The relevance of site energy distribution for the mixed alkali effect

Abstract The mixed alkali effect (MAE) can be explained as an interplay between network modifications induced by alkali ions of different size and an energy distribution of sites with a preferential occupancy of the low energy sites by the smaller cations. The underlying physical assumptions of this concept are discussed in the present paper in more detail and they are compared with results of Monte Carlo simulations of diffusion and the Anderson–Stuart model. It will be shown that the new concept is actually hidden in the Anderson–Stuart model. Applying the new concept to experimental data has been done so far with a box-type distribution of site energies. Besides the oversimplification, a major drawback of this distribution is that the diffusion coefficient of the smaller and minority cation is underestimated. It will be shown in the present study that by a different choice of distribution the concentration dependence of the diffusivity of the minority cations can be understood. In addition some results in quaternary glasses will be explained by a more general distribution function.