1: \documentstyle[preprint,aps]{revtex}
2: %\documentstyle[12pt,emlines]{article}
3: %\documentstyle[12pt]{artrus}
4: %\textwidth 159mm
5: %\textheight 190mm
6: %\topmargin -10mm
7: %\oddsidemargin -0.1mm
8: %\evensidemargin -0.1mm
9: %\renewcommand {\baselinestretch} {1.1}
10: %\input tcilatex
11:
12: \begin{document}
13: \author{ M.E. PALISTRANT}
14:
15: \title{SUPERCONDUCTIVITY IN TWO-BAND SYSTEM WITH LOW CARRIER DENSITY}
16: \maketitle
17:
18: \begin{center}
19: {\it Institute of Applied Physics, Academy of Sciences of Moldova,\\
20: 5 Akademiei Street, 2028 Chishinau, Moldova}
21: \end{center}
22:
23: \begin{center}
24: \sf{Received 25 octouber 2003}
25: \end{center}
26:
27: \begin{quotation}
28:
29: \sf{In the article a review over the theory of superconductors with energy bands
30: that overlap on Fermi surface at arbitrary densities of charge carriers
31: (including reduced and very low) is done. All pairings of electrons that
32: result in formation Cooper pairs of electrons from different energy bands as
33: well as in every energy band are considered.
34:
35: The system of equations for four order parameters $\Delta _{11}$, $\Delta
36: _{22}$, $\Delta _{12}$, $\Delta _{21}$ and chemical potential $\mu $ is
37: derived. Self-consistent approach is strictly necessary at $\mu \sim \Delta
38: _{nm}$ . Transition to the effective four-band model leading to the
39: temperature of superconducting transition is performed. Analytic and numeric
40: solutions have been performed for two mechanisms of superconducting
41: pairing -- non-phonon and phonon one. The account of the additional in
42: regards to Moscalenco model pairings of electrons from different bands or
43: their hybridization gives an option to obtain the high values of $T_{C}$
44: even at reduced density of charge carriers. This hybridization influences
45: greatly the form of the dependence of the quantity $T_{C}$ on the density of
46: charge carriers as well as the jump of the electron heat capacity at the
47: point $T=T_{C}$.
48:
49: High and low values of the relative jump of electronic heat capacity at the
50: point $T=T_{C}$ $\left( \frac{\textstyle C_{S}-C_{N}}{\textstyle C_{N}}
51: >1,43\right. $ and $\,\left. \frac{\textstyle C_{S}-C_{N}}{\textstyle C_{N}}
52: <1,43\right) $ are obtained with varying the density of charge carriers.
53:
54: These four effective energy bands in the considered system favors
55: essentially in experimental observation of the kink of the chemical
56: potential $\mu \left( T\right) $ at the point $T=T_{C}$ .
57:
58: The mechanism of superconductivity (non-phonon and phonon) is shown to
59: affect differently with varying the density of charge carriers as well as
60: the dependence of $T_{C}$ and on this density.
61:
62: Possible peculiarities in behavior of thermodynamic quantities in $MgB_{2}$
63: compound with reducing the density of charge carriers are predicted.
64:
65: Influence of the overlapping energy bands on Fermi surface at $T=0$ and very
66: low densities of charge carriers $\left( \mu <0\right) $ on the
67: superconductivity in the system that is in the state of Bose condensation of
68: localized pairs is studied.
69:
70: An application of the path integral method to the two-band model
71: is developed and on this basis, the process of transition from the Fermi to
72: the Bose pattern of elementary exitations at $T \not = 0$ in the presence of
73: a two-particll bound state in the system is demonstratied.The expression for
74: the temperature of Bose condensation $T_k$ is obtained and the contribution of
75: the residual boson interactions is estimated for systems with different
76: dimensions.
77:
78: The overlappining of the energy bands on the fermi surface is favorable for
79: superconductivity and intensify some peculiarities of thermodinamic
80: properties at low carrier density. This halps to their experimental
81: confirmation.}\\
82: \\
83: \it{Keywords}:\sf{Tow-band superconductor, hybridization, lowered and low
84: carrier density, thermodinamical propeties, magnezium borid, BCS-Bose
85: crossover.}
86:
87: \end{quotation}
88:
89: \begin{center}
90: \bf{1. INTRODUCTION}
91: \end{center}
92:
93: A rich body of experimental and theoretical research material on oxide
94: ceramics has been accumulated since the discovery of high - temperature
95: superconductivity. However, the description of the physical properties of these
96: materials remains one of the most difficult problems in low - temperature
97: physics today. This is because of the complexity of the objects of study: they
98: have a complex crystal structure, strong anisotropy, anomalies in the
99: electronic energy spectrum, a variable concentration of charge carriers, strong
100: electronic correlations, and so on. Models for the analysis of such systems
101: apparently should be based on the Hubbard model, which takes into account the
102: strong electronic correlations due to the Coulomb interaction of the electrons,
103: and should take into account the strong electron - phonon interaction. A
104: review of the different approaches to this problem and the approximations used
105: in them is given in Ref.1, for example. This theory contains dielectric and
106: magnetic phase transitions and the possible onset of superconductivity.
107: However, because of the great mathematical difficulties it is hard to obtain
108: any meaningful physical results without making some substantial
109: simplifications. Moreover, at a certain carrier density a metallic states
110: arises in the system, in which the electronic states are modified but not
111: destroyed by correlations. Consequently, there can be a transition to the
112: superconducting state, with the formation of Cooper pairs (the BCS scenario)
113: or local pairs (the Schafroth scenario). In this connection it is unquestionably
114: of interest to apply Fermi - liquid concepts to the study of the
115: superconducting properties of high - $T_c$ superconductors with allowance for
116: their peculiar features, such as overlapping of the energy bands at the Fermi
117: surface and the presence of various kinds of van Hove - Lifshitz singularities,
118: strong anisotropy, variable carrier density (including small values) etc., when
119: treating both phononic and nonphononic mechanisms of superconductivity.
120:
121: Many theoretical papers have by now been published on various aspects of the
122: two - band model \cite{Moskalenko_2},\cite{Suhl} . The great
123: interest in this model and in its generalizations is due, firstly, to band
124: calculations \cite{Krakauer}, \cite{Herman} showing that in metal - oxide
125: ceramics the energy bands on the Fermi surface overlap (a similar situation
126: obtains, apparently, also in systems with heavy fermions \cite{Baranov}),
127: and secondly to the possibility of using the aforementioned model to describe
128: the properties of systems with two groups of electrons (e.g., layered
129: compounds).
130:
131: The main development stages of the superconducting theory of the systems with
132: overlapping energy bands are presented in comment \cite{Palistrant_1}.
133:
134: The consideration of the overlapping of the energy bands leads not only to the
135: quantitative differenece of results from the case of one - band superconductor,
136: but in some cases to the qualitatively new results. For example:
137:
138: 1) In two - band superconductors high temperature superconductivity is possible
139: not only in the case of attractive interaction between the electrons, but even
140: if the interaction between the electrons has repulsive character $(\lambda_{n
141: m} < 0, n, m = 1 - 2)$, but relation $\lambda_{11} \lambda_{22} - \lambda_{12}
142: \lambda_{21} < 0$ is fulfilled.
143:
144: 2) In impurity two-band superconductors, for example, Anderson theorem is
145: violated at $\Delta_1 \not = \Delta_2$, and appears the dependence of
146: thermodynamical quantities on concentrations of non - magnetic impurity due to
147: the interband scattering of electrons on imputiry atoms.
148:
149: 3) In two - band superconductors appears collective oscillations of excitontype
150: Leggett mode,caused by the fluctuations of phase of order parameters for
151: different bands. In the three - band systems, and also in the two - band with
152: lowered density of charge carrier, which reduces to the the three - band model,
153: such oscillatory modes can be two.
154:
155: 4) On the basis of the theory of superconductivity with overlapping energetical
156: bands one can explane a great number of experimental results in High -$T_c$
157: materials.
158: In particular, by using the two - band model and assuming moderate values of
159: the coupling constants one can obtain high $T_c$, two energy gaps $2\Delta_1
160: /T_c > 3,5$ and $2 \Delta_2 / T_c < 3,5$, large values of negative $d ln T_c /
161: d ln V$ ($V$ is the volume), a positiv curvature of the upper critical field
162: near the transition temperature, and others \cite{Lee_1} - \cite{Volovik}.
163: Furthemore, in the two - band model it is possible to describe the
164: decrease of $T_c$ with increase of the oxygen disorder, as well as when copper
165: atoms are replaced by a nonmagnetic dopant (A1, Zn, etc)\cite{Moskalenko_5} -
166: \cite{Moskalenko_7}.
167:
168: 5). An important role in the determination of the thermodynamic and magnetic
169: properties of a two-band superconductor is played also by the location of the
170: Fermi level, which is changed by doping or by introduction of oxygen.
171: Having assumed a non-phonon pairing mechanism of superconductivity as
172: well as phonon mechanism in the multi-band systems with lowered densities of
173: charge carriers the account of the peculiarities mentioned above is very
174: crucial.Particular interest attaches to the possibility of obtaining a
175: bell-shaped dependence of $T_c$ and of the heat-capacity discontinuity $(C_S -
176: C_N)$ at the point $T = T_c$ on the carrier density \cite{Palistrant_2}-
177: \cite{Palistrant_4}. In the three-band model with a nonphonon sureconductivity
178: mechanism it is possible to obtain the ''step'' which is observed for $Y_1 Ba_2
179: Cu_3 O_{7 - \delta}$ in the dependence of $T_c$ on the carrier density
180: \cite{Kalalb_1}, \cite{Palistrant_5}. An investigation of the properties of
181: high-temperature ceramics, based on allowance for energy -band overlap of the
182: energy bands and for electronic topological transitions, was reviewed in Ref
183: \cite{Moskalenko_8}. This review contains the classical achievements of the
184: problem.On can find there references to experimental research results that can
185: be described by allowance for the singularitiesin the electron energy spectrum
186: of complex systems.
187:
188: An increase of the number of energy bands on the Fermi surface increases the
189: overall electron-state density and leads to the onset of an additional
190: interband electron-electron interaction that contributes to the onset of
191: superconductivity. This interaction violates the universal BCS relations and
192: leads to a substantial dependence of a number of physical characteristics on
193: the properties of an anisotropic system \cite{Moskalenko_8}-\cite{Hirch}.
194:
195: The discovery \cite{Nagamatsu} of superconductivity in $MgB_2$ with a high
196: transition temperature $T_c \sim 39 K$ has attracted considerable attention.
197: The theoretical investigations \cite{Bud'ko}-\cite{Choi_3} as well
198: experimental observations \cite{Buzea}, \cite{Ponomarev} have led to
199: the conclution that $MgB_2$ is multi-band superconductor with fonon-mediated
200: BCS superconductivity mechanism.
201:
202: Therefore, the known classical results for the system with
203: overlapping energy bands can be used as a basis to describe the
204: superconducting properties of $MgB_{2}$.
205:
206: Nowadays, many authors rediscover unfortunately well-known already
207: results, obtained early on the basis of Moskalenko \cite{Moskalenko_2}
208: two-band model.At the same time the method, taking into account the
209: overlapping ennergy bands as well as the anisotropy of energy gap, is developed
210: in a number of studies (see, for exampl \cite{Mishonov_1} -
211: \cite{Mishonov_3}). This method gives a good option to incorporate the
212: theoretical results to the experimental data about the specific heat in
213: $MgB_{2}$ .
214:
215: As mentioned above, an interesting feature of the two-band model is
216: independence of the superconducting-transition temperature of the sign of the
217: inter-and intraband interaction constants. This model can therefore be used for
218: the usual electron-phonon mechanism of superconductivity as well as for a
219: mechanism based on the repulsion between carriers. In all the references cited
220: above, the two-band model can be used to describe the properties of
221: superconductors for which the relation $\mu \gg T_c$ is satisfied ($\mu$ is the
222: chemical potential). This relation is satisfied in a number of cases also in
223: high- temperature compound. The existence of this relation between $\mu$ and
224: $T_c$ makes it possible to use in the calculations an approximation diagonal in
225: the band indices,\cite{Moskalenko_12},\cite{Moskalenko_13} which leads to
226: neglect of off-diagonal paramerers such as $\Delta_{12}$ and $\Delta_{21}$.
227:
228: In systems with low carrier density, however, the relation $\mu \gg T_c$ does
229: not hold. It becomes therefore necessary to develop a superconductivity theory
230: for two - band systems, without constraints on the Fermi energy. This is the
231: task of the present paper. We consider simultaneously two possible
232: superconductivity mechanisms - phonon and electron. A characteristic feature of
233: systems with low carrier density is a substantial dependence of the chemical
234: potential $\mu$ on the order parameter in the superconducting phase. This
235: circumstances has been noted in many papers, and the feasibility of
236: expererimentally observing anomalies in the temperature dependence of the
237: chemical potential was first suggest in Ref.40. It was shown there, with the
238: BCS model as the example, the $\mu (T)$ curve has an experimentally observable
239: bend at the point $T = T_c$. We shall show below that in the two - band case
240: this effect is enhanced by the presence of four order parameters
241: $(\Delta_{nm}; m = 1,2)$ and is manifested at $\mu$ values easier to observe
242: in experiment.
243:
244: In the articles referred to above, investigations were conducted on the basis
245: of the idea of Cooper pairing. In systems with small carrier concentrations,
246: such as semiconductors or metallic oxides, bound states
247: may arise following a decrease in the carrier concentration, and a transition
248: to a Bose condensate of local pairs with a final binding energy may occur (the
249: Schaffroth scenario \cite{Schaffroth}). The issue of such transitions
250: in one-band systems was discussed in a number of articles \cite{Migake} -
251: \cite{Loktev}.
252:
253: To realize the Schaffroth scenario, a bound two-particle state
254: \cite{Leggett_2} must exist in the system. The appearance of this state in
255: the presence of attractive interaction depends on the dimension of the system
256: \cite{Leggett_2} and is realized in the region of small carrier
257: concentrations. As is shown in \cite{Randeria}, \cite{Gordar} , the
258: change of sign in the chemical potential with a decrease in the carrier
259: concentration corresponds to a transition from the BCS to the Schaffroth
260: scenario. Condensation of local pairs occurs at concentrations of carriers for
261: which $\mu \leq 0$.
262:
263: This review article is devoted to investigations of superconducting ordering
264: in systems with two characteristic features - a small concentration of charge
265: carriers and overlapping energy bands on the Fermi surface. As mentioned
266: above, both of these features are characteristic of superconducting compounds.
267:
268: Our review article is organized as follows. In Sec.2 we prezennt a theory of
269: the superconductivity of two-band system that is valid for any carrier density
270: and takes into account all possible pairing of electrons due to intraband and
271: interband interactions on the basis of the idea of Cooper pairing. The critical
272: temperature $T_c$, the chemikal potential $\mu$, the heat capasity $(C_S -
273: C_N)$ at the point $T = T_c$ as function of carrier density are ilustrated.
274: Section 3 is devoted to a self-consistent discussion in the mean field
275: approximation of a system of equations in oreder parameters $\Delta_n$ and
276: $\mu$ at $T = 0$ and to revealing the influence of energy band overlapping on
277: these quantities and the carrier concentration at which the system experiences
278: a transition from Cooper pairing $(\mu > 0)$ to the Schaffroth $(\mu < 0)$
279: scenario. The equation for the binding energy $\varepsilon_b$ of a two-particle
280: state is also obtained and the relationship between $\varepsilon_b$ and $\mu$
281: is established. In Sec. 4 the path integral method as applied to the two-band
282: model is developed and, on this basis, the procedure for a transition from the
283: Fermi to Bose elementary excitations is given. The condensation temperature of
284: the Bose system $T_k$ is also determined. In the last section, the results of
285: the investigation are summarized.
286:
287: \begin{center}
288: \bf{2. THERMODINAMICAL PROPERTIES ON BASIS OF THE COOPER PAIRS OF THE
289: CHARGE CARRIERS}.
290: \end{center}
291:
292: {\bf 2. 1. System Hamiltonian and basic equations.}
293:
294:
295: The considered two-band system is described by the Hamiltonian \cite{Kochorbe_1}
296: $$
297: H = \sum_{n \vec k \sigma}\left[\varepsilon_n (\vec k) - \mu \right] a_{n \vec
298: k \sigma}^{+}\, a_{n \vec k \sigma} -
299: \frac{1}{V}\,\,\sum_{m_1...m_4,\vec k \vec k'}\,V_{m_1 m_2}^{m_3 m_4}\left(\vec
300: k - \vec {k'}; - \vec k \vec {k'}\right)
301: $$
302: \begin{equation}
303: a_{m_1 \vec k \uparrow} a_{m_2 - \vec k \downarrow} a_{m_3 - \vec k'\downarrow}
304: a_{m_4 \vec k'\uparrow}, \label{1} \end{equation} were $a_{n \vec k
305: \sigma}^{+}$ and $a_{n \vec k \sigma}$ are creation and annihilation
306: operators for a electron band $n$ with spin $\sigma$ and quasiwave vector {\bf
307: k}, $\mu$ is the chemical potential, and $V_{m_1 m_2}^{m_3 m_4}$ are the intra-
308: and interband interaction constants. Expression (1) is a generalization of the
309: BCS - Bogolyubov model Hamiltonian to include the two-band case. Account is
310: taken here of all possible methods of electron pairing within each band and of
311: electron pairing from different bands. If $m_1 = m_2$ and $m_3 = m_4$, the
312: Hamiltonian (1) is equal to that of the Moskalenko model \cite{Moskalenko_2},
313: wich considers only intraband pairing and transitions of Cooper pair as a whole
314: from one band to another are considered; this model is widely used to describe
315: the properties of high-temperature superconductors \cite{Lee_1} - \cite{Hirch}
316: and magnezium deboride \cite{Jiu_1} - \cite{Choi_3}; \cite{Mishonov_1} -
317: \cite{Mishonov_3}.Examination of the more general Hamiltonian (1) uncovers
318: additional possible onsets of superconductivity (on account of single-particle
319: hybridization and of all interband-interaction constants) and makes passible a
320: description of the properties of a system with a low density of states $(\mu
321: \sim T_c)$.Applying the Green's- function method \cite{Abricosov_1} to the
322: Hamiltonian (1) we obtain the set of equations for the order parameters
323: $\Delta_{np}$ \cite{Kochorbe_1}.
