Vitrification by joule-heating requires suitable electrode materials. Molybdenum electrodes are used often in high temperature (∼1500°C) joule-heated melters producing commercial glass while Inconel 690 has been the material of choice for lower temperature (∼1150°C) high-level nuclear waste vitrification. Vitrification of diverse waste streams at higher-temperatures places increasingly severe demands upon the electrode materials. Some commercially available ceramics possess sufficient conductivity at high temperatures to be used as electrodes. Many mixed- and low-level waste vitrification applications involve significant amounts of reducible species that accelerate the corrosion process for metallic electrodes. In addition, the corrosion rate of all electrode materials depends on the electrode current density. A special test rig was designed that measures the dependence of corrosion rate on both current density and temperature over extended periods of time. Molybdenum, Inconel 690, and three types of commercial ceramic coupons were tested over a range of conditions in simulated waste glasses containing key reducible species (e.g., PbO, CuO, ZnO, SO3). The corrosion rates show strong dependencies on the current density. Coupons were sectioned and the glass-coupon interface examined using Scanning Electron Microscopy (SEM) coupled with Energy Dispersive Spectroscopy (EDS). Cu, Pb, Ni and Sn were found to be reduced to their metallic states at the molybdenum coupon surfaces and were active in the corrosion process.

In many vitrification processes, refractory materials are used to contain the waste glass melt. The corrosive nature of the high-temperature melt consumes the waste feed materials but also limits refractory life. As vitrification is applied to more diverse waste streams, and particularly in higher-temperature applications, increasingly severe demands are placed on the refractory materials. A variety of potential refractory materials including Fused-cast AZS, Monofrax K3, Monofrax E, and the Corhart refractories ER1195, ER2161, C1215, C1215Z, Rechrome, and TI 186, were subjected to corrosion testing at 1450°C using the ASTM C-621 procedure. A series of simulated waste glasses was used which included F, Cl, S, Cu, Zn, Pb; these minor components were found to cause significant, and in some cases drastic, increases in corrosion rates. The corrosion tests were conducted over a range of time intervals extending to 144 hrs in order to investigate the kinetics of the corrosion processes. The change of the concentrations of constituents in the glass was monitored by compositional analysis of glass samples and correlated to the observed extent of corrosion; typically, components of the material under test increase with time while key minor components, such as Cu and Pb, decrease. The rate of corrosion of high-zirconia refractories was slowed considerably by adding zirconia to the waste glass composition; this has the added benefit of improving the aqueous leach resistance of the waste form that is produced.

The hypothesis of French nuclear waste glass disposal in a geological repository implies a comprehensive assessment of all the glass elements liable to participate in controlling the material alteration kinetics. The hypothetical existence of kinetically limiting elements other than silica could account for the observed R7T7 glass behavior in the presence of certain clays, and notably the continued high alteration rates observed even after silica saturation occurs. Flowing experiments with solutions near silica saturation but highly subsaturated with respect to aluminum hydroxide were defined to investigate the possible limiting role of aluminum. Experiments were conducted at different flow rates with the same constant steady-state H4 SiO4 activity for all the tests. The glass dissolution rate was observed to depend on the solution flow rate, indicating that under these conditions the kinetics are not controlled by dissolved silica alone. An additional experiment, in which only the Al(OH)4− activity in solution was allowed to vary, demonstrated the critical role of this element. Several interpretations are discussed.

The SON68 inactive “R7T7” composition is the French reference glass for the LWR nuclear waste glass. Vapor phase alteration was used to accelerate the reaction progress of glass corrosion and to develop the characteristic suite of secondary, alteration phases. Extensive solid-state characterization (AEM/SEM/HRTEM) was completed on six inactive R7T7 waste glasses which were altered in the presence of saturated water vapor (200 °C) for 91, 241, 908, 1000, 1013, and 1021 days. The AEM samples were examined in cross-section (lattice-fringe imaging, microdiffraction, and quantitative thin-film EDS analysis). The glass monoliths were invariably covered with a thin altered rind. The layer became thicker with time: 0.5μm for 22 days; 4 μm for 91 days; 6 μm for 241 days; 10 μm for 908 days; 26 μm for 1013 days; and <35μm for 1021 days. The composite alteration layer of the SON68 samples is at least four time less thick than that of the SRL 131 glass composition.

