Structures and functions of thin metal layers on semiconductor electrodes

Abstract

Metal-coated semiconductor electrodes such as Au/n-TiO 2 and Au/n-GaP show photovoltaic effect that cannot be explained by the conventional potential barrier model for metal—semiconductor contact. From experimental and theoretical investigations it has been concluded that the microscopically discontinuous structure of the metal layer is responsible for the anomalous photovoltaic effect. Several theoretical conclusions which are interesting from the point of view of solar energy conversion are derived. (1) Metal-coated semiconductor electrodes in electrolyte solutions generate high photovoltages at the metal—semiconductor interface in cases where the metal layer forms islands approximately 5 nm in diameter and separated by more than 20 nm from each other provided that the potential of the electrode is controlled so as to give band bending in the metal-free part of the surface. The maximum photovoltage can increase up to the equivalent of the band gap of the semiconductor in ideal cases. (2) The metal—semiconductor contact becomes ohmic when the potential of the electrode approaches the flat-band potential for the bare electrode in cases where the metal layer either has cracks or forms islands with gaps wider than 20 nm, say. Changes in the barrier height at the metal—semiconductor interface are not assumed in the theory, which can be applied to any metal—semiconductor pair. The conclusions provide a thoeretical basis for the explanation of the mechanisms of interesting properties of metal-coated semiconductor photoelectrodes and photocatalysts, such as enhanced hydrogen photoevolution on platinum-coated p-type semiconductor electrodes or on platinum-coated semiconductor particles in solution.