Traditional wet reprocessing technologies, such as the PUREX process, face challenges including high volumes of waste liquid generation and operational complexity. In contrast, dry reprocessing technology has garnered significant attention due to its high efficiency and reduced waste production. The conversion of spent oxide fuel into metallic fuel through dry reprocessing technology is a critical step in advancing the construction of integrated fast reactors in China and realizing a closed nuclear fuel cycle. However, few studies on the electrochemical reduction of oxide spent fuel have been reported in China, with relatively weak theoretical foundations, experimental methods, and associated technologies. Considering that the primary component of oxide spent fuel is UO2, UO2 pellets are converted into U3O8 powder via oxidation-volatilization techniques during the head-end process. Therefore this work focused on U3O8 solid powder, systematically investigating its electrochemical reduction mechanism in LiCl-Li2O molten salt at 650 ℃ through cyclic voltammetry(CV) and experiments of potentiostatic electrolysis. The phase composition, microstructure, and elemental distribution of electrolytic products were characterized using X-ray diffraction(XRD), scanning electron microscopy(SEM), and energy-dispersion spectroscopy(EDS). CV tests show that during the potential scan(1.50 V to 0.00 V vs. Li+/Li), six reduction peaks(c1-c6) and five corresponding oxidation peaks(a1-a5) are observed. The c1 peak(0.00 V) corresponds to Li deposition, while the c2-c6 peaks(0.15-0.84 V) are attributed to the stepwise reduction of U3O8 to UO2 and metallic uranium(U). Potentiostatic electrolysis experiment shows at 1.95 V(vs. Li+/Li), there is exclusively generated the UO2 phase. At 1.20 V, the products contain both UO2 and Li2UO4. However, no metallic U is detected at 0.70-0.10 V. These results indicate that the electrochemical reduction mechanism of U3O8 likely proceeds in three steps as following: (1) a redox reaction between U3O8 and Li+ to form UO2 and Li2UO4; (2) further electrochemical reduction of Li2UO4 to UO2; and (3) stepwise electro-deoxidation of UO2 to obtain metallic uranium(U). The presence of metallic U and UO2 in the products of pulsed constant-voltage electrolysis is confirmed by XRD and SEM/EDS analyses, demonstrating that U3O8 solid powder can be directly electrochemically reduced to metallic U in the system of LiCl-Li2O molten salt. Direct oxidation measurements reveal an average reduction rate of 45.47% for U3O8 and an average current efficiency of 22.86%. This work reveals the multi-step electrochemical reduction mechanism of U3O8 in LiCl-Li2O molten salt and uncovers the transformation pathways of key intermediates. These findings provide a theoretical basis for optimizing processes parameters in the electrochemical reduction of spent oxide fuels(e.g. potential control and molten salt composition design), offering significant implications for the efficient production of metallic fuels and the sustainable development of nuclear energy.

The development of advanced ceramic waste forms for nuclear waste immobilization requires precise control over phase evolution and chemical durability. Sr-90 and Ce-144 were employed in this study as simulants for typical nuclear waste fission products. SrCO3 and CeO2 were selected as precursors to systematically investigate their incorporation behavior and phase evolution within the CaZrTi2O7 (x = 0-1.0) ceramic matrix, with the objective of elucidating the composition-structure-property relationships of the material. Progressive embedding of the precursors induces a complete phase transformation from zirconolite to perovskite through coupled valence and size effects. Multimodal characterization reveals that Ce3+/Ce4+ redox dynamics govern site-specific incorporation (Ce3+-> Ca2+, Ce4+-> Zr4+), while Sr2+ substitution triggers lattice expansion via Vegard's law. Rietveld refinement and Raman spectroscopy show that the transformation pathway leads to complete phase conversion at the x = 0.8 embedding level. The optimized composition achieves exceptional chemical stability, evidenced by ultralow elemental leaching rates (10(-4) g m(-2) d(-1) for Sr and 10(-7) g m(-2) d(-1) for Ce) and >99% densification. Notably, the material maintains mechanical integrity (5-8 GPa hardness) despite phase transition-induced softening. This work establishes a fundamental framework for the rational design of new-generation radioactive waste forms with targeted immobilization capabilities.

