Symposium F.EN08—Scientific Basis for Nuclear Waste Management

Dr Eric (Lou) Vance was a trailblazer in developing novel treatment solutions for intractable nuclear wastes, who spent 32 years at the Australian Nuclear Science & Technology Organisation (ANSTO) in Sydney. Largely known for his contributions to Synroc technology, Lou’s experience spans multiple continents including time at Pennsylvania State University with McCarthy and Roy as well as time at the Atomic Energy of Canada Limited where he studied sphene glass-ceramics. In 1987 he was recruited to ANSTO to progress Synroc technology. Under Lou’s leadership, the original Synroc titanate full ceramic formulations invented by Ringwood evolved into a waste treatment technology platform. This platform can be applied to produce glass, glass-ceramic and ceramic wasteforms via hot-isostatic pressing.
Lou’s seminal contributions to the science of nuclear waste management are evident in his ~400 scientific articles in international journals/refereed conference proceedings and his co-authorship of 3 patents relating to Synroc. An entertaining and highly popular speaker, Lou was committed to attending two to three international conferences a year to showcase the Synroc technology and his innovative nuclear wasteform research. Lou was a steadfast member of the Scientific Basis for Nuclear Waste Management symposium for several decades and a highly enthusiastic mentor. In this paper, we celebrate the research career of the late Dr Vance and his substantial and essential contributions to the field of nuclear wasteform development and to the development of Synroc technology.

The capture and immobilization of long-lived radioactive iodine I-129 must be a priority for future nuclear fuel cycles involving reprocessing. This review summarizes work on waste forms for radioiodine, emphasizing the contributions of Dr. Lou Vance as well as recent work in other groups. As early as the 1980s, Dr. Vance recognized the need for iodine waste forms and was a leader in innovative waste form solutions for iodine and other volatile halides. His studies included immobilization in silicate minerals including zeolites and sodalite, encapsulation in metallic tin, partitioning in synroc mixed-phase ceramics, and consolidation of copper iodide. Here, we recognize his particular contributions to the understanding of iodosodalite, and note some recent work by others following in his footsteps.

Synroc-C, a multiphase ceramic wasteform based on stable and leach resistant titanate minerals, was invented by Australian scientists in the late 1970s. The formulation was directed toward the immobilization of PUREX wastes from the reprocessing of nuclear fuels and was originally to be produced via hot-uniaxial pressing (HUPing). Although glass was chosen as the preferred technology during the “Atlanta shoot-out” in the US in 1981, this was ultimately because Synroc lacked the technical maturity of vitrification. However, given the enhanced performance properties of Synroc-C as a wasteform, significant interest has been sustained in this Synroc formulation. The vast majority of work on Synroc-C has been undertaken following consolidation by HUP processing. There is very little reported on hot-isostatically pressed (HIPed) Synroc-C, where HIP canisters are filled with the calcined wasteform material (waste + additives), evacuated and sealed. The HIP canister is then loaded into the HIP chamber and processed through the desired time-temperature-pressure profile to achieve a dense and consolidated wasteform. The appropriate redox conditions within the HIP canister are vital to the formation of the correct phase composition and ultimately attaining a stable and leach resistant wasteform.
In this study, we investigate the effect of Ti-metal addition on the phase formation, microstructural characteristics and durability of HIPed Synroc-C. Dense Synroc-C samples containing 20 wt.% simulated high level waste and with varying Ti-metal additions (0%, 2%, 4% and 8 wt.%) were prepared by HIPing in stainless steel HIP canisters. An investigation of microstructure, mineral phases and elemental distribution in the samples was carried out using quantitative X-ray diffraction and electron microscopic analyses. Results from durability tests are also discussed.

The UK holds the world’s largest inventory of separated plutonium under civil safeguards, projected to reach 140 tons at the end of reprocessing options. A significant fraction of this material requires immobilisation for disposal since it is unsuitable for reuse in MOX fuel, which is HM Government’s preferred policy for disposition. Additionally, should it not prove possible to implement the policy for plutonium reuse in MOX fuel an alternative immobilisation and disposal option will be required for the greater inventory. This presentation will review the current status of wasteform engineering and process development for manufacture of zirconolite ceramic and glass-ceramic wasteforms for immobilisation of waste and inventory plutonium. We will demonstrate the development of Ce and U surrogate formulations, validated by targeted plutonium studies, with regard to phase assemblage and Pu / surrogate incorporation mechanism. The development of a novel containment system to enable hot isostatic pressing studies utilising U and Th surrogates will be presented, together with operational experience. The application of new, and potentially disruptive, wasteform manufacture technology will also be discussed.

Cs₃Bi₂I₉, a defect perovskite derivative, is a potential host phase to immobilize iodine and cesium with high waste loadings. In this work, two strategies are explored to form Cs₃Bi₂I₉-silica composites and core-shell structures in order to improve chemical durability of waste form materials meanwhile maintaining high waste loadings. Cs₃Bi₂I₉ loadings as high as 70 wt.% were incorporated into a silica matrix to form silica-ceramic composites, and 20 wt.% Cs₃Bi₂I₉ was encapsulated into silica to form a core–shell structure at low temperatures and high pressure. Semi-dynamic leaching experiments indicated Cs and I were incongruently released from waste form matrices depends on the bonding environment and surface passivating effect. A BiOI alteration layer formed during the surface reorganization, acting as a passivation layer to reduce the release of radionuclides. The long-term iodine release rate was low (30 mg m^-2 day^-1) for 70 wt.% Cs₃Bi₂I₉–silica composites leached in deionized water at 90 °C, which can be further reduced to 5×10^-3 mg m^-2 day^-1 for the 20 wt.% core-shell structure. This work highlights a robust way to immobilize the highly mobile radioactive nuclides and surface passivating mechanism through elemental transport across the liquid-solid interface for Cs₃Bi₂I₉.

Radioactive waste streams containing ⁹⁰Sr, from nuclear power generation and environmental cleanup operations, are immobilised in the UK in blended Portland cement (PC) grouts to limit radionuclide leaching. Due to poor compatibility of certain wastes with PC, alternatives such as alkali aluminosilicate ‘geopolymer’ binders are being investigated. Here, we used multinuclear solid-state nuclear magnetic resonance (NMR) spectroscopy probing ²³Na, ²⁷Al, ²⁹Si and ³⁹K nuclei, including high-field (20.0 T) and cross polarisation experiments, along with X-ray diffraction experiments, to investigate the incorporation mechanism of Sr²⁺ and Ca²⁺ in geopolymer wasteforms formed by alkali-activation of metakaolin. The main reaction product identified in gels cured at both 20°C and 80°C is a fully polymerised Al-rich alkali aluminosilicate hydrate ((N,K)-A-S-H) gel comprising Al and Si in tetrahedral coordination, with Si in Q⁴(4Al) and Q⁴(3Al) sites, and Na⁺ and K⁺ balancing the negative charge resulting from Al³⁺ in tetrahedral coordination. Faujasite-Na and partially Sr-substituted zeolite Na-A form within the gels cured at 80°C. Incorporation of Sr²⁺ or Ca²⁺ displaces some Na⁺ and K⁺ from the charge-balancing sites, with a slight decrease in the Si/Al ratio of the (N,K)-A-S-H gel. Ca²⁺ and Sr²⁺ induce essentially the same structural changes in the gels. It was concluded that the disordered geopolymers can readily accommodate Sr²⁺ and Ca²⁺ into their aluminosilicate framework structure via substitution of the divalent alkaline earth cations for monovalent alkali cations in charge balancing sites. This indicates that geopolymer gels are excellent candidates for immobilisation of radioactive waste containing ⁹⁰Sr and other alkaline earth radionuclides.

The UK’s nuclear sites contain stored wastes in ponds and silos as well as large volumes of contaminated concrete and soil. During storage decommissioning and deconstruction operations, there are significant risks of radionuclide transport via groundwater and airborne particulate releases. Colloidal silica grout is a nano-particulate material that has the potential to reduce these risks very significantly. In order to utilize such nanoparticulate material as a barrier, we require a detailed understanding of the grout properties. These include its injectability, mechanical behaiour, hydrauclic properties, durability and its impact on the geochemistry of radionuclides in the environment (e.g. fission products such as Sr and Cs). Over the past 5 years, experiments at Strathclyde have progressivley investigated colloidal silica grout properties. Colloidal silica-grouted sand shows an increased stiffness and an enhanced peak friction angle, while still having a very low hydraulic conductivity (~ 10-10 m/s), typical of intact clay. Even under drained conditions, the grout provides mechanical improvement to the soil and retains a very low hydraulic conductivity. The impact on radionuclide geochemistry has been investigated through combining leaching experiments with XAS analyses. These show that colloidal silica induces several competing effects on the mobility of Sr and Cs. First, cations within the colloidal silica gel compete with Sr and Cs for surface complexation sites. Second, an increased number of surface complexation sites is provided by the silica nanoparticles and finally, the elevated pH within the colloidal silica increases the surface complexation to clay minerals and the silica nanoparticles. Finally, experiments to investigate injectability and monitorability show that colloidal silica grout can be successfully injected in silty-sands, its gelling time can be accurately predicted in a full range of environmental conditions and the location of the subsurface barrier created can be accurately identified using Electrical Resistivity Tomography. Potential applications of colloidal silica on nuclear sites include inhibiting radionuclide migration from recently decommissioned structures during care and maintenance (e.g. silos and ponds from which wastes have already been retrieved); the development of primary or secondary barriers around existing radionuclide (or other) contaminated areas; and stabilisation of underground structures. Research work is continuing to improve the mechanical and chemical properties of colloidal silica grouts, such that they can be tailored to suit for indivdual nuclear site-based applications.

Machine learning (ML) is a valuable tool that can be used, in combination with domain expertise, to provide insight into nuclear waste vitrification processes. This talk summarizes our preliminary efforts in using ML methods to address important problems facing the nuclear waste vitrification mission at the Hanford site in Washington State. Three case studies using ML modeling will be described: nepheline crystallization, refractory corrosion, and sulfate solubility. In Hanford high-level waste, nepheline crystallization can occur in melt compositions containing high Na and Al contents when glass cools slowly in the canister, partitioning radioactive components to a less-durable residual glass fraction. In low-activity waste compositions, which are generally higher in alkali and volatile components like sulfur and halides, different critical issues present. Firstly, the low solubility of sulfate in borosilicate glass has led to many studies to predict glass composition dependence of this property, such that excess sulfur is not volatilized or partitioned to an alkali molten salt phase during processing. Second, these glass compositions tend to be harsh on refractories, leading to concerns over corrosion and general degradation of melter ceramic components and premature failure. For these three cases, existing datasets were used along with ML techniques to investigate the datasets and create predictive models for the output properties of interest. For all these studies, data science was used to identify underlying trends and correlations which were then applied to create supervised learning models. As a result, it has become evident that ML is a powerful tool that can be utilized to provide insight into complex issues in nuclear waste management, here specifically focused on immobilization by vitrification.

Vitrification is a key process to turn high-level nuclear waste into durable glasses. However, the virtually infinite compositional envelope accessible to glass wasteforms renders traditional design methods based on empirical knowledge and trial-and-error poorly efficient. In addition, the complex, disordered atomic structure of glasses makes it challenging to develop some mechanistic models relating their composition to their properties and performance. Here, I will discuss how machine learning offers a promising alternative route to predict the properties of glass wasteforms and guide/accelerate the design of novel glasses with enhanced performance for nuclear waste immobilization applications.

The development of stable actinide and other radionuclide hierarchical waste-form materials offer potential advantages of greater efficiency/loading, lower leach rates, ease of processing for so-called legacy nuclear wastes. Current waste-form discovery relies heavily on experiments that may or may not be successful in producing a viable product. Computational crystal discovery has aided in the search for new materials for different applications through use of high-throughput algorithms. However, these approaches have not been adapted to actinide and lanthanide compounds due to additional corrections required to properly handle f-orbital-containing elements. This presentation will consider the challenges and possible solutions to creating high-throughput computational methods to aide in the experimental discovery of novel hierarchical waste forms and offer tangible examples.

Pd and Ru are the main platinoid fission products formed during burnup of nuclear fuels. During fuel reprocessing, these element partly dissolve in nitric acid and they exhibit very low solubility into borosilicate glasses during the vitrification process. Due to their affinity towards tellurium from the fission product solution, palladium mainly precipitates under (Pd-Te) intermetallics [1] and ruthenium is likely to form hcp solid solutions [2]. The composition and microstructures of these phases are also sensitive to oxygen potential [3]; oxydation of ruthenium can lead to needle-shaped or polyhedral (Rh,Ru)O₂ oxide phases [4]. These precipitates can affect significantly the physico-chemical properties of the glass melt during melting in hot or cold crucible.
For decades, many studies were undertaken at CEA [5] to point out the high temperature interactions of the platinum group metals (PGM) issuing from the calcinated waste. In parallel to these studies, a Calphad database is being developed to address the thermodynamics of such systems to provide an assessment of the PGMs in nuclear waste.
This study presents the assessed binary systems. Some calculations are also performed to highlight how this flexible tool makes is used to predict the stability between metallic and oxide phases as a function of temperature and oxygen potential mainly driven by the glass.
1] S. Gossé, C. Guéneau, Intermetallics 19 (2011) 621-629
[2] S. Gossé, N. Dupin, C. Guéneau, J.-C. Crivello, J.-M. Joubert, J. Nuclear Mater 474 (2016) 163-173
[3] S. Gossé, S. Bordier, C. Guéneau, E. Brackx, R. Domenger, J. Rogez, J. Nuclear Mater 500 (2018) 252-264
[4] R. Pflieger, L. Lefebvre, M. Malki, M. Allix, A. Grandjean, J. Nuclear Mater 389 (2009)
J. Nuclear Mater, Vol. 389, N°3
[5] O. Pinet, S. Mure, Journal of Non-Crystalline Solids, 355 (2009) 221-227

The hollandite phase is a potential nuclear waste form for alkaline-earth metals such as Cs and Ba by chemical incorporation into the crystalline lattice. The general hollandite formula is A_xB₈O₁₆ with the A-site habitually occupied by large mono- or divalent cations (e.g., Ba, Sr, Na, K, Rb, Cs, Pb), and the B-site by small di-, tri-, or tetravalent cations (e.g., Ti, Al, Fe, Cr, Sc) and it crystallizes in a tetragonal or monoclinic structure depending on the composition. Recently, our group has developed a thermodynamic database to describe the BaO-Cs₂O-TiO₂-Cr₂O₃-Al₂O₃-Fe₂O₃-FeO-Ga₂O₃ system based on the Calphad method, where the hollandite solid-solution was modeled using the Compound Energy Formalism (CEF). Recent findings show the impact of varying A-site Cs stoichiometry of Zn-containing hollandite compositions and related thermodynamic stability. This consequently improved the limited performance of the hollandite phase in a multi-phase composite called SYNROC. We here report on our efforts to include the Zn in our waste form thermodynamic database, modeling the phase equilibria of the complex hollandite-forming oxide waste system.

Canada’s plan for the permanent storage of high level nuclear waste is to construct a Deep Geological Repository (DGR) at a depth of at least 500 m in a suitable host rock. The waste, which is composed almost entirely of CANDU fuel bundles, will be seal-welded into copper-coated steel “used fuel containers” (UFCs), which will be placed in bentonite clay within the DGR. Concurrent with site selection activities to find an informed, willing host community, the Nuclear Waste Management Organization (NWMO) is developing the UFC within a large technical and engineering program. The service life for the UFC is on the order of 1 million years, so research into container degradation, primarily via corrosion, is paramount to the success of the UFC development program. Research is done in partnership with many international waste management organizations, as well as through collaborations with national laboratories and academic research teams.
Overall, four types of corrosion comprise the total expected corrosion: oxic corrosion as a result of trapped oxygen during emplacement; radiation influenced corrosion owing to fuel decay; anoxic corrosion of copper and microbiologically-influenced corrosion. From these processes, some consideration must be made for non-uniform corrosion effects, and the service life must be supported by extensive modeling/simulation. This paper will provide experimental and modeling evidence for prediction of the long-term container corrosion performance, as well as outlining a path forward with respect to corrosion research as the Canadian program transitions through licensing activities and toward construction and operation.

Welded austenitic stainless steels (SS) are used to construct canisters for the dry storage of spent nuclear fuel. Exposure to ambient marine environments can lead to salt deposition and brine deliquescence, which, at long storage times, may lead to localized corrosion and stress corrosion cracking (SCC). Studies on the effects of atmospheric relative humidity (RH), which determines electrolyte concentration and chemistry, on corrosion and SCC and the effectiveness of several SCC repair strategies are being conducted on SS304, a common material for canister construction.

Studies examining the effects of RH quantify the effect of electrolyte composition and exposure time (1 w to 2 y) on pitting corrosion. RH was found to play an influential role in determining pit density and morphology. Specifically, pit density is increased at the 40% RH compared to 76% RH, reaching a limiting pit density of roughly 6X higher. Additionally, at 76 % RH, pits displayed an ellipsoidal surface geometry with smooth facets at the pit base, whereas, at 40 % pits showed evidence of attack along distinct features that appear to be slip bands. Moreover, surface micro-cracks were found to emanate from pits at 40 % RH alone for all exposure times with no applied external load. As such, a hypothesis is presented for the effects of pit morphology and surface cracking at 40% RH based on synergistic electrochemical and microstructural causes. At low RH, high [Cl⁻], mass transport restriction due to precipitated salts at the surface, and preferential dissolution along slip bands results in a fissure-like pit geometry that serves as a stress concentrator. Together with residual stresses from surface grinding possibly forming strain-induced martensite, hydrogen embrittlement is possible to due to cathodic hydrogen generation. Studies are ongoing to confirm this theory. Furthermore, the effect of electrolyte chemistry and electrolyte droplet size are being conducted to determine the cause of the unique pit morphology produced under 40 % RH exposure.

A companion study to investigate the effect of pit morphology on SCC initiation under atmospheric exposure is examining tension samples printed with ASTM sea salt, pre-exposed at 40 and 76 % RH, producing pre-corrosion that will serve as crack initiation sites once loaded. After pre-exposure from 1 w to 2 y, samples are mechanically loaded (at 1.2 X the nominal yield strength) under the same atmospheric conditions. A preliminary investigation for 40 % RH (after 6 months of pre-exposure) show surface cracks after 6 months of mechanical loading. Finite element models are being developed to understand 3D stress concentration profiles around each pit morphology (i.e. 40 % vs 76 %).

A study is also underway to identify which of several candidate repair strategies are effective at repairing already present SCC and maintaining future atmospheric and SCC resistance. ASTM G-36 boiling MgCl₂ testing will be used to downselect between tungsten-arc welding, friction stir welding, vaporizing foil actuator welding, cold spray, and soldering for further research.

This work is supported at The Ohio State University by Sandia National Laboratories (non-repair related research) and the Department of Energy [National Nuclear Security Administration] – Nuclear Energy University Program (DOE-NEUP) under Award Number DE-NE0008765 (repair related research). The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology and Engineering Solutions of Sandia LLC, a wholly owned subsidiary of Honeywell International Inc. for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525. SAND#2020-5962 A

The Nuclear Waste Management Organization plans to manage Canada’s used nuclear fuel by isolating it underground within a multiple-barrier system, in a deep geological repository (DGR), in keeping with internationally accepted best practice.¹ In the proposed design, used fuel bundles will be sealed in copper-coated carbon steel used fuel containers (UFC), encased in blocks of highly compacted bentonite clay, and emplaced ~500 m below ground in the DGR. Any gaps between the rock walls and the bentonite blocks will then be filled with a bentonite gapfill material (GFM). Due to its physical and chemical properties, the bentonite will significantly limit the transport of reactive species to the UFCs. The high swelling pressure of bentonite can suppress the action of sulphate reducing bacteria (SRB) by decreasing the clay pore space and lowering water activity, which then limits microbially influenced corrosion of the UFC.²
The goal of this work is to investigate the effects of bentonite dry density, the presence of oxygen, and the evolution of environmental conditions on the corrosion of copper in contact with bentonite, and on SRB viability. We are conducting a series of experiments in bentonite-filled vessels exposed to ocean conditions (ocean modules, OM) and more controlled laboratory-based conditions (pressure cells, PC). The OM are porous vessels containing copper coupons embedded in GFM that has been compacted to various dry densities. The modules are then placed in the Pacific Ocean at depths of up to 2.6 km for months-to-years at a time, exposed to seawater, with hydrostatic pressures representative of potential DGR pressures. PC experiments contain copper coupons embedded in GFM and are pressurized to 100 bar with Type-1 water for varying durations up to 1 year. After exposure, copper coupons are analyzed by SEM, XPS, FIB, and AES, the water activity of the bentonite is assessed, and 16s RNA analysis used to determine the presence of SRB.
Two OMs, both with cold-spray (Cu_cs) and wrought (Cu_w) copper coupons, with GFM (dry density of 1.25 or 1.45 g/cm³) were exposed to seawater at a depth of 90 meters for 6 months. The surfaces of the copper coupons experienced non-uniform corrosion, possibly due to non-uniform wetting and subsequent swelling of the clay. The copper speciation in the surface corrosion products consisted of ~85 % Cu₂O, ~ 5 % Cu₂S, ~3 % CuCl, and ~ 3 % CuO. Cu_w had higher amounts of Cu₂O and lower amounts of Cu₂S at both GFM dry densities. Both the Cu_w and Cu_cs exposed to the higher density GFM had lower amounts of Cu₂S and higher amounts of Cu₂O on the surface. This may be the result of non-uniform swelling, decreased mobility of the sulphide or a decrease in the number of metabolically active SRBs, or a combination of those factors.² Mass loss measurements determined that Cu_cs had a higher corrosion rate than Cu_w; however, Cu_w corrosion rates were less affected by the GFM density. Overall, the increase in bentonite dry density resulted in a decrease in copper corrosion rates, with the Cu_cs corrosion rate being decreased more significantly than that of Cu_w.
The experiments performed in PCs for 1 month duration provided similar results. Cross-sections of the film formed at the copper surface revealed the presence of copper, oxygen, and sulphur, unevenly distributed across the copper surface, accounting for the non-uniform corrosion observed. The corrosion rates were significantly higher than those in the OM experiments due to the higher oxygen concentration.
1. J. Noronha. Deep Geological Repository Conceptual Design Report Crystalline/Sedimentary Rock Environment, NWMO, Canada, APM-REP-00440-0015 R001, 2014.
2. S. Stroes-Gascoyne, C.J. Hamon, D. Priyanto, D. Jalique, C. Kohle, W. Evergen, A. Grigoryan, D. K. Kober, Microbial Analysis of a Highly Compacted Wyoming MX-80 Bentonite Plug Infused Under Pressure with Distilled Deionised Water over a Period of Almost Eight Years, NWMO, Canada, TR-2014-20, 2014.

The proposed plan for the long-term management of used nuclear fuel in Canada is to seal it in copper-coated steel containers (UFC) and emplace them in a deep geologic repository using a multi-barrier containment concept. While there is extensive literature providing confidence that UFCs will remain intact until fuel radiation levels decay to insignificant levels, it is judicious to investigate the consequences of container failure, when groundwater could come into contact with the fuel. The goal of this project is to investigate the effect of radiolytically produced oxidants on the corrosion/dissolution rates, and the influence of oxidant scavengers on the respective rates for fuel specimens that represent fuel used in Canada.
Within a failed UFC, water radiolysis could lead to the production of localized oxidizing conditions at the surface of the fuel, potentially leading to fuel corrosion and radionuclide release to the groundwater. The rate of such a process would be determined by the radiation dose rate from the fuel, and the reactivity of the fuel, which will be determined by its composition, in particular by the degree of non-stoichiometry and the concentration of fission products.^{1, 2} The corrosion rate at the fuel surface can be suppressed in the presence of oxidant scavengers produced at the steel-groundwater interface. These scavenging ions will consume radiolytically produced oxidants, largely through a catalytic reaction on noble metal particles.³
For undoped UO₂ (i.e., in the absence of fission products) the reactivity has been shown to be very dependent on the degree of non-stoichiometry (x in UO_2+x), which can be estimated by electrical conductivity measurements.⁴ Overall, specimen conductivities are being measured by electrical impedance measurements. The reactivity of characterized specimens is subsequently being determined using either electrochemical measurements or analytical measurements of the amounts of H₂O₂ consumed and U dissolved. Further, measurements of the oxidant scavenging properties of H₂ are being investigated in the absence of noble metal particles to observe the influence of localized non-stoichiometry. This research will work towards expanding the established model for the α-radiolytic corrosion of used nuclear fuel.⁵

1. He, H.; Qin, Z.; Shoesmith, D. W., Characterizing the relationship between hyperstoichiometry, defect structure and local corrosion kinetics of uranium dioxide. Electrochim. Acta 2010, 56 (1), 53-60.
2. Liu, N.; He, H.; Noël, J. J.; Shoesmith, D. W., The electrochemical study of Dy₂O₃ doped UO₂ in slightly alkaline sodium carbonate/bicarbonate and phosphate solutions. Electrochim. Acta 2017, 235, 654-663.
3. Razdan, M.; Shoesmith, D. W., The influence of hydrogen peroxide and hydrogen on the corrosion of simulated spent nuclear fuel. Faraday Discuss. 2015, 180 (0), 283-299.
4. Hyland, G. J.; Ralph, J., Electronic contributions to the high-temperature thermophysical properties of UO_2+x: a critical analysis. High Temp. - High Press. 1983, 15, 179-190.
5. Wu, L.; Liu, N.; Qin, Z.; Shoesmith, D. W., Modeling the radiolytic corrosion of fractured nuclear fuel under permanent disposal conditions. J. Electrochem. Soc. 2014, 161 (8), E3259-E3266.

The Nuclear Waste Management Organization is responsible for the implementation of Adaptive Phased Management (APM), Canada’s plan for the long-term management of used nuclear fuel in a deep geological repository (DGR). Currently in the site selection process, NWMO is seeking partnership with an informed and willing community to host the DGR while technical research is on-going in support of APM. An integral part of isolating and containing the used nuclear fuel is the engineered barrier system which consists of the bentonite clay, a long-lived steel container (with copper coating) and a durable waste form (zircaloy cladding and high-density ceramic fuel pellet). The copper coating consists of a 3 mm copper layer on a steel container and in this current work, a copper corrosion model was developed to predict the long-term uniform corrosion behaviour. The current model builds on the work by King and Kolar (2006) by extending the previous 1D model to include 3D geometry effects. The model was developed in COMSOL Multiphysics using a finite-element approach including 1D and 3D implementations. A step-wise approach was used in the development of the model, starting with a simple reaction scheme and increasing mechanistic detail while ensuring numerical stability and careful validation against experimental results. The developed copper corrosion model predicts the corrosion potential along with environmental parameters (ex. [Cl^-], [O₂]), copper species concentrations and mass transfer rates which can be used to determine, mechanistically, the rate-controlling processes for both the anodic and cathodic reactions. Corrosion potentials can also be used to determine the likelihood of pitting corrosion in the DGR environment and the corrosion current density can be used to predict the rate of uniform corrosion.

References
King, F. and M. Kolar. 2006. Simulation of the consumption of oxygen in long-term in situ experiments and in the third case study repository using the copper corrosion model CCM UC.1.1. Ontario Power Generation, Nuclear Waste Management Division Report, 06819REP0130010084R00

A multi-column adsorption system has been designed and installed to evaluate the independent capture and separation of xenon and krypton utilizing solid phase sorbents being developed by the DOE-NE Offgas Sigma team. The xenon and krypton are present in the offgas stream resulting from a used nuclear fuel reprocessing plant. The system is intended to provide greater operating flexibility and testing at a wide range of flow rates while maintaining desired pressure conditions. This system also allows for continuous analysis of effluent gas from each column. The system was set up by packing the first column with AgZ-PAN which has demonstrated a high selectivity for xenon over krypton at ambient temperatures. A second column was packed with HZ-PAN to capture the krypton. The AgZ-PAN column was maintained at ambient temperatures whereas the HZ-PAN column was maintained at 191K. A simulated offgas stream consisting of 1000 ppmv xenon and 150 ppmv krypton in a balance of air was introduced to the first column to capture the xenon and the resulting xenon free effluent was then routed into the second column for krypton capture. Flowrates of the feed stream were varied from 50-2000 sccm. The results of the multi-column testing demonstrated successful separation of xenon from krypton and indicated that increasing superficial velocity does not significantly impact sorbent capacity or initial breakthrough of the columns. This procedure, while not optimized, may be used to reduce the final waste volume of captured radioactive gas species emitted from a reprocessing facility.

New generation sorbents are being investigated for the removal and immobilization of radioiodine from the gas streams in a used nuclear fuel reprocessing facility. Silver-functionalized silica aerogel (Ag⁰-aerogel) exhibits high selectivity and sorption capacity for iodine. Moreover, a feasible consolidation of iodine-loaded Ag⁰-aerogel to a durable SiO₂-based waste form is an attractive choice for sequestering radioiodine. Extended-duration testing of Ag⁰-aerogel in highly oxidizing reprocessing off-gas streams indicated fragmentation of granules into small particles. These particles were held up in the column without being carried-over into the off-gas system. However, production of small particles during sorption can lead to compaction of the sorbent, restricting the flow of gas through the column. Our study targeted development of mechanically and chemically stable Ag⁰-aerogel through optimized synthesis process, which included using a patent-pending technology treatment to strengthen the granules. The reinforced Ag⁰-aerogel with bulk density ~ 0.5 g/cm³ went through rigorous testing at different off-gas streams followed by attrition testing with a vibratory sieve shaker. There were no fines generated even after 60 min exposure to shaking speeds of ~1800 revolutions per minute, with ~3 mm orbital shaker movement. The mechanically/chemically robust Ag⁰-aerogel retained high iodine sorption capacity of 400.5 ± 2.2 mg/g. The presentation will focus on various strategies that have been developed to improve the mechanical/chemical properties of Ag⁰-aerogel and conclude with a summary of highlights and a prognosis for future developments.

Molten salt systems provide highly efficient heat transfer options without pressurization, making them a potentially valuable tool throughout the energy sector, including solar power and nuclear energy. Molten salt reactors are of particular interest due to the advancements they can offer to the nuclear field, such as improved safety. However, the chemical behavior of key species anticipated in molten salt reactors is not well understood. Tools are needed to monitor and characterize the chemical composition of molten salts to better understand the behavior of actinides, fission products, and corrosion products. Furthermore, these tools are needed to allow for molten salt process control and to support the maintenance of materials accountancy. Of the options available, optical spectroscopy enables powerful routes to characterize these systems in situ. Utilization of these tools allows for the fast and robust identification and quantification of many chemical species of interest to molten salt processing, reactor control, and corrosion characterization. Most importantly, optical spectroscopy can enable the analysis of oxidation state ratios and species complexation, which can significantly impact system performance and corrosion behavior. Optical spectroscopy-based monitoring can be leveraged in fundamental and applied research. For example, optical spectroscopy can be paired with electrochemistry in spectroelectrochemistry to provide a powerful route to fundamental characterization of redox behavior, speciation, and chemical behavior dependence on temperature. Optical spectroscopy can also be paired with chemometric analysis to enable real-time quantification of target analytes within salt melts. Complex salt systems containing many chemical species will exhibit complex spectral data, presenting signal interferences and band overlap, which can hinder accurate analysis. However, the application of chemometric analysis can enable accurate characterization of multicomponent streams to provide real-time, automated monitoring of a process stream or batch. Optical spectroscopy-based on-line monitoring enables fast and robust identification and quantification of many actinides, fission products, and corrosion products. Here we demonstrate the application of optical monitoring approaches to the characterization of chloride-based salt melts containing lanthanides and actinides.

Self-radiation damage from the decay of radionuclides can affect microstructural evolution, phase stability, and thermodynamic properties in nuclear waste forms. The principal sources of radiation are beta-decay of the fission products and alpha-decay of the actinide elements. In general, beta-decay of the short-lived fission products leads to ionizing radiation that is the primary source of radiation and heat generation in high-level waste forms during the first 600 years of storage. Ionization can cause covalent and ionic bond rupture, valence changes, localized electronic excitations, and significant changes in ionic mobility; however, there is no evidence of significant long-term effects of fission product decay on the physical properties of nuclear waste forms. Because of the long half-lives of the actinides and their daughter products, alpha-decay is dominant over very long timescales. Alpha decay produces energetic alpha particles (4.5 to 5.5 MeV) and recoil nuclei (70 to 100 keV) that result in both radiation damage and the accumulation of helium. The evolution of radiation damage due to alpha-decay has been studied at ambient temperature using short-lived actinides, such as ²³⁸Pu and ²⁴⁴Cm, in many candidate glass and ceramic waste forms, and the early stages of evolution occur during the thermal period of the waste form. Ion beam irradiations of candidate ceramics are more frequently employed due to decreased costs, shorter irradiation timescales and the non-radioactive nature of the irradiated materials. Ion irradiations are performed over a wider range of irradiation conditions to implant helium, study radiation damage kinetics, understand the separate and combined effects of alpha particles and alpha recoils, and benchmark against alpha-decay damage due to short-lived actinides. At ion energies (keV to MeV) typically used to study alpha-decay damage in nuclear waste forms, the electronic and nuclear energy losses are both important, and the spatial coupling of electronic and nuclear energy loss can lead to reduced or enhanced damage production along the ion trajectory. While alpha particles are much less damaging than alpha recoils, the electronic energy loss from the alpha particles can also cause athermal annealing of alpha-recoil damage that can impact the evolution of radiation damage. Helium implantation studies have shown that helium bubbles form above a threshold helium concentration in candidate waste forms. Swift heavy ions (several hundred MeV to several GeV) have also been used to study radiation damage in nuclear waste forms; although the damage production mechanism and energy densities are substantially different from beta decay or alpha decay, large volumes are damaged.

In this study, we aimed to investigate the radiation effects in doped-fluorite structure ceramics. CeO₂ was chosen as the host matrix because of its fluorite crystallographic (Fm-3m) structure, which consists of a simple cubic oxygen sub-lattice with the cation ions occupying alternate cube centres, similar to UO₂, PuO_2, and ThO₂. The objective was first to quantify the defect chemistry/configurations when CeO₂ is doped with trivalent yttrium at different concentrations, and then to study the atomistic changes when the solid solutions were irradiated with 3 MeV gold ion under two different fluence, corresponding to a damage level of 6 dpa and 33 dpa values respectively.
Initial X-ray diffraction measurements of the prepared samples showed a contraction of the unit cell with Y-content attributed to a certain long-range ordering of the anion vacancies with randomly distributed Y₂O₃ micro-domains of ‘C-type’ structure which grows coherently within the fluorite matrix. This observation corroborates with Extended X-ray Absorption Fine Structure (EXAFs) analysis which indicated a change in the coordination number when the dopant is introduced in the CeO₂. On the other hand, Raman spectroscopy was used to quantify the presence of oxygen vacancy-type defects formed to compensate for the charge due to the substitution of Ce⁴⁺ with a trivalent dopant and the presence of reduced Ce³⁺.
In regards to the irradiated samples, the excellent radiation stability of Ce-based compounds was reported and it was observed that Au-incident ions tend to reduce Ce⁴⁺ to Ce³⁺. In addition, Positron annihilation spectroscopy measurements showed that the lower irradiation fluence increases considerably the concentration of lattice damages close to the disorder saturation level, where all the implanted positrons are trapped and a further increase in the irradiation dose initiates a dynamic damage recovery. Relevant defect mechanisms for the two considered irradiation fluence studies were identified and will be detailed in this presentation.

Ordered mesoporous silica still attracts considerable attention because of a myriad of industrial potential applications, such as catalysis, gas adsorption, electronics, optics, drug delivery and nuclear waste storage. More precisely concerning the latter example, alternative processes for the treatment of radioactive effluents containing actinides and fission products are developing. In this way, a precise knowledge of the damage induced in those materials, due to the radiation, is then required.
Apart from radiation tolerance of mesoporous metallic foam already reported by Bringa et al.[1], other studies[2] have highlighted the collapse of the mesoporous network of silica-based materials induced by the ballistic radiation damage for an energy deposited close to a few 10²¹ keV/cm³.
In this work, two kinds of samples have been irradiated to evaluate their dimensional effect in electronic regime. In one hand, thin films (~100 nm) of mesoporous silica deposited by dip-coating onto silicon wafer were irradiated at IRRSUD (GANIL) with ions having a stopping power in the 1-10 keV/nm range. In the second hand, pellets obtained by compaction of powders were irradiated at lower stopping power from 10^-3 keV/nm (2 MeV electrons, LSI) to 1 keV/nm (proton ~20 MeV, CEMHTI). These samples were analyzed before and after irradiation by X-ray reflectometry (thin films), nitrogen adsorption isotherm and small angle X-ray scattering measurements (pellets), electron microscopy (Scanning, Transmission) to investigate the evolution of the porous structure and also by spectroscopy (nuclear magnetic resonance, infra-red) to evaluate the evolution of the silica network. A mechanism is proposed to explain and describe the evolution of the mesoporous structure under irradiation.

