Mixed-alkali effect: some new results

Abstract

Most aspects of the mixed-alkali effect (MAE) are shown to be consistent with calculations employing critical rate analysis to hopping conduction in a random energy landscape. The temperature and frequency dependence of the conductivity can be understood when four factors are incorporated into theoretical models. These four factors are: existence of a spread of barrier heights and well depths, applicability of critical rate analysis, independence of low-lying energy states of different alkali ions, and relevance of dynamic Coulomb interactions. New calculations of the ac conductivity at lower frequencies, but still above the loss peak frequency of the dielectric relaxation, are given. Such frequencies are low enough to allow more quantitative comparison with experimental results for the stretching parameter describing stretched exponential relaxation. This comparison provides additional support for the theoretical interpretation.