X-ray Transitions in Broad Band Materials

Abstract

The general application of soft X-ray spectroscopy in the study of the electronic structure of materials is discussed, with particular emphasis on broad band materials. Several materials are studied using both soft X-ray spectroscopy and density functional theory to provide experimental and theoretical electronic structures, respectively. In particular, bonding, cation hybridization, and band gaps for several binary oxides (the alkali oxides: BeO, MgO, CaO, SrO, BaO; the post-transition metal oxides: ZnO, CdO, HgO; and the period 5 oxides In2O3, SnO, SnO2, Sb2O3, Sb2O5, and TeO2) are studied. The technique of using the peaks in the second derivatives of an X-ray emission and an X-ray absorption spectrum to estimate the band gap of a material is critically analyzed, and a more accurate ``semi-empirical'' method that involves both measured spectra and theoretical calculations is proposed.

The techniques used in the study of binary oxides are then applied to a more interesting (and industrially relevant) group of ternary oxides based on TiO2 (PbTiO3, Sn2TiO4, Bi2Ti4O11, Bi4Ti3O12, and ZnTiO3), and a general rule for the band gaps of these materials is suggested based on empirical data. This research may help direct efforts in synthesizing a hydrogen-producing photocatalyst with a band gap that can efficiently harness the bulk of the solar spectrum.

Finally, several layered pnictide superconductors and related compounds (CaFe2As2, Co-, Ni- and Cu-doped BaFe2As2, LiFeAs, LiMnAs, CaCu1.7As2, SrCu2As2, SrCu2(As0.84Sb0.16)2, SrCu2Sb2, and BaCu2Sb2) are studied. The X-ray spectra provide rather strong evidence that these materials lack strong on-site Hubbard-like correlations, and that their electronic structures are almost entirely like those of a broad band metal. In particular, it is shown that the notion that the transition metals are all divalent is completely wrong for copper in a layered pnictide, and that at best in these systems the copper is monovalent.