Pathway models for fast ion conductors by combination of bond valence and reverse Monte Carlo methods

Both the bond-valence (BV) method and the reverse Monte Carlo (RMC) method have proven to be useful tools for the investigation of the interplay between the microscopic structure and the transport properties in solid electrolytes. A combination of these two approaches opened a way for a deeper understanding of ion transport, especially in amorphous solid electrolytes, where the lack of reliable structural information so far have impeded minute atomistic descriptions of transport mechanisms. In this paper, we discuss both the abilities and limitations of this new approach. Special emphasis is put on the requirements for the consistent determination of bond-valence parameters that as necessary reflect the bond softness. For,a variety of Ag+ ion conducting glasses, we find, using bond softness sensitive bond-valence parameters, pathways for ion transport in the reverse Monte Carlo produced structural models of the glasses. From the volume fraction of these pathways, we are able to predict both the absolute value and the activation energy of the ionic conductivity. Simulations of ion transport as random walks within the bond-valence pathway network shed further light on the transport process and provide a complementary way to predict the conductivity from structural models. (C) 2002 Elsevier Science B.V. All rights reserved.