[en] A test was carried out in Atalante using a Bis-Triazinyl-Pyridine extractant (BTP) to separate actinides from lanthanides in a synthetic aqueous feed solution simulating a DIAMEX process stripping solution. The first CEA studies on BTP in 1998 showed that this extractant efficiently separates actinides(III) from highly concentrated nitric medium. Preliminary studies allowed the design of a first flowsheet with the following sections: extraction section, scrubbing to back-extract the lanthanides following actinides in the solvent back-extraction section to recover actinides with organic scrub to remove the palladium extracted by BTP. Preliminary experiments also showed that the extraction kinetics were slow, which can be ignored in flowsheet design. Accurate flowsheet design entails the determination of the number of stages in each section and the flowrates. This was done using the PAREX code. Specific models for distribution mechanisms and extraction kinetics were added to it. A model of distribution ratios for actinides, lanthanides and palladium was constructed. The extraction reaction can be written: M³⁺+i NO₃^- + nBTP - M(NO₃)i,BTP_n. The model is aimed to reproduce the influence of nitrate ion concentration and free nPr-BTP concentration on the distribution ratios. Figure I and 2 show the comparisons between calculated and experimental distribution ratios for different experimental conditions (nPr-BTP concentration, nitric acid concentration, feed). An experiment to study extraction kinetics was performed in a test tube using the same blade propeller as those of the mixer-settlers. Since diffusional resistance is negligible in the mixer, transfer resistance was attributed to interfacial reactions. The experimental results were reproduced using fitted interfacial transfer coefficients (Table 2). The performance specification for the test was actinide recovery over 99.9%. The actinide feed must also contain less than 5% lanthanides (by weight) and less than 1% palladium. The solvent is nPr-BTP: 0.04 mol.L^-1 in TPH/octanol 70/30%_vol. This extractant concentration was chosen as the highest concentration to enhance extraction compatible with a reasonable margin towards the solubility limit of extractant in the diluent. The flowsheet in Figure 5 was designed using the PAREX code. In the first battery (extraction-scrub), 5 stages are devoted to actinide extraction, and lanthanides are then back-extracted on 3 stages. The second battery is devoted to back-extraction of actinides (5 stages) and also features a palladium scrub section (3 stages). The calculation anticipates 99.5% actinide recovery rate and 0.02% lanthanide and 1 % palladium contamination. Transient calculations were also carried out to estimate and optimize the initial conditions to shorten the time to equilibrium. The test was carried out in 16 mixer-settlers with a volume of 25 ml each. The radioactive feed solution was a synthetic solution containing every element, at nominal concentration, which could be found in a DIAMEX back extraction solution. About 10 hours was needed to reach equilibrium with steady outflow concentrations of every element. This first hot test shows that the average targets were achieved, with over 99.8% actinide extraction yield. 98% of americium and 90% of curium were also recovered. Less than 2.3% (by weight) of lanthanides was back-extracted along with actinides, despite low aqueous scrub efficiency. Furthermore, only 1 % of palladium was measured in the actinide stripping solution. Ruthenium was slightly extracted, indicating its easy separation by this process. Part of the iron formed insoluble complexes with BTP solvent, damaging to stainless steel. Figures 6 to 8 show the experimental and calculated profiles. Experimental profiles of metallic species were correctly reproduced in the extraction sections. A comparison of experimental and calculated profiles tends to suggest that the kinetics is slower for back-extraction than for extraction. Division of the interfacial transfer coefficients by 15 helps correctly reproduce the experimental profiles of actinides in the back-extraction and scrubbing sections. This comparison offers interesting feedback for guiding future developments. The hot test yielded results demonstrating the real performance of the process, to validate the computation and support the actinide separation concept. (authors)

