Data-driven analysis of the electronic-structure factors controlling the work functions of perovskite oxides

Tuning the work functions of materials is practical interest for maximizing performance microelectronic and (photo)electrochemical devices, as efficiency these systems depends on ability to control electronic levels at surfaces across interfaces. Perovskites are promising compounds achieve such control. In this work, we examine more than 1,000 perovskite oxide (ABO$_3$) by data-driven (machine-learning) analysis identify factors that determine their magnitude. While BO$_2$-terminated sensitive energy hybridized oxygen p bands, AO-terminated exhibit a much less trivial dependence with respect filling d bands B-site atom its affinity. This study shows utility interpretable models in analyzing cubic perovskites from limited number electronic-structure descriptors.