Applicability of stretched exponential functions for describing dynamics in disordered solid materials

Investigating dynamics in a disordered solid material below, at, and well above glass transition temperature, we show that (1) to describe glass dynamics entirely it has to be regarded over a long range on logarithmic time scale, (2) a single stretched exponential function (Kohlrausch-Williams-Watts) can never describe the data, (3) stretching exponents do not cover the ranges previously suggested (from 0 to 1, e.g. as a sigmoid function). Optically recorded dynamics (measured by ellipsometry) is brought into connection with dielectric spectroscopy.