Efficient methods of X-ray diffraction pattern inversion

Abstract

An innovative, computational approach to the problem of X-ray diffraction pattern inversion is presented. The technique approximates the electron density of the trial structures encountered during the structure-fitting part of the inversion process with the superposition of the gas-phase electron densities of the component atoms of those structures. These latter electron densities were obtained using Density Functional Theory (DFT). The approach is shown to be computationally efficient, and to be effective when applied to both existing single-crystal x-ray diffraction data and to a series of simulated data sets obtained from periodic crystals. The effects of thermal fluctuation of the ion cores have been quantified, and the useful range of the technique has been discussed.