Six distinctive zones, based on phase chemistry and microstructure, were distinguished within the well-developed surface layers. Numerous crystalline phases such as analcime, tobermorite, apatite, and weeksite were identified on the surfaces of the reacted glasses as precipitates. Two crystalline phases, Ag2TeO3 and (Ca,Sr)Mo3O9(OH)2, were found within the inner zones of surface layers, and they must have nucleated in situ, indicating that Ag, Te, Sr, and Mo can be retained within the surface layer. The majority of the surface layer volume is composed of two morphologically and chemically different structures: one consists of well-crystallized fibrous smectite aggregates occurring along with cavities, the A-domain; and the other consists of poorlycrystallized regions containing needle-like smectite (montmorillonite) crystallites, a silica-rich amorphous matrix, and possibly ZrO2 particles, the B-domain. The retention of rare-earths and Zr mostly occurred within B-domains and that of transition elements, such as Zn, Cr, Ni, and Mn, in A-domains. The recrystallization of poorly-crystallized B-domains into well-crystallized Adomains may influence the long-term behavior of rare-earths, Zr, and transition elements. The mechanism of surface layer formation during vapor phase alteration is discussed based on the cross-sectional AEM studies of surface layers of the SON68 waste glasses.

Results from durability drip tests designed to simulate the unsaturated conditions in the proposed Yucca Mountain Repository are reported for two actinide-doped glasses used as model waste forms. These tests are being conducted with reference glass compositions doped with neptunium, plutonium, and americium from the Defense Waste Processing Facility (DWPF) and the West Valley Demonstration Project (WVDP), and have been ongoing for over 8 years. Solution compositions, including transuranics, have been periodically determined, and selected analyses of colloid formation and composition, glass corrosion layers, and solid alteration phases have been obtained by scanning and analytical transmission electron microscopies. The importance of integrated testing has been demonstrated, as complex interactions among the glass, the groundwater, and the sensitized stainless steel have been observed. The cumulative releases of both glassforming and dopant elements are presented along with identification of reaction phases and their partitioning between solution and solid phases. Alteration phases, including smectite clay, iron silicates, uranium silicates, and calcium thorium phosphate, have been observed forming on the glass and stainless steel and have occasionally been found suspended in solution as colloids. Actinides, except neptunium, concentrate into alteration phases or sorb onto the stainless steel. The subsequent transport of the actinides is then controlled by these phases.

Chemical durability of simulated radioactive low-level waste (LLW) baseline glass, 6- 5412, was found to decrease with the addition of either phosphate or sulfate, or chromium oxide. Scanning electron microscopic study of corroded samples suggests that the effect of these components on glass chemical durability may be attributed to cluster formation of: Na-O-P-O or Na-O-S-O, or (O0)3-Cr3+(O−)3 groups. Preferential corrosion attack was severe on certain corroded sample surface areas that are possibly enriched in these clusters. Durability tests in a solution with LiOH as a buffer suggest that chlorine in the glass slightly improved glass corrosion resistance, while fluorine significantly deteriorated glass corrosion resistance.

Laboratory tests are being conducted using two radionuclide-doped Defense Waste Processing Facility (DWPF) glasses (referred to as SRL 131A and SRL 202A) to characterize the effects of the glass surface area/solution volume (SN) ratio on the release and disposition of Tc and several actinide elements. Tests are being conducted at 90°C in a tuff ground water solution at S/V ratios of 10, 2000, and 20,000 m−1 and have been completed through 1822 days. The formation of certain alteration phases in tests at 2000 and 20,000 m−1 results in an increase in the dissolution rates of both glasses. The release of Tc parallels that of B and Na under most test conditions and its release increases when alteration phases form. However, in tests with SRL 202A glass at 20,000 m−1, the Tc concentration in solution decreases coincidentally with an increase in the nitrite/nitrate ratio that indicates a decrease in the solution Eh. This may have occurred due to radiolysis, glass dissolution, the formation of alteration phases, or vessel interactions. Technetium that was reduced from Tc(VII) to Tc(IV) may have precipitated, though the amount of Tc was too low to detect any Tc-bearing phases. These results show the importance of conducting long-term tests with radioactive glasses to characterize the behavior of radionuclides, rather than relying on the observed behavior of nonradioactive surrogates.

In-situ corrosion tests on nuclear waste glasses in Boom clay provided direct contact glass-clay at 90°C, for periods of 2, 3.5 and 7.5 years. The corroded reference glasses (two R7T7 type glasses, four Pamela type glasses), were studied in terms of SIMS (secondary ion mass spectroscopy) and mass losses.

The Al2O3 rich Pamela glasses appear to corrode in a selective-substitutional way, the other glasses dissolve almost congruently. Differences in the corrosion extent between the glasses are associated with compositional differences and secondary phase formation. SIMS analysis provides the reaction layer thickness and the relative element behaviour in this layer. Although relatively few, the data have provided a coherent picture of glass corrosion, in terms of corrosion mechanisms, time and glass composition dependence.

Mineralogical-geochemical investigation of a sample of nepheline syenite (urtite) as a natural analog of final radioactive waste form has been performed. The specimen of urtite consists of nepheline, alkali feldspar, pyroxene, sphene, apatite and minor magnetite and amphibole. As a first approximation, urtite simulates the mineral composition of waste slag produced in a plasma shaft furnace at SIA “Radon”. Determination of chemical compositions of the minerals by electron-probe microanalysis has shown that the main phases that hosted radionuclides and their geochemical analogs are as follows: nepheline (Rb and probably Cs), feldspar (Ba), sphene (Zr, Nb, REE, and actinides) and apatite (Sr, REE, and actinides).

Fractal structure domains for inhomogeneities in radwaste glasses were determined. SEM and TEM photos were analyzed for this aim as well as the Cs-137 leaching process for glass specimens. Unusual character of leaching was observed provided by anomalous diffusion of radionuclides through fractal cluster made of inhomogeneous particles.

Dissolution of powdered glass from Kilauea volcano, Hawaii (ca 51% SiO2) was studied in a fluidized-bed, flow-through reactor at room temperature in both dilute HCI and organic ligand-bearing solutions (citrate and oxalate) to determine the effects of pH and organic acids on the dissolution rate. Dissolution was non-stoichiometric in both HCI and organic solutions; however, the relative release rates of various ions and the composition of leached layers or secondary phases are fimctions of pH and organic ligand concentration and type. In HCl solutions, the minimum glass dissolution rate, as assessed from the Na leaching rate, was 7.4 × 10−12 gm cm−2 sec−1, comparable with previous results, and was virtually independent of pH. Addition of citrate and oxalate increased the non-stoichiometry of dissolution. At pH 7, the overall rate of glass dissolution decreased (by as much as 5 times) at low ligand concentrations (< 1 mM), but increased by as much as five times at higher concentration (3 mM). High ligand concentrations do increase the release rate of some elements, especially multivalent cations, such as Fe3+ which form strong organic complexes, by as much as 100 times.

Static corrosion tests of a powdered simulated waste glass were performed in the presence of magnetite under oxic and anoxic conditions, respectively. The corrosion tests under oxic conditions were performed in air, and those under anoxic conditions were performed in a glove box purged with mixed gas(Ar+5%H2). The experimental results showed that the presence of magnetite can enhance glass corrosion under both oxic and anoxic conditions, and the enhancement under anoxic conditions was small compared to that under oxic conditions. A large portion of the Si and insoluble elements were sorbed onto the magnetite surface, and the sorption under anoxic conditions was small compared to that under oxic conditions.

It was suggested that the enhancement of glass corrosion by magnetite results from the sorption or precipitation of silica on the magnetite surface, which can be greatly affected by redox condition. Under anoxic conditions, it was suggested that precipitation of amorphous silica on the magnetite surface may be the dominant process for enhancement of glass corrosion in addition to sorption.

Long term tests of vitrified intermediate level waste from nuclear power plants were performed in a shallow ground repository as well as on an open testing area. Borosilicate glasses were tested. Radionuclide leaching differs significantly in natural conditions in contrast to laboratory behaviour. Average leaching factors were obtained for open area tests as well as for repository tests. These factors should be used for long term predictions since they take into account the real character of leaching.

We present a Monte Carlo simulation method for modelling glass dissolution in aqueous solutions. This simulation method is consistent with transition state theory, and therefore also with the glass dissolution rate law, used for instance in the Grambow model. The simulation method allows to add dynamics (kinetics) to the existing thermodynamic models for glass dissolution. Using this method, it is possible to model non stoichiometric dissolution of the glass.

Besides, we introduce a simple, first version of a model in which we use the simulation method. In this model, we approximate the glass by a lattice. We assume that the glass contains two components: a network former and a network modifier. Bonds between two network formers are assumed to be much stronger than any other bond in the system. We observe that above a threshold value for the concentration of network modifiers, the glass dissolves fast. No surface layer develops and the dissolution rate is constant (linear stoichiometric dissolution). Below this threshold, the glass is more durable and surface layers are formed. As time goes on, the thickness of the surface layers grows. The dissolution of the glass is not stoichiometric. This behaviour agrees with experimental results.

