Technetium-99 (⁹⁹Tc) is one of radioactive tank waste components contributing the most to the environmental impacts associated with disposal of radioactive wastes currently stored in underground tanks at the Hanford site, WA. Cast Stone has been selected as the preferred waste form for solidification of aqueous secondary liquid effluents from the Hanford Tank Waste Treatment and Immobilization Plant (WTP) process condensates and low-activity waste (LAW) melter off-gas caustic scrubber effluents. Cast Stone is also being evaluated as a supplemental immobilization technology to provide the necessary LAW treatment capacity to complete the Hanford tank waste cleanup mission in a timely and cost effective manner. Research is being conducted to improve the retention of Tc in the Cast Stone waste forms.
One method to improve the performance of the Cast Stone waste forms is addition of “getters” that selectively sequester Tc. Getter materials that remove Tc from solution are expected to reduce Tc(VII) to the less mobile Tc(IV), In order to determine the effectiveness of the various getter materials prior to their solidification in Cast Stone, a series of batch sorption experiments was performed. Seven getter materials were tested for Tc. Testing involved placing getter materials in contact with spiked waste solutions for periods up to 45 days with periodic solution sampling. Two different solution media, 18.2 MOmega; DI H₂O and a 7.8 M Na LAW waste simulant, were used in the batch sorption tests. Each test was conducted at room temperature in an anoxic chamber containing N₂ with a small amount of H₂ (0.7%) to maintain anoxic conditions.
In this paper we present the results of the batch experiments conducted to determine potential Tc getter materials that will undergo continued testing, selection and subsequently incorporation into Cast Stone. Results indicate that most materials perform better in the DI H₂O (18.2 MOmega;) solution than in the 7.8 M Na LAW waste simulant. We have determined that Tc sequestration may be affected by the presence of other redox sensitive elements that are present in the waste simulant, such as Cr(VI). The Tc getter materials have been examined through various solid-state characterization techniques such as SEM/EDS and XANES. The results indicate that the Tc precipitate differs depending on the getter material and that Tc is reduced in most of the getters but at different extent, from Tc(VII) to Tc(IV).

The immobilization of ⁹⁹Tc and ¹³⁷Cs, two problematic nuclear waste isotopes, in an ABO₄-type structure (where A is an alkali metal and B is either Tc or Ru) has been investigated using atomic scale modelling techniques. The structural stability and free energies of several ABO₄ compounds, including that of CsTcO₄ were computed. Most perfect compositions were calculated to be scheelite structured. To understand the potential wasteforms&’ stabilities during and after transmutation of ⁹⁹Tc to ⁹⁹Ru, and ¹³⁷Cs to ¹³⁷Ba, we also computed the structures and energies of a range of defective ABO₄ compounds. Full and partial transmutation of the waste isotopes were considered, i.e., those of compounds such as A(Tc_1-xRu_x)O₄, (Cs_1-xBa_x)TcO₄ and ARuO₄. We present a number of compositions that may prove to be suitable for either ⁹⁹Tc waste as well as the simultaneous encapsulation of ⁹⁹Tc and ¹³⁷Cs.

⁹⁹Tc, a radioactive fission product, is discharged in nuclear fuel reprocessing operations. In its common form in the environment (TcO₄^minus;) it is a major concern for the remediation of contaminated waters. The difficulty with the removal of TcO₄^minus; is a result of its high mobility in solution and the presence of a high concentration of competing anions such as SO₄^2minus;. A functionalized material, developed at PNNL, known as self-assembled monolayers on mesoporous supports (SAMMS) has previously been found to selectively remove contaminant monovalent oxyanions even in the presence of the divalent SO₄^2minus;.
In this work we have constructed an atomistic model of the SAMMS material and validated the model by comparison to experiment. Density functional theory (DFT) calculations were used to parameterize a classical force field that accounts for the specific interactions of the competing oxyanions with the monolayer. Potentials of mean force for oxyanion adsorption were obtained using umbrella sampling and molecular dynamics (MD) simulations in order to study the material&’s preference for binding TcO₄^minus; over SO₄^2minus;. The results show that the pore structure is a key parameter governing the material&’s oxyanion selectivity. Finally, we suggest ways in which the structure of the material can be optimized in order to maximize TcO₄^minus; adsorption.

Among various radioactive nucleotides, ¹³⁷Cs is the most dangerous radioactive nucleotide because of its high fission yield (6.09 %), medium half-life (30.17 years), and very high solubility in water regardless of its counter anion. Once released into the environment, it easily spreads in nature and enters the food chain, causing enormous damage to human and animal health. In this respect, the effective removal of ¹³⁷Cs⁺ ions from contaminated groundwater, seawater and radioactive nuclear waste solutions is crucial for public health and for the continuous operation of nuclear power plants. However, it is an extremely difficult task because ¹³⁷Cs⁺ concentrations are usually incomparably lower than those of the co-existing competing cations (Na⁺, Ca²⁺, Mg²⁺, K⁺, and others).
Herein we report a novel microporous vanadosilicate K-SGU-45 with mixed valences of vanadium (IV and V), which shows an excellent capturing and immobilization of Cs⁺ from ground water, seawater and highly acidic and basic nuclear waste solutions. This material is superior to other known materials in terms of selectivity, capacity, and kinetics, in particular, at very low Cs⁺ concentrations, it was found to be the most effective material for the removal of radioactive Cs⁺.
This work will trigger the syntheses of various vanadium and other transition-metal silicates that capture various radioactive nuclides, such as ⁹⁰Sr²⁺ ions, and other toxic heavy-metal ions. Furthermore, the discovery of unprecedented hexadeca-coordinated Cs⁺ centers, which corresponds to the highest coordination number ever observed in chemistry, has been described.

Strontium-90 and caesium-137 are waste products produced by fission processes; both have long half-lives of 28 and 30 years respectively. Strontium in particular can have a severe biological impact as it has been shown to accumulate in bones after the intake of contaminated food or water.
Ion exchange materials, such as crystalline silicotitanite (CST, Na₂Ti₂O₃SiO₄middot;2H₂O) and commercially available IONSIV^[1], have been implemented in order to target and remove these harmful radioisotopes and have been shown to be somewhat effective. However Strontium and caesium have proven difficult to immobilise selectively in some cases as ion exchange uptake has been shown to be retarded by the presence of competing cations such as calcium or magnesium^[2] .
A range of new materials similar to CST but based on zirconium and tin silicates, such as NaKSnSi₃O₉.H₂O pictured below, have been investigated for their potential ion exchange applications. These materials are robust against thermal, chemical and radioactive conditions which would make them ideal for use in radioactive waste streams.
A range of materials and the ion exchanged heat treated waste forms have been studied using XRD, TGA, XRF, SEM and EDX analysis in order to characterise the structures and probe the ion exchange properties.
References
1.R.G. Anthony, C.V. Philip, R.G. Dosch, Waste Manage, 1993,13, 503
2.T. Möller, R. Harjula, M. Pillinger, A. Dyer, J. Newton, E. Tusa,S. Amin, M. Webb and A. Araya, J. Mater. Chem.,2001, 11, 1526

An extension of the Swedish final repository for short-lived radioactive waste (SFR) is planned and a safety assessment has been performed as part of the licensing process. Within this work, steps have been taken to advance the modelling environment to better integrate its individual parts. It is desirable that an integrating modelling environment provides the framework to set up and solve a consistent hierarchy of models on different scales. As a consequence, the consistent connection between software tools and models needs to be considered, related to the full assessment domain. It should also be possible to include the associated geometry and material descriptions, minimizing simplifications to source data. The usefulness of the analysis software Comsol Multiphysics as component of an integrating modelling environment has been tested and examples of development work are presented.
Geometry handling is an important part of the modelling process and is closely related to modelling assumptions and simplifications. For the SFR, the relevant geometry includes tunnel systems and storage vaults, as well as engineered structures and barriers. CAD geometries developed during planning and design work have been successfully imported into Comsol. The landscape above the repository also constitutes relevant geometrical input for assessment modelling. Development work has allowed the import of geographic information system (ArcGIS) data into Comsol, incorporating digital elevation models as well as soil and sediment domains into model geometries.
The ability to set up and solve a consistent hierarchy of models on different scales is an important capability of an integrating modelling environment. Extracting models for repository scale hydrology from regional hydrogeology models and regional surface hydrology models are two examples. The regional hydrogeology model of the SFR site covers several square kilometres of land and reaches depths of approximately one kilometre. The repository scale model is contained within the regional model and has dimensions one order of magnitude smaller. To calculate the detailed groundwater flow through the repository requires the proper boundary conditions from the regional hydrogeology. A consistent connection was achieved by programming an interface allowing Comsol to extract the near-field boundary conditions and bedrock property fields from the regional model, set up and solved in the DarcyTools software.
The repository scale hydrology models provided a basis for further model developments focused on coupled processes. An interface between Comsol the geochemical simulator PhreeqC has been developed to support reactive transport studies. An important test case involved radionclide transport in a 3D model of a catchment area. The dynamic surface hydrology was simulated with MIKE SHE and coupled to detailed chemical processes occurring in soils and sediments.

Reactive transport modelling entails the integration of hydrogeology and geochemistry. One of the challenges for such integration is the large amount of computational resources needed due to the high non-linearity of the resulting system of equations. Taking advantage of new developments of powerful numerical tools, and based on high performance parallel computing, the solution of large-scale hydro-thermal-geochemical-mechanical models has become possible. A software solution, denoted iDP, has been developed which serves as an interface between 2 standalone simulators: DarcyTools [for groundwater flow in fractured rocks] and PFLOTRAN [for reactive solute transport]. iDP has been applied for the first time to test the new update of Mare Nostrum, the main machine at the Barcelona Supercomputing Centre, the National Supercomputing Centre in Spain. An average of 8,000 processor cores during 15 days were used to solve a large-scale (100 Mcells), long-term (20,000 years) simulation to evaluate the degradation of cement grout that will be injected in the fractures of the granitic rocks during the construction of a deep geological repository for spent nuclear fuel in Forsmark (Sweden). The simulation integrates the complex 3D groundwater flow accounting for the Discrete Fracture Network (DFN) of the site, and the complexity of the geochemical system involved in cement grout dissolution and secondary minerals precipitation within the flowing fractures. Model results allow evaluating the expected durability of the injected cement grout, as well as to evaluate the risk of hyper-alkaline groundwater development and migration towards the depositional area of the repository. This work shows that High Performance Computing of reactive solute transport is a reliable and powerful tool for decision makers involved in the planning and constructions of deep geological repositories for nuclear waste.

A simple and effective model for a safety assessment of a conceptual repository system, in which low-level radioactive wastes that arises from nuclear power plants and other sources has been developed using the commercial GoldSim development tool. The repository system is assumed to be planned for construction on the surface area near the seashore. The computer program based on this model, developed as a GoldSim template, is ready for a total system performance assessment (TSPA), and is able to probabilistically evaluate a nuclide release from a repository and farther transport into the geosphere and biosphere under various normal, disruptive events, and scenarios that can occur after a failure of a waste drum with associated uncertainty. To quantify the nuclide release and transport through the various pathways possible in the near- and far-fields of the repository system under a normal groundwater flow and some alternative scenarios, illustrative evaluations are made and demonstrated through this study. Even though all parameter values associated with the repository system were assumed for the time being, the illustrative results should be informative since the evaluation of such releases is very important not only in view of the safety assessment of the repository, but also for the design feedback of its performance.
The 200L storage drums for low-level waste, which amounts to a total of 125,000 drums, are to be disposed of in concrete containers and then buffered by gravel or grouted with concrete. Impervious materials and multilayered covers for preventing water infiltration and some erosion as well as nuclide release are considered to place on the roof. In GoldSim modeling, a trench and its surrounding are discretized into several compartments ready for run-off, infiltration as well as diffusive and advective transport in and among them. Several principal release pathways from the trenches are set in place: the upper, side, and base pathways, all of which simultaneously reach to the far-field transport. All releases from the trenches are then later transported along with various unsaturated and saturated pathways including surface and subsurface groundwater flow pathways into the natural far-field area.
For trench type repositories at the surface or possibly subsurface depth, normally and commonly, once leakage from a damaged radioactive waste package of a drum, and through tiny holes, happens, the nuclides will spread out to the buffer material surrounding the drum, and then into other possible regions in the trench before farther transporting into the biosphere through various pathways. In the case of transport into the rock medium under the repository, the internal fractures and the major water conducting features (MWCFs) that are assumed to exist in the far-field area of the repository could be one of the main pathways through which the nuclides finally reach the human environment by passing over the geosphere-biosphere interfaces for exposure to human bodies.
The scenario mainly considered here for a probabilistic safety assessment is a normal case, under which nuclides are released by overflow and/or groundwater that normally flows along their own preferential pathways after release from each repository. Through this study, a probabilistic behavior of nuclide releases from a low-level waste trench type repository is illustrated with varying parameters, which were selected among many others in view of their possible consequences and probabilities.

The safe treatment of nuclear waste poses a lasting risk to the environment and has high costs. Buried repositories represent the long term storage which is required to be stable. The stability includes many aspects such as chemical and mechanical stabilities as well as impermeability. Clay minerals are excellent candidates to maintain a long lasting seal of the nuclear waste repository due to their large adsorption capacity and swelling characteristics in aqueous suspensions. However, the interaction and transport of radionuclides in clay minerals, including organic clay minerals, still need to be fully addressed.
Atom level simulations have not yet been fully exploited to investigate these processes not least because of the complexities involved. Here we present our recent work to gain atomistic insights into the factors controlling the interaction of heavy and radioactive ions at clay mineral - water interfaces. Quantum and potential based techniques are used to explore the evolution of systems of different sizes and for different lengths of time enabling us to efficiently evaluate structural and dynamical properties of this class of geosorbents. The interaction of these ions with clay minerals is generally thought to occur on the basal plane, which dominates their morphologies and has been the focus of many investigations. However, the edge surfaces are more reactive and with a greater range of compositions and charge states can indeed provide more efficient interaction sites.

Glass has been widely adopted as the first generation host material for the immobilisation of high level nuclear waste. It is intended that immobilised waste will go for permanent disposal in geological repositories and it is desirable that any wasteform should be durable under these conditions for an extended period. Atomic scale computer simulation can be used to provide a mechanistic basis for the structure and properties of glasses and as a result offers opportunities for the compositional optimisation of nuclear waste glasses.
Experimental studies have reported that zinc oxide improves the durability of nuclear glasses. Through the use of molecular dynamics, in conjunction with a simulated melt-quench procedure, atomic structures of sodium silicate glasses, have been generated with and without zinc. The structure of these glasses was studied through the use of pair distribution functions, ring size distributions and cluster analysis. Using the insights gained from these analyses the structural role of zinc oxide within silicate glass is discussed and consideration is given to reports of its differing roles as a network former and a network modifier. The effects of Zn addition on sodium ion distribution and clustering behaviour within the glasses is also reported. This is used to explain changes to intermediate-range structure and hence provide a possible explanation for the experimentally observed increase in durability obtained with the addition of Zn.

Radioactive iodine-bearing materials, such as spent silver adsorbent, are produced in nuclear reprocessing plants in Japan. According to Japanese disposal plan radioactive wastes that contain a certain quantity of iodine-129 are classified as Transuranic Waste Group 1 (TRU 1) for spent silver adsorbent or as Group 3 for bitumen-solidified waste and they should be disposed of by burial deep underground. Because the half-life of iodine-129 is 15.7 million years, it would be difficult to prevent release of iodine-129 from the wastes into the surrounding environment over such a prolonged time. Moreover, because iodine in its ionic forms is soluble and not readily adsorbed, its migration is not retarded significantly in engineered or natural barriers. Therefore the release of iodine-129 from nuclear wastes needs to be restricted to permit reliable safety assessment; this technique is called “controlled release”. It is desirable that iodine release period will be longer than 100,000 years.
Several techniques for immobilization of iodine have been developed for this purpose. These are narrowed down to three techniques such as synthetic rock, BPI (BiPbO₂I) glass and high performance cement. Iodine will be fixed as AgI in grain boundary of corundum or quartz through hot isostatic pressing (HIP) in the synthetic rock, as BPI in boron-lead based glass, or as some cement minerals such as ettringite in alumina cement. These techniques are assessed by three models such as the leaching model, the distribution equilibrium model, and the solubility-equilibrium model. In this paper current status of these techniques are described.

