Chlorite is an Fe(II)-containing phyllosilicate which is often present as a fracture filling mineral in e.g. granitic rocks. It may therefore be significant in influencing redox conditions and sorption processes in granitic groundwaters. The sorption properties of chlorite may therefore be important when modelling the migration of radionuclides under reducing conditions around nuclear waste repositories or in sites contaminated by mining waste.

The sorption behaviour of Ni(II) onto a natural chlorite (Karlsborg, Sweden) was investigated using a batch technique. The effects of three different background electrolyte concentrations (0.01 M, 0.1 M and 0.5 M NaClO4), different pH values (ranging from 4 to 11) and different Ni(II) concentrations (10−6 and 10−8 M) were studied under anoxic conditions in a glove-box. Ni(II) solutions were spiked with 63Ni and β-Liquid scintillation counting (LSC) was used to determine the concentration of nickel in the bulk solution, allowing the calculation of solid-water distribution coefficients for the metal ion.

The results of the sorption experiments show strong pH dependence at pH > 5, but the sorption is independent of ionic strength. The maximum adsorption is found in the pH range between 7 and 11 with Kd values ≍103 cm3/g. A diffuse double layer model has been used to describe the experimental results.

This article reviews current knowledge of the long-term behavior of R7T7-type glass during the thermal phase and in geological repository conditions (aqueous alteration). In interim storage R7T7 glass can be considered to conserve its integrity over time. In geological repository conditions, the mechanisms of glass alteration by water have been identified and parameter values have been assigned to the reaction kinetics for wide variations of the influential factors (temperature, pH, flow rate, S/V ratio, etc.). CEA has developed an operational model to obtain robust and reasonably conservative predictions of the glass quantities altered after disposal. Examples of applications of the operational model are discussed, future research topics are also proposed to consolidate this approach.

Radionuclide concentration limits are important parameters in performance assessments for geologic disposal of nuclear waste, but their statistical distributions are challenging to define. Thermodynamic solubilities provide an attractive theoretical constraint, but solubilities do not provide a basis for distribution functions, and concentration limits can exceed solubilities in irreversible systems. Distributions of natural concentrations are broad and do not correspond to distributions of concentration limits. Interpretations of natural nickel and lead concentrations suggest that they do not represent concentration limit distributions. Interpretation of groundwater data for dissolved calcium indicates a bimodal distribution, which hypothetically corresponds to a solubility-limited population and a population of values below the solubility limits.

Leaching experiments were done on UO2 pellets doped with alpha-emitters (238/239Pu, 1500 - year batch), in the presence of an external gamma irradiation source (A60Co = 260 Ci, γ dose rate = 650 Gy h−1). The effects of α and γ radiation and the nature of the cover gas (air or Ar + 4% H2) on water radiolysis, and on oxidizing dissolution of the UO2 matrix, are quantified and discussed. The cover gas clearly has a major role in the effect of γ radiolysis. The dissolution rate in an aerated medium is 83 mg m−2 d−1 compared with only 6 mg m−2 d−1 in Ar + 4% H2. The rate drop is accompanied by a reduction of about four orders of magnitude in the hydrogen peroxide concentrations in the homogeneous solution. The UO2 alteration rate under gamma radiation in Ar + 4% H2 is comparable to the rate obtained in the presence of alpha radiolysis alone for the doped UO2 pellets with the highest alpha activities (4.75 × 108 Bq/g UO2, 15 - year batch).

The kinetics of H2O2 reduction have been studied in slightly alkaline (pH 9.7 NaCl) solution on 1.5 at.% SIMFUEL. Using cyclic voltammetric techniques, we have shown that the cathodic reduction of H2O2 is kinetically facile on UIVUVO2+x surfaces. Carbonate ions are found to have a significant effect on the kinetics of H2O2 reduction. Suppression of the reaction rate is observed at very cathodic potentials, and has been attributed to a competition between carbonate and hydrogen peroxide for catalytic sites on the electrode surface. A small enhancement of the reduction current occurs between –200 and +100 mV vs. SCE, which appears to be due to a surface adsorbed carbonate complex. The large values of the Tafel slopes for H2O2 reduction have been interpreted in terms of a chemical-electrochemical mechanism involving surface UV species. The results are discussed in terms of a mixed potential model for the prediction of nuclear fuel dissolution rates under permanent disposal conditions.

