REAL-TIME MONITORING OF RHEED USING MACHINE VISION TECHNIQUES

A RHEED system has been developed that allows real-time monitoring of RHEED information throughout a multilayer growth run with rotation. Th e machine vision system consists of high-speed image capture hardware coupled with digital signal processing software that allows the real-t ime extraction/analysis of the RHEED intensity and width signals from the noise induced by substrate rotation. This system has been used to investigate the oxide desorption process on GaAs substrates, along wit h the specular spot intensity and width variation during the growth of a set of InGaAs/AlGaAs single quantum well structures with systematic ally varied process parameters. A strong correlation of the specular s pot intensity with growth parameters has been observed. It is also sho wn that the observed specular spot intensity can be used to predict th e quality of the InGaAs quantum well structures.