Abstract Pulse-potential electrodeposition is widely used to suppress dendrite growth during uranium electrorefining. However, the strongly coupled multi-physics nature of this process makes parameter optimization challenging. Here, a phase-field model was developed to simulate uranium dendrite growth under pulse-potential conditions, directly coupling dendrite morphology evolution with ion concentration distribution and electric potential field. The effects of duty cycle, pulse period, pulse-on potential, and pulse-off potential on dendrite morphology were systematically investigated, and optimal parameters were identified from the simulation results. Electrodeposition experiments under both constant-potential and pulse-potential conditions were performed to validate the model. The results show that the pulse-potential method markedly suppresses side-branch growth while leaving primary branches largely unaffected. Under the optimized parameters, uranium dendrites with substantially suppressed side branches were obtained experimentally. This work provides a framework for parameter optimization in pulse electrodeposition and offers actionable guidance for controlling dendrite growth during nuclear material processing.

During the electrorefining of spent nuclear fuel in molten salt for the recovery of U, Pu, and minor actinide (MA), fission products (FPs) such as rare earth and alkaline earth elements accumulate continuously, leading to the decrease of process efficiency. Thus, the purification of the spent salt and the stable immobilization of these volatile radioactive FPs chlorides still require further investigation. In this study, simulated FPs (140La, 144Ce, 90Sr) were precipitated from a LiCl-KCl molten salt system as low-volatility LaPO4, CePO4, and SrCO3, using K3PO4 and K2CO3 as precipitants. With LaPO4 as the host matrix, the precursor mixtures were sintered without pressure to produce a kind of monazite waste form with controlled stoichiometries (La1-2xSrxCexPO4). At 1350 degrees C with x <= 0.2, the crystalline component of the sintered waste pellets was found to consist exclusively of the monazite phase, while a secondary phase was notably enriched in Sr. Cerium was determined to exist in both +3 and + 4 valence states, indicating that a portion of Ce3+ substitutes for La3+, whereas another portion is oxidized to Ce4+ and forms charge-compensating pairs with Sr2+, collectively occupying two adjacent La3+ sites. MCC and PCT static leaching tests confirmed low normalized leaching rates (NRi) for La, Sr, and Ce in monazite. Furthermore, the addition of borosilicate glass (1, 10, 25, 50 wt%) to monazite reduced the leaching rates by 1-2 orders of magnitude, reaching NRi values of 4.83 & times; 10-6, 1.31 & times; 10-4, and 3.20 & times; 10-6 g m-2 d-1 for La, Sr, and Ce respectively with a 10 wt% glass addition. When the glass addition was increased to 50 wt%, the Vickers hardness attained 8.06 GPa. The monazite waste form in this work, with a maximum simulated FPs (140La, 144Ce, 90Sr) loading capacity of 57.6 wt%, exhibits high waste loading alongside excellent chemical durability and mechanical stability, demonstrating its potential as an advanced method for FPs immobilization.

137Cs has become one of the key challenges in the dry reprocessing of spent fuel due to its poor separability and high radioactivity. The synthesis of Cs2SnCl6 for the removal of CsCl from waste salts and the preparation of Cs2SnCl6-glass composites was investigated in this study. The optimal conditions for the hydrothermal synthesis of Cs2SnCl6 are a reaction at 150 degrees C for 6 h, and XRD results indicate that the product is a pure cubic phase. CsCl is removed from LiCl-KCl-CsCl waste salts by the synthesis of Cs2SnCl6. This method enables efficient and selective separation of Cs+ from waste salts with a Cs+ content of 2.5-10 wt.%, achieving a removal efficiency of over 96%. The glass composites have a maximum embedding rate of 30 wt.%. Cs+ and Cl- are uniformly distributed in the glass matrix, and the glass matrix exhibits good compatibility with Cs2SnCl6. The PCT test showed that the normalized leaching rates of Cs+ and Cl- on the 28th day were as low as 1.6 & times;10- 4 and 3.4 & times;10- 4 g & centerdot;m- 2 & centerdot;d- 1, respectively, demonstrating that the Cs2SnCl6-glass composite exhibits excellent leaching resistance.

