1:
2: \documentclass[prb,twocolumn,showpacs]{revtex4}
3: \usepackage{graphicx}
4:
5: \begin{document}
6:
7: \title{Effects of pressure on the superconducting properties of magnesium diboride }
8: \author{X. J. Chen,\thanks{Present address: Department of Physics, Kent State University,
9: Kent OH 44242} H. Zhang, and H.-U. Habermeier}
10: \affiliation{Max-Planck-Institut f\"{u}r Festk\"{o}rperforschung, D-70569 Stuttgart, Germany}
11: \date{Received 21 May 2001}
12:
13: \begin{abstract}
14: We discuss the effects of hydrostatic pressure on the superconducting properties of MgB$_{2}$
15: within the framework of Eliashberg theory. By considering the pressure dependences of all
16: parameters appearing in the McMillan formula, we show that the calculated pressure derivative of
17: $T_{c}$ as well as the variation of $T_{c}$ with pressure are in good agreement with recent
18: measurements. The pressure dependences of the energy gap $\Delta_{0}$, the effective interaction
19: strength $N(E_{F})v$, the critical magnetic field $H_{c}(0)$, and the electronic specific heat
20: coefficient $\gamma$ are also predicted for this system. A comparison of pressure effect in
21: non-transition elements clearly suggests that MgB$_{2}$ is an electron-phonon mediated
22: superconductor.
23: \end{abstract}
24: \pacs{74.62.Fj, 74.70.Ad }
25: \maketitle
26:
27: \section{INTRODUCTION}
28:
29:
30: The recent discovery of superconductivity in MgB$_{2}$ (Ref.\ \onlinecite{naga}) has attracted
31: considerable interest in the study of this material, both to understand the mechanism of
32: superconductivity and to explore other properties of MgB$_{2}$ and related materials. The high
33: transition temperature $T_{c}\approx$ 40 K in this material offers other possibility for finding
34: high-$T_{c}$ superconductivity in some binary intermetallic compounds besides cuprates and
35: $C_{60}$-based compounds. Meanwhile, the high critical currents observed in MgB$_{2}$ thin films
36: \cite{eom} and wires\cite{jin} reveal that MgB$_{2}$ belongs to a new class of low-cost,
37: high-performance superconducting materials for magnets and electronic applications.
38:
39:
40: Measurements of the isotope effect and of the influence of pressure on the transition temperature
41: and critical field of superconductors yield information on the interaction causing
42: superconductivity. Indeed, the pressure (or volume) and the mass number would seem to be the only
43: variables whose effect might be capable of immediate theoretical interpretation. By observing
44: how pressure changes the parameters of the lattice in the normal state and in the superconducting
45: state, and comparing the measurements with the theoretical predictions, one can test the validity
46: of some theoretical models. Olsen $et$ $al.$ \cite{olse1} have shown that the volume (V)
47: dependence of effective interaction $N(E_{F})v$, $d\ln N(E_{F})v/d\ln V$, can be scaled well with
48: the deviation $\xi$ from the full isotope effect where $\xi$ is defined by
49: $T_{c}\propto M^{-0.5(1-\xi)}$ in superconducting metals. Bud'ko $et$ $al.$ \cite{bud} and Hinks
50: $et$ $al.$ \cite{hink} reported a sizeable isotope effect for B ($\alpha_{B}$=0.26(3) or 0.30(1))
51: in newly discovered superconductor MgB$_{2}$. Although the total isotope coefficient
52: $\alpha$=0.32(1) (Ref.\ \onlinecite{hink}) is smaller than the canonical BCS value of 0.5,
53: it is the same as that in Cd (Ref.\ \onlinecite{palm}). The isotope effect along with other
54: measurements such as inelastic neutron scattering,\cite{sst,osbo} tunneling,\cite{kara} NMR
55: (Ref.\ \onlinecite{kote}), and specific heat\cite{krem,walt,bouq} confirmed that MgB$_{2}$ is
56: an electron-phonon mediated $s$-wave superconductor.
57:
58:
59: Soon after the discovery of superconductivity in MgB$_{2}$, the effect of pressure on $T_{c}$ was
60: studied by two groups\cite{mont,lore1} by resistivity or $ac$ susceptibility measurements. Both
61: groups observed a decrease of $T_{c}$ with increasing pressure, with initial pressure derivative
62: $dT_{c}/dP$ of --0.8 K/GPa (Ref.\ \onlinecite{mont}) or --1.6 K/GPa (Ref.\ \onlinecite{lore1}),
63: respectively. Moreover, Monteverde $et$ $al.$ \cite{mont} found that the superconductivity is
64: not destroyed by applying high pressure up to 25 GPa, at which point $T_{c}$ is as high as 21 K.
65: A somewhat larger $dT_{c}/dP$ of --2.0 K/GPa was late reported by Saito $et$ $al.$\cite{sait} from
66: high-pressure resistivity measurements. Using a He-gas apparatus, Tomita $et$ $al.$\cite{tomi}
67: determined a $dT_{c}/dP$ of --1.11 K/GPa under pure hydrostatic pressure conditions. In order to
68: find the reason why the reported values of $dT_{c}/dP$ are different among different groups,
69: Lorenz $et$ $al.$ \cite{lore2} carried out high pressure experiments on MgB$_{2}$ samples with
70: different $T_{c}$'s at ambient pressure and different pressure media. $T_{c}$ was found to
71: decrease linearly over the whole pressure range (0--1 GPa). In the He environment, the two samples
72: with the initial $T_{c}$=39.2 and 37.4 K yield the pressure derivatives of --1.07 and --1.45
73: K/GPa, respectively. The former is obviously very close to that of Tomita $et$ $al.$\cite{tomi}
74: The latter approaches their previous data,\cite{lore1} which was obtained by using the Fluorinert
75: FC77 as pressure medium. They therefore concluded that the variation in the value of $dT_{c}/dP$
76: by various groups results from the differences in sample preparation conditions. The value of
77: $dT_{c}/dP\simeq $--1.1 K/GPa is then confirmed to give the true hydrostatic pressure dependence
78: of $T_{c}$ in MgB$_{2}$.
79:
80:
81: Two theoretical models have been tried to describe the systematics of the behavior of $T_{c}$
82: under pressure in MgB$_{2}$. Based on the theory of hole superconductivity, Hirsch \cite{hirs}
83: predicted an increased $T_{c}$ with the decrease of B-B intraplane distance under the application
84: of in-plane biaxial pressure. However, this prediction has not been confirmed experimentally yet.
85: No uniaxial pressure measurement was reported due to the extreme difficulty in growing MgB$_{2}$
86: single crystal. The experiments of hydrostatic pressure effect on $T_{c}$ do not particularly
87: support this theory provided that no charge transfer between the Mg and B layers occurs.
88: Alternatively, the experimental results have been analyzed \cite{lore1,tomi,loa,vogt} by using the
89: McMillan formula \cite{mcmi} derived from Eliashberg theory,\cite{elia} supporting electron-phonon
90: mediated superconductivity. Interestingly, Loa and Syassen \cite{loa} analyzed the pressure effect
91: on $T_{c}$ from McMillan formula on the basis of their calculated elastic and electronic structure
92: data. Assuming that the electron-ion matrix element $I$ is pressure independent, they found that
93: the pressure effect on $T_{c}$ is in good agreement with experimental data by using a lattice
94: Gr\"{u}neisen parameter $\gamma_{G}$=1. These assumptions, however, deserve some refinements.
95: Recent band structure calculations suggest that MgB$_{2}$ is a traditional $sp$ metal
96: superconductor.\cite{kort,pick,kong} The pressure dependence of $I$ has long been an interesting
97: issue of the research of pressure effect in simple $sp$ metals.\cite{hodd,trof,coul,rott,daco}
98: Ziman's calculation of the electron-phonon interaction leads to $<I^{2}>\propto N(E_{F})^{-2}$ at
99: least in the limit of long wavelengths.\cite{zima} This then indicated that the consideration of
100: the pressure dependence of $I$ would be important for better understanding the superconducting
101: properties of MgB$_{2}$ under pressure. On the other hand, it has been found\cite{tomi}
102: that the choice of lattice Gr\"{u}neisen parameter $\gamma_{G}$ is crucially important in
103: explaining both the magnitude and the sign of the pressure derivative of $T_{c}$ when using
104: McMillan formula. The value of $\gamma_{G}$=1 in the calculation of Loa and Syassen is obviously
105: lower than those reported recently.\cite{roun,gonc} The pressure dependence of the effective
106: electron-electron Coulomb repulsion $\mu^{*}$ appearing in the McMillan formula is usually
107: neglected in previous studies due to the assumption of the small change of $\mu^{*}$ compared with
108: that of the electron-phonon coupling parameter $\lambda$ (Ref.\ \onlinecite{seid}). However, the
109: magnitude of $\mu^{*}$ is also of interest in connection with the possibility that superconductivity
110: may be destroyed by pressure.\cite{seid,bo} It was argued that the pressure dependence of $\mu^{*}$
111: makes a significant contribution to the behavior of $T_{c}$ under very high pressures and must
112: be handled carefully.\cite{daco,gubs,papa}
113:
114:
115: In this paper we discuss the pressure dependences of some interested superconducting properties in
116: MgB$_{2}$. The outline of this paper is as follows: In Sec. II we presented a theoretical
117: approach for pressure effects on the superconducting properties in the simple $sp$ metals
118: superconductors. Section III contains the theoretical results obtained and a comparison with
119: experiments for MgB$_{2}$. We draw conclusions in Sec. IV.
120:
121:
122: \section{THEORETICAL FORMULATION}
123:
124:
125: For our purposes, the relation between $T_{c}$ and microscopic parameters is given adequately by
126: the McMillan equation\cite{mcmi}
127: \begin{equation}
128: T_{c}=\frac{\Theta
129: _{D}}{1.45}\exp\left[-\frac{1.04(1+\lambda)}{\lambda-\mu^{*}(1+0.62\lambda)}\right]~~,
130: \end{equation}
131: which relates $T_{c}$ to the electron-phonon coupling parameter $\lambda$, the Coulomb repulsion
132: strength $\mu^{*}$, and a temperature $\Theta _{D}$ characteristic of the phonons.
133:
134:
135: Considering the variations of $\Theta _{D}$, $\lambda$, and $\mu^{*}$ with pressure or volume and
136: introducing parameters $\varphi = \partial\ln \lambda/\partial\ln V$ and $\phi =\partial\ln
137: \mu^{*}/\partial\ln V$, we can get the pressure coefficient of $T_{c}$
138: \begin{eqnarray}
139: \frac{d\ln T_{c}}{dP}=\frac{\gamma _{G}}{B_{0}} -
140: \frac{1.04\lambda(1+0.38\mu^{*})}{[\lambda-\mu^{*}(1+0.62\lambda)]^{2}}\frac{\varphi}{B_{0}} \nonumber \\
141: + \frac{1.04\mu^{*}(1+\lambda)(1+0.62\lambda)}{[\lambda-\mu^{*}(1+0.62\lambda)]^{2}}\frac{\phi}{B_{0}} ~~,
142: \end{eqnarray}
143: where $B_{0}\equiv 1/\kappa _{V}=-\partial P/\partial \ln V$ is the bulk modulus and $\Theta _{D}$
144: is assumed to be proportional to $<\omega^{2}>^{1/2}$ and
145: $\gamma _{G}=-\partial\ln <\omega^{2}>^{1/2}/\partial\ln V$ being the effective Gr\"{u}neisen
146: parameter.
147:
148:
149: It is well known that the usual BCS result for the energy gap can be expressed by\cite{bcs}
150: \begin{equation}
151: \Delta_{0}=2\Theta _{D}\exp\left[-\frac{1}{N(E_{F})v}\right]~~.
152: \end{equation}
153: Where $N(E_{F})$ is the electronic density of states at the Fermi energy $E_{F}$ and $v$ is the
154: pairing potential arising from the electron-phonon interaction. If we renormalize the
155: Morel-Anderson result \cite{moan} by introducing the renormalization parameter
156: $Z_{n}(0)\equiv 1+\lambda$ into their analysis, the effective interaction strength $N(E_{F})v$
157: can be rewritten as\cite{leav}
158: \begin{equation}
159: N(E_{F})v=\frac{\lambda-\mu^{*}}{1+\lambda}~~.
160: \end{equation}
161: The logarithmic volume derivative of $N(E_{F})v$ is then given by
162: \begin{equation}
163: \frac{d\ln N(E_{F})v}{d\ln V}=\frac{\lambda(1+\mu^{*})}{(\lambda-\mu^{*})(1+\lambda)}\varphi
164: -\frac{\mu^{*}}{\lambda-\mu^{*}}\phi~~.
165: \end{equation}
166: Considering the experimental observations of the pressure dependence of the energy gap of
167: superconductor,\cite{jpf} we differentiate Eq. (3) with respect to pressure
168: \begin{equation}
169: \frac{d \ln \Delta_{0}}{dP}=\frac{\gamma _{G}}{B_{0}}-\frac{1}{B_{0}}\left[
170: \frac{\lambda(1+\mu^{*})}{(\lambda-\mu^{*})^{2}}\varphi
171: -\frac{\mu^{*}(1+\lambda)}{(\lambda-\mu^{*})^{2}}\phi \right]~~.
172: \end{equation}
173:
174:
175: BCS expression for the critical field $H_{c}$ at absolute zero temperature is\cite{bcs}
176: \begin{equation}
177: \frac{H_{c}(0)^{2}}{8\pi}=2N(E_{F})\omega^{2}\exp\left[-\frac{2}{N(E_{F})v}\right]~~.
178: \end{equation}
179: Differentiating Eq. (7) with respect to the pressure, one obtains an expression of the pressure
180: coefficient of $H_{c}(0)$,
181: \begin{equation}
182: \frac{d\ln H_{c}(0)}{dP}=\frac{d\ln \Delta_{0}}{dP}-\frac{\gamma_{N}}{2B_{0}}+\frac{1}{2B_{0}}~~,
183: \end{equation}
184: where $\gamma _{N}=\partial \ln N(E_{F})/\partial \ln V$.
185:
186:
187: The expressions for $\gamma _{G}$, $\varphi$, and $\phi$ can be integrated to give
188: \begin{eqnarray}
189: \Theta _{D}(V) &=& \Theta _{D}(0)\left[V/V_{0}\right]^{-\gamma _{G}}\\ \nonumber
190: \lambda(V) &=& \lambda(0)\left[V/V_{0}\right]^{\varphi}\\ \nonumber
191: \mu^{*}(V) &=& \mu^{*}(0)\left[V/V_{0}\right]^{\phi} ~~.
