Thermal treatment of the UK’s intermediate level nuclear waste, such as vitrification, offers significant long-term storage savings compared to cementation. However, concerns exist surrounding the volatility of components present in intermediate level waste streams that must be addressed for thermal treatment to be a viable immobilisation option. Of particular interest are caesium radioisotopes, which account for a large proportion of the radioactivity in intermediate level waste.
This thesis aims to assess the volatility of caesium species during vitrification using a bespoke laboratory apparatus, allowing for in-situ assessment of the volatile species produced. This information can be fed back into the glass formulation to minimise/eliminate volatility during vitrification.
A bespoke, in-situ off-gas capture and analysis system has been designed and commissioned to enable this work, briefly comprising of a gas tight stainless steel system, a fibre coupled Raman spectrometer, connected to the steel apparatus using a quartz window Raman flow cell and gas washing bottles to allow for the analysis of off-gas solutions to aid a caesium mass balance.
To aid the identification of caesium bearing volatiles produced during the glass melts, trials were conducted in the system described above using caesium carbonate, caesium borate, and a mixed alkali borate. The purpose of these trials was to collect reference spectra of known volatile caesium species in the same experimental apparatus as the glass melts were to be conducted in.
Two glass series were designed to facilitate this work, an iron phosphate series and a sodium borosilicate series. Alkali borosilicate glasses are the most commonly used glass type for the immobilisation of high activity nuclear waste, in this work, the influence of additives (CaO and ZnO) on caesium retention during the vitrification process were investigated. The results presented in this thesis feed into the wider body of research conducted on the effect of CaO and ZnO on glass structure, which in turn influences important properties such as chemical durability and wastestream solubility. Iron phosphate glasses were selected for this study as a promising candidate for UK wastestreams that are poorly suited to borosilicate glasses. Both glass types were thoroughly analysed before being re-melted with varying caesium oxide waste loadings in the system described above.
The iron phosphate glass series investigated three additives, B2O3, MnO and ZnO. Each of the additives were substituted into a 40Fe2O3-60P2O5 iron phosphate system in place of Fe2O3. The main aim of studying a range of additives at varying concentration was to investigate their effect on caesium retention, however each additive was expected to influence the glass system in a different way. B2O3 was selected to investigate the relationship between B2O3 and caesium volatility at varying B2O3 concentrations, as the existence of caesium borate species above glass melts has been demonstrated in the literature. MnO was selected to investigate the relationship between the iron oxidation state and caesium retention, as Mn was expected to form a redox couple with the Fe present in the glass system. Finally, ZnO was selected as studies conducted on borosilicate glass systems have shown ZnO to have a positive influence on caesium retention and this study aimed to investigate if the same effect would be observed in an iron phosphate glass system. It was found that B2O3 and ZnO additions improve caesium retention, whilst MnO additions showed minimal improvement when compared to the 40Fe2O3-60P2O5 system. All glasses with additive concentrations of less than 7.5 mol% crystallised during the caesium doped remelt, with B2O3 melts crystallising regardless of the additive concentration. Glasses containing 7.5 and 10 mol% MnO or ZnO remained amorphous after the caesium doped remelt with Raman spectroscopy of the bulk glass indicating no significant structural changes.
The sodium borosilicate system investigated the influence of ZnO on caesium volatility, using a modified version of the UK’S CaZn composition, which is currently used to immobilise high level waste in the UK. Glasses with varying Zn:Ca ratios (100:0 to 20:80) were produced and remelted at Cs2O waste loadings of 2, 5, and 10 wt.%. A correlation was found between the measured Zn:Ca ratio and caesium retention. Glasses with low ZnO concentrations performed worse than the glasses with higher ZnO concentration. However, the NaBSZn100Ca0 composition did not perform as well as glasses containing CaO. Off-gas Raman analysis of the caesium vapours showed that at Cs2O loadings of 5 and 10 wt.%, a mixed alkali borate (NaCs(BO2)2) species was the dominant volatile identified. At the 2 wt.% Cs2O loading, the off-gas Raman spectra indicated that a caesium borate species was dominant. Importantly, this work showed that off-gas Raman spectroscopy can be used to detect differences in caesium volatiles during thermal treatments such as vitrification.
Overall, the NaBS series performed better than the IPG series in terms of caesium retention. However, it would be useful for further studies to compare the two glass systems at similar viscosities rather than arbitrary melting temperatures.
