In the classical molecular dynamics (CMD) technique, the role of the electronic system responsible for particle interactions is emulated by the use of empirical functions of individual particle coordinates. From these empirical functions it is possible to extract the forces among particles and then numerically solve the Newton equations of motion in the computer. The results of the simulation are the trajectories followed in time by each particle in the system. Empirical potentials include several parameters that are fitted to experimental measurements or more fundamental simulation results such as those obtained by ab initio techniques. Since the electronic system is not considered in the simulation, with empirical potentials it is possible to reach larger system sizes and time scales than those allowed by ab initio or tight-binding techniques. For example, it is possible to simulate systems of millions of atoms for times of the order of nanoseconds. If the empirical potential is adequate, useful information regarding the system dynamics can be obtained from simulations.
We use this technique to study defect morphology, diffusion and energetics, generation of damage due to irradiation, and recrystallization processes.
Animation showing the diffusion of the Si self-interstitial from a CMD simulation (L. A. Marqués et al., Phys. Rev. B 71, 085204 (2005) - Molecular dynamics study of the configurational and energetic properties of the silicon self-interstitial).
|Classical molecular dynamics|
|Binary collision approximation|
|Kinetic Monte Carlo|