Phase Transitions in Correlated Electron Systems. Chemical Potential Evidence

Abstract

We consider a model describing two kinds of correlated electrons characterized by the average occupation numbers n(1) = Σσ 〈n(1)a〉 and n(2) = Σσ 〈n(2)a〉 using the constraint n(1) + n(2) where n is the average number of electrons per magnetic ion. Depending on the model parameters, the model can describe second order phase transitions from ferromagnetic and normal phase to paramagnetic and normal (critical temperature TC), from paramagnetic and superconducting phase to paramagnetic and normal (critical temperature TS, as well as combined, reentrant, phase transitions with three critical temperatures TS1, TC and TS2 (TS1> TC > TS2). We consider two cases: (1) n(1) ≠ 0, n(2) ≠ 0 (two sorts of interacting electrons) and (2) n(1) ≠ 0, n(2) ≠ 0 (one sort of interacting electrons resulting from the shift of the center of gravity for the second band to higher energies). In both cases ((1) and (2)) the chemical potential μ exhibits small but distinct kinks at all critical temperatures of the system. This effect suggests a new possibility to detect all critical temperatures of a real solid (structural phase transitions included) exclusively from the experimental measurement of the chemical potential as function of temperature. In the case (1) the critical temperatures of the system are also visible from the kinks in the temperature dependence of the average occupation numbers n(1) and n(2) (the effect of critical electron redistribution at critical temperatures). Considering the phase transition from paramagnetic superconductor to normal system within our model, we can reproduce the same temperature dependence of the chemical potential as measured for the high-temperature superconductor YBa2Cu3O7—δ. This result entirely supports the statement about the applicability of the chemical potential as a detector of phase transitions in real solids.