192: \end{eqnarray}
193: Here $V$ and $V_{0}$ are the unit cell volumes under the applied pressure and at ambient
194: pressure, respectively. These two volumes can be related according to the first-order Murnaghan
195: equation of state $V(P)=V(0)(1+B_{0}^{\prime}P/B_{0})^{-1/B_{0}^{\prime}}$. The Eq. (9) is then
196: rewritten as
197: \begin{eqnarray}
198: \Theta _{D}(P) &=& \Theta _{D}(0)\left[1+\frac{B_{0}^{\prime}P}{B_{0}}\right]^{\gamma
199: _{G}/B_{0}^{\prime}}\\ \nonumber
200: \lambda(P) &=& \lambda(0)\left[1+\frac{B_{0}^{\prime}P}{B_{0}}\right]^{-\varphi/B_{0}^{\prime}}\\
201: \nonumber
202: \mu^{*}(P) &=& \mu^{*}(0)\left[1+\frac{B_{0}^{\prime}P}{B_{0}}\right]^{-\phi/B_{0}^{\prime}} ~~.
203: \end{eqnarray}
204:
205:
206: From Eqs. (1) and (10) we arrive at the expression for the pressure dependence of $T_{c}$
207: \begin{equation}
208: T_{c}(P)=T_{c}\left[\Theta _{D}(P), \lambda(P), \mu^{*}(P)\right]~~.
209: \end{equation}
210:
211:
212: Knowing $B_{0}$, $B_{0}^{\prime}$, $\gamma _{G}$, $\gamma _{N}$, $\phi$, and $\varphi$, one can
213: evaluate the pressure effects on the superconducting properties, especially the behavior of
214: $T_{c}$ under pressure. $B_{0}$ and $B_{0}^{\prime}$ can be obtained from the compressibility data
215: determined by neutron or synchrotron x-ray diffractions. A direct experimental determination of
216: $\gamma _{G}$ can be made by measuring electron tunneling \cite{keel,yama,zava} or inelastic
217: neutron scattering\cite{lech} under high pressure. In general, for metals in which different
218: techniques yield similar Gr\"{u}neisen constants, a good approximation to $\gamma _{G}$ is
219: provided by the room temperature value determined from the Gr\"{u}neisen equation
220: \begin{equation}
221: \gamma _{G}=\frac{\alpha _{V}V_{m}}{\kappa _{V}C_{p}}~~,
222: \end{equation}
223: where $\alpha _{V}$ is the volume coefficient of thermal expansion, $V_{m}$ is the molar volume,
224: and $C_{p}$ is the molar heat capacity at constant pressure. The approximation for $\gamma_{G} $
225: of Slater is derived from the pressure derivative of the bulk modulus
226: \cite{slat}
227: \begin{equation}
228: \gamma_{G} ^{S} \equiv \frac{B_{0}^{\prime}}{2}-\frac{1}{6}=-\frac{2}{3}
229: -\frac{1}{2}\frac{V\partial ^{2}P/\partial V^{2}}{\partial P/\partial V}~~.
230: \end{equation}
231:
232:
233: The formula for $\mu^{*}$ due to Morel and Anderson \cite{moan} used here is
234: \begin{equation}
235: \mu^{*}=\frac{\mu}{1+\mu \ln (E_{F}/\omega_{ph})}~~,
236: \end{equation}
237: with $\mu=0.5\ln[(1+a^{2})/a^{2}]$ and $a^{2}=\pi e^{2}N(E_{F})/k_{F}^{2}$, from which we
238: evaluate the volume dependence of $\mu^{*}$ as
239: \begin{equation}
240: \phi=\mu^{*}\left[\frac{2}{3}-\gamma_{G}-\frac{1-e^{-2\mu}}{2\mu^{2}}(\gamma_{N}+\frac{2}{3})\right]~~.
241: \end{equation}
242: Here the variation of $k_{F}$ with volume has been calculated from the fundamental definition
243: $k_{F}=(3\pi^{2}Z/V)^{1/3}$ with $Z$ the valency. Unfortunately, to the best of our knowledge
244: $\gamma_{N}$ has never measured directly for any superconductor in the case of free electron gas
245: it would have a value of $2/3$. Using the expression given by Migdal \cite{migdal} for the
246: electronic specific heat coefficient $\gamma$ one obtains for the electronic Gr\"{u}neisen
247: parameter
248: \begin{equation}
249: \gamma_{e}=\frac{\partial \ln \gamma}{\partial \ln V}=\gamma_{N}+\frac{\lambda}{1+\lambda}\varphi~~.
250: \end{equation}
251: The electronic Gr\"{u}neisen parameter $\gamma _{e}$ is usually deduced from measurements through
252: the simple relation \cite{barr}
253: \begin{equation}
254: \gamma _{e}=\frac{\alpha _{e}V_{m}}{\kappa _{V}C_{e}}~~.
255: \end{equation}
256: Here $\alpha _{e}$ is the contribution to the expansion coefficient from the electrons at lower
257: temperatures and $C_{e}$ is the electronic heat capacity. A theoretical estimate of $\gamma _{e}$
258: can also be given from the measurement of the volume dependence of the orbital susceptibility
259: \cite{fawc} or from band-structure considerations.\cite{flet}
260:
261:
262: The electron-phonon coupling parameter $\lambda$ can be expressed as
263: \begin{equation}
264: \lambda =\frac{N(E_{F})<I^{2}>}{M<\omega^{2}>}\equiv \frac{\eta}{M<\omega^{2}>}~~,
265: \end{equation}
266: where $<I^{2}>$ is the mean-square electron-ion matrix element and $M$ the ionic mass. The
267: McMillan-Hopfield parameter $\eta$ (or $N(E_{F})<I^{2}>$) has been regarded as a local
268: ``chemical'' property of an atom in a crystal. Allen and Dynes \cite{alle} pointed out that
269: $\eta $ is the most significant single parameter in understanding the origin of the high
270: $T_{c}$ of conventional superconductors. For strong coupling systems, variation in $\eta$ is
271: more important than variation of $<\omega^{2}>$ in causing $T_{c}$ to change. Softening
272: of $<\omega^{2}>$ often does enhance $T_{c}$, but very high $T_{c}$ should be caused more by
273: large $\eta$ than by small $<\omega^{2}>$.
274:
275:
276: The logarithmic volume derivative of $\lambda$, $\varphi$, is then obtained
277: \begin{equation}
278: \varphi=\frac{\partial \ln \eta}{\partial \ln V} + 2\gamma _{G} \equiv S+2\gamma _{G}~~.
279: \end{equation}
280: In order to understand how the electronic contribution $\eta=N(E_{F})<I^{2}>$ varies with volume,
281: we use the Gaspari-Gyoriffy theory \cite{gasp} for $\eta$, i.e.,
282: \begin{equation}
283: \eta=\frac{k_{F}^{2}}{\pi^{2}N(E_{F})}\sum_{l}\frac{2(l+1)\sin^{2}(\delta_{l+1}-
284: \delta_{l})N_{l}N_{l+1}}{N_{l}^{1}N_{l+1}^{1}}~~,
285: \end{equation}
286: where $N_{l}$ is the $l$th angular momentum component of the density of states, $N_{l}^{1}$ is the
287: $l$th component of the single scatterer density of states evaluated at $E_{F}$, and $\delta_{l}$
288: the phase shift.
289:
290:
291: For simple metals, the scatterers are assumed to be weak. We can take $N_{l}=N_{l}^{1}$ and
292: approximate $\sin^{2}(\delta_{l+1}-\delta_{l})$ by $(\delta_{l+1}-\delta_{l})^{2}$. Eq. (20) is
293: rewritten as
294: \begin{equation}
295: \eta=\frac{k_{F}^{2}}{\pi^{2}N(E_{F})}\sum_{l}2(l+1)(\delta_{l+1}-\delta_{l})^{2}~~.
