Ab initio thermodynamics of deposition growth: chemical composition of TiC/alumina

We present a method to theoretically determine and predict chemical compositions at surfaces and interfaces as they emerge during deposition growth in complex environments. The method is based on ab initio calculations and permits an improved account of as-grown reactive surfaces. It also increases our understanding of binding in as-grown interfaces. The method could guide choices of growth parameters to achieve, e.g., a prespecified nature of surface reactivity. Our method combines density functional theory calculations with thermodynamic concepts and rate equations describing the deposition process. We base our predictions on the Gibbs free energy of reaction G r for processes that lead to surfaces (interfaces) with different chemical compositions. The calculation of G r requires calculations of total energy differences and the specification of chemical potentials of all involved gas-phase species that enter or leave the reaction chamber. We use the ideal-gas approximation and thermochemical data to determine the chemical potentials as a function of temperature and the steady-state partial pressures. We do not assume equilibrium between the gases and the surface. Earlier implementations of ab initio thermodynamics [1] have focused on oxides and assume equilibrium between their surfaces (or interfaces) and a surrounding, O2-dominated environment. This leads to two main limitations. First, these implementations of ab initio equilibrium thermodynamics are by construction limited to materials that contain at least one constituent ’A’ that possesses a gas-phase counterpart ’A n ’ (n an integer). Second, asdeposited systems are, in general, not characterized by strict equilibrium. We exemplify our method for the TiC/alumina interface. Multilayers of these materials are commonly fabricated by chemical vapor deposition