324:
325: \begin{eqnarray}
326: \Delta_{np}&=& \frac{1}{4V}\,\sum_{\vec klr, r\not= l}\,V_{np}^{ll} \Delta_{ll}
327: \Bigg[\left(\frac{\xi_{l}^{2} -\xi_{r}^{2} + 2\Delta_{12} \Delta_{21}(1 +
328: 1/z_{(12)}}{d} + 1 \right)\nonumber\\
329: &\times& th \frac{\frac{\beta E_1}{2}}{E_1} - \left(\frac{\xi_{l}^{2}
330: - \xi_{r}^{2} + 2\Delta_{12} \Delta_{21}(1 + 1/z_{(12)}}{d} - 1 \right)
331: th \frac{\frac{\beta E_2}{2}}{E_2}\Bigg]+
332: \nonumber
333: \end{eqnarray}
334: \begin{equation}
335: + \frac{1}{4V}\,\sum_{\vec k}\left(V_{np}^{12}\Delta_{12}\frac{z_{(13)}}
336: {z_{(14)}}+ V_{np}^{21}\Delta_{21}\,\frac{z_{(14)}}{z_{(13)}}\right) \Gamma,
337: \label{2}
338: \end{equation}
339: where
340: \begin{equation}
341: E_{1,2} = \sqrt{\frac{a \pm d}{2}},\,\,a = \xi_{1}^{2} +\xi_{2}^{2} + 2
342: \Delta_{12} \Delta_{21},
343: \label{3}
344: \end{equation}
345: $$
346: d^2 = (\xi_{1}^{2} -\xi_{2}^{2}) + 4 \Delta_{12}\Delta_{21}\left[
347: (\bar \varepsilon_1 - \bar \varepsilon_2)^2 + (\Delta_{11} + \Delta_{22})^2
348: \right],
349: $$
350: $$
351: \Gamma = \left[\frac{(\bar \varepsilon_1 - \bar \varepsilon_2)^2 + (\Delta_{11}
352: + \Delta_{22})^2}{d} + 1 \right] \frac{th\frac{\beta E_1}{2}}{E_1} -
353: $$
354: \begin{equation}
355: - \left[\frac{(\bar \varepsilon_1 - \bar \varepsilon_2)^2 +
356: (\Delta_{11} + \Delta_{22})^2}{d} - 1 \right] \frac{th\frac{\beta
357: E_2}{2}}{E_2};
358: \label{4}
359: \end{equation}
360: $$
361: \xi_{n}^{2}=\xi_{n}^{2} \vec k = \bar \varepsilon_{n}^{2} +
362: \Delta_{nn}^{2},\,\,\bar \varepsilon_n =\varepsilon_n - \mu_{n},\,(n\,;
363: p\,;l\,;r = 1,\, 2) ,
364: $$
365: \begin{equation}
366: z_{(12)} = \Delta_{11}/\Delta_{22},\,\,z_{(13)} = \Delta_{11}/\Delta_{12},\,\,
367: z_{14} = \Delta_{11}/\Delta_{21}.
368: \label{5}
369: \end{equation}
370: The system (2) determines the order parameters $\Delta_{11}$ and $\Delta_{22}$
371: of an ordinary two-band superconductor (and can be simplified in this case by
372: putting $\Delta_{12} = \Delta_{21} = 0$) as well as of a superconductor with
373: low carrier density $\mu \sim T_c$. We supplement the system (2) with the
374: expression \cite{Kochorbe_1}.
375: \begin{equation}
376: N_0 = \sum_{knm,n\not=m} \left[1 - \left(\frac{\bar \varepsilon_n + \bar
377: \varepsilon_m}{2} + \frac{\bar \varepsilon_n - \bar \varepsilon_m}{2}
378: \frac{\xi_{1}^{2} - \xi_{2}^{2}}{d}\right) \frac{th\frac{\beta
379: E_n}{2}}{E_n} \right].
380: \label{6}
381: \end{equation}
382:
383: The self consistent set of Eqs. (2) and (6) determines the order parameters
384: $\Delta_{np}$ and the chemical potential $\mu$ for a specified temperature
385: $T$ and a carrier density $N_0$. A characteristic feature of the ground state
386: in a system with low carrier density is a substantial change of the position
387: of the Fermi level following the onset of the superconducting gap. The order
388: parameters $\Delta_{np}$ become of the same order as the chemical potential
389: $\mu\, (\mu \sim \Delta_{np})$. This leads to an anomalous behavior of the
390: chemical potential as a function of temperature. In particular, in the single-
391: band BCS model \cite{Van} and in the Hubbard model \cite{Robashkiewicz}
392: the chemical potential has in rarefied systems a kink at the point $T = T_c$.
393: As first noted by Van der Marel \cite{Van}, this kink is observable in
394: experiment and consequently its observation can help explain the
395: superconductivity mechanism. In Ref.\cite{Van} has demonstrated the
396: possibility of a kink at $\mu \leq 2\, me V$, which is the lower limit of
397: present-day accuracy \cite{Cardona}. Since the overlap of the energy bands
398: plays a major role in the explanation of the properties of high-temperature
399: supercondutors, it is of interest to investigate the anomalous behavior of the
400: chemical potential in a two-band model with low carrier density, and the
401: possible onset of a kink at values of $\mu$ more conducive to experimental
402: observation. We represent the order parameter $\Delta_{nm}$ near the
403: superconducting transition temperature $T \sim T_c$ in the form
404: \begin{equation}
405: \Delta_{nm} = c_{nm}(\beta - \beta_c)^{1/2} + c_{nm}^{1}
406: (\beta - \beta_c)^ {3/2} + ...\,\, .
407: \label{7}
408: \end{equation}
409: The expression for the chemical potential $\mu$ near $T_c$ is now
410: \begin{equation}
411: \mu(T) = \mu_0 (T) - R_0 \Delta_{11}^{2},
412: \label{8}
413: \end{equation}
414: were $\mu_0(T)$ is the chemical potential of the normal phase,
415: and $R_0$ is determined from the constancy of the carrier density in the
416: superconducting and normal phases. Substituting (7) in (8) we readily obtain a
417: jumplike change of $d \mu / d T$. Contributions to this change are made by all
418: the order parameters $\Delta_{nm}$, so that the results can differ from those
419: for the case of one band \cite{Van}.\\
420: \\
421: {\bf 2. 2. Self-consistent system of equations for T = Tc}
422:
423: To investigate the properties of a two-band system near the superconducting -
424: transition temperature we expand in Eqs.(2) and (6)in terms of the small
425: parameters $\Delta_{nm}\, \Delta_{mn}$ with account taken of expansions (7)
426: and (8).
427:
428: We introduce the dispersion law of the $n$th band
429: \begin{equation}
430: \varepsilon_n = - \zeta_n - \frac{k_{x}^{2} + k_{y}^{2}}{2 m_n},
431: \label{9}
432: \end{equation}
433: and change in these equations from summation over $\vec k$ to integration over
434: the energy in accordance with the dispersion law (9)$\left(\bar \varepsilon_{n
435: 0} = \varepsilon_n - \mu_{0 n}, \,\mu_{0 n} = \mu_0 - A_n \right)$:
436:
437: \begin{equation}
438: \frac{1}{V}\,\sum_{\vec k}\,\Phi (\varepsilon_n - \mu_{0 n}) = 2
439: N_n\,\int_{ - D_n}^{D_{cn}} \,dx \Phi(- x),
440: \label{10}
441: \end{equation}
442: where $N_n = m_n k_{zn}^{0} /2 (2\pi)^2$ is the density of the electron states
443: in the $n$th band, $A_n$ is a quantity that renormalizes the chemical potential
444: in the self-consistent-field approximation \cite{Kochorbe_1}. The integration
445: limits are chosen to be able to consider simultaneously two possible
446: superconductivity mechanisms: the values of $D_n = \eta_{0 n} - \zeta_n \leq
447: \omega_{D_n}$ and $D_{cn} = \omega_{D_n}$ ($\omega_{D_n}$ is the phonon cutoff
448: frequency in the $n$th band) correspond to the phonon mechanism of
449: superconductivity, and the quantities $D_n = \eta_{0n} - \zeta_n$ and $D_{cn}
450: =\zeta_{cn} - \eta_{0n}$ ($\zeta_{cn}$ is a cutoff energy of the order of that
451: of the electron;$\eta_n = - \mu_{0n}$) corresponds to the hole mechanism.
452:
453: Integrating next over the energy and equating the coefficients of equal powers
454: in the difference ($\beta - \beta_c$), we obtain for $c_{nm}$ and
455: $c_{nm}^{(1)}$ the set of equations \cite{Kochorbe_1}. In the "pseudoband"
456: representation these equations take formally the form of the four - band model
457: \cite{Moskalenko_2}, \cite{Moskalenko_9}.
458: \begin{equation}
459: c_{(n)} = - \sum_{(m) = 1}^{4}\,N_{(m)} J_m
460: U_{(n)(m)}c_{(m)},
461: \label{11}
462: \end{equation}
463: \begin{equation}
464: c_{(n)}^{1} = - \sum_{(m) = 1}^{4}\,N_{m} J_m
465: U_{(n)(m)}c_{(m)}^{(1)} - \sum_{(m) = 1}^{4}\,\,\Phi_{(m)}\,U_{(n)(m)},
466: \label{12}
467: \end{equation}
468: where
469: \begin{equation}
470: \Phi_{(m)} = \left(-\frac{\Theta_{(m)}}{\beta} + (\beta c_{(m)})^2 \bar F_{m})
471: \right) N_{(m)}.
472: \label{13}
473: \end{equation}
474: At $m = 1,2$ we have
475: $$
476: J_m = J^0 (- \beta D_{cm}) - J^0 (\beta D_m),\,J^0(x) = \int
477: \limits_{0}^{x}\,\,\frac {th\,y/2}{y}dy
478: $$
479: \begin{equation}
480: \Theta_m = th \frac{\beta\,D_{cm}}{2} - th \frac{\beta\,D_{m}}{2}\,,n = 1 - 4.
481: \label{14}
482: \end{equation}
483:
484: The functions $N_{(m)}\,\bar F_{(m)}$ at m = 1 - 4 and $N_{(m)}\,J_m,\,\,\,
485: N_{(m)}\,\Theta_m$ at $m = 3,4$ are complicated functions, contained the
486: summations on zon indexis \cite{Kochorbe_1}. We introduce quasiband indices
487: in accord with the rule
488: \begin{equation}
489: 11 \rightarrow (1),\,\,\,22 \rightarrow (2),\,\,\,12 \rightarrow (3),\,\,\,21
490: \rightarrow (4),
491: \label{15}
492: \end{equation}
493: and also the symbols:
494: $$
495: V_{nm}^{pr} \rightarrow V_{nm\,\,pr} \rightarrow U_{(n')(m')};\,\, n^{/},
496: m^{/} = 1 - 4.
497: $$
498: We omit hereafter the parentheses of the pseudoband subscripts. It is
499: convenient to rewrite Eqs.(11) and (12) in the matrix form
500: \begin{equation}
501: D_0 c = 0
502: \label{16}
503: \end{equation}
504: \begin{equation}
505: D_0 c^{(1)} = - \Phi,
506: \label{17}
507: \end{equation}
508: were $c,\,c^{(1)}$, and $\Phi$ are single - columm matrices in the indices 1
509: to 4, and
510: %\begin{displaymath}
511: \begin{equation}
512: D_0 =
513: \left( \begin{array}{cccc}
514: 1 + N_1J_1U_{11} & N_2J_2U_{12} & N_3J_3U_{13} & N_4J_4U_{14}\\
515: N_1J_1U_{12} & 1 + N_2J_2U_{22} & N_3J_3U_{23} & N_4J_4U_{24}\\
516: N_1J_1U_{31} & N_2J_2U_{32} & 1 + N_3J_3U_{33} & N_4J_4U_{34}\\
517: N_1J_1U_{41} & N_2J_2U_{42} & N_3J_3U_{43} & 1 + N_4J_4U_{44}
518: \end{array} \right).
519: \label{18}
520: \end{equation}
521: %\end{displaymath}
522: From the condition that the system (16) have a solution, we obtain an equation
523: for the critical temperature $T_c$:
524: \begin{equation}
525: ||D_0|| = 0
526: \label{19}
527: \end{equation}
528: where $||...||$ designates the determinant of a matrix. It follows from the
529: system (17) that
530: \begin{equation}
531: c_{1}^{(1)} = ||D_1||/||D_0||.
532: \label{20}
533: \end{equation}
534: Since $||D_0|| = 0$, we get
535: \begin{equation}
536: ||D_1|| = 0,
537: \label{21}
538: \end{equation}
539: where $D_1$ is a $4 \times 4$ matrix that differs from $D_0$ in that the first
540: column is replaced by the elements of the matrix $\Phi$.
541: We have on the basis of (20) the expression for $c_{1}^{2}$
542: \begin{equation}
543: c_{1}^{2} = \frac{1}{\beta_{c}^{3}}\,\frac{\sum_{n = 1} N_n \theta_n f_n
544: / z_{1n}^{2}}{\sum_{n = 1}^{4} N_n \overline{F}_n f_n / z_{1n}^{4}} = \frac
545: {\bar c_{1}^{2}}{\beta_{c}^{3}}.
546: \label{22}
547: \end{equation}
548:
549: Assuming the particle number $N_0$ to be fixed and using the expansion (7) for
550: the chemical potential near $T_c$ we obtain
551: \begin{equation}
552: \mu (T) = \mu_0 (T) - \frac{R_0}{\beta}\,{\bar c_{1}^{2}}(T_c - T),
553: \label{23}
554: \end{equation}
555: and $\mu_0$ is determined from the equation
556: \begin{equation}
557: N_0 = 2 \sum_{l} \left[D_{cl} + D_l - \frac{2}{\beta}ln \frac{ch\frac{\beta
558: D_{cl}}{2}}{ch\frac{\beta D_l}{2}}\right].
559: \label{24}
560: \end{equation}
561: The determinations of the functions $f_n / z_{1n}$ and $R_0$ see in
562: \cite{Kochorbe_1}.
563:
564: It is possible to calculate the dependence of the superconducting temperature
565: $T_c$ on the carrier density $N_0$ for all values of the interaction constants
566: $\lambda_{nm} = N_m U_{nm}\, (n,m = 1 - 4)$, and also the temperature dependence
567: of the chemical potential $\mu(T)$ on basis of (18) - (24).
568:
569: Figure 1 shows the dependence of the superconducting temperature $T_c$ on the
570: carrier density $N_0 / 2 N_1$ for different degrees of hybridization in
571: the casse of electron mechanizm of superconductivity: a) weak: $\lambda_{11} =
572: \lambda_{22} = 0.2,\,\lambda_{12} = \lambda_{21} = \lambda_{33} = \lambda_{44}
573: = 0.01,\,\lambda_{34} = \lambda_{43} = 0.105$, the remaining ones:$\lambda_{nm}
574: = 0.001\,(n, m = 1 - 4)$; b) strong $\lambda_{11} = \lambda_{22} = 0.2,
575: \lambda_{12} = \lambda_{21} = \lambda_{33}= \lambda_{44} = 0.1,\, \lambda_{34}
576: = \lambda_{43} = 0,15 ,$ the remaining ones $\lambda_{nm} = 0.01\,(n, m = 1 -
577: 4)$.
578:
579: Calculations are perfomed for the following values of the parameter
580: $$
581: \zeta_1 = 0\,eV,\,\,\zeta_2 = 0.03\,eV,\,\,\zeta_{c1} = 0.05\,eV,\,\,
582: \zeta_{c_2} = 0.08\,eV.
583: $$
584:
585: It follows from Fig.1 that one can obtaina bell-shaped dependence of $T_c$ on
586: $N_0$ (curves 1-3) as well as the weak dependence $T_c (N_0)$ curves(1 - 3).
587: Inclusion of strong hybridization raise $T_c$. The character of the
588: dependence of the transition temperature $T_c$ on the carrier density $N_0$ is
589: strongly influenced by the relation between the electronic - state densities of
590: the different bands. Lowering the ratio $N_2 / N_1$ slows down the growth of
591: $T_c$ with increase of $N_0$ (for case b) and accelerates the decrease for both
592: cases $(a,b)$
593:
594: In the case of weak hybridization (curves $1 - 3$) the $T_c\,(N_0)$ plot
595: acquires two maxima. The degree to which then become pronounced, given the
596: parameters $\lambda_{mn}$, is determined by the electron state-density ratio
597: $N_2 / N_1$. The presence of weakly pronounced maxima (curve 1) corresponds to
598: the case $N_1 = N_2$, and more strongly pronounced maxima (curves 2 and 3)
599: appear at $N_1 \not = N_2$ and are determined by the anisotropy of the system
600: (by the difference between the bands). Each of these maxima is connected with
601: the occupation of the corresponding band. In the absence of interband
602: interaction (energy - band overlap) the plot would consist of two nonoverlapping
603: curves. The onset of interband interactions produces simultaneos
604: superconductivity in both bands, with a single superconducting temperature
605: determined by all the interaction constants $\lambda_{nm}$. With increase of
606: the interband-interaction constants $(\lambda_{34},\lambda_{43} \sim
607: \lambda_{11},\lambda_{22})$ the contribution due to overlap of the two bands
608: begins to predominate over the individual contribution of each band, so that
609: the $T_c (N_0)$ plot is a single bell - shaped curve (4,5).
