Understanding the
nature of active sites is a non-trivial task,
especially when the catalyst is sensitively affected by chemical reactions
and environmental conditions. The challenge lies on capturing explicitly
the dynamics of catalyst evolution during reactions. Despite the complexity
of catalyst reconstruction, we can untangle them into several elementary
processes, of which surface diffusion is of prime importance. By applying
density functional theory–kinetic Monte Carlo (DFT–KMC)
simulation employed with cluster expansion (CE), we investigated the
microscopic mechanism of surface diffusion of Cu with defects such
as steps and kinks. Based on the result, the energetics obtained from
CE have shown good agreement with DFT calculations. Various diffusion
events during the step fluctuations are discussed as well. Aside from
the adatom attachment, the diffusion along the step edge is found
to be the dominant mass transport mechanism, indicated by the lowest
activation energy. We also calculated time correlation functions at
300, 400, and 500 K. However, the time exponent in the correlation
function does not strictly follow the power law behavior due to the
limited step length, which inhibits variation in the kink density.