References:
[1] E.M.Bringa, J.D.Monk and A.Caro,Nano Lett, 12 (2012) 3351-5
[2] L.Yu, S.Dourdain and C.Rey, Microporous and Mesoporous Materials, 251 (2017) 146-154

The stockpile of plutonium separated from used nuclear fuel amounts to many 100s of tonnes. The majority of this plutonium is stored as PuO₂ powder packaged in multi-can containers. The conditions during the packaging of reprocessed PuO₂ for long term storage are closely controlled to limit water uptake as radiolytic decomposition of any adsorbed water will lead to the formation of both a potentially flammable atmosphere containing molecular hydrogen as well as reactive oxygen species which may be incorporated into the oxide phase. Consequently, the safety case for the long-term storage of PuO₂ requires a complete understanding of the fundamental physical, chemical and materials degradation processes occurring at the water-oxide interface inside the storage canisters. This study investigates the effect of ionizing radiation on PuO₂ and various surrogate oxides under atmospheres potentially encountered in storage canisters.
To elucidate the effects of adsorbed water, samples of baked oxide powder were equilibrated in humidity chambers to attain a wide range of masses of water adsorbed onto the powder surface. In the case of PuO₂, once the surface water on the oxide had equilibrated with the humid atmosphere, the samples were sealed into a glass vessel held at constant humidity with either an argon or a nitrogen over-gas and left for up to 3 months. Periodic sampling of the headspace was undertaken and the atmosphere above the oxide analyzed using gas chromatography. In the case of the surrogate oxides, once the oxides had adsorbed the maximum quantity of water, the samples were flame sealed under argon into a glass tube and Co-60 gamma irradiated. Prior to and after gamma irradiation, the oxide powders were characterized using a variety of different microscopy and spectroscopy methods. Following gamma irradiation, the headspace was analyzed using gas chromatography. For PuO₂, the rate at which H₂ was produced increased with increasing water loading. This result contrasts with the experimental data in this study and from similar experiments in the literature utilizing UO₂, CeO₂ and ZrO₂ which showed a higher rate of production for lower water coverages. There was no evidence of a steady state hydrogen concentration being reached in the PuO₂ experiments, an outcome opposite of that observed in storage canisters and in experiments performed by other groups in which the PuO₂ was not held under constant humidity. Radiolytic production of molecular oxygen was not observed. Comparison of the properties of the surrogate oxide (ZrO₂) powder prior to and post irradiation did not reveal any differences in stoichiometry, surface functionalization or the formation of oxygen-centered free radicals, so the location of the sibling oxidant to the evolved molecular hydrogen currently remains unanswered.
Parallel experimental and modeling studies considering the chemistry of PuO₂ and surrogate oxides in H₂:O₂:Ar atmospheres are ongoing. To perform the experiments, a bespoke manifold was designed and commissioned to mix various ratios of H₂-O₂ and Ar gas. Gas chromatography was used to analyse post-irradiated samples. Experiments were performed in 10 ml volume 316 stainless steel canisters (with no degradable polymeric components). Experimental data will be presented showing the extent of recombination of H₂-O₂ mixtures in Ar under ⁶⁰Co γ rays and accelerated He ion irradiation and the influence an oxide surface on this recombination: radiolytic reaction of H₂ with O₂ to form H₂O is enhanced over what is expected from gas phase processes suggesting significant interfacial chemistry.
Acknowledgement: This work was supported by the Dalton Cumbrian Facility Project, a joint initiative of the University of Manchester and the Nuclear Decommissioning Authority, by the UK Engineering and Physical Sciences Research Council and by the US Department of Energy, Office of Nuclear Energy.

Complex microstructural changes occur in nuclear fuels due to the extreme environments of intense irradiation and high temperature during in-reactor operations. Several theoretical and experimental works are currently in progress to determine the effect of crystalline defects which has a major influence on the behaviour of fission products, mainly in regards to rare gases, that may impact the life-span of the rod-type uranium fuel element. For instance, the nucleation of insoluble gas bubbles is thought to act as trapping sites for migrating vacancy-type defects, and further increase in the damage levels induces significant microstructural modifications. In this study, we aimed to investigate the specific role of microstructure on the radiation stability of nuclear fuels.
For this purpose, UO₂ pellets with different microstructures were initially prepared through the wet chemistry route and sintered at different temperatures. Scanning electron microscopy and X-ray diffraction characterisations were performed to quantify the grain size/boundaries areas and crystallinity of the pristine samples. Raman spectroscopy also proved useful and allowed to investigate the effect of sintering temperature on the concentration of intrinsic lattice defects. The samples were then implanted with 1 MeV Krypton-ions at room temperature to induce irradiation damages in the matrix of the different grain-size UO₂ pellets.
Post-irradiation measurements were carried to highlight the excellent radiation stability of fluorite-type nuclear fuels. However, it was observed that the UO₂ microstructure has an important effect on the point defects produced by the displacement cascades induced by ionic implantation. In this presentation, an attempt is made to detail the mechanisms occurring near the defect-boundary interfaces in relation to radiation damage to highlight the precise role of grain boundaries modifying the defects behaviour and their subsequent evolution in fluorite structure relevant to nuclear fuels.

The safe disposal and long-term isolation of spent mixed oxide nuclear fuel poses significant challenges. Uranium and plutonium oxides enriched with rare earth and lanthanide fission products and under constant irradiation with alpha particles and neutrons must be stored in geologic formations for thousands of years. A detailed understanding of the microstructure and chemistry is required to optimize processing designed to recover fissionable products from used fuel and to improve the safety and volume of radioactive waste. Cerium oxide (ceria) has been identified as a suitable non-radioactive substitute to study the microstructure and its evolution under long term irradiation.
A number of techniques have been applied to uranium and cerium oxides to investigate basic properties and point defects. These include neutron scattering, X-ray diffraction and Raman spectroscopy of as produced, irradiated and annealed materials. These are sensitive to structural properties (XRD, neutrons) and oxygen vacancies (Raman). Positron annihilation spectroscopies (PAS) are sensitive to neutral and negatively charged vacancy-like point defects and impurity-vacancy complexes. Positrons can provide depth resolved information about uranium or cerium vacancies to complement Raman data. PAS has been applied previously to UO_2+x to investigate the properties of nuclear fuels.[1,2,3] Virtually no PAS work exists on CeO₂, doped or undoped. Currently, PAS are being used to investigate rare earth and lanthanide doped cerium oxide.
The work on uranium oxide by the group of MF Barthe[1,2] demonstrated the presence of significant concentrations of uranium vacancies or Schottky defects as well as the presence of negatively charged interstitial oxygen atoms in oxygen rich. They also investigated the effect of 1 MeV helium implantation and the interaction of the implanted helium with the formed vacancies during annealing. Xe bubbles and their thermal evolution could also be observed.[4]
In this presentation, the basics of positron annihilation spectroscopy will be presented. PAS work on undoped and doped cerium oxide will be shown and compared to results from uranium oxides. To simulate spent nuclear fuels, the samples were doped at different concentration with rare earth elements or yttrium, zirconium. Select samples were irradiated with heavy ions at different doses. Differences were observed depending on what simulated fission products are present. Preliminary results and literature searches show some correlation with Raman data.
References: [1] Barthe MF et al. EPJ Web of Conferences 115, 03004 (2016), and references therein; [2] Barthe MF et al. Phys. Stat. Solidi C4 3627-3632 (2007); [3] Wiktor J Physical Review B 90, 184101(2014); [4] Djourelov N. et al Journal of Nuclear Materials 432 287-293 (2013)
Acknowledgements: MW, JM, and SK thank the Institute for Materials Research and U.S. Department of Energy Nuclear Energy University Program – Used Nuclear Fuel Disposition program, award # DE-NE0008689. JM thanks the US-UK Fulbright Commission, and JM and SK thank the S.Y. Chung Centenary Fellowship for support during the research period. RM and CC thank Engineering and Physical Sciences Research Council, under the UK-US Nuclear Energy University Programme (EP/R006075/1), an Early Career Research Fellowship awarded to CC (EP/N017870/1), and the MIDAS Facility, at the University of Sheffield, established with support from the UK Department of Energy and Climate Change.

The Swedish strategy for spent nuclear fuel has, since around 1980, been direct geological disposal. Research about the properties and dissolution of spent fuel has been going on since then. An overview of this research program is given here with a focus on questions posed and answered along the way, ending with remaining open questions. At the beginning, some basic studies were performed to gain information about spent fuel as wasteform. Dissolution studies were at first mainly performed in contact with air, and often with aqueous solutions that mimicked the expected groundwater chemistry. However, complexities relating to precipitation of secondary phases indicated that simplified water compositions were preferred. Knowing that the radioactivity of the fuel caused radiolytic oxidation of uranium and thus controlled the dissolution rate, there were also attempts to achieve reducing conditions via bubbling a hydrogen/argon mixture in presence of a palladium-catalyst. The first attempts resulted in variable results, but notably, a reduction in apparent leach rate was observed for some experiments [1]. Later, possibilities to perform the leaching experiments in autoclaves, with controlled hydrogen atmosphere, allowed experimental conditions closer to the expected repository conditions. These experiments showed that hydrogen, during this type of experimental set-up, could suppress oxidative dissolution caused by radiolysis [2,3]. A new phase of research was thus started where the mechanistic explanation for the observed behaviour was sought. Experimental work, both with spent nuclear fuel and with analogue materials, was coupled with modelling efforts. At the time of the SR-Site safety assessment [4], which was a part of the license application submitted in 2011, a certain understanding had been reached, based on the effect of metallic particles in fuel. Still, questions remained and studies performed during the last decade (see for example [5]) can now be summarised to illustrate the knowledge gained, and remaining open questions. This summary is utilized in the next step in the licensing process: the application to start building the spent fuel repository, which is expected to be submitted within the next couple of years. In the coming decade the SKB research program will continue, with the goal to support the development of the safety assessment report, SAR, which will be submitted to the authorities as a part of the license application connected to start of trial operations.
During all these years, international collaboration has been a cornerstone of the SKB research program. This international collaboration has been carried out via European projects or bilateral agreements, and triggered the start-up of the Spent Fuel Workshops [6]. The research program has also been presented at this MRS symposium since the early 1980’s.
[1] Forsyth R, 1983. The KBS UO2 leaching program. Summary report 1983-02-01. KBS/SKBF Technical Report 83-26.
[2] Spahiu, K., Werme, L. and Eklund, U.B., 2000. The influence of near field hydrogen on actinide solubilities and spent fuel leaching. Radiochimica Acta, 88(9-11), pp.507-512.
[3] Werme, L.O., Spahiu, K., Johnson, L.H., Oversby, V.M., King, F., Grambow, B. and Shoesmith, D.W., 2004. Spent fuel performance under repository conditions: A model for use in SR-Can. SKB-TR-04-19. Swedish Nuclear Fuel and Waste Management Co.
[4] SKB, 2011. Long-term safety for the final repository for spent nuclear fuel at Forsmark. Main report of the SR-Site project. SKB TR-11-01, Swedish Nuclear Fuel and Waste Management Co.
[5] Barreiro Fidalgo, A., Kumagai, Y. and Jonsson, M., 2018. The role of surface-bound hydroxyl radicals in the reaction between H2O2 and UO2. Journal of Coordination Chemistry, 71(11-13), pp.1799-1807.
[6] Werme LO, 1984. Proceedings of the third fuel workshop. KBS/SKBF Technical Report 83-76

The addition of dopants, including chromium and aluminium, to uranium dioxide fuel results in the formation of larger grained microstructures, capable of higher fission gas retention and reduced swelling during operation within nuclear reactors. Despite their growing use by nuclear power plant operators, the long-term dissolution behaviour of spent doped fuels, when compared to typical spent fuel, is not yet well constrained. In a geological disposal facility, the release of radionuclides is known to be controlled by the oxidation of the UO₂ matrix and the subsequent dissolution of UO₂²⁺ species, therefore any changes to the matrix from the addition of dopants could be expected to influence the dissolution kinetics and/or mechanisms.

We here present the results of an investigation of how the addition of Cr₂O₃ to UO₂ fuel influences the crystal chemistry of UO₂ and its dissolution kinetics. Initial results, from experiments performed in air and a bicarbonate solution, show that the addition of Cr₂O₃ reduces the normalised mass loss of uranium. The potential influencing factors that give rise to increased durability with Cr addition, e.g. defect concentration, Cr-oxidation state etc. are discussed.

The “back end” of the nuclear fuel cycle corresponds to the direct disposal or reprocessing of the spent nuclear fuel (SNF), and to the conditioning and disposal in an appropriate repository of the nuclear wastes. In most of these processes, the behavior of actinides and fission products is governed by dissolution reactions at solid/ liquid interface. The design of reprocessing process and of nuclear waste disposal strategy requires the determination of well constrained kinetic and thermodynamic data. The acquisition of reliable kinetic and thermodynamic data is a tough issue considering on the one hand, the complexity of SNF in terms of chemical composition and microstructure and on the other hand, the challenge posed by the manipulation of highly radioactive samples.

For several years, we developed and improved a methodology that aims to determine dissolution rates of uranium-based oxide materials and thermodynamic data of phases of interest for final disposal of spent fuel in deep-mined geological repositories. This methodology is based on the synthesis of model compounds with well defined composition and microstructure, on the measurement of macroscopic dissolution rates in controlled chemical and hydrodynamic conditions, and on the monitoring of the properties of the solid/ liquid interface at the microscopic scale. For that purpose, several surface characterization techniques were used and developed, such as 3D reconstruction of Environmental SEM images, AFM or Grazing Incidence-XRD. To illustrate the progress made beyond the state of the art by using this methodology, several examples have been selected among the studies that we reported in the literature. The selected examples deal either with the reprocessing of uranium based oxides in nitric acid conditions, or with the behavior of UO₂ in SNF under the weathering conditions of geologic disposal.

The mechanisms that control the kinetics of dissolution of U-based oxides in nitric acid solutions are illustrated through the study of samples showing different compositions and microstructures. The use of oriented UO₂ single crystals allowed to evidence the enhanced reactivity of the surface at defect sites during a first kinetic regime. The second kinetic step was attributed to a catalyzed dissolution mechanism involving species produced at the solid/ liquid interface during the first step. The key role of catalytic species in the strong increase of dissolution rate was demonstrated not only for UO₂ single crystals, but also for sintered pellets of (U, Th) mixed oxides and UO₂-doped with platinum group metals. The monitoring of solid/ liquid interface by 3D-ESEM during uncatalyzed and catalyzed regimes revealed specific dissolution sites of enhanced reactivity and specific dissolution features related to UO₂ crystallographic orientation, which traduced a structural control of the topographic evolution of the material.

Most of the geologic sites under investigation for underground repositories are located in undisturbed clay-rich rock or granite, with silica-rich groundwaters, deep enough to have reducing conditions. Understanding the interaction of used fuel with the silicate-rich groundwaters is critical to evaluate the safety of different disposal strategies. For several years, we investigated the role of silicates in the mechanisms of UO₂ weathering. For the first time, we were able to show the formation of coffinite (USiO₄) from UO₂ in the presence of solution saturated with respect to SiO₂(am) under conditions typical of deep-mined geologic repositories for SNF. Through dedicated experiments, we have constrained the conditions of formation of coffinite in an Eh-pH diagram. Finally, we showed that coffinite precipitation could lower the uranium release from the UO₂ matrix of SNF through oxidative weathering in the presence of oxygen in the geological repository.

In an effort to incorporate high levels of Mo associated with the post-operational clean-out of high active liqor storage tanks at the Sellafield reprocessing facility in the UK, a modified version of the "MW" sodium aluminoborosilicate base glass has been adopted. The addition of Ca favours the formation of CaMoO₄, while Zn infers improved viscosity. This new Ca/Zn MW base glass is currently being used to immobilise reprocessing wastes from oxide fuels and also magnox fuels, which contain a significant proportion of Mg. The divalent cations Ca²⁺, Zn²⁺ and Mg²⁺ are known to exert an influence on the durability of vitrified wastes, however the nature of their additive, or competitive effects are yet to be fully understood.

We here describe the results of a range of experiments that aim to elucidate the influence of divalent cations on the mechanims and kinetics of glass dissolution. The addition of Zn to the glass was found to enhance durability in dilute, Si-undersaturated conditions, due to an increase in glass polymerisation with increasing Zn content, as determined by ²⁹Si MAS NMR. However, under conditions of Si-saturation, the presence of Zn prevented the formation of a passivating silica gel layer, and promoted the formation of Zn-containing smectite clays, significantly enhancing the dissolution rate with increasing Zn content. Mg had a similar effect to Zn, which was additive, however glasses containing Ca generally showed lower dissolution rates, in agreement with other studies. The formation of Zn- and Mg-bearing silicates was concurrent with a significant increase in the dissolution rate, similar to Stage III "rate resumption" behaviour, which has previously only been associated with the formation of zeolites.

Safety assessment of Spent Nuclear Fuel (SNF) disposal in a Deep Geological Repository (DGR) requires evidences of radionuclides release after groundwater eventually gets into a breached container and contacts the fuel. Within the frame of EU H2020 program into the DISCO project, the impact of additives (Cr, Cr/Al and Gd) is being investigated on the long term DGR dissolution performance and compared to conventional UO₂. In particular, the effect of these dopants is studied on the long term matrix dissolution rate of the SNF, the so-called Advanced Tolerant Fuels (ATF), such as Cr₂O₃-doped UO₂, and Cr₂O₃-Al₂O₃-doped UO₂ fuels.
ATF have demonstrated to reach higher burn-up than conventional UO₂ fuels because dopants provide enhanced safety against Pellet-Clad Interaction failure (PCI), higher density, enlarged grain size, lower fission gas release and higher creep rate. Besides, the difference between those ATF and standard UO₂ in thermo-physical properties is almost negligible [1-4]. However, the dopants influence upon the matrix leaching performance at long term under DGR conditions is still largely unknown when compared to traditional UO₂ SNF.
The combinatory effects of two uncertain parameters (dopants influence and H₂) on UO₂ matrix leaching resistance are investigated. In the context of WP2 and WP4 of the DISCO project, UO₂-based pellets doped with a nominal concentration of 0.06^w/_oCr₂O₃, 0.05^w/_oCr₂O₃-0.02^w/_oAl₂O₃ and 4.5^w/_oGd₂O₃, were systematically produced by homogenizing stoichiometric proportions of UO₂ and dopants, and carefully characterized in order to understand the effects of the addition of Cr- and/or Al-oxide into the fuel matrix dissolution behaviour. Gd-doped UO₂ was also included, as a burnable poison and a trivalent rare-earth dopants (RE^III). The results are the basis for potential more complex systems which would correspond to irradiated doped fuels.
Previous studies suggest that RE^III play an important role on the UO₂ oxidation, revealing that they prevent the fuel from a complete oxidation to U₃O₈ [5], so in order to better understand the effect of trivalent RE^III, a set of dissolution experiments is carried out. The presence of reducing species, such as H₂ (produced by anaerobic iron corrosion and radiolytically), affects the reactivity of UO₂; hence, it must be considered in a long-term safety assessment. On the other hand, it is well known that HCO₃^- (one of the key complexing constituents in the groundwater) is highly likely to influence fuel dissolution by enhancing the solubility of U^VI thus preventing the secondary phase formation [6].
To experimentally evaluate the effect of additives (doped fuels) together with H₂(g) and different leachants on UO₂ matrix dissolution rate, leaching experiments were performed under different representative aqueous systems. Prepared pellet/disks were characterized by means of XRD, Raman spectroscopy and SEM-EDX. In order to understand the effect of the presence/absence of HCO₃- that could be considered as relevant to those coming from the interaction of water between natural barriers (groundwater) and EBS (Engineering Barrier System), specific synthetic groundwater was used in the dissolution tests. Static leaching experiments of doped-UO₂ pellets were run in stainless steel autoclaves with reaction vessels made of PEEK autoclaves under a N₂/4.7%H₂ atmosphere. The U and dopants concentration released into the aqueous solution was quantified by ICP-MS. pH and redox measurements were also tracked during the experiments.
[1] Thomas, G.A. et al. (2013) LWR Fuel Performance Meeting. TopFuel 201.
[2] Gentet, G. et al. (2015) Top Fuel 2015 Conference Proceedings A0198, 170 – 176.
[3] Backman, K. et al. (2009) INIS-XA--1074. International Atomic Energy Agency (IAEA).
[4] “Technical Exchange Meeting: EnCore® Accident Tolerant Fuel ADOPT Fuel Pellets”.
[5] Travis A. et al. (2020). J. Nucl. Mater. 530, 151959.
[6] Hossain, M.M.et al. (2006). J. Nucl. Mater., 358, (2-3), 202-208.