610:
611: Figure 2 shows the temperature dependence of $\eta = - \mu$ for a non - phonon
612: superconductivity mechanism at various carrier densities $N_0$ and with weak
613: hybridization. We see that this plot has at $T = T_c$ (curves 1 - 3) a kink
614: that becomes less peaked with increase of the carrier density and vanishes at
615: $\eta \sim 8$ me V (curve 4). The behavior of $\eta (T)$ under strong
616: hybridization is similar. The anomaly of the temperature dependence of the
617: chemical potential $\eta = - \mu$ at the point $T = T_c$ is due to the
618: appearance, on the Fermi surface, of a superconducting gap that does not differ
619: excessively from the chemical potential. This gap influences substantially the
620: chemical potential $\mu$ at $T < T_c$, since the values of $\eta$ and
621: $\Delta_{n m}$ are self-consistently determined from the set of equations (2)
622: and (6).
623:
624: Allowance for the energy-band overlap leads to a kink on the temperature
625: dependence of the chemical potential at $T = T_c$ for the sufficiently high
626: values $\eta \leq 8 meV$ me ($\eta \leq 2 meV$ in the single band case
627: \cite{Van}, which undoubtedly facilitates experimental verification of this
628: effect).
629: \\\\
630: {\bf 2. 3. Critical temperatur $T_c$ and ratio $2\Delta_n / T_c$ \\
631: \,\,\,\,\,\,in the limiting case $\Delta_{12} = \Delta_{21} = 0$.}
632:
633: In the limiting case $\Delta_{12} = \Delta_{21}\,\,\,(N_3 = N_4 = 0)$ the
634: problem is simplify essentialy . Such model corresponds to \cite{Moskalenko_2},
635: \cite{Moskalenko_9} the equation for $T_c$ has the following form
636: \cite{Palistrant_2}:
637:
638: \begin{equation}
639: 1 + \lambda_{11} J_1 + \lambda_{22}J_2 + a J_1\,J_2 = 0,
640: \label{25}
641: \end{equation}
642: \begin{equation}
643: \lambda_{nm} = V_{nm} N_m,\,\,a = \lambda_{11} \lambda_{22} - \lambda_{12}
644: \lambda_{21}.
645: \label{26}
646: \end{equation}
647: Besides, the $\eta$ in the case of nonfonon mechanism of superconductivity is
648: defined from
649: \begin{equation}
650: N_0 = 2\,\sum_n N_n \left[\zeta_{cn} -
651: \zeta_n - |\zeta_{cn} - \eta| + |\zeta_{n} - \eta| - 2 T_c ln \frac{1 +
652: exp(-\beta_c |\zeta_{cn} - \eta|)}{1 + exp (- \beta_c|\zeta_n - \eta|)}\right].
653: \label{27}
654: \end{equation}
655:
656: From the relation (22) we have
657: \begin{equation}
658: c_{1}^{2} = \frac{1}{\beta_{c}^{3}}\,\frac{N_1\,\theta_1 \,+
659: N_2\,\theta_2 / z^2}{N_1 F_1 + N_2 F_2 / z^4} ,
660: \label{28}
661: \end{equation}
662: where you $J_m$ and $\Theta_m$ are determined by (14), and
663: \begin{equation}
664: F_m = -\frac{1}{4} \int \limits_{-\beta_c D_m}^{-
665: \beta_c D_{cm}}\, dx \, \frac{shx - x}{x^3 ch^2 x/2}\,\,,z = \Delta_1/\Delta_2.
666: \label{29}
667: \end{equation}
668:
669: For the sake of definiteness we choose $\zeta_1 = 0 < \zeta_2 <
670: \zeta_{c1} < \zeta_{c2}$ and consider the equation for the
671: superconducting transition temperature (25). On the basis of this equation one
672: can obtain the analytic expressions for the transition temperature $T_c$ if
673: the following conditions are satisfied: $|\eta - \zeta_n|/ T_c \gg 1$ and
674: $|\zeta_{cn} - \eta|/ T_c \gg 1$
675: So, for definite values of $\eta$ we obtain the following expressions:\\
676: (1) For $0 < \eta < \zeta_2$ :
677: \begin{equation}
678: T_c = \frac{2 \gamma}{\pi}\,\sqrt{\eta(\zeta_{c1} - \eta)}
679: exp\,\left[- \frac{1}{2} \frac{1 + \lambda_{22} ln ((\zeta_2 -
680: \eta)/(\zeta_{c2} - \eta))}{\lambda_{11} + a\,ln((\zeta_2 -
681: \eta)/(\zeta_{c2} - \eta))}\right] ,
682: \label{30}
683: \end{equation}
684:
685: $N_0 = 4 N_1 \eta$.
686:
687: (2) For $\eta = \zeta_2$:
688: $$
689: T_c = \frac{2 \gamma}{\pi}\,\eta^{1/4}(\zeta_{c_1} -
690: \eta)^{1/4}\sqrt{\zeta_{c_2} - \eta}\,
691: $$
692: \begin{equation}
693: exp\,\left[- \frac{2 \lambda_{11} + \lambda_{22}}{4a} \pm \frac{1}{4a}
694: \sqrt{\left(a\,ln \frac{\eta (\zeta_{c_1} - \eta)}{(\zeta_{c_2} -
695: \eta)^2} + 2 \lambda_{11} - \lambda_{22}\right)^2 + 8\lambda_{12}
696: \lambda_{21}}\,\,\right],
697: \label{31}
698: \end{equation}
699:
700: $N_0 = 4N_1 \eta$.
701:
702: (3) For $\zeta_{c2} < \eta < \zeta_{c1}$:
703:
704: $$
705: T_c = \frac{2 \gamma}{\pi}\left[\eta (\zeta_{c_1} - \eta)(\zeta_{c_2}
706: - \eta)(\eta - \zeta_2)\right]^{1/4}
707: $$
708: \begin{equation}
709: exp\,\biggl\{- \frac{\lambda_{11} + \lambda_{22}}{4 a} \pm \frac{1}{4 a}
710: \sqrt{ a\,ln\, \frac{\eta (\zeta{c_1} - \eta)}{(\zeta_{c_2} -
711: \eta)(\eta - \zeta_2)} + (\lambda_{11} -\lambda_{22})^2 + 4 \lambda_{12}
712: \,\lambda_{21}} \biggr\},
713: \label{32}
714: \end{equation}
715:
716: $N_0 = 4 N_1\,\eta_1 + 4 N_2 (\eta - \zeta_2)$.
717:
718: (4) For $\eta = \zeta_{c1}$
719:
720: $$
721: T_c = \frac{2\gamma}{\pi} \sqrt{\eta}\,(\zeta_{c2} - \eta)^{1/4} (\eta -
722: \zeta_2)^{1/4}
723: $$
724: \begin{equation}
725: exp\,\left[ -\frac{\lambda_{11} + 2 \lambda_{22}}{4a} \pm \frac{1}{4 a}
726: \sqrt{\left(a\,ln\,\frac{\eta^2}{(\zeta_{c_2} - \eta) (\eta -
727: \zeta_2)} + \lambda_{11} - 2 \lambda_{22} \right) +
728: 8\,\lambda_{12} \lambda_{21}}\right],
729: \label{33}
730: \end{equation}
731:
732: $N_0 = 2 N_1 (\eta + \zeta_{c_1}) + 4N_2(\eta - \zeta_2)$.
733:
734: (5) For $\zeta_{c_1} < \eta < \zeta_{c_2}$
735:
736: \begin{equation}
737: T_c = \frac{2 \gamma}{\pi} \sqrt{(\zeta_{c_2} - \eta)(\eta -
738: \zeta_2)} \,exp \left[ - \frac{1}{2}\,\frac{1 + \lambda_{11}\, ln ((\eta
739: - \zeta_{c1}) / \eta)}{\lambda_{22}+ a\,ln ((\eta - \zeta_{c1})
740: /\eta)}\right],
741: \label{34}
742: \end{equation}
743:
744: $N_0 = 4 N_1\,\zeta_{c_1} + 4 N_2 (\eta - \zeta_2)$.
745:
746: The expressions $(31) \sim (33)$ contain two solutions wich correspond to the
747: sign $"\pm"$ in the brackets. A negative value of the quantity in the exponent
748: is a necessary condution for the solution's selection. In the case where both
749: solutions satisfy the above necessary condition, the solution giving the
750: greater $T_c$ must be selected. Note that we will obtain the results of Ref.
751: \cite{Moskalenko_2} provided $V_{11} = V_{22} = 0$ and $\zeta_{c_1} =
752: \zeta_{c_2}$.
753:
754: The analysis of the above $T_c$ expressions permits us to make the conclusion
755: that high $T_c$ values can be achived both for $\lambda_{nm} > 0$ (carrier
756: attraction) and $\lambda_{nm} < 0$ (repulsion). In the latter case the
757: condition $ a = \lambda_{11} \lambda_{22} - \lambda_{12} \lambda_{21}< 0$,
758: imposing restrictions on the constants $\lambda_{nm}$, must be satisfied.
759: The dependence of $T_c$ on $\eta$ is demonstrated in fig. 3 at definite values
760: of the parameters $\lambda_{nm}$. Here and hereafter in computations we choose
761: the following values of the parameters: $\zeta_1 = 0 \,e V,\,
762: \zeta_2 = 0.11 \,e V,\,\zeta_{c_1} = 0.2 \,e V$\,and
763: $\zeta_{c_2} = 0.3 \,e V\, $. Curves 1 and 2 correspond to the case
764: $\lambda_{nm} > 0$ and to the case $\lambda_{nm} < 0$, respectively. As follows
765: from this figure, high values of $T_c$ can be achieved as the chemical
766: potential changes. The $T_c$ dependence on $\bar N_0 = N_0 / 4 N_1$ can be
767: presented easily. In this case the ratio $N_2 / N_1$ must be given. Then the
768: $T_c$ dependence will be defined by this ratio. The $\eta$ dependence on $\bar
769: N_0$ at different values of $N_2 / N_1$ is shown in fig.4. As follows from this
770: figure, the rate of growth of the quantity $\eta$ is defined by the ratio
771: $N_2 / N_1$ as $N_0$ increases.
772: The critical temperature $T_c$ as well as the order paramiters $\Delta_m$ as
773: function of $\eta$ in this limit case is studied detaly in
774: \cite{Palistrant_2}.
775:
776: This investigations permit us to define the ratios $2 \Delta_1 / T_c$ and
777: $2 \Delta_2 / T_c$ as function of $\eta$ (see fig.5). The curves 1 and 2
778: correspond to the ratios $2 | \Delta_1 (0) | / T_c$ and
779: $2 | \Delta_2 (0) | / T_c$ respectively for $\lambda_{nm} > 0$, and the curves $1^{/},2^{/}$
780: correspond to the ones for $\lambda_{nm} < 0$. As is seen from
781: this figure the behavior of the quantities $2 | \Delta_1 (0) | / T_c$ and
782: $2 | \Delta_2 (0) | / T_c$ as functions of $\eta$ depends essentially on the
783: values of the parameters $\lambda_{nm}$.When $\lambda_{nm} > 0 $ we have a
784: step decrease of $2 |\Delta_1(0)|/T_c$ and a smooth growth of $2 |\Delta_2
785: (0)| / T_c$ as $\eta$ increases. When $\lambda_{nm} < 0 $ the ratio $2 |
786: \Delta_1 (0) | / T_c$ essentially increases and can achieve values near 7.5
787: and the ratio $2 | \Delta_2 (0) | / T_c$ slowly decreases, achieving the value
788: 3.5 as $\eta$ increases. So the values of $2 | \Delta_1 (0) | / T_c$ and $2 |
789: \Delta_2 (0) | / T_c$ depend on $\eta$ and can be essentially different from
790: the BCS theory ones where $2 | \Delta (0) | / T_c$ = 3.5 and from the two -
791: band theory with phonon superconductivity mechanism \cite{Moskalenko_2},
792: \cite{Moskalenko_9}, where these ratios are independent of the chemical
793: potential (carrier concentration). Note that for $T_c$, as for $| \Delta_n
794: (0)|$, the ratio $2 | \Delta_n (0)| / T_c$ can be shown in dependence on the
795: carrier concentration, by using relations between $\eta$ and $\bar N_0$ (see
796: fig.4).
797: \\\\
798: {\bf 2.4. Heat - capacity jump at the point T = Tc.}
799:
800: In the section 2.2 we changed over to the pseudo-band representation, which
801: allowed us to write down Eqs. (11) and (12), as it were, for a four-band
802: model.
803:
804: It can also be shown that the difference between the free energies in the
805: superconducting and normal phases in the pseudoband representation generalizes
806: the corresponding expression of the Moskalenko two-band
807: model \cite{Moskalenko_2}, \cite{Moskalenko_9}. We obtain
808:
809: \begin{equation}
810: \frac{\Psi_S - \Psi_N}{V} = \sum_{nmp}\,\int_{0}^{\Delta_p}\,\Delta_n \Delta_m
811: \frac{(\delta U^{-1} )_{nm}}{\delta \Delta_p} \delta \Delta_p,
812: \label{35}
813: \end{equation}
814:
815: where $n, m$ and $p$ are the pseudoband nambers $(n, m, p = 1 - 4)$, and $U^{-
816: 1}$ is the inverse of the interaction matrix $U$ in (11). We expand, in the
817: pseudoband formalism, the set of Eqs. (2) for the order parameters $\Delta_n$
818: in powers of the small quantity $(\beta \Delta_n)^2$ in the vicinity of the
819: critical temperature $T_c$:
820: \begin{equation}
821: \Delta_m = - \sum_n N_n\,U_{nm} (J_n + (\beta \Delta_n)^2 \overline F_n + ...),
822: \label{36}
823: \end{equation}
824: where $J_n$ and $\overline F_n$ are defined in [49]. Using the calculation method of
825: Refs.\cite{Moskalenko_9} and \cite{Palistrant_2} for Eqs.(35) and (36),
826: we obtain for the heat - capacity jump at the point $T = T_c$
827: $$
828: \frac{C_S - C_N}{V} = - T \frac{\partial^2}{\partial\, T^2} \frac{\Psi_S -
829: \Psi_N}{V} = \beta_{c}^{5}\sum_{n = 1}^4 N_n\,\overline F_n\,c_{n}^{4} =
830: $$
831: \begin{equation}
832: = T_c \left[\frac {\Sigma_{n = 1}^{4}\,N_n \theta_n f_n / z_{1n}^{2}}{
833: \Sigma_{n = 1}^{4}\,N_n \overline F_n f_n / z_{1n}^{4}}\right]^2\,\sum_{n = 1}^
834: {4}\,\,\frac{N_n\,\overline F_n}{z_{1n}^{4}}.
835: \label{37}
836: \end{equation}
837:
838: The equation for the electronic heat capacity in the normal phase is
839: \cite{Palistrant_2}
840: \begin{equation}
841: C_N = 4 T\,\sum_{n = 1}^{2}\,N_n \phi_n (\eta),
842: \label{38}
843: \end{equation}
844: where
845: \begin{equation}
846: \phi_n (\eta) = \int_{- \beta (\eta - \zeta_{cn})}^{\infty}\,\,\frac{x^2 d x}
847: {(1 + e^x)(1 + e^{- x})}.
848: \label{39}
849: \end{equation}
850: In accordance with Eqs. (37) and (38) we have for the relative heat - capacity
851: jump
852: \begin{equation}
853: \frac{C_S - C_N}{C_N} = \left[\frac {\Sigma_{n = 1}^{4}\,N_{n} \theta_
854: {n} f_{n} / z_{1n}^{2}}{ \Sigma_{n = 1}^{4}\,N_{n} \overline F_{n}
855: f_{n}/ z_{1n}^{4}}\right]^2\, \frac {\Sigma_{n =
856: 1}^{4}\,N_{n} \overline F_n / z_{1n}^{4}}{4 \Sigma_{n = 1}^{2} N_n \,\phi_n
857: (\eta)}.
858: \label{40}
859: \end{equation}
860: Figures 6 shows the dependences of the absolute and relative jump of the
861: electron heat capacity at the point $T = T_c$ on the carrier density $N_0$,
862: obtained by numerical methods using the equations given above. The numbers of
863: the curves in these figures are the same as in Fig.1.
864:
865: As seen from Fig. 6a, the dependence of $(C_S - C_N)_{T = T_c}$ on $N_0$ has a
866: maximum. At the same time, this dependence does not duplicate the behavior
867: of $T_c (N_0)$. This circumstance indicates that a substantial contribution
868: to the dependence of $(C_S - C_N)_{T = T_c}$ on $N_0$ is made not only by
869: $T_c$ but also by the complicated function in the right-hand side of (37).
