Theoretical study of large-scale graphene on Cu(111) surface using machine learning potential†

Abstract

The properties of graphene supported on a metal surface are highly dependent on its size. However, for graphene with dimensions of several nanometers or larger, high-precision studies remain limited. This study investigates the interaction of large-scale graphene with the Cu(111) surface, focusing on the average adsorption and average formation energies of graphene configurations with varying carbon atom numbers. The force field parameters for the graphene-Cu(111) system were derived using the High-Dimensional Neural Network Potential, and its applicability and accuracy were validated. The study evaluates the average adsorption and average formation energies for graphene nanosheet, as well as zigzag and armchair graphene nanoribbon configurations and establishes relationships between these energies and the number of carbon atoms. The findings provide valuable insights into the structural evolution, stability, and adsorption behavior of graphene on Cu(111) from nanoscale to mesoscale.