lammps-graphene/abstract

2. Abstract

Molecular Dynamics of Bicrystalline Graphene to Determine Fracture Strength

Otho Ulrich1, Joseph Gonzalez2, Kien Cong Nguyen2, Ivan Oleynik2


In its pristine form, graphene is one of the strongest materials measured, and possesses a wide range of technologically appealing characteristics.  Several recent experiments have explored the mechanical properties of graphene which contains grains, which seem to show contradictory results. To explore the atomic structure of grain boundaries in graphene, we employ a complex of computational approaches. Unit cells of the crystal structure of graphene bicrystals with grain misorientations of differing angles are generated by applying a conjugate gradient method with periodic boundary conditions using the SEDREBO potential for carbon-carbon interaction. Structures are classified by formation energy and atomic coordination, and identification of physically viable samples is achieved using these statistics. The defective regions constituting the grain boundaries are defined using the atomic energy distribution. Formation energies of any viable structures are normalized according to cell height and compared by indexing misorientation angles. Lack of a functional relationship between misorientation angle and formation energy indicates a greater complexity in the mechanisms of the grain boundaries.




n its pristine form, graphene is one of the strongest materials measured, and possesses a wide range of technologically appealing characteristics.  Several recent experiments have explored the mechanical properties of graphene which contains grains, which seem to show contradictory results. In this work, we utilize classical molecular dynamics to investigate the response of bicrystalline graphene subjected to uniaxial and biaxial stress.  Atomic-scale modeling of the material elucidate trends in the strength of the material as a function of misorientation angle and provide clues into the mechanism responsible for the strengthening of the grain boundaries. We also show the validity of the results is highly dependent on the potential chosen to describe the carbon-carbon interactions.  This is achieved by comparing the three most commonly used potentials with our own potential developed previously.