Rapid Performance Optimization Method for Photoelectrodes

We report the development of a rapid method for the performance optimization of photoelectrodes. The method requires as an input incident photon-to-current efficiency (IPCE) measurements of a material at two or more wavelengths, an estimation of the complex refractive index, the permittivity, the doping concentration, the flatband potential, and the photoelectrode thickness, and estimates in return the diffusion length, the optical loss, the bulk and surface recombination losses, and the space charge region loss. The diffusion optical number, defined as the product of the absorption coefficient and the diffusion length at 500 nm, was used to quantify the performance of photoelectrodes. The method was validated using planar Cu2O water-splitting photoelectrodes. Subsequently, it was applied to planar water-splitting photoelectrodes made of Cu2O, Si, Fe2O3, BiVO4, Cu2V8O3, and CuFeO2, and to nanostructured photoelectrodes made of Fe2O3 and LaTiO2N. The projected diffusion optical number of Fe2O3 was improved by one order of magnitude when nanostructuring compared to the diffusion optical number of flat Fe2O3 photoelectrodes. Thus, a modification of the synthesis or deposition method should be prioritized instead of developing nanostructuring techniques for any photoelectrode with a diffusion optical number two orders of magnitude below the one required to obtain an internal quantum efficiency of 95 %. Using this benchmark, the investigated Si photoelectrode performed well without the need of nanostructuring. Cu2O and LaTiO2N photoelectrodes were found to benefit from nanostructuring. In contrast, nanostructuring is not advised for Fe2O3, BiVO4, Cu2V8O3, and CuFeO2 photoelectrodes, rather their synthesis method should be modified. Approaches for performance improvements by nanostructuring, doping concentration optimization, surface passivation, or by changing the photoelectrode thickness were presented for all investigated photoelectrode materials and nanostructures. We predicted—consistent with previous reporting—that nanostructuring improves the projected diffusion length but also increases the surface recombination losses, which partially counteract the performance improvement, exemplified by the investigated nanostructured Fe2O3. The validated tool is well suited to evaluate if nanostructuring can bring a photoelectrode to high performance and/or if the deposition or synthesis methods should be optimized. This tool can also be used for developing new approaches to passivate surfaces and bulk defects, modify doping concentration and investigate their impact on the photoelectrode performance and, more specifically, on the surface and bulk losses. This tool is not restricted to water-splitting reaction but can be applied to any photoelectrochemical reaction.