Process-Specific Point Defect Modeling in CdTe with KROGER—Accounting for Finite Sample Dimensions and Cooling Rates

Semicondcutor majority and minority carrier properties are determiend by impurity and native defects, which in turn are determined by each sample's microstructure and specific procesing history. Calculation of defect equilibrium is fairly straightforward based on pair formation equilibria or on formation energies computed from density functional theory. However, examples of real world samples in which defects are fully equilibrated spatially throughout the entire volume are few and far between. Therefore raw defect formation energies provide only part of the story and can not explain differences between bulk, thin film, and polycrystalline samples. For example, frequently substitutional dopants are kinetically trapped near the growth temperature while native defects and other interstitals may continue to equilibrate to lower temperatures becasue they face smaller migration barriers. In addition to incorporating a nearly exhaustive list of temperature dependent effects, our calculation framework KROGER allows either chemical potentails or concentrations to be specified for each element, allowing cases of kinetic trapping to be modeled correctly. Recently, we have generalized this flexability to perform sequential freeze-in calculations, automatically for ranges of sample dimensions and cooling rates. These abilities combined with the numerical efficiency of KROGER allow defect modellers to quantitatively predict defect ensembles for slow-cooled or quenched samples of different dimensions. We will illustrate these capabilities and achieving quantitative agreement with experiments throguh examples in Ga₂O₃ and CdTe.