2025/09/10 by Benedict Saunders, Lukas Hörmann, Saunders, Benedict +3
Materials Science · #2D Materials and Applications #FOS: Physical sciences #Graphene research and applications #Machine Learning in Materials Science #Materials Science (cond-mat.mtrl-sci) #Statistical Mechanics (cond-mat.stat-mech)
paper · pdf · doi:10.48550/arxiv.2509.08352
openalex publication_date 2025/09/10 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Graphene has been studied in detail due to its mechanical, electrical, and thermal properties. It is well documented that the introduction of dopants or defects in the lattice can be used to tune material properties for a specific application, such as in electronics, sensors, or catalysis. To design graphene with specific properties, one must achieve control over the composition and concentration of defects. This requires a fundamental understanding of the stability of defects and their interaction in a superstructure. We present a comprehensive defect structure determination approach that enables close to exhaustive enumeration of all relevant defect structures. The approach uses a combination of Density Functional Theory and machine learning to build a transferable energy model for defect formation. Henceforth, we show the capabilities of our approach for a proof-of-principle application on free-standing graphene with heteroatom defects. This allows us to provide physical insights into defect interactions and to establish a thermodynamic model to investigate how temperature affects the configuration space of doped graphene.