For decades, scientists have relied on borosilicate glass to immobilize radioactive waste, a method selected by the U.S. Department of Energy (DOE) as early as 1982 for its stability and long-term effectiveness.
Until recently, much of the experimental data behind this approach remained scattered across institutional archives and technical reports. That changed in 2019, when Savannah River and Pacific Northwest National Laboratories, supported by the DOE’s Office of Environmental Management, launched the first online, open-access nuclear waste glass database.
The project has since grown with contributions from Dr. Trivelpiece (SRNL), Dr. DP Aidhy (Materials Science & Engineering, Clemson University) and Dr. Collin Wilkison (Alfred University), creating a centralized, searchable platform where users can filter data by composition, viscosity, durability, and more. Importantly, the database also integrates quality assurance metadata, helping researchers evaluate the reliability of each dataset.
In 2024, the platform expanded dramatically with the integration of SciGlass Next, a resource of more than 400,000 glass entries, and machine-learning models capable of predicting glass properties. Together, these additions enabled the creation of the Virtual Center for Nuclear Waste Glass Science, which now features information on over 6,000 compositions, along with predictive statistical modeling via the “Stat Mech Glass” Python module.
Currently in its final beta testing phase, this open-access environment is poised to transform nuclear waste research—accelerating discovery, guiding vitrification strategies, and fostering global collaboration.
Read the full article on ACerS Bulletin
