The pitfalls of empirical fitting of glass relaxation data with stretched exponents

Abstract

A stretched exponent is commonly used to fit experimental relaxation data for glasses, which typically exhibit a range of time constants. While it has been supported by theory for various processes, mostly near and above the glass transition temperature, Tg, it is also commonly used in phenomenological fits below Tg without a mechanistic model. The properties of the stretched exponent, and sample data for the enthalpy release and dynamic-mechanical response of metallic glasses, are used to show that in the absence of a mechanistic model, stretched-exponent fits easily lead to artifacts and flawed conclusions.