296: \end{equation}
297: This expression is identical to the pseudopotential formula of McMillan.\cite{mcmi,evan} Assuming
298: that the phase shift $\delta_{l}$ does not vary very much under pressure for simple $sp$
299: superconductors, we then obtain
300: \begin{equation}
301: S=-\gamma_{N}-\frac{2}{3}~~.
302: \end{equation}
303: The form in Eq. (22) is the same as that of Baryakhtar and Makarov,\cite{bary} who used the
304: constant of the electron-phonon interaction of Fr\"{o}hlich and Mitra.\cite{floh} The expression
305: is an improvement over the expressions of $S$=0 and $S=-$4/3 obtained by Olsen $et$ $al.$
306: \cite{olsen1} and Seiden,\cite{seid} respectively. It is interesting to notice that
307: substitution of $\gamma _{N}$=2/3 into Eq. (22) yields $S=-4/3$. Eq. (22) reduces to the
308: expression of Seiden,\cite{seid} who modified McMillan's expression for $\lambda$ somewhat by
309: considering the effects of a real lattice spectrum as opposed to the Jellium model. Since the
310: electronic Gr\"{u}neisen parameter $\gamma _{e}$ usually varies among different metals even in
311: the simple non-transition elements,\cite{palm,barr,flet} we believe that Eq. (22) should
312: provide a more reasonable value of $S$ compared with Seiden's formula.
313:
314:
315: \section{RESULTS AND DISCUSSION}
316:
317:
318: Using the experimental value of $T_{c}$=39.25 K (Refs.\ \onlinecite{tomi,lore2}), and the
319: theoretical estimates of $\lambda$=0.87 and $\mu^{*}$=0.10 (Ref.\ \onlinecite{kong}), we got
320: $\Theta_{D}$=860 K from Eq. (1) for MgB$_{2}$. We believe that all these parameters, which will
321: enter our calculations, are reliable. For example, the inelastic neutron scattering measurements
322: \cite{osbo} provide an estimate of $\lambda\sim $0.9, which is close to that we used. The
323: calculated value of $\Theta_{D}$=860 K is in the range from 746 to 1050 K determined from the
324: specific heat measurements.\cite{krem,walt,bouq}
325:
326:
327: We took the structural parameters $B_{0}$=147.2 GPa and $B_{0}^{\prime}$=4 from the measurements
328: under the pure hydrostatic pressures up to 0.62 GPa (Ref.\ \onlinecite{jorg}) and under high
329: pressures up to 15 GPa (Refs.\ \onlinecite{vogt,gonc}), respectively. To our knowledge no
330: inelastic neutron scattering or tunneling data exist for MgB$_{2}$ under hydrostatic pressure.
331: We have to use Eq. (12) or (13) for estimating the lattice Gr\"{u}neisen parameter $\gamma_{G}$.
332: The measurements of heat capacity\cite{swif} give a $C_{p}$ of 47.80 J/(K mol) at $T$=298.16 K.
333: $V_{m}$=1.75$\times 10^{-5}$ m$^{3}$/mol, $\kappa _{V}$=6.79$\times 10^{-12}$ Pa$^{-1}$, and
334: $\alpha _{V}$=2.22$\times 10^{-5}$ K$^{-1}$ can be drawn from the neutron diffraction data.
335: \cite{jorg} We therefore obtained $\gamma_{G}$=1.2 by using Eq. (12). Based on the first-order
336: Murnaghan equation for $V(P)$ and the Slater expression of Eq. (13), we got a somewhat larger
337: $\gamma_{G}$ of 1.83 compared to that from Eq. (12). For most simple metals, there is no much
338: difference between the room temperature lattice Gr\"{u}neisen parameter given through Eq. (12)
339: and the Slater relation.\cite{seid,gsch} It was found that the Slater expression usually can
340: yield the reasonable values of $\gamma_{G}$ for most metals.\cite{moru} The only uncertainty
341: entering Eq. (12) in our calculation comes from the indirect measurements of the linear
342: coefficients of thermal expansion.\cite{jorg} Roundy $et$ $al.$ \cite{roun} reported a value of
343: $\gamma_{G}\approx$2.3 from $ab$ $initio$ calculations, which is close to our calculated
344: $\gamma_{G}^{S}$ according to Eq. (13). Meanwhile, Goncharov $et$ $al.$ \cite{gonc}
345: determined a large $E_{2g}$ mode Gr\"{u}neisen parameter of 2.9$\pm$0.3 from the measurements
346: of Raman spectra under pressure. This value is obviously larger than those derived from Eqs.
347: (12) and (13).
348:
349:
350: \begin{figure}[tbp]
351: \vspace{-1.0cm}
352: \begin{center}
353: \includegraphics[height=9.0cm,width=7.5cm,angle=270]{figure1.eps}
354: \end{center}
355: \caption{ Pressure derivative of $T_{c}$ as a function of the lattice Gr\"{u}neisen parameter
356: $\gamma_{G}$ in MgB$_{2}$. The circles show the calculation from the four different values of
357: $\gamma_{G}$=1.2, 1.83, 2.3, and 2.9. }
358: \end{figure}
359:
360: In the calculations of electronic density of states, Loa and Syassan\cite{loa} found that
361: $N(E_{F})$ decreases with pressure at a rate of $d\ln N(E_{F})/dP=$--3.1$\times$10$^{-3}$
362: GPa$^{-1}$. Combining this calculated value and experimental value of $B_{0}$, we got
363: $\gamma_{N}$=0.46. The volume dependence of $\mu^{*}$ is then derived from (15) once having
364: the values of $\gamma_{G}$ and $\mu$. For simple $sp$ metals, $a^{2}$ has a typical value of
365: 0.4 (Ref.\ \onlinecite{moan}), which yields $\mu=$0.63. The volume dependence of $\lambda$,
366: $\varphi$, is therefore determined from Eqs. (19) and (22). Using $\gamma_{N}$=0.46, we have
367: $S$=--1.13 for MgB$_{2}$, which is smaller in magnitude than --4/3 in Seiden's formula for
368: simple metals.\cite{seid} For the transition metals, Hopfield \cite{hopf} commented that $S$
369: is a relatively constant quantity with a value of about --3.5. The values of
370: $S$=--3.5$\sim$--3.1 obtained by inverting the measured $dT_{c}/dP$ for
371: YNi$_{2-x}$M$_{x}$B$_{2}$C (M=Co and Cu) (Ref.\ \onlinecite{loon}) are comparable to that of
372: the transition metals, but are larger in magnitude compared to that of MgB$_{2}$.
373:
374:
375: With the parameters determined above, we have calculated the pressure derivatives of $T_{c}$ for
376: MgB$_{2}$ by using Eq. (2). In Fig. 1 we plotted $dT_{c}/dP$ as a function of $\gamma_{G}$ in
377: the interested range. It is interesting to note that $\gamma_{G}$ plays a predominant role
378: for the pressure effect of $T_{c}$. For the four different $\gamma_{G}$'s considered here
379: $dT_{c}/dP$ are negative. The values obtained from $\gamma_{G}$=1.83 and 2.3 are --0.78 and
380: --1.12 K/GPa, respectively. These are close to the hydrostatic pressure value of --1.1 K/GPa
381: (Refs.\ \onlinecite{tomi,lore2}). Thus the hydrostatic pressure results can be reproduced in
382: terms of our present model by using the values of $\gamma_{G}$ obtained from either the Slater
383: relation or $ab$ $initio$ calculation. It is difficult to obtain the measured results by using
384: $\gamma_{G}$=1 as suggested by Loa and Syassen.\cite{loa} We noticed that a $\gamma_{G}$ of
385: 2.27 is necessary so as to account for the pressure effect on $T_{c}$ for MgB$_{2}$. As
386: emphasized above, all quantities entering Eq. (12) are experimental values and only $\alpha _{V}$
387: was taken from indirect measurements. Thus it is highly expected to operate the thermal expansion
388: measurement to yield a direct $\alpha _{V}$. The present results indicate that the range from
389: $\gamma_{G}$=1.83 to 2.3 should cover the reasonable choices for the lattice Gr\"{u}neisen
390: parameters.
