AVS 52nd International Symposium
    Surface Science Thursday Sessions
       Session SS1-ThA

Paper SS1-ThA2
Diffusion of Two Dimensional Cu Adatom Islands on Cu(111) using a "Self Learning" Kinetic Monte Carlo Technique

Thursday, November 3, 2005, 2:20 pm, Room 202

Session: Transport and Structural Stabilization of Surfaces
Presenter: T.S. Rahman, Kansas State University
Authors: T.S. Rahman, Kansas State University
O. Trushin, Russian Academy of Sciences
A. Kara, Kansas State University
A. Karim, Kansas State University
P. Vikulov, Russian Academy of Sciences
Correspondent: Click to Email
Diffusion of two dimensional Cu adatom islands, containing 1 to 100 atoms, on Cu (111) has been studied using a newly developed "self learning" Kinetic Monte Carlo (SLKMC) technique in which the standard KMC method is combined with procedures for automatic generation of a table of microscopic events and calculation of their activation energy barriers. Nontrivial paths thus revealed and fully characterized are permanently recorded in a database for future usage, through a pattern recognition scheme. The system thus automatically builds all possible single and multiple atom processes that it needs for sustained simulation and evolves according to processes of its choosing. The diffusion coefficients calculated for 300K, 500K, and 700K, show that the effective diffusion barriers increase almost monotonically with increasing island size. From the tabulated frequencies of events we find that concerted motion and multiple particle processes play a key role in the diffusion of small islands, however, periphery diffusion and single particle processes dominate for the larger sized islands. Contrary to the case of small Cu islands on Cu (100), we do not find any oscillatory behavior for the effective diffusion barrier or for the diffusion coefficients as a function of island size.