Commercially produced thin films are grown in a variety of ways such as by evaporation, with or without ion beam assist, or by magnetron sputtering. There are a lot of different parameters that control the growth process, such as the energy of the impacting particles, the stoichiometry of the arriving species and in the case of magnetron sputtering the pulse frequency and current density. However it is not possible to include these parameters into a molecular dynamics simulation (MD) because the deposition times involved are beyond those reachable computationally by MD.
Recently however there have been significant advances in multi-time scale methods whereby the impacting particles are modelled using MD and the diffusion events between impacts with an on-the-fly kinetic Monte Carlo technique. This methodology has been incorporated into a computer code which can be used to model thin film growth over realistic time scales. Saddle point searches between impacts are carried out in parallel at the same time as a new impact event and a transition or new impact event chosen according to their relative probabilities.
The method has been used to investigate the growth of TiO2, ZnO, Ag and Al thin films with and without the simultaneous bombardment of Ar. The methodology can explain how stacking faults and twinning can occur in thin film growth, the role of defects in the growth process but more importantly it can be used quantitatively to advise experimentalists of the conditions by which the films can be grown to minimise defects and achieve the best crystallinity.