Developing an understanding of the relationship between the durability of simplified and complex UK radioactive waste glasses

Safely disposing of radioactive waste glass within a geological disposal facility requires a thorough understanding of the kinetics and mechanisms of their aqueous dissolution over geological timescales. Whilst a significant number of studies have taken place on international waste glass compositions, major compositional differences between UK Magnox and international waste glasses render many of these studies inapplicable to UK Magnox glasses. Notably, one of the major compositional differences between UK and international glasses is the presence of Li. Further, the presence of excess Li concentrations in some UK vitrified products has meant the effects of Li on the aqueous durability of UK waste glasses needed to be investigated. Additionally, dissolution experiments commonly take place at a higher temperature than the expected temperature of groundwater within a geological disposal facility. As such, the effects of dissolution temperature on the mechanisms of dissolution was also investigated. As the Li contents of complex waste glasses cannot be varied, two seven-component analogues of molar Li:Na ratios of 1.0 and 1.5 based upon a 25 wt.% loading UK Magnox waste glass with and without excess Li contents were fabricated. After characterising these glasses, they were leached in a mechanistic study to investigate the effects of Li on chemical durability. In addition to 7Li MAS-NMR and 6Li-1H CP-NMR, comprehensive 11B and 23Na MAS-NMR studies and SEM imaging of the leached glass surfaces then took place to investigate the effects of dissolution on the structures of the glasses. Here we show that the proportion of IIIB to IVB units did not evolve as the Li:Na ratio was varied; suggesting that Na preferentially charge compensates the B network. However, the B network was shown to leach incongruently at 90 °C. Despite Li being shown to be detrimental to durability during the earlier dissolution regimes, the residual rates of alteration implied excess Li contents had no long-term effects on chemical durability. The observed incongruent dissolution of the B network and initially decreased chemical durability could be attributed to Li preferentially modifying the Si network, thereby promoting glass hydration and B network dissolution whilst the Na compensated IVB units were less affected than IIIB units. Additionally, Li and Na were shown to be incorporated into secondary phases at 90 °C, but these Na-bearing phases were not observed at 40 °C. To compare the effects of Li contents on international and UK waste glass compositions, Li was substituted for Na in the well-studied Li-free French analogue “International Simple Glass” at molar Li:Na ratios of 0.4 and 0.9. These fabricated glasses were then leached and characterised in the same manner as the Magnox waste glasses. It was shown that the IIIB/IVB ratio and the role of Na in the pristine glasses varied insignificantly with the Li:Na ratio. Further, the B network of these glasses was shown to leach congruently at both 40 and 90 °C. However, Li, Na and Mg were shown to not be incorporated into secondary phase precipitates for these glasses and the leached glass surfaces displayed only minimal evidence of surface alteration. Contrastingly, whilst Li was shown to be detrimental to aqueous durability, further substituting Li for Na improved long-term aqueous durability. This was attributable to a Li-Na mixed alkali effect which was not evident in the Magnox waste glasses. The effects of dissolution temperature on the kinetics and mechanisms of dissolution of Magnox waste glasses still needed to be investigated. As such, a Magnox waste glass of 25 wt.% simulant waste loading was leached in static batch experiments at 40, 70, 80 and 90 °C to investigate the Arrhenius dependence of dissolution. Leached samples were characterised by EDX, SEM and XRD. It was shown that changing the dissolution temperature changed the rate of hydrolysis relative to interdiffusion. At higher temperatures, the initial release of Na deviated from Arrhenius-type behaviour and instead displayed an almost flat Arrhenius plot; demonstrating changes in temperature affect Na differently to other glass species. Whilst the activation energies of the Li and B releases were in the range of a mixed reaction, the higher activation energy of Na at lower temperatures combined with its non-Arrhenius behaviour suggested the dissolution processes of Li and Na differed. These observations were attributed to the preference of Na to charge compensate the B network. These results highlighted a need for an additional dimension with which to probe the mechanisms by which glasses dissolve. As such, a simplistic proof of concept dissolution experiment took place at 90 °C on a simplified analogue of a complex waste glass to investigate whether the temporal evolution of the isotopic signatures of the glass leachates could provide information on glass dissolution mechanisms. Li, B and Mg isotope ratio analysis took place on the glass leachates. It was shown that dissolution was initially congruent but diffusive processes were rate limiting whilst the glass dissolved at its residual rate. These isotopic techniques were then applied to investigate the temperature dependence of dissolution of a complex Magnox waste glass. Contrastingly, it was shown that diffusive processes initially controlled dissolution but such diffusive processes were not visible at longer durations. These results suggested the same dissolution processes were taking place at both 40 and 90 °C.

This work was supported by EPSRC under an Industrial CASE award with the Nuclear Decommissioning Authority (NDA) and Radioactive Waste Management (RWM) (Grant Ref: EP/M507350/1).