In significant contrast to the tremendous research efforts mostly geared to addressing the severe hole accumulation at the back contact of a p-type Cu2O photocathode with a fluorine-doped tin oxide (FTO) substrate, sluggish electron transfer from an n-type Cu2O photoanode to a tin-doped indium oxide (ITO) substrate has been largely overlooked. To tackle this issue that has been reported to largely limit the photoelectrochemical performance of n-type Cu2O photoanodes at a low bias, the present contribution puts forward a strategy to introduce oxygen vacancies into the ITO substrate via an unprecedented yet facile electrochemical approach. Such defect engineering turns out to decrease the work function of the ITO substrate, which in turn approaches the conduction band extremum of n-Cu2O to highly efficiently extract the photoexcited electrons therein. Moreover, the dendritic growth of n-Cu2O is, in the meantime, interfered by the oxygen vacancy manifested as pinholes distributed over the ITO substrate, which is thereby crystallized into several small grains with augmented surface roughness that is in favor of the injection of the photoexcited hole into the electrolyte. Such facile interfacial charge-transfer kinetics leads to a significant cathodic shift amounting to 200 mV of the onset potential to 0 VAg/AgCl, whereat the n-Cu2O photoanode deposited on the defective ITO substrate delivers the maximum photocurrent density reaching 2 mA cm-2 and, more significantly, its applied bias photon-to-current efficiency (ABPE) reaches 1.1%, which is among the highest performance reported to date for a variety of state-of-the-art metal oxide-based photoanodes in the literature.
Keywords: Ohmic back contact; charge-transfer kinetics; defective ITO; n-type Cu2O; onset potential shift.