| Theoretical descriptions of ion transport processes involving real glass systems are often idealized. Below their glass transition temperatures, ion exchange glasses are usually assumed to have a spatially rigid network with only monovalent cations as the mobile components. The effect of one cation's movement on another is presumed to be purely electrostatic. This thesis shows site relaxation can exist transmitting into additional cross interactions between cations for certain oxide glasses. The oxide glasses are categorized by cation environments, being non-bridging oxygen (NBO) rich and bridging oxygen (BO) rich. Macroscopically, the relaxation is manifest by time dependent interdiffusion and radiotracer diffusion coefficients in a BO rich glass. The interdiffusion coefficient for a NBO rich glass is found to be time independent.; By using irreversible thermodynamics, phenomenological expressions for the diffusion process can be written. However, it is not immediately obvious how to connect proper theoretical treatment of the more complex relaxing system to experiment. In contrast, the traditional approach can still be applied if the appropriate selection and determination of phenomenological coefficients are made at a fixed diffusion time. We demonstrate the novel approach of using dilute co-tracer diffusion in conjunction with dilute single tracer diffusion to experimentally determine the appropriate phenomenological coefficients to use for ionic diffusion in a relaxing glass system. Non-reciprocal cross phenomenological coefficients are realized for the relaxing BO rich glass. To our knowledge this represents the first known measurement of cross phenomenological coefficients for dilute cation diffusion in glass. Concentration dependence is accounted for by using a series of uniform samples with different compositions. Time dependence is measured by using different diffusion times.; The interdiffusion coefficient describing non-dilute ion exchange is defined in terms of the phenomenological coefficients (including cross coefficients). The corresponding thermodynamic term, relating dilute parameters to non-dilute ones, is measured using the salt bath equilibrium technique. In our time dependent system, the thermodynamic term is measured in the base samples represented by zero diffusion time. Thus, the thermodynamic term is better suited for short diffusion times and less suited for longer diffusion times. |
