Frank Osterloh1,Sahar Daemi1,Anna Kundmann1,Kathleen Becker1
University of California, Davis1
Frank Osterloh1,Sahar Daemi1,Anna Kundmann1,Kathleen Becker1
University of California, Davis1
The efficiency of photocatalysts and photoelectrodes for the conversion of solar energy into fuels is closely related to the electrochemical potentials of the electrons and holes and the quasi-Fermi level splitting energy (photovoltage) under illumination. While electrochemical measurements can provide the electrochemical potential of the majority carriers at the semiconductor back-contact, the potential of the minority carriers at the front semiconductor/electrolyte is more difficult to determine experimentally, due to the lack of a direct electrical connection. Here we show that surface photovoltage spectroscopy on photoelectrode films in contact with electrolytes can reveal information about the quasi-Fermi level splitting energy in them and on the electrochemical potential of the minority carriers at the solid-liquid interface. For these measurements the semiconductor films (BiVO<sub>4</sub>) or wafers (n-GaP or p-GaP) are immersed in aqueous electrolyte and placed underneath a vibrating Kelvin probe to monitor the light induced contact potential difference change. For BiVO<sub>4</sub> films, we find that the photovoltage (quasi-Fermi level splitting) strongly depends on the electrolyte, the back contact, and the light intensity. For example, in water and aqueous Na<sub>2</sub>SO<sub>4</sub> electrolyte, the photovoltage shows a near ideal logarithmic dependence on the illumination intensity (62 mV dec<sup>-1</sup>) and reaches 1.03 V under 73 mW cm<sup>-2</sup>(400 nm) illumination. Using the experimental majority carrier potential (from open circuit measurements on a BiVO<sub>4 </sub>photoelectrode), the quasi- Fermi level hole potential is estimated at 1.1 – 1.2 V vs RHE, nearly independent of light intensity. This value is not sufficient for water oxidation and it agrees with the general observation of a 0.3-0.4 V vs RHE potential bias requirement for water oxidation with BiVO4 photoelectrodes. The implications of these findings and related observations on gallium phosphide wafers on the understanding of charge separation in solar fuel photoelectrodes will be discussed.