To support the future expansion of nuclear energy an effective method is needed for the capture and safe storage of radioisotopes released during reprocessing of spent nuclear fuel. The Department of Energy Office of Nuclear Energy (DOE-NE) is currently investigating alternative waste forms for ¹²⁹I. DOE is interested in new waste forms that can provide higher waste loadings, more efficient consolidation routes, lower costs, etc. PNNL has been developing non-oxide aerogels made with metal sulfides, termed chalcogels, for iodine immobilization and thus far, the materials do show promise as a potential replacement avenue for AgZ. These chalcogels are stable in aqueous solutions. Scientists at the university lead on this proposal who area the inventors of the chalcogel class of materials have already demonstrated selective affinity with chalcogels for metal ions in aqueous media such as Cs⁺, Sr²⁺, and Co²⁺. Aerogels have been studied for confinement of radioactive wastes in recent years and are under investigation as waste forms for ¹²⁹I. Aerogels can act as precursors to the final glass matrix that actually immobilizes the wastes. Use of silica aerogels for the purpose has been limited by their brittleness in the presence of water that is commonly present in off-gas treatment and also due to their low permeability to nuclear waste.
Recently, we reported a new type of aerogel made with metal chalcogenides (where chalcogen is S, Se, and/or Te) and is referred as chalcogel. Non-oxide materials such as the chalcogels have different properties than oxide materials and, in this case, some of those differences are actually advantages. For example, the high polarizability of the chalcogens (over oxygen) can be used to capture iodine. We will report the exploration of chalcogels as high affinity materials for capturing iodine and the conversion of the loaded materials to glass forms. The efficiency of a chalcogel-based waste form is expected from strong complex formation based on the high chemical affinity of chalcogen atoms for iodine gas. The strong chemical affinity is due to the soft Lewis acid/soft Lewis base complex formation, according to Pearson&’s Hard/Soft Acid-Base (HSAB) principle. We also report that chalcogels can be chemically tailored to exhibit additional strong I₂ capture mechanisms.

¹²⁹I is a major byproduct generated from nuclear fission of uranium fuel. Due to its adverse health effects in humans, safe removal and storage of ¹²⁹I is of utmost importance across various nuclear energy plants. The sorbents for the absorption of radioiodine has to be stable during the treatment process and also it should be capable of sorbing large amounts of ¹²⁹I. The most commonly used sorbents are silver-loaded zeolites and Ag-loaded silica aerogel. However, due to the poor mechanical stability of silica aerogels in an aqueous environment there is a need to develop new material with better mechanical stability. The chalcogen-based aerogels called “chalcogels” are highly porous and have showed good affinity towards heavy metal ions. Herein we report the use of chalcogels and silver functionalized analogues as host materials for capture and immobilization of ¹²⁹I. Iodine capture was studied with different chalcogels (Sb₄Sn₄S₁₂, Zn₂Sn₂S₆, NiMoS₄ and CoMoS₄), their silver functionalized analogues, and binary metal sulfides. All the chalcogels showed high uptake reaching up to 200 mass% and the iodine chemically reacted with the sorbents to form metal -iodide complexes. We will also report the consolidation of various iodine loaded chalcogels with different glass-forming additives into a final waste form.

Reprocessing of spent nuclear fuel is being considered in the U.S. In that case, the release of volatile ¹²⁹I from reprocessing plants and its safe storage would have to be controlled to meet the Environmental Protection Agency emissions regulations (which require capture of 99.4% of ¹²⁹I) and disposal restrictions. Currently, a silver-loaded zeolite (AgZ) is the baseline material for removing ¹²⁹I. However, recent studies indicate limitations in the sorption performance and long-term stability of AgZ. Also, AgZ requires addition of low-temperature glass to immobilize trapped radioiodine. To avoid these drawbacks, silver-functionalized silica aerogel is being developed for the efficient capture and immobilization of ¹²⁹I. This novel sorbent has a high affinity for iodine at the low concentrations expected in the off-gas and a high sorption capacity, and, after loading with iodine, it can be consolidated into a dense and leach-resistant SiO₂-based waste form. It was demonstrated to have a sorption capacity for I₂ of 480 mg/g, decontamination factors in excess of 10 000, good sorption performance after long-term exposures to dry and humid air, and retention of more than 92% of iodine in the densified product. The presentation will highlight the results from a series of sorption and consolidation studies.

Since 1991, the potential of several specific inorganic host matrices was studied at CEA to ensure a durable immobilization of iodine 129 in the frame of a possible disposal in a deep geological repository.
Due to evidence of retention of xenon 129, decay product of iodine 129, over geological time scales in apatites, these phases were the first materials to be considered. Specifically, a lead-bearing apatite with a good chemical durability was initially developed. Its composition can be written as Pb₁₀(VO₄)_4.8(PO₄)_1.2I₂. At 90°C, in pure water, its leach rate is of 2.28 10^-3 g.m².j^#8209;1 on the basis of iodine release and this rate decreases with time as the progressive replacement of iodide ions by hydroxyl groups along the channels of the crystalline structure occurs. This replacement obeys to a diffusive law and the transformation of iodoapatite grains into hydroxyapatite can be qualified as being pseudomorphic. Current studies are devoted to the shaping of such an iodoapatite in order to get a dense monolith. In so doing, it was found that a reactive sintering by spark plasma sintering at 400°C under a pressure range of 40-70 MPa could offer a clear benefit over sintering techniques in sealed environment (e.g., HIP) of which the use could be seen as more complicated for a reprocessing plant. This allows pellets of more than 92% of the theoretical density to be obtained. However, this process also appears to be very sensitive to scaling effects and it requires a subsequent optimization.
Other apatite compositions were also studied to avoid the use of toxic elements like lead. These apatites were developed on a phospho-calcic basis. They have the noticeable ability of incorporating iodine under its iodate form. Depending on phases constitutive of the geological barrier around the repository site, iodate ions could be less mobile in comparison with iodide which could delays the return of iodine to the biosphere. It was demonstrated that the incorporation mechanism of iodate into such an apatite relies on a substitution of hydroxyls groups. The chemical durability of this apatite is currently evaluated.
Together with ceramics, some glass compositions were also considered. They belong to the AgI-Ag₂O-P₂O₅-Al₂O₃ system. Close compositions were already proposed by Japanese teams for a similar goal. Here, the idea was to improve their properties by addition of a cross-linking reagent of the phosphate network like alumina. These glasses have an intrinsic compatibility with silver iodide which is the form adopted by iodine in most of the iodine capture processes on solid filters. They can incorporate high iodine amounts and their leaching behavior depends on phosphate chain length, iodine amount and alumina content.
Beyond the development of each matrix, the desired goal would be to have a correct opinion on the strengths and drawbacks of these solutions to face with future industrial and regulatory needs.

Apatite structure type, with a typical chemical composition of A₁₀(BO₄)₆C₂ (e.g. , A=Ca, Na, Pb, rare earth, fission product, actinides; B=P or V; C=F, Cl, I.) shows tremendous potentials as advanced waste forms for effective nuclear waste management. A wide range of radionuclides can be incorporated into its crystal structure by coupled substitutions at both cation and anion sublattices. Of particular importance, iodine-bearing apatite with chemical composition of Pb₁₀(VO₄)₆I₂ is proposed to confine extremely mobile and highly volatile I-129, a fission product of uranium fission. However, iodine-bearing apatite are typically synthesized and densified at elevated temperatures, resulting in evitable iodine loss. In this work, Pb₁₀(VO₄)₆I₂ powder samples are synthesized by solid state reaction at room temperature by using High energy ball milling (HEBM) followed by thermal annealing at 200 oC to control the crystallinity. Dense iodine-loaded apatite ceramic pellets were consolidated by state-of-art Spark plasma sintering (SPS) at various temperature (350 ^oC to 700 ^oC) and very short durations (0 ~20 mins). Iodine retention and the microstructure tenability, especially grain size, were investigated as functions of different SPS parameters (temperature, holding time, and pressure). No significant iodine loss was identified by high energy ball milling and during densification during SPS process. The thermal stability, thermal conductivity, and mechanical properties of the densified apatite pellets as durable iodine waste forms were studied. These results highlights that the SPS combining with high energy ball milling is a promising method to consolidate durable ceramic waste forms for confining highly volatile iodine for effective nuclear waste management.

The synthetic rock solidification is a HIPing technique to immobilize radioactive iodine (I-129) in the fuel reprocessing off-gas systems collected by silver nitrate impregnated onto an alumina base sorbent. Although iodine on the sorbent as a form of silver iodide (AgI) is unstable under the reducing condition, the α-alumina matrix of the synthetic rock is expected to fix the AgI physically in the grain boundary to be controlled iodine release after the geological disposal.
In the present study, the MCC-1 type immersion tests have been performed as a function of pH and hydrosulfide ion (HS^-) concentration as a reductant. The synthetic rock sample has been prepared by HIPing at 175MPa and 1473K for 3hours from a simulated spent sorbent saturated with stable iodine. Leached iodine has been under detection limit of an ICP-MS measurement below the HS^- concentration of 10^-5 M due to the stability of AgI itself. As the HS^- concentration of 10^-3 M, the iodine leaching rapidly increases within 100 days due to the AgI dissolution located at the surface and in open pore. The cross-section observation after immersion by EPMA and XRD suggests the following reaction: 2AgI + HS^- = Ag₂S + 2I^- + H⁺. Effect of pH has been clarified after 100 days that the both aluminum and iodine leaching decrease as pH decreases from 12.5 to 8. This indicates that the alumina matrix reasonably controls the iodine release. However the normalized leaching rate of iodine is 10 to 1000 times larger than that of aluminum. The incongruent leaching behavior would be due to the internal pore and grain boundary dissolution.
This research is a part of “Research and development of processing and disposal technique for TRU waste (FY2013)” under a grant from the Japanese Ministry of Economy, Trade and Industry (METI).

For safety assessment analyses of the disposal of spent nuclear fuel (SNF) in deep geological repositories it is indispensable to evaluate the contribution of activation and fission products to the instant release fraction (IRF). This fraction consists of soluble elements with low sorption tendency and contributes significantly to the calculated dose rates.
The IRF is controlled by the segregation of a part of the radionuclide inventory to the gap interface between the cladding and the pellet, to the fractures as well as to grain boundaries. The radionuclides that segregate are the fission gases (Kr and Xe), volatile radionuclides (³⁶Cl, ⁷⁹Se, ¹²⁹I, ¹³⁵Cs and ¹³⁷Cs) and metallic radionuclides (⁹⁹Tc and ¹²⁶Sn). The degree of segregation of the various radionuclides depends highly on in-reactor fuel operating parameters such as linear power rating, fuel temperature, burn-up, ramping processes and interim storage time. In the case of the fission gases, the gas release occurs by diffusion to grain boundaries, grain growth accompanied by grain boundary sweeping, gas bubble interlinkage and intersection of gas bubbles by cracks in the fuel.
During the last three years a EURATOM FP7 Collaborative Project, “Fast / Instant Release of Safety Relevant Radionuclides from Spent Nuclear Fuel (CP FIRST-Nuclides)” was carried out to get a better understanding of the IRF. Within CP FIRST-Nuclides, an irradiated UO₂ fuel pellet with cladding was sampled from a fuel rod segment with an average burn-up of 50.4 MWd/kg_HM. The cladded SNF pellet was leached in 19 mM NaCl + 1 mM NaHCO₃ solution under 40 bar Ar + H₂ atmosphere (p_H2: 4 bar). In the multi-sampling autoclave experiment, gaseous and liquid samples were taken periodically. The gaseous samples were analysed for fission gases by means of gas mass-spectrometry, the liquid samples were analysed for ¹²⁹I, ¹³⁷Cs and other dissolved radionuclides by means of gamma spectrometry, LSC and ICP-MS. In the present communications we focus on the behaviour of fission gases, ¹²⁹I and ¹³⁷Cs as a function of leaching time. After 177 days of leaching experiment, the percentage of fraction of the inventory released into the gaseous and aqueous phases was: 12.8 for the fission gases (Kr + Xe), 9.2 for ¹²⁹I and 2.8 for ¹³⁷Cs, respectively.

MoO₃, which can be found at elevated levels in some high level nuclear waste streams in the UK and France, is one of the most challenging oxides to immobilise in the borosilicate glasses conventionally used for nuclear waste vitrification. MoO₃ usually has a low solubility (le;1 mol%) in silicate glasses and excess MoO₃ in nuclear glass causes the formation of “yellow phase” which is detrimental to the vitrification process. Glass compositions with greater MoO₃ solubility limits are therefore desirable. In this work the solubility and incorporation of MoO₃ in an alkaline earth aluminosilicate glass system (AeAS, Ae = Mg, Ca, Sr, Ba or two of these combined) have been investigated, showing that MoO₃ solubility steadily increases in the order Ba < Sr < Ca < Mg. Up to 5.3 mol% (12.3 wt%) MoO₃ can be retained in magnesium aluminosilicate glass (MAS) without phase separation while only 2.0 mol% (2.5 wt%) MoO₃ can be completely dissolved in barium aluminosilicate glass (BAS). The high MoO₃ solubility in MAS glass provides the possibility of using it as an alternative wasteform for the vitrification of nuclear waste containing high levels of MoO₃. The changes in glass structure and properties caused by MoO₃ incorporation are also assessed. Glass density increases whereas glass transition and crystallisation temperatures decrease with increasing MoO₃ addition. The prepared glasses reveal good thermal stability until glass transition. All visibly homogeneous glasses are X-ray amorphous while the partially crystallised glasses exhibit some small X-ray diffraction peaks which are probably due to corresponding molybdates. In Raman spectra, MoO₃ addition contributes two broad bands which are assigned to vibrations of MoO₄^2#8210; tetrahedra. The intensities of these bands increase along with MoO₃ incorporation until saturation. In the Raman spectra of partly crystallised glasses with combined alkaline earths, the crystalline bands are in accordance with the molybdate with lowest solubility whenever possible, indicating that MoO₃ solubility in glass is controlled by the cation whose molybdate salt has the highest crystallisation tendency. Electron microscopy shows that these separated particles are spherical, with sub-micron diameters and are randomly dispersed within glass. The separated phases are formed through liquid-liquid separation and thereafter crystallisation. Overall AeAS glasses look quite promising for molybdate immobilisation with MAS glasses being particularly attractive.

Currently in Russia some compositions of spent nuclear fuels (SNF) such as molybdenum-bearing SNF of uranium-graphite reactors (AMB) are not reprocessed yet but their reprocessing is under consideration now. High level waste (HLW) from AMB SNF reprocessing is suggested to be incorporate in sodium aluminophosphate (SAP) based glass similarly to different HLW. Mo is one of the troublesome components of HLW causing liquid/liquid phase separation in borosilicate glasses and crystallization of phosphate glasses and reduction of chemical durability of vitrified waste. Therefore the effect of Mo solubility, its valence state and speciation on chemical durability of glasses has to be studied. Incorporation of Mo in SAP glass favors its crystallization and annealing increases the degree of crystallinity. Valence state and local environment of Mo in the materials containing ~2 wt.% MoO₃ were characterized by X-ray absorption fine structure (XAFS). In the quenched samples composed of major vitreous and minor AlPO₄ crystalline phase nearly all Mo is located in the vitreous phase as [Mo⁶⁺#1054;₆] units whereas in the annealed samples Mo is partitioned among vitreous and one or two orthophosphate crystalline phases. The spectra of annealed markedly crystallized samples contain weak response in pre-edge range which can be assigned to the line due to contribution of [#1052;#1086;⁶⁺#1054;₄] units located in the crystalline phase. Mo predominantly exists in a hexavalent state in distorted octahedral environment. Three oxygen ions are positioned at a distance of ~1.70 Å and three - at a distance of ~2.04 Å. In the highly-crystalline annealed samples especially contained 5.4 wt.% MgO the best fit is achieved on the assumption of three different Mo-O distances: 2.5-3 oxygen ions are positioned at a distance of ~1.73 Å, 2.3-3 oxygen ions - at a distance of ~2.05 Å and 1-1.5 oxygen ions - at a distance of ~2.3 Å from Mo⁶⁺ ions. This may be attributed to contribution due to minor Mo in complex orthophosphate. Formation of Mo-bearing phosphates could be the reason of deteriorating of chemical durability of the materials especially after their annealing resulting in increase of the degree of crystallinity.

