Directed design of new photocatalysts remains a challenging task in materials chemistry. One approach in metal oxides is to engineer the bulk electronic structure to achieve enhanced visible-light absorption and favorable hole transport through the incorporation of heavy main group cations with lone pair electron configurations (Sn2+, Pb2+, Bi3+). Here, we exemplify this strategy with layered SnTiO3, which displays an extremely wide valence band resulting in a favorable band gap (E-g = 1.90 eV), a hole effective mass which is among the lowest known for oxides (mh*/m0 = 0.37 in-plane, 0.60 out-of-plane) and visible-light H-2 evolution activity. While bulk probes suggest powders of SnTiO3 are stable in air, microscopy, photoemission and 2D nuclear magnetic resonance spectroscopies, and comparison of Mott-Schottky analysis with ab initio band positions indicate the formation of a thin, oxidized passivation layer at the surface which prevents efficient extraction of photoexcited holes and acts as a sink for photoexcited electrons. Extending these ab initio calculations to several Sn(II) oxide semiconductors leads to a remarkably wide range of valence band maxima (spanning similar to 4 eV) and suggests an inherent competition between a favorably dispersive valence band and stability against oxidation. In addition to the covalency-driven influence of the valence band width, electrostatics are shown to play a decisive role, with open, low-dimensional structures and cations of lower formal charge leading to smaller ionization potentials. These ab initio band alignments reconcile the experimental observations of (in)stability across several Sn(II) oxides and underscore the utility of this technique as an early, inexpensive screening tool for assessing the suitability of new materials for (opto)electronic and catalytic applications.
