Abstract
Non-empirical calculations of the electronic structure of NiO and related compounds have proven to be problematic, in that NiO is an insulator both above and below the antiferromagnetic ordering temperature (TN approximately 525K) with a gap of approximately 4 eV. This is so despite the partly filled d-band on the Ni ion. Conventional hand theories predict that NiO is a metal for T>525K and a narrow-gap semiconductor or a metal for T<525K. It is argued here that the failure of band theory is due to the use of an over restricted form of the Hartree Coulomb potential, and the use of Koopmans theorem for the narrow, partly occupied 3d band. In this letter the author considers an approach based upon systematic improvements to a valence bond description for the narrow 3d states. He predicts a wide gap insulator for all T and find other results which are in reasonable agreement with experiment.

This publication has 19 references indexed in Scilit: