Superconductivity and the High-T _c Materials

Abstract

Superconductivity, one of the most fascinating phenomena in solid-state physics, was discovered in 1911 by Kamerlingh Ormes [14.1], but it was not until 1957 that a satisfactory microscopic theory of the effect became available, i.e., the one by Bardeen, Cooper and Schrieffer (BCS) [14.2]. The major obstacle theorists were faced with earlier can be summarized as follows. The superconducting transition temperature T _c is usually of the order of 10 K, which corresponds to an energy of order 1 meV (we are not considering for the moment the more recent high-temperature superconductors). Provided that superconductivity is based on electron correlations and taking into account that the correlation energy of electrons in a metal is of the order of 1 eV per electron, is it then necessary to compute that energy to an accuracy of order l%o to find a superconducting ground state? This would indeed be an impossible task and eliminate any hope for a microscopic theory. Fortunately, a very special correlation leads to the phenomenon of superconductivity and the treatment of the remaining correlation contributions is unnecessary. All the correlations that are difficult to treat — and thus have been left out — enter the theory only in the form of renormalization parameters. An example are the heavy-fermion systems, in which the characteristic strong correlations manifest themselves in the form of quasiparticle energies; the (pair) correlations responsible for superconductivity are added and treated separately. This explains why reliable calculations of the superconducting transition temperature have so far remained an unsolved problem.