391:
392:
393: To verify these results, and also to study the behavior of $T_{c}$ as a function of pressure, we
394: have performed explicit calculation based on Eq. (11). The theoretical results in the pressure
395: range from 0 to 1.0 GPa are shown in Fig. 2. The experimental data points of Tomita $et$
396: $al.$ \cite{tomi} and Lorenz $et$ $al.$ \cite{lore2} measured under hydrostatic pressure
397: conditions are also plotted for comparison. It is clearly seen that our calculations agree
398: well with the experiments.
399:
400: \begin{figure}[tbp]
401: \vspace{-1.5cm}
402: \begin{center}
403: \includegraphics[height=9.0cm,width=7.5cm,angle=270]{figure2.eps}
404: \end{center}
405: \caption{ Variation of $T_{c}$ with pressure in the region of 0 to 1.0 GPa of MgB$_{2}$ for
406: $\gamma_{G}$=1.83 and 2.3, respectively. The circles and squares represent the hydrostatic
407: pressure experimental data taken from the works of Tomita $et$ $al.$ \cite{tomi} and Lorenz $et$
408: $al.$,\cite{lore2} respectively. }
409: \end{figure}
410:
411:
412: In Fig. 3 we presented the calculated results as well as the experimental data points of
413: Monteverde $et$ $al.$ \cite{mont} and Deemyad $et$ $al.$\cite{deem} measured in the relatively
414: high pressure region. Here we assume that phase transitions of all kinds do not occur under
415: pressure range that we consider. We noticed that the experimental data points of Deemyad $et$
416: $al.$\cite{deem} and the sample 4 of Monteverde $et$ $al.$\cite{mont} are situated well between
417: the two theoretical curves calculated by using $\gamma_{G}$=1.83 and 2.3, respectively.
418: Interestingly, the agreement between our theoretical curve calculated by using $\gamma_{G}$=1.83
419: and the experimental data points of other samples of Monteverde $et$ $al.$\cite{mont} is seen to
420: be reasonable, although there are some scatters among different samples and the reason is not
421: clear. Furthermore, although the pressure measurements are limited to the region below 33 GPa,
422: it is seen from the inset of Fig. 3 that Eq. (11) continues to describe the pressure dependence
423: of $T_{c}$ as high as 100 GPa. Even at this point, the superconductivity is not destroyed by
424: pressure in newly discovered superconductor MgB$_{2}$. There was a discrepancy on whether pressure
425: can destroy superconductivity.\cite{seid,bo,olsen2,smithc} However, our results support the
426: conclusion of Olsen and collaborators\cite{bo,olsen2} that the possibility of destruction of
427: superconductivity by the application of sufficiently high pressure most likely does not exist.
428: It follows from the comparison of Figs. 2 and 3 that the the pressure effect on $T_{c}$ indeed
429: provides a support to the electron-phonon mediated superconductivity in MgB$_{2}$.
430:
431: \begin{figure}[tbp]
432: \begin{center}
433: \includegraphics[width=8.2cm,height=6.5cm,angle=0]{figure3.eps}
434: \end{center}
435: \caption{ Pressure dependence of the transition temperature in MgB$_{2}$ up to 40 GPa. Experimental
436: data are from the works of Monteverde $et$ $al.$\cite{mont} and Deemyad $et$ $al.$\cite{deem} The inset
437: is a calculation of $T_{c}$ under pressure up to 100 GPa. }
438: \end{figure}
439:
440: In Fig. 4 we presented the normalized $\lambda$ and $\mu^{*}$ as a function of pressure up to
441: 30 GPa, calculated from Eq. (10) by using $\gamma_{G}$=2.3. The Coulomb pseudopotential
442: $\mu^{*}$ increases slightly with pressure. Whereas $\lambda$ changes significantly with
443: pressure. The contribution from $\mu^{*}$(P) to the variation of $T_{c}$ with pressure is
444: much less important than that of $\lambda$(P). Thus in the range from 0 to 30 GPa the
445: pressure effect of $T_{c}$ for MgB$_{2}$ is dominated by the competition of $\lambda$ and
446: $\Theta _{D}$ (or $<\omega^{2}>^{1/2}$).
447:
448: \begin{figure}[tbp]
449: \vspace{-1.0cm}
450: \begin{center}
451: \includegraphics[height=9.0cm,width=7.5cm,angle=270]{figure4.eps}
452: \end{center}
453: \caption{ Pressure dependence of normalized Coulomb ($\mu^{*}$) and electron-phonon coupling
454: ($\lambda$) coupling strengths in MgB$_{2}$ calculated by using $\gamma_{G}$=2.3. }
455: \end{figure}
456:
457: \begin{table}[b]
458: \caption{Pressure dependences of superconducting state parameters in MgB$_{2}$. The units of
459: $d\ln X/dP$ (X=$T_{c}$, $\Delta_{0}$, are $H_{c}(0)$) are in 10$^{-2}$ GPa$^{-1}$. }
460: \label{table. 1}
461: \begin{ruledtabular}
462: \begin{tabular}{cccccc}
463: $\gamma_{G}$ & $\gamma_{e}$ & $\frac{d\ln N(E_{F})v}{d\ln V}$ & $\frac{d\ln
464: T_{c}}{dP}$ & $ \frac{d \ln \Delta_{0}}{dP}$ & $\frac{d\ln H_{c}(0)}{dP}$ \\
465: \hline
466: 1.83 & 1.64 & 1.71 & -1.99 & -1.58 & -1.39\\
467: 2.3 & 2.07 & 2.34 & -2.86 & -2.30 & -2.11\\
468: \end{tabular}
469: \end{ruledtabular}
470: \end{table}
471:
472: Table I contains the calculated values of pressure dependences of superconducting parameters for
473: MgB$_{2}$ from Eqs. (5), (6), (8) and (16) by using $\gamma_{G}=1.83$ and $\gamma_{G}=2.3$,
474: respectively. The reliable values of $\gamma_{e}$ is readily determined using Eqs. (16), (19),
475: and (22). We obtained $\gamma_{e}$=1.64 and 2.07 for MgB$_{2}$, which are close to
476: $\gamma_{e}$=1.7 for Pb and for Sn, $\gamma_{e}$=2.0 for Al (Ref.\ \onlinecite{smithc}). The
477: negative sign for $d\ln H_{c}(0)/dP$ predicted for MgB$_{2}$ is in agreement with the
478: measurements for all simple elements superconductors with the exception of thallium.\cite{levyo}
479: For the simple $sp$ metals superconductors, Rohrer\cite{rohrer} has demonstrated that
480: $d\ln N(E_{F})v/d\ln V$ must have approximately a value of 2.0. However, it was realized
481: \cite{levyo} that the transition metals fail to show such simple behavior. Our estimated
482: $d\ln N(E_{F})v/d\ln V$=1.71 and 2.34 for MgB$_{2}$ are comparable to those obtained for simple
483: $sp$ metals superconductors.\cite{levyo,rohrer} Early measurements for most simple metals
484: \cite{trof,jpf,zava} show that there is difference between the quantities in $d\ln \Delta_{0}/dP$
485: and $d\ln T_{c}/dP$. This can be understood with the aid of the results by Ge\v{i}likman and
486: Kresin,\cite{beili} that is,
487: $2\Delta_{0}/k_{B}T_{c}=3.52[1+5.3(T_{c}/\omega_{ph})^{2}\ln\omega_{ph}/T_{c}]$. The calculated
488: data of MgB$_{2}$ listed in Table I make it possible to support this theory. Since the phonon
489: spectrum shifts under pressure, it follows that for all superconductors with
490: $2\Delta_{0}/k_{B}T_{c}>3.52$ a change of $2\Delta_{0}/k_{B}T_{c}$ under pressure can be
491: expected. It is interesting from the viewpoint of experiment to investigate the tunnel
492: characteristics of MgB$_{2}$ under hydrostatic pressure.