The Hanford site in southeastern Washington State is the largest repository of nuclear waste in the United States, where 177 underground tanks stored in excess of 50 million gallons of waste. This waste will be mixed with glass-forming additives and vitrified at the Waste Treatment and Immobilization Plant (WTP). A large fraction of the anticipated waste streams are simultaneously high in both Na and Al, leading to frequency crystallization of aluminosilicate phases such as nepheline upon cooling. This aluminosilicate crystallization has been previously shown to be deleterious to chemical durability due to the extraction of alumina and silica from the glass-forming matrix, leaving a residual glass of less durable components. The long-term corrosion resistance is of significance because vitrified waste must tolerate subterranean storage for ge;10⁶ years.
A challenge in the formulation of nuclear waste glasses arises from maintaining a sufficiently low addition of glass-forming additives to maximize waste loading, while ensuring that the addition is sufficiently high to prevent crystallization. In this study a glass composition, designated as A4, was selected due to its particular crystallization behavior. This composition was formulated for a high-alumina waste stream (>25 wt% Al₂O₃) with 45 wt% waste loading. A4 glass was batched from powder precursors and subjected to air quenching, isothermal heat treatments, and canister-centerline cooling (CCC) to observe the crystallization behavior, with the goal of obtaining time-temperature-transformation curves.
Crystal fractions were obtained by x-ray diffraction (XRD) and crystallite structure was observed through scanning electron microscopy (SEM) with composition measured through wavelength dispersive spectroscopy (WDS) and energy dispersive spectroscopy (EDS). Nepheline, a feldspathoid with sodium end-member composition NaAlSiO₄, is the prominent aluminosilicate phase in the CCC high-alumina waste glass, but it forms in unusually large isolated dendritic crystals in the presence of a complex assemblage of crystals of phosphate, spinel, and a residual glass system enriched in Ca, Mg, Zr, and B.
The ultimate goal of this ongoing work is to obtain a kinetic model for crystallization of nepheline and enable compositional design to inhibit rapid crystallization of nepheline in high Na and Al wastes.

To limit foaming and improve the melting rate of simulated high-level waste melter feeds during the vitrification process, the importance of silica grain size was investigated. A high-alumina high-level waste melter feed formulated by Vitreous State Laboratory was prepared using various silica grain sizes. Feed samples were heated at 5°C/min up to 1200°C. To observe volume expansion, feed pellets were created and photographed during heating. Quenched samples from heat treatments were prepared for scanning electron microscopy and crystalline phases were determined with X-ray diffraction. By eliminating fine particles contained in most silica sources, the volume expansion caused by foaming was decreased due to a delay in silica dissolution that improved the overall melting rate.

To efficiently vitrify Hanford waste, the melting process (i.e., melter feed turning into waste glass) must be modeled and optimized. The rate of heat transfer to the melter feed in a waste glass melter, and thus the rate of melting, is strongly affected by the melter feed porosity, especially in the final stages where the glass-forming melt produces foam that insulates the feed from the molten glass. The volume expansion test allows the determination of the melter feed porosity as a function of temperature. This test measures the profile area of the feed pellet as it turns into glass. This contribution presents the calculation of the void fraction (porosity) of the melter feed as a function of temperature, heating rate, and material parameters. The process of finding the void fraction is described as well as results from the application of this process.

Nepheline, (Na,K)AlSiO₄, is an aluminosilicate mineral that can crystallize in waste glass containing a high fraction of Al₂O₃ and Na₂O when it is slow cooled from a glass melt. The formation of nepheline alters the residual glass composition and lowers durability by taking key glass-forming components of alumina and silica out of the glass. Understanding what glass compositions limit or encourage nepheline formation and having a model to predict nepheline formation in waste glasses is critical to achieve the maximum waste loading. Early ternary models only included Na₂O, Al₂O₃, and SiO₂ as predictive variables for nepheline formation and were too conservative. A new model includes Na₂O, Li₂O, CaO, Al₂O₃, B₂O₃, and SiO₂ and appears to show a clear delineation in the data between glasses that do and do not form nepheline upon slow cooling. Details for how this model was constructed and future glass formulation work to evaluate the model will be discussed.

The Department of Energy-Office of River Protection (DOE-ORP) is constructing the Hanford Tank Waste Treatment and Immobilization Plant (WTP) to treat radioactive waste currently stored in underground tanks at the Hanford site in Washington. The WTP that is being designed and constructed by a team led by Bechtel National, Inc. (BNI) will separate the tank waste into High Level Waste (HLW) and Low Activity Waste (LAW) fractions with the majority of the mass (~90%) directed to LAW and most of the activity (>95%) directed to HLW. The pretreatment process, envisioned in the baseline, involves the dissolution of aluminum-bearing solids so as to allow the aluminum salts to be processed through the cesium ion exchange and report to the LAW Facility. There is an oxidative leaching process to affect a similar outcome for chromium-bearing wastes. Both of these unit operations were advanced to accommodate shortcomings in glass formulation for HLW inventories. A by-product of this are a series of technical challenges placed upon materials selected for the processing vessels. There exist additional questions on the adequacy of the design to ensure nuclear safety requirements are met during across the entire spectrum of possible operating conditions.
At the heart of the treatment process is the vitrification of the HLW and LAW waste streams in Joule Heated Ceramic Melters (JHCMs). The JHCM is typically operated at a melt pool temperature of 1150°C. The slurry feed is introduced from the top of the melter and during operation the melt pool is almost entirely covered with unmelted feed termed the cold cap. The Hanford JHCMs are fitted with a patented bubbler system to agitate the melt pool, thus improving heat transfer to the cold cap and, therefore, feed processing rate. The Office of River Protection undertook an extensive investigation focused upon glass formulation improvements and enhancements of operating efficiencies in the vitrification facilities. The outcomes have the potential of profound, but positive, impacts on the baseline flow sheet for the Pretreatment Facility. Not forgetting the impact on more rapid realization of successfully emptying the waste tanks and treating the waste.
The WTP will process and treat approximately 53 million gallons of mixed hazardous wastes (i.e., radioactive and chemical waste). This presentation provides an overview of the project status and technical challenges facing the process, design, and construction of the WTP facilities.

The storage of radioactive waste is an on-going problem around the world. Yucca Mountain with its tuff and granite rocks was declined by the US government. In Germany the salt stocks are again in discussion and the exploration of a suitable repository is starting over again. In discussion are clay deposits - which e.g. Switzerland favors - or granite or other rock material.
Several factors have to be considered for the suitability of such places: geo-mechanical behavior, fluid-tightness, no earthquake region etc. As the nuclear waste might heat up the surroundings, the thermal conductivity of the surrounding materials should be high enough to avoid too much accumulation of the waste heat.
Another question is what happens to the repository material when heated up accidentally. In this contribution, granite, rock salt and clays were investigated by TG-DSC, dilatometry, LFA and evolved gas analysis and results for thermal stability, thermal expansion and thermal diffusivity will be reported.

High-level waste (HLW) from reprocessing of spent nuclear fuel of uranium-graphite reactors (Russian AMB type) is suggested to be vitrified with production of aluminophosphate based glass similarly to current PWR type (Russian WWER). Some of the AMB fuel compositions contain copper and therefore behavior of copper ions in sodium alumonophosphate glasses has to be investigated. Target glass formulations contained ~2.4-2.5 mol.% CuO. The mixtures of chemicals were dried in a dessicator, placed in Pt crucibles, heated to 1000 °C, kept at this temperature for 0.5 hr, and melts were poured onto a stainless steel plate followed by annealing of the materials at a temperature of 500 °C for 14 hrs. The quenched sample was composed of major amorphous phase and minor aluminum orthophosphate (20-30 vol.% of total). The quenched MgO bearing (13.2 mol.%) sample was predominantly amorphous (<5 vol.% AlPO₄). The annealed MgO free sample had higher degree of crystallinity than the annealed MgO-bearing sample but both them contained orthophosphate phases. Cu in the materials was partitioned in favor of the vitreous phase. In all the samples copper is present as major Cu(II) and minor Cu (I) forms. Cu²⁺ ions form planar square complexes (CN=4) with a Cu²⁺-O distance of 1,93-1,95 Å. Two more ions are positioned at a distance of 2,76-2,86 Å from Cu²⁺ ions. So the Cu²⁺ environment looks like a strongly elongated octahedron as it also follows from the absence of the pre-edge peak due to 1s→3d transition in Cu K edge XANES spectra of the materials. Cu⁺ ions form two collinear bonds at Cu⁺-O distances of 1,80-1,85 Å. Thus average Cu coordination number (CN) in the first shell was found to be 2.7-3.0.

Until the early 1970s, borosilicate glass was the reference waste form for immobilizing high-level (reprocessing) nuclear waste. But in the early 1970s, Penn State University (PSU) researchers (Rustum Roy, Will White and Greg McCarthy) introduced the potential use of synthetic minerals which were known to be leach resistant in hot, wet conditions through geological studies and could incorporate key fission product elements in their crystal lattices. These minerals were phosphates and silicates and produced by standard ceramic methods. Ted Ringwood at the Australian National University soon became aware of the PSU work and put forward assemblages of titanate minerals for the same purpose and showed them to be much more durable in water than the phosphates and silicates (and borosilicate glass). Bob Dosch at Sandia also worked on titanate ion exchangers for the same reasons. This presentation will highlight the early ground breaking work which still significantly influences current waste form proposals.

The disposal of high level radioactive waste is one of the most pressing and demanding challenges. With respect to long-term safety aspects of geological disposal, the minor actinides (MA) such as Am, Cm and Np and long-lived fission products such as ³⁵Cl, ¹³⁵Cs, ⁷⁹Se, ⁹⁰Sr and ¹²⁹I may be of particular concern due to their long half-lifes, their high radiotoxicity and mobility in a repository system, respectively. Ceramic waste forms for the immobilisation of these radionuclides have been investigated extensively in the last decades since they seem to exhibit certain advantages compared to other waste forms (incl. borosilicate glasses and spent fuel) such as high loadings and chemical durability. Currently, most on-going nuclear waste management strategies do not include ceramic waste forms. However, it is still important to study this option, e.g. with respect to specific waste streams and certain constraints regarding deep geological disposal.
In the present communication we report on the research program in Jülich regarding ceramic waste forms for the conditioning of MA. It is based on fundamental science and follows an integral approach that covers the separation of elements or elemental groups with similar chemical properties from a waste stream by liquid/liquid extraction as well as the immobilization in ceramic materials as hosts. Various aspects with the focus on single phase waste forms, such as monazite and zirconates with pyrochlore structure will be discussed:
1.) Development and optimisation of synthesis routes suitable for immobilisation of MA into ceramic waste forms and the handling of radionuclides such as sol-gel route, hydrothermal synthesis and co-precipitation,
2.) structural and microstructural characterisation using state of the art spectroscopic (Raman, TRLFS, EXAFS), diffraction (powder and single crystal XRD) and microscopic (SEM, FIB/TEM) techniques,
3.) determination of thermodynamic data (calorimetry) and reactivity under conditions relevant to geological disposal, in particular with respect to dissolution in aqueous environments (static & dynamic dissolution experiments on powders and pellets) as well as
4.) studies on radiation damages (irradiation with α-particles and/or heavy ions, and incorporation of short-lived actinides such as ²³⁸Pu, ²⁴¹Am or ²⁴⁴Cm).
Finally, a fundamental understanding of the long-term behaviour on the atomic scale will help to improve the scientific basis for the safety case of deep geological disposal concepts using ceramic materials.

The complexity and diversity in the chemistry of legacy UK radioactive wastes has necessitated the development and validation of a toolkit of advanced wasteforms. This approach will expand the range of materials to be consigned to a future geological disposal facility, to include new glass, ceramic, and cement wasteforms. This presentation will examine recent advances in the fundamental understanding of these wasteforms and their interaction with conceptual disposal environments, to support the disposal system safety case, including:
* The design, processing and disposability of glass and ceramic products from thermal treatment of plutonium and intermediate level wastes; where we have achieved: control over partitioning of radionuclides between component phases, an understanding of mechanisms of wasteform alteration in the hyperalkaline environment of a cementitious GDF; and the mechanism of the crystalline to amorphous phase transition induced by alpha recoil damage.
* The development of new low pH potassium magnesium phosphate cement systems suitable for encapsulation of reactive metals; where we have achieved a state of the art understanding of the mechanisms of binder phase formation, its impact on the product mechanical properties, radiation stability, and interaction with reactive metals, leading to hydrogen production.
* The application of advanced radio-imaging methodology for determining radionuclide transport in cement backfill, and mechanistic understanding of the immobilisation of problematic radionuclides in new functional cement barrier materials

From X-ray diffraction, scanning electron microscopy, and X-ray near edge structure and diffuse reflectance spectroscopic studies, it was found that approximately 0.03 formula units of mixed hexavalent and pentavalent U, but <0.001 formula units of Pu or Np if any, can be dilutely incorporated into the Ti site of rutile, TiO₂, sintered at 1400^oC in air. The valence of 0.01 formula units (f.u.) of U in Zr_(1-x)Y_xO_1-x/2 (x = 0.23-0.5) fired in air is mainly +6 from X-ray near edge spectroscopy and contradicts earlier X-ray photoelectron spectroscopy data. Some U⁵⁺ is also present from diffuse reflectance work. The valence of 0.01 f.u. of Np in the Zr_(1-x)Y_xO_1-x/2 (x = 0.23-0.5) is found to be +6 from diffuse reflectance study.

Sodalite (Na₈[AlSiO₄]₆Cl₂), a naturally occurring Cl-containing mineral, has long been regarded as a potential immobilisation matrix for the chloride salt wastes arising from pyrochemical reprocessing operations, as it allows for the conditioning of the waste salt as a whole without the need for any pre-treatment. Here the consolidation and densification of Sm-doped sodalite (as an analogue for AnCl₃) has been investigated with the aim of producing fully dense (i.e. > 95 % t.d.) ceramic monoliths via conventional cold-press-and-sinter techniques at temperatures of < 1000 °C. Microstructural analysis of pressed and sintered sodalite powders under these conditions is shown to produce poorly sintered, porous, inhomogeneous pellets. However, by the addition of a sodium aluminophosphate glass sintering aid, fully dense Sm-sodalite ceramic monoliths can successfully be produced by sintering at temperatures as low as 800 °C.

Multiphase ceramic waste forms show promise as a viable alternative to borosilicate glasses for nuclear waste treatment and disposal. These waste forms include hollandite, perovskite, pyrochlore, and zirconolite phases. We synthesized high phase purity pyrochlore (Nd₂Ti₂O₇) via. solid-state reaction. Praseodymium (Pr) and samarium (Sm) were chosen as dopants for our study. These dopants were added from x = 0.1-0.5 according to Nd_2-x(Pr,Sm)_xTi₂O₇. X-ray diffraction (XRD) was used to confirm phases present in the samples and determine lattice parameters. Scanning electron microscopy (SEM) and energy-dispersive spectrometer (EDS) were used to characterize the powders and morphology. SEM micrographs of fractured samples confirm the presence of single, homogenous phase and agree with the XRD and EDS data. XRD data showed the solid solution limit was not yet reached in this system. Lattice volume increased linearly with dopant concentration. Changes in the lattice parameters suggested expansion of the unit cell of the Pr-doped Nd₂Ti₂O₇ in the a and b directions. Selected samples were sintered via spark plasma sintering (SPS) to dense microstructures, which were examined using SEM and EDS. High-density values of about 95-98% of theoretical density were obtained for comparable grain size. Our results show the doping did not impact sinterability of the pyrochlore.