In subsurface environments, colloids may play a role in the transport of contaminants to the biosphere because of their high capacity to adsorb them. The electrostatic interactions between colloids and the heterogeneous rock surfaces will determine the fate and transport of both colloids and contaminants adsorbed on the particles. Thus, the aim of the present work was to study the electrostatic interactions between colloidal particles and a crystalline rock surface. Since different minerals are expected to act as preferential reactive areas, the spatial distribution of colloids on the mineral surfaces was studied, accounting several conditions, by means of the nuclear ion beam technique μ-Particle Induced X-ray Emission (μPIXE).

In order to obtain Pu-doped zircon, (Zr,Pu)SiO4, with a maximum Pu content in the form of solid solution, zircon single crystals have been grown using the flux method from starting materials overloaded with Pu. The crystals obtained ranged from 0.2-0.5 to 3.5-4.5 mm in size, are transparent, and characterized by deep pink-brown color. No inclusions of separate Pu phases were observed in the crystals. The distribution of Pu in crystals is zoned and the Pu content varying from approximately 5 to 14 wt.% el. The zircon unit cell parameters calculated from XRD data of bulk powdered sample were: a=6.620(1); c=5.989(2). The results obtained allow us to conclude that the capacity of the zircon structure to incorporate Pu exceeds 10 wt.% el.; however, additional research is required to study the extent of solid solution, (Zr,Pu)SiO4, at higher Pu contents.

Furniture contaminated with uranium will be disposed of, together with U-bearing waste, in ageological repository such as the WIPP site in New Mexico. It is important to understand the effect of the wooden furniture on the migration of uranium in order to predict long-term behavior of uranium in a geological repository environment. In this paper, we present natural uraninite-bearing carbonized wood pieces from a sandstone-hosted roll-type uranium ore deposit in NW China. Results from SEM and TEM observations show that there are nanometer sized, and micron-sized, uraninite crystals that have accumulated on cell walls of the carbonized wood. Some uranitite crystrals display oval and round shapes that may be the result of microbial-induced reduction of uranium from groundwater. The wood carbonized fragments are the most uranium-rich “phase.” In some areas, aggregates of pyrite crystals occur withthe uraninite. It is proposed that organic components from the decay of the wood cells provide nutrientsfor the anaerobic bacteria to grow. The wood pieces with the bacteria inside may serve as scavengers of uranium because of the local reducing chemical environment.

Electrochemical and surface analytical techniques are being used to determine the mechanismand kinetics of Cu corrosion in the presence of sulphide, with the primary goal of developing a mixed potential model for Cu corrosion under disposal conditions. Corrosion potential (Ecorr) measurements indicate the formation of copper sulphide films is rapid leading to rate control by the cathodic reduction of water. Impedance spectroscopy suggests the anodic charge transfer process is rapid leading to film growth limited by diffusive transport. Experiments are underway to determine the current-potential relationships necessary in the development of a mixed potential model.

Garnet-based ceramics doped with Gd, Th, and Ce as waste surrogates and compositions: (Ca1.5GdTh0.5)ZrFeFe3O12 (G3) and Ca2.5Ce0.5Zr2Fe3O12 (G21) were prepared at 1300 °C in air. Corrosion rates in deionized water measured using an MCC-2 procedure were found to be (in g × m-2 × day-1): 10-6 for Ce (G21), 3×10-5 for Gd and 10-6 for Th (G3). Corrosion rates of Ce, Gd, and Th in an acid solution (0.01 M HCl) were 2 × 10-1 (G21), 4 × 10-3 and 10-3 (G3), respectively. Chemical durability of the garnet ceramics is comparable to durability of titanate zirconolite and pyrochlore samples under similar conditions. Scanning electron microscopy has revealed no alteration of the sample surfaces after their contact with water, and the presence of newly formed iron-enriched phases of 1 to 3 micrometers in size, after interaction of the G21 sample with acid solution for 30 days. High leaching of the elements from the Ce-loaded garnet may be due to reduction of Ce in acid solution to Ce(III) or may be explained by lower stability of the structure of garnets (Ca2.5An0.5)Zr2Fe3O12 compared to (Ca1.5GdAn0.5)(ZrFe)Fe3O12, An = Ce4+ and Th4+. It was found that the (Ca2.5Ce0.5)Zr2Fe3O12 and (Ca2.5Th0.5)Zr2Fe3O12 garnets have distorted the elementary unit cell with decreasing lattice symmetry to tetragonal.

Geopolymers are made by adding aluminosilicates to concentrated alkali solutions for dissolution and subsequent polymerization to form a solid. They are amorphous to semicrystalline three dimensional aluminosilicate networks. Although they have been used in several applications their widespread use is restricted due to lack of long term durability studies and detailed scientific understanding. Three important tools for the study of geopolymers are transmission electron microscopy (TEM), solid state magic angle spinning nuclear magnetic resonance (MAS NMR) and infra red (IR) spectroscopy.

Cs and Sr are two of the most difficult radionuclides to immobilize and are therefore suitable elements to study in assessing geopolymers as matrices for immobilization of radioactive wastes. In this study Cs or Sr was added to geopolymer samples prepared using fly ash precursors. A commercial metakaolinite geopolymer was studied for comparison.

The geopolymers were mainly amorphous as shown by TEM, whether they were made from fly ash or metakaolinite. In the fly ash geopolymer, Cs preferentially inhabited the amorphous phase over the minor crystalline phases, whereas Sr was shared in both. The MAS NMR showed that Cs is held mostly in the geopolymer structure for both fly ash and metakaolinite geopolymers. The IR spectra showed a slight shift in antisymmetric Si-O-Al stretch band to a lower wavenumber for the fly ash geopolymer, which implies that more Al is incorporated in this geopolymer structure than in the metakaolinite geopolymer.

An initial uranium phase that has been observed to form during the corrosion of spent nuclear fuel is the uranium oxy-hydroxide metaschoepite. It has been proposed that neptunium(V) solubility can be limited by its association with this uranium phase. Metaschoepite has been synthesized in the presence of neptunium(V) over the pH range modeled in the proposed Yucca Mountain geologic repository. Uranium (VI) phases were synthesized by varying pH and neptunium concentrations. Results of neptunium association with the uranium alteration phases are presented and the relationship to dissolved neptunium concentrations discussed.

A mixed-potential model to predict the corrosion behaviour of nuclear fuel inside a failed carbon steel-lined copper waste container in a granitic repository is briefly described. A number of experiments underway to improve the mechanistic form of the model and to provide the necessary input data are discussed. A primary emphasis is placed on the consequences of the accumulation of corrosion product deposits on the fuel surface on the development of aggressive local chemistries, the cathodic reduction of H2O2 and potential for scavenging of H2O2 by the products of carbon steel corrosion (in particular H2).

The solubility of niobium was investigated for Ca and pH conditions relevant for cementequilibrated solutions. For the pH range considered (9.5-13.2), the dissolved Nb concentration decreases with increasing pH. Overall, experiments lead to Nb concentrations between 2·10-5 M and 2·10-9 M. For all pH values, the dissolved Nb concentration also decreases systematically with increasing Ca concentration. X-ray diffraction measurements of selected experiments confirmed the presence of a solid Ca-Nb-oxide phase, with CaNb4O11·8H2O (hochelagaite) being the most likely composition. On the basis of these findings an empirical regression model for the prediction of Nb solubility data as a function of pH and Ca concentration was derived. This empirical relation is consistent with the presence of a solubility limiting Ca-Nb solid phase and permits to predict aqueous Nb solubility values in cementitious environments over a relatively wide range of conditions. Predicted values are in good agreement with independent experimental results.

In the framework of the research conducted on the long term evolution of spent nuclear fuel in geological disposal conditions, a source term model has been developed to evaluate the instantaneous release of RN (Instant Release Fraction IRF) and the delayed release of the RN which are embedded within the matrix. This model takes into account all the scientific results currently available in the literature except the hydrogen effect. IRF was assessed by considering the evolution with time of the RN inventories located within the fuel microstructure to which no confinement properties can be allocated on the long term (rim, gap, grain boundaries). It allows to propose some reference bounding values for the IRF as a function of time of canister breaching and burnup. The matrix radiolytic dissolution was modeled by a simple kinetic model neglecting the radiolytic species recombination and the influence of aqueous ligands and radiolytic oxidants were supposed to completely react with the fuel surface. Spent fuel performance was therefore demonstrated to deeply depend on the reactive surface area.

An approach to study the diffusion of plutonium (Pu) waste into the material of the storage containers is presented. The delta phase of Pu has been examined, in this article, using a spin and orbitally polarized ab-initio (full potential LMTO) approach [1, 2]. These computations are relatively inexpensive, and it is the intention to see if the results obtained would be satisfactory (compared to the accurate SIC-LDA computations [1-2]) for diffusive studies of Pu into the material of storage containers (for example, into FCC iron as a surrogate for steel [2]). The combination of LDA and spin and orbitally polarized LMTO calculations that we use successfully predict a delocalized 5f, (i.e. bonding) monoclinic ground state and a localized FCC excited state with a density about 8% greater than experiment (compared to the 30% error from pure LDA calculations by various groups [3-6]). The density of states is computed and compared to experimental photoemission studies; there is qualitative agreement. The spin, orbital and total magnetic moment for the FCC phase is computed; the results are qualitatively in agreement with those of other studies. An analytical method for studying the time-dependent diffusion of Pu across an interface with another atomic species is presented.

The in-situ diffusion experiment was conducted at AECL's Underground Research Laboratory (URL) to improve the understanding of diffusive solute transport in sparsely fractured or intact granitic rock (SFR). The experimental program used a comparative series of laboratory and in-situ field experiments to evaluate the ability of laboratory measurements to estimate in-situ rock properties and to explore issues surrounding the influence of stress relaxation, rock texture, porosity, pore geometry, and anisotropy on derived effective diffusion coefficients (De). In-situ experiments yielded iodide Debetween 1.4 × 10−13 and 1.1 × 10−12 m2/s. Unlike laboratory results, the in-situ De estimates did not exhibit correlation with sample depth or varied stress regime. Laboratory-derived measurements of De, porosity and permeability were found to systematically increase for samples removed from greater depths and higher stress regimes. Laboratory-derived iodide De values consistently trended higher than in-situ values by a factor of 1 to 15, except on the shallowest 240-m Level (σ1 ≍ 30 MPa) where differences were negligible. Laboratory-derived estimates of permeability were consistently higher than in-situ derived values by a factor of 2 to 100. This experimental program provides evidence that laboratory steady-state diffusion experiments are most likely to yield conservative values of De for simulation of diffusive mass transport in SFR.

Any environmental performance assessment of spent nuclear fuel disposal requires a prediction of the spent fuel matrix alteration rate when contact with groundwater is established. Taking into account the disposal design, the groundwater at the expected depth of the repository are generally reducing. These reducing redox conditions near the spent fuel surface may be changed to oxidising conditions by the radiolysis of water, due to radiation associated with the waste. This study is focussed on determining the influence of the γ radiation on the spent fuel matrix alteration rate under simulated repository conditions. A series of sequential leaching experiments under controlled conditions have been done. The experiments simulate the repository environmental conditions when engineering barriers are degraded and groundwater arrives at the surface of the spent fuel. In these experiments SIMFUEL (SIMulated FUEL) was used as analogue of spent fuel. Two types of leachant (i.e., granitic water and bentonitic-granitic water) were used in the leaching experiments. These experiments were irradiated using γ- radiation at an industrial irradiation facility with a 60Co source. The results obtained in this study show different behaviour as a function of both the leachant used in the leaching experiments and the initial atmosphere used in the experiments. As a preliminary conclusion of this experimental work, under γ radiation and a small partial pressure of H2 (1 bar) in the first steps of the leaching process has a beneficial effect since U in solution is lower than any other conditions and overall alteration rate is minimized. For experimental times longer than one day, a clear increase of the U in solution is revealed in the experimental run products.

Selection of solid matrix for immobilization of nuclear waste and of geological sites for its disposal represents a complex and consuming task. Previously, we demonstrated that the ion-ion interaction potential (IIIP), representing the main term of the cohesive energy, can be used for estimation of stability of solid matrix/impurity systems. Here we demonstrated a strong correlation between the weathering properties of minerals and their IIIP. Based on these results we proposed a simple theoretical criterion which can be used for assessment of the long term stability of solid matrix for immobilization of radioactive and toxic metals. This criterion was applied in analysis of stability of autunites and some possible practical consequences of results obtained for immobilization of uranium by apatite are discussed.

The Electric Power Research Institute (EPRI) has conducted independent total system performance assessments (TSPAs) of the proposed Yucca Mountain spent fuel and high-level radioactive waste repository for 15 years. EPRI uses its TSPA code, IMARC, currently in its eighth version. The major results of the IMARC-8 analyses are presented in this paper. In all of the situations evaluated by EPRI using IMARC-8, the results for the reference repository concept for Yucca Mountain are well within the regulatory criteria established in the applicable regulations, 10 CFR 63 and 40 CFR 197. Further analyses indicate that at least two of the repository barriers must fail to function as anticipated for dose risks to rise as much as one millirem per year (1/15th of the regulatory limit).