The theory of electrode matching in photoelectrochemical cells for the production of hydrogen

Abstract

The problem of matching the properties of the semiconductors in a two photon photoelectrochemical cell is addressed. The effect of the semiconductor properties on the performance of the whole cell has been calculated for the first time. Calculations of the cell potential and photocurrent have been made for the first time. Computations have been made for specific combinations of electrodes, e.g., n-SrTiO 3/p-GaP and n-TiO 2/p-GaP. Taking different p-type semiconductors as photocathodes, calculations have been made to determine the properties of n-type semiconductors which would be expected to form efficient (>10%) light driven PECs for producing hydrogen. It was found that without electrocatalysts, no combination of oxide semiconductors will give efficiency >4.0%. However, when the effects of the deposition of submonolayer amounts of different electrocatalysts on both the semiconductor surfaces are taken into consideration, the predicted efficiencies of the conversion of light to chemical energy improve substantially. Successful n-type electrodes cannot be oxides and therefore need protective coatings. Calculations suggest that it should be possible to attain efficiency for the conversion of light up to 16% with p-InP (Pt)/n-GaAs(e) and up to 14% with p-Si (Pt)/n-InP (e), using appropriate electrocatalysts on the n-Si and on the n-InP electrodes, together with Mn-oxide protective coating.