Lanthanum zirconate pyrochlore (La₂Zr₂O₇) have been extensively studied as a model for radioactive waste hosts, primarily due to its ability to recover from heavy ion radiation damage [1]. However, compartively little work has been undertaken examining the stability when substituted by radioactive elements, e.g. U/Pu.
Characterisation of uranium substitution based on the general formulation La₂Zr_2-xU_xO_7+δ , in both air and reducing H₂/N₂ atmospheres has been completed. The stability in both air and H₂/N₂ atmospheres has been shown by X-ray diffraction to be x = 0.8 in H₂/N₂ and 0.6 in air. At these upper limits of the substitution electron diffraction has been used to confirm the existence of the pryochlore superstructure, along with determining the presence of any other ordering. For example, in the H₂/N₂ samples the electron diffraction shows the presence of pyrochlore superstructure at x = 0.8, whereas X-ray diffraction indicates a fluorite structure has been formed. This difference is discussed with respect to both electron and X-ray diffraction, coupled with Raman spectroscopy, and the implications for use of this material as a host matrix for the stable storage of U/Pu.
References:
[1] G.R. Lumpkin et al. J. Phys.: Condens. Matter, 16 (2004) 8557

Pyrochlores (A₂B₂O₇) have been studied extensively for the immobilization of actinides. ^{1,2, 3} This project focuses on studying the effects of helium gas build-up and radiation damage due to alpha decay in ceramic nuclear waste form materials, specifically Gd₂Ti₂O₇ and Gd₂Zr₂O₇. Alpha decay in waste forms produces both helium atoms and atomic displacements that can be replicated using ion implantation and ion-beam irradiation techniques. The accumulation of helium can result in helium platelets and bubbles, and the accumulation of defects can result in the formation of dislocation loops, bubbles and phase transformations. It is well known that Gd₂Ti₂O₇ undergoes a pyrochlore to amorphous transformation at ~0.2 dpa at room temperature and that Gd₂Zr₂O₇ undergoes an order-to-disorder transformation from the pyrochlore to defect fluorite structure at ~0.4 dpa at low temperatures. ¹ Gd₂Ti₂O₇ and Gd₂Zr₂O₇ were synthesized by solid-state synthesis. These materials were irradiated with 7 MeV Au ions at room temperature to create a thick (~1 micron) amorphous state in Gd₂Ti₂O₇ and transform Gd₂Zr₂O₇ to the defect fluorite structure (~1 micron thickness), which are the structures of interest for long-term evaluation. These samples were subsequently implanted with 200 keV He ions to fluences of 2x10¹⁵ and 2x10¹⁶ ions/cm² at room temperature, which at the implantation peak correspond to expected He concentrations at 1000 and 100,000 years, respectively. The helium implanted samples have been irradiated with 7 Mev Au ions slightly evaluated temperatures to radiation doses corresponding to 50,000 years or more. The higher irradiation temperatures are used to accelerate the defect and helium interaction kinetics, simulating irradiation-induced microstructure evolution at longer time scales. After irradiation, the evolution of microstructure (dislocation loops, bubbles, etc.) were characterized using transmission electron microscopoy (TEM), scanning electron microscopy (SEM), and x-ray diffraction (XRD) techniques. The results of this study will be discussed.
1. Wang, S.-X. et al. Radiation stability of gadolinium zirconate: a waste form for plutonium disposition. Journal of materials research14, 4470-4473 (1999).
2. Lian, J. et al. Radiation-induced amorphization of rare-earth titanate pyrochlores. Physical Review B68, 134107 (2003).
3. Ewing, R.C., Weber, W.J. & Lian, J. Nuclear waste disposal—pyrochlore (A2B2O7): Nuclear waste form for the immobilization of plutonium and “minor” actinides. Journal of Applied Physics95, 5949-5971 (2004).

The immobilisation of actinides within the crystalline structure of ceramic waste forms seems to offer certain advantages over other waste forms (incl. borosilicate glasses and spent fuel). Monazite, LnPO₄ (Ln=La-Gd) is a promising ceramic as a waste form for actinides related to long-term safety aspects. For a reliable assessment of their long-term stabilty under conditions relevant to nuclear waste disposal deeper fundamental studies on these materials are necessary.
In the present communication we report on the atomic scale investigation of solid solution formation and on microstructural studies of the synthesis dependent sintering behaviour of monazite-type ceramics. A combined understanding concerning the structural and microstructural properties is of a great importance with regard to key parameters guiding the long-term stability of ceramic materials for safe nuclear disposal. Cluster formation in non ideal solid solutions in the case of minor actinides immobilisation could influence irradiation damages resistance, criticality aspects and dissolution behaviour. Microstructure impacts mechanical properties and corrosion resistance. Certain porosity in accordance with waste loading degree is needed to avoid crack formation due to swelling processes caused by He-evolution from α-decay reaction.
In recent studies we investigated the structural incorporation of Eu(III) in synthetic Eu(III) doped LaPO₄, and GdPO₄, as well as mixtures thereof by site-selective time-resolved laser fluorescence spectroscopy (TRLFS). Eu(III) was taken as an analogue for the long-lived trivalent actinides Pu(III), Am(III) and Cm(III) found in spent nuclear fuel. In the pure LaPO₄ and GdPO₄ monazites, Eu³⁺ substitutes the host cation sites in the highly ordered ceramic materials independent of the ionic radius of the host cation. However, excitation spectra of the mixed Eu(III)-doped (La,Gd)PO₄ monazite phases indicate a slight disordering of the crystal structure.
Additionally the present work was focused on the elaboration of (La,Eu)PO₄ solid solutions through wet chemistry routes then on the study of their densification. In this aim, investigations on in-situ sintering phenomena were carried out by the joint use of dilatometry and high temperature environmental scanning electron microscopy. Particularly, it allowed us to precise the conditions required for the densification of monazite but also to provide new insights on the microstructure development during heat treatment, including grain growth rate.

Coffinite (USiO₄) and associated solid solutions are expected to play an important role in the field of direct storage of spent nuclear fuels in underground repository since they could control the concentration of actinides in groundwaters. However, the thermodynamic properties associated with coffinite, especially the solubility, remain poorly defined. The few thermodynamic data related to coffinite formation or solubility reported in the literature are hardly reliable since none of them were determined experimentally from solubility measurements [1]-[3]. Solubility studies require pure single-phase USiO₄. Most of the natural samples contain coffinite as very small grain crystals [4] and in intimate intergrowths with large amounts of associated minerals. Moreover, for several decades persistent difficulties have been encountered in the preparation of pure single-phase synthetic coffinite. The precipitation of coffinite from a mixture of U(IV)-containing acidic solution and sodium metasilicate appeared as the most promising method to provide USiO₄ samples [5]. In this context, a thorough multiparametric study of the formation of synthetic coffinite was achieved. In this aim, the effect of various parameters such as pH, heating time, U/Si mole ratio and temperature were investigated to point out the optimal operating conditions for the preparation of coffinite.
The optimized protocol allowed the preparation of polyphased samples that contained mainly USiO₄ associated with oxide side products (amorphous SiO₂ and nanoparticles of UO₂). A purification process was developed that conduct to pure synthetic coffinite sample suitable for solubility experiments. The ion activity product in solution equilibrated with USiO₄ was determined by dissolution experiments conducted in 0.1 mol L^-1 HCl under Ar atmosphere at room temperature. The dissolution was congruent and a constant composition of the aqueous solution was reached after 50 day. The solubility product of coffinite was then determined (log*K_S,USiO₄ (298 K) = minus;6.14 ± 0.08). At low temperatures, coffinite appears to be less stable than the mixture of binary oxides, which is consistent with qualitative evidence from petrographic studies of uranium ore deposits. Finally a tentative mechanism was proposed to explain the formation of USiO₄ providing new insights concerning the formation of coffinite in environmental conditions.
[1] Guillaumont, R.; Fanghänel, T.; Fuger, J.; Grenthe, I.; Neck, V.; Palmer, D. A.; Rand, M. H., Chemical Thermodynamics Vol. 5. North Holland Elsevier Science Publishers B.V.: Amsterdam, The Netherlands, 2003, p 919. [2] Langmuir, D., Geochim. Cosmochim. A. 1978, 42, 547-569. [3] Hemingway, B. S., USGS Open file Report 82-619, 1982, p 89. [4] Deditius, A. P.; Utsunomiya, S.; Ewing, R. C., Chem. Geol. 2008, 251, 33-49. [5] Costin, D. T.; Mesbah, A.; Clavier, N.; Dacheux, N.; Poinssot, C.; Szenknect, S.; Ravaux, J., Inorg. Chem. 2011, 50, 11117-11126.

The radiation stability of candidate wasteforms is one of the greatest uncertainties when considering the long term geological disposal of HLW. Ceramic wasteforms have attracted a lot of interest due to their inherently low leach rates combined with high thermal and mechanical stability. Unlike glass wasteforms, there are often natural mineral analogues available (they may contain up to 30 wt. % of
U and Th impurities)¹ which can give an insight into the radiation stability of these phases on a geological timescale. Xenotime (YPO₄) and fluorapatite (Ca₅(PO₄)₃F) are two such phases of interest where both are rarely found to be metamict in nature unlike zircon.
Total scattering techniques are a powerful tool for studying amorphous and disordered materials; these use both the Bragg and diffuse scattering to give information on the long range ordering and local structure through the analysis of the pair distribution function (PDF). Samples of both xenotime and fluorapatite were irradiated with swift heavy ions (2.3 GeV Pb) to simulate the damage caused by daughter recoil nuclei from fission reactions and subsequently the PDFs were analysed. To support the experimental results, semi-empirical pair potentials were used to simulate intrinsic and extrinsic defect properties within these phases. Molecular dynamics simulations use these potentials to predict the extent of the damage cascade caused by a recoil nucleus and the degree of annealing that takes place at the periphery. Through a combination of experimental and computational techniques the radiation damage structure of xenotime and fluorapatite can be characterised.
References
[1] W. J. Weber et al. J. Mater. Res., 1998, 13, 1434-1484

Pyroprocessing of used nuclear fuel to separate out actinides during reprocessing creates radioactive salt waste, and sodalite-based glass-ceramics are strong candidates for immobilisation of these salts [1]. As such the leaching behavior of sodalite in water is of strong interest. Although a considerable amount of work has been done on the aqueous leaching of sodalite made by ceramic means, little work exists on the leachability of natural sodalite [2], thus uncertainties remain because of the difficulty of making sodalite free of other phases. A natural sodalite sample almost completely devoid of impurities has been studied in some detail using the PCT and MCC-1 type tests. PCT leach tests were indicative of near-congruent leaching and were well below the typical values obtained for HIPed sodalite ceramic samples.
Further, X-ray diffraction, scanning electron microscopy and Raman spectroscopy have been used to study the amorphization of sodalite following irradiation with gold and iodine ions. The amorphous regions produced by masking and irradiation have been progressively leached and analyzed to determine the effect of radiation damage on the leaching of sodalite.
[1] E. R. Vance, J. Davis, K. Olufson, I. Chironi, I. Karatchetvseva and I. Farnan, J. Nucl. Mater., 420, 396-404 (2012).
[2] T. Nakazawa, H. Kato, K. Okeda, S. Ueta and M. Mihara, in Scientific Basis for Nuclear Waste Management, eds., Materials Research Society, Warrendale, PA, USA, pp. 51-7 (2001).

Aqueous durability enhancement in borosilicate glasses containing 1 and 4 wt% Zn with increasing Zn content was evident in PCT-B tests, mainly via the normalised Zn leaching being much lower than those of other elements. In 14-day MCC-1 type leach tests conducted at 90^oC, surface alteration was very clear in the undoped glass via the formation of strongly altered amorphous material which tended to spall off the surface, but these effects were minimal in the doped glasses.
Surface layers on the leached glasses in the MCC-1 type tests were studied by grazing angle X-ray diffraction, transmission and electron microscopy and soft positron annihilation lifetime spectroscopy (PALS). No sign of crystallinity was detected by grazing incidence X-ray diffraction or electron microscopy of surface layers, and the surface material was very rich in silica for the Zn-free glass but although surface alteration was quite severe for the Zn-bearing glasses, no large compositional changes were evident. Bulk positron annihilation lifetime spectra (PALS) of glasses containing 0-4 wt% Zn could be analysed with three distinct lifetimes but did not show significant differences and the densities increased only slightly with increasing Zn content, indicating that the glass structure was not greatly influenced by the Zn content and did not contain voids due to foaming. Changes in surface PALS were evident as leaching progressed, indicating the formation of surface alteration layers with significant porosity. The results are compared with those obtained on the recently formulated Simple International glass.

The nuclear waste glass produced in the UK contains magnesium (Mg). This arises from the Mg cladding on the natural uranium fuel of Magnox reactors. The magnesium is entrained in the fission product fraction following fuel re-processing operations and is incorprated into the glass. At least two studies [1, 2] have shown that full component simulant UK glasses containing magnesium have poorer aqueous durability compared with similar calcium (Ca) based glasses. A series of simplified 6/7 component UK simulant waste glasses was prepared where the separate effect of Ca for Mg substitution in the glass could be studied. Five glasses with an alkaline earth (Ca/Mg) content of 6.5 mol% were prepared two with pure end member compositions and three with compositions of 25%, 50% and 75% Mg. All glasses showed similar macroscopic properties and glass transition temperatures. The only difference was their Mg/Ca content. The underlying glass dissolution, as measured by boron in solution following PCT-B tests over 7, 14, 28 and 112 days, was proportional to the Mg content and approximately one order of magnitude higher for Mg vs Ca glasses. Boron-11 NMR studies of the series of unleached glasses also showed a systematic increase in the amount of three-coordinated boron (B3) with increasing amounts of Mg. Similar ¹¹B NMR measurements of the leached material showed a null result in that the additional B3 was not preferentially leached from the Mg containing samples and the boron in the glasses dissolved ‘congruently&’ with respect to B3/B4 ratio. Quantitative proton NMR experiments indicated substantially more protonation on the surface than can be accounted for by direct replacement of leached ions and again there was approximately an order of magnitude greater number of protons on the surface of the Mg compared with the Ca glasses. NMR spectra of glasses leached with isotopically enriched water (H₂¹⁷O) showed substantial re-precipitation onto the glass surface together with significant retention of Si and Mg determined from solution measurements. Preliminary identification of the nature of the amorphous altered layers have been made. However, no crystalline phases were formed after 112 days of leaching.
1. P. K. Abraitis et al., The kinetics and mechanisms of simulated British Magnox waste glass dissolution as a function of pH, silicic acid activity and time in low temperature aqueous systems, Applied Geochemistry 15, 1399 (2000).
2. E. Curti, J. L. Crovisier, G. Morvan, A. M. Karpoff, Long-term corrosion of two nuclear waste reference glasses (MW and SON68): A kinetic and mineral alteration study, Applied Geochemistry 21, 1152 (2006).

The contribution of energetically reactive surface sites to the dissolution rate of spent nuclear fuel remains a key uncertainty in the safety case for geological disposal. We report a systematic study of how such surface features influence the dissolution rate of CeO₂ and ThO₂ ceramics, manufactured with a microstructure designed to approximate that of fuel grade UO₂. These model systems enable study of the subtle effects of reactive surface features without the complication of redox sensitivity as in the case of UO₂ The morphology of grain boundaries (natural features), surface facets (specimen preparation-induced features), and artefacts of surface polishing were investigated during dissolution. We show that preferential dissolution occurs at grain boundaries, resulting in grain boundary decohesion and enhanced dissolution rates. A strong crystallographic control was observed, with high misorientation angle grain boundaries retreating more rapidly than those with low misorientation angles, which may be due to the accommodation of defects in the grain boundary structure. Study of these simplified analogue systems provide evidence to support the hypothesis that grain boundaries play an important role in mediating the "instant release fraction" of spent fuel, and hence should be recognised in safety performance assessements for the geological disposal of spent fuel. Surface facets formed during the sample annealing process also exhibited a strong crystallographic control and were found to dissolve rapidly on initial contact with dissolution medium. Defects and strain induced during sample polishing caused an overestimation of the dissolution rate, by up to 3 orders of magnitude.

Based on the importance of the carnotite phase, K₂(UO₂)₂(VO₄)₂.xH₂O in some oxidized U ore deposits [1], low solubility uranyl vanadates might control the mobility and the ultimate distribution of U in oxidative environment [2]. The determination of the thermodynamic data associated with this phase thus appears to be a crucial step toward understanding the origin of uranium deposits or to forecast the fate of uranium either in mine tailing or in natural media. A parallel approach based on the study of both synthetic and natural carnotite samples was set up to evaluate its solubility constant. The synthetic carnotite sample was obtained by dry chemistry route [3]. The natural sample comes from the Montrose mine (Colorado), and contains carnotite associated with navajoite, V₂O₅.3H₂O. The two solids were first extensively characterized and compared by means of XRD, SEM, X-EDS analyses, Raman spectroscopy and BET measurements. The synthetic sample was suitable for solubility experiments undertaken in several acidic media (1 M HCl, HNO₃ and H₂SO₄) between 277 K and 333 K. The kinetic of dissolution was studied far from equilibrium under dynamic conditions. The apparent activation energy of the dissolution mechanism reached 49 ± 2 kJ mol^-1, whatever the dissolution medium. Under static conditions, an apparent equilibrium was reached for experiments conducted at 277 K. At higher temperatures, the precipitation of vanadate was observed. The analysis of the solid phase revealed the presence of vanadium oxide formed during the dissolution process. Due to kinetic hindering, such secondary phase was not formed at 277 K and an apparent equilibrium constant was derived for carnotite (log K°_app (277 K) = -62.5 ± 0.4). The behaviour of the synthetic sample during dissolution was compared to the natural one. The dissolution of the natural sample was found to be un-congruent due to the presence of the additional navajoite phase. In both cases, a pseudo-equilibrium was reached after 10 days of dissolution, controlled by the solubility of the two phases. Such results highlighted the importance of using synthetic samples in order to better constrain solubility experiment and derive reliable thermodynamic interpretation.
[1] Dahlkamp, F.J., Uranium deposits of the world, Springer Reference, Eds Springer, 2013, p 2400. [2] Gorman-Lewis, D.; Burns, P. C.; Fein, J. B., J. Chem. Thermo. 2008, 40, (3), 335-352. [3] Abraham, F.; Dion, C.; Saadi, M., J. Mater. Chem. 1993, 3, (5), 459-463.