870: Analysis of the curves of Fig.6 shows that the charater of the plot of $(C_S -
871: C_N)_{T = T_c}$ versus $N_0$ is determined by the type of hybridization (strong
872: or weak) and by the ratio $N_2 / N_1$ of the electron - state densities. It is
873: possible for $T_c$ and $(C_S - C_N)_{T = T_c}$ to have maxima at one and the
874: same value of $N_0$ (curves 4 and 5). This situation is observed in
875: experiment, for example in $La_{2 - x}Sr_x Cu_2 O_4 $ \cite{Hongshum}.
876:
877: The possibility of obtaining small $(C_S - C_N) / C_N < 1.43$, as well as large
878: $(C_S - C_N) / C_N > 1.43$ values of the relative jump of the electron
879: specific heat is demonstrated by Fig. 6b. This picture is observed in high -
880: temperature ceramics \cite{Junod} -\cite{Akis_2}. The complicated dependence
881: of the relative electron-heat-capacity jump, shown in Fig. 6a, is determined by
882: the competition between the behavior of the diference $C_S - C_N$ shown in
883: Fig.6a as a function of $N_0$ and the quantity $C_N$ which increases as $N_0$
884: increases.
885:
886: We have considired in the present study quasi-two-dimensional systems with a
887: simple dispersion law (9). This approach is dictated, in particular, by the
888: lower dimensionality of a number of high-temperature ceramics. Since, however,
889: the electron-state densities $N_n (\varepsilon) , n = 1, 2$ have no
890: singularities for the dispersion law (9), we obtain the very same equations
891: also for a three-dimensional system. Only the values of $N_n$ will differ. Just
892: as in the case of single-band superconductors, \cite{Van} in our case
893: the ratio $C_S-C_N > 1.43$ is goverened to a considerable degree by the small
894: $C_N$ in the considered range of $N_0$ compared with the case of ordinary
895: metals (or by the faster increase of $C_S-C_n$ with increase of $N_0$ compared
896: with the increase of $C_S-C_N$ with increase of $N_0$ compared with the
897: increase of $C_N$).
898:
899: In the case $N_{3} = N_{4} = 0$ (this limit corresponds to
900: ujualy two-band model \cite{Moskalenko_2}), we presents the specific heat jump
901: in the form Ref. (\cite{Palistrant_2}):
902: \begin{equation}
903: \frac{C_S - C_N}{V} = T_c
904: \frac{\left[ N_1\theta_1 (\eta) + N_2 \theta_2 (\eta) / z^2 \right]^2}{N_1
905: F_1(\eta) + N_2 F_2 (\eta) / z^4},
906: \label{41}
907: \end{equation}
908: \begin{equation}
909: \frac{C_S - C_N}{C_N} = T_c \frac{\left[ N_1\theta_1 (\eta) + N_2 \theta_2
910: (\eta) / z^2 \right]^2}{4 \left[N_1 F_1(\eta) + N_2 F_2 (\eta) / z^4\right]
911: \left[N_1 \varphi_1 (\eta) + N_2 \varphi_2 (\eta)\right]}.
912: \label{42}
913: \end{equation}
914: The presence of the functions $\theta_n (\eta)$ and $F_n (\eta)$ distinguishes
915: this expression from the two-band theory with phonon superconductivity. In the
916: above approximation $\eta >> T_c$, values of these functions at the different
917: values of $\eta$ are given in table 1 Ref.\cite{Palistrant_2}. So the specific
918: heat jump depends on $\eta$ (on $N_0$ through the $T_c$ dependence on this
919: quantity and by the suplementary function in the right hand part of eq. (41).
920: By using table 1 \cite{Palistrant_2} and the above values of $T_c$ the
921: specific heat jump dependence $(C_S - C_N))$ on $N_0$ can be easily presented
922: on the basis of eq. (41). This dependence is shown in fig.7 at different values
923: of the ratio $N_2 / N_1$. The curves with non-primed numbers correspond to
924: $\lambda_{nm} > 0$ and the ones with primed nambers correspond to $\lambda_{nm}
925: < 0$. We obtain the result that the quantity $(C_S - C_N) / V N_1$ increas as
926: the carrier concentration increases. The maximum of this quantity is achieved
927: at the values of $N_0$ wich corespond to the superconducting transition
928: temperature. Then this quantity decreases as $N_0$ increases for all values of
929: $N_0 / N_1$.It is interesting to note also that the same dependence of the
930: absolute specific heat jump at the point $T = T_c$ on $S_r$ concentration has
931: been observed in $La_{2-x} S_{rx} Cu_2 O_4$ \cite{Hongshum}. By using table 1
932: of Ref. \cite{Palistrant_2} we present expressions for the relative jump
933: of electron spesific heat in some points.\\
934: 1) For $0 < \eta < \zeta_2$:
935: \begin{equation}
936: \frac{C_S - C_N}{C_N} = \frac{12}{7 \zeta(3)} = 1,43.
937: \label{43}
938: \end{equation}
939: 2) For $\eta = \zeta_2$:
940: \begin{equation}
941: \frac{C_S - C_N}{C_N} = \frac{12}{7 \zeta(3)} \frac{(N_1 + N_2 / z^2)^2}
942: {(N_1 + N_2 / z^4)(N_1 + N_2 / 2)}.
943: \label{44}
944: \end{equation}
945: 3)For $\zeta_2 < \eta < \zeta_{c1}$:
946: \begin{equation}
947: \frac{C_S - C_N}{C_N} = \frac{12}{7 \zeta(3)} \frac{(N_1 + N_2 / z^2)^2}
948: {(N_1 + N_2 / z^4)(N_1 + N_2)}.
949: \label{45}
950: \end{equation}
951: 4)For $\eta = \zeta_{c1}$:
952: \begin{equation}
953: \frac{C_S - C_N}{C_N} = \frac{12}{7 \zeta(3)} \frac{(N_1 + N_2 / z^2)^2}
954: {(N_1/2 + N_2 / z^4)(N_1/2 + N_2)}.
955: \label{46}
956: \end{equation}
957: 5)For $\zeta_{c1} < \eta < \zeta_{c2}$:
958: \begin{equation}
959: \frac{C_S - C_N}{C_N} = \frac{12}{7 \zeta(3)}.
960: \label{47}
961: \end{equation}
962:
963: The formulae (43)-(46) implicity demonstrate the consequence of inclusion and
964: turning-off of the overlapping bands as the parameter $\eta$ increases. The
965: expression (45) corresponds to the existence of both bands and to the maximum
966: value of the superconducting transition temperature.This expression coinsides
967: in form with the one obtained in the model with phonon superconductivity
968: \cite{Moskalenko_2},\cite{Moskalenko_9}.
969:
970: By using the above formulae the relative jump of the electron specific heat one
971: has $(C_S - C_N) / C_N < 1.43$ at any values of the parameter $\eta$(or $N_0$)
972: and the former can achieve sufficiently small values. Note that the small
973: values of the quantity $(C_S - C_N)/C_N = 0.43 - 1.14$ have been considered in
974: the thallium ceramics \cite{Junod}.
975:
976: The specific heat jump dependence on $N_0$ corresponding to $\lambda_{n m} > 0$
977: and to $\lambda_{n m} < 0$ is presented in fig. 8(a) and (b),respectively. In
978: both cases the ratio $(C_S - C_N)/ C_N$ is equal to 1.43 in the range where
979: only one energy band exists $(\eta < \zeta_2$ and $\eta > \zeta_{c1})$. This
980: value corresponds to the BCS theory. This ratio decreases after inclusion of
981: both bands.In the case $\lambda_{nm} > 0$ we obtain a smooth dependence with a
982: slightly spread minimum. Nevertheless, in the case $\lambda_{nm} < 0$ we
983: observe a quick decrease of this ratio as $N_0$ increases followed by a slower
984: rate of growth in the range with overlapping energy bands.
985:
986: The mentioned above formulas for the jump of heat capacity (42),
987: and (43)-(47) as well, correspond to the isotropic two-band superconductor
988: at $\Delta_{12}=\Delta_{21}=0$. In particular, for $\Theta_n=F_n=1$ formula
989: (42) gives the case of electon-phonon mechanism of superconductivity
990: \cite{Moskalenko_2}, \cite{Moskalenko_9} in two-band isotropic
991: system. In case of anisotropic electron-phonon interaction the result is
992: presented in Ref. \cite{Mishonov_1}-\cite{Mishonov_3} and have succeeded to
993: describe the behavior of the heat capacity at $T=T_c$ in $MgB_2$ compound.
994:
995: Note that our investigations were made in the meanfield approximation, and it
996: is in this approximation that the proposed superconductivity theory, with two
997: overlapping energy bands, describes the properties of the system for an
998: arbitrary ratio of $T_c$ to $N_0$. The mean-field approximation
999: itself, however, may turn out to be insufficient when very low carrier densities
1000: are considered. It becomes necessary here to take into account the
1001: fluctuations of the order parameters near the superconducting transition
1002: temperature.Our numerical calations were made mainly for values $T_c
1003: /\varepsilon_F \sim 10^{- 1} - 10^{-2}$(cf.the data of Figs. 7 and 9), and in
1004: this respect we can be assumed to have a physical picture that is qualitatively
1005: close to the real one. To be sure, we are still faced here with the question
1006: of the nature of the superconductivity, namely, will it be based on the Cooper-
1007: pair production mechanism or will it be determined by the Bose condensate.Good
1008: results in the ground state are obtained in the BCS theory for $\Delta \ll
1009: |\mu|$, but a condensate of noninteractiong bosons is produced in the opposite
1010: limit.Analysis of the intermediate region at finite temperature is still an
1011: unsolved problem, and our results are an interpolation of the BCS mechanism to
1012: this intermediate region.
1013: \\\\
1014: {\bf 2.5 Electron-phonon mechanism of superconductivity}
1015:
1016: In the precious sections we has derived the equations and the analytical
1017: expressions in particular cases for the quantities $T_{C}$ , $\mu \left(
1018: T\right) $, and for the absolute and relative jump of heat capacity at the
1019: point $T=T_{C}$. The equations are valid both for electron and phonon
1020: mechanism of superconducting pairing. Formally speaking, this feature
1021: appears in a way of cut-off while having performed the integration over
1022: energy in the main equations (see section 2.2). The graph plots of the
1023: presented here dependences of various quantities (fig. 1-9) correspond the
1024: non-phonon superconductivity.
1025:
1026: We present also the results for the electron-phonon superconductivity,
1027: having considered as earlier the weak and strong hybridization, and chosen
1028: the same values of parameters $\lambda _{nm}$ as in section 2.2.
1029:
1030: We choose also the following values .$\varsigma _{1}=0$ eV, $\varsigma
1031: _{2}=0.01$ eV, $\varsigma _{c1}=0.02$ eV, $\varsigma _{c2}=0.03$ eV. The
1032: results of numerical calculations are presented on the fig.1 - 10. The shape
1033: of dependences of the quantity $T_{C}$ on $N_{0}$ presented on the fig.1 and
1034: 10 is easily observed not to be the same. This difference appears clearer
1035: especially in the case of weak hybridization. In the case of strong
1036: hybridization there is a bell-shape dependence of the quantity $T_{C}$ on
1037: the density of charge carriers.
1038:
1039: The kink of chemical potential $\mu \left( T\right) $ at the point $T=T_{C}$
1040: (fig. 11) is well distinguished and disappears at the values $\mu \sim 6$
1041: meV what is as in the case of electron pairing mechanism (fig. 2) achievable
1042: to the experimental observation.
1043:
1044: The dependence of the jump of electron heat capacity $\left(
1045: C_{S}-C_{N}\right) |_{T=T_{C}}$ on the density of charge carriers differs
1046: distinctly at the weak hybridization as well (compare fig.6a and 12). In the
1047: case of phonon pairing mechanism of superconductivity and weak hybridization
1048: the maximum of the quantity $\left( C_{S}-C_{N}\right) /\,VN_{1}$ is shifted
1049: towards low densities of charge carriers in comparison with the non-phonon
1050: pairing mechanism. The curves (fig.6a and 12) that correspond to the strong
1051: hybridization differ as well. The dependence of the relative jump of
1052: electron heat capacity on the density of charge carriers in both cases of
1053: hybridization has complex shape ( see fig.6b and 13) and depends essentially
1054: upon theory parameters. Regarding the value of $N_{0}$ the quantity $\left(
1055: C_{S}-C_{N}\right) /\,C_{N}$ at $T=T_{C}$\ can be as greater as lower than
1056: 1.43 (the value that corresponds to the isotropy BCS model with phonon
1057: pairing mechanism of superconductivity). Detailed analysis of these results
1058: and their comparison with experimental data might shed light upon the
1059: mechanism of superconductivity in two-band system.
1060:
1061: As was noticed above the discovery of high-$T_{C}$ superconductivity in
1062: magnesium boride $MgB_{2}$ $T_{C}\sim 39K$ \ has lead to rapid development
1063: in experimental and theoretical studies. Confirmation of two-band nature of
1064: this system (presence of two energy gaps) \cite{Ponomarev} and electron-phonon
1065: mediated pairing mechanism of superconductivity \cite{Bud'ko} - \cite{Choi_3} is
1066: very important discovery. The singularities of two-band system in temperature
1067: dependence of heat capacity, penetration depth of magnetic field, upper
1068: critical magnetic field and other physical quantities are observed. The
1069: background of all these studies is the model of Moscalenco that was
1070: suggested by him in 1959 \cite{Moskalenko_2}. On the basis of this model
1071: Moscalenco and his colleagues have made a great number of studies long before
1072: the discovery the compound $ MgB_{2}$. Many modern authors to our pity
1073: represent their own results as originally new ones not making any references on
1074: the pioneer works of Moscalenco and his colleagues (see \cite{Palistrant_1},
1075: \cite{Moskalenko_8}).
1076:
1077: In order to describe the physical properties of $MgB_{2}$ it is necessary to
1078: determine the parameters of theory on the basis of present experimental
1079: data, to take into account anisotropy of matrix element of electron-phonon
1080: interaction \cite{Mishonov_1} - \cite{Mishonov_3}, important role of Van
1081: Hove singularities \cite{Loram}, it is apt to take into consideration strong
1082: electron-phonon interaction and other singularities as well as multi-band
1083: nature of the system.
1084:
1085: Because parameters of theory in the two-band model presented in this article
1086: are unknown, our studies performed here permit us to make qualitative
1087: conclusions about superconducting properties of $MgB_{2}$.
1088:
1089: 1. The quantity $T_{C}$ as a function of the density of charge carriers has
1090: bell-shape dependence. At small hybridization there is wide field of
1091: densities of charge carriers $N_{0}$ at which $T_{C}$ depends slightly upon $
1092: N_{0}$. At great hybridization the maximum in the dependence of $T_{C}$ on $
1093: N_{0}$ is notably distinguished. The height and position of this maximum is
1094: determined by relationship $N_{1}/N_{2}$ as well as by values of coupling
1095: constants $\lambda _{nm}$.
1096:
1097: 2. The chemical potential at the point $T=T_{C}$ has kink at low densities
1098: of charge carriers. This kink disappears with increasing $N_{0}$. The
1099: overlapping of energy bands favors the experimental confirmation of this
1100: kink.
1101:
1102: 3. The positions of maximums of $T_{C}$ and difference $\left(
1103: C_{S}-C_{N}\right)$ at $T=T_{C}$ as a functions of density of charge carriers
1104: do not coincide.
1105:
1106: 4. The relative jump of electron heat capacity $\left( C_{S}-C_{N}\right)
1107: /C_{N}$ at the point $T=T_{C}$ depends on the density of charge carriers and
1108: can get as great (greater than 1.43) as small (less than 1.43) values. The
1109: more detailed study of the two-band model considered in this article with
1110: reduced density of charge carriers and account of all possible pairings of
1111: electrons at phonon mechanism of superconductivity can be found in the
1112: second article of the Ref. \cite{Kochorbe_1}.
1113:
1114: Note that in the most of works about $MgB_{2}$ theoretical and experimental
1115: studies of superconducting properties are made at given value of density of
1116: charge carriers. But there are studies (see for example,\cite{Akis_1},
1117: \cite{Akis_2}) where the dependence of the quantity $T_{C}$ on the density
1118: of charge carriers is investigated on the basis of simplified Moscalenco
1119: model assuming $ V_{11}=V_{22}=0$, $V_{12}\neq 0$ . In order to achieve high
1120: values of $T_{C}$ it was assumed $V_{12}>1$.
1121:
1122: The change of density of charge carriers can be done by doping as it was
1123: suggested in works \cite{Mishonov}, \cite{Rodriguez} where compounds
1124: $MgB_{2-x}C_{x},Mg_{1-y}D_{y} B_{2}(D=Li,Al\;{\rm and\;others})$. have been
1125: studied. Having introduced impurity in such compounds, the interband scattering
1126: of electrons on a impurity, that leads to the decrease of the quantity $T_{C}$
1127: with increase of impurity concentration \cite{Lagos}, occurs as well as the change
1128: of chemical potential. So, in order to confirm the properties 1 - 4 in the
1129: experimental perspective we have to exclude the interband impurity scattering,
1130: to neutralize it somehow. It can be done , for example, by introducing the
1131: impurity outside the plane that is responsible for superconductivity.
1132: Therefore, in this plane the arrangement of electrons is preserved and the
1133: scattering on the impurity potential is absent \cite{Lagos}.\\
1134:
1135: {\bf 2.6 Conclusion}
1136:
1137: We have developed a superconductivity theory for a system with two overlapping
1138: energy bands on the Fermi surface.This theory is valid in the weak-field
1139: approximation for any carrier density, including a low one $(\mu \sim T_c)$.