For a future deep repository for spent nuclear fuel, the rate and mechanism for dissolution of fission products and actinides from spent fuel is a key parameter for the safety assessment of the disposal concept. Although the majority of the fuel to be placed in the deep repository is expected to be intact at the time of disposal, some failed fuel rods will also be present. In a failed fuel rod, the uranium matrix may have been exposed to water in the reactor and during interim storage for several years or decades prior to disposal. This exposure can lead to oxidation of the matrix and the formation of secondary phases under conditions of low uranium solubility. Upon barrier failure and ground water exposure of the spent fuel in the deep repository, these pre-oxidized layers are expected to dissolve at a higher rate compared to unaffected UO₂ and hence increase the rate of radionuclide release. For this reason, knowledge about the characteristics of failed fuel and its behavior under deep repository conditions is required for the evaluation of the long-term safety of geological repositories. The present work aims to add to this knowledge.

Although the behavior of spent nuclear fuel under deep repository conditions has been studied extensively, studies of failed fuel are, in this context, relatively scarce, in particular studies of failed fuel that has been exposed to water during decades of wet interim storage. However, at the Studsvik laboratory in Sweden, leaching tests of spent fuel have been performed since the 1980s. This provides an opportunity to investigate the effects of long term contact with water under aerated conditions, such as those occurring in a wet interim storage.

In this study, XRD investigations have been performed on the surfaces of two fuel pellet samples which have been exposed to leaching solutions (synthetic ground water and deionized water, respectively) for over 30 years. Furthermore, the fuel surfaces have been exposed to carbonate solutions in an effort to dissolve the oxidized phases formed during the extended leaching time. The aim is to characterize the fuel matrix alteration.

The development of robust safety cases for geological disposal of spent fuel requires a solid understanding of its dissolution over very long time scale (up to a million years). While the dissolution of standard spent fuel has reached a good level of comprehension, the development of modern LWR fuels (Cr/Al-doped UO₂ or MOX fuels) raises the question whether these fuels will behave similarly to standard fuels in geological repository conditions. Within the EU-DisCo project, systematic dissolution studies have been initiated with carefully prepared and characterized, simplified UO₂-based model materials to understand the matrix dissolution of these modern LWR fuels under geological repository relevant conditions. As DisCo partner SCK CEN has performed such dissolution tests to study the effect of Cr-doping on the UO₂ matrix dissolution rate in reducing conditions_. Tests were started on four model materials : UO₂, Cr-UO₂, Pu-UO₂ and Pu/Cr-UO₂ with a density close to 94% of the theoretical density and average grain size around 10 µm for UO₂ and Pu-UO₂ , 35 µm for Cr-UO₂ and 25 µm for Pu/Cr-UO_{2 .} The Cr doped pellets contain 1600 µg Cr₂O₃ per gram of UO₂, and the alpha doping level of the Pu-doped pellets (~ 18 MBq/gUO₂) is representative for a fuel age of 10,000 years. Static dissolution experiments with these model materials have been performed in autoclaves in two reference waters (synthetic young cementitious water at pH 13.5 and bicarbonate water at pH 9) under 10 bar H₂ pressure and in presence of Pt/Pd catalyst. After a predissolution in 0.01 M NaCl solution for 3 weeks to remove the pre-oxidized UOX layer, samples were taken from the solution over a period up to 2 months. For the materials UO₂ and Cr-UO₂, the Cr-doping leads to lower uranium concentrations in solution, both in synthetic cementitious and in bicarbonate water. In the synthetic cementitious water, this effect is constant over the entire experiment, but the concentrations, increasing immediately after the replacement of the NaCl solution by the cementitious solution, remain high at ~10^-7 M for UO₂ and ~10^-8 M for Cr-UO₂. In bicarbonate water, the initial uranium concentrations (after replacement of the NaCl solution by the bicarbonate solution) are the same as for the cementitious water, but after 2 months, all concentrations are ~10^-9 M. These preliminary results suggest that the Cr doping decreases the U release in solution, but the long-term effect depends on the leaching solution. Experiments with the Pu doped materials (Pu-UO₂ and Pu/Cr-UO₂) are now carried on to study the Cr-effect in combination with alpha-radiolysis of the solution.

Programs for deep geological repositories are being developed in many countries to safely dispose of spent nuclear fuel. Therefore, understanding the behavior of spent nuclear fuel under relevant conditions is crucial for the safety assessment of these facilities. During the last decades, new types of nuclear fuel are being introduced in commercial reactors. To allow these new fuels in the spent fuel repository, it must be shown that they do not dissolve at a higher rate than standard fuel. Fuel additives such as chromium oxide (Cr₂O₃) have an impact on the UO₂ microstructure and therefore might affect properties relevant to the safety assessment.

The main goal of this investigation is to understand the effects of Cr-doping on the instant release fraction of fission products and the matrix dissolution under oxidative conditions. A sample consisting of fuel fragments and separated cladding is leached in simplified groundwater under aerated conditions at Studsvik’s Hot Cell Laboratory. The fuel used in this study was manufactured by AREVA and has been irradiated in a Swedish commercial pressurized water reactor (PWR) to a local burnup at the sampling position of ~59 MWd/kgU.

Data from the leaching experiments are presented and compared to other commercial fuels’ leaching performance: Cr-doped fuel with lower doping levels (Advanced DOped Pellet Technology, ADOPT) and standard UO₂ fuel.

The direct disposal of spent modern nuclear fuel (SNF) in a deep geological repository (DGR) requires an in-depth understanding of the long-term corrosion behavior of the SNF and the associated radionuclide source term. These are indispensable requirements for demonstrating the long-term safety of the DGR. Today’s modern Cr- and Al-doped light water reactor (LWR) fuels have already successfully demonstrated their improved in-reactor performance, but it is still not known, whether the corrosion behavior of such fuels in a DGR is similar to conventional spent LWR-fuels. Due to the chemical and structural complexity of SNF and its high beta and gamma radiation field, experiments with SNF cannot unravel all of the various concurring dissolution mechanisms entirely. Thus within the EU-DisCo project single-effect dissolution studies are carried out with systematically produced and carefully characterized UO₂-based model materials, to provide additional insights into the behavior of modern doped LWR-fuels under the post-closure conditions expected in a DGR, complementary to dedicated dissolution experiments with SNF.
Here, we present recent results on the dissolution behavior of tailor made UO₂ model materials in accelerated static batch experiments using H₂O₂ as simulant for radiolytic oxidants, present in long-term disposal scenarios for SNF in failed container conditions due to the alpha-irradiation of water. In these dissolution experiments pure UO₂ reference pellets exhibiting different densities and grain sizes, as well as Cr-doped UO₂ pellets with various Cr-doping levels, produced using different doping methods having different grain sizes, were used. In addition, Nd-doped and industrially produced Cr- and Cr/ Nd-doped UO₂ pellets were used to determine the influence of these parameters on the dissolution rates. The dissolution experiments were performed under strictly controlled conditions with respect to exclusion of oxygen, temperature control, and exclusion of light. During the dissolution experiments, solutions were systematically sampled and the H₂O₂ (UV-Vis) and the uranium (ICP-MS) concentrations were determined.
The initial dissolution rates of all materials were determined from the concentrations of dissolved U(VI) obtained within the first 24 hours of the experiments. In experiments with pure UO₂ reference pellets a clear linear dependence of the initial dissolutions rates from the density of the pellets was observed, with higher initial dissolution rates at lower densities. No dependence on grain size was found. The initial dissolution rates of pure, Cr-doped, and Nd-doped UO₂ pellets fall into three groups: i) pure UO₂ pellets with the fastest dissolution rates, independent of the oxygen potential during sintering (and therefore the grain size), ii) Cr-doped UO₂ pellets, independent of doping method, doping level, and oxygen potential during sintering (grain size), and iii) Nd-doped UO₂ pellets with the lowest dissolution rates. The following initial dissolution rates were calculated for the different materials:
pure UO₂: 1.3 +/- 0.1 * 10^-6 mol/m²*s
Cr-doped UO₂: 0.65 +/- 0.2 * 10^-6 mol/m²*s
Nd-doped UO₂: 1.63 +/- 0.7 * 10^-7 mol/m²*s
The slight difference between the dissolution rates of pure UO₂ and Cr-doped UO₂ pellets (i.e. factor 2) can partly be explained by the slightly lower density of pure UO₂ pellets compared to Cr-doped UO₂ pellets. A very strong influence of Nd, representative of fission lanthanides in SNF, on the dissolution rates in this type of experiments is clearly visible. Although the density of the Nd-doped UO₂ pellets is similar to the Cr-doped pellets (i.e. 96-98% relative density), the initial dissolution rate is more than half an order of magnitude lower than for Cr-doped pellets. Overall, the results of this study indicate that the Cr-doping of modern nuclear fuels seems to have no significant effect on the corrosion behavior of UO₂-based SNF.
.

The dissolution rates of six calcium aluminosilicate glasses were determined experimentally at pH(25 °C) = 7.5 and a temperature of 75 °C in a Single-Pass Flow-Through (SPFT) system. Glasses studied in this work can be categorized into two sets of three compositions each. In the first (metaluminous) set, the molar proportion of CaO to Al₂O₃ remained the same (1:1), while SiO₂ increased (40, 50 and 60 mol.%). The second (peralkaline) set was characterized by constant Al₂O₃ (10 mol.%), with decreasing CaO and increasing SiO₂ (40, 50 and 60 mol.%). Measurements of ²⁷Al by MAS-NMR indicated a slight increase in Al^[5] at the expense of Al^[4] with decreasing Si concentration. One to four grams of powdered glass (300 to 425 mm diameter) were placed into 10 mL volume reactors with solution input through the bottom and effluent solution output through the top. Solutions were buffered by 0.05 M TRIS and flow-rates varied between ~38 and 500 mL/d. A plot of the dissolution rate vs. flow-rate (q) to surface area (S) showed that all dissolution rates were maximized at q/S ≈ 45 × 10⁸ m/s. Dissolution rates were from 4.2, 1.5 and 0.62 × 10^-7 mol/(m² sec) for the metaluminous and 1.0, 0.65 and 0.50 × 10^-7 mol/(m² sec) in the peralkaline glasses, respectively. Interestingly, the data indicate that rates for the metaluminous trio are generally faster than those of the peralkaline set, which runs contrary to expectations. As a set of glasses, the log₁₀ rates [mol/(m² sec)] are inversely proportional to the fraction of Si in the glasses, forming a single array of values. These data indicate that rates are strongly tied to the Si content of glasses, and that other glass composition parameters— including Ca/Al (molar) ratio—exert a much weaker effect on dissolution rate in this medium.

The UK intends to ultimately dispose of its nuclear waste by storing it deep underground within a geological disposal facility (GDF). The planned setup is to vitrify high-level waste (HLW) within glass, encase this in a metal canister, and store this using a buffer within the host rock. Currently, there is no agreed site for a UK GDF, and therefore, accurate modelling is required despite significant uncertainties in the modelling parameters. This study aims to build upon previous work that simulated HLW radionuclide migration within both clay and granite host rocks. Significant updates have been made to the original GoldSim Monte-Carlo models, including updating the HLW inventory, radionuclide sorption, diffusion, and solubility coefficients, canister degradation failure rates, number of waste canisters, granite/clay density, sensitivity analysis, and Monte-Carlo parameters. In addition, the effect of HLW dissolution has been examined by varying the source-terms used (initial dissolution, residual, and resumption rates) through variation on the original simplistic source-terms, and by coupling GoldSim with the analytical GRAAL model (via MATLAB). Results demonstrate that radionuclide migration behaviour is similar between the original and updated GoldSim models, despite cumulative activities being approximately one to two orders of magnitude lower in the upper part (crown) of the GDF. Variation on the original simplistic glass dissolution source terms has an understand but insignificant effect on GDF performance, with there being no substantial difference having used the original model glass dissolution rates versus those applied by the UK nuclear industry. Finally, radionuclide activities are strongly correlated with systematic modifications in analytical GRAAL model parameters through changes in glass degradation rates.

Predicting the release rates of radionuclides from nuclear waste glasses during aqueous dissolution within a disposal facility over geological timescales requires robust models of dissolution built upon an extensive understanding of glass dissolution mechanisms. However, many of the dissolution models presented in the past decade conflict regarding the mechanisms predicted to control dissolution. Whilst nanoscale post-process analyses of altered glasses have afforded insights into altered layer structures, these results highlight the need for an additional dimension to understand the mechanisms of glass dissolution beyond the conventionally employed techniques presented in the literature.
In a previous study, lithium isotope fingerprinting techniques were applied to the leachates of a simplified Magnox waste glass leached at 90 °C. It was shown that whilst dissolution was initially congruent, lithium isotopes leached incongruently through diffusive processes across a passivating altered layer component at longer durations. However, further evidence was required to rank the dissolution models presented in the literature. Here, boron isotope fingerprinting techniques were applied to the same Magnox glass to further understand its dissolution processes, understand how lithium and boron were leached and distinguish between dissolution models. Additionally, the same isotopic techniques were applied to leachates of an international benchmark glass composition to compare the dissolution processes of the two compositions and allow for interdiffusion models to be fitted to the results. Solution renewal experiments took place to further investigate the passivating nature of the altered layers and their role as a diffusive barrier to dissolution at long durations.
Boron isotopes leached from the Magnox glass were consistent with congruent models of dissolution in the initial stages of dissolution, but at later stages were consistent with diffusive isotopic fractionation across a passivating altered layer component about an incongruent dissolution front. Whilst boron isotopes in the leachates of International Simple Glass were consistent with diffusive, incongruent dissolution models throughout dissolution, boron isotopes in its solution renewal experiment were most consistent with congruent models of dissolution. As such, these isotopic methodologies demonstrate that no single dissolution model may be correct for all glass compositions and experimental conditions; providing an improved understanding of the mechanisms governing aqueous glass dissolution and supporting the idea of a model which unifies the mechanisms presented in the literature.

The rates of aqueous dissolution and radionuclide release from glasses used to immobilise high-level waste that will be disposed of within a geological disposal facility are strongly dependent on composition. Glasses are challenging to reliably fabricate because after liquid cooling, they are considered to be in non-thermodynamic equilibrium, and therefore their state variable, for example volume, cannot be predicted solely from temperature and composition, but also requires consideration of factors such as thermal history. Other material properties therefore prove valuable as descriptors of glass state, for example, glass viscosity or density. In addition, phase separation can occur depending on the glass composition, particularly with increased waste loadings. This study uses machine learning to predict glass viscosity, viscosity model fitting parameters, density, and phase separation, aiming to both predict glass properties and extract relevant compositional dependencies from the analytics. A series of algorithms have been applied both in isolation and in combination, for example, via voting or stacking. Data has been provided by the nuclear industry or by making use of large-scale international datasets. Overall, results indicate that machine learning can be valuable either in accurately predicting the properties of nuclear waste glasses (viscosity, viscosity model fitting parameters, density, phase separation) or can be used to analyse the importance of different features either for determining particular properties directly or indirectly by correlating features and prediction errors.

Layered uranium oxyhydroxide (LUOH) phases are important minerals which occur as weathering products of uraninite deposits and also importantly, are encountered during the paragenetic sequence of secondary phases that form during an accident scenario involving spent nuclear fuel (SNF) in the near-field (1). Understanding the structural-chemistry of these phases is of importance as via interaction mechanisms, they can potentially retard the transport of harmful fission daughters from SNF towards groundwater and the biosphere. Particularly it has been previously shown that the secondary LUOH phases, such as metaschoepite, uranophane and studite, can act to impede and retard the transport of fission daughters and SNF products such as NpO₂⁺, Cs⁺ and Sr⁺ (2,3). Chemically, the origin of these favourable interactions is predominantly due to favourable intercalation chemistry between cationic SNF species with that of anionic lattices of encountered phases. However in the case of anionic fission daughters, such as ¹³¹I and ⁷⁹Se, these present considerable concern during release from SNF as the overwhelming majority of near-field SNF secondary phases and minerals possess anionic lattices, including LUOH phases, which are typically not expected to react and intercalate well with. Although some evidence is present for partial intercalation of anionic iodine into layered mineral structures through ionic exchange mechanisms in model system studies (4-6) the current consensus is the retention of anionic SNF release products in the near field is not proven and best considered in the far-field of SNF repository performance and accident scenario assessment. Layered double hydroxide (LDH) phases are considered prime candidate materials and minerals for the retention of these anionic phases and received considerable attention (7), however in the context of an accident scenario with SNF their occurrence is not certain. We have recently undertaken a comprehensive systematic investigation of actinide solid state systems relevant to SNF and nuclear waste management. In the course of our investigations we have examined the chemical interactions that may occur between the anionic fission daughter species, I and Se in different chemical forms with some potential LUOH phases associated with SNF. Using a combination of batch synthesis experiments and in situ exchange experiments, we provide evidence for the inclusion and retention of these anionic species in LUOH phases. Importantly, using single crystal X-ray diffraction we show the ability for anionic species to be incorporated into certain LUOH phases is a consequence of their respective topological arrangement of uranyl and hydroxyl groups. Furthermore, by considering the topology and structural chemistry of LUOH phases, it can allow for a predictive basis to be established towards which the relative interaction affinity between certain anionic fission daughter species and specific LUOH phases can be understood. The systematic experimental studies used in this investigation will be discussed in detail w.r.t potential interactions that occur between anionic fission daughters and LUOH derived secondary phases that occur during an accident scenario involving SNF.

References:
1 P. C. Burns, R. C. Ewing, and A. Navrotsky, Science 335, 1184 (2012).
2 J. I. Friese, M. Douglas, E. C. Buck, S. B. Clark, and B. D. Hanson, MRS Proceedings 824, CC9.2, Cc9.2 (2004).
3 A. Espriu-Gascon, J. Giménez, I. Casas, and J. de Pablo, Journal of Hazardous Materials 353, 431 (2018).
4 S. Wu, S. Wang, A. Simonetti, F. Chen, and T. E. Albrecht-Schmitt, Radiochim. Acta 99, 573 (2011).
5 S. Wu, F. Chen, M. Kang, Y. Yang, and S. Dou, Radiochim. Acta 97, 459 (2009).
6 S. J. Wu, F. R. Chen, A. Simonetti, and T. E. Albrecht-Schmitt, Environ. Sci. Technol. 44, 3192 (2010).
7 N. Daniels et al., Applied Clay Science 176, 1 (2019).

Stuffed pyrochlore ceramics, with the general stoichiometry Ln₂TiO₅, have received significant interest for application within the nuclear industry due to their high radiation tolerance, good thermal stability, and structural flexibility. These ceramics have been proposed as a potential wasteform for actinide wastes, for incorporation into inert matrix fuels (IMF), and as inclusion in oxide dispersion strengthened (ODS) steels. These materials can be viewed as derivatives of pyrochlore structured Ln₂Ti₂O₇ phases and are often more conveniently written in the form Ln₂(Ln_0.67Ti_1.33)O_6.67, where the pyrochlore B-site now contains mixed occupancy. These materials can adopt a variety of different crystal structures depending on the size of the lanthanide cation, and the temperature and pressure used during synthesis. The three reported crystal structures are orthorhombic (Pnma), hexagonal (P6₃/mmc) and cubic defect-fluorite (Fm-3m). Compounds that are nominally assigned as defect-fluorite by X-ray diffraction have been probed by neutron diffraction (ND) and transmission electron microscopy (TEM) revealing features consistent with the presence of short range ordered regions with the (Fd-3m) cubic pyrochlore structure.

