Optimal methodology for explicit solvation prediction of band edges of transition metal oxide photocatalysts

Abstract

The conduction and valence band edges (EC and EV) of a material relative to the water redox potential levels are critical factors governing photocatalytic water splitting activity. Here we discuss the large discrepancy in the experimentally measured EC and EV of various transition metal oxides (TMOs) in vacuum and in an aqueous solution. We speculate that the discrepancy stems from the different degree of electron transfer across the surface due to the different environment at the surface of the TMOs in vacuum and water. Accurately modeling the electronic structure at TMO/water interfaces is a significant challenge, however. Using first-principles density functional theory calculations on rutile titanium dioxide and cobalt monoxide model systems, here we identify the optimal approaches to accurately predict the band edge positions in vacuum and water. We then validate the optimized schemes on other TMOs, demonstrating good agreement with experimental measurements in both vacuum and water.