d hybridization. 
275: Weaker p-d hybridization 
276: (less intense p$_z$ peak) 
277: for ScB$_2$
278: and CrB$_2$ results in a larger anisotropy,  $\Delta n_p$,
279: and the boron EFG's are larger for these compounds than for  TiB$_2$.
280: Since 4$d$ and 5$d$ states are less localized
281: than are 3$d$ states, the corresponding  B $p_z$ PDOS are broadened for
282: MoB$_2$ and especially for TaB$_2$ (Fig.2)
283: (the strong hybridization of B 2p and Ta 5d states was shown also in
284: Ref.\cite{Pick}), that leads to the small EFG's. 
285: While a Mulliken analysis is not an accurate approach for
286: the calculation of orbital charges, especially for delocalized
287: p-orbitals, it still allows one to describe general trends in EFG's and to
288: correlate them with  peculiarities of  the electronic structure.
289: Thus, we conclude  that  M-B$p$
290: hybridization is the main factor controlling the boron EFG variation.
291: 
292: The effect of hydrostatic pressure on the EFG at the B site in MgB$_2$ 
293: was investigated for 5 and 10 GPa with
294: lattice parameters taken from  the extrapolation formula
295: \cite{c23} 
296: $a=a_0(1-0.00187P)$ and $c=c_0(1-0.00307P)$.
297: We found a very slow increase of EFG with pressure
298: (Table 1),
299: that also demonstrates that the boron EFG's 
300: in diborides do not have a strong dependence on
301: the interatomic distances. 
302: As  the EFG is a very sensitive characteristic, 
303: no large changes are expected in the anisotropy of the B charge distribution 
304:  under pressure.
305: 
306: As expected, the boron $p$  bands  widen under pressure (Fig.1a).
307: One can see that the band shifts relative to E$_F$ 
308: are different for different directions of BZ;
309: the main changes in the occupied $p_{x,y}$ and
310: $p_z$ bands occur in the low energy range 
311: at the M, K and A points. 
312: These bands  
313: move down with pressure relative to E$_F$ along  $\Gamma$-M-K-$\Gamma$
314: and  A-L  and 
315: the overall shift of the PDOS to lower energies 
316: leads to the loss of these states in the energy range from E$_F$ to -2 eV, 
317: as stated  in Ref. \cite{c24}. 
318: The decrease of $p_{x,y}$ and  $p_z$  PDOS 
319: near  E$_F$ is
320: partly compensated by its increase at lower energies,
321: and as a result, the changes in the partial $p$
322: occupations are small. 
323: We found the increase of  $p$ occupancies with pressure
324: to be  anisotropic -- the larger growth of $p_{x,y}$ occupancy 
325: compared to $p_z$  giving  
326: an $\Delta n_p$ increase by 0.02 for 10 GPa. 
327: The accurate calculation
328: gives smaller EFG than it follows from $\Delta n_p$. 
329: As seen from the Table,
330: the  increase of B p electron anisotropy
331: is partly compensated by the increase of  the core charge contribution,
332: and as a result
333: the boron EFG in MgB$_2$ is weakly dependent on applied pressure.
334: Thus, we  conclude, that 
335: the  charge distribution at B site 
336: shows more  isotropic change under pressure
337: than do the B $p$ valence electrons. 
338: due to the compensating behavior of electron and core systems.
339: NMR measurements under pressure would be important
340: to confirm  this conclusion.
341:  
342: The PDOS changes  near E$_F$ under a pressure of 10 GPa is shown in Fig.3. 
343: The hole concentration in the bonding $\sigma$ bands 
344: decreases by 0.005 (within an energy interval up to 0.8 eV).
345: The changes in $p_{x,y}$ PDOS  is due to  
346: the behavior of these bands with pressure along 
347: $\Gamma$-A:
348: the bands move down at $\Gamma$ and up at A, 
349: resulting, respectively,  in the loss of holes 
350: in the energy range up to 0.5 eV
351: and their increase for higher energies.
352: Although  the hole $p_z$  PDOS change 
353: is negative in the energy interval up to 0.2 eV,
354: its concentration is almost constant under pressure 
355: for energies  up to 0.8 eV.  
356: The loss of both $p_{x,y}$ and $p_z$ states near E$_F$
357: results in a small decrease of N(E$_F$) by 0.02 states/eV.
358: Thus, we showed that the carrier concentration  decreases 
359: and so the observed decrease of the resistance with pressure 
360: is likely  to be connected  
361: with better coupling between the sintered grains, 
362: as  suggested in Ref. \cite{Mont}.
363: 
364: Finally, we estimated the pressure dependence of
365: the Hopfield parameter, $\eta$, which is an electronic part of the
366: electron-phonon coupling
367: $\lambda$ = $\eta$/M$<\omega^2>$, where
368: $\eta$= N(E$_F$)$<I^2>$.   
369: The calculation of 
370: the averaged electron-ion matrix element squared, $<I^2>$, 
371: performed  within the rigid muffin tin
372: approximation\cite{foot} gave a faster increase of $<I^2>$ with pressure 
373: than the N(E$_F$) decrease.
