Kinetic Monte Carlo simulation of small vacancy clusters electromigration on clean and defective Cu(100) surface

Electromigration of small vacancy clusters on clean and defective Cu(100) surface is investigated on the atomic scale by performing self-learning kinetic Monte Carlo simulations. Drift velocity dependencies of vacancy clusters on their size, the substrate temperature, the direction and the absolute value of current density are obtained. The drift velocity dependence on the size of vacancy cluster has an oscillatory behavior. The nature of these oscillations is connected with the difference in diffusion mechanisms of “fast” and “slow” vacancy clusters. The presence of point defects leads to the monotonic decrease of the drift velocity of vacancy clusters. The drift velocity drops down if the diameter of the vacancy cluster is larger than the average distance between the point defects.