Actinides mixed dioxides are considered as reference fuel materials in several nuclear reactors concepts (including Gen III and Gen IV) or as matrices for the recycling of minor actinides, either directly in the core or in fertile blankets. Based on their potential reprocessing or their long-term storage, the consequences of structural and microstructural parameters coming from their life cycle (including preparation) on their chemical durability have to be carefully considered.
This study was first focused on the dissolution of (M^IV,Ln^III)O₂ and (M^IV,M^IV)O₂ samples (with M^IV = Th, U, Ce ; Ln^III = La-Yb), as model compounds for future mixed oxide fuels, and highlight the often described role of conventional parameters (temperature, acidity or nature of dissolution medium, hellip;). Moreover, a particular attention was also paid to the less studied structural parameters (crystal structure, crystal defects, crystallite size, role of oxygen vacancies associated to the incorporation of aliovalent elements, hellip;) and microstructural ones (chemical composition and homogeneity, crystallization state, density, pore size and distribution, density and cohesion of the grain boundaries, hellip;) [1]. Among them, the important role of chemical homogeneity at the solid/liquid interface was clearly evidenced for (M^IV,Ln^III)O₂ and Th_1-xU_xO₂ systems ; the kinetics of alteration being slowing down when improving the distribution of cations at the microscopic scale, namely by using wet chemistry routes of preparation [2]. The presence of crystal defects in the studied compounds, influences the kinetics as much as the acidity of the leachate while crystallite size, grain size or densification rate remain second order parameters [3]. Nevertheless, these latter parameters must be taken into account carefully when studying the evolution of the solid/solution interface. Indeed, ESEM observations performed in operando during the dissolution process allowed imaging the preferential alteration zones for several solids which can be located either at the grain boundaries, triple junctions or within the grains leading to the formation of intragranular corrosion pits. Also, it allowed pointing out the role of surface heterogeneities on the dissolution kinetics as well as the strong evolution of the reactive surface area during the dissolution of the ceramics [4]. This information appears of main importance when working with normalized dissolution rates and led us to consider limit cases for using such variables.
[1] D. Horlait et al., Inorg. Chem., 50, 7150 (2011) & 51, 3868 (2012) ; J. Nucl. Mater., 429, 237 (2012). [2] N. Hingant, et al., J. Nucl. Mater., 385, 400 (2009). [3] L. Claparede et al., Inorg. Chem., 50, 9059 (2011) & 50, 11702 (2011). [4] S. Szenknect et al., J. Phys. Chem. C, 116, 12027 (2012) & D. Horlait et al., J. Mater. Chem. A, 2, 5193 (2014).

In all potential European deep geological repositories for spent nuclear fuel, radium becomes a major dose contributor at long time perspective (after ~300 k years). Thermodynamic calculations indicate that the formation of solid solutions can reduce the Ra solubility by several orders of magnitude compared to the solubility of pure RaSO₄. However, due to a lack of reliable thermodynamic and experimental data for the expected scenario at close-to equilibrium conditions, the solid solution system RaSO₄-BaSO₄-H₂O has so far not been considered in long term safety assessments for nuclear waste repositories.
Here, we have combined a macroscopic experimental approach with micoanalytical techniques (SEM, TOF-SIMS, FIB/TEM, LEAP), atomistic simulations and thermodynamic modeling (GEMS) to study in detail how a Ra containing aqueous solution will equilibrate with solid BaSO₄ at room temperature. Batch recrystallization experiments were carried out with two types of barite at an initial Ra/Ba ratio of 0.3 (5 x 10^-6 mol/L Ra) at near neutral pH-conditions. Depending on the solid/liquid ratio, a significant decrease of the Ra concentration by more than 99 % occurred within the first 70 days of the experiment. A final steady state was reached after more than 800 days for all experiments. Micro-analytical characterization of the formed radiobarite indicates, that the entire barite crystals have equilibrated with the Ra containing solution.
The thermodynamic mixing parameters for the solid solution have been derived from DFT calculations to be W_BaRa = 2.50 ± 1.00 kJ/mol indicating a non-ideal solid solution. The results of thermodynamic modeling indicate a good agreement of the apparent final Ra(aq) equilibrium concentration from experimental data with the computed W_BaRa and a solubility product of the RaSO₄ end-member of log K_SP(RaSO₄) = -10.41. In summary, we provide a coherent microscopic - macroscopic picture of radiobarite solid solution formation as a scientific bais for its application in long-term safety assessments.

Mixed actinide dioxides are currently used as fuels in Pressurized Water Reactors (PWR) (including Gen III, EPR) and also stand as potential candidates for several Gen IV concepts including Sodium-cooled Fast Reactor (SFR) or Gas-cooled Fast Reactor (GFR). In this field, the reprocessing of minor actinides coming from nuclear spent fuel into mixed-oxide fuels or in UO₂-based blankets surrounding the core is often considered. In this context, it is thus necessary to gain insight into the mechanisms of dissolution of the new compounds in the frame of reprocessing operations. In addition, the chemical durability of such compounds must be also tested under the chemical conditions encountered in the vicinity of direct disposal repository setting, as several countries choose the “open cycle” option.
While uranium-based mixed oxides are interesting materials for several concepts of nuclear reactors, the effects of trivalent elements in the fluorite structure and of the redox reactions at the solid/solution interface during the dissolution of such compounds remain widely unknown. In order to underline the influence of those parameters, various dissolution experiments were carried out on different compositions (i.e. U_0.75Ce_0.25O₂, U_0.75Th_0.25O₂, Th_0.75Nd_0.25O_1.875 and U_0.75Ln_0.25O_1.875 with Ln = Nd, Gd). These tests were performed on sintered pellets in nitric acid solutions (from 10^-2 M to 4 M) at different temperatures (22°C-90°C) under dynamic conditions. Therefore, a multiparametric study of the dissolution kinetics was then achieved in order to determine the partial order of dissolution reaction regarding to H₃O⁺ activity and the activation energy.
The obtained results gave evidence of the strong dependency of the dissolution rate with the nitric acid concentration. In fact, for concentrations higher than 2 M, the dissolution process is almost controlled by the oxidation of U(IV) to U(VI). On the contrary the effect of oxygen vacancies becomes predominant for acid concentrations lower than 0.5 M. Under these conditions, the systems containing trivalent elements exhibited the lowest chemical durability. The partial order of the reaction regarding to the H₃O⁺ activity reached n = 1.7 for U_0.75Th_0.25O₂ and was found to be constant over the entire acidity range. A variation of the partial order of the dissolution reaction regarding to H₃O⁺ was observed along the nitric acid concentration range for U_0.75Ln_0.25O_1.875 solid solutions, that underlines the predominance of different controlling reactions in the dissolution mechanism: surface-controlling reaction involving H₃O⁺ at low pH (i.e. n < 1) for C_HNO3 le; 0.5 M and uranium(IV) oxidation (i.e. n > 1) for C_HNO3 ge; 1 M.

It is well known that the radiation emitted by spent nuclear fuel (SNF) will produce radiolysis products in the presence of water vapor or a thin-film of water (including OHbull; and Hbull; radicals, O₂, e_aq, H₂O₂, H₂, and O₂). However, for these products to increase or change the rate of SNF degradation and result in the release of radionuclides requires understanding the processes that might occur in these interfacial regions. The mixed potential model (MPM) is used to calculate the SNF corrosion rates for a wide range of disposal environments to provide the source term radionuclide release rates for generic repository concepts. The SNF corrosion rate is calculated for chemical and oxidative dissolution mechanisms using mixed potential theory to account for all relevant redox reactions at the SNF surface, including those involving oxidants produced by solution radiolysis and provided by the radiolysis model (RM). The RM calculates the concentration of species generated at any specific time and location from the surface of the SNF.
We have developed a strategy for coupling three process level models to produce an integrated SNF degradation model. The G-value for H₂O₂ production (G_cond) to be used in the MPM needs to account for intermediate spur reactions. The effects of these intermediate reactions on [H₂O₂] are accounted for in the Radiolysis Model. We describe the methods for applying RM calculations that encompass the effects of these fast interactions on [H₂O₂] as the solution composition evolves during successive MPM iterations and then represent the steady-state [H₂O₂] in terms of an “effective instantaneous or conditional” generation value (G_cond). It is anticipated that the value of G_cond will change slowly as the reaction progresses through several iterations of the MPM as changes in the nature of fuel surface occur. Sensitivity runs with RM indicate significant changes in G-value can occur over narrow composition ranges.

We report on a subtle global feature of the mass action kinetics equations for water radiolysis that results in predictions of a critical behavior in H₂O₂ and associated radical concentrations. While radiolysis kinetics has been studied extensively in the past, it is only in recent years that high speed computing has allowed the rapid exploration of the solution behavior over widely varying dose and compositional conditions. We explore the radiolytic production of H₂O₂ under various externally fixed conditions of molecular H₂ and O₂ that have been regarded as problematic in the literature - specifically, “jumps” in predicted concentrations, and inconsistencies between predictions and experiments have been reported for alpha radiolysis. We computationally map-out a critical concentration behavior for alpha radiolysis kinetics using a comprehensive set of reactions. We then show that all features of interest are accurately reproduced with 15 reactions. An analytical solution for steady-state concentrations of the 15 reactions reveals regions in [H₂] and [O₂] where the H₂O₂ concentration is not unique - both stable and unstable concentrations exist. The boundary of this region can be characterized analytically as a function of G-values and rate constants independent of dose rate. Physically, the boundary can understood as separating a region where a steady-state H₂O₂ concentration exists, from one where it does not exist without a direct decomposition reaction. We show that this behavior is consistent with reported alpha radiolysis data and that no such behavior should occur for gamma radiolysis. We suggest experiments that could verify or discredit a critical concentration behavior for alpha radiolysis and could place more restrictive ranges on G-values from derived relationships between them.

The waste forms internationally mostly considered are glass, spent fuel and cement (as a matrix). The chemical durability of these waste forms, in particular in the long-term (thousands of years and longer) and in conditions of relevance for the envisaged disposal, has been studied for several decades. Over the past years an international common agreement has grown on the existence of a residual or long-term, very small dissolution rate for glass and spent fuel. The origin of these long-term rates is different, and not fully understood. In case of glass amongst other water diffusion into the glass surface or precipitation reactions onto the glass surface might be rate controlling, whereas in case of spent fuel the hydrogen formed due to container corrosion would very much limit the corrosion. The very small long-term dissolution rates would lead to lifetimes in many disposal conditions of hundreds of thousands of years and more, but an in-depth understanding of the underlying hypotheses and validation under realistic in situ conditions has not been achieved yet. In case of cement, the knowledge of the long-term durability is quite different, amongst other because cement is constituted of different mineral phases, each of them having a different reaction kinetics. Yet, there is no common experimental methodology or approach to study the long-term dissolution of cement in disposal conditions.
This paper will focus on the status of the understanding of the processes controlling the long-term dissolution of these waste forms, and on the status of the modeling of the long-term dissolution. The impact of the interacting environment will be discussed: pure solutions, solutions loaded with corrosion products or clay. The impact of a cementitious disposal environment will also be discussed.
Next, the contribution of these final dissolution rates to the safety assessment and safety case of candidate disposal sites will be discussed. We will also formulate recommendations for further R&D.
Reference:
P. Van Iseghem, " Corrosion issues of radioactive waste packages in geological disposal systems". In "Nuclear corrosion science and engineering", Edited by D. Feron, Woodhead Publishing Series in energy No 22, 2012, 939-987.

The radionuclide migration from repository and their expansion into the biosphere depend on sorption properties of backfill materials. These properties determine the forms of stored radionuclide, especially in long-time contact between them. With time the physicochemical forms of radionuclides change by chemical actions of adsorption and coprecipitation with barrier materials. Consequently radionuclides precipitate on solid phase and migration retained.
The experiments were carried for the evaluation of Cs¹³⁷, Sr⁹⁰ and Co⁶⁰ forms in natural materials, which serve as backfill in near surface repositories for radioactive wastes. The experiments were carried on with covering silt and sand of glaciolacustrine origin. The covering silt contained quartz, feldspar and clay minerals (kaolinite group, illite, vermiculite, montmorillonite). The sand contained quartz and small amounts of clay minerals, feldspar, ferrum hydroxides and manganese compounds. In the experiments, the samples of materials were treated by solutions of different compositions. Before then these samples were contact with radionuclides solutions during 2 weeks, 1, 2, 4 and 6 months accordingly.
In the samples of backfill materials radionuclides of Cs¹³⁷, Sr⁹⁰ and Co⁶⁰ were in different forms: mobile and fixed. The ratio of these forms also differs in different samples. After the long-term contact of radionuclides with covering silt and sand occurs the transformation of radionuclides forms. The main feature for all radionuclide was the decreasing of ratio of mobile forms during their contact with materials increase. Simultaneously increased tightly bound forms of radionuclides. With time the fixed forms become dominate.
The time of transformations processes in the materials differed, which could be connected with mechanism of radionuclide absorption by the solid phase: intrusion to scale of clay minerals, ion exchanging and chemical coprecipitation. The time of transformation of radionuclide forms increases in such way in line Sr < Co < Cs and in line sand< covering silt. A great part of cesium ran to tightly bound form in a first weeks of contact. The ratio of cesium in tightly bound form achieved its maximum (95%) already after two weeks of contact. Strontium and cobalt compared to cesium were more mobile and their transformation into tightly bound form occurred gradually during the experiment. The maximal abundance of these radionuclides in tightly bound form was 68-71 % and 74-85 % respectively.
The results of experiments allowed to estimate hardness of radionuclide fixation by barrier materials for near surface repositories and to conclude that radionuclide migration in backfill increases in operational period.

The disposal of nuclear wastes accumulated in large underground tanks at the Hanford Nuclear Reservation in Richland, WA is a priority goal of the United States Department of Energy (U.S. DOE). DOE plans to first separate the high level waste (HLW) from the low activity waste (LAW), and isolate these wastes in vitrified (glass) waste forms for long term disposal. It has been shown that a ratio of 0.1 to 0.5 of Fe²⁺/Fe³⁺ is ideal for glass melting conditions. Foaming may occur if the melt is too oxidizing and a metallic phase interfering with instruments may form if the melt is too reductive. In this study wet chemical and UV-Vis spectroscopy were performed to quantify Fe²⁺/Fe³⁺ ratios and total iron content of quenched alkali alumino-boro-silicate (simulated nuclear waste glasses), applying a well-established colorimetric method. 1,10 phenanthroline (C₁₂H₈N₂, ortho-phenanthroline or o-Phen), a tri-cyclic nitrogen heterocyclic compound, reacts with Fe²⁺ to form a strongly colored complex with a molar absorptivity of 11,100 L/mol#8729;cm. In this work, 1,10 phenanthroline was mixed with dissolved simulated nuclear waste glass powder. The resulting solution was equilibrated, and the absorbance was measured at 520 nm. The solution was then equilibrated with a mild reducing agent, and the solution absorbance was measured again at 520 nm. The absorbance values allowed for calculation of the Fe²⁺/Fe³⁺ ratio and the total iron content in the glasses. All quenched glasses analyzed showed a Fe²⁺/Fe³⁺ ratio between 0.06 (± 0.01) and 0.04(± 0.01). These values are consistent with those obtained for similar glass compositions melted under analogous conditions, indicating a composition of ca. 94-96% Fe³⁺.

