Multiscale Modeling of the Atomic Layer Deposition of HfO2 Thin Film Grown on Silicon: How to Deal with a Kinetic Monte Carlo Procedure

An original integrated approach developed within a multiscale strategy, which combines first-principles quantum simulations and kinetic Monte Carlo (KMC), is presented to investigate the atomic layer deposition (ALD) of HfO2 on Si(100) surface. Density functional theory within the hybrid functional is used to determine the detailed physicochemical mechanisms and associated energetics of the two half cycles taking place during the initial stage of film growth. A kinetic Monte Carlo model is then proposed that deals with the stochastic nature of the calculated DFT mechanisms and barriers. Beyond the chemical information emanating from DFT calculations, the lattice-based KMC approach requires preliminary physical considerations issued from the crystal structures that the system is intended to adopt. This is especially critical in the case of heterogeneous systems like oxides deposited on silicon. We also describe (i) how atomistic configuration changes are performed as a result of local events consisting in elementary reaction mechanisms occurring on specific lattice sites, (ii) the temporal dynamics, governed by transition probabilities, calculated for every event from DFT activation barriers, and (iii) the relation of KMC with the ALD experimental procedure. Some preliminary validation results of the whole multiscale strategy are given for illustration and pertinence with regard of the technological main issues.