Assessment of a W:BiVO₄–CuBi₂O₄Tandem Photoelectrochemical Cell for Overall Solar Water Splitting

We assess a tandem photoelectrochemical cell consisting of a W:BiVO₄ photoanode top absorber and a CuBi₂O₄ photocathode bottom absorber for overall solar water splitting. We show that the W:BiVO₄ photoanode oxidizes water and produces oxygen at potentials ≥0.7 V vs RHE when CoPi is added as a cocatalyst. However, the CuBi₂O₄ photocathode does not produce a detectable amount of hydrogen from water reduction even when Pt or RuO_x is added as a cocatalyst because the photocurrent primarily goes toward photocorrosion of CuBi₂O₄ rather than proton reduction. Protecting the CuBi₂O₄ photocathode with a CdS/TiO₂ heterojunction and adding RuO_x as a cocatalyst prevents photocorrosion and allows for photoelectrochemical production of hydrogen at potentials ≤0.3 V vs RHE. A tandem photoelectrochemical cell composed of a W:BiVO₄/CoPi photoanode and a CuBi₂O₄/CdS/TiO₂/RuO_x photocathode produces hydrogen which can be detected under illumination at an applied bias of ≥0.4 V. Since the valence band of BiVO₄ and conduction band of CuBi₂O₄ are adequately positioned to oxidize water and reduce protons, we hypothesize that the applied bias is required to overcome the relatively low photovoltages of the photoelectrodes, that is, the relatively low quasi-Fermi level splitting within BiVO₄ and CuBi₂O₄. This work is the first experimental demonstration of hydrogen production from a BiVO₄-CuBi₂O₄-based tandem cell and it provides important insights into the significance of photovoltage in tandem devices for overall water splitting, especially for cells containing CuBi₂O₄ photocathodes.