[9a-A24-1~12] 11.3 Analog applications and their related technologies

9:00 AM - 9:15 AM JST(12:00 AM - 12:15 AM UTC)

[9a-A24-1] Virtual Experiments and Optimization of NbN Thin-Film Growth Conditions Using a Machine Learning Model

〇Hirotake Yamamori^1,2, Daiki Matsumaru¹, Kaho Koyanagi¹, Michitaka Maruyama¹, Kazumasa Makise^1,2, Hirokazu Ishino³ (1.AIST, 2.NAOJ, 3.Okayama Univ.)

Keywords:

NbN,Bayesian optimization

Data-driven materials exploration using AI was applied to NbN thin films to optimize sputtering conditions. Machine learning and iterative virtual experiments identified pressure as a key parameter. By reducing Ar flow, conditions for forming the superconducting δ-phase were achieved. Further optimization of nitrogen flow and substrate temperature yielded (111)-oriented NbN films with a high critical temperature of Tc = 16.2 K.