Comparative study of the artificial and natural, indoor ageing of crosslinked, polyurethane ester closed-cell foams

• Artificial ageing cannot fully replicate the chemical and morphological complexity of long-term NA. • The comprehensive comparison of naturally and artificially aged PU-ES closed-cell foams revealed fundamental differences between them. • Observed natural ageing phenomena included blooming, pox, and dried fluids, all linked to hydrolysis but distinct in form and/or chemical composition. • While general mechanical embrittlement was successfully mimicked upon artificial ageing, the specific adipic acid – related ageing phenomena were not. • Leveraging naturally aged materials from industrial heritage collections allows a rare glimpse into processes taking place at long times otherwise difficult to replicate. This study compares naturally aged (NA) closed-cell polyester urethane (PU-ES) foams with an artificially aged (AA) chemically crosslinked, comparable PU-ES. The foams are shoe soles part of a museum collection from a well-defined time, place, and use, and were thoroughly characterised visually, chemically, and, where possible, mechanically. Under museum conditions, naturally aged materials displayed hydrolysis related phenomena namely blooming, pox, and dried fluids, consisting either of adipic acid deposits, or resulting from reactions between polyester polyol oligomers and shoe metal components. Shore hardness, only method applicable to both sets of samples, showed that 56 days at 70°C and 98 %RH produced a loss of elasticity in AA material comparable to ∼25 years of natural ageing. Yet, artificial ageing was not able to replicate NA surface phenomena, signalling a fundamental difference between material undergoing both kinds of ageing. Particular effort was devoted to the mechanical and chemical characterization, and to evaluating pyrolysis-GCMS for monitoring ageing in PU-ES. While detailed MS interpretation enabled proposal of several structures of pyrolysis products, certain key issues intrinsically affected the repeatability of py-GCMS analysis, making MDA, a proposed marker for PU-ES degradation, unsuitable for following ageing. Alternative methods for this purpose included the infrared band ratio 1727 cm⁻¹ / 1706 cm⁻¹, which increased consistently with accelerated ageing but was reliable only when adipic acid did not obscure the region. In contrast, GCMS of material extracts showed higher repeatability and promise for identifying molecular ageing markers. Overall, the results indicate that while artificial ageing can reproduce certain mechanical changes, it cannot fully replicate the chemical and morphological complexity of long-term natural ageing, highlighting the importance of complementary analytical strategies and the potential of the study of naturally aged materials from industrial heritage collections. By leveraging their repetitive nature and well-established, industrially manufactured formulations from collections, this study demonstrates how such materials can uniquely bridge natural and artificial ageing research— not only refining conservation strategies to ensure their safeguarding for future generations, but also allowing the study of long-term polymer degradation processes otherwise difficult to access within current material science strategies.