In this contribution we present a study probing the local structure of a series of Ln₂TiO₅ (Ln = La, Gd, Dy, Yb, Y) compounds. They were synthesised via a solid state synthesis route under different cooling regimes to prepare samples of identical chemistry but different short range ordering. Fabrication of high quality sintered specimens of Ln₂TiO₅ phases is a particular challenge due to their refractory nature. Therefore, a systematic study of sintering parameters was undertaken to optimise the fabrication of sintered ceramic specimen. Characterisation of these materials was performed using techniques including XRD, SEM/EDX and XAS. Initial XAS results show that some of these materials do have a cubic pyrochlore-like local structure but show a deviation from the ideal coordination environment of cubic pyrochlore.

Oxides with generic stoichiometry A₇O₁₂ occupy, in a generic phase diagram, an intermediate place between fluorite and bixbyite structure types. Their structure is derived from the ideal fluorite structure, and it is characterized by a more or less pronounced ordering of the O vacancies within the average fluorite sublattice. We believe that ternary oxides with this generic formula display some structural flexibility relevant to fission product stabilization in nuclear fuels and to the understanding of actinide waste forms as well. We would like to discuss the structural characteristics of the chemical bonds in these ternary systems consisting of trivalent and hexavalent cations studying the electron density using the tools of the quantum theory of atoms in molecules (Bader's AIM theory and ELF). Eventually, in these compounds with generic formula A^III₆B^VIO₁₂, the A cation can be either an actinide or a rare earth element, whereas the B cation can be either a transition metal (W, Mo) or hexavalent uranium. We believe that studying the characteristics of the polyhedron of the sevenfold coordinated A ion is particularly interesting: it is often described in the literature as a mono-capped trigonal prism, with bond lengths exhibiting an extremely large dispersion, ranging for instance from 2.19 to 2.70 Å in Y₆WO₁₂. We would discuss the implications of this large dispersion of distances on the chemical bond characteristics, compare DFT models with experiments on pristine and irradiated specimens, and discuss the competitive metastability of the oxygen deficient cubic fluorite phase that can be observed in some of these systems, as for instance Y₆MoO₁₂.

The analysis of individual micrometre- and submicrometre-sized particles collected by IAEA safeguards inspectors on swipe samples during in-field verification activities requires the implementation of a sustainable quality control system. This system needs to be further advanced by developing mass spectrometric analytical measurements and of suitable reference materials. In this context pure and compound reference materials in microparticulate form such as U, Ln/U, Th/U and Pu/U microparticles are of special interest. For the design of these microparticulate reference materials, according to their required properties, the understanding of morphology, chemistry, and structure is essential.
To this end, the production of pure and neodymium doped uranium-oxide based microparticles utilising an aerosol-based particle production process which has been developed and established in the laboratories of Forschungszentrum Juelich since 2012 will be presented. This method allows for the generation of microparticles with monodisperse particle size distribution. SEM/EDX measurements confirmed the monodispersity of the produced microspheres as well as the incorporation of 15 mol% Nd into the compound particles. Additionally, results from In-SEM µ-Raman (CEA, DAM) and U L₃-edge XAFS measurements in U L_α fluorescence yield detection mode (INE-Beamline @ KIT synchrotron radiation source) will be discussed pointing to the formation of U₃O₈ microparticles for both compounds. The position and shape of the X-ray absorption spectrum is insignificantly affected by the incorporation of 15 mol% Nd, indicating the structural flexibility and stability of the U₃O₈ microparticles. Additionally, a minor fraction of schoepite phase was identified by Raman measurements, which is most probably caused by ageing processes during storage under laboratory conditions.
The first batch of U₃O₈ microparticles produced in Juelich was successfully applied in an international laboratory exercise NUSIMEP9 organised by JRC-Geel and finally certified regarding the isotopic composition and the U amount per particle.
These results demonstrate that the process established in Juelich allows for a very flexible and reliable preparation of reference materials for particle analysis methods applied in safeguards.

Advanced materials have been made using a variety of elements, however, uranium has not seen much impact. The common fear of working with “nuclear material” has prevented its chemical use, even though it has been known since 1911 to be an efficient catalyst in the widely applied Haber-Bosch process. A possible strategy to synthesize advanced materials based on uranium is the chemical vapor deposition (CVD). To date, only a few suitable uranium precursors have been reported.
Here we report new volatile uranium(IV) complexes based on β-ketoenamine ligands, which have been structurally characterized and which can be synthesized via a simple salt metathesis. Functionalization of the N atom of the chelating ligand changes the coordination motif, and subsequent NMR studies offer a deeper insight into the coordination chemistry. Sterically demanding ligands (L) selectively lead to heteroleptic complexes of the form UCl₂L₂ while an additional degree of freedom in the substituent allows for the isolation of homoleptic complex of the form UL₄ with significantly increased volatility. These are suitable candidates for CVD towards uranium oxide thin films. Uranium oxides with band-gap energies in a range of 1.8 – 2.7 eV coupled with their good electrical and catalytic properties mostly driven by facile valence dynamics of the uranium cations are promising electrode materials in photoelectrochemical water splitting reactions.
We have been able to synthesize a variety of such ligands and present four novel uranium complexes which we are currently investigating with respect to the formation of novel uranium based thin film materials. Furthermore, using β-heteroarylthioketamine ligands leads to similar complexes that may be used as precursors for uranium sulfide thin films.

Radioactive waste exists in the form of gas, liquid and solid. Among them, the disposal method of collecting radioactive waste, melted and solidifying it and burying it in the ground is common. However, in the current method, after the radioactive waste is collected, it is mixed with glass as it is to be melted and solidified, or it is mixed with cement and solidified, so that the volume of the waste becomes huge. Therefore, in this study, we investigated a process in which a chemically adsorbed monolayer having an imidazole group at the end was formed on a glass fiber, and only radioactive substances were adsorbed from a solution in which waste was dissolved, and then melted and solidified at once. As a result, we succeeded in adsorbing radioactive substances with high efficiency in solution. By using this method, if the melting point of the glass fiber is set low, the melted and solidifying process will be extremely simple, and it will be possible to manage the radioactive substances in the necessary minimum volume.

Copper is a candidate for use as an overpack material in deep underground nuclear waste disposal. Copper, however, is susceptible to corrosion following closure of the repository and migration of the corrosion products through the buffer material may affect the migration of redox-sensitive radionuclides. Electromigration experiments were performed whereby a copper coupon, which was in contact with compacted bentonite, served as the working electrode and was held at a constant potential of +300 mV vs. Ag/AgCl electrode for up to 48 h. The chemical form of copper that migrated into the bentonite specimens were estimated by using Cu K-edge X-ray absorption near-edge structure (XANES) at Saga-Light Source, Japan. The results revealed that the oxidation state of Cu in the bentonite would be Cu(II) as shown the correlation between the amount of Cu migrated into bentonite and the corrosion current. On the other hand Cu(I) also observed on the copper coupon not in bentonite. Copper could corrode as Cu(II) then react with Cu metal and precipitate as Cu₂O on the copper metal.

In the composition space of interest for Hanford low-activity nuclear waste glass, the waste loading of some formulations is limited by the incorporation of sulfate into the glass (i.e., its solubility). Numerous studies have been undertaken to understand the mechanism of sulfate incorporation and its affect by various other additives. In particular, the model by Vienna et al (2014, J. Amer. Ceram. Soc. 97, 3135) identified elements which tend to raise or lower sulfate solubility in borosilicate glass. In the current experimental study, we focus on representative additives that purportedly decrease (chromium) or increase (vanadium) the sulfate solubility. A baseline aluminoborosilicate composition containing Ca and Na was melted with target composition of either i) 1 wt% SO₃ and up to 2 wt% Cr₂O₃ or ii) 2 wt% SO₃ and up to 5 wt% percent V₂O₅. Thermal and physical properties were measured and structural analysis by Raman spectroscopy was performed to investigate changes in the glass structure with transition metal addition. Electron probe microanalysis was used to measure elemental compositions in melted glass for comparison with batched compositions. Increasing Cr₂O₃ concentration caused saturation of Cr concentration within the glass and formation of crystalline eskolaite (Cr₂O₃); glass transition temperature change was only dependent on the Cr which remained in glass, while measured density changes depended on both Cr in glass and eskolaite. No apparent change in sulfate solubility was observed, though not all the targeted sulfate was incorporated in the glasses containing Cr. On the other hand, all the targeted V was incorporated into the glass, and as the concentration of V₂O₅ increased, the glass transition temperature decreased. Again, not all of the batched sulfur was incorporated into the glass, but V₂O₅ showed a weak tendency to increase the observed sulfur content in the glass. Some incorporation mechanisms to explain these trends are proposed.

Uranium dioxide prepared from depleted uranium is commonly used as an analogue to study used nuclear fuels. While actual used nuclear fuel is highly complex and contains a large number of fission products, synthetic analogues are often simplified by single species doping. Previous studies involving trivalent RE doped UO₂ show reduced dissolution rates compared to undoped UO₂. When U^IV is replaced with a RE^III cation, defect clusters containing U^V resulting from local charge compensation form which can serve to limit ability to form U₃O₈. Uranyl ions (UO₂)²⁺, which are allowed to form as a result of U^VI in U₃O₈, have a solubility much higher than stoichiometric UO₂. To predict the long-term stability of used nuclear fuel in oxidizing conditions it is necessary to understand the relationships between fuel chemistry, microstructure and environment.

Here, a series of UO₂ pellets doped with Ce, Nd, Yb, Zr and Th at 1 at. % and 5 at. % were prepared via uranyl nitrate + RE-nitrate co-precipitation to ensure homogenous distribution of dopant ions before pelletizing and sintering at 1700° C in H-Ar reducing atmosphere. Defect signatures in the resulting near-stoichiometric (U,RE)O_2+x ceramic pellets were analyzed using Raman spectroscopy at multiple excitation wavelengths pre and post oxidation treatment. Through collaboration with the University of Sheffield, we have also obtained Raman data on our samples after irradiation with 1 MeV Kr ions at room temperature, fluence 5×10¹⁵ ions/cm² at the Dalton Cumbrian Facility. Deconvolution of Raman spectra reveals distinct signatures corresponding to defects such as interstitial or vacant oxygen sites or reduced lattice symmetry. The primary Raman peak observed in UO₂ at 445 cm^-1 corresponds to the symmetric U-O stretching bond, and indicates the presence defects in the UO₂ structure by losing intensity and broadening while shifting towards higher wavenumber. Thermogravimetric analysis up to 900° C in air show a delayed transition from UO₂ to U₃O₈ in the doped samples, the effect being greatest in Yb-doped UO₂. Using Raman mapping with in-situ heating we can track the evolution of defect bands during oxidation and correlate spectral features to the orientation, size and distribution of grains and grain boundaries measured by electron backscatter diffraction (EBSD).

Self-radiation damage from the decay of radionuclides can affect microstructural evolution, phase stability, and thermodynamic properties in nuclear waste forms. The principal sources of radiation are beta-decay of the fission products and alpha-decay of the actinide elements. In general, beta-decay of the short-lived fission products leads to ionizing radiation that is the primary source of radiation and heat generation in high-level waste forms during the first 600 years of storage. Ionization can cause covalent and ionic bond rupture, valence changes, localized electronic excitations, and significant changes in ionic mobility; however, there is no evidence of significant long-term effects of fission product decay on the physical properties of nuclear waste forms. Because of the long half-lives of the actinides and their daughter products, alpha-decay is dominant over very long timescales. Alpha decay produces energetic alpha particles (4.5 to 5.5 MeV) and recoil nuclei (70 to 100 keV) that result in both radiation damage and the accumulation of helium. The evolution of radiation damage due to alpha-decay has been studied at ambient temperature using short-lived actinides, such as ²³⁸Pu and ²⁴⁴Cm, in many candidate glass and ceramic waste forms, and the early stages of evolution occur during the thermal period of the waste form. Ion beam irradiations of candidate ceramics are more frequently employed due to decreased costs, shorter irradiation timescales and the non-radioactive nature of the irradiated materials. Ion irradiations are performed over a wider range of irradiation conditions to implant helium, study radiation damage kinetics, understand the separate and combined effects of alpha particles and alpha recoils, and benchmark against alpha-decay damage due to short-lived actinides. At ion energies (keV to MeV) typically used to study alpha-decay damage in nuclear waste forms, the electronic and nuclear energy losses are both important, and the spatial coupling of electronic and nuclear energy loss can lead to reduced or enhanced damage production along the ion trajectory. While alpha particles are much less damaging than alpha recoils, the electronic energy loss from the alpha particles can also cause athermal annealing of alpha-recoil damage that can impact the evolution of radiation damage. Helium implantation studies have shown that helium bubbles form above a threshold helium concentration in candidate waste forms. Swift heavy ions (several hundred MeV to several GeV) have also been used to study radiation damage in nuclear waste forms; although the damage production mechanism and energy densities are substantially different from beta decay or alpha decay, large volumes are damaged.

In this study, we aimed to investigate the radiation effects in doped-fluorite structure ceramics. CeO₂ was chosen as the host matrix because of its fluorite crystallographic (Fm-3m) structure, which consists of a simple cubic oxygen sub-lattice with the cation ions occupying alternate cube centres, similar to UO₂, PuO_2, and ThO₂. The objective was first to quantify the defect chemistry/configurations when CeO₂ is doped with trivalent yttrium at different concentrations, and then to study the atomistic changes when the solid solutions were irradiated with 3 MeV gold ion under two different fluence, corresponding to a damage level of 6 dpa and 33 dpa values respectively.
Initial X-ray diffraction measurements of the prepared samples showed a contraction of the unit cell with Y-content attributed to a certain long-range ordering of the anion vacancies with randomly distributed Y₂O₃ micro-domains of ‘C-type’ structure which grows coherently within the fluorite matrix. This observation corroborates with Extended X-ray Absorption Fine Structure (EXAFs) analysis which indicated a change in the coordination number when the dopant is introduced in the CeO₂. On the other hand, Raman spectroscopy was used to quantify the presence of oxygen vacancy-type defects formed to compensate for the charge due to the substitution of Ce⁴⁺ with a trivalent dopant and the presence of reduced Ce³⁺.
In regards to the irradiated samples, the excellent radiation stability of Ce-based compounds was reported and it was observed that Au-incident ions tend to reduce Ce⁴⁺ to Ce³⁺. In addition, Positron annihilation spectroscopy measurements showed that the lower irradiation fluence increases considerably the concentration of lattice damages close to the disorder saturation level, where all the implanted positrons are trapped and a further increase in the irradiation dose initiates a dynamic damage recovery. Relevant defect mechanisms for the two considered irradiation fluence studies were identified and will be detailed in this presentation.

The stockpile of plutonium separated from used nuclear fuel amounts to many 100s of tonnes. The majority of this plutonium is stored as PuO₂ powder packaged in multi-can containers. The conditions during the packaging of reprocessed PuO₂ for long term storage are closely controlled to limit water uptake as radiolytic decomposition of any adsorbed water will lead to the formation of both a potentially flammable atmosphere containing molecular hydrogen as well as reactive oxygen species which may be incorporated into the oxide phase. Consequently, the safety case for the long-term storage of PuO₂ requires a complete understanding of the fundamental physical, chemical and materials degradation processes occurring at the water-oxide interface inside the storage canisters. This study investigates the effect of ionizing radiation on PuO₂ and various surrogate oxides under atmospheres potentially encountered in storage canisters.
To elucidate the effects of adsorbed water, samples of baked oxide powder were equilibrated in humidity chambers to attain a wide range of masses of water adsorbed onto the powder surface. In the case of PuO₂, once the surface water on the oxide had equilibrated with the humid atmosphere, the samples were sealed into a glass vessel held at constant humidity with either an argon or a nitrogen over-gas and left for up to 3 months. Periodic sampling of the headspace was undertaken and the atmosphere above the oxide analyzed using gas chromatography. In the case of the surrogate oxides, once the oxides had adsorbed the maximum quantity of water, the samples were flame sealed under argon into a glass tube and Co-60 gamma irradiated. Prior to and after gamma irradiation, the oxide powders were characterized using a variety of different microscopy and spectroscopy methods. Following gamma irradiation, the headspace was analyzed using gas chromatography. For PuO₂, the rate at which H₂ was produced increased with increasing water loading. This result contrasts with the experimental data in this study and from similar experiments in the literature utilizing UO₂, CeO₂ and ZrO₂ which showed a higher rate of production for lower water coverages. There was no evidence of a steady state hydrogen concentration being reached in the PuO₂ experiments, an outcome opposite of that observed in storage canisters and in experiments performed by other groups in which the PuO₂ was not held under constant humidity. Radiolytic production of molecular oxygen was not observed. Comparison of the properties of the surrogate oxide (ZrO₂) powder prior to and post irradiation did not reveal any differences in stoichiometry, surface functionalization or the formation of oxygen-centered free radicals, so the location of the sibling oxidant to the evolved molecular hydrogen currently remains unanswered.
Parallel experimental and modeling studies considering the chemistry of PuO₂ and surrogate oxides in H₂:O₂:Ar atmospheres are ongoing. To perform the experiments, a bespoke manifold was designed and commissioned to mix various ratios of H₂-O₂ and Ar gas. Gas chromatography was used to analyse post-irradiated samples. Experiments were performed in 10 ml volume 316 stainless steel canisters (with no degradable polymeric components). Experimental data will be presented showing the extent of recombination of H₂-O₂ mixtures in Ar under ⁶⁰Co γ rays and accelerated He ion irradiation and the influence an oxide surface on this recombination: radiolytic reaction of H₂ with O₂ to form H₂O is enhanced over what is expected from gas phase processes suggesting significant interfacial chemistry.
Acknowledgement: This work was supported by the Dalton Cumbrian Facility Project, a joint initiative of the University of Manchester and the Nuclear Decommissioning Authority, by the UK Engineering and Physical Sciences Research Council and by the US Department of Energy, Office of Nuclear Energy.

Complex microstructural changes occur in nuclear fuels due to the extreme environments of intense irradiation and high temperature during in-reactor operations. Several theoretical and experimental works are currently in progress to determine the effect of crystalline defects which has a major influence on the behaviour of fission products, mainly in regards to rare gases, that may impact the life-span of the rod-type uranium fuel element. For instance, the nucleation of insoluble gas bubbles is thought to act as trapping sites for migrating vacancy-type defects, and further increase in the damage levels induces significant microstructural modifications. In this study, we aimed to investigate the specific role of microstructure on the radiation stability of nuclear fuels.
For this purpose, UO₂ pellets with different microstructures were initially prepared through the wet chemistry route and sintered at different temperatures. Scanning electron microscopy and X-ray diffraction characterisations were performed to quantify the grain size/boundaries areas and crystallinity of the pristine samples. Raman spectroscopy also proved useful and allowed to investigate the effect of sintering temperature on the concentration of intrinsic lattice defects. The samples were then implanted with 1 MeV Krypton-ions at room temperature to induce irradiation damages in the matrix of the different grain-size UO₂ pellets.
Post-irradiation measurements were carried to highlight the excellent radiation stability of fluorite-type nuclear fuels. However, it was observed that the UO₂ microstructure has an important effect on the point defects produced by the displacement cascades induced by ionic implantation. In this presentation, an attempt is made to detail the mechanisms occurring near the defect-boundary interfaces in relation to radiation damage to highlight the precise role of grain boundaries modifying the defects behaviour and their subsequent evolution in fluorite structure relevant to nuclear fuels.

