Improving Polymer UV-Aging Resistance: Insights from Degradation Mechanisms and High-Throughput Screening

Photodegradation of high-performance fibers critically limits their utility in aerospace and heat-resistant applications, particularly under acidic photoirradiation conditions. A representative example is poly(p-phenylene benzobisoxazole) (PBO), which degrades under photoirradiation, particularly in acidic conditions. To address this challenge, we present a machine learning framework for high-throughput screening of UV-resistant functional groups to enhance PBO’s photostability. First, density functional theory (DFT) calculations elucidate the degradation mechanism, revealing that phosphoric acid catalysis significantly lowers reaction barriers for oxazole ring opening. Molecular dynamics (MD) simulations with a reactive force field (ReaxFF) further demonstrate that mechanical performance declines when degraded repeating units exceed a critical threshold of 2.50%. Building on these mechanistic insights, a Transformer-based spectral prediction model is developed to screen UV-resistant candidates, identifying three optimal functional groups that preserve PBO’s intrinsic mechanical properties. This integrated computational approach─combining DFT, ReaxFF-MD, and machine learning─provides a scalable strategy for designing UV-resistant polymers, with broader applicability to advanced material systems.