Predictions of polymer thermal degradation: relevance of selecting the proper kinetic model

Abstract

Making predictions, such as lifetime estimations, is one of the main objectives of kinetic studies. Thus, from conventional thermal analysis experiments, the behavior of polymeric materials under processing or application conditions, usually far away from those used in the laboratory experiments, could be estimated. Conventional prediction procedures usually make use of oversimplified equations based on simple approaches. One of the most common approaches is the assumption of a first, or n-order, kinetic model for the process. However, recent studies have shown, for a number of polymers, that random scission kinetic models are not only physically sound, but more reliable in terms of describing the degradation kinetics. In this paper, the consequences of selecting an erroneous kinetic model on lifetime predictions is discussed. It is demonstrated, using both simulated and experimental data, that any kinetic analysis of a chain scission driven reaction performed assuming a first-order model entails enormous deviations in predictions. This occurs despite the fact that the first-order kinetic model can fit experimental data from chain scission driven reactions with significant correlation coefficients, and even lead to a reasonably good reconstruction of the original experimental curves.