The investigations on enhanced reprocessing of nuclear spent fuel, and notably on separating the long-lived minor actinides, such as Am and Cm, from the other fission products have led to the development of highly durable specific matrices such as glass-ceramics for their immobilization. This study deals with the characterization of zirconolite (CaZrTi2O7) based glass-ceramics synthesized by devitrification of an aluminosilicate parent glass. Trivalent actinide ions were simulated by neodymium, which is a paramagnetic local probe. Glass-ceramics with Nd2O3 contents ranging from 0 to 10 weight % were prepared by heat treatment of a parent glass at two different growth temperatures: 1050° and 1200°C. X-ray diffraction (XRD), energy dispersive X-ray analysis (EDX) and electron spin resonance (ESR) measurements clearly indicate that Nd3+ ions are partly incorporated in zirconolite crystals formed in the bulk of the glass-ceramic samples. The amount of neodymium in the crystalline phase was estimated using ESR results and was found to decrease with increasing either heat treatment temperature or total Nd2O3 content.

Zirconolite (CaZrTi2O7) based glass-ceramics designed for the specific immobilization of plutonium wastes or minor actinides (Np, Am, Cm) from high level radioactive wastes were investigated. To reach an efficient double containment, actinides must be preferentially located in the crystalline phase, which is homogeneously dispersed in a calcium aluminosilicate residual glass. Several heat treatments (between 950° and 1350°C) of a parent glass belonging to the SiO2-Al2O3-CaO system and containing TiO2 and ZrO2 were performed to prepare glass-ceramics. Trivalent minor actinides were simulated introducing Nd2O3 in the glass composition. Electron microscopy, X-ray diffraction (XRD) and thermal analysis have shown that devitrification processes in the bulk and on glass surface are different. They lead to the crystallization of zirconolite in the bulk and to a mixture of titanite (CaTiSiO5) and anorthite (CaAl2Si2O8) near the surface. For heat treatment temperatures greater than or equal to 1250°C, baddeleyite (m-ZrO2) crystals form at the expense of zirconolite in the bulk of glass-ceramics. XRD indicates that the order in zirconolite Ca/Zr planes increases with heating temperature. At the same time, extended defects density decreases.

Several glasses were melted using laboratory-scale microwave (MW) equipment to study the vitrification of Pu-containing sludges from the Mining and Chemical Combine in Zheleznogorsk, Russia. This work is part of an effort to identify glass compositions that can be adopted for industrial scale production. The glass must meet durability requirements and have good technological properties for long-term storage and geologic disposal.

Electron Probe Analysis (EMPA), scanning electron microscope (SEM) and x-ray methods were used to characterize glasses containing 50 wt% solid sludge residues with high sodium content.

Leach tests of 3, 7, 14, and 28 day durations were performed using the MCC-1 method at 90°C. The surface of the glass was analyzed before and after leaching. Based on the observed results, there are differences in the behavior of Pu and U for the glasses tested, and glass durability is dependent on homogeneity, the uranium oxidizing state, and the SiO2/Na2O+Li2O ratio.

A Gd-rich crystalline phase precipitated in a sodium gadolinium alumino-borosilicate glass during synthesis. The glass has a chemical composition of 45.4-31.1 wt% Gd2O3, 28.8-34.0 wt% SiO2, 10.8-14.0 wt% Na2O, 4.3-5.9 wt% Al2O3, and 10.8-14.9 wt% B2O3. Backscattered electron images revealed that the crystals are hexagonal, elongated, acicular, prismatic, skeletal or dendritic, tens of μm in size, some reaching 200 μm in length. Electron microprobe analysis confirmed that the crystals are chemically homogeneous and have a formula of NaGd9(SiO4)6O2 with minor B substitution for Si. The X-ray diffraction pattern of this phase is similar to that of lithium gadolinium silicate apatite. Thus, this hexagonal phase is a rare earth silicate with the apatite structure. We suggest that this Gd-silicate apatite in a Gd-borosilicate glass is a potential glass-ceramic nuclear waste form for actinide disposition. Am, Cm and other actinides can easily occupy the Gd-sites. The potential advantages of this glass-ceramic waste form include: 1) both the glass and apatite can be used to immobilize actinides, 2) silicate apatite is thermodynamically more stable than the glass, 3) borosilicate glass-bonded Gd-silicate apatite is easily fabricated, and 4) the Gd is an effective neutron absorber.

