Use of genetic algorithms for the development and optimization of crystal growth processes

Abstract

This paper demonstrates that the combination of a genetic algorithm (GA) and a thermal modeling tool can be useful and efficient for the development and optimization of crystal growth processes. Results are shown for two typical semiconductor crystal growth models: VGF-GaAs and Si-Czochralski. The VGF case demonstrates the automatic development of an entire growth process, including optimization of the furnace itself and of the heater temperature settings under various constraints. In the Czochralski model, an optimization of the heat shield configuration has been conducted in order to demonstrate how the GA can be applied to geometry optimization tasks.