Electronic structure of 3<i>d</i>-transition-metal compounds by analysis of the 2<i>p</i>core-level photoemission spectra

Abstract

The electronic structures of a wide range of transition-metal compounds, including Cu, Ni, Co, Fe, and Mn oxides and sulfides, with metal valences ranging from 2+ to 4+, have been investigated by a cluster-type configuration-interaction analysis of the core-level 2p x-ray photoemission spectra. We show that by including the d-d exchange interaction (retaining only diagonal terms) and an anisotropic metal-ligand hybridization in the model, these spectra can be well reproduced, and so can be used to deduce quantitatively values for the ligand-to-metal charge-transfer energy \ensuremath{\Delta}, the on-site d-d Coulomb repulsion energy U, and the metal-ligand transfer integrals T. Systematics for \ensuremath{\Delta} and U are generally consistent with those found from previous valence-band studies and follow expected chemical trends. By using values of \ensuremath{\Delta} and U found from this model, we show that most of the transition-metal compounds studied in this work can be classified in the charge-transfer regime of the Zaanen-Sawatzky-Allen diagram. A few exceptions to these systematics have been found. Small U values found for pyrite-type ${\mathrm{CoS}}_{2}$ and ${\mathrm{FeS}}_{2}$ and large T values for Mn perovskite oxides, as well as the neglect of other mechanisms such as exciton satellites, may indicate a limitation of the local-cluster model.