[FMC6-3L]Bayesian Optimization-Driven Improvement of Thermoelectric Properties of Polycrystalline III-V Semiconductor Films Synthesized at Low Temperature
*Takamitsu Ishiyama1,2, Koki Nozawa1, Takashi Suemasu1, Kaoru Toko1(1. University of Tsukuba (Japan), 2. JSPS Research Fellow (Japan))
Keywords:
Bayesian optimization,Thermoelectric materials,Low-temperature synthesis,Semiconductors,Thin-film growth
We utilized Bayesian optimization to improve the thermoelectric performance of Sn-doped InGaAsSb thin films. We obtained a figure of merit ZT = 0.033, which represent a 2.5-fold improvement compared to random initial experimental results. Additionally, the optimal process temperature used for synthesizing the film was suitable for deployment on a flexible substrate.