[en] The aim of nuclear fuel reprocessing is to separate reusable elements, uranium and plutonium from the other elements, fission products and minor actinides. PUREX process uses liquid-liquid extraction as separation method. Numerical codes for modelling the extraction operations of PUREX process use a semi-empirical model to represent the partition of species. To improve the precision and precision and predictive nature of the models, we looked for a theoretical tool which permits to quantify medium effects, especially in the organic phase, for which few models are available. The Sergeivskii-Dannus model permits to quantify deviations from ideality in organic phase equilibrated with aqueous phase, but with parameters depending on extractant/diluent ratio. We decided to investigate UNIQUAC and UNIFAC models which permit to estimate activity coefficients in non-electrolytic phases taking account of the mutual interactions of molecules and their morphology. UNIFAC is based on UNIQUAC but molecules are considered as structural groups assemblies. Before applying these model to extraction systems, we investigate their abilities to describe simple systems, binary and ternary systems. UNIQUAC has been applied to TBP/diluent mixtures and permits to estimate activity coefficients for diluents whose interactions with TPB are very different in nature and strength. Group contribution (UNIFAC) applied to TBP/alkane mixtures permits to represent the effect of lengthening alkane chain but not the effect of branching. UNIQUAC fails to describe the TBP/diluent/water/non-extractable-salt systems in case of strong TBP diluent interactions. In order to obtain a correct description of these systems, we used the Chem-UNIFAC model, where the INIFAC equation is supplemented with chemical equilibria allowing explicitly for complexes formation and where group contribution is used to describes complexes. We have with Chem-UNIFAC a model available which can take the effect of the diluent into account and permits to represent partition isotherms with the same parameters whichever is TBP/diluent ratio. (author)

[en] Chem-UNIFAC model is applied to TBP/dodecane/water/UO₂(NO₃)₂ and TBP/dodecane/water/HNO₃ systems, following our previous studies on TBP/diluent (alkane, HCCl₃, CCl₄) and TBP/diluent/water/salting out agent (same diluents) systems. New Chem-UNIFAC parameters for TBP/water pair are calculated and an unique set of parameters is proposed to describe the TBP/dodecane/water/UO₂(NO₃)₂ system, for a wide range of TBP/diluent volume proportions (10%, 30%, 50%, 100%) and a wide range of uranyl nitrate aqueous concentration. Nitric acid and water extraction in TBP/dodecane/water/HNO₃ is also calculated and an improvement is found in the description of the corresponding isotherms. (authors)

[en] A test was carried out in the ATALANTE facility at Marcoule using a Bis-Triazinyl-Pyridine extractant to separate actinides(III) from lanthanides(III) out of a synthetic aqueous feed solution simulating a DIAMEX process stripping solution. This extractant molecule was proposed by Z. Kolarik(2) in the framework of the European contract 'NEWPART'(1), coordinated by the CEA. The first CEA studies on BTP in 1998, showed that this extractant was very efficient to separate actinides(III) from a highly concentrated nitric medium. Preliminary studies allowed to design a first flowsheet with the following sections: an extraction section, a scrubbing section to back-extract the part of lanthanides(III) that follows the actinides(III) in the solvent, a back-extraction section to recover actinides(III) with an organic scrubbing to remove the palladium(II) well extracted by BTP. Preliminary experiments also showed that extraction kinetics were slow, which must be taken into account in the flowsheet design. The PARER code, with specific models for metal nitrate distribution phenomena and extraction kinetics, was used to elaborate precisely the SANEX/BTP process flowsheet (stage number of each section and flow rates). The test was carried out using 2 batteries of 8 mixers-settlers with a volume of 25 mL for each stage. The radioactive feed solution was a synthetic solution containing every element, at nominal concentration, which could be found in a DIAMEX back extraction solution. This first test showed that the mean objectives have been reached. More than 99.8% of actinides(III) were extracted and 98% of americium(III) and 90% of curium(III) were recovered. Less than 2.3% (in mass) of lanthanides(III) were back extracted along with actinides, in spite of a bad aqueous scrubbing efficiency. Furthermore, only 1% (in mass) of palladium(II) was measured in the actinides(III) stripping solution. Ruthenium was slightly extracted, so the separation of that element is easy by this process. A part of iron has made insoluble complexes with BTP solvent which damages stainless steel. Experimental concentration profiles of metallic species were correctly reproduced by calculation for extraction sections. The comparison of experimental and calculated concentration profiles led to believe that kinetics are slower for back-extraction than for extraction. This comparison gives an interesting feedback for orientation of future developments. (authors)

[en] In this chapter of the DCC 1999 scientific report, the following theoretical studies are detailed: electronic structure of lanthanides or actinides complexes, forecasting of the stoichiometry of europium nitrate complexes, actinides aqueous solutions analytical and thermodynamical chemistry, actinides complexes structural determination. It also provides experimental studies: actinides and lanthanides separation, radioactive wastes processing and conditioning, plasma torch vitrification process, simulation of the wastes packages characterization, wastes storage with concrete behaviour and biodegradation. (A.L.B.)