Title: Cluster Diffusion and Coalescence on Metal Surfaces: applications of a Self-learning Kinetic Monte-Carlo method

Abstract: The Kinetic Monte Carlo (KMC) method has become an important tool for examination of phenomena like surface diffusion and thin film growth because of its ability to carry out simulations for time scales that are relevant to experiments. But the method generally has limited predictive power because of its reliance on predetermined atomic events and their energetics as input. We present a novel method, within the lattice gas model in which we combine standard KMC with automatic generation of a table of microscopic events, facilitated by a pattern recognition scheme. Each time the system encounters a new configuration, the algorithm initiates a procedure for saddle point search around a given energy minimum. Nontrivial paths are thus selected and the fully characterized transition path is permanently recorded in a database for future usage. The system thus automatically builds up all possible single and multiple atom processes that it needs for a sustained simulation. Application of the method to the examination of the diffusion of 2-dimensional adatom clusters on Cu(111) displays the key role played by specific diffusion processes and also reveals the presence of a number of multiple atom processes, whose importance is found to decrease with increasing cluster size and decreasing surface temperature. Similarly, the rate limiting steps in the coalescence of adatom islands are determined. Results are compared with those from experiments where available and with those from KMC simulations based on a fixed catalogue of diffusion processes.