The safe disposal and long-term isolation of spent mixed oxide nuclear fuel poses significant challenges. Uranium and plutonium oxides enriched with rare earth and lanthanide fission products and under constant irradiation with alpha particles and neutrons must be stored in geologic formations for thousands of years. A detailed understanding of the microstructure and chemistry is required to optimize processing designed to recover fissionable products from used fuel and to improve the safety and volume of radioactive waste. Cerium oxide (ceria) has been identified as a suitable non-radioactive substitute to study the microstructure and its evolution under long term irradiation.
A number of techniques have been applied to uranium and cerium oxides to investigate basic properties and point defects. These include neutron scattering, X-ray diffraction and Raman spectroscopy of as produced, irradiated and annealed materials. These are sensitive to structural properties (XRD, neutrons) and oxygen vacancies (Raman). Positron annihilation spectroscopies (PAS) are sensitive to neutral and negatively charged vacancy-like point defects and impurity-vacancy complexes. Positrons can provide depth resolved information about uranium or cerium vacancies to complement Raman data. PAS has been applied previously to UO_2+x to investigate the properties of nuclear fuels.[1,2,3] Virtually no PAS work exists on CeO₂, doped or undoped. Currently, PAS are being used to investigate rare earth and lanthanide doped cerium oxide.
The work on uranium oxide by the group of MF Barthe[1,2] demonstrated the presence of significant concentrations of uranium vacancies or Schottky defects as well as the presence of negatively charged interstitial oxygen atoms in oxygen rich. They also investigated the effect of 1 MeV helium implantation and the interaction of the implanted helium with the formed vacancies during annealing. Xe bubbles and their thermal evolution could also be observed.[4]
In this presentation, the basics of positron annihilation spectroscopy will be presented. PAS work on undoped and doped cerium oxide will be shown and compared to results from uranium oxides. To simulate spent nuclear fuels, the samples were doped at different concentration with rare earth elements or yttrium, zirconium. Select samples were irradiated with heavy ions at different doses. Differences were observed depending on what simulated fission products are present. Preliminary results and literature searches show some correlation with Raman data.
References: [1] Barthe MF et al. EPJ Web of Conferences 115, 03004 (2016), and references therein; [2] Barthe MF et al. Phys. Stat. Solidi C4 3627-3632 (2007); [3] Wiktor J Physical Review B 90, 184101(2014); [4] Djourelov N. et al Journal of Nuclear Materials 432 287-293 (2013)
Acknowledgements: MW, JM, and SK thank the Institute for Materials Research and U.S. Department of Energy Nuclear Energy University Program – Used Nuclear Fuel Disposition program, award # DE-NE0008689. JM thanks the US-UK Fulbright Commission, and JM and SK thank the S.Y. Chung Centenary Fellowship for support during the research period. RM and CC thank Engineering and Physical Sciences Research Council, under the UK-US Nuclear Energy University Programme (EP/R006075/1), an Early Career Research Fellowship awarded to CC (EP/N017870/1), and the MIDAS Facility, at the University of Sheffield, established with support from the UK Department of Energy and Climate Change.

Layered uranium oxyhydroxide (LUOH) phases are important minerals which occur as weathering products of uraninite deposits and also importantly, are encountered during the paragenetic sequence of secondary phases that form during an accident scenario involving spent nuclear fuel (SNF) in the near-field (1). Understanding the structural-chemistry of these phases is of importance as via interaction mechanisms, they can potentially retard the transport of harmful fission daughters from SNF towards groundwater and the biosphere. Particularly it has been previously shown that the secondary LUOH phases, such as metaschoepite, uranophane and studite, can act to impede and retard the transport of fission daughters and SNF products such as NpO₂⁺, Cs⁺ and Sr⁺ (2,3). Chemically, the origin of these favourable interactions is predominantly due to favourable intercalation chemistry between cationic SNF species with that of anionic lattices of encountered phases. However in the case of anionic fission daughters, such as ¹³¹I and ⁷⁹Se, these present considerable concern during release from SNF as the overwhelming majority of near-field SNF secondary phases and minerals possess anionic lattices, including LUOH phases, which are typically not expected to react and intercalate well with. Although some evidence is present for partial intercalation of anionic iodine into layered mineral structures through ionic exchange mechanisms in model system studies (4-6) the current consensus is the retention of anionic SNF release products in the near field is not proven and best considered in the far-field of SNF repository performance and accident scenario assessment. Layered double hydroxide (LDH) phases are considered prime candidate materials and minerals for the retention of these anionic phases and received considerable attention (7), however in the context of an accident scenario with SNF their occurrence is not certain. We have recently undertaken a comprehensive systematic investigation of actinide solid state systems relevant to SNF and nuclear waste management. In the course of our investigations we have examined the chemical interactions that may occur between the anionic fission daughter species, I and Se in different chemical forms with some potential LUOH phases associated with SNF. Using a combination of batch synthesis experiments and in situ exchange experiments, we provide evidence for the inclusion and retention of these anionic species in LUOH phases. Importantly, using single crystal X-ray diffraction we show the ability for anionic species to be incorporated into certain LUOH phases is a consequence of their respective topological arrangement of uranyl and hydroxyl groups. Furthermore, by considering the topology and structural chemistry of LUOH phases, it can allow for a predictive basis to be established towards which the relative interaction affinity between certain anionic fission daughter species and specific LUOH phases can be understood. The systematic experimental studies used in this investigation will be discussed in detail w.r.t potential interactions that occur between anionic fission daughters and LUOH derived secondary phases that occur during an accident scenario involving SNF.

References:
1 P. C. Burns, R. C. Ewing, and A. Navrotsky, Science 335, 1184 (2012).
2 J. I. Friese, M. Douglas, E. C. Buck, S. B. Clark, and B. D. Hanson, MRS Proceedings 824, CC9.2, Cc9.2 (2004).
3 A. Espriu-Gascon, J. Giménez, I. Casas, and J. de Pablo, Journal of Hazardous Materials 353, 431 (2018).
4 S. Wu, S. Wang, A. Simonetti, F. Chen, and T. E. Albrecht-Schmitt, Radiochim. Acta 99, 573 (2011).
5 S. Wu, F. Chen, M. Kang, Y. Yang, and S. Dou, Radiochim. Acta 97, 459 (2009).
6 S. J. Wu, F. R. Chen, A. Simonetti, and T. E. Albrecht-Schmitt, Environ. Sci. Technol. 44, 3192 (2010).
7 N. Daniels et al., Applied Clay Science 176, 1 (2019).

Stuffed pyrochlore ceramics, with the general stoichiometry Ln₂TiO₅, have received significant interest for application within the nuclear industry due to their high radiation tolerance, good thermal stability, and structural flexibility. These ceramics have been proposed as a potential wasteform for actinide wastes, for incorporation into inert matrix fuels (IMF), and as inclusion in oxide dispersion strengthened (ODS) steels. These materials can be viewed as derivatives of pyrochlore structured Ln₂Ti₂O₇ phases and are often more conveniently written in the form Ln₂(Ln_0.67Ti_1.33)O_6.67, where the pyrochlore B-site now contains mixed occupancy. These materials can adopt a variety of different crystal structures depending on the size of the lanthanide cation, and the temperature and pressure used during synthesis. The three reported crystal structures are orthorhombic (Pnma), hexagonal (P6₃/mmc) and cubic defect-fluorite (Fm-3m). Compounds that are nominally assigned as defect-fluorite by X-ray diffraction have been probed by neutron diffraction (ND) and transmission electron microscopy (TEM) revealing features consistent with the presence of short range ordered regions with the (Fd-3m) cubic pyrochlore structure.

In this contribution we present a study probing the local structure of a series of Ln₂TiO₅ (Ln = La, Gd, Dy, Yb, Y) compounds. They were synthesised via a solid state synthesis route under different cooling regimes to prepare samples of identical chemistry but different short range ordering. Fabrication of high quality sintered specimens of Ln₂TiO₅ phases is a particular challenge due to their refractory nature. Therefore, a systematic study of sintering parameters was undertaken to optimise the fabrication of sintered ceramic specimen. Characterisation of these materials was performed using techniques including XRD, SEM/EDX and XAS. Initial XAS results show that some of these materials do have a cubic pyrochlore-like local structure but show a deviation from the ideal coordination environment of cubic pyrochlore.

Oxides with generic stoichiometry A₇O₁₂ occupy, in a generic phase diagram, an intermediate place between fluorite and bixbyite structure types. Their structure is derived from the ideal fluorite structure, and it is characterized by a more or less pronounced ordering of the O vacancies within the average fluorite sublattice. We believe that ternary oxides with this generic formula display some structural flexibility relevant to fission product stabilization in nuclear fuels and to the understanding of actinide waste forms as well. We would like to discuss the structural characteristics of the chemical bonds in these ternary systems consisting of trivalent and hexavalent cations studying the electron density using the tools of the quantum theory of atoms in molecules (Bader's AIM theory and ELF). Eventually, in these compounds with generic formula A^III₆B^VIO₁₂, the A cation can be either an actinide or a rare earth element, whereas the B cation can be either a transition metal (W, Mo) or hexavalent uranium. We believe that studying the characteristics of the polyhedron of the sevenfold coordinated A ion is particularly interesting: it is often described in the literature as a mono-capped trigonal prism, with bond lengths exhibiting an extremely large dispersion, ranging for instance from 2.19 to 2.70 Å in Y₆WO₁₂. We would discuss the implications of this large dispersion of distances on the chemical bond characteristics, compare DFT models with experiments on pristine and irradiated specimens, and discuss the competitive metastability of the oxygen deficient cubic fluorite phase that can be observed in some of these systems, as for instance Y₆MoO₁₂.

Dr Eric (Lou) Vance was a trailblazer in developing novel treatment solutions for intractable nuclear wastes, who spent 32 years at the Australian Nuclear Science & Technology Organisation (ANSTO) in Sydney. Largely known for his contributions to Synroc technology, Lou’s experience spans multiple continents including time at Pennsylvania State University with McCarthy and Roy as well as time at the Atomic Energy of Canada Limited where he studied sphene glass-ceramics. In 1987 he was recruited to ANSTO to progress Synroc technology. Under Lou’s leadership, the original Synroc titanate full ceramic formulations invented by Ringwood evolved into a waste treatment technology platform. This platform can be applied to produce glass, glass-ceramic and ceramic wasteforms via hot-isostatic pressing.
Lou’s seminal contributions to the science of nuclear waste management are evident in his ~400 scientific articles in international journals/refereed conference proceedings and his co-authorship of 3 patents relating to Synroc. An entertaining and highly popular speaker, Lou was committed to attending two to three international conferences a year to showcase the Synroc technology and his innovative nuclear wasteform research. Lou was a steadfast member of the Scientific Basis for Nuclear Waste Management symposium for several decades and a highly enthusiastic mentor. In this paper, we celebrate the research career of the late Dr Vance and his substantial and essential contributions to the field of nuclear wasteform development and to the development of Synroc technology.

The capture and immobilization of long-lived radioactive iodine I-129 must be a priority for future nuclear fuel cycles involving reprocessing. This review summarizes work on waste forms for radioiodine, emphasizing the contributions of Dr. Lou Vance as well as recent work in other groups. As early as the 1980s, Dr. Vance recognized the need for iodine waste forms and was a leader in innovative waste form solutions for iodine and other volatile halides. His studies included immobilization in silicate minerals including zeolites and sodalite, encapsulation in metallic tin, partitioning in synroc mixed-phase ceramics, and consolidation of copper iodide. Here, we recognize his particular contributions to the understanding of iodosodalite, and note some recent work by others following in his footsteps.

Synroc-C, a multiphase ceramic wasteform based on stable and leach resistant titanate minerals, was invented by Australian scientists in the late 1970s. The formulation was directed toward the immobilization of PUREX wastes from the reprocessing of nuclear fuels and was originally to be produced via hot-uniaxial pressing (HUPing). Although glass was chosen as the preferred technology during the “Atlanta shoot-out” in the US in 1981, this was ultimately because Synroc lacked the technical maturity of vitrification. However, given the enhanced performance properties of Synroc-C as a wasteform, significant interest has been sustained in this Synroc formulation. The vast majority of work on Synroc-C has been undertaken following consolidation by HUP processing. There is very little reported on hot-isostatically pressed (HIPed) Synroc-C, where HIP canisters are filled with the calcined wasteform material (waste + additives), evacuated and sealed. The HIP canister is then loaded into the HIP chamber and processed through the desired time-temperature-pressure profile to achieve a dense and consolidated wasteform. The appropriate redox conditions within the HIP canister are vital to the formation of the correct phase composition and ultimately attaining a stable and leach resistant wasteform.
In this study, we investigate the effect of Ti-metal addition on the phase formation, microstructural characteristics and durability of HIPed Synroc-C. Dense Synroc-C samples containing 20 wt.% simulated high level waste and with varying Ti-metal additions (0%, 2%, 4% and 8 wt.%) were prepared by HIPing in stainless steel HIP canisters. An investigation of microstructure, mineral phases and elemental distribution in the samples was carried out using quantitative X-ray diffraction and electron microscopic analyses. Results from durability tests are also discussed.

The UK holds the world’s largest inventory of separated plutonium under civil safeguards, projected to reach 140 tons at the end of reprocessing options. A significant fraction of this material requires immobilisation for disposal since it is unsuitable for reuse in MOX fuel, which is HM Government’s preferred policy for disposition. Additionally, should it not prove possible to implement the policy for plutonium reuse in MOX fuel an alternative immobilisation and disposal option will be required for the greater inventory. This presentation will review the current status of wasteform engineering and process development for manufacture of zirconolite ceramic and glass-ceramic wasteforms for immobilisation of waste and inventory plutonium. We will demonstrate the development of Ce and U surrogate formulations, validated by targeted plutonium studies, with regard to phase assemblage and Pu / surrogate incorporation mechanism. The development of a novel containment system to enable hot isostatic pressing studies utilising U and Th surrogates will be presented, together with operational experience. The application of new, and potentially disruptive, wasteform manufacture technology will also be discussed.

Cs₃Bi₂I₉, a defect perovskite derivative, is a potential host phase to immobilize iodine and cesium with high waste loadings. In this work, two strategies are explored to form Cs₃Bi₂I₉-silica composites and core-shell structures in order to improve chemical durability of waste form materials meanwhile maintaining high waste loadings. Cs₃Bi₂I₉ loadings as high as 70 wt.% were incorporated into a silica matrix to form silica-ceramic composites, and 20 wt.% Cs₃Bi₂I₉ was encapsulated into silica to form a core–shell structure at low temperatures and high pressure. Semi-dynamic leaching experiments indicated Cs and I were incongruently released from waste form matrices depends on the bonding environment and surface passivating effect. A BiOI alteration layer formed during the surface reorganization, acting as a passivation layer to reduce the release of radionuclides. The long-term iodine release rate was low (30 mg m^-2 day^-1) for 70 wt.% Cs₃Bi₂I₉–silica composites leached in deionized water at 90 °C, which can be further reduced to 5×10^-3 mg m^-2 day^-1 for the 20 wt.% core-shell structure. This work highlights a robust way to immobilize the highly mobile radioactive nuclides and surface passivating mechanism through elemental transport across the liquid-solid interface for Cs₃Bi₂I₉.

Radioactive waste streams containing ⁹⁰Sr, from nuclear power generation and environmental cleanup operations, are immobilised in the UK in blended Portland cement (PC) grouts to limit radionuclide leaching. Due to poor compatibility of certain wastes with PC, alternatives such as alkali aluminosilicate ‘geopolymer’ binders are being investigated. Here, we used multinuclear solid-state nuclear magnetic resonance (NMR) spectroscopy probing ²³Na, ²⁷Al, ²⁹Si and ³⁹K nuclei, including high-field (20.0 T) and cross polarisation experiments, along with X-ray diffraction experiments, to investigate the incorporation mechanism of Sr²⁺ and Ca²⁺ in geopolymer wasteforms formed by alkali-activation of metakaolin. The main reaction product identified in gels cured at both 20°C and 80°C is a fully polymerised Al-rich alkali aluminosilicate hydrate ((N,K)-A-S-H) gel comprising Al and Si in tetrahedral coordination, with Si in Q⁴(4Al) and Q⁴(3Al) sites, and Na⁺ and K⁺ balancing the negative charge resulting from Al³⁺ in tetrahedral coordination. Faujasite-Na and partially Sr-substituted zeolite Na-A form within the gels cured at 80°C. Incorporation of Sr²⁺ or Ca²⁺ displaces some Na⁺ and K⁺ from the charge-balancing sites, with a slight decrease in the Si/Al ratio of the (N,K)-A-S-H gel. Ca²⁺ and Sr²⁺ induce essentially the same structural changes in the gels. It was concluded that the disordered geopolymers can readily accommodate Sr²⁺ and Ca²⁺ into their aluminosilicate framework structure via substitution of the divalent alkaline earth cations for monovalent alkali cations in charge balancing sites. This indicates that geopolymer gels are excellent candidates for immobilisation of radioactive waste containing ⁹⁰Sr and other alkaline earth radionuclides.

The UK’s nuclear sites contain stored wastes in ponds and silos as well as large volumes of contaminated concrete and soil. During storage decommissioning and deconstruction operations, there are significant risks of radionuclide transport via groundwater and airborne particulate releases. Colloidal silica grout is a nano-particulate material that has the potential to reduce these risks very significantly. In order to utilize such nanoparticulate material as a barrier, we require a detailed understanding of the grout properties. These include its injectability, mechanical behaiour, hydrauclic properties, durability and its impact on the geochemistry of radionuclides in the environment (e.g. fission products such as Sr and Cs). Over the past 5 years, experiments at Strathclyde have progressivley investigated colloidal silica grout properties. Colloidal silica-grouted sand shows an increased stiffness and an enhanced peak friction angle, while still having a very low hydraulic conductivity (~ 10-10 m/s), typical of intact clay. Even under drained conditions, the grout provides mechanical improvement to the soil and retains a very low hydraulic conductivity. The impact on radionuclide geochemistry has been investigated through combining leaching experiments with XAS analyses. These show that colloidal silica induces several competing effects on the mobility of Sr and Cs. First, cations within the colloidal silica gel compete with Sr and Cs for surface complexation sites. Second, an increased number of surface complexation sites is provided by the silica nanoparticles and finally, the elevated pH within the colloidal silica increases the surface complexation to clay minerals and the silica nanoparticles. Finally, experiments to investigate injectability and monitorability show that colloidal silica grout can be successfully injected in silty-sands, its gelling time can be accurately predicted in a full range of environmental conditions and the location of the subsurface barrier created can be accurately identified using Electrical Resistivity Tomography. Potential applications of colloidal silica on nuclear sites include inhibiting radionuclide migration from recently decommissioned structures during care and maintenance (e.g. silos and ponds from which wastes have already been retrieved); the development of primary or secondary barriers around existing radionuclide (or other) contaminated areas; and stabilisation of underground structures. Research work is continuing to improve the mechanical and chemical properties of colloidal silica grouts, such that they can be tailored to suit for indivdual nuclear site-based applications.