This study investigated the electrolytic reduction of kilogram-scale U₃O₈ pellets in molten LiCl and focused on optimizing the process for nuclear fuel cycle applications. Critical developments included resolving carbon impurity issues in LiCl, implementing a sequential kerosene-cyclohexane washing process at 150 °C for lithium anode pretreatment, and establishing optimal pellet fabrication parameters. By adjusting the sintering atmosphere, the pellet density (4.5–8.5 g/cm3) and chemical composition (U₃O₈ or U₄O₉) could be controlled. The electrolytic reduction used a stainless-steel cathode basket loaded with sintered pellets and a lithium metal counter electrode. The experiment was conducted in an inert-atmosphere glove box for 181h. The results showed reduction extents of 99.09

With the rapid growth of scientific literature and engineering data， knowledge extraction and structured modeling from multimodal data have become key challenges for intelligent engineering transformation.In recent years， large language model （LLM） have demonstrated strong capabilities in cross-modal understanding and knowledge modeling， providing new technical pathways for multimodal knowledge base construction.However， in domains characterized by high specialization， dense terminology， and strict safety requirements， general-purpose LLM still face multiple challenges such as hallucinated outputs， knowledge distortion and deployment constraints.To enable effective application of LLM in spent nuclear fuel reprocessing， we first review the development of LLM and their core techniques， and analyzes their roles in professional knowledge base construction.We examine the representative studies of LLM in three domains， say， symbolic regression， chemical materials and healthcare， to summarize a technical framework and key technical components of LLM-driven multimodal knowledge base construction， including structured knowledge extraction， knowledge representation and indexing， retrieval-augmented generation and reliability control.Then we focus on the knowledge base construction of spent nuclear fuel reprocessing， which is subject to strict safety constraints.An engineering validation is conducted on the structured extraction and knowledge ingestion stages.A case study on equation–reaction structured extraction is presented to demonstrate the transferability and practical feasibility of the obtained technical framework.Finally， current challenges in this domain are discussed， and future directions are outlined in terms of safe and controllable deployment， few-shot adaptation， multimodal fusion， and expert knowledge integration， providing technical references for enabling LLM-driven intelligence in the nuclear industry.

Because of the complex mechanisms in pulsed disk and doughnut columns (PDDCs), traditional empirical functions often fail to make accurate predictions in new datasets, such as different experimental conditions or different PDDC structures, indicating a lack of generalizability. In this work, some machine learning techniques such as random forest regression (RFR), least absolute shrinkage and selection operator, support vector regression (SVR), and artificial neural network are developed to predict dispersed phase holdup based on experimental data collected from numerous studies. Two training methods were used: One is to randomly divide the collected data into groups for training and testing, and the other is to separate the data of one study for testing and training in data from other studies. These methods were used to compare and analyze the accuracy, generalizability, and stability of these models, using the mean relative error (MRE) as the performance evaluation criterion. SVR has an MRE of 15.0% in the test set and 11.0% in the entire dataset, outperforming other alternative models in both efficiency and ability to mitigate overfitting. Furthermore, the relative importance of each parameter in influencing holdup was analyzed by RFR.

