High throughput theoretical and experimental screening of photocatalysts for water splitting

We apply a high-throughput computational screening process with feedback from experimental results to identify promising photocatalysts for water splitting. Using an automated linear response framework for predicting Hubbard parameters, we overcome the limitations of local and semilocal density-functional theory calculations in predicting band gaps. From our computational screening, we identify 28 promising photocatalysts, many of which have not been previously reported. We then synthesized some along with chemical derivatives from our selection, for a total of 18 synthesized materials using solid-state synthesis techniques. We report hydrogen production in a number of the materials tested for photocatalysis using gas chromatography; many of which have been previously unreported for water splitting. We further measure cyclic voltammograms, demonstrating high levels of photoactivity in several samples.