Machine learning method for determining chemical vapor deposition conditions for large-area graphene growth

The synthesis of wafer-scale single-crystal graphene sheets has become essential to realize future electronic device applications. To synthesize large-area boundary-free graphene, it is effective to use chemical vapor deposition (CVD) on copper (Cu) surfaces that possess a thin oxide layer. In this study, we constructed machine learning (ML) modeling to design experimental CVD conditions for the formation of large-area graphene. The constructed ML model predicted the graphene domain size from the experimental CVD growth conditions and the spectral information of the Cu surface. Furthermore, we demonstrated the formation of large-area graphene domain on the Cu surface using the CVD conditions determined by the constructed ML model, which provided faster graphene growth compared to previously reported strategies.