Machine learning (ML) is a valuable tool that can be used, in combination with domain expertise, to provide insight into nuclear waste vitrification processes. This talk summarizes our preliminary efforts in using ML methods to address important problems facing the nuclear waste vitrification mission at the Hanford site in Washington State. Three case studies using ML modeling will be described: nepheline crystallization, refractory corrosion, and sulfate solubility. In Hanford high-level waste, nepheline crystallization can occur in melt compositions containing high Na and Al contents when glass cools slowly in the canister, partitioning radioactive components to a less-durable residual glass fraction. In low-activity waste compositions, which are generally higher in alkali and volatile components like sulfur and halides, different critical issues present. Firstly, the low solubility of sulfate in borosilicate glass has led to many studies to predict glass composition dependence of this property, such that excess sulfur is not volatilized or partitioned to an alkali molten salt phase during processing. Second, these glass compositions tend to be harsh on refractories, leading to concerns over corrosion and general degradation of melter ceramic components and premature failure. For these three cases, existing datasets were used along with ML techniques to investigate the datasets and create predictive models for the output properties of interest. For all these studies, data science was used to identify underlying trends and correlations which were then applied to create supervised learning models. As a result, it has become evident that ML is a powerful tool that can be utilized to provide insight into complex issues in nuclear waste management, here specifically focused on immobilization by vitrification.

Vitrification is a key process to turn high-level nuclear waste into durable glasses. However, the virtually infinite compositional envelope accessible to glass wasteforms renders traditional design methods based on empirical knowledge and trial-and-error poorly efficient. In addition, the complex, disordered atomic structure of glasses makes it challenging to develop some mechanistic models relating their composition to their properties and performance. Here, I will discuss how machine learning offers a promising alternative route to predict the properties of glass wasteforms and guide/accelerate the design of novel glasses with enhanced performance for nuclear waste immobilization applications.

The development of stable actinide and other radionuclide hierarchical waste-form materials offer potential advantages of greater efficiency/loading, lower leach rates, ease of processing for so-called legacy nuclear wastes. Current waste-form discovery relies heavily on experiments that may or may not be successful in producing a viable product. Computational crystal discovery has aided in the search for new materials for different applications through use of high-throughput algorithms. However, these approaches have not been adapted to actinide and lanthanide compounds due to additional corrections required to properly handle f-orbital-containing elements. This presentation will consider the challenges and possible solutions to creating high-throughput computational methods to aide in the experimental discovery of novel hierarchical waste forms and offer tangible examples.

Pd and Ru are the main platinoid fission products formed during burnup of nuclear fuels. During fuel reprocessing, these element partly dissolve in nitric acid and they exhibit very low solubility into borosilicate glasses during the vitrification process. Due to their affinity towards tellurium from the fission product solution, palladium mainly precipitates under (Pd-Te) intermetallics [1] and ruthenium is likely to form hcp solid solutions [2]. The composition and microstructures of these phases are also sensitive to oxygen potential [3]; oxydation of ruthenium can lead to needle-shaped or polyhedral (Rh,Ru)O₂ oxide phases [4]. These precipitates can affect significantly the physico-chemical properties of the glass melt during melting in hot or cold crucible.
For decades, many studies were undertaken at CEA [5] to point out the high temperature interactions of the platinum group metals (PGM) issuing from the calcinated waste. In parallel to these studies, a Calphad database is being developed to address the thermodynamics of such systems to provide an assessment of the PGMs in nuclear waste.
This study presents the assessed binary systems. Some calculations are also performed to highlight how this flexible tool makes is used to predict the stability between metallic and oxide phases as a function of temperature and oxygen potential mainly driven by the glass.
1] S. Gossé, C. Guéneau, Intermetallics 19 (2011) 621-629
[2] S. Gossé, N. Dupin, C. Guéneau, J.-C. Crivello, J.-M. Joubert, J. Nuclear Mater 474 (2016) 163-173
[3] S. Gossé, S. Bordier, C. Guéneau, E. Brackx, R. Domenger, J. Rogez, J. Nuclear Mater 500 (2018) 252-264
[4] R. Pflieger, L. Lefebvre, M. Malki, M. Allix, A. Grandjean, J. Nuclear Mater 389 (2009)
J. Nuclear Mater, Vol. 389, N°3
[5] O. Pinet, S. Mure, Journal of Non-Crystalline Solids, 355 (2009) 221-227

The hollandite phase is a potential nuclear waste form for alkaline-earth metals such as Cs and Ba by chemical incorporation into the crystalline lattice. The general hollandite formula is A_xB₈O₁₆ with the A-site habitually occupied by large mono- or divalent cations (e.g., Ba, Sr, Na, K, Rb, Cs, Pb), and the B-site by small di-, tri-, or tetravalent cations (e.g., Ti, Al, Fe, Cr, Sc) and it crystallizes in a tetragonal or monoclinic structure depending on the composition. Recently, our group has developed a thermodynamic database to describe the BaO-Cs₂O-TiO₂-Cr₂O₃-Al₂O₃-Fe₂O₃-FeO-Ga₂O₃ system based on the Calphad method, where the hollandite solid-solution was modeled using the Compound Energy Formalism (CEF). Recent findings show the impact of varying A-site Cs stoichiometry of Zn-containing hollandite compositions and related thermodynamic stability. This consequently improved the limited performance of the hollandite phase in a multi-phase composite called SYNROC. We here report on our efforts to include the Zn in our waste form thermodynamic database, modeling the phase equilibria of the complex hollandite-forming oxide waste system.

A multi-column adsorption system has been designed and installed to evaluate the independent capture and separation of xenon and krypton utilizing solid phase sorbents being developed by the DOE-NE Offgas Sigma team. The xenon and krypton are present in the offgas stream resulting from a used nuclear fuel reprocessing plant. The system is intended to provide greater operating flexibility and testing at a wide range of flow rates while maintaining desired pressure conditions. This system also allows for continuous analysis of effluent gas from each column. The system was set up by packing the first column with AgZ-PAN which has demonstrated a high selectivity for xenon over krypton at ambient temperatures. A second column was packed with HZ-PAN to capture the krypton. The AgZ-PAN column was maintained at ambient temperatures whereas the HZ-PAN column was maintained at 191K. A simulated offgas stream consisting of 1000 ppmv xenon and 150 ppmv krypton in a balance of air was introduced to the first column to capture the xenon and the resulting xenon free effluent was then routed into the second column for krypton capture. Flowrates of the feed stream were varied from 50-2000 sccm. The results of the multi-column testing demonstrated successful separation of xenon from krypton and indicated that increasing superficial velocity does not significantly impact sorbent capacity or initial breakthrough of the columns. This procedure, while not optimized, may be used to reduce the final waste volume of captured radioactive gas species emitted from a reprocessing facility.

New generation sorbents are being investigated for the removal and immobilization of radioiodine from the gas streams in a used nuclear fuel reprocessing facility. Silver-functionalized silica aerogel (Ag⁰-aerogel) exhibits high selectivity and sorption capacity for iodine. Moreover, a feasible consolidation of iodine-loaded Ag⁰-aerogel to a durable SiO₂-based waste form is an attractive choice for sequestering radioiodine. Extended-duration testing of Ag⁰-aerogel in highly oxidizing reprocessing off-gas streams indicated fragmentation of granules into small particles. These particles were held up in the column without being carried-over into the off-gas system. However, production of small particles during sorption can lead to compaction of the sorbent, restricting the flow of gas through the column. Our study targeted development of mechanically and chemically stable Ag⁰-aerogel through optimized synthesis process, which included using a patent-pending technology treatment to strengthen the granules. The reinforced Ag⁰-aerogel with bulk density ~ 0.5 g/cm³ went through rigorous testing at different off-gas streams followed by attrition testing with a vibratory sieve shaker. There were no fines generated even after 60 min exposure to shaking speeds of ~1800 revolutions per minute, with ~3 mm orbital shaker movement. The mechanically/chemically robust Ag⁰-aerogel retained high iodine sorption capacity of 400.5 ± 2.2 mg/g. The presentation will focus on various strategies that have been developed to improve the mechanical/chemical properties of Ag⁰-aerogel and conclude with a summary of highlights and a prognosis for future developments.

Molten salt systems provide highly efficient heat transfer options without pressurization, making them a potentially valuable tool throughout the energy sector, including solar power and nuclear energy. Molten salt reactors are of particular interest due to the advancements they can offer to the nuclear field, such as improved safety. However, the chemical behavior of key species anticipated in molten salt reactors is not well understood. Tools are needed to monitor and characterize the chemical composition of molten salts to better understand the behavior of actinides, fission products, and corrosion products. Furthermore, these tools are needed to allow for molten salt process control and to support the maintenance of materials accountancy. Of the options available, optical spectroscopy enables powerful routes to characterize these systems in situ. Utilization of these tools allows for the fast and robust identification and quantification of many chemical species of interest to molten salt processing, reactor control, and corrosion characterization. Most importantly, optical spectroscopy can enable the analysis of oxidation state ratios and species complexation, which can significantly impact system performance and corrosion behavior. Optical spectroscopy-based monitoring can be leveraged in fundamental and applied research. For example, optical spectroscopy can be paired with electrochemistry in spectroelectrochemistry to provide a powerful route to fundamental characterization of redox behavior, speciation, and chemical behavior dependence on temperature. Optical spectroscopy can also be paired with chemometric analysis to enable real-time quantification of target analytes within salt melts. Complex salt systems containing many chemical species will exhibit complex spectral data, presenting signal interferences and band overlap, which can hinder accurate analysis. However, the application of chemometric analysis can enable accurate characterization of multicomponent streams to provide real-time, automated monitoring of a process stream or batch. Optical spectroscopy-based on-line monitoring enables fast and robust identification and quantification of many actinides, fission products, and corrosion products. Here we demonstrate the application of optical monitoring approaches to the characterization of chloride-based salt melts containing lanthanides and actinides.

The Swedish strategy for spent nuclear fuel has, since around 1980, been direct geological disposal. Research about the properties and dissolution of spent fuel has been going on since then. An overview of this research program is given here with a focus on questions posed and answered along the way, ending with remaining open questions. At the beginning, some basic studies were performed to gain information about spent fuel as wasteform. Dissolution studies were at first mainly performed in contact with air, and often with aqueous solutions that mimicked the expected groundwater chemistry. However, complexities relating to precipitation of secondary phases indicated that simplified water compositions were preferred. Knowing that the radioactivity of the fuel caused radiolytic oxidation of uranium and thus controlled the dissolution rate, there were also attempts to achieve reducing conditions via bubbling a hydrogen/argon mixture in presence of a palladium-catalyst. The first attempts resulted in variable results, but notably, a reduction in apparent leach rate was observed for some experiments [1]. Later, possibilities to perform the leaching experiments in autoclaves, with controlled hydrogen atmosphere, allowed experimental conditions closer to the expected repository conditions. These experiments showed that hydrogen, during this type of experimental set-up, could suppress oxidative dissolution caused by radiolysis [2,3]. A new phase of research was thus started where the mechanistic explanation for the observed behaviour was sought. Experimental work, both with spent nuclear fuel and with analogue materials, was coupled with modelling efforts. At the time of the SR-Site safety assessment [4], which was a part of the license application submitted in 2011, a certain understanding had been reached, based on the effect of metallic particles in fuel. Still, questions remained and studies performed during the last decade (see for example [5]) can now be summarised to illustrate the knowledge gained, and remaining open questions. This summary is utilized in the next step in the licensing process: the application to start building the spent fuel repository, which is expected to be submitted within the next couple of years. In the coming decade the SKB research program will continue, with the goal to support the development of the safety assessment report, SAR, which will be submitted to the authorities as a part of the license application connected to start of trial operations.
During all these years, international collaboration has been a cornerstone of the SKB research program. This international collaboration has been carried out via European projects or bilateral agreements, and triggered the start-up of the Spent Fuel Workshops [6]. The research program has also been presented at this MRS symposium since the early 1980’s.
[1] Forsyth R, 1983. The KBS UO2 leaching program. Summary report 1983-02-01. KBS/SKBF Technical Report 83-26.
[2] Spahiu, K., Werme, L. and Eklund, U.B., 2000. The influence of near field hydrogen on actinide solubilities and spent fuel leaching. Radiochimica Acta, 88(9-11), pp.507-512.
[3] Werme, L.O., Spahiu, K., Johnson, L.H., Oversby, V.M., King, F., Grambow, B. and Shoesmith, D.W., 2004. Spent fuel performance under repository conditions: A model for use in SR-Can. SKB-TR-04-19. Swedish Nuclear Fuel and Waste Management Co.
[4] SKB, 2011. Long-term safety for the final repository for spent nuclear fuel at Forsmark. Main report of the SR-Site project. SKB TR-11-01, Swedish Nuclear Fuel and Waste Management Co.
[5] Barreiro Fidalgo, A., Kumagai, Y. and Jonsson, M., 2018. The role of surface-bound hydroxyl radicals in the reaction between H2O2 and UO2. Journal of Coordination Chemistry, 71(11-13), pp.1799-1807.
[6] Werme LO, 1984. Proceedings of the third fuel workshop. KBS/SKBF Technical Report 83-76

The addition of dopants, including chromium and aluminium, to uranium dioxide fuel results in the formation of larger grained microstructures, capable of higher fission gas retention and reduced swelling during operation within nuclear reactors. Despite their growing use by nuclear power plant operators, the long-term dissolution behaviour of spent doped fuels, when compared to typical spent fuel, is not yet well constrained. In a geological disposal facility, the release of radionuclides is known to be controlled by the oxidation of the UO₂ matrix and the subsequent dissolution of UO₂²⁺ species, therefore any changes to the matrix from the addition of dopants could be expected to influence the dissolution kinetics and/or mechanisms.

We here present the results of an investigation of how the addition of Cr₂O₃ to UO₂ fuel influences the crystal chemistry of UO₂ and its dissolution kinetics. Initial results, from experiments performed in air and a bicarbonate solution, show that the addition of Cr₂O₃ reduces the normalised mass loss of uranium. The potential influencing factors that give rise to increased durability with Cr addition, e.g. defect concentration, Cr-oxidation state etc. are discussed.

The “back end” of the nuclear fuel cycle corresponds to the direct disposal or reprocessing of the spent nuclear fuel (SNF), and to the conditioning and disposal in an appropriate repository of the nuclear wastes. In most of these processes, the behavior of actinides and fission products is governed by dissolution reactions at solid/ liquid interface. The design of reprocessing process and of nuclear waste disposal strategy requires the determination of well constrained kinetic and thermodynamic data. The acquisition of reliable kinetic and thermodynamic data is a tough issue considering on the one hand, the complexity of SNF in terms of chemical composition and microstructure and on the other hand, the challenge posed by the manipulation of highly radioactive samples.

For several years, we developed and improved a methodology that aims to determine dissolution rates of uranium-based oxide materials and thermodynamic data of phases of interest for final disposal of spent fuel in deep-mined geological repositories. This methodology is based on the synthesis of model compounds with well defined composition and microstructure, on the measurement of macroscopic dissolution rates in controlled chemical and hydrodynamic conditions, and on the monitoring of the properties of the solid/ liquid interface at the microscopic scale. For that purpose, several surface characterization techniques were used and developed, such as 3D reconstruction of Environmental SEM images, AFM or Grazing Incidence-XRD. To illustrate the progress made beyond the state of the art by using this methodology, several examples have been selected among the studies that we reported in the literature. The selected examples deal either with the reprocessing of uranium based oxides in nitric acid conditions, or with the behavior of UO₂ in SNF under the weathering conditions of geologic disposal.

The mechanisms that control the kinetics of dissolution of U-based oxides in nitric acid solutions are illustrated through the study of samples showing different compositions and microstructures. The use of oriented UO₂ single crystals allowed to evidence the enhanced reactivity of the surface at defect sites during a first kinetic regime. The second kinetic step was attributed to a catalyzed dissolution mechanism involving species produced at the solid/ liquid interface during the first step. The key role of catalytic species in the strong increase of dissolution rate was demonstrated not only for UO₂ single crystals, but also for sintered pellets of (U, Th) mixed oxides and UO₂-doped with platinum group metals. The monitoring of solid/ liquid interface by 3D-ESEM during uncatalyzed and catalyzed regimes revealed specific dissolution sites of enhanced reactivity and specific dissolution features related to UO₂ crystallographic orientation, which traduced a structural control of the topographic evolution of the material.

Most of the geologic sites under investigation for underground repositories are located in undisturbed clay-rich rock or granite, with silica-rich groundwaters, deep enough to have reducing conditions. Understanding the interaction of used fuel with the silicate-rich groundwaters is critical to evaluate the safety of different disposal strategies. For several years, we investigated the role of silicates in the mechanisms of UO₂ weathering. For the first time, we were able to show the formation of coffinite (USiO₄) from UO₂ in the presence of solution saturated with respect to SiO₂(am) under conditions typical of deep-mined geologic repositories for SNF. Through dedicated experiments, we have constrained the conditions of formation of coffinite in an Eh-pH diagram. Finally, we showed that coffinite precipitation could lower the uranium release from the UO₂ matrix of SNF through oxidative weathering in the presence of oxygen in the geological repository.

In an effort to incorporate high levels of Mo associated with the post-operational clean-out of high active liqor storage tanks at the Sellafield reprocessing facility in the UK, a modified version of the "MW" sodium aluminoborosilicate base glass has been adopted. The addition of Ca favours the formation of CaMoO₄, while Zn infers improved viscosity. This new Ca/Zn MW base glass is currently being used to immobilise reprocessing wastes from oxide fuels and also magnox fuels, which contain a significant proportion of Mg. The divalent cations Ca²⁺, Zn²⁺ and Mg²⁺ are known to exert an influence on the durability of vitrified wastes, however the nature of their additive, or competitive effects are yet to be fully understood.

We here describe the results of a range of experiments that aim to elucidate the influence of divalent cations on the mechanims and kinetics of glass dissolution. The addition of Zn to the glass was found to enhance durability in dilute, Si-undersaturated conditions, due to an increase in glass polymerisation with increasing Zn content, as determined by ²⁹Si MAS NMR. However, under conditions of Si-saturation, the presence of Zn prevented the formation of a passivating silica gel layer, and promoted the formation of Zn-containing smectite clays, significantly enhancing the dissolution rate with increasing Zn content. Mg had a similar effect to Zn, which was additive, however glasses containing Ca generally showed lower dissolution rates, in agreement with other studies. The formation of Zn- and Mg-bearing silicates was concurrent with a significant increase in the dissolution rate, similar to Stage III "rate resumption" behaviour, which has previously only been associated with the formation of zeolites.