Experimental alteration results are presented in which a French SON 68 (R7T7-type) nuclear containment glass specimen was first altered under static conditions for 600 days at an S/V ratio of 5000 m−1 before transferring it to pure water. The experiment was designed to assess the diffusion barrier properties of the gel formed during the preliminary alteration phase. Contrary to predictions by kinetic models based only on chemical affinity, the renewed alteration of the specimen in pure water was very limited. Measurements at close intervals showed that the maximum alteration rate under these conditions was 7 ×10−3 g·m−2d−1, or about r0/300, whereas a pristine glass coupon in contact with pure water is altered at r0.

The renewed alteration behavior is attributed to partial (5–10%) dissolution of the existing gel, resulting in a slight degradation of its protective properties. Saturation conditions, at a different level than in the preliminary phase, were reached within a few days. The saturation of the solution with respect to silicon, generally interpreted as a glass/solution (Grambow) or gel/solution (Bourcier) equilibrium, is shown to depend not only on the alteration conditions, but also on the specimen leaching history, and is thus not an inherent glass property.

This experiment confirms the idea that the gel formed under saturation conditions controls the kinetics of SON 68 glass alteration by means of a diffusion barrier effect. It also raises a number of issues concerning the concept of a “residual rate” and the possible relations between the quasi steady-state Si concentrations observed in solution and the protective properties of the gel.

Most nuclear glass alteration models are based on a first-order kinetic law of the general type r = r0 (1–Csi/C*) or r = r0 (1–Q/K). It is generally assumed that the establishment of quasi steady-state concentrations in solution corresponds to an intrinsic glass solubility limit that may be expressed either in terms of the dissolved silicon concentration C* or as an ion activity product at saturation K. Experimental research over the last five years in France has shown that the quasi steady-state concentrations observed in solution do not correspond to a thermodynamic solubility limit with respect to the glass. This assertion is supported by a large body of experimental evidence that is discussed in this paper. The signification of the C* parameter is therefore reconsidered, and its implications on long-term behavior modeling are discussed.

The effect of γ-radiation on the dissolution of candidate high-level waste glasses was investigated in potential disposal environments at 90°C. The media consisted of mixtures of Boom Clay, bentonite clay and cement with clay water. During the experiments the pH of Boom Clay decreased, probably mainly by radiolytical oxidation of pyrite. The addition of bentonite, cement and glass buffered the pH decrease. Under radiation the glass mass losses decreased, whereas the leach rate of soluble elements was not influenced or appeared to increase. This is explained through the radiolytical acidification, and possibly by bubble formation in the glass.

One presents the results of numerical simulations of glass leaching. The glass is modelled as a random mixture of partly and totally soluble species, which represent silica, and boron or alkali oxides, respectively. It is shown that the dissolution rate and the thickness of the altered surface layer are strongly dependent on the glass composition, whereas the equilibrium solubility is not. The dependence of the layer thickness on the glass surface area to solution volume ratio is also emphasized. The protective role of the surface layer is shown to arise from its restructuring after the extraction of the soluble species. The simulation results are compared to an experimental study performed on series of SiO2-B2O3-Na2O glasses.

Zirconolite-glass and sphene -glass specimens, doped with REE as simulants for trivalent actinides, were leached under two conditions. The first was a solubility test using a powdered sample in deionized water at an SA/V ratio of 200 cm−1, to examine the long-term leaching behavior of the composite materials. The other test was carried out in the presence of moist clay, to assess the degree of surface alteration of the composites in the presence of potential geological repository materials. Both tests were carried out at 90°C.

