In situ characterisation of epiready III–V substrates for MOVPE

Abstract The use of the term epiready in relation to III–V substrates in MOVPE is ill defined and poorly understood. In this paper we attempt to clarify some of the issues associated with the term epiready, in particular, the thickness and distribution of native oxide on the surface of a GaAs wafer. The surface quality of a wafer is established at three stages, fresh from the packet, oxide removed, and after growth. The surface is assessed by means of atomic force microscopy on a microscopic level and laser light scattering and oxide thickness mapping on a macroscopic scale. GaAs substrates from long-term storage are also examined. It is shown that even long-term stored wafers (in excess of six years) with quite thick native oxide layers can be successfully deoxidised to give atomically flat terraces and can subsequently be used for successful homoepitaxial growth provided that atomic hydrogen (in this case, from the arsenic precursor) is used in the deoxidation stage. No difference between various manufacturers substrates has been found in respect to storage and subsequent use, nor has any difference been established between doped and undoped wafers.