Safety assessment of Spent Nuclear Fuel (SNF) disposal in a Deep Geological Repository (DGR) requires evidences of radionuclides release after groundwater eventually gets into a breached container and contacts the fuel. Within the frame of EU H2020 program into the DISCO project, the impact of additives (Cr, Cr/Al and Gd) is being investigated on the long term DGR dissolution performance and compared to conventional UO₂. In particular, the effect of these dopants is studied on the long term matrix dissolution rate of the SNF, the so-called Advanced Tolerant Fuels (ATF), such as Cr₂O₃-doped UO₂, and Cr₂O₃-Al₂O₃-doped UO₂ fuels.
ATF have demonstrated to reach higher burn-up than conventional UO₂ fuels because dopants provide enhanced safety against Pellet-Clad Interaction failure (PCI), higher density, enlarged grain size, lower fission gas release and higher creep rate. Besides, the difference between those ATF and standard UO₂ in thermo-physical properties is almost negligible [1-4]. However, the dopants influence upon the matrix leaching performance at long term under DGR conditions is still largely unknown when compared to traditional UO₂ SNF.
The combinatory effects of two uncertain parameters (dopants influence and H₂) on UO₂ matrix leaching resistance are investigated. In the context of WP2 and WP4 of the DISCO project, UO₂-based pellets doped with a nominal concentration of 0.06^w/_oCr₂O₃, 0.05^w/_oCr₂O₃-0.02^w/_oAl₂O₃ and 4.5^w/_oGd₂O₃, were systematically produced by homogenizing stoichiometric proportions of UO₂ and dopants, and carefully characterized in order to understand the effects of the addition of Cr- and/or Al-oxide into the fuel matrix dissolution behaviour. Gd-doped UO₂ was also included, as a burnable poison and a trivalent rare-earth dopants (RE^III). The results are the basis for potential more complex systems which would correspond to irradiated doped fuels.
Previous studies suggest that RE^III play an important role on the UO₂ oxidation, revealing that they prevent the fuel from a complete oxidation to U₃O₈ [5], so in order to better understand the effect of trivalent RE^III, a set of dissolution experiments is carried out. The presence of reducing species, such as H₂ (produced by anaerobic iron corrosion and radiolytically), affects the reactivity of UO₂; hence, it must be considered in a long-term safety assessment. On the other hand, it is well known that HCO₃^- (one of the key complexing constituents in the groundwater) is highly likely to influence fuel dissolution by enhancing the solubility of U^VI thus preventing the secondary phase formation [6].
To experimentally evaluate the effect of additives (doped fuels) together with H₂(g) and different leachants on UO₂ matrix dissolution rate, leaching experiments were performed under different representative aqueous systems. Prepared pellet/disks were characterized by means of XRD, Raman spectroscopy and SEM-EDX. In order to understand the effect of the presence/absence of HCO₃- that could be considered as relevant to those coming from the interaction of water between natural barriers (groundwater) and EBS (Engineering Barrier System), specific synthetic groundwater was used in the dissolution tests. Static leaching experiments of doped-UO₂ pellets were run in stainless steel autoclaves with reaction vessels made of PEEK autoclaves under a N₂/4.7%H₂ atmosphere. The U and dopants concentration released into the aqueous solution was quantified by ICP-MS. pH and redox measurements were also tracked during the experiments.
[1] Thomas, G.A. et al. (2013) LWR Fuel Performance Meeting. TopFuel 201.
[2] Gentet, G. et al. (2015) Top Fuel 2015 Conference Proceedings A0198, 170 – 176.
[3] Backman, K. et al. (2009) INIS-XA--1074. International Atomic Energy Agency (IAEA).
[4] “Technical Exchange Meeting: EnCore® Accident Tolerant Fuel ADOPT Fuel Pellets”.
[5] Travis A. et al. (2020). J. Nucl. Mater. 530, 151959.
[6] Hossain, M.M.et al. (2006). J. Nucl. Mater., 358, (2-3), 202-208.

For a future deep repository for spent nuclear fuel, the rate and mechanism for dissolution of fission products and actinides from spent fuel is a key parameter for the safety assessment of the disposal concept. Although the majority of the fuel to be placed in the deep repository is expected to be intact at the time of disposal, some failed fuel rods will also be present. In a failed fuel rod, the uranium matrix may have been exposed to water in the reactor and during interim storage for several years or decades prior to disposal. This exposure can lead to oxidation of the matrix and the formation of secondary phases under conditions of low uranium solubility. Upon barrier failure and ground water exposure of the spent fuel in the deep repository, these pre-oxidized layers are expected to dissolve at a higher rate compared to unaffected UO₂ and hence increase the rate of radionuclide release. For this reason, knowledge about the characteristics of failed fuel and its behavior under deep repository conditions is required for the evaluation of the long-term safety of geological repositories. The present work aims to add to this knowledge.

Although the behavior of spent nuclear fuel under deep repository conditions has been studied extensively, studies of failed fuel are, in this context, relatively scarce, in particular studies of failed fuel that has been exposed to water during decades of wet interim storage. However, at the Studsvik laboratory in Sweden, leaching tests of spent fuel have been performed since the 1980s. This provides an opportunity to investigate the effects of long term contact with water under aerated conditions, such as those occurring in a wet interim storage.

In this study, XRD investigations have been performed on the surfaces of two fuel pellet samples which have been exposed to leaching solutions (synthetic ground water and deionized water, respectively) for over 30 years. Furthermore, the fuel surfaces have been exposed to carbonate solutions in an effort to dissolve the oxidized phases formed during the extended leaching time. The aim is to characterize the fuel matrix alteration.

The development of robust safety cases for geological disposal of spent fuel requires a solid understanding of its dissolution over very long time scale (up to a million years). While the dissolution of standard spent fuel has reached a good level of comprehension, the development of modern LWR fuels (Cr/Al-doped UO₂ or MOX fuels) raises the question whether these fuels will behave similarly to standard fuels in geological repository conditions. Within the EU-DisCo project, systematic dissolution studies have been initiated with carefully prepared and characterized, simplified UO₂-based model materials to understand the matrix dissolution of these modern LWR fuels under geological repository relevant conditions. As DisCo partner SCK CEN has performed such dissolution tests to study the effect of Cr-doping on the UO₂ matrix dissolution rate in reducing conditions_. Tests were started on four model materials : UO₂, Cr-UO₂, Pu-UO₂ and Pu/Cr-UO₂ with a density close to 94% of the theoretical density and average grain size around 10 µm for UO₂ and Pu-UO₂ , 35 µm for Cr-UO₂ and 25 µm for Pu/Cr-UO_{2 .} The Cr doped pellets contain 1600 µg Cr₂O₃ per gram of UO₂, and the alpha doping level of the Pu-doped pellets (~ 18 MBq/gUO₂) is representative for a fuel age of 10,000 years. Static dissolution experiments with these model materials have been performed in autoclaves in two reference waters (synthetic young cementitious water at pH 13.5 and bicarbonate water at pH 9) under 10 bar H₂ pressure and in presence of Pt/Pd catalyst. After a predissolution in 0.01 M NaCl solution for 3 weeks to remove the pre-oxidized UOX layer, samples were taken from the solution over a period up to 2 months. For the materials UO₂ and Cr-UO₂, the Cr-doping leads to lower uranium concentrations in solution, both in synthetic cementitious and in bicarbonate water. In the synthetic cementitious water, this effect is constant over the entire experiment, but the concentrations, increasing immediately after the replacement of the NaCl solution by the cementitious solution, remain high at ~10^-7 M for UO₂ and ~10^-8 M for Cr-UO₂. In bicarbonate water, the initial uranium concentrations (after replacement of the NaCl solution by the bicarbonate solution) are the same as for the cementitious water, but after 2 months, all concentrations are ~10^-9 M. These preliminary results suggest that the Cr doping decreases the U release in solution, but the long-term effect depends on the leaching solution. Experiments with the Pu doped materials (Pu-UO₂ and Pu/Cr-UO₂) are now carried on to study the Cr-effect in combination with alpha-radiolysis of the solution.

Programs for deep geological repositories are being developed in many countries to safely dispose of spent nuclear fuel. Therefore, understanding the behavior of spent nuclear fuel under relevant conditions is crucial for the safety assessment of these facilities. During the last decades, new types of nuclear fuel are being introduced in commercial reactors. To allow these new fuels in the spent fuel repository, it must be shown that they do not dissolve at a higher rate than standard fuel. Fuel additives such as chromium oxide (Cr₂O₃) have an impact on the UO₂ microstructure and therefore might affect properties relevant to the safety assessment.

The main goal of this investigation is to understand the effects of Cr-doping on the instant release fraction of fission products and the matrix dissolution under oxidative conditions. A sample consisting of fuel fragments and separated cladding is leached in simplified groundwater under aerated conditions at Studsvik’s Hot Cell Laboratory. The fuel used in this study was manufactured by AREVA and has been irradiated in a Swedish commercial pressurized water reactor (PWR) to a local burnup at the sampling position of ~59 MWd/kgU.

Data from the leaching experiments are presented and compared to other commercial fuels’ leaching performance: Cr-doped fuel with lower doping levels (Advanced DOped Pellet Technology, ADOPT) and standard UO₂ fuel.

The direct disposal of spent modern nuclear fuel (SNF) in a deep geological repository (DGR) requires an in-depth understanding of the long-term corrosion behavior of the SNF and the associated radionuclide source term. These are indispensable requirements for demonstrating the long-term safety of the DGR. Today’s modern Cr- and Al-doped light water reactor (LWR) fuels have already successfully demonstrated their improved in-reactor performance, but it is still not known, whether the corrosion behavior of such fuels in a DGR is similar to conventional spent LWR-fuels. Due to the chemical and structural complexity of SNF and its high beta and gamma radiation field, experiments with SNF cannot unravel all of the various concurring dissolution mechanisms entirely. Thus within the EU-DisCo project single-effect dissolution studies are carried out with systematically produced and carefully characterized UO₂-based model materials, to provide additional insights into the behavior of modern doped LWR-fuels under the post-closure conditions expected in a DGR, complementary to dedicated dissolution experiments with SNF.
Here, we present recent results on the dissolution behavior of tailor made UO₂ model materials in accelerated static batch experiments using H₂O₂ as simulant for radiolytic oxidants, present in long-term disposal scenarios for SNF in failed container conditions due to the alpha-irradiation of water. In these dissolution experiments pure UO₂ reference pellets exhibiting different densities and grain sizes, as well as Cr-doped UO₂ pellets with various Cr-doping levels, produced using different doping methods having different grain sizes, were used. In addition, Nd-doped and industrially produced Cr- and Cr/ Nd-doped UO₂ pellets were used to determine the influence of these parameters on the dissolution rates. The dissolution experiments were performed under strictly controlled conditions with respect to exclusion of oxygen, temperature control, and exclusion of light. During the dissolution experiments, solutions were systematically sampled and the H₂O₂ (UV-Vis) and the uranium (ICP-MS) concentrations were determined.
The initial dissolution rates of all materials were determined from the concentrations of dissolved U(VI) obtained within the first 24 hours of the experiments. In experiments with pure UO₂ reference pellets a clear linear dependence of the initial dissolutions rates from the density of the pellets was observed, with higher initial dissolution rates at lower densities. No dependence on grain size was found. The initial dissolution rates of pure, Cr-doped, and Nd-doped UO₂ pellets fall into three groups: i) pure UO₂ pellets with the fastest dissolution rates, independent of the oxygen potential during sintering (and therefore the grain size), ii) Cr-doped UO₂ pellets, independent of doping method, doping level, and oxygen potential during sintering (grain size), and iii) Nd-doped UO₂ pellets with the lowest dissolution rates. The following initial dissolution rates were calculated for the different materials:
pure UO₂: 1.3 +/- 0.1 * 10^-6 mol/m²*s
Cr-doped UO₂: 0.65 +/- 0.2 * 10^-6 mol/m²*s
Nd-doped UO₂: 1.63 +/- 0.7 * 10^-7 mol/m²*s
The slight difference between the dissolution rates of pure UO₂ and Cr-doped UO₂ pellets (i.e. factor 2) can partly be explained by the slightly lower density of pure UO₂ pellets compared to Cr-doped UO₂ pellets. A very strong influence of Nd, representative of fission lanthanides in SNF, on the dissolution rates in this type of experiments is clearly visible. Although the density of the Nd-doped UO₂ pellets is similar to the Cr-doped pellets (i.e. 96-98% relative density), the initial dissolution rate is more than half an order of magnitude lower than for Cr-doped pellets. Overall, the results of this study indicate that the Cr-doping of modern nuclear fuels seems to have no significant effect on the corrosion behavior of UO₂-based SNF.
.

The dissolution rates of six calcium aluminosilicate glasses were determined experimentally at pH(25 °C) = 7.5 and a temperature of 75 °C in a Single-Pass Flow-Through (SPFT) system. Glasses studied in this work can be categorized into two sets of three compositions each. In the first (metaluminous) set, the molar proportion of CaO to Al₂O₃ remained the same (1:1), while SiO₂ increased (40, 50 and 60 mol.%). The second (peralkaline) set was characterized by constant Al₂O₃ (10 mol.%), with decreasing CaO and increasing SiO₂ (40, 50 and 60 mol.%). Measurements of ²⁷Al by MAS-NMR indicated a slight increase in Al^[5] at the expense of Al^[4] with decreasing Si concentration. One to four grams of powdered glass (300 to 425 mm diameter) were placed into 10 mL volume reactors with solution input through the bottom and effluent solution output through the top. Solutions were buffered by 0.05 M TRIS and flow-rates varied between ~38 and 500 mL/d. A plot of the dissolution rate vs. flow-rate (q) to surface area (S) showed that all dissolution rates were maximized at q/S ≈ 45 × 10⁸ m/s. Dissolution rates were from 4.2, 1.5 and 0.62 × 10^-7 mol/(m² sec) for the metaluminous and 1.0, 0.65 and 0.50 × 10^-7 mol/(m² sec) in the peralkaline glasses, respectively. Interestingly, the data indicate that rates for the metaluminous trio are generally faster than those of the peralkaline set, which runs contrary to expectations. As a set of glasses, the log₁₀ rates [mol/(m² sec)] are inversely proportional to the fraction of Si in the glasses, forming a single array of values. These data indicate that rates are strongly tied to the Si content of glasses, and that other glass composition parameters— including Ca/Al (molar) ratio—exert a much weaker effect on dissolution rate in this medium.

Predicting the release rates of radionuclides from nuclear waste glasses during aqueous dissolution within a disposal facility over geological timescales requires robust models of dissolution built upon an extensive understanding of glass dissolution mechanisms. However, many of the dissolution models presented in the past decade conflict regarding the mechanisms predicted to control dissolution. Whilst nanoscale post-process analyses of altered glasses have afforded insights into altered layer structures, these results highlight the need for an additional dimension to understand the mechanisms of glass dissolution beyond the conventionally employed techniques presented in the literature.
In a previous study, lithium isotope fingerprinting techniques were applied to the leachates of a simplified Magnox waste glass leached at 90 °C. It was shown that whilst dissolution was initially congruent, lithium isotopes leached incongruently through diffusive processes across a passivating altered layer component at longer durations. However, further evidence was required to rank the dissolution models presented in the literature. Here, boron isotope fingerprinting techniques were applied to the same Magnox glass to further understand its dissolution processes, understand how lithium and boron were leached and distinguish between dissolution models. Additionally, the same isotopic techniques were applied to leachates of an international benchmark glass composition to compare the dissolution processes of the two compositions and allow for interdiffusion models to be fitted to the results. Solution renewal experiments took place to further investigate the passivating nature of the altered layers and their role as a diffusive barrier to dissolution at long durations.
Boron isotopes leached from the Magnox glass were consistent with congruent models of dissolution in the initial stages of dissolution, but at later stages were consistent with diffusive isotopic fractionation across a passivating altered layer component about an incongruent dissolution front. Whilst boron isotopes in the leachates of International Simple Glass were consistent with diffusive, incongruent dissolution models throughout dissolution, boron isotopes in its solution renewal experiment were most consistent with congruent models of dissolution. As such, these isotopic methodologies demonstrate that no single dissolution model may be correct for all glass compositions and experimental conditions; providing an improved understanding of the mechanisms governing aqueous glass dissolution and supporting the idea of a model which unifies the mechanisms presented in the literature.

The analysis of individual micrometre- and submicrometre-sized particles collected by IAEA safeguards inspectors on swipe samples during in-field verification activities requires the implementation of a sustainable quality control system. This system needs to be further advanced by developing mass spectrometric analytical measurements and of suitable reference materials. In this context pure and compound reference materials in microparticulate form such as U, Ln/U, Th/U and Pu/U microparticles are of special interest. For the design of these microparticulate reference materials, according to their required properties, the understanding of morphology, chemistry, and structure is essential.
To this end, the production of pure and neodymium doped uranium-oxide based microparticles utilising an aerosol-based particle production process which has been developed and established in the laboratories of Forschungszentrum Juelich since 2012 will be presented. This method allows for the generation of microparticles with monodisperse particle size distribution. SEM/EDX measurements confirmed the monodispersity of the produced microspheres as well as the incorporation of 15 mol% Nd into the compound particles. Additionally, results from In-SEM µ-Raman (CEA, DAM) and U L₃-edge XAFS measurements in U L_α fluorescence yield detection mode (INE-Beamline @ KIT synchrotron radiation source) will be discussed pointing to the formation of U₃O₈ microparticles for both compounds. The position and shape of the X-ray absorption spectrum is insignificantly affected by the incorporation of 15 mol% Nd, indicating the structural flexibility and stability of the U₃O₈ microparticles. Additionally, a minor fraction of schoepite phase was identified by Raman measurements, which is most probably caused by ageing processes during storage under laboratory conditions.
The first batch of U₃O₈ microparticles produced in Juelich was successfully applied in an international laboratory exercise NUSIMEP9 organised by JRC-Geel and finally certified regarding the isotopic composition and the U amount per particle.
These results demonstrate that the process established in Juelich allows for a very flexible and reliable preparation of reference materials for particle analysis methods applied in safeguards.

Advanced materials have been made using a variety of elements, however, uranium has not seen much impact. The common fear of working with “nuclear material” has prevented its chemical use, even though it has been known since 1911 to be an efficient catalyst in the widely applied Haber-Bosch process. A possible strategy to synthesize advanced materials based on uranium is the chemical vapor deposition (CVD). To date, only a few suitable uranium precursors have been reported.
Here we report new volatile uranium(IV) complexes based on β-ketoenamine ligands, which have been structurally characterized and which can be synthesized via a simple salt metathesis. Functionalization of the N atom of the chelating ligand changes the coordination motif, and subsequent NMR studies offer a deeper insight into the coordination chemistry. Sterically demanding ligands (L) selectively lead to heteroleptic complexes of the form UCl₂L₂ while an additional degree of freedom in the substituent allows for the isolation of homoleptic complex of the form UL₄ with significantly increased volatility. These are suitable candidates for CVD towards uranium oxide thin films. Uranium oxides with band-gap energies in a range of 1.8 – 2.7 eV coupled with their good electrical and catalytic properties mostly driven by facile valence dynamics of the uranium cations are promising electrode materials in photoelectrochemical water splitting reactions.
We have been able to synthesize a variety of such ligands and present four novel uranium complexes which we are currently investigating with respect to the formation of novel uranium based thin film materials. Furthermore, using β-heteroarylthioketamine ligands leads to similar complexes that may be used as precursors for uranium sulfide thin films.