The Role of Cements in Radioactive Waste Immobilization: Chemistry and Physics

Portland cements are used to immobilize low-level radioactive wastes and have several advantages over other matrix materials. They are tolerant of wet wastes, are nonflammable and inert and have a well proved process technology. Moreover. they are reasonably durable in underground repositories and, of course, are compatible with concrete vaults, backfills, etc. The short-term immobilization potential is well documented from leach test studies. However, it is not clear how the future performance of cement based immobilization systems can be extrapolated over several centuries, based on leach tests of relatively brief duration (1-10 years). Therefore, more science based studies of the immobilization mechanisms have been supported. Cements are intrinsically microporous materials and the pores are normally water filled. This so-called pore water is chemically active, so that aqueous solution concepts can be applied to cements. Characterization of the coexisting solids and pore fluids are described: the main bonding phase is a calcium silicate hydrogel, abbreviated as C-S-H. The Ca : Si molar ratio of the C-S-H gel can range between about 1.7 to 0.9: it is 1.7 in Portland cements, but reactive blending agents such as fly ash or slag may give a lower ratio C-S-H product. The pore fluid p H is normally in the range 12.5-13.0, but as soluble alkali is leached, the p H is increasingly buffered by the C-S-H. Computer-based speciation codes are being modified to accept data on solubility and speciation in the CaO-Si02H 2 0 system, as well as solubility data for part of the Ca0-Na20-Si02-HZ0 system. The problem of developing models is not simply one of adding more data; C-S-H is incongruently soluble and most standard geochemical programmes permit only congruent solubility. The redox potential function (Eh) of cement pore fluids is examined. Most ordinary Portland cements are slightly oxidizing, with Eh values in the range from +lo0 to +200 mV. However, blends containing slag cements may exhibit much lower E h values, ranging as low as -400 mV. The lowered E h is considered to arise as a consequence of the S*content of slags. The nature of interactions between cement systems and radwaste constituents is analysed. Sorption on the high surface area C-S-H is important, e.g., by iodide. However, such sorption can be highly metastable. Some of the more crystalline hydration products exhibit weak structural incorporation, e .g . , of ions XO3"and XOJin the calcium aluminate sulphate hydrate phases, of which 103is an example. The high pH of cements also leads to immediate precipitation of many cationic species, e.g., U6f species. However, in the longer term, the initially formed hydrous oxide precipitate reacts with cement components: in the still longer term, uranium forms crystalline uranophane, etc. This leads to a slow but steady decrease in solubility as, increasingly, solubility is controlled by thermodynamic stable phases rather than by metastable precipitates. An experimental approach can determine the fate of radwaste constituents in cement. Once specific mechanisms are identified, the appropriate physicochemical data can be obtained and, in conjunction with models of cement performance, used 'to explore quantitatively the flux of radwaste species from a repository. The application of specialized data bases, in conjunction with geochemical modelling codes, are likely to give reliable predictions of the future performance of radioactive waste species immobilized in cements.