Glass matrices for immobilizing nuclear waste containing molybdenum and phosphorus

https://doi.org/10.1016/j.jnucmat.2007.11.013Get rights and content

Abstract

Vitrification has been selected in France as the process for immobilizing high-level waste arising from spent fuel reprocessing. Some high-level solutions generated by reprocessing legacy fuel contain high molybdenum concentrations. Molybdenum is known to be sparingly soluble in conventional borosilicate glass, and work is in progress to find suitable glass formulations for such waste. The results of a basic study to identify borosilicate glasses composition zones of potential interest are discussed. A vast composition range was investigated by defining a fine mesh. The limits considered to delimit the range of the study were intentionally extended to identify formulations such as SiO2–B2O3–Al2O3–Na2O–P2O5 that are of interest for vitrifying molybdenum-rich waste. Observation of more than 50 tested mixtures revealed two composition zones of potential interest. One forms a homogeneous glass after melting at 1300 °C and rapid cooling; the other vitreous material comprises unconnected microbeads uniformly dispersed in a borosilicate glass.

Introduction

Nuclear waste containment in glass has several advantages: it guarantees the stability of the waste package over very long time periods and reduces the waste volume. Vitrification has been used for several decades in France and other countries to immobilize high-level nuclear waste. The first difficulty in fabricating a package is to find glass formulations capable of chemically incorporating the minerals contained in the waste. The glass formulations must be compatible with synthesis by vitrification processes suitable for industrial production in a radioactive environment, and must also produce a matrix that meets long-term behavior requirements. Borosilicate glass compositions are particularly suited to these requirements.

Among the chemical elements in reprocessing waste, molybdenum is one of the most difficult to incorporate in borosilicate glass. In fission product solutions generated by reprocessing UMo fuel, which was used in the past in gas-graphite reactors, molybdenum is the major chemical element in the waste, and is accompanied by phosphorus, also at high concentrations. In order to develop a glass formulation for this wasteform, it is therefore of particular interest to investigate the SiO2–B2O3–Na2O–Al2O3–P2O5–MoO3 phase diagram in greater detail. This study focuses on composition ranges with potential for containment of nuclear waste rich in molybdenum and phosphorus.

Section snippets

State of the art of molybdenum immobilization in glass

The results obtained to date with borosilicate glass compositions such as those generally formulated for nuclear waste containment show that separate molybdate phases appear if the molybdenum concentration exceeds a few percent. The maximum MoO3 concentration in a silicate glass appears to be 4 wt%; this limit has been reached in R7/T7 glass samples at laboratory scale [1]. This solubility limit could be slightly increased by fabricating the glass under more reductive conditions than usual.

Glass composition range investigated

For the purposes of this study, a tradeoff had to be found between the potential number of chemical elements and their concentration ranges and a reasonable number of glass compositions to be fabricated. We considered six elements: SiO2, P2O5, B2O3, Al2O3, MoO3, Na2O.

The weight percentages of MoO3 and Na2O were specified as 10 and 12 wt%, respectively, for all the compositions. A mass fraction of 10% MoO3 in a nuclear containment glass is well within the molybdenum-rich glass range. A fixed Na2O

Experimental procedure

As the vitrification behavior of these mixtures was difficult to predict in the absence of prior studies covering this composition range, we decided to synthesize 52 compositions uniformly distributed over the range. After heating the constituents, the resulting mixtures were visually examined to identify zones within the range that could be promising for vitrifying molybdenum-rich waste.

Table 1 shows the mixture compositions (in oxide wt%). The precursors used were powder samples of oxides (SiO

Results

After cooling, all the samples formed a massive vitreous material. The cooled samples were classified by visual observation into four categories:

  • homogeneous transparent samples,

  • homogeneous transparent samples with a few beige inclusions,

  • opaque beige samples,

  • samples comprising two superimposed layers.

All the mixtures were observed to be melted on removal from the furnace. In the samples consisting of two layers after cooling, the phase separation was observed in the molten state on removal from

Discussion

All these results confirm the difficulty of loading high molybdenum concentrations in borosilicate glass. The compositions studied exhibited a strong phase separation tendency despite heat treatment at 1300 °C. However, significant differences were observed within the composition range: some compositions yielded stratified samples on removal from the furnace while others were fully homogeneous. The homogeneous glasses are situated in the composition ranges with the highest alumina concentrations

Conclusion

Composition ranges of interest for the containment of nuclear waste rich in molybdenum and phosphorus were identified. A vast composition range was investigated by defining a fine mesh. The limits considered to delimit the range of the study were intentionally extended to identify formulations such as SiO2–B2O3–Al2O3–Na2O–P2O5 that are of interest for vitrifying molybdenum-rich waste. The relevance of these limits was confirmed by the fact that compositions near the limits result in mixtures

Acknowledgement

The authors gratefully acknowledge the support and cooperation of AREVA NC for this research.

References (14)

  • R.J. Short et al.

    J. Nucl. Mater.

    (2005)
  • O. Deruelle et al.

    J. Non-Cryst. Solids

    (2000)
  • M.J. Toplis et al.

    J. Non-Cryst. Solids

    (1998)
  • O. Pinet et al.

    J. Non-Cryst. Solids

    (2006)
  • R. Do Quang, V. Petitjean, J.F. Hollebecque, O. Pinet, T. Flament, A. Prod’homme, in: Waste Management 2003 Conference,...
  • A. Horneber, B. Camara, W. Lutze, in: Proceedings of the Material Research Society, Fifth International Symposium on...
  • O. Pinet, E. Baudrey, C. Fillet, J.L. Dussossoy, J.F. Hollebecque, in: Glass Technology Volume of the XIX International...
There are more references available in the full text version of this article.

Cited by (45)

  • Chemical models of molybdenum-calcium phosphate glasses

    2023, Journal of Non-Crystalline Solids
View all citing articles on Scopus
View full text