ON THE VALIDITY OF THE BAND MODEL FOR HIGH Tc SUPERCONDUCTORS

Abstract

The discovery of high T/sub c/ superconductivity in cupric oxides has led, amongst other things, to numerous electronic bandstructure studies of these compounds. The foundation of these calculations is Density Functional Theory (DFT) and they are implemented with the Local Density Approximation (LDA) or Local Spin Density (LSD) approximation. In this article we will review what we have learned from these calculations regarding the electronic structure of the family of high T/sub c/ materials. To answer this we have to find out how relevant the LSD/LDA is for the electronic properties of cupric oxides. It is well known that the application of LDA DFT for transition metal oxides is fraught with difficulty and controversy. Although the LDA describes the variation of the equilibrium volume through the 3d transition metal oxides, including the volume expansion associated with Mott insulators, the theory fails to describe FeO and CoO as antiferromagnetic insulators, and in the case of NiO and MnO where the LDA does yield an insulating ground state, the band gap is grossly underestimated. Evidently the high-T/sub c/ materials are systems that are on the verge of undergoing a Mott transition; consequently, it is necessary to be circumspect regarding the applicability of LDA DFT both in the metallic (superconducting) and magnetic insulating states. 23 refs., 12 figs.