The determination of the long-term stability and corrosion of vitrified nuclear waste is an important aspect of research for the U.S. Department of Energy (DOE). It is necessary to understand the rate and mechanisms of corrosion in Nuclear Waste Glass (NWGs) to determine whether or not the glassy matrix will be able to retain radionuclides for the appropriate repository performance time period. Glass corrosion and the rates of glass corrosions are typically determined by chemical, optical and spectrometric methods. Energy-dispersive X-ray spectrometry (EDS) and Wavelength-dispersive C-ray spectrometry (WDS) are common and powerful methods utilized in the examination of the chemographic difference between corroded and uncorroded NGWs. In this work, a new sampling method utilizing EDS and WDS for the measurement of Si, Al, Fe, Mg, Ca, Na, K and B in NWGs was created, calibrated using a set of natural and synthetic standard glasses and minerals that have compositions similar to the glasses under examination, and verified for both accuracy and precision by repeated measurements of pristine samples of Savannah River International Standard Glass. Pristine monolithic samples were cut along their diagonal, mounted vertically in low temperature and quick setting epoxy, and then polished to fine (3 mu;m) grit before being analyzed using a JEOL JXA-8500F Field Emission Electron Probe Microanalyzer (FE-EPMA), to achieve representative sampling of the glass&’s entire surface. The particular instrument used was equipped with a Schottky-type field emission electron source, five WDS spectrometers, and a silicon-drift energy dispersive X-ray spectrometer. The cross-sectioned surfaces were digitally gridded, and three to five spots per each grid were selected at random. Scanning Electron Microscopy (SEM) was employed to measure the thickness of the corrosion layers on the corroded glasses. The accuracy and precision of analyses samples mounted and analyzed following the protocol described above were compared to those obtained from simple surface analysis of the same sample.

Due to accidents in Fukushima I Nuclear Power Plant, Japan on March, 2011, several kinds of radioactive elements caused by depressurization of the containments of the three reactors were emitted and distributed mainly in Fukushima prefecture, Japan. Among the radioactive elements, the decontamination of ¹³⁷Cs is urgently necessary to make volume reduction of radioactive waste. In order to develop volume reduction techniques, chemical bonding states of Cs adsorbed in clay minerals, e.g. vermiculite, has been studied by using synchrotron radiation x-ray photoemission spectroscopy (SR-XPS) in our research group. In this presentation, an interpretation of data obtained by surface charge modulation using an electron flood gun during SR-XPS is proposed to discuss chemical bonding states of Cs in the vermiculite.
SR-XPS experiments were conducted at the surface chemistry experimental station (SUREAC2000) of BL23SU in SPring-8. The natural vermiculite produced in Ono town, Fukushima prefecture, Japan was processed to adsorb Cs by dipping in a non-radioactive CsCl solution. The Cs concentration in the vermiculite was estimated to be 2.1 wt%. The other Cs compounds, e.g. CsClO₄, CsNO₃, CsI and so on, were also used as experimental samples. The energy of synchrotron radiation was 1486.6 eV, identical with the Al-K_α line for convenience&’ sake of comparison with experiments in laboratories. Surface charge on the samples was changed by using an electron flood gun during SR-XPS.
In general, an Auger parameter is not influenced by surface charging because of cancelling out of XPS peak shift. So Auger parameters of Cs in Cs-contained vermiculite and Cs compounds were evaluated and compared to investigate similarity of chemical bonding states. The Auger parameter of Cs of CsClO₄ was closest to that of Cs-contained vermiculite. This implies the similarity of chemical bonding state of Cs between Cs-contained vermiculite and CsClO₄. Indeed, the steric structure of perchloric acid ion ClO₄^- is tetrahedral. The Cl atom is surrounded by four O atoms so that the Cs atom interacts with the O atom. Cs atoms in the vermiculite may also interact with O atoms contained in the vermiculite. In order to confirm it, chemical shift of Cs-3d core level was measured by high resolution SR-XPS. A Cs-3d photoemission spectrum includes four components. As a matter of fact, binding energies of these components are close to that of CsClO₄. The component having the highest binding energy was only shifted by using an electron flood gun, suggesting most ionic chemical bonding. The other three components were not shifted as well as K atoms originally-contained between phyllosilicate layers. As a consequence, the shifted component may be originated from Cs hydrated in water of weathered wide crevice between phyllosilicate layers. The other unshifted three components are corresponding to covalent interaction with O and Si atoms in narrow phyllosilicate interlayers.

In rock matrix radionuclides migrate via diffusion in the pores (fractures, inter- and intragranular pores) . Typically, in crystalline rock, the majority of pore sizes are in nano-scale and have a significant effect on the migration of radionuclides. Small-scale structures are also closely linked to the surface area available for sorption of radionuclides to take place. For this study, we chose two types of rock for nano-tomographic imaging: one from Olkiluoto, close to a future repository site for high-level nuclear waste, and one from Sievi, a previous candidate for a repository site. We chose the samples so that the tonalitic deep-rock sample from Sievi was more altered than the mica-gneiss sample from Olkiluoto. Our aim was to see the effects of alteration on the microstructure and pore structure in nanometer scale.
Out of each sample we separated three subsamples consisting of a single mineral using heavy-liquid separation and imaged them for a total of 15 times with both micro- and nano-tomography. As the feldspars we chose plagioclase from the Olkiluoto sample and potassium feldspar from the Sievi sample. As the dark minerals we chose biotite from the Olkiluoto sample and amphibole from the Sievi sample. Cordierite was also separated from the Olkiluoto sample and Quartz from the Sievi sample.
We were able to see the fine structure of these samples. The Sievi amphibole contained more interlamellar pores and alteration products than the Olkiluoto biotite, both a sign of heavier alteration. Both clearly showed the typical lamellar structure of a mica mineral. Both feldspars showed alteration products, but there were more of those in the Sievi feldspar, as well as more pores. Neither had a connected pore network visible with the resolution used (voxel size 65 nm, spatial resolution 150 nm). However, it is expected that the visible pores are connected via smaller ones. The cordierite sample showed a typical structure of transformed cordierite, and the quartz sample, which was identified as albite in the elements study, was almost homogeneous, as expected. As a result of this study we can conclude that nano-tomography is a relatively good tool for analyzing effects of alteration on minerals and pores.

In long-term safety assessment of geological disposal system, it is necessary to evaluate the impact on the radionuclides migration where groundwater flow and water composition are changed with decreasing a depth of the repository by uplift (two types: uniform and tilted) and denudation. We developed an integrated safety assessment methodology for uplift and denudation where radionuclides migration analysis was combined with calculation results of groundwater flow and salt concentration distribution in a disposal site and of long-term transition of engineered barriers to estimate their properties and migration parameters. This methodology was applied to the assumed disposal site composed of sedimentary rocks with uplift and denudation. Migration parameters such as migration distance and mean velocity along its migration pathway are determined based on the particle trajectory analysis from an assumed repository to the outlet of natural barrier where the depth is 40m from ground surface, under the condition of distributions of groundwater velocity and salt concentration at a given time. Distribution coefficients for elements along the migration pathway, which is characteristic with the combination of geology (three types of sedimentary rock) and water composition (two types of rainwater and salt water), are determined with reference to the existing distribution coefficient database. Migration parameters in engineered barrier such as diffusion coefficient and distribution coefficient in buffer material were determined based on systemized procedure using coupled mass transport and chemical reaction for each engineered barrier material such as carbon steel, bentonite clay and cement under surrounding pore water condition of excavation disturbed zone. The migration fluxes at the outlet of natural barrier were evaluated in combination of denudation, anisotropic water permeability, permeability variation due to the depth change by denudation in case study in addition to the type of uplift. Compared with calculated results of peak time of migration fluxes at the outlet of the natural barrier between uniform and tilted uplift, the peak time of key radionuclides, Se-79, Cs-135 and Np-237, in the case of tilted uplift is earlier than those in the case of uniform uplift. Maximum fluxes of all radionuclides are 6 orders of magnitude greater than those at uniform lift. This is because groundwater velocity is approximately five times larger due to the increase of hydraulic gradient by the tilted uplift which leads to 25% longer migration distance, while the migration distance is 25% shorter and groundwater velocity does not change for the uniform uplift. In a paper, the results of other case studies will be shown in addition to the above mentioned and then the requirements for future investigation of natural barrier related to uplift and denudation at a candidate site will be presented.

Calcium chlorosilicate (Ca₃(SiO₄)Cl₂) is seen as a potential host phase for the immobilisation of Cl-rich wastes arising from pyrochemical reprocessing, a waste stream often containing a mix of both di- and trivalent cations. Substitution of trivalent cations into the lattice requires some form of charge compensation to ensure the lattice remains charge neutral overall. Whilst previous work has only examined this through the formation of Ca vacancies, this study investigates the feasibility of charge-balancing via the substitution of a monovalent cation onto the Ca sites of the lattice. To that end, a series of static lattice calculations were performed to determine the site selectivity of monovalent cations of differing size when substituted onto the Ca sites of the calcium chlorosilicate lattice and the solution energies for the overall substitution processes compared with those for charge compensation via vacancy formation. In all cases the monovalent charge-balancing species shows a clear preference for substitution onto the Ca1 site in the calcium chlorosilicate lattice. The solution energy of the substitution process increases with the increasing ionic radii of both the mono- and trivalent species as the steric stresses associated with substitution of cations larger than the Ca²⁺ host increase. As such, only charge-balancing using Li⁺, Na⁺ or K⁺ is more favourable than via formation of a Ca vacancy.

Apatites are often seen as good potential candidates for the immobilisation of halide-rich wastes. In particular, phosphate apatites have received much attention in recent years, however, their synthesis often produces complicated multi-phase systems, with a number of secondary phases forming. Calcium vanadinite (Ca₅(VO₄)₃Cl) demonstrates a much simpler phase system, with only a single Ca₂V₂O₇ secondary phase which can easily be retarded by the addition of excess CaCl₂. However, when doping with SmCl₃ (as an inactive analogue for AnCl₃) the Sm forms a wakefieldite (SmVO₄) phase rather than being immobilised within the vanadinite, a result of having to form an energetically unfavourable Ca vacancy in order for the lattice to remain neutral overall. It has been postulated that charge-balancing the lattice via co-substitution of a monovalent cation will be less disfavoured and therefore help stabilise formation of a (Ca_5-2xSm_xA_x)(VO₄)₃Cl solid solution (A = monovalent cation). This has been investigated using a combined modelling and experimental approach. Static lattice calculations performed using Li⁺, Na⁺ and K⁺ as charge-balancing species have shown the energy cost to be less than half that of charge-balancing via formation of a Ca vacancy. As a result, solid state synthesis of (Ca_5-2xSm_xLi_x)(VO₄)₃Cl, (Ca_5#8209;2xSm_xNa_x)(VO₄)₃Cl and (Ca_5-2xSm_xK_x)(VO₄)₃Cl solid solutions have been trialled, and analysis of the resulting products has shown a significant reduction in both the SmVO₄ and Ca₂V₂O₇ secondary phases across all dopant levels.

Calcium Silicate Hydrate (CSH) is the main component of cementitious materials, and it is important for the migration assessment of radionuclides to clarify the interaction of CSH and radionuclides. Especially, Iodine-129 would provide large contributions to the dose evaluation because of its long half-life and its low ability to adsorb on engineered barrier. However, authors have reported that iodide ions are incorporated into CSH with the formation process of CSH. Besides, since hydrotalcite also has the ability of anion exchange, such mineral might be used as a concrete aggregate. In this study, the conflict of iodide ions in the sorption processes was examined under a co-existing condition of CSH and hydrotalcite.
In the experiment, the weight ratio of CSH and hydrotalcite was 1.0. CSH was synthesized with CaO and fumed silica, and Ca/Si molar ratios of CSH were set to 0.4, 0.8, 1.2 and 1.6. The immersion solutions were pure water or 0.6 M NaCl solution containing 0.5 mM iodide ions. The total volume of the solution was 30 ml, and the liquid/solid weight ratios were set to 20, 15 and 10. After 14 days, the concentration of iodide ions in liquid phase and Raman spectra for solid phase were measured.
As a result, iodide ions sorbed on solid phase mixed CSH and hydrotalcite. The distribution coefficients, K_d, were in the range from 7 to 23 L/kg for the condition of pure water. Besides, for 0.6 M NaCl, K_d were 1 to 2 L/kg. This significant decrease of K_d is due to the influence of Cl ions competing with iodide ions. The results of Raman spectra showed Al-O-Al and Al-O-Si bond of hydrotalcite, and Si-O-Si bond of CSH, respectively. These mean the structures of hydrotalcite and CSH are maintained during the hydration of solids and the sorption of iodide ions. Furthermore, K_d for CSH only were almost the same as the co-existing condition. These results suggest that hydrotalcite undergoes the retardation effect of anion nuclides such as I-129 even if in a condition co-existing with CSH.

Zirconia based pyrochlores are very promising waste forms for the safe disposal of minor actinides (MA = Np, Am, Cm) and plutonium as they are highly durable against aqueous alteration and radiation damages [1,2]. Pyrochlores are chemically very flexible structures with the general formula A₂B₂X₆Y, which are known to form solid solutions. Here, we have focused on ZrO₂ based pyrochlores with A = La, Nd, Eu, Pu, Cm; B = Zr; X = O; Y = O.
Based on a wet chemical synthesis route, which was developed using non-radioactive surrogates, zirconia based pyrochlores with Pu and Cm were studied. A simultaneous co-precipitation of the Nd-, Zr- and Pu-hydroxides was applied to yield highly homogeneous pyrochlores. The crystallisation of the (Nd,Pu)₂Zr₂O₇ pyrochlore solid solution was carried out during a sintering step under reducing conditions. Several samples with a Pu concentration of 5 mol% and 10 mol% were synthesised. Powder X-ray diffraction (XRD) measurements showed a systematic shift of the reflexes of the pyrochlore with the different Pu contents, indicating the structural uptake of Pu. However, the presence of additional PuO₂ could not be excluded by XRD measurements. Furthermore, scanning electron microscopy (SEM) was carried out on a Nd_1.9Pu_0.1Zr₂O₇ pellet. The SEM images and energy-dispersive X-ray spectroscopy (EDX) confirmed a homogeneous distribution of Pu within the whole pellet.
An insight into the structural environment of MA within the pyrochlore was obtained using time resolved laser fluorescence spectroscopy (TRLFS) [3]. Fluorescence spectra were obtained by UV and direct excitation of Cm and Eu doped La₂Zr₂O₇. The observed splitting of the major species after direct excitation indicates the presence of Cm or Eu at the A position of the pyrochlore crystal structure. The maximum splitting of the ⁵D₀ - ⁷F_1,2 transition for the minor species in the pyrochlore structure is caused by disordered vacancies in the nearby environment of the A position, which are typical for the defect fluorite structure. Zirconia based pyrochlores are known to transform to the defect fluorite structure due to radiation damage. In future studies, the distinction of the defect fluorite and pyrochlore environment may allow the quantification of radiation damages in zirconia based pyrochlores via TRLFS.
[1] R.C. Ewing, C. R. Geosci., 343 (2011) 219-229. [2] G.R. Lumpkin, Elements, 2 (2006) 365-372. [3] K.S. Holliday, et al., (2013) J. Nucl. Mater. 433, 479-485.

