Artificial neural networks for Cu surface diffusion studies

Kinetic Monte Carlo (KMC) is a powerful method for simulation of diffusion processes in various systems. The accuracy of the method, however, relies on extent of details used for the parameterization of the model. Migration barriers are often used to describe diffusion on atomic scale, but the full set of these barriers may become easily unmanageable in materials with increased chemical complexity or a large number of defects. In this work, we apply a machine learning approach for Cu surface diffusion. We train an artificial neural network on a subset of the large set of $2^{26}$ barriers needed to describe correctly the surface diffusion in Cu. Our KMC simulations using the obtained barrier predictor shows sufficient accuracy in modelling {100} and {111} surfaces. The {110} surface could be modelled by overriding a limited set of barriers given by the network.