Understanding photocatalytic overall water splitting on CoO nanoparticles: Effects of facets, surface stoichiometry, and the CoO/water interface

Abstract

Unlike CoO micropowder, which is unable to split water, CoO nanoparticles have been observed to photocatalytically split water into H2 and O2 at room temperature without an externally applied potential or co-catalyst. The photocatalytic activity of CoO nanoparticles has been suggested to stem from an upwards shift in the band edges relative to bulk CoO such that the conduction and valence band edges span the water redox potentials. However, the origin of this shift in the band edges is unknown. In this study, we use first-principles density functional theory (DFT) calculations to explore the thermodynamically stable surface configurations of CoO as a function of oxygen chemical potential. We show that the band edge positions are sensitive to surface chemistry which is determined by surface orientation, adsorbates, and stoichiometry, and thus growth conditions and operating environment. In particular, we predict that CoO nanoparticles have fully hydroxylated CoO(1 1 1) facets, with band edges spanning the water redox potentials, while larger CoO particles (such as CoO micropowders) have a full monolayer of hydrogen on the CoO(1 1 1) facets, with a band alignment that favors water oxidation but not water reduction. Furthermore, we demonstrate that explicit inclusion of liquid water is crucial for accurately predicting the band edge positions, and thus photocatalytic behavior, of CoO in an aqueous solution. Our work explains why photocatalytic overall water splitting has only been observed on CoO nanoparticles, and provides new insights into the relationships between environmental conditions, surface structure, and band alignment, which may lead to new approaches for optimizing activity in CoO and other oxide photocatalysts.