One of the principal challenges for solar-driven hydrogen production via water splitting is to improve the solar-to-hydrogen conversion efficiency. We have employed combinatorial chemistry using a materials inkjet printer, and selected CuBi 2 O 4 as a promising p-type material. The steady-state photocurrent corresponding to water reduction for a 280 nm film at 0.2 V (RHE) was about 0.12 mA cm −2 , significantly lower than that attainable for a 2 eV band gap semiconductor. We have applied intensity-modulated photocurrent spectroscopy (IMPS) to distinguish between the photoelectrochemical processes involved and to determine the associated time constants, in order to gain insight into the loss processes responsible for the low efficiency. The charge separation efficiency reaches up to 0.66 at sufficiently negative potential, however, the recombination rate constant is larger than that corresponding to electron transfer to the solution. This results in a relative charge transfer efficiency of 0.2−0.4, explaining the low photocurrent. At low light intensity, the relative charge transfer efficiency increases up to 0.8, indicating the promise of the material. Interestingly, at sufficiently positive applied potential, the IMPS spectrum of the CuBi 2 O 4 photoelectrode switches sign, indicating a net modulated positive photocurrent. However, the rate constant for hole transfer to the solution is small resulting in a negligible steady-state anodic photocurrent. Strategies to improve the efficiency are discussed.

No abstract

No abstract