PNNL

Abstract

Nuclear waste glasses are generated from the processing of various types of nuclear waste. By extension, the composition of the glass product changes with waste type. The various glasses will corrode upon contact with water. Though it is known that the rate of degradation of the glass matrix is a function of glass composition, the impact of the various chemical constituents in the glass on the glass corrosion rate, in particular the structural role that the various components have, is not known. The aim of the present work is to identify the effect of various chemical species on the glass dissolution rate. Our focus in on glass dissolution in dilute conditions so that only the impact of the chemical constituents in the glass is taken into account without the complexity of solution feedback effects and the precipitation of secondary phases. The results are compared to those of the International Simple Glass (ISG), the reference simple glass that is the focus of this special issue. Glasses dissolve at the forward rate (rf) when exposed to very dilute aqueous solutions. It has long been known that pH and T strongly impact rf, while the impacts of glass composition have remained uncertain. Typically, rf is modeled as: rf =k0aH+-?exp(-Ea/RT). Correlating the model parameters (k0, ?, and Ea) with composition of the glass has long been challenging. We recently reported a strong positive correlation between the log[k0] and Ea parameters fitted to rf data. This strong correlation (>85%) suggests that neither parameter can be uniquely defined given the scale of uncertainties in measure rf data. In the present paper, a model was fitted to rf data generated as part of this study and those found in literature. It was determined that only a relatively small composition effect was seen when fitting the combined data (e.g., 90% of the variation in data was described only by T and pH effects). Relatively weak composition effects where found associated with the log[k0] term and not associated with either pH or T. The rf differences between test glasses, after normalizing for differences in pH and T, were found with variation in the fraction of glass forming tetrahedra contributed by four coordinated boron (fTb). The impact of fTb was non-linear and, based on the current data, it is most likely abrupt threshold with higher rf with high fTb glasses. On either side of the threshold, no composition effects could be identified.