Predictability of ion transport properties from the structure of solid electrolytes

The concept of bond valence (BV) is widely used in crystal chemical considerations, e.g. to assess equilibrium positions of atoms in crystal structures from an empirical relationship between bond length RM-X and bond valence s(A-X) = exp [(R-O - RM-X) / b] as sites where the BV sum V(A) = Sigma s(M-X) equals the formal valence V-id of the cation M+. Our modified BV approach that systematically accounts for the softness of the bond may then be effectively used to study the interplay between structure and properties of solid electrolytes. This is exemplified for correlations to experimental data from IR, NMR, and impedance spectroscopy. Combining the bond valence approach with reverse Monte Carlo (RMC) modeling or molecular dynamics (MD) simulations provides a deeper understanding of ion transport mechanisms, especially in highly disordered or amorphous solids. Local structure models for crystalline electrolytes are derived by combining crystallographic structure information with simulations. A method for the prediction of the activation energy of the ionic conductivity from the bond valence analysis of the crystal structure is proposed. Taking into account the mass dependence of the conversion factor from bond valence mismatch into an activation energy scale, we could establish a correlation that holds for different types of mobile ions. The strong coupling of the H+ transfer to the anion motion in proton conductors requires a special treatment. For glassy solid electrolytes RMC structure models are BV-analyzed to assess the total number of equilibrium sites and to identify transport pathways for the mobile ions. Recently, we have reported a correlation between the pathway volume fraction and the transport proper-ties that permits to predict both absolute value and activation energy of the dc ionic conductivities of disordered solids (including mixed alkali glasses) directly from their structural models. Here we discuss a corresponding BV analysis of molecular dynamics simulation trajectories that allows quantifying the evolution of pathways in time and the influence of temperature on the transport pathways.