Prediction of optical band gap of β-(Al x Ga 1-x ) 2 O 3 using material informatics

Abstract In this study, the optical band gap of β-(Al x Ga 1-x ) 2 O 3 versus the Al composition x is predicted using principal component regression and a Gaussian stochastic process. Properties were sourced from other mature Al-alloyed compound semiconductors to form a band gap model. It is found that the electronic band gap, the thermal conductivity, and the Al composition have the greatest influences on the optical band gap. A final relation is generated from a hybrid informatics approach combining information gained from multiple models. The optical band gap of β-(Al x Ga 1-x ) 2 O 3 versus the Al composition is predicted and agrees well with measured optical band gap.