ABSTRACT
InGaNP, lattice-matched to Si or GaP, is predicted to possess direct bandgaps. The problem of low solubility limit of N in InP and GaP is overcome by low-temperature gas-source molecular beam epitaxial growth. We have achieved a N composition of about 1% in InP and 16% in GaP, the highest values ever reported. A large bandgap bowing is observed but suggests no semimetallic region. Quaternary InGaNP layers lattice-matched to GaP have been achieved, but so far they have an indirect bandgap.
