Computational scheme for ab-initio predictions of chemical compositions interfaces realized by deposition growth

We present a novel computational scheme to predict chemical compositions at interfaces as they emerge in a growth process. The scheme uses the Gibbs free energy of reaction associated with the formation of interfaces with a specific composition as predictor for their prevalence. It explicitly accounts for the growth conditions by rate-equation modeling of the deposition environment. The Bell-Evans-Polanyi principle motivates our emphasis on an effective nonequilibrium thermodynamic description inspired by chemical reaction theory. We illustrate the scheme by characterizing the interface between TiC and alumina. Equilibrium thermodynamics favors a nonbinding interface, being in conflict with the wear-resistant nature of TiC/alumina multilayer coatings. Our novel scheme predicts that deposition of a strongly adhering interface is favored under realistic conditions.