As the level of Pu4+ substituted on the Zr-site in CaZr1-xPuxTi2O7 zirconolite increased, from x=0.1 to 0.6, a series of structural transitions occurred from zirconolite-2M to zirconolite-4M and subsequently from zirconolite-4M to pyrochlore. The solid-solution limit for Pu4+ substituted on the Zr-site in zirconolite-2M was ~ 0.15 formula units. Zirconolite-4M was only stable over a narrow compositional range, centered about CaZr0.59Pu0.41Ti2O7, whilst the pyrochlore structure was stabilized with CaZr0.4Pu0.6Ti2O7 stoichiometry. The stability of the zirconolite polytypes is therefore sensitive to the average effective ionic size of the ions occupying the seven-coordinated Zr-site. The reduction in Pu from Pu4+ to Pu3+ destabilized the zirconolite-4M, producing a mixture of perovskite and possibly zirconolite-3T. The CaZr0.4Pu0.6Ti2O7 pyrochlore was also predominantly transformed to perovskite as a result of this reduction of Pu.

Synthetic zirconolite (CaZrTi2O7) doped with rare earth elements (REEs) and Hf has been subjected to corrosion tests in a closed system at elevated temperature and pressure in fluids with various compositions. Together with previous studies, the results indicate only a very weak corrosion below 250°C at a pressure of 50 MPa. Above that temperature and up to 500°C zirconolite suffers from relatively rapid corrosion and, depending on the fluid composition, it may be covered by various secondary phases. Above 500°C in Na-rich fluids, zirconolite is replaced by perovskiteand calzirtite (nominally CaTiO3 and Ca2r5Ti2O16, respectively), but theREEs and Hf (acting as actinide analogues and/or neutron absorbers) are almost quantitatively incorporated into the secondary phases. The breakdown of zirconolite and its replacement by other phases in the laboratory tests are comparable to reactions observed in natural systems. Additional experiments with U-doped zirconolite revealed differences in the behavior of the used actinide analogues (Nd, Ce, Gd)and U during corrosion. The results of this study, together with observations on natural samples, strongly support the use of zirconolite-based ceramic waste forms for actinide-rich wastes.

The heat treatment conditions are a key factor in fabricating zirconolite ceramics and glass-ceramics following high-temperature melting. An oxide mixture melted at 1450°C and subsequently heat-treated at 1200°C yielded a glass-ceramic containing crystallized zirconolite–2M. The silica-enriched residual glass represented about 60-70 vol% of the total; the actinide surrogates (Nd, Ce)were equally distributed between the residual glass and the zirconolite crystals. Zirconolite ceramics obtained after melting an oxide mixture at 1600–1700°C consisted of zirconolite, perovskite and rutile. Rapid cooling rates (> 100°#x00B0;··min-1) were obtained by pouring the melt into ingot molds; the resulting zirconolite ceramics were characterized by crystals of zirconolite-2M ranging from 1 to no more than 20 μm. Slow cooling (< 25°C#x00B0;··min-1 produced ceramics with crystals several hundred micrometers long. Despite the microstructural differences, the chemical durability of the zirconolite ceramics was identical. The initial alteration rates r0 were about two orders of magnitude lower than those measured for the residual aluminosilicate glass of the zirconolite glass-ceramics. Moreover, during long-term leach tests at high S/V ratios to obtain advanced degrees of reaction progress, the alteration rates of all the materials diminished by over 3 to 4 orders of magnitude below r0.

