Cluster-Based Semi-Empirical Model for Dopant Activation in Silicon Carbide

Abstract

We present a cluster-based semi-empirical model for dopant activation in silicon carbide (SiC). We model the following species: dopants on lattice points, point defects, dopant-defect pairs, and small clusters of different sizes. We define the possible reactions between these species, add their reaction kinetics, and use a system of ordinary differential equations (ODEs) to model the time evolution of the concentration of the different species during annealing. We use the MaxLIPO+TR optimizer to obtain the post-implant conditions of the SiC film, including the various cluster concentrations. These concentrations are not measurable and can only be calculated through time-intensive atomistic simulations, which we apply to verify and calibrate our model. The framework presented here, consisting of an ODE model generator, an ODE solver and an optimizer, gives a practical solution to predict as-implanted defect concentrations, which is missing from previous works.