Abstract
Measurements of the electrical impedance of LiPO3 glass at frequencies between 1 Hz and 106 Hz were analyzed using two popular procedures: (1) fitting the conductivity and permittivity to power laws in frequency, and (2) fitting the electrical modulus to a stretched exponential. The parameter, β, which expresses the non-exponential nature of the relaxation of electrical perturbations, is substantially different for these two procedures. The value obtained from the frequency dependence of the conductivity via the coupling model, βn, is similar to that seen for mechanical and nuclear magnetic resonance relaxations, suggesting that it may be more physically meaningful, whereas that from the electrical modulus, βm, is significantly larger. This dissimilarity is surprising since others often use these two βs interchangeably when testing theoretical predictions. This result will significantly after previous conclusions drawn from such tests.