On the (110) orientation as the preferred orientation for the molecular beam epitaxial growth of GaAs on Ge, GaP on Si, and similar zincblende-on-diamond systems

Abstract
Recent successful (110) growth of GaAs on Ge has prompted a reevaluation of the (110) orientation for the MBE growth of zincblende‐on‐diamond–type heterostructures. It is argued that the atomic geometry at a (110) interface should be particularly favorable for defect‐free heteroepitaxy in such systems, for two reasons: (a) Recent work by Harrison et al. has shown that interfaces other than (110) interfaces in such systems must reconstruct. Such reconstruction will be incomplete, leaving behind hard‐to‐control interface charges. (b) The freesurface reconstruction of the diamond‐structure (110) surface is such that it should favor subsequent growth of zincblende structures without antiphase domain boundaries.