AI-enabled construction and prediction of atomic models for thin-film heterostructures via materials genome approach

This study leverages an Artificial Intelligence-enabled materials genome approach to predict and analyze atomic models for thin-film heterostructures. A novel computational strategy was employed to investigate the structural and electronic properties of wurtzite Gallium Nitride thin films grown on flexible muscovite mica substrates. Among the configurations studied, the N-polar GaN/K-terminated Muscovite model was the most stable, exhibiting the lowest interface energy of -0.281 eV/& Aring;2. This result aligns closely with experimental observations, wherein N-polar GaN formation is favored under nitrogen-rich growth conditions. Our findings demonstrate the efficacy of combining Artificial Intelligence with the materials genome approach to accurately predict interfacial energies and optimize heterostructure designs. This innovation accelerates the discovery of advanced materials and paves the way for breakthroughs in optoelectronics and related technologies by enabling precise and efficient structural analysis.