Building Monte Carlo Models of Glasses Using Neutron and/or X-ray Diffraction Data

Abstract

Neutron and X-ray diffraction are key techniques that are used to investigate the atomic and nanometric mesoscale structure of glasses and amorphous materials. These experimental methods probe the nuclear (neutron) or atomic (X-ray) pair correlation func- tions between atoms. For a glass containing N atom types, the information content of the data is low, considering that the data are a weighted sum of N(N+1)/2 partial pair correlation terms. This complexity can often make direct interpretation of results difficult or impossible. Modern computational methods can now rapidly refine atomistic models of disordered materials that satisfy the constraints imposed by diffraction data. These models can then be used to investigate how the partial pair correlation functions contribute to the total scattering data, given a chosen set of underlying physico-chemical constraints, and allow us to extract many structural functions of interest such as bond angle distributions and coordination number histograms. To illustrate these capabilities the technique of Empirical Potential Structure Refinement (EPSR), has been applied to a range of results from a selection of oxide-glass systems and the results provide a set of reference parameters that can be used in future studies on similar glass systems where EPSR is the goal.