This study delineates the intrinsic composition of naturally occurring radioactive material (NORM) waste and affirms the viability of the carbothermal reduction method for the transformation of Ba(226Ra)SO4 into Ba(226Ra)S. The waste was solubilized using ethylenediaminetetraacetic acid, and its constituents were determined employing X-ray diffraction and inductively coupled plasma-atomic emission spectrometry, identifying barium sulfate (BaSO4) as the predominant component at a weight percentage of 67.13

The local structure and physical properties of LiCl-Li2O-Li molten salt, the reaction medium for lithium thermal and electrolytic reduction, are very important for the study of spent fuel pyroprocessing process. In this work, the machine-learned deep potential (MLDP) was trained using dataset based on first-principle molecular dynamics (FPMD) and was used to predict the changes in the physical properties of molten LiCl with the addition of different concentrations of Li2O and Li between 923 K and 1323 K. Deep potential molecular dynamics (DPMD) calculations were performed for properties including shear viscosity, electrical conductivity, thermal conductivity, and specific heat capacity. It was revealed that the addition of Li significantly reduces the diffusion activation energies (Ea) of Li+ and Cl-in the molten salt. By comparison with the experimental data of pure LiCl, it can be concluded that the MLDP can describe the inter-atomic interactions of molten salt correctly, overcome the problem of missing potential parameters in the classical inter-atomic empirical potentials. Finally, DPMD allows to simulate large systems with comparable accuracy of FPMD, thus provide theoretical guidance for the optimization of the pyroprocessing technology.

A cost-effective anode material for uranium oxide electrolytic reduction in lithium chloride is still in deficiency. In this work, the application of liquid lithium-bismuth alloy anode was investigated. In the LiCl electrolyte at 923 K, UO2 was reduced electrochemically in cathode, while Li-Bi alloy served as counter electrode. Partial reduction of UO2 was verified by X-ray powder diffraction when the cathode potential was intentionally controlled above the lithium reduction potential. In contrast, when the precipitation of lithium metal was intentionally controlled, the reduction of UO2 was significantly improved. The charge transfer coefficient of UO2/U reaction was also calculated. Regeneration of Li-Bi alloy in LiCl-Li2O through electrolysis was proposed. Carbon, gold, and platinum had been investigated as anode materials. According to the potential variation curve, lithium was not effectively reduced into bismuth as in pure LiCl when oxygen ion was present in the molten salt. These three materials failed to demonstrate advantage in the regeneration of Li-Bi alloy.

Artificial graphite, a key material in numerous industries, is typically produced through processes involving coaltar pitch and coke filler particles. The baking process, integral to its production, influences the final product's structural integrity. Despite extensive research on coal-tar pitch pyrolysis, there is a gap in understanding the carbonization behavior within the green body, as the coal-tar pitch is carbonized in confined space and shows location-dependent pyrolysis behavior. This study investigates the location-dependent pyrolysis behavior of green bodies by introducing a novel parameter, baking degree (eta). Experimental and simulation studies reveal delayed pyrolysis behavior within the central part of the green body, confirmed through examination of porosity development. Such delayed pyrolysis is not caused by temperature gradient, but should be attributed to the kinetic factors associated with carbonization behavior in the confined space. Introducing the eta helps to grasp more precisely the pyrolysis process of the binder in the green body. By optimizing process parameters to minimize the difference in eta between the center and surface of the green bodies during the baking process, the delayed carbonization of binders within the green bodies can be effectively reduced, thereby enhancing the performance of the artificial graphite.

