Mechanisms of ionic conduction in glass
The effect on conductivity of progressively substituting foreign alkali for host alkali in a series of lithium aluminoborate glasses has been studied. The shape of the conductivity isotherm in the dilute foreign alkali region was found to be the same for both Na+ and K+ dopants. Assuming a 'regular interstitialcy model' for ionic conduction conductivity data for a variety of glasses were subjected to isothermal analysis, allowing concentrations of mobile species and mobilities to be calculated. Conductivity changes were found to be largely attributable to changes in mobility, the concentration term remaining remarkably constant. The a.c. behaviour of aluminoborate glasses has also been studied and modulus spectra fitted using the Kohlrausch-Williams-Watts function for various beta-values. Decreasing the total alkali content and substituting foreign alkali for host alkali was observed to cause some narrowing of the modulus, mixed-alkali glasses behaving much as though only 'host' cations were present. 'Microscopic' activation energies (E_a) have been calculated from beta-values (Ea=beta Ead.c.). These are found to follow the trend in experimentally observed activation energies indicating that the mixed alkali effect occurs as the result of decreases in both long- and short-range ionic mobilities. A cluster-bypass model has been proposed in an attempt to reconcile the above results. The model assumes that the mobile species are to be found in a tissue material surrounding the clusters. A brief examination of the site preferences of spectroscopic probe ions (T1+ and Pb2+) in mixed cation glasses by u.v. spectroscopy indicated a dependence of the frequency of the absorption band on the reagents used in glass preparation. This result is thought to point towards some structural effects persisting in the melt which in turn may be important with respect to cluster formation in the glass.