A simple model is presented to describe resonant photoemission due to super-Coster-Kronig decay of ${3p}^5{3d}^{N+1\mathrm{}}$ states in transition-metal compounds. An explanation is given for the puzzling behavior observed in recent experiments on NiO and Ni${\mathrm{Cl}}_2$ in which the multielectron satellite is enhanced rather than the main valence band. Several experiments are proposed to test the theory.

• Received 14 September 1981

DOI:https://doi.org/10.1103/PhysRevB.25.2912