493:
494: \begin{table}[b]
495: \caption{Experimental values of $d\ln N(E_{F})v/d\ln V$, isotope effect exponent $\alpha$ and
496: its derivative $\xi=1-2\alpha$ in nine simple $sp$ metals superconductors. }
497: \label{table. 2}
498: \begin{ruledtabular}
499: \begin{tabular}{ccccc}
500: Element & Z & $\frac{d\ln N(E_{F})v}{d\ln V}$ & $\alpha$ & $\xi$ \\
501: \hline
502: Zn & 2 & 2.0 & 0.37 & 0.26 \\
503: Cd & 2 & 2.9 & 0.32 & 0.36 \\
504: Hg($\alpha$) & 2 & 1.7 & 0.50 & 0 \\
505: Al & 3 & 3.4 & 0.325 & 0.35 \\
506: Ga & 3 & 1.8 & 0.41 & 0.18 \\
507: In & 3 & 2.3 & 0.466 & 0.068 \\
508: Tl & 3 & 0 & 0.49 & 0.02 \\
509: Sn & 4 & 2.3 & 0.47 & 0.06 \\
510: Pb & 4 & 2.1 & 0.478 & 0.044\\
511: \end{tabular}
512: \end{ruledtabular}
513: \end{table}
514:
515: \begin{figure}[tbp]
516: \vspace{-1.0cm}
517: \begin{center}
518: \includegraphics[height=9.0cm,width=7.5cm,angle=270]{figure5.eps}
519: \end{center}
520: \caption{Relation between the logarithmic volume derivative of $N(E_{F})v$ and the deviation
521: $\xi$ from the full isotope effect exponent $\alpha=0.5(1-\xi)$ in nine simple $sp$ metals
522: superconductors and MgB$_{2}$. The two squares are the values for MgB$_{2}$. }
523: \end{figure}
524:
525: Figure 5 is a plot of $d\ln N(E_{F})v/d\ln V$ versus the deviation $\xi$ from the full isotope
526: effect for nine simple $sp$ metals superconductors as well as MgB$_{2}$. The experimental values
527: of $d\ln N(E_{F})v/d\ln V$ for simple metals are chosen from the work of Olsen, Andres, and
528: Geballe.\cite{olsen1} The experimental results for isotope effect exponent $\alpha$ and its
529: deviation $\xi$ are taken from the works in Refs. [\ \onlinecite{palm,maxw,hake,fass1,fass2,meser}].
530: There are no experimental data available for Al and In now, we took the calculated results from
531: Leavens and Carbotte.\cite{leav} We summarized these results for simple $sp$ metals in Table II.
532: Olsen $et$ $al.$ \cite{olse1} suggested that $d\ln N(E_{F})v/d\ln V$ is related to the isotope
533: effect exponent $\alpha$ in metals superconductors. As seen from Fig. 5 the relation between
534: $d\ln N(E_{F})v/d\ln V$ and $\xi$ is not very clear when more data included. An isotope effect
535: $\alpha$=0.32(1) (Ref.\ \onlinecite{hink}) in MgB$_{2}$ is consistent with what appears to
536: be a systematic variation of $\alpha$ across the non-transition elements. It is well known that
537: deviations of the isotope effect exponent from $1/2$ are a measure of the relative strengths
538: of the Coulomb and phonon-mediated electron-electron interactions. It is indicated, from the
539: compared values of $\xi$ for MgB$_{2}$ (Ref.\ \onlinecite{hink}) with Zn (Ref.\
540: \onlinecite{fass1}), Cd (Ref.\ \onlinecite{palm}), and Al (Ref.\ \onlinecite{leav}), that
541: MgB$_{2}$ should be a medium coupling superconductor.
542:
543:
544: \section{SUMMARY AND CONCLUSIONS}
545:
546:
547: The major conclusions given by present investigation can be summarized as follows:
548:
549:
550: (i) A simple expression was derived for the pressure dependences of superconducting properties in
551: simple $sp$ superconductor on the basis of McMillan equation. The logarithmic volume derivatives
552: of $\lambda$, $\mu^{*}$, and $\Theta _{D}$ can be self-consistently determined from experiments
553: and theories. We gave an expression for $\varphi$ from the theory of Gaspari and Gyorffy.
554: \cite{gasp} The theory of Morel and Anderson \cite{moan} was used to obtain $\phi$, which makes
555: it possible to investigate the pressure dependence of $\mu^{*}$. Neglecting the pressure
556: dependence of $\mu^{*}$, the present theoretical model can be reduced to the two popular models
557: of Seiden \cite{seid} and Baryakhtar and Makarov \cite{bary} when taking $\gamma_{N}=$ 2/3 and
558: neglecting the direct electron-electron interaction, respectively. Furthermore, we obtained an
559: explicit expression for the change of $T_{c}$ as a function of pressure with the help of
560: Murnaghan equation. The present model enables us to study the pressure behaviors of some
561: interested superconducting parameters such as the zero temperature energy gap $\Delta_{0}$, the
562: critical field at absolute zero temperature $H_{c}(0)$, the effective interaction strength
563: $N(E_{F})v$, and the electronic specific heat coefficient $\gamma$.
564:
565:
566: (ii) We investigated the pressure effects on superconducting properties in the newly discovered
567: superconductor MgB$_{2}$ using our simple approach. It was found that the hydrostatic pressure
568: derivative of $T_{c}$ can be reproduced by using the values of $\gamma_{G}$ obtained from
569: either the Slater relation or $ab$ $initio$ calculation. The calculated
570: $d\ln N(E_{F})v/d\ln V\approx$2.0 in MgB$_{2}$ is close to those obtained in simple $sp$
571: superconductors. The quantitative agreement for the variation of $T_{c}$ with pressure in the
572: low pressure region as well as high pressure region is very good when comparing our theoretical
573: results with experimental data measured by three groups. The predicted values of
574: $d\ln H_{c}(0)/dP$, $d\ln \Delta_{0}/dP$, and $\gamma_{e}$ are also comparable to those in simple
575: $sp$ metals superconductors. All these characteristic pressure behaviors allow us to conclude
576: that MgB$_{2}$ should be a simple electron-phonon mediated $sp$ superconductor and the mechanism
577: in simple $sp$ metals superconductors is also responsible for the superconductivity in MgB$_{2}$.
