ABSTRACT Modeling of glass transition temperature ( T g ), chemical durability, and crystallization was carried out using the glass structure gene modeling (GSgM) approach for phosphate‐based glasses containing simulated high‐level waste (HLW) enriched in MoO 3 , ZrO 2, and RE 2 O 3 . In this approach, the structural groups of phosphate glasses were derived from deconvolution of FTIR spectra, and high‐accuracy structure‐property models for the target properties were developed using limited experimental data. The chemical durability model also revealed the difference in the leaching behavior of Li + relative to Na + and K + , as well as La 3+ relative to Ce 3+ and Nd 3+ . Results showed that these structure‐property models are preferable for property predictions and reliable for fast screening of new glass compositions.
Zhang et al. (Fri,) studied this question.