Static dissolution tests on Nd-bearing zirconolite were conducted in deionized water at 150°C for up to 6 months. Surfaces, both before and after aqueous dissolutions, were examined using X-ray Photoelectron Spectroscopy (XPS) and Scanning Electron Microscopy (SEM). Individual submicron-sized crystals wereobserved only on some zirconolite grains after hydrothermal treatment for one week, and were identified as brookite (possibly plus anatase) by Transmission Electron Microscopy (TEM). The number of secondary crystals present on the zirconolite surface was, however, reduced significantly after 6 months of durability testing. The results of electron microscopy studies are consistent with those obtained from XPS in that the average Ca and Ti concentrations of the top surface layer (< 5 nm) decreased slightly with respect to Zr, while the average Al concentration increased after durability testing.

Cesium, iodine and strontium ions have been implanted into yttria-stabilized cubic zirconia (YSZ) to determine the effects of fission product incorporation in YSZ that is considered as an inert nuclear fuel matrix. The ion implantation was conducted at room temperature to 1 × 1021ions/m2 for each ion with ion energies ranging from 70 to 400 keV. The peak displacement damage level and the peak ion concentration in YSZ reached 205-330 dpa and 11-26 at%, respectively. The microstructure of the implanted YSZ was studied by in situ and cross-sectional transmission electron microscopy. In the iodine and strontium implanted samples, a damaged layer with a high density of defect clusters was observed, while in the cesium implanted specimen, the damaged layer is amorphous. Nanocrystalline precipitates were observed in the strontium implanted specimen after annealing at 1000°C. The results are discussedin terms of the ionic size, mobility and the solubility of the implanted species in YSZ.

Cerium is a well-known surrogate element for Pu that is widely used in experiments on the synthesis and study of ceramic waste forms for actinide immobilization. Cathodoluminescence (CL) images and emission spectra from synthetic zircon, (Zr,…)SiO4, cubic zirconia, (Zr,Y,Gd,…)O2 and pyrochlore, Ca(Hf,Gd,…)Ti2O7, individually doped with Ce and Pu were investigated. It was determined that some of the Ce, which was added as Ce4+ in the form of CeO2, was converted to the 3+ valence state during synthesis of zircon-based ceramics by sintering in air and in argon gas. However, Pu-doped zircon synthesized in air under the same conditions does not exhibit the CL emission of Pu3+. Analysis of CL spectra of cubic zirconia doped with Ce as well as Pu suggests that both ions are in the tetra-valent state in these synthetic materials. However, the CL peak of Ce-doped zirconia is almost identical to that of pure zirconia (2.6 eV). The CL peak for Pu-doped zirconia is shifted to 2.4 eV. No CL emission of Ce3+ or Pu3+ was observed in pyrochlore samples. The CL spectra for Ce-and Pu-doped pyrochlore samples were almost identical. We suggest that under oxidizing conditions Ce is good surrogate of Pu for experiments on the synthesis of pyrochlore-based and acceptable enough for zirconia- based ceramics. The use of Ce as Pu surrogate for zircon-based ceramics should be considered as limited.

Zircon, ZrSiO4, as well as its Hf-analogue hafnon, HfSiO4, have been proposed for use as durable Pu host phases for the immobilization of weapons grade Pu and other actinides. Four samples of Pu-doped ceramics based on the zircon and hafnon structures were synthesized through sintering in air using precursors containing 5-6 and 10 wt% 239Pu. Synthesized ceramic samples were studied by X-ray powder diffraction (XRD), scanning electron microscopy (SEM), microprobe method, MCC-1 leach test at 25 and 90°C. Inclusions of separated a PuO2phase in the matrix of zircon-based ceramic and presumably, (Pu,Hf)O2 phase in the hafnon-based ceramic were observed for samples obtained from precursors doped with 10 wt% Pu. No separated Pu-phases in significant amounts were identified in the matrices of both ceramics obtained from the precursors doped with 5-6 wt% Pu. It was found that normalized Pu mass losses (without correction on ceramic porosity) for samples doped with 10 wt% Pu which contain separated inclusions of PuO2 or (Pu,Hf)O2 after 14/28 days were approximately (in g/m2) - for zircon: 0.2/0.2 - at 90°C and 0.03/0.04 - at 25°C and for hafnon: 0.02/0.04 - at 90°C and 0.01/0.01 - at 25°C. The losses of Pu from samples doped with 5-6 wt% are 1-2 order of magnitude less. It was suggested that optimal amount of Pu which could be incorporated by zircon and hafnon lattices does not exceed 7 wt%. An important additional conclusion is that Pu- doped ceramic based on zircon or hafnon can be successfully fabricated excluding hot pressing method.

Three melted samples of pyrochlore-brannerite-based ceramics produced under different redox conditions were examined. Two of the samples were produced using cold crucible melting at ~1600 C. The third sample was obtained via melting in a microwave oven at 1700-1800 C. All the samples are composed of major pyrochlore and brannerite phases, and minor rutile and UO2-based solid solution or pseudobrookite phases. Pyrochlore-structured phases predominate in all three samples and account for 50-60% of the total bulk. Two pyrochlore varieties – Ca-pyrochlore (predominant) and Ba-pyrochlore have been found in these samples. The latter phase is more stable at high temperatures than the Ba-hollandite present in sintered pyrochlore-rich Synroc-F ceramics. Decomposition of the Ba-hollandite results in rutile formation in the melted samples.

The chemical durabilities of two Pu-doped pyrochlore samples were studied by Single-Pass-Flow-Through (SPFT) tests at 70°C. The dissolution of pyrochlore is incongruent with preferential releases of Ca and Gd over Ti, close to stoichiometric releases of U and Ti, and lower releases of Hf and Pu than Ti. Altered pyrochlore and polymorphs of TiO2 (brookite and probably anatase) have been identified on the surface of the leached sample and the principal secondary phase is an unknown polymorph of TiO2 containing Hf and varying amounts of Gd and Pu. These surface alteration phases are consistent with reported studies of natural samples. The releases of U, Gd, Ca and Ti into solution follow linear kinetics, whereas the releases of Pu and Hf exhibit non-linear behavior. The presence of ∼5% PuO2 and trace amounts of glass does not appear to have an effect on the overall durability of the material. Further, the low Pu release rate and the similar kinetics for Pu and Hf releases limit the possibility of nuclear criticality under repository conditions. Overall, this study provides useful information on the lower bounds of durabilities of the materials.

Six different formulations of pyrochlore-rich synroc waste forms were designed to contain 32-50wt% simulated actinides. Their physical properties were determined. Phase identification was carried out by using X-ray diffractometer (XRD), and backscattered electron image (BEI) in scanning electron microscopy (SEM). Phase microanalyses were carried out using energy dispersive spectrometer (EDS) in SEM. Their durability properties were determined by PCT leach tests, the leachate was analyzed by using inductive coupled plasma mass spectroscopy (ICP/MS) and atomic absorption spectroscopy (AAS).

Brannerite, as a minor phase, exists in the pyrochlore-rich titanate ceramic formulations designed for immobilization of surplus weapons Pu. The dissolution of synthetic brannerite was studied at 90°C using static tests in pH 4 solution, deionized water and Finnsjön synthetic groundwater.

After 140 days the normalized U release rates into a pH 4 solution and deionized water reach similar values, ~10−2 g m−2 d−1, and are about 2 orders of magnitude higher than those in Finnsjön synthetic groundwater. The normalized Ti release rate into Finnsjön synthetic groundwater is about an order of magnitude higher than those in pH 4 solution and deionized water. The dissolution of brannerite is incongruent in the pH 4 solution and deionized water (preferential release of U over Ti), and nearly congruent in Finnsjön synthetic groundwater. SEM observations of the samples after 140 days in pH 4 solution and deionized water revealed minor surface alteration, in the form of a thin surface layer, probably TiO2, as a result of preferential releases of U in both cases.