Process modeling of indium arsenide/aluminum antimonide materials for high electron mobility transistors grown by molecular beam epitaxy.

摘要

This research demonstrates process modeling of InAs/AlSb HEMT structures produced by MBE using neural network technology. The process models developed employ MBE process conditions to predict HEMT structure metrics. This research effort enhances the ability to: (1) better understand the MBE growth process for these particular, complex structures; and (2) achieve reproducible results. The performance of InAs/AlSb structures depend greatly on the ability to control the physical properties in thin films produced via MBE. Therefore, the influence of the process conditions on these properties should be realized. The approach used in this research involves using carefully designed experiments and in-situ diagnositc techniques to evaluate thin film quality. The experiments in this thesis demonstrate the range of transport properties attainable and the significance of the parameters studied.