DOPING STRATEGIES TO IMPROVE CHARGE SEPARATION AND COLLECTION IN HEMATITE PHOTOANODES FOR SOLAR HYDROGEN PRODUCTION

Semiconducting metal-oxides, such as TiO₂, Fe₂O₃, WO₃ and SrTiO_3, are leading candidates for hydrogen production through solar water splitting. One of the most promising materials for the oxygen evolution reaction (OER) is iron oxide (α-Fe₂O₃, hematite), which is stable in alkaline solutions, has a favourable bandgap energy of 2.1 eV (λ ≈ 600 nm), is a good catalyst for water oxidation, and is cheap and abundant. However, hematite also possesses two major drawbacks: the mobility of the charge carriers in α-Fe₂O₃ is low (~0.1cm²V^-1s^-1) and the lifetime of minority charge carriers is very short (~100 ps). Thus leads to short (~ 2-5 nm) diffusion length of photogenerated minority carriers (holes) and high bulk recombination. Consequently, hematite photoanodes photocurrent and efficiency are significantly lower than their theoretical limit.

Possible approach to reduce bulk recombination is by tailoring built-in electrical fields, besides the one at the photoanode/electrolyte interface. In this work, we explore this approach by several doping schemes aiming to supress the bulk recombination by generating internal electric fields. Specifically, we investigate p-n and p-i-n structures similar to conventional photovoltaic devices rather than homogenously doped layers as in conventional α-Fe₂O₃ photoanodes.

Ultrathin films of 1% Zn-doped, undoped, and 1% Ti-doped α-Fe₂O₃ were deposited on FTO-coated glass substrates by pulsed laser deposition (PLD). The use of thin films reduces recombination in the photoanode and provides a high degree of structural and compositional control. We observe enhancement in the plateau photocurrent by almost 20% and cathodic shift of the onset potential by 150 mV in p-n and p-i-n structures compared to their homogenously doped counterparts. These results are encouraging for the development of highly efficient α-Fe₂O₃ photoanodes for water splitting.