Historic glass artefacts, especially those submerged in marine environments for centuries, serve as natural testbeds for long-term material stability. Understanding how glasses corrode over decades or centuries is critical for a wide range of applications, from cultural heritage conservation to the safe immobilisation of nuclear waste.
Until now, lab-based accelerated corrosion tests have been used as proxies – but how well do they mirror real-world behaviour? Researchers from the University of Sheffield and their marine archaeology and materials science collaborators tackled this question in their paper, published in Applied Geochemistry. The team analysed ancient lead-silicate glass ingots recovered from the 18th century shipwreck of the HMS Albion and compared them with laboratory-weathered analogues. Using micro-X-ray absorption spectroscopy at Diamond’s I18 beamline, they probed element redistribution and corrosion layer development.
When glass is immersed in saline, i.e. oxygenated seawater for extended periods, its surface layers evolve through ion leaching, new phase formation (such as clays or zeolites) and structural collapse – all of which affect long-term durability. Accelerated lab tests simulate this but often differ in time scale, solution chemistry or mechanical stress, which raises the question: Do accelerated tests capture the full complexity of real glass corrosion?
In this study, the team analysed shipwreck-altered lead-silicate glass (with visible corrosion layers up to several hundred microns thick) and compared the leached layers, element migrations (e.g., Pb, Si, Ca) and secondary mineral formation with artificially corroded analogues in lab immersion tests.
At Diamond’s I18 beamline, the team used micro-focused XAS and XRF to map elemental speciation and spatial distribution across corrosion layers.
Read more on the Diamond website