Atomistic simulations of graphite etching at realistic time scales

Abstract

We demonstrate that long time-scale events in atomistic ion-surface bombardment simulations can be essential and need to be accounted for.

Mechanisms of elementary hydrogen ion-surface interactions during multilayer graphene etching at high surface temperature as a function of flux

“…Despite the reduced H uptake, formation of weakly bound CH3 groups and erosion are initiated after fewer impacts (after ~300 impacts for both cases, rather than ~500 and ~600, respectively). This can attributed to a similar mechanism as was found in previous work [49]: a higher probability for C-C bond breaking due to prolonged exposure to thermal stress resulted in more potential binding sites [49] (this is also supported by the observed transition towards hydrocarbon release by thermally-induced C-C bond breaking, see Fig. S2 in the supplementary information).…”

“…2): thermally-induced hydrocarbon erosion (which was already discussed in Ref. [49]), surface diffusion, hydrogen desorption, and diffusion- where both H atoms hopped at least once before recombining.…”

“…The typical high surface temperature of 1000 K enabled us to explore the transition towards thermal H-surface interactions. An elaborate description of the simulation method is provided in our previous article [49], and briefly summarized below. The simulations were performed using the LAMMPS package [56], and modified Colvars module [57].…”

“…Long time-scale processes (of the order of µs-ms) that become important at elevated surface temperatures and low fluxes can thus not be accessed. Examples of these processes are: hydrogen surface diffusion [46], Langmuir-Hinshelwood recombination, desorption of hydrogen and weakly bound species [47], and other surface relaxation phenomena [48,49].…”

“…In order to reach these desirable longer time-scales of thermal processes, we have presented a simulation approach in our previous work [49]; the applications of a recently developed acceleration procedure, collective variable-driven hyperdynamics (CVHD) [50]. This is a generic implementation of the hyperdynamics method [51], in which the waiting time between minima-to-minima transitions is reduced by adding a bias potential (ΔV) to the global potential energy surface (PES) of the system.…”