CST was developed and engineered by Universal Oil Products (UOP) as the active ingredient in a a commercially available ion exchanger known as IE-911 (IONSIV^®).¹ IONSIV^® is highly selective for caesium over a wide pH range in the presence of a number of competing ions and therefore large amounts of IONSIV^® have been employed for caesium clean up¹ and await final disposal or storage in intermediate depositories around the globe.² It was found in a recent study conducted by Tzu-Chen (2013)³ that hot isostatic pressing (HIPing) of Cs-loaded IONSIV^® creates a robust immobilised waste form suitable for final geological storage.³
The major Cs containing phase was identified as Cs₂TiNb₆O_18, a pyrochlore analogue.⁴ Studies suggested that Cs₂TiNb₆O₁₈ had the potential to be an excellent waste form, demonstrating superior leach rate properties to other waste forms.³ The aim of this study is to elucidate whether the Ba²⁺ transmutation product of ¹³⁷Cs⁺ will remain in the Cs₂TiNb₆O₁₈ structure and importantly will the leach rates demonstrated in the parent phase remain consistent after transmutation. The Ba²⁺ cation is somewhat smaller that Cs⁺ and therefore the Ba²⁺ ion may not be compatible with the Cs⁺ site in the structure. Furthermore there must be some sort of charge compensating mechanism to balance the charge in the structure after the Cs⁺ has decayed. It is reported that hollandite, the major Cs-containing phase in SYNROC is able to accommodate for Ba²⁺ and balance for the change in charge through the simultaneous reduction of Ti⁴⁺ to Ti³⁺ or Fe³⁺ to Fe²⁺.⁵ It was envisaged a similar process could take place in Cs₂TiNb₆O₁₈ where the Nb⁵⁺ and Ti⁴⁺ could reduce to Nb⁴⁺ and Ti³⁺ respectively, acting as 'electron traps' for the β^- particle released in transmutation rendering the overall charge neutral.³
A number of Cs_2-xBa_xTi^(IV)_1+xNb^(V)_6-xO₁₈, Cs_2-xBa_xTiNb^(V)_6-xNb^(IV)_xO₁₈ and Cs_2-xBa_xTi^(IV)_1-xTi^(III)_xNb₆O₁₈ phases with various Ba doping were synthesised using both sol gel and solid state methods. Techniques such as X-ray fluorescence (XRF), X-ray diffraction (XRD) and scanning electron microscopy/energy dispersive x-ray spectroscopy (SEM/EDX) were used to try and confirm whether Ba could be incorporated into the phase of interest. Once confirmed further leach tests could help predict the long term behaviour of the waste form in intermediate depositories.
1. B. Yu, J. Chen and C. Song, J. Mater. Sci. Technology, 2002, 18, 206-210
2. E. Maddrell, Chem. Eng. Res. Des., 2013, 91, 735-741
3. T.-Y. Chen, J.A. Hriljac, A. S. Gandy, M. C. Stennett, N. C. Hyatt and E. R. Maddrell, Scientific Basis for Nuclear Waste Management XXXVI, MRS Symp. Proc. 2013, 1518, 67-72
4. G. Desgardin, C. Robert, D. Groult and B. Raveau, J. Solid State Chem., 1977, 22, 101-111
5. A. Y. Leinekugel-le-Cocq, P. Deniard, S. Jobic, R. Cerny, F. Brt and H. Emerich, J. Solid State Chem, 2006,179, 3196-3208

In the context of nuclear wastes disposal facilities, NEEDS project (Nuclear, Energy, Environment, Wastes and Society) aims for the study of long term evolution of nuclear waste packages. This French project federates ANDRA (the French National Radioactive Waste Management Agency), French research centers CNRS and CEA, EDF (the French Electricity Agency), AREVA (a world leader of nuclear energy), and IRSN (the French Institute of Radioprotection and Nuclear Safety). At the end of the light water reactors operation, the Zircaloy (zirconium alloy) cladding tubes have been activated and oxidized. After processing, these cladding tubes are compacted in a wafer form, placed into a steel container, and then into a concrete over-pack with a view of being disposed of in geological repository. These wastes are mostly composed of activated oxidized metal pieces which also contain traces of fission, activation products and actinides. Our study focuses on carbon-14, an activation product from nitrogen-14 and oxygen-17, contained in cladding tubes. In the disposal, the radioactive wastes are exposed to humid air on a first phase. A fundamental study has thus been defined on the influence of wet air radiolysis on carbon-14 mobility.
To simulate real conditions, zirconia samples were performed by Zircaloy-4 oxidation under air pressure in a furnace at 600°C. Carbon atoms were introduced into oxidized Zircaloy-4 using ionic implantation with IMIO400 at the “Institut de Physique Nucléaire de Lyon”. SRIM simulations were performed so as to define implantation parameters. This study leads to the carbon 13 stable isotope implanted at a 260 nm depth and a concentration at 3.6% atomic carbon.
On a first side, carbon diffusion into zirconia is investigated under temperature conditions and without radiolysis. Annealing experiments of zirconia samples were performed at 100°C (simulating disposal temperature) and 500°C to accelerate diffusion rate. For these experiments, samples were placed in a cell especially developed at IPNL and the same one is used in radiolysis experiment studied on a second part.
Regarding radiolysis experiment devices, control of humidity in the irradiation chamber is established using a glycerol solution. The irradiation experiments are performed using a 2 MeV energy proton beam at the Van de Graaff accelerator of IPNL. Before and after radiolysis and annealing treatments, Secondary Ion Mass Spectrometry allows us to determine the carbon depth distribution. The oxygen profile distribution on the zirconia surface is determined by Nuclear Backscattering Spectrometry, also performed with the Van de Graaff accelerator. Physico-chemical and structural techniques of characterization are used to analyze the oxide film on the samples (DRX, MEB).
In accordance with these investigations with and without radiolysis, carbon mobility in oxidized Zircaloy-4 is discussed.

According to the Planning Act of 28^th June 2006, Andra is in charge of ensuring the sustainable management of all radioactive waste generated in France, especially the high-level and long-lived vitrified waste produced by spent fuel reprocessing.
According to the current design, nuclear glass is poured into a stainless steel container which is in turn put in a carbon steel overpack to prevent water from leaching the waste as long as the temperature is greater than 70 °C (about a thousand years). The overpacks are then laid down one behind the other but separated by spacers in order to comply with the thermal constraints in tunnel-like cell equipped with a carbon steel lining.
In the so-called “Dossier 2005”, Andra demonstrated the feasibility of a deep geological disposal in the Callovo-Oxfordian claystone. At that time, an operational model called “V₀→V_R” was built in order to compute glass alteration and radionuclides release. This operational model integrated an initial alteration rate (V₀) which is maintained as long as reactive materials are in contact with the glass, and a residual alteration rate (V_R) assuming that glass alteration during the rate drop phase is negligible. All the model parameters were acquired in pure water without explicitly taking into account the environmental effects.
Since 2006, all the studies and research related to the components of HLW cells have been incorporated into a broader R&D program which aims at characterizing and modelling (i) the dissolution of glass matrix, (ii) the corrosion of the overpack and the lining, and (iii) the claystone evolution in the near field, considering all the interactions between these materials. This program, coordinated by Andra, has involved up to eighteen laboratories.
The major part of the work focused on the influence of the environment on glass alteration. The effect of clay pore water on glass alteration rates (initial rate, rate drop and residual rate) was determined and particularly that of pH and magnesium. The nature of corrosion products and their interactions with glass alteration were also investigated. All these studies relied on experiments in surface laboratories, in Andra&’s underground laboratory, together with natural or archeological analogs and modelling studies.
In addition, after closure of disposal cells and overpack failure, glass alteration is expected to begin in partially saturated conditions due to hydrogen production resulting from carbon steel corrosion in anoxic conditions. Therefore, part of the glass should be hydrated by water vapor during thousands of years until complete saturation. A part of the studies aimed to determine the glass behavior in such conditions, the influence of the main parameters (temperature, relative humidity) and consequences of vapor hydration on subsequent radionuclides release by water leaching.

A complete understanding of radioactive waste glass interactions with near-field materials is essential for appropriate nuclear waste repository performance assessment. In many geologic repository designs, iron (Fe) is present in both the natural environment and in the stainless steel containers that will hold the waste glasses. Fe materials have been known to accelerate glass corrosion, but the mechanisms for this acceleration are not well understood. The effect of the initial oxidation state of Fe and its effect on the performance of International Simple Glass (ISG) has been in investigated in this work. The two Fe-bearing materials that have been studied are Fe⁰ foil and hematite (Fe₂O₃). Larger corrosion product concentrations were observed for systems containing Fe⁰ rather than hematite. Additionally, the effect of the distance between Fe materials and glass has been investigated. Minimal glass corrosion was observed for distances between Fe materials and ISG greater or equal to 2 mm; substantial glass corrosion was observed for systems exhibiting full contact between Fe material and ISG. Additionally, the efficiency of selected corrosion suppressants was tested. The silica-based minerals, fumed silica and diatomaceous earth, were investigated in order to suppress ISG corrosion initiated by Fe. Glass corrosion rates were calculated in each experiment.

Experiments were conducted to address the uncertainty in dissolution rates calculated from tests using crushed glass. Reaction rates are commonly represented on a per-area basis to compare responses of different glasses under various test conditions and to upscale measured rates to represent full-sized waste forms. Samples of crushed glass are prepared for testing by isolating size fractions by sieving, and the specific surface area calculated by relating the mean sieve opening to the dimension of a geometric model of the particle, such as the diameter of a sphere. These tests address the uncertainty in the test response due to the distribution of particle sizes relative to the repeatability of the test. Analyses have shown the distributions of effective particle dimensions in -100 +200 mesh size fractions prepared for PCTs are nearly Gaussian. This means the distribution of particle surface area in a test sample, which is proportional to the square of the particle dimension, will be skewed to smaller particles. The single value of the specific surface area selected to represent the size fraction should minimize the error in the calculated the dissolution rate. The average dimension based on the sieve fraction will not provide the best representation of the surface area, but the error it imposes may be within the uncertainty of the test method.
A series of PCTs was conducted to measure the sensitivity of the response to the size fraction of particle sizes used to determine the sensitivity of PCT-A response to the skewed-Gaussian area distributions within each test sample. ARM-1 glass was crushed and dry sieved to isolate the -100 +200 mesh size fraction, which was then repeatedly washed to remove fines by agitation and decanting. A portion was then re-sieved to isolate the -100 +120, -120 +140, and -140 +200 mesh size fractions, which were again washed to remove fines. All fractions were analyzed with SEM and a Microtrac particle size analyzer to measure the size distributions. All size fractions had a Gaussian size distribution slightly skewed to smaller particles. Samples of the -100 +200 mesh size fraction were used in an inter-laboratory study at the nominal water/glass mass ratio (10:1) and in separate tests conducted at the extreme water/glass mass ratios allowed in PCT-A (9.5:1 and 10.5:1). Triplicate tests were conducted with each sub-fraction at the nominal water/glass mass ratio (10:1). Another set of samples is being prepared by wet sieving to compare the size distributions and PCT responses.
The responses of triplicate PCTs are sensitive to the size fraction and water/glass ratio; the repeatability of tests under different conditions is about the same. The relationships between the PCT responses, particle size distributions, and preparation method and the uncertainty associated with up-scaling to full-sized glass waste forms will be discussed. The results of the inter-laboratory study, which is in progress, will also be presented.

Borosilicate glass is the leading candidate for a disposal matrix in permanent geologic repositories. Since corrosion of nuclear waste-containing borosilicate glass by (ground)water cannot be ruled out for long-term storage over geological time scales, scientists have conducted numerous experiments in an effort to identify the key mechanism that leads to glass breakdown when contacted with aqueous solutions. In most of these studies it has been appreciated that in advanced stages of glass-water reaction, a nano- to micro-porous corrosion layer (or “leach layer”, “gel”, or “hydrated residual glass”) develops on the glass surface that has the potential to govern the long-term reactivity of glass. However, there is still no consensus how this layer forms and how it influences the long-term corrosion behaviour. At least three glass corrosion models used to predict silicate glass degradation in aqueous environments have gained widespread acceptance. The chemical affinity model is based on the notions of “chemical reaction affinity” and “deviation from equilibrium”, defined by the degree of silica saturation in solution. It commonly neglects the importance of the corrosion layer until discrete secondary phases begin to precipitate, affecting the long-term solution saturation state and thus the dissolution kinetics. In the alkali-proton exchange model, the corrosion layer is regarded as the product of rate-limiting ion exchange reactions between protons and network modifiers such as Na, producing silanol (Si-OH) groups which are believed to re-polymerize in the solid state, forming the residual, hydrated glass layer (“leach layer”). The third model postulates that the corrosion layer constitutes a protective “gel” layer that governs the long-term corrosion kinetics. Similar to the alkali-proton exchange model, the corrosion layer is regarded as a dynamically restructured, residual, hydrated glass that with increasing reaction time becomes so dense and the pore spaces so constricted that a relative impermeable zone develops that slows down the corrosion process. I will introduce and discuss an alternate mechanistic model for silicate glass corrosion in the context of the above mentioned corrosion models. This model is based on novel results of Mg, O, and Si isotope tracer and exchange experiments as well as on in situ, real-time, and spatially resolved Raman spectroscopic measurements that were performed while a borosilicate glass interacts with an aqueous solution at elevated temperatures. In this model, the formation of glass corrosion rims is explained by congruent glass dissolution (i.e., all glass constituents are released into solution), which is coupled in space and time to the precipitation/deposition of amorphous silica at a moving dissolution-precipitation front. Such coupled reaction produces a corrosion zone that is composed of spherical silica aggregates of variable composition, size, and porosity.

Dissolved silica is known to affect borosilicate glass alteration mechanisms and rate, firstly by reducing the affinity for the reaction of hydrolysis of Si-O bonds and secondly, by favoring the gel polymerization which eventually slows down the transport of water toward the pristine glass. The latter process is related to the so called ‘passivating effect&’, which accounts for the rate drop of several orders of magnitude compared to the initial dissolution rate. Depending on glass composition, temperature and pH conditions, Si-bearing minerals (clay-type minerals, calcium silicate hydrates, zeolites, etc.) can precipitate during glass alteration, modifying the distribution of Si between the solution, the amorphous layer and crystalline phases. In case of zeolite precipitation, the glass dissolution rate can suddenly resume. In order to better understand the relationships between the phenomena involving silicon, a series of experiments have been undertaken with the International Simple Glass, a six oxide borosilicate glass selected by the international nuclear glass community to improve the understanding of glass corrosion mechanisms and kinetics. The ISG glass was altered at 90°C in solutions initially saturated with respect to amorphous silica and at fixed pH of 2, 7, 9, 9.5, 9.8, 10.1 and 10.3. Glass alteration was monitored during 56 days, following the release of boron. Our first results show that, between pH 7 and pH 9, i) a homogenous amorphous layer, mainly formed with Si, Al and Zr, has replaced an equivalent volume of glass, ii) no Si-crystalline phases have precipitated, iii) the formation of the alteration layer has not consumed Si initially present in solution, iv) the passivating zone was found. These results are in agreement with the GRAAL model, a mechanistic model used to predict the dissolution rate of R7T7-type glass in various environments.

The pH-dependence of glass corrosion rates has a well-known U-shaped form with minima for near-neutral solutions. This paper analyses the change of U-shaped form with time and reveals that the pH dependence evolves even for solutions kept at constant pH. The U(t) dependence is due to changes of concentration profiles of elements in the near-surface layers of glasses in contact with water and is most evident within initial stages of glass corrosion at relatively low temperatures of experiment.
Fresh glasses in contact with water have minimal corrosion rates shifted to higher pH. Aged glasses which were in contact with water for a substantial time are characterised by relative symmetric U-shape. Older glasses have a U-form squeezed to the acidic range of solutions.
Numerical examples are given for the nuclear waste borosilicate glass K-26 which is experimentally characterised by an effective diffusion coefficient of caesium D_Cs = 4.5 10^-12 cm²/day and by a rate of glass hydrolysis in non-saturated groundwater as high as r_h = 100 nm/year The changes of U-shaped form need to be accounted when assessing the performance of glasses in contact with water solutions.

The dissolution rates of many surrogate waste glasses increase suddenly and significantly coincidentally with the precipitation of secondary phases under a range of experimental conditions. The increased rate is commonly referred to as Stage 3. Because the release of radionuclides from waste glasses in a disposal system will be controlled by glass degradation, Stage 3 behavior will deminish waste form performance. Not all glasses and not all secondary phases result in Stage 3 behavior. Understanding what triggers the increase and what controls the rate thereafter will either provide confidence that the dissolution rate of a particular waste glass will not be triggered under anticipated disposal conditions or lead to reliable predictions of when and to what extent the rate will increase in a disposal system so the phenomenon can be taken into account in performance calculations. Results of recent tests to study the process and conditions that trigger Stage 3 behavior and consideration of recent advances in geochemical mineralogy have led to a new conceptual model for glass dissolution. The model is a modification of the Grambow reaction affinity model in which the partial equilibrium assumption used to account for secondary phase effects is replaced by a coupled kinetics model. Most importantly, the glass dissolution kinetics are coupled with the precipitation kinetics of secondary phases to establish a steady-state condition rather than decreasing as if the glass is itself approaching a pseudo-equilibrium. The Stage 3 rate reflects the balance between the thermodynamic drive to covert glass to increasingly stable alteration phases and kinetic limitations to the transfer of material from the glass to those phases through the solution. The precipitation reactions coupled to glass dissolution and the steady-state rate will change as the suite of alteration phases evolves from an amorphous silica-like surface alteration layer formed by initial dealkalization reactions to a complex assemblage of precipitated phyllosilicates. Aspects of geochemical models that provide the foundation of the new conceptual model and key experimental results supporting application to waste glasses will be summarized. How the model addresses other experimental observables will also be discussed, such as the negligible effects of clay formation on the measured glass dissolution rate and why the addition of mineral secondary phases to test systems fails to trigger Stage 3 behavior. Approaches being developed to incorporate the Stage 3 model into the glass source term model for disposal system performance assessments and laboratory tests to measure effective parameter values will also be presented.

