Nitride Semiconductors: Why they Work in Optoelectronic Devices

Abstract

GaN and its alloys with AlN and InN, grown heteroepitaxially on sapphire and other substrates, have a high concentration of point and line defects. In this paper I am addressing the fundamental question as to why such semiconductors produce very efficient minority carrier optoelectronic devices (LEDs, lasers etc). One fundamental difference between the traditional semiconductors (silicon and III-V compounds) and nitride semiconductors is that the chemical bonds in the former are mostly covalent while are strongly ionic in the later. This results in the bunching of the surface states in GaN and its alloys with AlN and InN near the band edges. This implies that that the surface states in free surfaces and the dangling bonds in edge dislocations are not expected to be non-radiative recombination centers but traps. Besides this unique property the InGaN and AlGaN alloys can be made to have strong band structure potential fluctuations, leading to exciton localization and thus efficient radiative recombination.