Changes to the occupied and unoccupied electronic structure of TiO₂(100) upon alkali metal reduction

Abstract

Direct and inverse ultraviolet photoemission spectroscopy (UPS and IPS) are used to measure the occupied and unoccupied electronic structure of TiO2(100)1×1 and TiO2(100)c(2×2)K. UPS results from stoichiometric TiO2(100)1×1 are in good agreement with earlier studies. IPS data recorded from TiO2(100)1×1 show two distinct features with energy positions 1.7 eV and 5.3 eV above the Fermi level (EF). These features are associated with the Ti 3d derived t2g and eg levels. Upon formation of TiO2(100)c(2×2)K, achieved by annealing K treated TiO2(100)1×1, several changes are observed to occur to the surface electronic structure. Population of a band gap state pins EF and removes band bending at the surface. The valence band of TiO2(100)c(2×2)K is noted to be substantially narrower (of the order of 0.5 eV) than that of TiO2(100)1×1. In addition to this, a sharp feature lying to the high binding energy side of the valence band is populated. Concurrently, intensity in the Ti t2g derived conduction band level is lost. This redistribution of spectral intensity observed between TiO2(100)1×1 and TiO2(100)c(2×2)K is interpreted in terms of a K induced redistribution of charge at the surface. We use resonant photoemission to test the idea that this charge redistribution involves mainly the Ti d levels.