Band gap anomalies of the ZnM2IIIO4 (MIII = Co, Rh, Ir) spinels

Development of high figure-of-merit p-type transparent conducting oxides has become a global research goal. ZnM2IIIO4 (MIII = Co, Rh, Ir) spinels have been identified as potential p-type materials, with ZnIr2O4 reported to be a transparent conducting oxide. In this article the geometry and electronic structure of ZnM2IIIO4 are studied using the Perdew-Purke-Ernzerhof generalized gradient approximation (PBE-GGA) to density functional theory and a hybrid density functional, HSE06. The valence band features of all the spinels indicate that they are not conducive to high p-type ability, as there is insufficient dispersion at the valence band maxima. The trend of increasing band-gap as the atomic number of the MIII cation increases, as postulated from ligand field theory, is not reproduced by either level of theory, and indeed is not seen experimentally in the literature. GGA underestimates the band-gaps of these materials, while HSE06 severely overestimates the band-gaps. The underestimation (overestimation) of the band-gaps by GGA (HSE06) and the reported transparency of ZnIr2O4 is discussed.