578:
579: \begin{acknowledgments}
580: The authors acknowledge useful discussions with O. K. Andersen, O. Jepsen, Y. Kong, R. K.
581: Kremer, and K. Syassen. We are indebted to J. S. Schilling and J. D. Jorgensen for allowing us
582: to use their experimental data prior to publication as well as their valuable comments on the
583: manuscript. XJC thanks the MPG for financial support.
584: \end{acknowledgments}
585:
586: {\it Note added.--} After submission of this manuscript, the authors have learned that
587: the superconductivity is not destroyed up to 44 GPa where $T_{c}$ is still as high as 12 K.
588: \cite{struz} The intrinsic $dT_{c}/dP\approx -1.1$ K/GPa under hydrostatic pressure
589: conditions has recently been reported by other three groups.\cite{struz,schla,tang} We also
590: have learned a possible explanation given by Tissen $et$ $al.$\cite{tissen} for the large
591: $-dT_{c}/dP$ observed in the low $T_{c}=37.4\pm0.1$ samples.
592:
593:
594: \begin{thebibliography}{99}
595:
596: \bibitem{naga} J. Nagamatsu, N. Nakagawa, T. Muranaka, Y. Zenitanl, and J. Akimitsu, Nature
597: (London) {\bf410}, 63 (2001).
598:
599: \bibitem{eom} C. B. Eom, M. K. Lee, J. H. Choi, L. J. Belenky, X. Song, L. D. Cooley, M. T.
600: Naus, S. Patnaik, J. Jiang, M. Rikel, A. Polyanskii, A. Gurevich, X. Y. Cai, S. D. Bu, S. E.
601: Babcock, E. E. Hellstrom, D. C. Larbalestier, N. Rogado, K. A. Regan, M. A. Hayward, T. He,
602: J. S. Slusky, K. Inumaru, M. K. Haas, and R. J. Cava, Nature (London) {\bf411}, 558 (2001).
603:
604: \bibitem{jin} S. Jin, H. Mavoori, C. Bower, and R. B. van Dover, Nature (London) {\bf411},
605: 563 (2001).
606:
607: \bibitem{olse1} J. L. Olsen, E. Bucher, M. Levy, J. Muller, E. Corenzwit, and T. Geballe, Rev.
608: Mod. Phys. {\bf36}, 168 (1964).
609:
610: \bibitem{bud} S. L. Bud'ko, G. Lapertot, C. Petrovic, C. E. Cunningham, N. Anderson, and P. C.
611: Canfield, Phys. Rev. Lett. {\bf86}, 1877 (2001).
612:
613: \bibitem{hink} D. G. Hinks, H. Claus, and J. D. Jorgensen, Nature (London), {\bf411}, 457 (2001).
614:
615: \bibitem{palm} C. Palmy, Phys. Lett. A {\bf29}, 373 (1969).
616:
617: \bibitem{sst} T. J. Sato, K. Shibata, and Y. Takano, cond-mat/0102468 (2001).
618:
619: \bibitem{osbo} R. Osborn, E. A. Goremychkin, A. I. Kolesnikov, and D. G. Hinks, Phys. Rev. Lett.
620: {\bf87}, 017005 (2001).
621:
622: \bibitem{kara} G. Karapetrov, M. Iavarone, W. K. Kwok, G. W. Crabtree, and D. G. Hinks, Phys.
623: Rev. Lett. {\bf86}, 4374 (2001).
624:
625: \bibitem{kote} H. Kotegawa, K. Ishida, Y. Kitaoka, T. Muranaka, and J. Akimitsu, Phys. Rev. Lett.
626: {\bf87}, 0127001 (2001).
627:
628: \bibitem{krem} R. K. Kremer, B. J. Gibson, and K. Ahn, cond-mat/0102432 (2001).
629:
630: \bibitem{walt} Ch. W\"{a}lti, E. Felder, C. Degen, R. Monnier, B. Delley, and H. R. Ott, Phys.
631: Rev. B {\bf64}, 172515 (2001).
632:
633: \bibitem{bouq} F. Bouquet, R. A. Fisher, N. E. Phillips, D. G. Hinks, and J. D. Jorgensen, Phys.
634: Rev. Lett. {\bf87}, 047001 (2001).
635:
636: \bibitem{mont} M. Monteverde, M. N\'{u}\~{n}ez-Regueiro, N. Rogado, K. A. Regan, M. A. Hayward,
637: T. He, S. M. Loureiro, and R. J. Cava, Science {\bf292}, 75 (2001).
638:
639: \bibitem{lore1} B. Lorenz, R. L. Meng, and C. W. Chu, Phys. Rev. B {\bf64}, 012507 (2001).
640:
641: \bibitem{sait} E. Saito, T. Taknenobu, T. Ito, Y. Iwasa, K. Prassides, and T. Arima, J. Phys.:
642: Condens. Matter {\bf13}, L267 (2001).
643:
644: \bibitem{tomi} T. Tomita, J. J. Hamlin, J. S. Schilling, D. G. Hinks, and J. D. Jorgensen, Phys.
645: Rev. B {\bf64}, 092505 (2001).
646:
647: \bibitem{lore2} B. Lorenz, R. L. Meng, and C. W. Chu, cond-mat/0104303 (2001).
648:
649: \bibitem{hirs} J. E. Hirsch, Phys. Lett. A {\bf282}, 392 (2001).
650:
651: \bibitem{loa} I. Loa and K. Syassen, Solid State Commun. {\bf118}, 279 (2001).
652:
653: \bibitem{vogt} T. Vogt, G. Schneider, J. A. Hriljac, G. Yang, and J. S. Abell, Phys. Rev. B
654: {\bf63}, 220505 (2001).
655:
656: \bibitem{mcmi} W. L. McMillan, Phys. Rev. {\bf167}, 331 (1968).
657:
658: \bibitem{elia} G. M. Eliashberg, Sov. Phys. JETP {\bf11}, 696 (1960).
659:
660: \bibitem{kort} J. Kortus, I. I. Mazin, K. D. Belashenko, V. P. Antropov, and L. L. Boyer, Phys.
661: Rev. Lett. {\bf86}, 4656 (2001).
662:
663: \bibitem{pick} J. M. An and W. E. Pickett, Phys. Rev. Lett. {\bf86}, 4366 (2001).
664:
665: \bibitem{kong} Y. Kong, O. V. Dolgov, O. Jepsen, and O. K. Andersen, Phys. Rev. B {\bf64}, 020501
666: (2001).
667:
668: \bibitem{hodd} R. Hodd, Phys. Rev. {\bf180}, 530 (1969).
669:
670: \bibitem{trof} P. N. Trofimenkoff and J. P. Carbotte, Phys. Rev. B {\bf1}, 1136 (1970).
671:
672: \bibitem{coul} M. A. Coulthard, J. Phys. F: Metal Phys. {\bf1}, 195 (1971).
673:
674: \bibitem{rott} H. R. Ott and R. S. Sorbello, J. Low Temp. Phys. {\bf14}, 73 (1974).
675:
676: \bibitem{daco} M. M. Dacorogna, M. L. Cohen, and P. K. Lam, Phys. Rev. B {\bf34}, 4865 (1986).
677:
678: \bibitem{zima} J. M. Ziman, {\it Electrons and Phonos} (Clarendon Press, Oxford, 1960).
679:
680: \bibitem{roun} D. Roundy, H. J. Choi, H. Sun, S. G. Louie, and M. L. Cohen (unpublished).