Organophosphorus ligands such as TBP, TiAP, and DMHMP have exhibited excellent performance in recovering actinides from spent fuel. In this work, the molecular geometries and properties of TBP, TiAP and DMHMP were investigated using density functional theory calculations. Furthermore, the extraction mechanism of ligands for actinides (Np(VI), Pu(IV)) was further elucidated by simulating the microstructures and extraction reactions of metal-ligand complexes. The results demonstrate that the complexation ability of the three ligands on actinide cations (NpO22+ and Pu4+) follows the order of DMHMP > TiAP > TBP. The electrostatic potential (ESP) analysis indicates that the nucleophilic ability of DMHMP is stronger than that of the other two ligands. The frontier molecular orbital analysis of the three ligands represents that DMHMP has the highest HOMO energy, suggesting that it has the strongest electron-donating capability and is more likely to bond with metal ions. The values of Wiberg bond indices (WBI) suggest that the MO bonds in DMHMP complexes have more covalency. According to the QTAIM analysis, the interactions between actinide cations and the ligands are predominantly ionic in nature. The molecular orbital analysis of the complexes shows that the M(NO3)(n)center dot 2DMHMP (MNpO22+ and Pu4+) complexes are more stable, which is supported by thermodynamic energy analysis. This work has clarified the complexing properties of actinide cations with three ligands, shedding light on the extraction mechanisms of organophosphorus ligands for actinide cations. It is anticipated to lay the theoretical foundation for the efficient recovery of critical actinide elements in spent fuel reprocessing, which will also provide innovative approaches for the design and development of related separation processes.

Tuning the extraction performance of phenanthroline-derived ligands for Am 3+ by combining different functional groups.

The purpose of this study is to improve the concentration of U(VI) in carbonate solution reasonably, which to improve the application potential of the alkaline reprocessing processes. The dissolution behavior of U3O8 in carbonate peroxide solutions was investigated under different conditions, including pH, carbonate concentration, and solid-liquid ratio. The results showed that the dissolution rate of U3O8 increased with the increase of pH from 8 to 11 in the mixed carbonate solution containing 0.5 mol/L H2O2. The role of carbonate ions in the dissolution of U3O8 was further elucidated by observing the dissolution of UO4·4H2O in carbonate solutions. Furthermore, the concentration of U(VI) in 3 mol/L Na2CO3 solution was successfully increased to 350 g/L under ultrasonic-assisted conditions at 60 °C and a solid-liquid ratio at 1/2 g/mL. Meanwhile, it is suggested that increasing the concentration of carbonate ions can improve the stability of the dissolved solution containing uranyl peroxycarbonate complex.

A comprehension of the electrochemical characteristics of uranium ions is crucial for the efficient separation of uranium from other actinides and fragment elements during electrorefining in pyroprocessing. In this study, cyclic voltammetry and square wave voltammetry were used to study LiCl-KCl-UCl3 molten salt system at 773 K. The BET adsorption model was employed in this study to provide a more comprehensive description of the kinetic process of the reduction of U3+ to uranium metal on the working electrode surface. By comparing the simulated results with experimental data, it is observed that there is a strong agreement between them, indicating an accurate depiction of the redox process of uranium ion. Additionally, this model facilitates the analysis of uranium metal deposition on the electrode surface, thereby further substantiating the rationality and efficacy of the model.

In the nuclear spent fuel reprocessing, zirconium dibutyl phosphate (Zr-DBP) is deposited on the surface of stainless steel equipment, making spent fuel reprocessing difficult. Therefore, in this study, hydrazine carbonate ((N2H5)(2)CO3) was used to dissolve Zr-DBP. The effects of solid-liquid ratio, different mass fractions of (N2H5)(2)CO3 solutions, and (N2H5)(2)CO3 solutions containing different concentrations of hydrogen peroxide (H2O2) on the dissolution of Zr-DBP were investigated. The corrosion rates of different dissolution solutions on stainless steel were also investigated. When the mass fraction of (N2H5)(2)CO3 in solution was increased from 1 % to 20 %, the concentration of Zr(IV) at dissolution equilibrium increased from 1.05 g/L to 1.61 g/L, respectively. When the dissolution temperature was kept in the range of 25 degrees C to 65 degrees C, the increasing temperature promoted the dissolution of Zr-DBP. The presence of H2O2 induced the production of the white precipitates zirconium hydroxide (Zr(OH)(4)). In addition, the corrosion rate of stainless steel gradually increased from 0.00153 mm/a to 0.0138 mm/a when the concentration of H2O2 in the (N2H5)(2)CO3 solution was increased from 0 M to 1.5 M. The dissolution ability of Zr-DBP in different carbonate solutions was (N2H5)(2)CO3 >= ammonium carbonate ((NH4)(2)CO3) > sodium carbonate (Na2CO3) >= potassium carbonate (K2CO3).

