Friction Study of Polycrystalline Graphene Using Accelerated Molecular Dynamics

Graphene has a great potential as a solid lubricant for both macroscopic and small-length scale devices such as micro/nano-electromechanical systems (MEMS/NEMS). While the molecular dynamics (MD) simulation has been widely used to study the frictional properties of graphene by mimicking experimental processes such as the atomic force microscopy (AFM), its small time-scale limits the available sliding velocity at best to meters per second, which is several orders of magnitude larger than experimental values ranging typically from hundreds of nanometers to micrometers per second. Here, an accelerated MD simulation method based on hyperdynamics is applied to study the friction of polycrystalline graphene by decreasing the sliding velocity approaching those in AFM experiments. A bias potential is added to reduce the energy barrier and expedite the thermally-activated transition, based on the bond-boost algorithm which utilizes the bond breaking event during the sliding process. The accelerated simulation is validated with the direct comparison with the unaccelerated MD simulation at intermediate velocities and then is applied at lower velocities. The simulation results reveal that the friction force increase linearly with the logarithm of the sliding speed, which agrees with the prediction of the well-known Prandtl-Tomlinson model.