681:
682: \bibitem{gonc} A. F. Goncharov, V. V. Struzhkin, E. Gregoryanz, J. Hu, R. J. Hemley, H.-K. Mao,
683: G. Lapertot, S. L. Bun'ko, and P. C. Canfield, Phys. Rev. B {\bf64}, 100509 (2001).
684:
685: \bibitem{seid} P. E. Seiden, Phys. Rev. {\bf179}, 458 (1969).
686:
687: \bibitem{bo} R. I. Boughton, G. Br\"{a}ndli, J. L. Olsen, and C. Palmy, Helv. Phys. Acta
688: {\bf42}, 587 (1969).
689:
690: \bibitem{gubs} D. U. Gubser and A. W. Webb, Phys. Rev. Lett. {\bf35}, 104 (1975).
691:
692: \bibitem{papa} D. A. Papaconstantopoulos and B. M. Klein, Physica B {\bf107}, 725 (1981).
693:
694: \bibitem{bcs} J. Bardeen, L. N. Cooper, and J. R. Schrieffer, Phys. Rev. {\bf108}, 1175 (1957).
695:
696: \bibitem{moan} P. Morel and P. W. Anderson, Phys. Rev. {\bf125}, 1263 (1962).
697:
698: \bibitem{leav} C. R. Leavens and J. P. Carbotte, Can. J. Phys. {\bf49}, 724 (1971).
699:
700: \bibitem{jpf} J. P. Franck and W. J. Keeler, Phys. Rev. Lett. {\bf20}, 379 (1968).
701:
702: \bibitem{keel} J. P. Franck and W. J. Keeler, Phys. Lett. A {\bf25}, 624 (1967).
703:
704: \bibitem{yama} J. Yamashita and S. Asano, Jpn. J. Phys. Soc. {\bf29}, 264 (1970).
705:
706: \bibitem{zava} N. V. Zavaritski\v{i}, E. S. Itskevich, and A. N. Voronovski\v{i}, Sov. Phys. JETP
707: {\bf33}, 762 (1971).
708:
709: \bibitem{lech} R. Lechner and G. Quittner, Phys. Rev. Lett. {\bf17}, 1259 (1966).
710:
711: \bibitem{slat} J. C. Slater, {\it Introduction to Chemical Physics} (McGraw-Hill, New York, 1939).
712:
713: \bibitem{migdal} A. B. Migdal, Sov. Phys. JETP {\bf7}, 996 (1958).
714:
715: \bibitem{barr} T. H. K. Barron, J. G. Collins, and G. K. White, Adv. Phys. {\bf29}, 609 (1980).
716:
717: \bibitem{fawc} E. Fawcett and G. K. White, J. Appl. Phys. {\bf39}, 576 (1968).
718:
719: \bibitem{flet} G. C. Fletcher and M. Yahaya, J. Phys. F: Metal Phys. {\bf9}, 1529 (1979).
720:
721: \bibitem{alle} P. B. Allen and R. C. Dynes, Phys. Rev. B {\bf12}, 905 (1975).
722:
723: \bibitem{gasp} G. D. Gaspari and B. L. Gyorffy, Phys. Rev. Lett. {\bf28}, 801 (1972).
724:
725: \bibitem{evan} R. Evans, G. O. Gaspari, and B. L. Gyorffy, J. Phys. F: Metal Phys. {\bf3}, 39
726: (1973).
727:
728: \bibitem{bary} V. G. Baryakhtar and V. I. Makarov, Sov. Phys. JETP {\bf22}, 1320 (1966).
729:
730: \bibitem{floh} H. Fr\"{o}hlich and T. K. Mitra, Proc. Phys. Soc. {\bf1}, 544 (1968).
731:
732: \bibitem{olsen1} J. L. Olsen, K. Andres, and T. H. Gebelle, Phys. Lett. A {\bf26}, 239 (1968).
733:
734: \bibitem{jorg} J. D. Jorgensen, D. G. Hinks, and S. Short, Phys. Rev. B {\bf63}, 224522 (2001).
735:
736: \bibitem{swif} R. M. Swift and D. White, J. Am. Chem. Soc. {\bf79}, 3641 (1957).
737:
738: \bibitem{gsch} K. A. Gschneidner, Jr., in {\it Solid State Physics}, edited by F. Seitz and D.
739: Turnbull (Academic Press Inc., New York, 1964), Vol. 16.
740:
741: \bibitem{moru} V. L. Moruzzi, J. F. Fanak, and K. Schwarz, Phys. Rev. B {\bf37}, 790 (1988).
742:
743: \bibitem{hopf} J. J. Hopfield, Physica {\bf55}, 41 (1971).
744:
745: \bibitem{loon} C. Looney, A. K. Gangopadhyay, and A.-K. Klehe, and J. S. Schilling, Physica C
746: {\bf252}, 199 (1995).
747:
748: \bibitem{deem} S. Deemyad, J. S. Schilling, J. D. Jorgensen, and D. G. Hinks (unpublished).
749:
750: %Physica C {\bf361}, 227 (2001); J. S. Schilling (private communication).
751:
752: \bibitem{olsen2} J. L. Olsen and H. Rohrer, Helv. Phys. Acta {\bf33}, 872 (1960); M. Levy and J.
753: L. Olsen, Solid State Commun. {\bf2}, 137 (1964).
754:
755: \bibitem{smithc} T. F. Smith and C. W. Chu, Phys. Rev. {\bf159}, 353 (1967).
756:
757: \bibitem{levyo} M. Levy and J. S. Olsen, in {\it Physics of High Pressures and the Condensed
758: Phase}, edited by A. Van Itterbeek (North-Holland Pub., Amsterdam, 1965).
759:
760: \bibitem{rohrer} H. Rohrer, Helv. Phys. Acta {\bf33}, 675 (1960).
761:
762: \bibitem{beili} B. T. Ge\v{i}likman and V. Z. Kresin, Sov. Phys. Solid State {\bf7}, 2659
763: (1966).
764:
765: \bibitem{maxw} E. Maxwell, Phys. Today {\bf5}, No. 12, 14 (1952).
766:
767: \bibitem{hake} R. R. Hake, D. E. Mapother, and D. L. Decker, Phys. Rev. {\bf112}, 1522 (1958).
768:
769: \bibitem{fass1} R. E. Fassnacht and J. R. Dillinger, Phys. Rev. {\bf164}, 565 (1967).
770:
771: \bibitem{fass2} R. E. Fassnacht and J. R. Dillinger, Phys. Lett. A {\bf28}, 741 (1969).
772:
773: \bibitem{meser} R. Meservey and B. B. Schwartz, in {\it Superconductivity}, edited by
774: R. D. Parks (M. Dekker Inc., New York, 1969).
775:
776: \bibitem{struz} V. V. Struzhkin, A. F. Goncharov, R. J. Hemley, H.-K. Mao, G. Lapertot, S.
777: L. Bud'ko, and P. C. Canfield, cond-mat/0106576 (2001).
778:
779: \bibitem{schla} S. I. Schlachter, W. H. Fietz, K. Grube, and W. Goldacker, cond-mat/0107205
780: (2001).
781:
782: \bibitem{tang} J. Tang, L. C. Qin, A. Matsushita, Y. Takano, K. Togano, H. Kito, and H.
783: Ihara, Phys. Rev. B {\bf64}, 132509 (2001).
784:
785: \bibitem{tissen} V. G. Tissen, M. V. Nefedova, M. N. Kolesnikov, and M. P. Kulakov, Physica C
786: {\bf363}, 194 (2001).
787:
788: \end{thebibliography}
789:
790:
791: \end{document}
792: