A three-dimensional self-learning kinetic Monte Carlo model: application to Ag(111).

Area of Science:

• Computational Materials Science
• Surface Science
• Chemical Physics

Background:

• Kinetic Monte Carlo (KMC) simulations are vital for modeling materials processes.
• Simulation reliability hinges on accurate transition rates, often computationally expensive to obtain.
• Existing self-learning KMC methods offer a balance between accuracy and efficiency.

Purpose of the Study:

• To extend the two-dimensional self-learning KMC method to three dimensions.
• To develop strategies for managing the increased computational cost of 3D KMC simulations.
• To validate the enhanced 3D method with relevant materials science case studies.

Main Methods:

• Expansion of the self-learning KMC algorithm from 2D to 3D.
• Implementation of an initial database of activation energies derived from a simpler KMC model.
• Application of a pattern recognition scheme for efficient rate cataloging.

Main Results:

• Successful extension of the self-learning KMC method to three dimensions.
• Demonstrated avoidance of excessive on-the-fly rate calculations through database pre-computation.
• Validated the 3D method for Ag diffusion and homoepitaxial growth on Ag(111).

Conclusions:

• The 3D self-learning KMC method provides an accurate and efficient approach for materials simulations.
• Pre-computation of activation energies in a database significantly reduces computational burden.
• The method is applicable to complex surface processes like diffusion and epitaxial growth.