Poly(ethylene-terephthalate), with its high dielectric breakdown strength, is widely used in photovoltaic module backsheets as a core layer. However, it is vulnerable to ultraviolet light and humidity and degrades during outdoor service. Stabilizer additives to polymeric materials can withstand these stresses to some extent, but even these stabilizers are subject to degradation. Towards the purpose of investigating these phenomena, the photolytic and hydrolytic degradation of three PET grades were studied under different accelerated weathering exposures and changes in the polymers were monitored by various evaluation techniques.
It was found that yellowing was mostly caused by light exposures and hazing was predominant in humidity exposures. The UV stabilized grade was found to be slightly more stable than the unstabilized grade. Although the hydrolytically stabilized grade was degraded more severely than the unstabilized grade, the intended stabilization was found to be effective when comparing its degradation between different exposures. The formation of light absorbing chromophores and UV stabilizer bleaching were recorded through optical spectroscopy. FTIR analysis showed that chain scission and crystallization are common mechanisms under both photolytic and hydrolytic conditions, based on the IR absorptions of the carbonyl (C=O) band at 1711 cm-1 and the trans ethylene glycol unit at 975 cm-1, respectively. Even though similar yellowing values were reported for both light and humidity exposures, the reduced fluorescence intensity in the former was attributed to oxidation of fluorescing species; however, the increased intensity in the latter was attributed to the presence of moisture.
The degradation mechanisms determined from optical and chemical evaluations were then used to construct a set of degradation pathway network models using semi-supervised generalized structural equation modeling (semi-gSEM) approach using a stress|mechanism-mode|response framework. This method provided temporal evolution of the degradation by capturing statistically significant relationships between applied stressors, mechanistic variables, and performance level responses. Equations that correspond to these pathways were, lastly, derived and contributions from mechanistic variables were determined. It was found that contributions from mechanistic variables closely predicted the final response under constant exposure conditions; however, cyclic conditions caused complex interactions of multiple degradation mechanisms that are difficult to differentiate and model accurately.