The necessity to a priori predict the durability of high level nuclear waste (HLW) glasses on extended time scales has led to a variety of modeling approaches based primarily on solution (leachate) concentrations. An examination of the {glass-gel-solution} system from long-term accelerated leach test data has demonstrated that the {gel-leachate} system, along with the bulk glass composition and structure, determines if a glass will resume leaching at long times. The glass composition and structure control both the leachate and the gel compositions which in turn control what reaction products form. A database, Accelerated Leach Testing of GLASS (ALTGLASS), was developed. ALTGLASS contains 213 glasses of which 74 are HLW glasses. From the 74 HLW glasses a population of 12 HLW {glass-gel-solution} systems were chosen: some known to resume leaching at an accelerated rate and some that continued at a steady state rate. This HLW glass population had been durability tested for one to two decades and these results were coupled with recent accelerated durability studies on SON68 and the AFCI glass. This population was the longest HLW durability available in the DOE complex. When zeolites, such as analcime, formed on the glass surfaces, the glass resumed dissolution at an accelerated rate. Gels that aged into clay mineral assemblages continued leaching at constant long-term rates. Analytic interrogation of the gel layer chemistry for the HLW glasses indicated that glasses that resumed leaching had gel compositions of ~Na_0.84Al_0.98Si_2.0O_5.89middot;xH₂O, hydrated analcime, while glasses that dissolved at a constant long-term rate were found to have gel compositions of ~Fe₂Si₂O₅(OH)₄middot;2H₂O (hisingerite) and amorphous SiO₂. Hisingerite possesses the same local structure as nontronite clay and is a poorly-crystallized precursor of ferric smectite clays. Glasses with excess molar alkali, (Na,Li,K)₂O, over the sum of molar (Na,Li,K)AlO₄ + (Na,Li,K)FeO₄ + (Na,Li,K)BO₄ resumed leaching and contained excess alkali in the gel and excess strong base, [SB], in solution. Glasses without excess alkali did not resume leaching and the leachate continued to generate excess weak acids, [WA], in solution with reaction progress favoring gel aging into clays. Identification of these rate limiting precursor complexes and their dependency on leachate [SB] and [WA], allowed preliminary rate-determining leach layer forming exchange reactions to be written. New accelerated leach data has been generated on HLW glass standards often used in Performance Assessments such as the Defense Waste Processing Facility (DWPF) Environmental Assessment (EA) glass and the DWPF Waste Compliance Plan (WCP) glasses. This data will be analytically interrogated to determine the validity of the preliminary rate-determining leach layer forming exchange reactions and assess the impact of these findings on Stage 3 glass degradation.

Geological disposal of HLW and used nuclear fuel (UNF) in very deep boreholes is a concept whose time has come. The alternative - disposal in a mined, engineered repository is beset with difficulties not least of which are the constraints placed upon the engineered barriers by the high thermal loading. The deep borehole concept offers a potentially safer, faster and more cost-effective solution. Despite this, international interest has been lukewarm, largely due to perceived problems with retrieveability and uncertainty about the ability to drill accurate vertical holes to a depth of 4-5 km with diameters greater than 0.5 m. The closure of Yucca Mountain and the subsequent recommendations of the Blue Ribbon Commission have lead to a renewed interest in deep borehole disposal (DBD) and the US DOE has commissioned Sandia National Labs, working with industrial and academic partners (including the University of Sheffield), to undertake a program of R&D leading to a demonstration borehole being drilled somewhere in the USA.
In this presentation, we focus on some of the key safety and engineering features of DBD including methods of sealing the boreholes, sealing and support matrices for the waste packages and options for deployment of the packages. Numerical modeling has, and continues to play, a significant rocirc;le in expanding and validating the DBD concept. In this presentation we report on progress in the use of modeling in the above contexts, paying particular attention to constraints on the engineering materials resulting from high heat loading. Our modelling covers both UNF and alternative wastes; the latter including Cs and Sr capsules from the Hanford site in the USA

The use of deep borehole disposal (DBD) for spent nuclear fuel (SF) and high level nuclear waste (HLW) streams is now seen as a viable alternative to disposal in geologically shallow repositories [1]. Significant advantages could arise by placing and sealing SF and HLW wasteforms within boreholes drilled several kilometers into granitic basement rock compared to disposal in a repository a few hundred meters deep [2].
The University of Sheffield in the UK has played a pivotal role in developing the DBD concept. Recent research has focused on sealing and support of the waste packages within the borehole as well as providing mechanical support for the load of the overlaying containers [3]. Any such dual purpose material must retain its key physical and chemical properties in a harsh environment; the temperature at the bottom of a borehole will be around 120°C (depending on borehole depth and the local geothermal gradient), with radiogenic heating from the waste packages increasing this value. One type of sealing/supporting matrix being developed in Sheffield is a cementitious grout for use when borehole wall temperatures are lower than approximately 200°C [3]. The exposure of a cementitious grout to these temperatures will dictate the properties of the wet paste before setting and its properties after hardening [1].
In this publication we report the latest research into the design of a high performance cementing system suitable for providing down-hole sealing and support for waste packages. The performance of suitable cementitious systems is reviewed and development of an initial bespoke grout is described. This grout contains a siliceous component to provide the desirable physical/chemical properties at the high temperatures to which the grout will be exposed, as well as components to retard hydration of the cement during curing and retain the grout mix water during grout deployment through the water in the borehole. The rheological properties of the grout are being studied at pressures and temperatures representative of those at the bottom of a 5 km deep borehole focusing on assessing grout consistency, viscosity, yield stress and flow. The data presented here are being used to optimize the composition of the grout. This work is a precursor to assessing the performance of the grout in the hardened state in order to predict long-term durability in the challenging down-borehole environment.
1. B. Arnold, P. Brady, S. Altman, P. Vaughn, D. Nielson, J. Lee, F. Gibb, P. Mariner, K. Travis, W. Halsey, J. Beswick, J. Tillman, Deep Borehole Disposal Research: Demonstration Site Selection Guidelines, Borehole Seals Design, and RD&D Needs, FCRD-USED-2013-000409, SAND2013-9490P, Sandi National Laboratories report for U.S. Department of Energy, October 25, 2013.
2. F. Gibb, Looking Down the Bore, Nuclear Engineering International, 21-22, February 2010.
3. J. Beswick, F. Gibb, K. Travis, Proceedings of the ICE - Energy, 167 (2014) 47 -66.

Most waste disposal programs envisage that the waste will be sealed inside metallic containers and placed in a deep geologic repository. To meet safety assessment requirements it is essential to demonstrate the margins of safety against potential failure mechanisms. In Canada and Scandinavia the current reference container design consists of a thick-walled corrosion resistant copper shell over an inner steel vessel or cast iron insert. However, models predict that the maximum corrosion penetration will be < 1.27 mm over a million years. With this in mind, and to overcome design and fabrication issues, thinner-walled copper-coated steel vessels are being investigated in Canada. Since the available copper corrosion barrier would be significantly reduced this has stimulated a re-evaluation of container corrosion performance.
This presentation will describe some of the studies underway to evaluate this performance. A primary goal of these studies is to investigate whether the fabrication process has any significant influence on corrosion behaviour. With this goal in mind the corrosion of cold spray and electrodeposited copper coatings are being compared to that of standard wrought copper. A range of electrochemical techniques are being applied under both anoxic and aerated conditions as well as in the presence of sulphide which could be present in a waste repository as a consequence of mineral dissolution or remote microbial activity in repository materials. Also, a re-evaluation of the influence of gamma radiation, emitted by the wasteform, is underway, since the radiation fields on the exposed container wall will be higher than for the reference design. The nature and distribution of corrosion products and damage is being investigated using techniques such as scanning electron microscopy, energy dispersive X-ray analysis, X-ray photoelectron spectroscopy, profilometry and confocal laser scanning laser microscopy.
The ultimate goal of these studies is: (i) to demonstrate whether the extensive available database on the corrosion of copper waste containers can be applied to copper-coated steel; and (ii) to assess the extent to which the copper shell can be thinned without compromising container corrosion performance.

The use of copper on the exterior of ferrous based used fuel containers (UFCs) has been explored by SKB, Posiva and others as a means of de-coupling the structural and corrosion performance requirements of a UFC in a deep geological repository (DGR). The strong inner cast iron (or steel) insert of such a vessel allows it to withstand geological pressures, including glacial loads during future ice ages, while the selection of copper is based on a thermodynamic understanding of the stability of copper over the long anaerobic period required for long-term containment. Analysis of many decades of copper corrosion data has demonstrated that a suitable corrosion allowance for copper in a Canadian DGR is just 1.27 mm, and it is expected that substantially less copper corrosion will actually occur. However, fabrication of SKB&’s KBS-3 vessels requires significantly thicker copper so that the cylindrical vessels will not collapse during fabrication, machining and assembly; the current reference copper thickness is 50 mm for the KBS-3 shell.
In recent years, Canada&’s Nuclear Waste Management Organization (NWMO) has been developing methods for depositing copper directly on steel-based used fuel container (UFC) materials in an effort to remove all structural requirements from the copper form. While weld cladding methods have been preliminarily examined, efforts have been predominantly focussed on electrodeposition and cold-spray methodologies. The former is a widely used commercially available method and is being considered to produce coatings, while the latter has been used primarily as a repair process over the past two decades within selected niche markets (i.e. aerospace). Both processes have demonstrated success within research environments, and are currently being developed for small scale manufacturing. Electrodeposition is the leading technology to pre-coat containers and lids, while cold spray will be used over the seal weld, once the container is loaded with fuel in the used fuel packaging plant.
To verify suitable performance, extensive testing on corrosion properties is underway. In addition, finite element modelling is being combined with mechanical testing to verify performance of copper coated containers. To date, a high degree of success has been shown within this analysis, verifying the integrity of copper coated UFCs for long term storage of used nuclear fuel in Canadian DGRs. This paper will focus on developments within the manufacturing programs of copper coated UFC components, as well as highlighting results from structural analysis.

The corrosion of spent fuel inside a failed nuclear waste container will be controlled by the redox conditions prevailing at the fuel surface. These conditions will be dictated by competition between alpha radiolytically produced oxidants, of which H₂O₂ is the dominant one, and reactions which consume H₂O₂ and prevent its reaction with the fuel surface.
Peroxide consuming reactions include those with steel corrosion products, Fe²⁺and H₂, and H₂O₂ catalyzed decomposition on the fuel surface. There is now a considerable amount of evidence to show all these reactions have a significant influence on redox reactions and, hence, on the fuel corrosion/radionuclide release processes. If the predictions of models developed to describe fuel corrosion are to be meaningful it is necessary to quantify the relative importance of these reactions.
Electrochemical studies show that the interaction of H₂O₂ with a UO₂ surface is very dependent on the condition of the surface; i.e., whether it is catalytic (U^IV_1-2xU^V_2xO_2+x) or partially blocked by the presence of insulating U^VI oxides/hydroxides. In addition the reactions between H₂O₂ and H₂are strongly influenced by the presence of noble metal particles produced during in-reactor fission.
This presentation will describe our recent electrochemical studies of H₂O₂ reactions on UO₂ (SIMFUEL) surfaces with an emphasis on the competition between decomposition and corrosion. Additionally, we have been studying the reaction between H₂O₂ and H₂ on fuel surfaces. The presentation will address the importance of surface composition and the importance of bicarbonate/carbonate in controlling it.

Since 1994, the Korea Hydro & Nuclear Power Co., Ltd. (KHNP) has developed a vitrification technology to treat the LILW (Low-and Intermediate-Level radioactive Waste) generated from the Korean nuclear power plant. To vitrify the LILW including Ion Exchange Resin, Zeolite and combustible Dry Active Waste (DAW), two borosilicate glasses were formulated. The formulated glass DG2 is for the DAW vitrification solely and the other formulated glass AG8W1 for the blended wastes. The physico-chemical properties of two glasses were evaluated as described below.
To evaluate the processability of the glasses, the viscosities and electrical conductivities of glass melts were measured in the laboratory within a temperature range of 950 to 1350 degree C, respectively. The liquidus temperatures of the glasses were evaluated using data from heat treatment. The phase stability of the glasses was examined by using SEM/EDS and XRD. The Mossbauer spectroscopy for the blended waste glass(AG8W1) was employed to evaluate relations between redox equilibria of iron. About 42% Fe²⁺ state existed in the glass produced from the pilot scale CCIM test.
To verify the chemical durability of the glasses, several tests such as PCT, ISO, VHT, Soxhlet, MCC-1, and ANS16.1 were performed as follows.
The PCT shown the leach rates of B, Na, Li and Si were much less than those of the benchmark glass (SRL-EA). To examine a long-term leaching behavior of the glasses, the ISO test was performed at 90 degree C for 1022 days and Cumulative Fraction Leached of major elements in the glasses were analyzed. According to the VHT, the glasses had an outstanding chemical resistance to the humid environment at 200 degree C for 7 days. The corrosion rates of the glasses were in the range of 2-10 g/m²/day and met the DOE specification. The Soxhlet leaching was accomplished on a rectangular glass samples of size about 5x5x10 mm. The water temperature in the overflow vessel containing the glass sample was 98 degree C. For the determination of the weight losses of the glasses during the test time, the glass samples were taken out of the equipment after 1, 3, 6, 10, 17 and 30 days. To analyze the normalized mass losses and the forward dissolution rates of major glass elements, the MCC-1 leaching method were conducted at temperatures of 40, 70, and 90 degree C for three weeks in pH buffer solutions spanning in range from pH 4 to pH 11. The ANS16.1 test was also conducted to evaluate the leachability indexes for major elements in the glasses.
In addition, the compressive strengths of the glasses were tested after 90 days leaching tests, thermal cycling tests, and radiation irradiation tests.
The processability of the glasses was in the desired ranges. And the product quality of the glasses met all regulatory guidelines. Using the glasses, the CCIM commissioning tests in a commercial vitrification facility of Korea were successfully performed and showed the good workability.

In a future Swedish deep repository for spent nuclear fuel, irradiated control rods from PWR nuclear reactors are planned to be stored together with the spent fuel. The control rod absorber consists of an 80% Ag, 5% Cd, 15% In alloy. Upon in-reactor irradiation ¹⁰⁸Ag is produced by neutron capture. Release of ¹⁰⁸Ag has been identified as potential source term for release of radioactive substances from the deep repository.
Under deep repository conditions, the Ag corrosion rate is however expected to be low, which would imply that the release rate of ¹⁰⁸Ag should be low under these conditions.
The aim of this study is to investigate the dissolution of PWR control rod absorber material under conditions relevant to a future deep repository for spent nuclear fuel. The experiments include tests using irradiated control rod absorber material form Ringhals 2, Sweden. Furthermore, un-irradiated control rod absorber alloy has been tested for comparison.

Inert matrix fuels for minor actinide burning have been under study for over 20 years and Kleykamp [1] gave an early summary of the candidacy of various metallic and ceramic materials. At that time stabilized zirconia was indicated to be a leading candidate and this is still true today. Some of the major attributes of the best candidates were high melting point, high thermal conductivity, low neutron absorption and stability in aqueous environments. Apart from some work by Japanese workers [2,3], who used multiphase ceramics consisting of mineral-like compounds such as magnetoplumbite (CaAl₁₂O₁₉), Al₂O₃, ZrO₂ and MgAl₂O₄, little focus has been given to the influence of fission products on the behavior of inert matrix fuels. We have looked at the disposition of Cs, Sr, Ba, La, Mo and I as representative fission products in stabilized zirconia that was hot isostatically pressed in a Ni can at 1400°C/100MPa. HIPing resulted in a material with density of ~92% of theoretical and with ~3% of open porosity. Cs₂MoO₄ and I-bearing material was formed even though much of the Mo was reduced to metal. Even the finely dispersed Mo metal appeared to be leachable in PCT tests when open porosity was present, as found from powder XRD before and after leaching and analysis of the leach liquids.
As alternative inert matrix fuels, a sample of NZP-structured CaZr₄(PO₄)₆ material, as well as a magnetoplumbite, Al₂O₃, ZrO₂ formulation, each containing equivalent quantities of fission products, were HIPed similarly. The disposition and leachability of the fission products for all three potential inert matrix fuel materials will be reported.
References
[1] H. Kleykamp, J. Nucl. Mater. 275, 1-11 (1999).
[2] N. Nitani et al., J. Nucl. Mater. 247, 59-62 (1997).
[3] T. Yamashita et al., J. Nucl. Sci. Tech. 39, 865-871 (2002).