1140:
1141: The main results are the following.
1142:
1143: 1. We have introduced a system Hamiltonian (1) which can account for
1144: superconducting pairing of electrons both within each band and from different
1145: bands. A system of self-consistent equations was derived for the order
1146: parameter $\Delta_{nm}$ (2) and for the chemical potential $\mu$ (6).
1147:
1148: 2. We have used a sub-band representation, in which the basic equations at
1149: temperatures close to critical can be expressed as the set (16)-(19) for the
1150: four-band model. This set can be used to consider both the phonon and
1151: non-phonon superconductivity mechanisms and can be used to describe
1152: superconductivity in a system with low carrier density $(\mu \sim T_c)$.The low
1153: carrier density notwithstanding, this system can lead to rather high
1154: $T_c$ in view of the inclusion of more interband interactions than in
1155: Ref.\cite{Moskalenko_2}, which are connected with formation of superconducting
1156: pairs of electrons from different bands.
1157:
1158: 3. Our model offers more possibilities of describing various two - band
1159: systems, since a major role is played in the theory by the ratio $N_2 / N_1$
1160: of state densities of electrons from different bands, as well as by interation
1161: constants $\lambda_{nm}(n; m = 1 - 4)$. The foregoing is in fact clearly
1162: demonstrated in Figs.1 and 6a which show respectively the dependences of the
1163: critical temperature $T_c$ and of the electronic heat capacity $(C_S - C_N)_{T
1164: = T_c}$ on the carrier density for a nonphonon superconductivity mechanism and
1165: strong (curves 4 - 5) hybridization.
1166:
1167: 4. The plot of the chemical potential of a superconductor with low carrier
1168: density versus temperature has at the point $T = T_c$ a sharp kink that becomes
1169: less peaked and vanishes when the carrier density is increased. It vanishes at
1170: $\mu \approx 2$ meV in the BCS model \cite{Van} and at $\mu \approx 8$ meV in
1171: the two - band case (Fig. 2). Overlap of the energy band produces thus
1172: favorable conditions for revealing anomalies in the $\mu (T)$ dependence and
1173: hence for elucidating the superconductivity mechanisms.
1174:
1175: 5. Analytic expressions were obtained for the absolute $(C_S - C_N)$ (37)
1176: and relative $(C_S - C_N) / C_N$ (40) jumps of the electron heart capacity at
1177: $T = T_c$. The behavior of $(C_S - C_N)$ as a function of the carrier density
1178: $N_0$ does not, generally speaking, duplicate the behavior of $T_c (N_0)$ (see
1179: Figs. 1 and 6a) and is determined by the parameters of the theory. Situations
1180: are possible in which $T_c$ and $(C_S - C_N)_{T = T_c}$ have maxima at one and
1181: the same density $N_0$ (e.g., curves 3 - 5). The character of the dependence of
1182: the ratio $(C_S - C_N)/C_N$ at the point $T = T_c$ on $N_0$ is also determined
1183: by the parameters of the theory and depend substantially on the carrier
1184: density. This quantity can be either large or small: $(C_S - C_N) / C_N >
1185: 1.43$ or $(C_S - C_N) / C_N < 1.43$.
1186:
1187: 6. At large carrier densities $(\mu \gg T_c)$ we have $\Delta_{12}, \Delta_{21}
1188: \ll \Delta_{11}, \Delta_{22}$, so that simpler equations for $T_c, \Delta_{nm},
1189: (C_S - C_N)_{T = T_c}$ and $(C_S - C_N)/ C_N$ can be obtained by putting
1190: $\Delta_{12} =\Delta_{21} = 0$ in (2) and $N_{3} = N_{4} = 0$ in (19),
1191: (37) and (40). The ensuing results agree with the corresponding equations for
1192: phonon \cite{Moskalenko_2}, \cite{Moskalenko_9} and non - phonon
1193: \cite{Palistrant_2} superconductivity mechanisms.
1194:
1195: We have considered two cases: (a) all constants both for the interband and for
1196: the intraband interaction are positive: ($\lambda_{nm} > 0$, this corresponds
1197: to effective attraction between carriers); (b)$\lambda_{nm} < 0$, (this
1198: corresponds to repulsion). We suppose that case (2) corresponds to the strong
1199: Coulomb interaction when any attraction between carriers is suppressed. The
1200: interband interaction coupling constants which are involved in the $T_c$ square
1201: definition are here the unique reason for generating a superconductivity
1202: mechanism, with the intraband interactions preventing one. Nevertheless, high
1203: values of $T_c$ can be achieved provided
1204: $\lambda_{11}\lambda_{22}<\lambda_{12}\lambda_{21}$. In the case of
1205: $\lambda_{nm}>0$ the high values of $T_c$ are easily acquired without imposing
1206: any restrictions on the values of the parameters $\lambda_{nm}$. The results
1207: concerning the dependences of $2\vert\Delta_n(0)\vert/T_c$ on the carrier
1208: concentration are of great interest to us. These dependences can be
1209: interpreted as the dependences on the oxygen content in the yttrium ceramics
1210: or on the doping content in the lanthanum ones. The supplemental dependence on
1211: the carrier concentration results in our model in a difference from that with a
1212: two-band phonon superconductivity mechanism \cite{Moskalenko_2},
1213: \cite{Moskalenko_9}.
1214:
1215: The theory, represented in this work, gives the essential dependence of the
1216: relative specific heat jump $(C_s - C_N)/C_N$ on the carrier concentration.
1217: In particular, sufficiently small values of this ratio are possible (see
1218: fig. 6a and 6b).
1219: As has been noted before, small values of this quantity have been observed,
1220: for example, in the thallium ceramics \cite{Junod}. So the specific heat
1221: jump dependence on the carrier concentration which has a maximum (see fig. 7)
1222: is of great interest to us. The value of $N_0$ which corresponds to the above
1223: maximum coincides with the one which defines the maximum of the
1224: superconductivity transition temperature. The specific heat dependence on the
1225: doping concentration has been observed, for instance, in $La_{2-x}Sr_xCu_2O_4$
1226: \cite{Hongshum}.
1227:
1228: In the case of electron-phonon mechanism of superconductivity the results
1229: obtained in this work can be used to describe the thermodynamic properties
1230: of superconducting $MgB_{2}$ compound in all scope of densities of charge
1231: carriers (see section 2.5). In this way we have to determine the parameters
1232: of theory for this compound and find suitable experimental technique to
1233: lower the density of charge carriers preferably up to values where $\mu \sim
1234: T_{C}$ . As it was shown in many works the superconductivity in $MgB_{2}$
1235: compound with its inherent density of charge carriers is described by
1236: Moscalenco two-band model \cite{Moskalenko_2}.
1237:
1238: \begin{center}
1239: \bf{3. SUPERCONDUCTIVITY IN TWO-BAND SYSTEM PASSED TO THE STATE OF A BASE
1240: CONDENSATE OF LOCALIZED PAIRS AT T = 0 }.
1241: \end{center}
1242:
1243: {\bf 3.1. Analitical solutions for the basic equations in the meanfield
1244: approximation at T = 0}.
1245:
1246: The two-band superconducting system with an arbitrary carrier concentration
1247: is described by the Hamiltonian (1). Using this Hamiltonian and the Green
1248: functions method \cite{Abricosov_1}, we have obtain the self consistent
1249: system of equations for the order paramiters $\Delta_{nm}(n; m = 1,2)$ and the
1250: chemical potential $\mu\,(2),(6)$. These equations we will study at $T = 0$.
1251:
1252: Since this system of equations cannot be solved in the general form, our goal is
1253: to find these solutions for the region of low carrier concentrations when the
1254: system passes to the state of Bose condensate of localized pairs
1255: \cite{Leggett_2}, \cite{Randeria} $(\mu_n < 0)$ at $(\Delta_{mn}/
1256: \mu_{n,m})^2 \ll 1$. For these conditions, in expressions (2)-(6),
1257: expansion in terms of $\Delta_{nm}^2$ can be performed.
1258:
1259: For the sake of simplicity, redefine the indices $11 \rightarrow 1; 22
1260: \rightarrow 2; 12 \rightarrow 3$. As a result, the system of equations (2) and
1261: (6) can be rewritten as a system of equations for the order parameters of the
1262: three-band model \cite{Palistrant_6}.Limiting ourselves to terms of order
1263: $\Delta_{n}^{2}$, we obtain the system of equations for finding the values ot
1264: the chemical potential $\mu$ and order parameters $\Delta_n (n = 1, 2, 3)$.
1265:
1266: This system of equations and theirs solutions have study in paper
1267: \cite{Palistrant_7}. As follows from this paper the solutions for $\mu$ and
1268: $\Delta_{n}^{2}$ are rather cumbersome and contain a complicated dependence on
1269: the carrier concentration and the intra- and interband interaction constants.
1270: To clarify the role or the overlapping energy bands, some particular cases will
1271: be considered below. Let us consideder two simpler cases here.
1272:
1273: A. The Moskalenko model \cite{Moskalenko_2} assumes the existence of
1274: intraband $V_{11}$ and $V_{22}$ and interband $V_{12}$ electron-electron
1275: interactions corresponding to electron pairing inside one band and transitions
1276: of pairs as a whole from one band to the other. In thise case, $\Delta_3 = 0$
1277: and in the region of a small concentrations $(\mu < 0, \Delta_{n}^{2}/ \mu^2
1278: \ll 1)$ we reduce the equations for $\Delta_1, \Delta_2$ and $\mu$ to the
1279: following form:
1280: \begin{equation}
1281: \mu (1 - \exp \alpha_1) + \frac{\Delta_{1}^{2}}{4} f_1 =\tilde
1282: \zeta_{c_1},\,\,\,\,\,\mu (1 - \exp \alpha_2) + \frac{\Delta_{2}^{2}}{4} f_2
1283: =\tilde \zeta_{c_2},\,\,\,\,\,n_1 F_1 \frac{\Delta_{1}^{2}}{4} + n_2 F_2
1284: \frac{\Delta_{2}^{2}}{4} = \tilde n,
1285: \label{48}
1286: \end{equation}
1287: where
1288: $$
1289: \alpha_1 = \frac{V_{22} - V_{12} \Delta_2 / \Delta_1}{N_{1}^{'}(V_{11}V_{22} -
1290: V_{12}V_{21})},\,\,\,\,\,\,\,\,\, \alpha_2 = \frac{V_{11} - V_{21} \Delta_1 /
1291: \Delta_2}{N_{1}^{'}(V_{11}V_{22} - V_{12}V_{21})},
1292: $$
1293: \begin{equation}
1294: N_{n}^{'} = \frac{N_n}{2},\,\,\,\,\tilde n = \frac{N}{2(N_1 + N_2)},\,\,\,\,
1295: n_n = \frac{N_n}{N_1 + N_2},
1296: \label{49}
1297: \end{equation}
1298: $$
1299: f_n = \frac{(1 - \exp \alpha_n)(1 - \exp 2 \alpha_n)}{\zeta_{cn} \exp
1300: \alpha_n},\,\,\,\,F_n = \frac{(1 - \exp \alpha_n)^2}{\zeta_{cn} \exp \alpha_n},
1301: $$
1302: $$
1303: \tilde \zeta_{cn} = \zeta_{cn} + \zeta_n(1 - exp \alpha_n),
1304: $$
1305: and where $N_n$ is the density of electron states in the $n$th sheet of the
1306: Fermi surface and $\zeta_{cn}$ is the cutoff energy for the integrals. The
1307: solutions of. Eqs.(48) have the form
1308: \begin{equation}
1309: \mu =\left(n_1 \tanh \frac{\alpha_1}{2} + n_2 \tanh
1310: \frac{\alpha_2}{2}\right)^{-1}\,\Biggl\{-\frac{1}{2}\left[n_1 \tilde \zeta_{c_1}
1311: \left(1 - \tanh \frac{\alpha_1}{2}\right) + n_2 \tilde \zeta_{c_2}\left(1 -
1312: \tanh \frac{\alpha_2}{2}\right)\right] + \tilde n\Biggl\},
1313: \label{50}
1314: \end{equation}
1315: $$
1316: \Delta_{1}^{2} = \frac{\zeta_{c_1}}{sinh \frac{\alpha_1}{2}\,\cosh\,
1317: \frac{\alpha_1}{2}} \left( n_1\,\tanh\, \frac{\alpha_1}{2} + n_2\,\tanh\,
1318: \frac{\alpha_2}{2}\right)^{-1} \times
1319: $$
1320: \begin{equation}
1321: \times \Biggl\{\tilde n + n_2 \left[\zeta_{c_1}\tanh \frac{\alpha_2}{2}\left
1322: (\coth \frac{\alpha_1}{2} - 1 \right) + \zeta_{c_2} \left(\tanh
1323: \frac{\alpha_2}{2} - 1 \right)\right]\Biggl\}.
1324: \label{51}
1325: \end{equation}
1326: The expression for $\Delta_2$ can be obtained from Eq.(51) by replacing indices
1327: $1 \leftrightarrow 2$.On the basis of Eq.(51) we obtain the equation for the
1328: ratio $\Delta_1 / \Delta_2$,
1329: \begin{equation}
1330: \left(\frac{\Delta_1}{\Delta_2}\right)^2 =\frac{\zeta_{c_1}\tilde n + n_2
1331: \zeta_{c_1} \left[\zeta_{c_1} \tanh \frac{\alpha_2}{2}(\coth
1332: \frac{\alpha_1}{2} - 1) + \zeta_{c_2}(\tanh \frac{\alpha_2}{2} - 1)\right]
1333: \sinh \alpha_2}{\zeta_{c_2}\tilde n + n_1
1334: \zeta_{c_2} \left[\zeta_{c_2} \tanh \frac{\alpha_1}{2}(\coth
1335: \frac{\alpha_2}{2} - 1) + \zeta_{c_1}(\tanh \frac{\alpha_1}{2} - 1)\right]
1336: \sinh \alpha_1}.
1337: \label{52}
1338: \end{equation}
1339:
1340: Let us pass to the limit of symmetrical bands. In this case,
1341: $\alpha_1=\alpha_2$, $\zeta_{c_1}=\zeta_{c_2}$, $n_1 + n_2=1$ and formulas
1342: (51) transform into
1343: \begin{equation}
1344: \mu=-\frac{\tilde \zeta_{c_1}}{2}\left[\coth\frac{\alpha_1}{2} - 1\right] +
1345: \tilde n\coth\frac{\alpha_1}{2},\ \ \ \Delta_1^2=\Delta_2^2=\frac{\tilde n
1346: \zeta_{c_1}}{\sinh^2\alpha_1/2}.
1347: \label{53}
1348: \end{equation}
1349: Formulas (53) coincide in form with analogous expressions for one-band
1350: superconductors \cite{Gordar}, and the difference is in the determination
1351: of $\alpha_1$ (49), including interband interactions in addition to intraband
1352: interactions. This fact leads to a decrease in the denominator in the second
1353: formula in (53), and consequently, to an increase in $\Delta_1$, in comparison
1354: with the one-band case. At the same time, a more essential shift of $\mu$
1355: towards the negative region occurs under the condition of weak changes in the
1356: value of $\zeta_{c_1}$. As follows from (51), in the case of an anisotropic
1357: two-band system, the formulas are modified due to the difference in the
1358: parameters of the considered systems. So, for example, in the determination of
1359: $\Delta_1$ for the carrier concentration $\tilde n$, besides renarmalization of
1360: the coefficient, an additional term arises due to the presence of the second
1361: band ($n_2 \not = 0$).
1362:
1363: As follows from formulas (50), (51), the dependence of $\mu$ and $\Delta_n$
1364: on the carrier concentration has a complex character due to the ratio
1365: $\Delta_1/\Delta_2$ entering the expressions determining $\alpha_1$ and
1366: $\alpha_2$. The $\Delta_1/\Delta_2$ ratio itself is determined by Eq. (52).
1367: From this formula it is easy to see that for a weakly anisotropic system (the
1368: bands are nearly similar), the dependence of $\Delta_1/\Delta_2$ on the carrier
1369: concentration can be neglected. As a result, $\mu$ and $\Delta_n$ are
1370: determined by the explicit dependence on $\tilde n$ (51), as in the case of a
1371: one-energy band. In the case of a strongly anisotropic system, the dependences
1372: of $\mu$ and $\Delta_n$ must be determined from formulas (50), (51) and
1373: equation (52). The result can be esentially different from the case of one-band
1374: superconuctors.
1375:
1376: In the behavior of a superconducting system with a small carrier concentration,
1377: the existence of a two-particle bound state plays an important role
1378: \cite{Leggett_2}, \cite{Gorbatsevich_1}. The equation for determining
1379: the binding energy $\varepsilon_b$ in a two-band system is obtained by
1380: considering the two-particle "density-density" Green's function at
1381: $\varepsilon_F=0$ and the corresponding equation for the vertices
1382: \cite{Palistrant_6}, \cite{Palistrant_8},
1383: $$
1384: \Gamma_{nm} = - V_{nm} + \sum_{n_1} V_{nn_1} \xi_{n_1}(|\varepsilon_b|)
1385: \Gamma_{n_1 m}.
1386: $$
1387: \begin{equation}
1388: \xi_n (|\varepsilon_b|) = \sum_{\vec k}\left[2 \varepsilon_n (\vec k) +
1389: |\varepsilon_b| \right]^{-1}.
1390: \label{54}
1391: \end{equation}
1392:
1393: This system is written for the case of $\Delta_n = 0$ and $\varepsilon_F = 0$
1394: and corresponds to the two - particle interaction without accounting for
1395: multiparticle effects.The bound state of two particles arises when the vertex
1396: diveges. Consequently, on the basis of Eq.(54), the following equation for the
1397: bound-state energy $\varepsilon_b$ is obtained:
1398: \begin{equation}
1399: 1 - V_{11} \xi_1 (|\varepsilon_b|) - V_{22} \xi_2 (|\varepsilon_b|) +
1400: (V_{11} V_(22) - V_{12 }^2) \xi_1(|\varepsilon_b|) \xi_2 (|\varepsilon_b|)
1401: = 0.
1402: \label{55}
1403: \end{equation}
1404: From the condition $N_0 \rightarrow 0$, we obtain \cite{Palistrant_9}
1405: \begin{equation}
1406: 1 - V_{11} \xi_1 (2 \eta) - V_{22} \xi_2 (2 \eta) +
1407: (V_{11} V_(22) - V_{12 }^2) \xi_1(2 \eta) \xi_2 (2 \eta) = 0.
1408: \label{56}
1409: \end{equation}
1410: Comparing the two last equations, we can state that as $N_0 \rightarrow 0$,
1411: \begin{equation}
1412: |\varepsilon_b| = 2 \eta,\,\,\,\,\,(\eta = - \mu).
1413: \label{57}
1414: \end{equation}
1415:
1416: B. All the interaction constants are equal to zero $(V_{nm}^{ke} = 0)$,
1417: except $V_{12}^{V_{21}}$. In this case $\Delta_1 = \Delta_2 = 0$ and we
1418: have the following expression for $\Delta_3$ and $\mu$:
1419: \begin{equation}
1420: \Delta_{3}^{2} = \Delta_{12}^{2} = \frac {\tilde n (1 + m_1 / m_2)
1421: (\zeta_{c_1} + \zeta_{c_2})}{8 n_1 \sinh^2 \beta_1 / 2}
1422: \label{58}
1423: \end{equation}
1424: \begin{equation}
1425: \mu = - (\zeta_{c_1} + \zeta_{c_2}) \frac{1}{4} \left(\coth
1426: \frac{\beta_1}{2} - 1 \right) + \frac{\tilde n (1 + m_1 / m_2)}{4 n_1}\coth
1427: \frac{\beta_1}{2},
1428: \label{59}
1429: \end{equation}
1430: Where $m_n$ is the effective mass of the electron in $n$ - band.
1431: \begin{equation}
1432: \zeta_{c_2} = \frac{m_1}{m_2} \zeta_{c_1},\,\,\,\,\,\beta_1 = \frac{1 + m_1
1433: / m_2}{4 N_{1}^{'} V_{12}^{21}}.
1434: \label{60}
1435: \end{equation}
1436: If we assume that the bands are symetric, i.e., $\zeta_{c_1} =
1437: \zeta_{c_2}$,$m_1 = m_2, n_1 = 1 / 2$, the above formulas are modifided
1438: into the expressions
1439: \begin{equation}
1440: \mu = - \zeta_{c_1}\frac{1}{2} \left(\coth \frac{\beta_1}{2} - 1 \right)
1441: + \tilde n \coth \beta_1 / 2 ,\,\,\,\,\,\Delta_{3}^{2}
1442: =\frac{\zeta_{c_1} \tilde n}{\sinh^2 \beta_1 / 2},
1443: \label{61}
1444: \end{equation}
1445: whose form resembles (53).
1446:
1447: The above considered case of the presence of one order parameter
1448: ($\Delta_{12}$) in the system leads to a concentration dependence analogous to
1449: the case of symmetrical bands.
1450:
1451: For a system with Hamiltonian (1), the expression for the difference in
1452: thermodynamic potentials for the states with $\Delta_{nm} \not = 0$ and
1453: $\Delta_{nn} = 0$ can be written as (35). Using the calculation technique
1454: reported in \cite{Palistrant_2} and the system of equations for
1455: $\Delta_n$, we can reduce the difference (35) to the form
1456: \begin{equation}
1457: \frac{\Psi (\Delta) - \Psi (0)}{V} = \frac{1}{8} N_{1}^{'}\left[ f_{1}
1458: \Delta_{1}^{4} + f_2 \,\Delta_{2}^{4} \frac{N_2}{N_1} + 2 f_3\,\Delta_{3}^{4}\,
1459: \right],
1460: \label{62}
1461: \end{equation}
1462: where $f_n < 0$ \cite{Palistrant_7}. As a concequence, the state with $\Delta_n
1463: \not = 0$ is more advantageous than the trival solution $\Delta_n = 0$. The
1464: advantage increases as the number of order parameters increases in connection
1465: with accounting for all kinds of inter- and intraband interactions.
1466:
1467: In the case of symetrical bands, expression (62) can be rewritten in the form
1468: \begin{equation}
1469: \frac{\Psi (\Delta) - \Psi (0)}{V} = -2 N^{'}\coth \frac{\alpha_1}{2}\,\tilde
1470: n^2
1471: \label{63}
1472: \end{equation}
1473: This formula coincides in form with an analogous expression for the one - band
1474: superconductor \cite{Gordar}.
1475:
1476: {\bf 3.2 Application of the path integral method.}
1477:
1478: In this section the path integral method as applied to the two - band model is
1479: developed and, on this basis, the procedure for transition from the Fermi to
1480: Bose elementary exitations is given.
1481:
1482: We start from following Hamiltonian describing the two-band systems:
1483: $$
1484: H = \int\,dr\,\sum_{n \sigma} \Psi_{n \sigma}^{+}(\tau) H_{0n} \Psi_{n \sigma}
1485: (\tau) - \frac{1}{2} \sum_{nm} V_{n m} \int dr \sum_{\sigma\,\sigma^{'}}
1486: $$
1487: \begin{equation}
1488: \Psi_{n \sigma}^{+}(\tau)\Psi_{n \sigma^{,}}^{+} (\tau) \Psi_{n
1489: \sigma^{'}} (\tau) \Psi_{n \sigma}(\tau)(1 - \delta_{\sigma \sigma^{'}}),
1490: \label{64}
1491: \end{equation}
1492: where the first term corresponds to the kinetic energy, the second is
1493: responsible for the superconductivity, the band indices $n; m = 1, 2, V_{nm}$
1494: are the constants of intra- and interband interaction, $\Psi_{n
1495: \sigma}^{+}(\tau)$ is the particle creation operator in the band $n$ with
1496: spin $\sigma$, and $H_{0n} = \frac{\bigtriangledown^2}{2 m_n} - \mu_n$.
1497: Hamiltonian (1) correspond to expression of the Moskalenko model
1498: \cite{Moskalenko_2} considering only intarband pairind and the transitions of
1499: Cooper pairs as a whole from one band to another.
1500:
1501: The expression for the statistical sum of a two-band superconductor after
1502: introducting additional scalar fields $\Delta_n$, Hubbard-Stratonovich
1503: transformations, integrations under Fermi fields, and transition into $n k
1504: \Omega$ representation has the following form \cite{Palistrant_10} (the
1505: detailed version see \cite{Palistrant_9}):
1506: \begin{equation}
1507: Z = C \int \left(\prod_{i = 1,2} d \Delta_{i}^{*} d \Delta_{i}\right) \exp {-
1508: S_{eff}^{(2)}},
1509: \label{65}
1510: \end{equation}
1511: where
1512: $$
1513: S_{eff}^{(2)} = \frac{1}{\beta}\,\sum_{nm}\,\sum_{\vec q
1514: \omega}\,\Delta_{n}^{*} (\vec q \omega)
1515: $$
1516: \begin{equation}
1517: \biggl\{ V_{nm}^{-1} + \Pi_n(\vec q\omega)\delta_{nm}\biggl\} \Delta_m (\vec
1518: q \omega),
1519: \label{66}
1520: \end{equation}
1521: \begin{equation}
1522: \Pi_n (\vec q, \omega) = \frac{1}{\beta}\sum_{\vec k \Omega} G_{- n} (\vec k
1523: \Omega) G_{+ n} \left(\vec k - \vec q, \Omega - \omega \right)
1524: |\chi_n (\vec k, \vec k - \vec q)|^2,
1525: \label{67}
1526: \end{equation}
1527: \begin{equation} G_{\pm
1528: n}^{-1}(k \Omega) = - i \Omega \pm \frac{k^2}{2 m_n} \mp \mu_n.
1529: \label{68}
1530: \end{equation}
1531:
1532: Here $V_{nm}^{- 1}$ is the matrix element of inverse to the interaction matrix
1533:
1534: $$ \hat V = \left( \begin {array}{ccc}
1535: V_{11} & V_{12}\\
1536: V_{21} & V_{22}
1537: \end {array} \right). $$
1538:
1539: and $\chi_n (\vec k, \vec k - \vec q)$ are the integrals over the elementary
1540: cell under the Bloch functions.
1541:
1542: Further calculations are connected as in the case of a single band
1543: \cite{Gorbatsevich_1} with the existence of a nonzero solution of Eq. (55),
1544: determining the energy of the bound state $\varepsilon_b$, and we consider
1545: the case $T \ll |\varepsilon_b|$.
1546:
1547: In the mean-field approximation $\mu = \varepsilon_b / 2$ (57).Let us suppose
1548: that $\mu$ is near $\varepsilon_b / 2$. After integrating under $\Omega$ in
1549: formula (67) we make the expansion under $\omega, \vec q$ and the difference
1550: $(2 \mu - \varepsilon_b)$. After that we acquire
1551: \begin{equation}
1552: \Pi_n (\vec q, \omega) = -\xi_n (|\varepsilon_b| ) + \gamma_n \left[ - i \omega
1553: + \frac{q^2}{2 m_{n}^{*}} - \mu^{*}\right] ,
1554: \label{69}
1555: \end{equation}
1556: were
1557: \begin{equation}
1558: \gamma_n =- \frac{\partial \Pi_n}{\partial(2 \mu)}\,\,\,\,\mu^{*} = 2 \mu -
1559: \varepsilon_b,\,\,\,\frac{1}{m_{n}^{*}} = \gamma_{n}^{- 1} \frac{\partial^{2}
1560: \Pi_n} {\partial \vec q^2}.
1561: \label{70}
1562: \end{equation}
1563:
1564: Let us substitute (66) into (65) and perform the functional integration under
1565: additional Bose fields, leading to the canonical form by orthogonal
1566: transformation. We obtain
1567: \begin{equation}
1568: \ln Z = - \sum_{\vec q \omega} \ln \biggl\{1 + V_{11} \Pi_1 (\vec q, \omega) +
1569: V_{22}\Pi_2 (\vec q, \omega) + \left(V_{11} V_{22} - V_{12}^{2}\right) \Pi_1
1570: (\vec q, \omega) \Pi_2 (\vec q, \omega) \biggl\}.
1571: \label{71}
1572: \end{equation}
1573: We substitute in this formula expressions for the $\Pi_n (\vec q, \omega)$ [see
1574: Eq. (69)] and take into account Eq. (55), determining the energy of the bound
1575: state $\varepsilon_b$, and also relationships
1576: \begin{equation}
1577: V_{11} - \left(V_{11} V_{22} - V_{12}^{2} \right) \xi_2 = V_{21} \Delta_1 /
1578: \Delta_2,\,\, V_{22} - \left(V_{11} V_{22} - V_{12}^{2} \right) \xi_1 = V_{12}
1579: \Delta_2 / \Delta_1,
1580: \label{72}
1581: \end{equation}
1582: resulting from the system of equations for the order parameters $\Delta_n$.
1583: Thus we obtain
1584: \begin{equation}
1585: \ln Z = -\sum_{\vec q \omega} \ln \Biggl\{ \gamma_1 V_{21}
1586: \frac{\Delta_1}{\Delta_2} \left(- i \omega + \frac{q^2}{2 m_{1}^{*}} - \mu^*
1587: \right) + \gamma_2 V_{12} \frac{\Delta_2}{\Delta_1} \left( - i \omega +
1588: \frac{q^2}{2 m_{2}^{*}} - \mu^* \right)\Biggl\}.
1589: \label{73}
1590: \end{equation}
1591: For the charge carrier density we have
1592: \begin{equation}
1593: n = \frac{1}{\beta}\,\frac{\partial}{\partial \mu^*} \ln Z = \frac{1}{\beta}
1594: \sum_{\vec q \omega} \left(- i \omega + \frac{q^2}{2M^*} - \mu^{*} \right)^{-1}
1595: = \sum_{\vec q} \left(\exp \left[ \beta \left( \frac{q^2}{2 M^*} - \mu^*
1596: \right)\right] - 1 \right)^{-1},
1597: \label{74}
1598: \end{equation}
1599: were
1600: \begin{equation}
1601: (M^*)^{-1} = \frac{1}{m_{1}^{*}} \left(\gamma_1 \frac{\Delta_1}{\Delta_2} +
1602: \gamma_2 \frac{\Delta_2}{\Delta_1} \frac{m_{1}^{*}}{m_{2}^{*}}\right)
1603: \left(\gamma_1 \frac{\Delta_1}{\Delta_2}+\gamma_2 \frac{\Delta_2}{\Delta_1}
1604: \right)^{-1}.
1605: \label{75}
1606: \end{equation}
1607: Expression (75) has been obtained in the approximation in which only the
1608: quadratic part of complete action is taken into account. In this approximation
1609: we obtain the ideal Bose gas with effective mass $M^*$ and chemical potential
1610: $\mu^* = 2 \mu - \varepsilon_b$.In distinction from the case of a
1611: single band \cite{Gorbatsevich_1}, here $\varepsilon_b$ and $\mu$ are
1612: determined on the two-band basis (55), (50) and $M^*$ is defined by (75). In
1613: the case where the effective masses from different bands are equivalent
1614: $(m_{1}^{*} = m_{2}^{*})$, the relationship $M^* = m_{1}^{*} = m_{2}^{*} $
1615: results from the last formula. If the energy band with effective mass
1616: $m_{1}^{*}$ overlaps with the wider band $(m_{1}^{*} >m_{2}^{*})$, in the field
1617: with low charge carrier density a lighter Bose gas arises $(M^* < m_{1}^{*})$
1618: in comparison with the case of a single-band superconductor effective electron
1619: mass of which is equal to $m_{1}^{*}$.
1620:
1621: The system undergoes transition to the state of Bose condensation at the point
1622: $\mu^* = 0$ and this relationship is the condition for determining the
1623: condensation temperature $T_k$.After summation under $\vec q$ in (74) at $\mu^*
1624: = 0$, we obtain
1625: \begin{equation}
1626: T_k = a_D \frac{n^{2 / D}}{M^*},
1627: \label{76}
1628: \end{equation}
1629: where $D$ is the dimension of system and $a_D$ is a constant. We see that
1630: $T_k$ depends on the charge carrier density according to the law $n^{2 / D}$ as
1631: in the case of a single - band superconductor but only for systems with
1632: equivalent bands $(M^* = m_{1}^{*} = m_{2}^{*})$. If the system is strongly
1633: anisotropic $(m_{1}^{*} \not = m_{2}^{*})$, the quantity $M^*$ has the
1634: additional inexplicit dependence on the charge carrier density through the
1635: same dependence of the relation $\Delta_1 / \Delta_2$. From the formulas (75)
1636: and (76) we see that the existence of a second wider energy band can lead to
1637: an increase in the condensation temperature $T_k$ in comparison with the case
1638: of a single energy band. The expression for the effective action (66) can be
1639: considered as an improved mean-field approximation, which can be also applied
1640: at $T \not = 0$.
1641:
1642: Taking into account the residual interaction between bosons, the effective
1643: action has the following form:
1644: \begin{equation}
1645: S_{eff} = S_{eff}^{(2)} + S_{eff}^{(4)}+... ,
1646: \label{77}
1647: \end{equation}
1648: where
1649: \begin{equation}
1650: S_{eff}^{(4)} = \frac{1}{\beta}\,\sum_{\vec q \omega}\biggl\{|\Delta_1|^4 \pi_1
1651: +|\Delta_2|^4 \pi_2\biggl\} ,
1652: \label{78}
1653: \end{equation}
1654: \begin{equation}
1655: \pi_n = \frac{1}{2 \beta} \sum_{\vec k \Omega}\left[ G_{+ n}(k \Omega)\right]^2
1656: \left[G_{- n}(k \Omega)\right]^2 = \sum_{\vec k} \left[2 \varepsilon_n (\vec k)
1657: + \varepsilon_b \right]^{- 3}.
1658: \label{79}
1659: \end{equation}
1660: After renormalization of the Bose fields
1661: \begin{equation}
1662: \Delta_n = \left(\frac{\partial \Pi_n}{\partial (2|\mu|)}\right)^{- 1/2}
1663: \varphi_n
1664: \label{80}
1665: \end{equation}
1666: we have
1667: \begin{equation}
1668: S_{eff}^{(4)}= \frac{1}{\beta}\sum_{n}\,\sum_{\vec q
1669: \omega}|\varphi_{n}(\vec q \omega)|^2 \pi_{n}^{'}(\vec q,
1670: \omega),
1671: \label{81}
1672: \end{equation}
1673: \begin{equation} \pi_{n}^{'} = \left(\frac{\partial
1674: \Pi_n}{\partial(2|\mu|)}\right)^{- 2}\pi_n.
1675: \label{82}
1676: \end{equation}
1677: For the quantity $\pi_{n}^{'}$ determining the potential of two-particle
1678: interaction in the $n th$ band we obtain
1679: \begin{equation}
1680: \pi_{n}^{'}\sim m_n^{-D / 2}|\varepsilon_b|^{1 - D / 2}
1681: \label{83}
1682: \end{equation}
1683:
1684: This quantity depends essentially on the system dimension and is negligible
1685: in the case of three-edimensional. This conclusion consides with the case
1686: of a single band \cite{Gorbatsevich_1}. Thus , omiting the residual
1687: interaction, which is equivalent to the mean-field approximation, is valid only
1688: for three-dimensional systems.
1689: \\\\
1690: {\bf 3.3 Conclusions and discussion}
1691:
1692: In the section 3.1 the theory of superconductivity for a two - band system
1693: with a low carrier density at $T = 0$ was constructed. All types of
1694: electron-electron interband and intraband interactions were accounted for . As
1695: a result, three diferent order parameters, $\Delta_{11}, \Delta_{22}$, and
1696: $\Delta_{12}$,\cite{Palistrant_7} arise in the sistem. The sistem of
1697: equations for the order parameters and the chemical potential $\mu$ was
1698: obtained. The state of the system in the region of low carrier concentrations,
1699: where $\mu_n < 0 \,\,(n = 1,2)$ (corresponding to a Bose condensate of
1700: localized pairs), was considered. Under the assumption that
1701: $\Delta_{nm}^{2}/\mu_{n,m}^{2} \ll 1 $, the expansion in terms of $\Delta_
1702: {nm}^{2}$ was conducted. This allowed the analytical expressions for $\mu$ and
1703: $\Delta_n$ to be obtained. The expression for the difference in thermodynamic
1704: potentials for $\Delta_{nm} \not = 0$ and $\Delta_{nn} = 0$ was also obtained.
1705:
1706: On the basis of these investigations the following conclusions are made:
1707:
1708: 1. The overlaping of energy bands on the Fermi surface in a system with low
1709: carrier density gives rise to three order parameters. In the limit of the
1710: symmetrical bands $(\Delta_{11} = \Delta_{22} = \Delta_{12})$, the values of
1711: these parameters are higher than in the case of the one-energy band, i.e., the
1712: overlapping of energy bands is favorable to setting up the superconducting
1713: properties. At the same time, the value of $\mu$ shifts towards the negative
1714: value region, accelerating the transition of the system from the BCS pattern to
1715: the mode of a Bose condensate of localized pairs.
1716:
1717: 2. The order parameters $\Delta_{nm}$ and the chemical potential $\mu$ are
1718: complex functions of inter- and intraband interaction constants, as well as of
1719: carrier concentrations. In the weak anisotropy case (the bands are nearly
1720: symmetrical) the dependences of $\Delta_{nm}$ and $\mu$ on the carrier
1721: concentrations are analogous to those for the case of one-band superconductors.
1722: For strongly anisotropic systems (the bands are different), the concentration
1723: dependences of $\Delta_{nm}$ and $\mu$ can differ from the case of one-band
1724: superconductors. This is because, in addition to the explicit dependence on the
1725: carrier concentration, a nonexplicit dependence exists through the ratios
1726: $\Delta_{11} /\Delta_{22},\,\Delta_{12} /\Delta_{22}$, etc.
1727:
1728: 3. The difference in the thermodynamic potentials (62) shows that the
1729: considered condensed state is advantageous at $\Delta_{nm} \not = 0$. The
1730: advantage of this state increases as the number of order parametrs and,
1731: consequetly, the namber of intraelectron constants, increases.
1732:
1733: To make the results more illustrative, two simpler cases - the limiting case of
1734: the Moskalenko model [2], where $\Delta_{11}, \Delta_{22} \not = 0,
1735: \Delta_{12} = 0$, as well as the case where $\Delta_{11} = \Delta_{22} = 0,
1736: \Delta_{12} \not =0$ - were considered. We have show that overlapping of the
1737: energy bands facilitate superconductivity and increase the critical
1738: concentration of the carriers at which a transition from the BCS scenario to
1739: the local-pair condensate scenario occurs. In the case of different bands
1740: (especially, for strong anisotropy), the dependence of $\Delta_{nm}$ and $\mu$
1741: on the carrier concentration may be different from the one-band superconductor
1742: case because of their additional implicit dependence on $\tilde n$ through the
1743: ratio $\Delta_1 / \Delta_2$. We also obtained an equation for the bound-state
1744: energy $\varepsilon_b$ on a two-band basis and established the relation
1745: $\varepsilon_b = 2 \mu$.
1746:
1747: In section 3.2 is developed the path integral method for two-band
1748: superconductors and, on this basis, inprove the mean field approximation to
1749: make it applicable for $T \not = 0$.
1750:
1751: To do this, we introduced action, wich takes into account the existence of
1752: intra - and interband interactions and additional boson fields. In the equation
1753: for the statidtical sum, we performed the Hubbard-Stratanovich transformation
1754: generalized to the two-band case. Having perfomed the integration over Fermi
1755: fields, and expansion in terms of $\hat \Delta_n$ with an accuracy to quadratic
1756: terms, we came to effective action (66),(67). We connected the calculation of
1757: $\Pi_n (\vec q, \omega)$ with the existence in the system of a bound state with
1758: the bound-state energy $|\varepsilon_b| \gg T$ and performed the expansion of
1759: $\Pi_n (\vec q, \omega)$ in powers of $\omega,\,q^2$, and
1760: $\left(\varepsilon_b - 2 \mu \right)$. Along the way, we came to statistical
1761: sum for the ideal Bose gas with a renormalized chemical potential $\mu^* =
1762: 2 \mu - \varepsilon_b$ and effective mass $M^*$ (75).
1763:
1764: We also determined the condensation temperature $T_k$, Eq. (76). In the case
1765: of different energy bands, considerable renormalization connected with the
1766: existence of two energy bands occurs. At the same time, the dependence of $T_k$
1767: on the carrier concentration is explicity determined by the dimension of the
1768: system and is implicitly determined by the dependence of $M^*$ on the carrier
1769: concentration.
1770:
1771: The contribution of the residual Bose interaction to the effective action was
1772: also calculated. In analogy to the case of a one-band superconductor, this
1773: contribution is unessential only in three-dimensional systems
1774: \cite{Gorbatsevich_1}. For such systems, the mean field approximation can be
1775: applied, while for systems of reduced dimension, the mean field approximation
1776: appears to be insufficient.
1777:
1778: The main results of the section 3.2 are following:
1779:
1780: 1. On the basis of the functional integration method, the procedure of the
1781: transition from the BCS scenario to the Schafroth one is developed for the
1782: two-band superconductor, when changing the charge carrier density. In the
1783: mean-field approximation the ideal Bose gas with renormalized chemical
1784: potential $\mu^* = 2\mu - \varepsilon_b$ and effective mass $M^*$ is
1785: obtained.These quantities are dettermined on the basis of the two -
1786: band model. The anisotropy of the energy bands plays the critical role in the
1787: renormalization of the effective mass. In particular, at $m_1 > m_2$ the easier
1788: Bose gas $( M^* < m_{1}^{*})$ appears in comparison with the case of a single
1789: band with effective mass $m_1$.
1790:
1791: 2.In the case of strongly anisotropic two-band systems $(\Delta_1 \not =
1792: \Delta_2)$ the possibility of increasing the temperature of the Bose
1793: condensation of localized pairs $T_k$ in comparison with the case of the
1794: single energy band is shown.
1795:
1796: 3. The contribution of residual interaction between bosons to the effective
1797: action depends essentially on the system dimension and has a low value only
1798: for a three-dimensional systems.Therefore for systems with reduced dimension
1799: it is necessary to depart from the mean-field approximation.\\
1800: \begin{center}
1801: \bf{4. Summary}.
1802: \end{center}
1803:
1804: High-$T_{C}$ superconductors as was mentioned in Introduction are very
1805: complicated compounds, and it is impossible to take into account all their
1806: features simultaneously. So, the theoretical studies are
1807: performing on the basis of the models, which take into account some
1808: characteristic features of these systems.
1809:
1810: The goal of this articl is to make the review of the
1811: papers, in which two peculiarities are taken into consideration
1812: - the overlapping of the energy bands on the Fermi surface and the reduced
1813: (or low) density of the charge carriers.
1814:
1815: In the two-band model proposed by prof. Moskalenko \cite{Moskalenko_2} the
1816: Cooper pairs appears in every energy band and pass as a whole from one
1817: enrgy band to another.
1818:
1819: In the systems with the low density of the charge carriers there is the
1820: necessity to take into consideration all possible additional pairings
1821: of electrons from different energy bands \cite{Kochorbe_1} and go beyond the
1822: approximation, in which are taken into account only the diagonal over the
1823: band indices Green functions \cite{Moskalenko_9}. The basic equations of the
1824: theory of superconductivity in such two-band systems is reduced to the
1825: representation of four energy pseudo-bands. In this case there appears three
1826: energy gaps $\Delta_{11}$, $\Delta_{22}$ and $\Delta_{12}=\Delta _{21}$ and it
1827: is also possible to obtain the values of the temperature of the superconducting
1828: transition $T_c$, characteristic to the high-$T_C$ materials, even at
1829: the low density of the charge carriers as in the case of the electron
1830: attraction ($\lambda_{nm}>0$), as in the case of their repulsion
1831: ($\lambda_{nm}<0$). The main attention in this paper is devoted to the
1832: dependence of the thermodynamic quantities on the density of charge carriers.
1833: In particular, the behavior of the quantity $T_c$ and the jump of electron
1834: heat capacity ($C_S-C_N$) at the point $T=T_c$ as a function of the density of
1835: the charge carriers is determined by the degree of the overlapping of the
1836: energy bands, of the filling of these bands, relation between the constants
1837: of the electron-electron interaction $\lambda_{nm}$ ($n; m=1-4$) and the
1838: ratio of electron density states $N_1 / N_2$.
1839:
1840: As, for example, in the case of strong hybridization $T_c$ and
1841: $(C_S-C_N)$ as function of the concentration of the charge carriers
1842: are represented by the bell-shaped dependence, which is
1843: observed in the experiment in the number of the oxide ceramics. The
1844: represented theory allows to obtain as small $(C_s-C_N)/C_N<1.43$, as
1845: large $(C_s-C_N)/C_N>1.43$ values for the relative jamp of the electron heat
1846: capacity. This picture is observed in the high-temperature materials
1847: \cite{Junod}-\cite{Akis_2}. These studies have fulfiled in the
1848: representation of the BCS superconducting Cooper pairs. At decreasing the
1849: density of charge carriers at the point $\mu =0$ there takes place the
1850: crossover BCS state-Bose condensation of the localized pairs. In the state of
1851: the deep Bose-condensation, when $\mu < 0$ and the relation
1852: $\Delta_{n}^{2}/\mu^2 \ll 1$ takes place at $T=0$, we obtain that analytical
1853: expressions for the quantities $\Delta_n$ and $\mu$ contain additional
1854: inexplicit dependence on the density of charge carriers in comparison with the
1855: single-band system, due to the consideration of the overlapping of the energy
1856: bands on the Fermi surface.
1857:
1858: Given in the article the method of the functional integration applied to the
1859: two-band system with low density of the charge carriers demonstrates the
1860: transition procedure from the Fermi to the Bose elementary excitations at $ T
1861: \not = 0$. We have obtain ideal Bose-gas with the effective mass $M^{*}$ and
1862: chemical potential $\mu^*=2\mu- \varepsilon_b$ ($\varepsilon_b$ is the energy
1863: of the bound two-particle state). These quantities are determined on the
1864: two-band basis.
1865:
1866: The important results of the given work are: the overlapping of the energy
1867: bands on the Fermi surface promote to the
1868: appearing of the superconductivity, to explain some
1869: experimental data in the investigations of the thermodynamical properties of
1870: the system, intensity the number of the effects,
1871: characteristical to the systems with lowered density of the charge carriers,
1872: and assist their experimental confirmation. These effects, for example, are
1873: the appearing of the kink in the temperature dependence of the chemical
1874: potential $\mu(T)$ in the point $T=T_c$, and also crossover the state
1875: BCS - Bose condensation of the local pairs. The theory proposed in the paper
1876: can be applied to the oxide ceramics and also to the high-temperature
1877: composition $MgB_2$ and other compositions, in which takes place the
1878: overlapping of the energy bands on the Fermi surface and there is possibility
1879: to change the concentration of the charge carriers.
1880:
1881: In addition to the overlapping of the energy band, an important factor of the
1882: band structure of high-temperature materials is the existence of high symmetry
1883: points in the momentum space that lead to singularies in the electron densitty
1884: of states (Van Hove singularities), as well as to topological electron
1885: transitions. The behavior of these singularities dependends, to a large extent,
1886: on the dimension of the system.
1887:
1888: Numerouse theoretical and experimental investigations have confirmed that in
1889: quasi two-dimensional systems (the majority of high-temperature materials are
1890: just such system), a logarithmic divergence of the density of states occurs in
1891: the vicinity of half filling of the energy band at the point where the Fermi
1892: surface intersects the Brillouin zone boundary. This corresponds to the change
1893: from electron to hole topology on the Fermi surface.
1894:
1895: This singularity gives rise to high superconducting transition temperatures
1896: $T_c$ and anomalies in the isotope effect and specific heat(see the review in
1897: \cite{Markiewicz}), as well as providing an explanation for experimental
1898: data on the temperature dependence of the resistance and thermal EMF in high-
1899: $T_c$ materials \cite{Palistrant_11}, etc. Photoemission experiments have shown
1900: that logaritmic divergence is a characteristic of the $La_{2 -x} Sr_x CuO_4$
1901: compound \cite{Gofron}. At the same time, the densities of
1902: electron states may have stronger divergences, such as $E^{-1 / 4}$ and $E^{-1
1903: / 2}$, in $Y Ba_2 Cu_3 O_{7 - \delta}$ and $BiSrCaCu_2O_8$ compounds,
1904: respectively ($E$ is energy counted from the Van Hove singularity), and
1905: correspond to the so-called extented Van Hove sigularities
1906: \cite{Abricosov_2}.
1907:
1908: In a number of articlies, several attempts have been made to explain high-$T_c$
1909: by accounting only for these singularities of the density of electron states
1910: and assuming that the pair interaction is universal for all compounds. In the
1911: weak-coupling approximation, accounting for these singularities results in a
1912: larger increase in the superconducting transition temperature $T_c$ in
1913: comparison with the case of logarithmic divergence . In the
1914: case of strong coupling, the picture changes, leading to an essential decrease
1915: in the role of the extented Van Hove singularities that are spread apart due to
1916: the inelastic scattering of electrons \cite{Radtke}.
1917:
1918: It should be noted that we investigated the influence of the density of the
1919: state singularities of type $E^{- 1 / 2}$ on the thermodynamic and kinetic
1920: proterties of superconductors long before the discovery of high-$T_c$
1921: superconductivity \cite{Palistrant_12}, \cite{Palistrant_13}. In
1922: particular, we showed that the scattering of electrons on the impurity
1923: potential decreased the role of that singularity in the determination of $T_c$.
1924:
1925: In studying the properties of high-temperature materials, it is probably
1926: insufficient to consider only the existence of extended Van Hove singularities.
1927: It is necessary to use a more complex approach that accounts for energy band
1928: overlapping,reduced corrier concentrations, nonadiabacity and the other factors
1929: characteristic of high-temperature materials, especially becous the two-band
1930: model is capable of explaining a large amount of experimental a great number
1931: \cite{Moskalenko_5} - \cite{Moskalenko_8}, \cite{Kochorbe_1},
1932: \cite{Palistrant_14}, \cite{Palistrant_15}, in pure and doped high - $T_c$
1933: superconductors including $MgB_2$ \cite{Todor}.
1934:
1935: The discovery of high $T_c$ superconductivity in $MgB_2$ and in some other new
1936: materials with the lowered density of the charge curriers persuade us that
1937: given reviw article is well useful for theoretical and experimental
1938: investigations.
1939:
1940: We note that there are many authors at present which carry out the theoretical
1941: investigations of the physical properties of $MgB_2$ on the bases of two - band
1942: model (see, for example, in book \cite{Rare}, also in
1943: \cite{Bud'ko}-\cite{Choi_3},\cite{Rodriguez}, \cite{Lagos} and al. and
1944: references in them).
1945:
1946: The information about theory investigations of thermodynamic and kinetic
1947: properties of many-band superconductors, elaborated long before discovery of
1948: superconductivity in $MgB_2$, are in arXiv: Cond. Mat.
1949: \cite{Palistrant_1},\cite{Moskalenko_8}, \cite{Kon} and in this review article.
1950: The comparison of this theory with the investigations of last years permits to
1951: physisists, working with above-mentioned problems, to do conclusions about
1952: situation in given field.
1953:
1954: {\bf Acknowledgments.}
1955:
1956: The author gratefully acknowledges Prof. to V.A. Moskalenko, for his interest
1957: in this problem and is thankful to Dr. F.G.Kochorbe for long-term collaboration
1958: and to Prof. T.Mishonov for his active position in the recognition of the
1959: priority accomplishment of Moldavian physicists to the development of the
1960: theory of multi-band superconductors.
1961:
1962: \begin{thebibliography}{98}
1963:
1964: \bibitem{Moskalenko_1} V. A. Moskalenko, P. Entel, and D. F. Digor,
1965: {\it Phys. Rev.B}, {\bf 59}, 619 (1999).%1
1966:
1967: \bibitem{Moskalenko_2} V. A. Moskalenko, {\it Fiz. Met.Metalloved};
1968: {\bf 8}, 503 (1959); {\it Phys. Met. and Metallog.} {\bf 8}, 25 (1959) %2
1969:
1970: \bibitem{Suhl} H. Suhl, B. T. Matthias, and L. R. Walker,
1971: {\it Phys. Rev. Lett.} {\bf 3}, 552 (1959). %3
1972:
1973: \bibitem{Krakauer} H. Krakauer and E. Pickett, {\it ibid}, {\bf 60}, 1665
1974: (1988). %4
1975:
1976: \bibitem{Herman} J. F. Herman, R. V. Kasowski, and W. G. Hsu, {\it
1977: Phys. Rev. B.}, {\bf 36}, 6904 (1987).%5
1978:
1979: \bibitem{Baranov} M. A. Baranov and Yu. M. Kagan, {\it Zh. Eksp. Teor. Fiz.},
1980: {\bf 102}, No. 1, 313 (1992) [{\it Sov. Phys. JETP} {\bf 75}, 165 (1992)]. %6
1981:
1982: \bibitem{Palistrant_1} M. E. Palistrant, arXiv :{\it cond - mat / 0305496},
1983: 26 june 2003.%7
1984:
1985: \bibitem{Lee_1} D. H. Lee and J. Ihm, {\it Sol. State Commun.}, {\bf 62}, 81
1986: (1987).%8
1987:
1988: \bibitem{Moskalenko_3} V. A. Moskalenko, M. E. Palistrant, and V. M. Vakalyuk,
1989: {\it Mechanisms of High - Temperature Superconductivity} [in Russian],
1990: {\it Joint Inst. for Nucl. Research}, Dubna, (1988), p. 34.%9
1991:
1992: \bibitem{Moskalenko_4} V. A. Moskalenko, M. E. Palistrant, and V. M. Vakalyuk,
1993: {\it Fiz. Nizk. Temp.}, {\bf 15}, 378 (1989) [{\it Sov. J. Low - Temp. Phys}].
1994: {\bf 15} (1989).%10
1995:
1996: \bibitem{Galaiko} V. P. Galaiko, E. V. Bezuglyi, and V. S. Shumeiko {\it
1997: ibid.} {\bf 13},1301 (1987) %11
1998:
1999: \bibitem{Konsin} P. Konsin, N. Kristoffel, and T. Ord, {\it Phys. Lett. Ser.A}
2000: {\bf 129}, 399 (1988)%12
2001:
2002: \bibitem{Volovik} E. G. Volovik, {\it Pism'ma v Zh. Eksp. Teor. Fiz.}, {\bf
2003: 49}, 185 (1989) [{\it JETP Lett.} {\bf 49}, 214 (1989)].%13
2004:
2005: \bibitem{Moskalenko_5} V. A. Moskalenko, M. E. Palistrant, V. M. Vakalyuk,
2006: and I. V. Padure, {\it Solid. St. Commun.}, {\bf 69}, 747 (1989).%14
2007:
2008: \bibitem{Moskalenko_6} V. A. Moskalenko, M. E. Palistrant, and V. M. Vakalyuk,
2009: {\it 10th Internat. Symp. on the Jahn-Teller Effect},Kishinev (1989), p.88.%15
2010:
2011: \bibitem{Moskalenko_7} V. A. Moskalenko, M. E. Palistrant, and V. M. Vakalyuk,
2012: {\it Procced. Eigh International Conference on Ternary and Multinary
2013: Compounds, Kishinev.}, p. 57, (1990).%16
2014:
2015:
2016: \bibitem{Palistrant_2} M. E. Palistrant and F. G. Kochorbe {\it Physica C},
2017: {\bf 194}, 351 (1992).%17
2018:
2019: \bibitem{Palistrant_3} M. E. Palistrant and Vakalyik, {\it Sverhprovodimost':
2020: Fiz. Khim. Tekh.}, {\bf 3}, 557 (1990).%18
2021:
2022: \bibitem{Palistrant_4} M. E. Palistrant and F. G. Kochorbe {\it Quantum-Field
2023: Methods of Investigating High} - {\it Temperature Supercanductors and Disordered
2024: Systems [in Russian] Shtiintsa}, Kishinev (1992).%19
2025:
2026:
2027: \bibitem{Kalalb_1} M. G. Kalalb, F. G. Kochorbe, and M. E. Palistrant {\it
2028: Teor. Mat. Fiz.}, {\bf 91}, 483 (1992).%20
2029:
2030: \bibitem{Palistrant_5} M. E. Palistrant and M. G. Kalalb, {\it Izvest. Akad.
2031: Nauk, Resp. Moldova}, {\bf 1(7)},70 (1992).%21
2032:
2033: \bibitem{Moskalenko_8} V. A. Moskalenko, M. E. Palistrant, and V. M. Vakalyuk,
2034: {\it Usp. Fiz. Nauk}, {\bf 161}, 155 (1991) {\it [Sov. Phys. Usp.}{\bf 34},
2035: 717 (1991), arXiv:{\it cond-mat / 03099671}%22
2036:
2037: \bibitem{Moskalenko_9} V. A. Moskalenko, L. Z. Kon, and M. E. Palistrant,
2038: {\it Low-Temperature Properties of Metals with Band-Spectrum Singularities}
2039: [in Russian], Shtiintsa, Kishinev (1989).%23
2040:
2041: \bibitem{Moskalenko_10} V. A. Moskalenko and M. E. Palistrant,
2042: {\it Statistical Physics and Quantum Field Theory} [in Russian],Nauka, Moscow
2043: (1973), p. 262.%24
2044:
2045: \bibitem{Moskalenko_11} V. A. Moskalenko, {\it Electromagnetic and Kinetic
2046: Properties of Superconducting Alloys with Overlapping Energy Bands} [in
2047: Russian],Shtiintsa, Kishinev (1976).%25
2048:
2049: \bibitem{Kresin_1} V. Kresin and S.A. Wolf, {\it Phys. Rev B}, {\bf 41}, 4278
2050: (1990); {\it Physica C} {\bf 169}, 476 (1990)%26
2051:
2052: \bibitem{Hirch} J. E. Hirch and F. Marsiglio {\it Phys. Rev B}, {\bf 43}, 424
2053: (1991).%27
2054:
2055: \bibitem{Nagamatsu} J. Nagamatsu, N. Nakagawa, T. Muranaka, Y. Zenitani and J.
2056: Akimutsu, {\it Nature (London)}, {\bf 410}, 63 (2001).%28
2057:
2058: \bibitem{Bud'ko} S. L. Bud'ko, G. Lapertot, C. Petrovich at al. , {\it
2059: Phys. Rev. Lett.}, {\bf 86},1877 (2001).%29
2060:
2061: \bibitem{Jiu_1} A. Y. Liu, I. I. Mazin, and I. Kartus,{\it Phys. Rev. Lett.}
2062: {\bf 87}, 087005 (2001).%30
2063:
2064: \bibitem{Kong} Y. Kong, O. V. Dolgov, O. Jepsen and O. K. Andersen, {\it
2065: Phys. Rev. B}, {\bf 64}, 020501-1 (2001).%31
2066:
2067: \bibitem{Choi_3} H. J. Choi, D. Roundy, H. Sun et al., {\it Nature},{\bf 418},
2068: 758 (2002).%32
2069:
2070: \bibitem{Buzea} C. Buzea and T. Yamashita,{\it Supercond. Sci. Technol.} {\bf
2071: 14} R 115 (2001).%33
2072:
2073: \bibitem{Ponomarev} Y. U. Ponomarev, S. A. Kuzmichev at al, (unpublished).%34
2074:
2075: \bibitem{Mishonov_1} T. Mishonov and E. Penev, arxiv: {\it cond-mat / 0206118 V
2076: 2} 24 june (2002); {\it International Journal of Modern Physics B}.%35
2077:
2078: \bibitem{Mishonov_2} T. Mishonov, S. L. Drechsler and E. Penev,{\it arxiv:
2079: cond-mat / 0209192 V 1} 8 sep. (2002).%36
2080:
2081: \bibitem{Mishonov_3} T. Mishonov, E. Penev, J. O. Indekeu and V. I.
2082: Pokrovsky arxiv:{\it cond-mat / 0209342 V 2} 17 Mar. (2003);{\it Phys. Rev. B}
2083: {\bf 68}, 104517 (2003).%37
2084:
2085:
2086: \bibitem{Moskalenko_12} V. A. Moskalenko, and M. E. Palistrant,
2087: {\it Dokl. Akad. Nauk SSSR} {\bf 162}, 539 (1965) [Sov. Phys. Dokl].%38
2088:
2089: \bibitem{Moskalenko_13} V. A. Moskalenko and M. E. Palistrant,
2090: {\it Zh. Eksp. Teor. Fiz.} {\bf 49}, 770 (1965) {\it [Sov. Phys. JETP}{\bf
2091: 22}, 536 (1965)].%39
2092:
2093: \bibitem{Van} V. Van der Marel, {\it Physica C}, {\bf 165}, 35 (1990).%40
2094:
2095: \bibitem{Schaffroth} M. R. Schafroth , {\it Phys.Rev.}, {\bf 111}, 72
2096: (1958).%41
2097:
2098: \bibitem{Migake} K. Migake, {\it Prog. Teor. Phys.},{\bf 69}, 1784 (1983).%42
2099:
2100: \bibitem{Randeria} M. Randeria, J. M. Duan, and L. I Shich, {\it Phys. Rev.
2101: Lett.}, {\bf 62}, 981 (1989); {\it Phys. Rev. B.}, {\bf 41}, 372 (1990).%43
2102:
2103: \bibitem{Michas} R. Michas, J. Rannienger, and S. Robaszkevicz, {\it Rev.
2104: Mod. Phys.}, {\bf 62}, 113 (1990).%44
2105:
2106: \bibitem{Leggett_2} A. J. Leggett,{\it Modern Trends in the Theory of Condensed
2107: Matter}, {\bf 115}, 13 (1980).%45
2108:
2109: \bibitem{Gorbatsevich_1} A. A. Gorbatsevich and I. V. Tokatly, {\it JETP},
2110: {\bf 76}, 347 (1993).%46
2111:
2112: \bibitem{Gordar} E. V. Gorbar, V. P. Gusynin, and V. M. Loktev, {\it
2113: Sverhprovodimost': Fiz. Khim. Tekh.,}, {\bf 6}, 483 (1993); {\it Fiz. Nizk.
2114: Temp.,} {\bf 19}, 1171 (1993).%47
2115:
2116: \bibitem{Loktev} V.M. Loktev, R. M. Quick, S. G. Sharapov, {\it Physics
2117: Reports},{\bf 349}, 1 - 123 (2001).%48
2118:
2119: \bibitem{Kochorbe_1} F. G. Kochorbe, M. E. Palistrant, {\it Zh. Eksp. Teor.
2120: Fiz.}, {\bf 104}, 3084 (1993);{\it JETP}, {\bf 77} 442 (1993); {\it [ Teor.
2121: Mat. Fiz.} {\bf 96}, 459 (1993); {\it Theoret. Mathemat. Phys.}{\bf 96},1083
2122: (1993);
2123: {\it [ 4th Internat. Conference M2S - HTS IV Grenoble (France), FR-PS}, 494
2124: (1994).%49
2125:
2126: \bibitem{Abricosov_1} A. A. Abrikosov, L. P. Gor'kov and I. E. Dzyaloshinskii,
2127: {\it Quantum Field Theoretical Methods in Statistical Physics};{\it [Prentice
2128: - Hall, Englewood Cliffs}, N 7; (1963); {\it Moscow, Nauka} (1962)].%50
2129:
2130: \bibitem{Robashkiewicz} S. Robashkiewicz, R. Mikhas, and G. A. Chao, {\it
2131: Phys.Rev. B}, {\bf 26}, 3915 (1982).%51
2132:
2133: \bibitem{Cardona} M. Cardona and L. Ley, eds., {\it Photoemission in Solids,
2134: Springer} (1978).%52
2135:
2136: \bibitem{Hongshum},H. Hongshum, X. Zhan, et al, {\it Physica C},
2137: {\bf 172},71 (1990)%53
2138:
2139: \bibitem{Junod} A. Junod, D. Eckert, at al.,{\it ibid} {\bf 159}, 215
2140: (1989).%54
2141:
2142: \bibitem{Loram} J. W. Loram and K. A. Mirza, {\it ibid},{\bf 153 - 155},
2143: 1020 (1988).%55
2144:
2145: \bibitem{Akis_1} R. Akis, F. Morsiglio, and J. P. Carbotte, {\it Phys. Rev. B},
2146: {\bf 39},2722 (1989)%56
2147:
2148: \bibitem{Akis_2} R. Akis and J. P. Carbotte, {\it Physica C},
2149: {\bf 159}, 395 (1989)%57
2150:
2151: \bibitem{Mishonov}T. Mishonov, S. Drechsler and E. Penev, {\it Modern Physics
2152: letter B}, {\bf 17}, 755 (2003)..%58
2153:
2154: \bibitem{Rodriguez} J. J. Rodriguez-Nunez and A. A. Schmidt, {\it Phys. Rev.B
2155: B}, {\bf 68}, 224512 (2003).%59
2156:
2157: \bibitem{Lagos} R. E. Lagos and G. G. Cabrera, {\it Brazilian Juornal of
2158: Physics}, {\bf 33}, 713 (2003).%60
2159:
2160: \bibitem{Jemma}S. Jemma Balaselvi; A. B. Bharthi at al, {\it arXiv: cond-mat
2161: / 0209200 \,; 0303022}.%61
2162:
2163: \bibitem{Kazakov} S. M. Kazakov, J. Karpinski at al, {\it arXiv: cond-mat
2164: / 0304686 \,; 0303022}.%62
2165:
2166: \bibitem{Palistrant_5a} M. E. Palistrant, {\it Low Temperature Physics} {\bf
2167: 26}, 799 (2000).%63
2168:
2169: \bibitem{Palistrant_6} M. E. Palistrant, {\it Teoret. Matemat. Fiz.},
2170: {\bf 95},101 (1993).%64
2171:
2172: \bibitem{Palistrant_7} M. E. Palistrant, {\it Teoret. Matemat. Fiz.},
2173: {\bf 105}, 491 (1995) ; {\it Teoret. and Mathem. Phys.} {\bf 105}, 1593
2174: (1995).%65
2175:
2176: \bibitem{Palistrant_8} M. E. Palistrant, V. M. Vackalyuk, M. G. Calalb, {\it
2177: Physika C}, {\bf 208}, 170 (1993).%66
2178:
2179: \bibitem{Palistrant_9} M. E. Palistrant, {\it Teoret. Matemat. Fiz.},
2180: {\bf 109}, 137 (1996) ; {\it Teoret. and Mathem. Phys.} {\bf 109}, 1352
2181: (1996).%67
2182:
2183: \bibitem{Palistrant_10} M. E. Palistrant, {\it J. of Superconductivity},
2184: {\bf 10}, 19 (1997).%68
2185:
2186: \bibitem{Markiewicz} R. S. Markiewicz, {\it Int. J. Mod. Phys.},
2187: {\bf 5}, 2037 (1991).%69
2188:
2189: \bibitem{Palistrant_11} M. E. Palistrant and F. G. Kochorbe, {\it Izv. Akad.
2190: Nauk Resp. Moldova}, 2(5), {\bf 7} (1991).%70
2191:
2192: \bibitem{Gofron} K. Gofron , J. C. Campuizano, and H. Ding at al.,{\it Phys.
2193: Chem. Solids} {\bf 54} 1193 (1993).%71
2194:
2195: \bibitem{Abricosov_2} A. A. Abrikosov, J. C. Compuizano and H. Gofron,
2196: {\it Physica C}, {\bf 214}, 73 (1993).%72
2197:
2198: \bibitem{Radtke} R. J. Radtke and M. R. Norman, {\it Phys. Rev. B} {\bf 50},
2199: 9594 (1994).%73
2200:
2201: \bibitem{Palistrant_12} M. E. Palistrant and Trifan, {\it Fiz. Nizk. Temp.,}
2202: {\bf 3}, 241 (1977); {\bf 3}, 976 (1977).%74
2203:
2204: \bibitem{Palistrant_13} M. E. Palistrant and O. P. Bezzub, {\it Fiz. Nizk.
2205: Temp.,} {\bf 6}, 1146 (1980); {\bf 9}, 357 (1983).%75
2206:
2207: \bibitem{Palistrant_14} M. E. Palistrant, {\it Teor. Matemat. Fiz.}{\bf 111},
2208: 289 (1997), {\it Theoretical Mathem. Phys.} {\bf 111}, 621 (1997).%76
2209:
2210: \bibitem{Palistrant_15} M. E. Palistrant and F. G. Kochorbe, {\it Fiz. Nizk.
2211: Temp.}{\bf 26}, 1077 (2000); {\it Low Temp. Phys.} {\bf 26}, 799 (2000).%77
2212:
2213: \bibitem{Todor} T. M. Mishonov, V. L. Pokrovsky and H. Wei, arXiv: {\it cond.
2214: mat. / 0312210}, Dec. 2003.%78
2215:
2216: \bibitem{Rare} Rare Earth Transition Metal Borocarbides (Nitrides):
2217: Superconducting, Magnetic and Normal State Properties (edited by Karl-Hartmut
2218: Muller and Vladimir Narozhnyi) Nato Sciences Series II. Mathematics, Physics
2219: and Chemistry - Vol. 14.%79
2220:
2221: \bibitem{Kon} L. Z. Kon, arXiv: {\it cond. mat. / 0309707}%80
2222: \end{thebibliography}
2223: \end{document}
2224:
2225: