Thermochemical Modeling of Glass: Application to High-Level Nuclear Waste Glass

• Published in: MRS bulletin 1999, Vol.24 (4), p.37-44
• Main Authors: Spear, Karl E., Besmann, Theodore M., Beahm, Edward C.
• Format: Article
• Language: English
• Subjects: Computer Simulations from Thermodynamic Data: Materials Production and Development; Radioactive wastes
• ISSN: 0883-7694; 1938-1425
• DOI: 10.1557/S0883769400052179

Despite the obvious importance of understanding the chemistry of oxide glass materials, predictive thermochemical modeis of complex glasses have not yet been developed. Such modeis are important for technologies such as the disposal of high-level nuclear and transuranic waste (HLW), which are currently fore-seen as being incorporated in a host glass for permanen t Sequestration. A large number of glasses have been explored, with a borosilicate glass being the typical base composition. An example of the complexity of such a HLW glass is given in Table I. This article discusses our at-tempts to develop an accurate, easy to understand and use glass Solution model for describing the thermodynamic stability of such HLW glasses. Critical for such a model is the availability of reliable thermodynamic data that can be used in generating accurate values for thermodynamic activities of glass components as a function of temperature and glass composition. Therefore, a major part of this article focuses on developing reliable sets of thermodynamic data for complex HLW glass Systems and Subsystems. With such Information and a model, we can make predictions of the stability of these waste forms, including their volatility, leaching behavior, and corrosion reactions, and understand crystallization behavior during both the initial glass processing and long-term storage. Using an equilibrium thermodynamic model is offen questioned, since HLW is to be stored as part of a glass phase, and glass is a nonequilibrium material. Our model uses a pseudoequilibriu map-proach in which we thermochemically treat the glass as a supercooled liquid. This is a more accurate approach than assuming a global System equilibrium, as it describes the behavior of the metas-table glass phase using thermodynamic data for the liquid phase and excludes the formation of crystalline species. As a result, developing an accurate model and data for representing the thermodynamic properties of oxide liquid phases is critical to understanding the limiting chemi-cal behavior of the nuclear waste glass. The methodology requires that a critically assessed thermodynamic database be created for binary and ternary combinations of the major constituents in a typical waste glass. These data can then be combined to represent the thermodynamic behavior of the more complex multicomponent HLW glass Systems. If a crystalline phase is experimentally observed to precipitate from the glass under certain conditions, a therm