In some radwaste applications, such as immobilization of U.S. Tank wastes using vitrification, Tc and Cs/Sr are likely to be separated out from HLW supernates. Simplified Synroc preparations can be devised for the immobilization of separated Tc and Cs/Sr, either together or individually. Under suitably reducing processing conditions, Tc can be immobilized, as metal or Tc4+. The volatility of Tc and Cs/Sr in Synroc processing is very restricted.

In Synroc-C, designed for Purex high-level reprocessing wastes, Tc exists as a metal alloy. In air-saturated water, normalized MCC-1 type total differential leach rates at 90°C decreased with leaching time and after 90 days were <10-4 g/m2/d, decreasing by a factor of ∼100 in anoxic conditions. The corresponding results in a pH=6.1 buffer solution were fairly similar to those in deionised water, but in anoxic conditions, the leach rates were higher in the buffer solution than in the deionised water.

A single perovskite phase was loaded with ∼40 wt% of TcO2 to form CaTc0.5Ti0.5O3 by a combination of graphite-die hot-pressing and heating in argon. Rutile, intended to be doped with ∼35wt% TcO2, was fabricated by bellows hot-pressing at 1200°C; aproximately 75% of the Tc formed a solid solution with rutile, but some metallic Tc was also present due to imperfect redox control. Results of both MCC-1 and PCT-type leach tests are presented on a preparation containing 70 wt% hollandite + 20 wt% perovskite + 10 wt% rutile containing about 5 wt% of Cs and Sr respectively. Phase distributions resulting from incorporating Cs and Sr in Synroc-B precursor were studied.

For the HLW fractions of certain Hanford tank wastes, wasteforms consisting mainly of crystalline Synroc phases in an aluminosilicate matrix can be formed by melting at 1350–1450°C in a neutral atmosphere, with waste loadings of up to 70 wt%, depending on the waste composition. The leachability of all measured elements in 7-day PCT tests was <0.1 g/m2/d at 90°C in the best materials.

The titanate based nuclear waste immobilization medium, Synroc-B, has been synthesized at atmospheric pressure from freeze dried nitrate precursors. Complete formation of the phase assemblage (CaTiO3, CaZrTi2O7, and BaAl2Ti5O14) occurred upon calcination of the nitrate precursor after only 10 minutes at 1100°C. This improvement in the preparation conditions may lead to practical application of the material in the safe disposal of high level nuclear waste and the immobilization of other strategic nuclear materials.

We investigated the interaction between Synroc-C and deionized water or Boom clay disposal media. We used powdered Synroc-C to achieve high SA/V (surface area to volume) conditions (100, 1000, 10000 m-1). The temperature was 90°C. Reaction progress up to 10+6 days/mi was reached. We conclude that dissolution in DW is mainly controlled by initial ion exchange, followed by matrix dissolution. In both Boom clay media (a 500 g/l and a 2000 g/l clay/claywater mixture) matrix dissolution is dominating. The depletion depth of the main Synroc constituent Ti is below 250 nm in the clay media, and below 2 nm in DW after 110 days corrosion at 100 m-1. The corrosion rates are very small, though we cannot present meaningful values. The effect of Boom clay is mainly to increase the solubility of Ti, Zr and the rare earths in solution.

Np-doped Synroc was subjected to modified MCC-1 leach tests in demineralised water at 70°C. Normalised total differential Np leach rates were about 10−4 g m−2 d−1 initially, decreased to about 10−5 g m−2 d−1 after 60 days and stayed approximately constant for periods totalling up to 2283 d. Scanning electron microscope (SEM) and analytical transmission electron microscope (AEM) examination of discs leached for periods totalling 56 d or longer showed that they were covered with crystalline anatase, brookite and ilmenite. The crystalline secondary phases initially nucleated in polishing scratches and on perovskite grains. However, after leaching for a total of 422 d most of the surface was covered. None of the secondary phases contained observable amounts of Np (< 0.2 wt%). These data are discussed in relation to previous work.

A ceramic waste form based on Synroc-D is under development for the incorporation of the mineral residues from molten salt oxidation treatment of mixed low-level wastes. Samples containing as many as 32 chemical elements have been fabricated, characterized, and leach-tested. Universal Treatment Standards have been satisfied for all regulated elements except two (lead and vanadium). Efforts are underway to further improve chemical durability.