Experimental and Computational Characterization

Abstract

This chapter briefly describes and assesses the strengths and weaknesses of various experimental and computational methods for understanding defect behavior. Defect properties, including thermodynamic ionization levels, can often be derived from direct experimental techniques such as electron paramagnetic resonance, deep level transient spectroscopy experiments, and scanning tunneling microscopy. As defects often exist at low concentrations, indirect methods such as diffusion measurements are frequently required. Also, quantum calculations by density functional theory can give valuable insight into the electronic and structural properties of charged defects, and indeed comprise a large portion of the overall literature. These methods are briefly described; however, the various methods often yield findings that differ greatly from each other and from experiments. For this reason, systems-based methods such as maximum-likelihood estimation are especially valuable for distilling the results of disparate methods into a single “best” value for a defect formation energy, ionization level, or activation energy of diffusion. This chapter describes the implementation of maximum likelihood estimation as used elsewhere in the book.