374: As a result, 
375: despite a small decrease of N(E$_F$), ($dN(E_F)/dP$ = -0.51\%/GPa), 
376: the Hopfield parameter increases with pressure as
377: $d\eta/dP$ = + 0.55\%/Gpa. 
378: Hence the decrease  of N(E$_F$) cannot be considered 
379: as the reason for  the T$_c$ reduction,
380: which is known \cite{chu} to behave 
381: as $dT_c/dP$ = - 1.6 K/GPa. 
382: Thus,
383: according to the McMillan formula, the main reason for the
384: reduction of T$_c$ under pressure is the strong pressure
385: dependence of  phonon frequencies, which is sufficient to
386: compensate for the electronic effects. 
387: 
388: 
389: Work at Northwestern University supported by the U.S. Department
390: of Energy (grant No. DE-F602-88ER45372)
391: 
392:  \begin{references}
393: 
394:  \bibitem{c1} J. Akimitsu, Symposium on Transition Metal Oxides, Sendai,
395:  Januaru 10, 2001; J. Nagamatsu et al,  Nature, {\bf 410}, 63 (2001).
396:  
397:  \bibitem{c3} A.L. Ivanovskii and N.I. Medvedeva, Russian J. Inorganic
398:  Chemistry {\bf 45}, 1234 (2000).
399: 
400:  \bibitem{c4} J. Kortus, I.I. Mazin, K.D. Belashenko, V.P. Antropov and
401:  L.L.Boyer, Phys.Rev.Lett., {\bf 86},4656 (2001). 
402: 
403:  \bibitem{c5} K.D. Belahschenko, M.van Schilfgaarde and V.P. Antropov,
404:  Phys. Rev. B ?  (2001).
405: 
406:  \bibitem{c6} G. Satta, G. Profetta, F. Bernardini, A. Continenza
407:  and S. Massidda, cond-mat/0102358 (2001).
408: 
409:  \bibitem{c7} J.M. An and W.E. Pickett, Phys. Rev. Lett. {\bf 86}, 4366 (2001).
410: 
411:  \bibitem{c8} N.I. Medvedeva, A.L. Ivanovskii, J.E. Medvedeva and A.J. Freeman,
412:   Phys. Rev. B, {\bf 64}, 020502 (2001).
413: 
414:  \bibitem{c9}  A. Gerashenko et al, cond-mat/0102421 (2001).
415: 
416:  \bibitem{c10} H. Kotegawa, K. Ishida, Y. Kitaoka, T. Muranaka and J. Akimitsu
417:   cond-mat/0102334 (2001).
418: 
419:  \bibitem{c11} S.L. Bud'ko, G. Lapertot, C. Petrovic,
420:  C.E. Cunningham, N. Anderson and P.C. Canfield,
421:  Phys.Rev.Lett., {\bf 86}, 1877 (2001).
422: 
423:  \bibitem{c12} M. Methfessel and M. Scheffler, Physica, {\bf B172}, 175 (1991).
424: 
425:  \bibitem{Loa} I. Loa and  K. Suassen,
426:   Solid State Commun.,  (2001), in press.
427: 
428:  \bibitem{c24} T. Vogt, G. Schneider, J.A. Hriljac, G. Yang and J.S. Abell
429:   Phys. Rev. B, {\bf 63}, 0220505 (2001).
430:  
431:  \bibitem{Islam}  A.K.M.A. Islam, F.N.Islam, and S. Kabir, J.Phys. 
432:   Cond.Matter, {\bf 13}, L641 (2001).
433: 
434:  \bibitem{Rabin} P. Rabindran,   et al, cond-mat/0104253, (2001).
435: 
436:  \bibitem{c23}  J.D. Jorgensen,  D.G. Hinks and S. Short, Phys. Rev. B,
437:   {\bf 63}, 4222  (2001).
438:  
439:  \bibitem{Hirsh} J.E. Hirsh,   cond-mat/0102115, (2001).
440: 
441:  \bibitem{kong} Y. Kong, O.V. Dolgov, O. Jepsen and O.K. Andersen, Phys. Rev. B,
442:   {\bf 64}, 020501 (2001).
443: 
444:  \bibitem{c13} I. Felner, Physica C, {\bf 353}, 1 (2001).
445: 
446:  \bibitem{c14} J.S. Slusky et al, Nature, {\bf 411}, 6833 (2001).
447: 
448: 
449:  \bibitem{c16} A.L. Ivanovskii, N.I. Medvedeva et al,
450:  Metallofiz. Nov. Tecknol., {\bf 20}, 41 (1998).
451: 
452:  \bibitem{c17} A.L. Ivanovskii, N.I. Medvedeva and J.E. Medvedeva,
453:   Metallofiz. Nov. Tecknol., {\bf 21}, 19 (1999).
454: 
455:  \bibitem{c17a} A.L. Ivanovskii, N.I. Medvedeva and J.E. Medvedeva,
456:   Mendeleev Commun., {\bf 4}, 129 (1998).
457: 
458:  \bibitem{c17b} A.L. Ivanovskii, N.I. Medvedeva and J.E. Medvedeva,
459:   Dokl.Akad. Hauk (in Russian), {\bf 361}, 642 (1998).
