Bond valence analysis of ion transport in reverse Monte Carlo models of mixed alkali glasses

Abstract

ABSTRACTAn analysis of RMC structure models of ion conducting glasses in terms of our bond softness sensitive bond-valence method enables us to identify the conduction pathways for a mobile ion as regions of sufficiently low valence mismatch. The strong correlation between the volume fraction F of the “infinite pathway cluster” and the transport properties yields a prediction of both the absolute value and activation energy of the dc ionic conductivities directly from the structural models. Separate correlations for various types of mobile cations can be unified by employing the square root of the cation mass as a scaling factor. From the application of this procedure to RMC models of mixed alkali glasses, the mixed alkali effect, i.e. the extreme drop of the ionic conductivity when a fraction of the mobile ions is substituted by another type of mobile ions may be attributed mainly to the blocking of conduction pathways by unlike cations. The high efficiency of the blocking can be explained by the reduced fractal dimension of the pathways on the length scale of individual ion transport steps.