To enhance the extraction of remaining fissile nuclides from spent nuclear fuel through pyrochemical reprocessing, the operational lifetime of such fuels can be extended, ultimately leading to increased cost efficiency and a reduction in the amount of radioactively contaminated waste generated by fast-neutron reactors. Understanding the electrochemical behavior of fissile nuclides in LiCl–KCl is pivotal to the success of molten salt electrorefining pyrochemical reprocessing. In this pursuit, a comprehensive study of Pu(III) salt was conducted to comprehend its electrochemical characteristics within molten chloride salt mixtures. To achieve this, PuCl 3 was meticulously prepared by reacting PuO 2 with HCl in a LiCl–KCl mixture. Subsequently, we investigated the reduction mechanism, the diffusion coefficient of Pu(III) ( D Pu(III) ), and the apparent standard reduction potential of Pu(III)/Pu(0) ( E 0* Pu(III)/Pu(0) ) in situ, using a Mo working cathode. Our findings revealed that Pu(III) undergoes a single-step reduction to Pu(0), involving the exchange of three electrons. Furthermore, the rate of diffusion governs the reduction of Pu(III) at the Mo cathode. The relationship between the diffusion coefficient and temperature was described by ln D = − 5.51 to 4244.2/ T , with an activation energy of 35.28 kJ/mol. Additionally, we examined the temperature-dependent variations of E 0* Pu(III)/Pu(0) and the Gibbs free energy of formation for PuCl 3 (Δ G PuCl3 ). These dependencies were found to be E 0* Pu(III)/Pu(0) = − 3.194 + 6.4 × 10 −4 T and Δ G PuCl3 = − 924.5 + 0.185 T , respectively.

Metal-oxide nanoparticles (NPs) are of particular interest in various industrial applications. However, current liquid-phase synthetic routes for metal-oxide NPs generally involve complicated processes. Herein, we demon-strate a simple direct route to metal-oxide NPs through the pyrolysis of their metal-organic frameworks (MOFs). As illustrated by the synthesis of ZrO2 and CeO2 NPs, we employed Zr(IV)-and Ce(IV)-based CAU-24 MOFs as precursors, and pyrolysis was performed in both air and molten NaCl environments. Pure oxide NPs with sizes of less than 200 nm were obtained via both routes. Compared with direct pyrolysis in air, the use of molten salt as a liquid reaction medium leads to much smaller ZrO2 NPs but facilitates the growth of CeO2 NPs. The results can be explained by the presence of different phases in the nanocrystal growth process.

In spent fuel reprocessing, UO2(DBP)2 (U-DBP) can be deposited in stainless steel equipment. U-DBP must be removed by dissolution and the process must not cause corrosion to stainless steel. This study was conducted to find the best scheme for dissolution. U-DBP was manufactured by the titrimetric sedimentation method. The effects of different factors on the dissolution of U-DBP were investigated. For example, solid-liquid ratio, hydrazine carbonate solutions with different mass components, mixed solutions containing different concentrations of H2O2, and different carbonates. The results indicated that U-DBP does not have a regular crystal morphology. With the increase of the solid-liquid ratio and the mass fraction of hydrazine carbonate, the concentration of U(VI) at the dissolution equilibrium increases gradually. The addition of H2O2 has a great promotion effect on the dissolution. However, when the concentration of H2O2 is greater than 0.5 M, the dissolution solution may have an erosive effect on the stainless steel. (NH4)2CO3 can increase the dissolution capacity of dissolved u-DBP, but it may also accelerate the corrosion of stainless steel.