460: 
461:  \bibitem{c18} L. Leyarovska and E. Leyarovski, J. Less Common Metals,
462:  {\bf 67}, 249 (1979).
463: 
464:  \bibitem{c15} G.V. Samsonov and I.M. Vinitsky, Refractory Compounds
465:  (in Russian),  Metallurgija, Moskva, 1976.
466:  
467:  \bibitem{kacz} D. Kaczorowski  et al, cond-mat/0103571, (2001).
468: 
469: 
470:  \bibitem{Wimmer}  E. Wimmer, H. Krakuaer, M. Weinert and A. Freeman, Phys. Rev. B,
471:  {\bf 24}, 864 (1981).
472: 
473:  \bibitem{c19}  K. Schwarz, H. Ripplinger and P. Blaha. Z. Naturforsch.,
474:  {\bf 51a}, 527 (1996).
475: 
476:  \bibitem{c21} J.P. Kopp and R.G. Barnes, J. Chem. Phys., {\bf 54}, 1840 (1971).
477: 
478:  \bibitem{c22} A.H. Silver and T. Kushida, J. Chem. Phys, {\bf 38}, 865 (1963).
479: 
480:  \bibitem{Pick} H. Rosner and W.E.Pickett,   cond-mat/0106092, (2001).
481: 
482:  \bibitem{foot} For zero pressure we obtained the Hopfield parameter
483:  $\eta$ = 129.2 mRy/$a^2_B$ which
484:  is very close to value 135 mRy/$a^2_B$ from \cite{c4}
485:  and leads to approximately the same
486:  T$_c$ $\sim$ 20 K for the Coulomb pseudopotential $\mu$=0.1
487:  and phonon frequency $<\omega_{log}>$=500 cm$^{-1}$.
488: 
489:  \bibitem{Mont} M. Monteverde, et al, Science, {\bf 292}, 75 (2001).
490:  
491:  \bibitem{chu} B. Lorentz, R.L. Meng and C.W. Chu, Phys. Rev. B, {\bf 64},
492:   012507 (2001).
493: 
494: 
495:  
496: 
497: 
498: 
499: \end{references}
500: 
501:  \begin{figure}
502:  \caption{Band structures of (a) MgB$_2$ (dot lines for zero pressure, 
503:   solid lines for 10 GPa),
504:   (b) BeB$_2$, (c) AlB$_2$, (d) TiB$_2$, (e) ScB$_2$,
505:   and (f) TaB$_2$. \label{fig1}}
506:  \end{figure}
507: 
508: 
509: 
510: \begin{figure}
511: \caption{
512: Boron partial densities of $p_{x,y}$ (dash lines) and $p_z$
513:   (dot lines) states (PDOS) and their anisotropy (solid lines) for
514:    (a)  MgB$_2$, (b) BeB$_2$, (c) TiB$_2$, (d) TiB$_2$. 
515:  The Fermi level corresponds to the zero energy. \label{fig2}}
516:  \end{figure}
517:  
518: \begin{figure}
519: \caption{
520: The change in the boron $p_{x,y}$ (solid  line) and $p_z$ (dash line)
521:   PDOS in Mg$B_2$ under  pressure 10 GPa. 
522:    The Fermi level corresponds to  the zero energy. \label{fig3}}
523: \end{figure}
524: 
525: \begin{table}
526: \caption{
527: Theoretical and experimental boron EFG, $V_{zz}$ (in
528: $10^{21}V/m^2$),
529:   for $s$-, $p$-, $d$-diborides}
530: \begin{tabular}{cccccc}
531: Diboride   & $V_{zz}^{el}$ &  $V_{zz}^{lat}$ & $V_{zz}$ & $|V_{zz}^B|$\cite{c19} & $|V_{zz}^B|$, exp \\ \hline
532: MgB$_2$     & -1.94 & 0.06 &  -1.88 &  &  1.69\cite{c9}  \\
533: MgB$_2$\footnotemark[1] & -2.00 & 0.10 & -1.90  &  &   \\
534: BeB$_2$     & -2.43 & 0.33 & -2.10 &  &            \\
535: AlB$_2$     & -1.17 & 0.18 & -0.99 &   &  1.08\cite{c21}  \\
536: ScB$_2$     & -0.75 & 0.13 & -0.60 &   &           \\
537: TiB$_2$     & -0.66 & 0.31 & -0.35 & 0.38   &  0.37\cite{c22}  \\
538: VB$_2$      & -0.76 & 0.38 & -0.38 & 0.39   &  0.43\cite{c22}  \\
539: CrB$_2$     & -1.01 & 0.42 & -0.59 & 0.60   &  0.63\cite{c22}  \\
540: MoB$_2$     & -0.55 & 0.32 & -0.23 & 0.22   &  0.23\cite{c22}  \\
541: TaB$_2$     & -0.21 & 0.25 &  0.04 & $<$0.05  &  0.02\cite{c22}  \\ 
542: \end{tabular}
543: \footnotetext[1]{under pressure 10 GPa}
544: \end{table}
545: 
546: \end{document}
547: