- Email: [email protected]
Solubility of uranium oxide in molten salt electrolysis bath of LiF–BaF2 with LaF3 additive
Journal of Nuclear Materials 470 (2016) 90e96
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Solubility of uranium oxide in molten salt electrolysis bath of LiFeBaF2 with LaF3 additive Nagaraj Alangi a, b, *, Jaya Mukherjee a, L.M. Gantayet b a b
Laser & Plasma Technology Division, Bhabha Atomic Research Centre, Mumbai, India Homi Bhabha National Institute, Mumbai, India
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 April 2015 Received in revised form 28 November 2015 Accepted 2 December 2015 Available online 12 December 2015
The solubility of UO2 in the molten mixtures of equimolar LiFeBaF2(1:1) with LaF3 as additive was studied in the range of 1423 Ke1523 K. The molten fluoride salt mixture LiFeBaF2 LaF3 was equilibrated with a sintered uranium oxide pellet at 1423 K, 1473 K, 1523 K and the salt samples were collected after equilibration. Studies were conducted in the range of 10%e50% by weight additions of LaF3 in the equimolar LiFeBaF2(1:1) base fluoride salt bath. Solubility of UO2 increased with rise in LaF3 concentration in the molten fluoride in the temperature range of 1423 Ke1523 K. At a given concentration of LaF3, the UO2 solubility increased monotonously with temperature. With mixed solvent, when UF4 was added as a replacement of part of LaF3 in LiFeBaF2(1:1)-10 wt% LaF3 and LiFeBaF2(1:1)-30 wt% LaF3, there was an enhancement of solubility of UO2. © 2015 Elsevier B.V. All rights reserved.
Keywords: Molten fluoride UO2 Solubility LaF3 additive
1. Introduction Electrolytic reduction of uranium oxide to uranium by molten salt electrolysis above the melting point of uranium is an attractive method of producing uranium metal for metallic nuclear fuel. This route for producing uranium metal assumes more significance in the context of growing importance of metallic nuclear fuels in the nuclear fuel cycle and the role of atomic vapour laser isotope separation (AVLIS) in it [1]. The pyro-electrochemical process is capable of accepting the feed material in all the forms of oxides and fluorides of uranium, and even partially oxidized metal. Being a single step low pressure process, this process is amenable for operation in radiochemical environment. The electrolyte is generally composed of fluorides of Li, Ba, Ca, etc. which provides the electrochemical window of operation with UF4 as the functional electrolyte [2e6]. The addition of UF4 enhances dissolution of UO2 and its decomposition to U4þ and O2 ions. The tetravalent uranium ions migrate to the cathode, and first get converted to trivalent ion and then to uranium metal at the cathode. Oxygen ion migrates to the graphite anode and gets converted to CO and leaves the cell. The electrolysis parameters such as current density, anode effect and
* Corresponding author. Laser & Plasma Technology Division, Bhabha Atomic Research Centre, Mumbai, India. E-mail address: [email protected] (N. Alangi). http://dx.doi.org/10.1016/j.jnucmat.2015.12.002 0022-3115/© 2015 Elsevier B.V. All rights reserved.
rate of addition of UO2 depend on the solubility of UO2 in the electrolyte. Electrolytic reduction of uranium oxide in the integrated nuclear fuel cycle plant would involve working with different isotopes of uranium. In this case UF4 solvent has to be made from uranium with a specific isotopic composition, which in itself is a difficult job. It will be particularly difficult to manage the various uranium isotopes generated in the thorium fuel cycle. Thus, an electrolytic reduction process with a non-uranium additive in the molten salt electrolyte will remove this constraint. The data of uranium oxide solubility with UF4 as the solvent in different supporting electrolyte mixtures is reported in the literature [7e12]. Varwig et al. [10]. have studied the solubility of UO2 in UF4 containing LiFeBaF2(1:1) electrolyte, which has been extensively used by Mallinckrodt Chemical Works for electrolysis [13]. They have reported that the UO2 solubility is a function of UF4 content in LiFeBaF2(1:1) electrolyte and it varies from 0.1 wt% of UO2 at 0% UF4 to ~2.3 wt% UO2 at 45 wt% UF4 at 1423 K. Haas et al. [12] have also studied the solubility of UO2 in UF4eLiFeBaF2 electrolyte. They found that the UO2 solubility at 1473 K is a function of UF4 content in LiFeBaF2(1:1) electrolyte, varying from 0.08 wt% of UO2 at 0% UF4 to ~4.47 wt% UO2 at 45 wt% UF4. Typically 25wt% of UF4 is used as solvent in LiFeBaF2(1:1) electrolyte for electrolysis which has solubility of ~2 wt%. LaF3 additive is a potential candidate for replacing functional electrolyte UF4. Among the rare earth fluorides LaF3 is easily
N. Alangi et al. / Journal of Nuclear Materials 470 (2016) 90e96
available and abundant. The cathodic deposition potential of lanthanum being higher than that of uranium, it provides a comfortable electrochemical window. However, there is no reported data on UO2 solubility in molten fluoride salt with LaF3 additive. The present work was to measure the solubility of uranium oxide in the LiFeBaF2(1:1) electrolyte with LaF3 additive in the temperature range of 1423 Ke1523 K, towards development of a new electrolyte for electrolytic reduction of UO2 above the melting point of uranium. The solubility of oxides in fluoride melts is predominantly due to the formation of complexes of uranium, fluorine and oxygen. It is reportedly enhanced with increasing amount of UF4 solvent [12].. Although a number of uranium oxyefluoride complexes exist, the complex prevailing at high temperature has never been proven in the molten fluoride system. Greenfield and Hyde have reported compound formation of UF4 with diluent salt, K2UF6 or 2KFUF4 in KF containing electrolytes and 4LiF.UF4, 7LiF.6UF4 in LiF rich electrolytes, which reduce the total availability of free UF4 which associates with UO2 to increase its solubility [7]. In a similar fashion, Ba2LaF7 phase has been observed at room temperature by XRD of samples from LiFeBaF2(1:1)-LaF3 mixtures. It can be hypothesized that LaF3 may exhibit complexes similar to that of UF4 in LiFeBaF2(1:1) bath, which will affect solubility of UO2.
2. Materials LiF and BaF2 of reagent grade (>99% assay) were sourced from M/s Sigma Aldrich with CAS No. 7789-24-4 and 7787-32-8 respectively. LaF3 (CAS No. 13709-38-1) with minimum purity of 99.5% was obtained from M/s Prabhat chemicals. Nuclear grade UF4 having a minimum assay of 99% was procured internally. Fresh defect free sintered UO2 pellets (~10.4 g/cc density) were obtained from Nuclear Fuel Complex, Hyderabad and were of nuclear fuel grade, which had a purity of 99.8%. The impurities in the sintered pellet were as shown in Table 1.
91
3. Experimental setup An experimental setup was designed and fabricated for this study. The setup consisted a recrystallized alumina tube of 80 mm ID and 90 mm OD 800 mm long, closed at one end was used as a muffle. Viton O-ring sealed water cooled stainless steel flanges were used on the open end of the alumina tube. The cover flange had openings for gas inlet and outlet connections for high purity argon with <0.1 ppm of oxygen. Other feed throughs on the flange were used for the pellet holder and the thermocouple. Heating of the alumina tube was done by a 6 kW resistive heating furnace controlled by a programmable PID controller with an accuracy of ±3 K. A pack of radiation shields of 78 mm diameter, supported from the flange was used to prevent heat losses and to maintain a uniform temperature zone. A high density graphite crucible of ~50 mm diameter was used to contain the salt mixture. The crucible containing the electrolyte rested on an alumina support tube of 45 mm outer dimension in the uniform temperature zone, which in turn rested on the floor of the alumina tube. The UO2 pellet holder was made out of high density graphite, which had a provision to hold one end of a UO2 pellet of ~13 mm diameter and ~19 mm height. The pellet holder had a provision to thread to a 8 mm diameter molybdenum rod. The high temperature structural members were made of either graphite or molybdenum. The furnace was housed inside a fumehood and the purge gas released into the fumehood was passed through a series of HEPA filters to separate the fine radioactive particulate mass and a water scrubber to remove fluoride vapour before exhausting to the environment. The schematic diagram of the experimental set-up is shown in Fig. 1 and the photograph in Fig. 2a. Fig. 2b shows the photographs of uranium oxide pellet and its holder. Due safety clearance from the regulatory body was obtained to carry out experiments.
4. Salt mixture preparation and purification All the salts used in the experiment were dried individually in vacuum at 1073 K for 3 h in a graphite crucible. They were retained as stock in desiccators. Around 30 g batch fluoride salt mixture was
Table 1 List of impurities in sintered UO2 pellet. Sl. No.
Element
Value, ppm
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.
Ag Al B Ca Cd Ce Co Cr Cu Dy EBC Er Eu Fe Gd Mg Mn Mo Na Ni Si Sm Zn
<1.0 <25.0 0.12 13.0 <0.1 <0.5 <2.0 <10.0 <6.0 <0.1 ~0.71 <0.1 <0.1 47.0 ~0.08 9.0 2.0 5.0 35.0 1.0 <30.0 <0.1 <1.0
EBC-Equivalent boron content. SSA-Specific surface area of UO2 powder 2.55 m2/g in case where UO2 powder was used to prepare UO2þC pellets.
Fig. 1. Schematic of the experimental setup UO2 for solubility study in fluoride melts (a) Immersion with pellet holder (b) Full immersion of the pellet in the electrolyte.
92
N. Alangi et al. / Journal of Nuclear Materials 470 (2016) 90e96
3.
4.
5.
6.
7.
Fig. 2. a Photograph of the experimental setup for UO2 solubility inside fumehood. b Uranium oxide pellet and pellet holder used in the solubility experiments.
prepared by weighing the individual salts of LiF, BaF2 and LaF3 in the desired proportions and mixing by mortar and pestle. The pretreatment involved first equilibrating the electrolyte mixture in a graphite crucible at 673 K for 24 h and then at 1473 K for another 24 h in a flowing pure argon atmosphere. The composition of the salt mixtures thus prepared and characterized consisted of equimolar LiF: BaF2(1:1) base salt with 10, 20, 30 and 50 wt% LaF3. The mixed electrolyte of LaF3 and UF4 will be produced during the course of electrolysis. Hence, in a few experiments 8.1 wt%LaF31.9 wt% UF4 and 28.1 wt%LaF3-1.9 wt% UF4 mixtures were used instead of 10 wt% LaF3 and 30 wt% LaF3, respectively.
5. Methods of determination of saturation solubility The method of determination of saturation solubility of the oxide in an electrolyte bath in this study was selected from among several known methods described below. 1. In the fluoride melts, oxide is added to the salt bath in a transparent quartz container and laser based visual techniques are used to determine the point beyond which oxide is not soluble. Since the fluoride melt attacks quartz this method was not chosen. 2. The second method involves equilibrating the salt bath with different amounts of oxide, quickly quenching and observing the undissolved oxide metallographically. Quenching the salt
mixtures was avoided as it involves potential release of radioactive powders and vapors. Potentiometric measurement of the electrode potentials with respect to a reference electrode with addition of different amounts of oxide and the point beyond which the potential remains constant may be considered as the limiting solubility. The electrode potential measurement is possible in fluoride media. However, reliable high temperature reference electrode was not available to adopt this method. The potential at which the anode effect occurs depends on the concentration of dissolved uranium oxide. In this technique, a known quantity of uranium oxide powder is dissolved and the point beyond which the anode effect potential (determined by linear sweep voltammetry) does not vary is considered the limit of solubility. Here too, the non availability of a high temperature reference electrode was a handicap. The fifth method involves, collecting a sample of the fluoride melt that has been saturated with oxide and conducting the chemical analysis to determine the extent of solubility after filtering out the suspended undissolved oxide fully. Several attempts of collecting a sample through a porous graphite filter from the surface of the fluoride bath again did not yield reliable results. In the sixth method, a solid shape of the oxide is immersed in the fluoride bath and after sufficient time has elapsed, the weight change is measured. A sintered UO2 pellet was immersed in the salt bath. However, the mechanical operation to remove the adhering salt led to chipping of the pellets, which introduced errors. Similar problem of fluoride salt adhesion to solid rods also have been reported by Greenfield and Hyde [7]. A variant of the last mentioned is to measure the concentration of uranium in the electrolyte instead of measuring weight change in the UO2 pellet. The major mass of the solidified electrolyte was easily separable from the pellet. The minor quantity of salt adhering to UO2 pellet, if any, was removed separately and was not mixed with the bath sample. The uranium concentration in the sample was measured and the total uranium in the fluoride bath was estimated. The procedure for determination of uranium in fluoride salt mixture was also developed and described in the experimental method section. It was ensured that the pellet did not have any sharp edges, which could chip off during the experiment. Around fifteen trials were conducted to qualify this method. In all the trials the integrity of the sintered pellet remained intact. This method which was free of the problems of the other methods, was adopted
It is to be noted that uranium concentration can be unambiguously related to the dissolution of UO2. However, oxygen concentration will not be exactly equivalent to the oxygen from UO2 dissolution, as fluoride salts invariably pickup oxygen from the moisture and oxygen in the environment [12]. During this work, authors have observed presence of substantial quantities of complex oxy-fluorides while preconditioning. 6. Experimental methods Two experimental procedures were used in the solubility studies. In the first, a pre-weighed sintered UO2 pellet was held in a graphite pellet holder, and after preconditioning of the fluoride bath the pellet was immersed in the molten salt and removed before cooling of the salt (as shown in Fig. 1a). In the second procedure the pellet was immersed fully without a holder in the molten salt and was allowed to cool down along with the bath (Fig. 1b) after equilibration. Since the total surface area available was higher with the UO2 pellet fully immersed in the salt bath, this
N. Alangi et al. / Journal of Nuclear Materials 470 (2016) 90e96
procedure was adopted for all the experiments. All the experiments, particularly the safety related ones, were checked before commencing each experiment. The sintered UO2 pellet was weighed and placed in the high density graphite crucible. The fluoride salt mixture was added to the annular gap between the crucible and the pellet. Before commencing the flow of argon purge gas, the complete system was evacuated to 1E-2 mbar and back filled with argon at least 8e10 times to remove the trapped air. The next step involved equilibrating the bath and the sintered UO2 pellets for 48 h at the temperature of study. Initially, the time required for reaching saturation was determined by conducting total 8 experiments with immersion duration of 12, 24, 36 and 48 h. It was observed that the concentration reached saturation after 24 h of immersion of the pellet. In all the reported experiments in this study the sintered UO2 pellets were equilibrated with the salt bath for 48 h, which ensured saturation even with the experimental uncertainties. In an attempt to study the effect of increase in the surface area available for dissolution, in one of the experiments equilibrating a UO2þC pellet with the salt bath was tried. It was observed that the UO2þC pellet had disintegrated and no meaningful determination of solubility limit was possible. Hence using UO2þC pellet was not continued. The equilibration temperatures chosen were 1473 K, 1523 K and 1573 K. In the experiments the argon purge was continued throughout till the system was cooled and ready for dismantling. The oxygen concentration in the inlet gas was continuously monitored by an online oxygen gas analyzer (Systech 800 Series analyzer). After cooling to room temperature the frozen salt mixture was removed. The photograph in Fig. 3 shows the colors developed in various samples. The graphite crucible had a draft angle of 3 to remove the solidified salt mixture easily. The salt mixture was separated from the pellet and the entire sample was ground for homogenizing and collecting a sample for analysis of the total uranium concentration. All the uranium in the sample was assumed to be obtained from the dissolved UO2. Analysis of uranium in the fluoride salt posed a challenge. The standard method of preparing sample by dissolution in HNO3 (15.8 M) or HCl (11.6 M) always led to an undissolved residue. Hence, microwave dissolution technique was developed.
93
heating were used. A standard microwave oven of 1600 W in continuous mode supplied by M/s Milestone was used for heating. The acids tried for dissolution were pure HNO3(15.8 M), pure HCl (11.6 M) and combination HNO3 and HCl, and HCl with different proportions of boric acid. The microwave heating was carried out for 1 h. A summary of the dissolution studies is given in Table 2. The combination of 6 ml Conc HCl, 6 ml H3BO3 solution (0.7 M) was selected. The time was then optimized to 20 min, which ensured complete dissolution of the sample. Inductively coupled plasma mass spectrometry (ICP-MS) Model JY238 Ultrace was used for the determination of the total uranium present in the dissolved sample. All the frozen samples were also characterized by X-ray powder diffraction (XRD), to identity the phases of LiF, BaF2, UF4, UO2, LaF3 present. A Rigaku Miniflex II diffractometer with Cu Ka1 radiation (l ¼ 1.5405 Å) was used. The XRD data for indexing and cell-parameter calculations were collected in the scanning mode with a step of 0.02 in the 2q range from 20 to 80 and a scanning rate of 2.0 min1 with silicon as the internal standard. 7. Results and discussions In the high temperature molten salt based experiments several unavoidable errors occur in the measurements. Some of them are due to not so clean separation of the frozen electrolyte from the pellet, liquid phase segregation, inhomogeneous precipitation of the dissolved UO2, incomplete dissolution and error in the ICP-MS. Therefore, each experiment was carried out 3 times for obtaining reliable data. Thus, the individual solubility data presented here is the mean of at least 3 observations. The uranium oxide solubility data reproducibility is around ±5% of the value reported. This limit includes the errors enumerated above along with preparing the samples for analysis. At first the experiments were done with 10 wt % LaF3 and 50 wt% LaF3. Reporting significantly different result with ±5% reproducibility, the number of data points in the range of interest, that is 10 wt% to 30 wt% LaF3 for electrolysis was limited. It was then decided to study the linearity of the solubility between 10 wt% and 30 wt% LaF3. As the results were reassuredly linear, 20 wt% LaF3 was studied to cover the range of electrolysis. 7.1. Effect of quantity of LaF3
6.1. Development of microwave dissolution technique for preparing sample for ICPMS For complete dissolution of the solid sample we tried a noninterfering complexing agent such as boric acid (H3BO3). High pressure dissolution and high temperature operation is necessary to dissolve fluoride salt. High pressure TFM-fluropolymer based proprietary containers (M/s Milestone) of capacity 100 ml, designed for 100 bar and 573 K and compatible with microwave
Fig. 3. Colors developed due to uranium concentration in different salt baths. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Studies were conducted with 10 wt%, 20 wt%, 30 wt% and 50 wt% LaF3 additions in LiFeBaF2(1:1) base salt. The solubility of UO2 in equimolar LiFeBaF2(1:1) at 1473 K without any solvent is reported as 0.08% by weight [12]. In our experiments of measurement of solubility in equimolar LiFeBaF2(1:1) at 1473 K, the dissolved uranium was found to be below detection limit (BDL) (<20 mg/1.5 g sample) when measured by Energy Dispersive X-ray Fluorescence Spectrometer (EDXRF). The dissolved uranium in LiFeBaF2(1:1)LaF3 melts was determined by ICP-MS following the procedure outlined above. The results of solubility at different LaF3 concentrations are presented in Table 3 and graphically in Fig. 4. The solubility of UO2 was higher with increase in LaF3 in the melt at all the temperatures of the experiments, namely 1423 K, 1473 K and 1523 K. The solubility variation between 30 wt% LaF3 and 50 wt% LaF3 was within the standard deviation of the analysis. The reason for not having a large variation of UO2 solubility in melts containing 30 wt% LaF3 and 50 wt% LaF3 may be owing to the reduction in the total availability of free LaF3, which could form Ba2LaF7 compound with BaF2 of the base salt mixture. At 1523 K probably the compound is less stable thereby freeing some LaF3 which increases the solubility of UO2 at that temperature. In XRD investigations BaF2 and Ba2LaF7 phases were seen in XRD of frozen salt mixture. Probably, all the dissolved uranium was present as solid solution.
94
N. Alangi et al. / Journal of Nuclear Materials 470 (2016) 90e96
Table 2 Parametric study for microwave dissolution of fluoride samples. Salt and solvent
A
D
E
Quantity, mg
50
100
50
B 100
50
C 100
100
50
100
200
>300
LiF CaF2 LiFeCaF2 LiFeBaF2 LiFeBaF2-10 wt%UF4 LiFeBaF2-25 wt%UF4 LiFeBaF2-10wt%LaF3 LiFeBaF2-20wt%LaF3 UO2
√ √ √ X X X X X e
√ √ √ X X X X X e
√ √ √ X X X X X e
√ √ √ X X X X X e
X X X X X X X X e
X X X X X X X X e
X X X X X X X X e
e e √ √ √ √ √ √ √
e e √ √ √ √ √ √ √
e e √ √ √ √ √ √ √
e e √# √# √# √# √# √# √#
A: 7 ml Conc HCl, 1 ml Conc HNO3 solvent, for 20 min at 220 C. B: 7 ml Conc HCl, 1 ml Conc HNO3 solvent, for 30 min at 220 C. C: 5 ml Conc HCl, 5 ml Conc HNO3 solvent, for 30 min at 220 C. D: 8.7 ml Conc HCl, 1.3 ml Conc HNO3, 0.5 gm H3BO3 solvent, for 40 min at 220 C. E: 6 ml Conc HCl, 6 ml H3BO3 solution (0.7 M), for 20 min at 220 C. √ - Complete dissolution. X - Incomplete dissolution. #-with proportionate increase of solution required for dissolution of 200 mg.
Table 3 Solubility of UO2 in LiFeBaF2(1:1)-LaF3 melts. S.no
Sample No and weight % LaF3
UO2 solubility at 1423 K, weight %
UO2 solubility at 1473 K, weight %
UO2 solubility at 1523 K, weight %
1 2 3 4
10 20 30 50
0.11 e 0.2 0.19
0.415 0.482 0.57 0.56
0.76 e 0.8 0.984
wt% wt% wt% wt%
LaF3 LaF3 LaF3 LaF3
substantially when the temperature was raised from 1423 K. In the LiFeBaF2(1:1)- 10 wt% LaF3 electrolyte the solubility of UO2 was 0.11 wt% at 1423 K, which increased to 0.145 wt% at 1473 K and further increased to 0.76 wt% at 1523 K. Similar increasing trend with temperature was observed for LiFeBaF2(1:1)- 30 wt% LaF3 and LiFeBaF2(1:1)- 50 wt% LaF3 electrolytes. The trend can be clearly seen in Fig. 6. The maximum solubility of UO2 in LaF3 containing electrolytes observed at 1523 K in LiFeBaF2(1:1)- 50 wt% LaF3 electrolyte, was 0.984 wt%. 7.3. Effect of addition of UF4 to LaF3eLiFeBaF2(1:1) electrolyte (mixed electrolyte)
Fig. 4. Variation of solubility of UO2 in LaF3 solvent based molten fluoride salt.
Perhaps, for the same reason, UO2 or UF4 were also not seen in the XRD plot. As the dissolved uranium was less than 1 wt%, the uranium bearing phase could not be seen. The XRD plot of LiFeBaF2(1:1)-10 wt% LaF3 is shown in Fig. 5.
1.9 wt% of UF4 was added as a replacement of LaF3. Experiments were conducted with LiFeBaF2(1:1)- 8.1 wt%LaF3-1.9 wt% UF4 and LiFeBaF2(1:1)-28.1 wt%LaF3-1.9 wt% UF4 at 1473 K and results are shown in Table 4. With the addition of 1.9 wt% UF4 to the LaF3, the combined effect of the mixed electrolyte resulted in an increase in UO2 solubility in comparison to the UF4 free electrolyte. The total uranium was analyzed and the contribution from UF4 was accounted for to arrive at the solubility of UO2. The UO2 solubility at 1473 K in LiFeBaF2(1:1)- 8.1 wt%LaF31.9 wt% UF4 and LiFeBaF2(1:1)-28.1 wt%LaF3-1.9 wt% UF4 electrolyte was 1.04 wt% and 0.92 wt% respectively, corresponding solubilities in UF4 free electrolyte were 0.415 wt% and 0.57 wt %. The higher dissolution due to the addition of UF4 is shown in Fig. 7. This could be due to greater probability of complex formation between UO2 and UF4.
7.2. Effect of temperature
8. Conclusions
Data of UO2 solubility as a function of temperature in different electrolytes is important as the temperature of electrolysis plays an important role. As anticipated intuitively, at a given percentage of LaF3 in the molten fluoride mixture the UO2 solubility increased
The solubility of UO2 in ternary mixtures of LiFeBaF2(1:1)-LaF3 was determined over a range of composition in the temperature range 1423 Ke1523 K. The measurement confirmed that the UO2 dissolves in LaF3 and its solubility increases with increase in the
95
Ba2 La F7
N. Alangi et al. / Journal of Nuclear Materials 470 (2016) 90e96
5.0e+003
40
50
60
Ba2 La F7 Li Ba F3
Li Ba F3 Ba2 La F7
Ba2 La F7
Ba2 Ba2 La La F7 F7
70
LiLiBa Ba F3 F3 Ba2 La F7
Ba2 La F7 Li Ba F3 Ba2 La F7
Ba2 La F7
Ba2 La F7
Li Ba Ba2 LaF3 F7
Ba2 La F7
Ba2 La F7
Li Ba F3 Ba2 La F7
Ba2 La F7
Ba2 La F7
30
Ba2 La F7 Li Ba F3
0.0e+000 20
Li Ba F3 Ba2 La F7
1.0e+003
Li Ba F3
2.0e+003
Ba2 La F7
Ba2 La F7
3.0e+003
Ba2 La F7 Li Ba F3
Intensity (cps)
4.0e+003
80
Barium Lanthanum Fluoride Lithium Barium Fluoride
Fig. 5. Typical XRD plot of sample from frozen LiFeBaF2 1:1e10% LaF3 salt mixture.
Fig. 6. Variation of solubility of uranium oxide with temperature in LaF3 based fluoride bath.
Fig. 7. Solubility of uranium oxide in mixed fluoride melt containing LaF3/UF4 in LiF:BaF2 base salt.
Table 4 Solubility of UO2 in mixed electrolytes.
Division, Bhabha Atomic Research Centre for providing facilities and permission to publish this work, and specially thank Dr A K Suri incisive discussions. Authors thank Mr Sunil Dehade and Mr Mukesh Verma for their support in carrying out the experiments and acknowledge the help of Mr Hareendran and Dr Sujoy Biswas in the analysis by ICP-MS.
Electrolyte
Weight % UO2 solubility at 1473 K
LiFeBaF2(1:1)- 9.4 wt%LaF3-0.6 wt% UF4 LiFeBaF2(1:1)- 8.1 wt%LaF3-1.9 wt% UF4 LiFeBaF2(1:1)-28.1 wt%LaF3-1.9 wt% UF4
0.46 1.04 0.92
References LaF3 quantity. The UO2 solubility is also found to increase at higher temperature for a given concentration of LaF3. Replacing a small amount of LaF3 with UF4 has a net positive effect and it enhances the solubility. The exact reason for the increased solubility needs separate investigation. These results provide guidance for the selection of uranium free solvent for the electrolytic cells to produce liquid uranium having any isotopic composition without the need of a separate synthesis step for isotopically tailored UF4. Acknowledgment Authors are grateful to Head, Laser and Plasma Technology
[1] R.S. Hargrove, J.B. Knighton, R.S. Eby, J.H. Pashley, R.E. Norman, Integration of the avlis process into the nuclear fuel cycle, in: Summer Meeting of the AIChE, August 24-27, 1986. Boston, Massachusetts. [2] C.D. Harrington, A.E. Ruehle, Uranium Production Technology, 1959. [3] L.W. Niedarch, B.E. Dearing, The preparation of uranium metal by the electrolytic reduction of its oxides, J. Electrochem. Soc. 105 (6) (1958) 353e358. [4] R.D. Piper, R.F. Leifield, Electrolytic reduction of uranium metal from uranium oxides, I & EC Process Des. Dev. 1 (3) (1962) 208e212. [5] T.A. Henrie, Electrowinning rare-earth and uranium metals from their oxides, J. Metals (1964) 978e981. [6] P.A. Haas, P.W. Adcock, A.C. Coroneos, D.E. Hendrix, Small cell experiments for electrolytic reduction of uranium oxides to uranium metal using fluoride salts, Metall. Mater Trans. B 25 (1994) 505e518. [7] B.F. Greenfield, K.R. Hyde, The Solubility of Uranium Dioxide in Fluoride Melts, United Kingdom Atomic Energy Authority Report AERE R-6463, 1983.
96
N. Alangi et al. / Journal of Nuclear Materials 470 (2016) 90e96
[8] L.W. Niedrach, B.E. Dearing, Electrowining of Uranium from its Oxides e I. Laboratory Studies, Knolls Atomic Power Laboratory, Rep No KAPL-1761, 1957. [9] P.A. Haas, P. Cooper, Solubility of Uranium Oxide in Fluoride Salt at 1200 C, U.S. Department of Energy Report K/AIS-243, 1989. [10] J.W. Varwig, Solubility of UO2 in Molten Salt, Mallinckrodt Chemical Works, Rep No. MCW-1490, 1964.
[11] B. Porter, E.A. Brown, Determination of Oxide Solubility in the Molten Fluorides, Bureau of Mines Report of investigation 5878, 1961. [12] P.A. Haas, Stanley P. Cooper, Solubility of Uranium Oxides in Fluoride Salts at 1200 C, Am. Chem. Soc. 38 (No. 1) (1993) 26e30. [13] R.D. Piper, R.F. Leifield, Electroreduction of Uranium Oxides to Massive Uranium Metal, Mallinckrodt Chemical Works, Rep No. MCW-1447, 1960.
Contents lists available at ScienceDirect
Journal of Nuclear Materials journal homepage: www.elsevier.com/locate/jnucmat
Solubility of uranium oxide in molten salt electrolysis bath of LiFeBaF2 with LaF3 additive Nagaraj Alangi a, b, *, Jaya Mukherjee a, L.M. Gantayet b a b
Laser & Plasma Technology Division, Bhabha Atomic Research Centre, Mumbai, India Homi Bhabha National Institute, Mumbai, India
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 April 2015 Received in revised form 28 November 2015 Accepted 2 December 2015 Available online 12 December 2015
The solubility of UO2 in the molten mixtures of equimolar LiFeBaF2(1:1) with LaF3 as additive was studied in the range of 1423 Ke1523 K. The molten fluoride salt mixture LiFeBaF2 LaF3 was equilibrated with a sintered uranium oxide pellet at 1423 K, 1473 K, 1523 K and the salt samples were collected after equilibration. Studies were conducted in the range of 10%e50% by weight additions of LaF3 in the equimolar LiFeBaF2(1:1) base fluoride salt bath. Solubility of UO2 increased with rise in LaF3 concentration in the molten fluoride in the temperature range of 1423 Ke1523 K. At a given concentration of LaF3, the UO2 solubility increased monotonously with temperature. With mixed solvent, when UF4 was added as a replacement of part of LaF3 in LiFeBaF2(1:1)-10 wt% LaF3 and LiFeBaF2(1:1)-30 wt% LaF3, there was an enhancement of solubility of UO2. © 2015 Elsevier B.V. All rights reserved.
Keywords: Molten fluoride UO2 Solubility LaF3 additive
1. Introduction Electrolytic reduction of uranium oxide to uranium by molten salt electrolysis above the melting point of uranium is an attractive method of producing uranium metal for metallic nuclear fuel. This route for producing uranium metal assumes more significance in the context of growing importance of metallic nuclear fuels in the nuclear fuel cycle and the role of atomic vapour laser isotope separation (AVLIS) in it [1]. The pyro-electrochemical process is capable of accepting the feed material in all the forms of oxides and fluorides of uranium, and even partially oxidized metal. Being a single step low pressure process, this process is amenable for operation in radiochemical environment. The electrolyte is generally composed of fluorides of Li, Ba, Ca, etc. which provides the electrochemical window of operation with UF4 as the functional electrolyte [2e6]. The addition of UF4 enhances dissolution of UO2 and its decomposition to U4þ and O2 ions. The tetravalent uranium ions migrate to the cathode, and first get converted to trivalent ion and then to uranium metal at the cathode. Oxygen ion migrates to the graphite anode and gets converted to CO and leaves the cell. The electrolysis parameters such as current density, anode effect and
* Corresponding author. Laser & Plasma Technology Division, Bhabha Atomic Research Centre, Mumbai, India. E-mail address: [email protected] (N. Alangi). http://dx.doi.org/10.1016/j.jnucmat.2015.12.002 0022-3115/© 2015 Elsevier B.V. All rights reserved.
rate of addition of UO2 depend on the solubility of UO2 in the electrolyte. Electrolytic reduction of uranium oxide in the integrated nuclear fuel cycle plant would involve working with different isotopes of uranium. In this case UF4 solvent has to be made from uranium with a specific isotopic composition, which in itself is a difficult job. It will be particularly difficult to manage the various uranium isotopes generated in the thorium fuel cycle. Thus, an electrolytic reduction process with a non-uranium additive in the molten salt electrolyte will remove this constraint. The data of uranium oxide solubility with UF4 as the solvent in different supporting electrolyte mixtures is reported in the literature [7e12]. Varwig et al. [10]. have studied the solubility of UO2 in UF4 containing LiFeBaF2(1:1) electrolyte, which has been extensively used by Mallinckrodt Chemical Works for electrolysis [13]. They have reported that the UO2 solubility is a function of UF4 content in LiFeBaF2(1:1) electrolyte and it varies from 0.1 wt% of UO2 at 0% UF4 to ~2.3 wt% UO2 at 45 wt% UF4 at 1423 K. Haas et al. [12] have also studied the solubility of UO2 in UF4eLiFeBaF2 electrolyte. They found that the UO2 solubility at 1473 K is a function of UF4 content in LiFeBaF2(1:1) electrolyte, varying from 0.08 wt% of UO2 at 0% UF4 to ~4.47 wt% UO2 at 45 wt% UF4. Typically 25wt% of UF4 is used as solvent in LiFeBaF2(1:1) electrolyte for electrolysis which has solubility of ~2 wt%. LaF3 additive is a potential candidate for replacing functional electrolyte UF4. Among the rare earth fluorides LaF3 is easily
N. Alangi et al. / Journal of Nuclear Materials 470 (2016) 90e96
available and abundant. The cathodic deposition potential of lanthanum being higher than that of uranium, it provides a comfortable electrochemical window. However, there is no reported data on UO2 solubility in molten fluoride salt with LaF3 additive. The present work was to measure the solubility of uranium oxide in the LiFeBaF2(1:1) electrolyte with LaF3 additive in the temperature range of 1423 Ke1523 K, towards development of a new electrolyte for electrolytic reduction of UO2 above the melting point of uranium. The solubility of oxides in fluoride melts is predominantly due to the formation of complexes of uranium, fluorine and oxygen. It is reportedly enhanced with increasing amount of UF4 solvent [12].. Although a number of uranium oxyefluoride complexes exist, the complex prevailing at high temperature has never been proven in the molten fluoride system. Greenfield and Hyde have reported compound formation of UF4 with diluent salt, K2UF6 or 2KFUF4 in KF containing electrolytes and 4LiF.UF4, 7LiF.6UF4 in LiF rich electrolytes, which reduce the total availability of free UF4 which associates with UO2 to increase its solubility [7]. In a similar fashion, Ba2LaF7 phase has been observed at room temperature by XRD of samples from LiFeBaF2(1:1)-LaF3 mixtures. It can be hypothesized that LaF3 may exhibit complexes similar to that of UF4 in LiFeBaF2(1:1) bath, which will affect solubility of UO2.
2. Materials LiF and BaF2 of reagent grade (>99% assay) were sourced from M/s Sigma Aldrich with CAS No. 7789-24-4 and 7787-32-8 respectively. LaF3 (CAS No. 13709-38-1) with minimum purity of 99.5% was obtained from M/s Prabhat chemicals. Nuclear grade UF4 having a minimum assay of 99% was procured internally. Fresh defect free sintered UO2 pellets (~10.4 g/cc density) were obtained from Nuclear Fuel Complex, Hyderabad and were of nuclear fuel grade, which had a purity of 99.8%. The impurities in the sintered pellet were as shown in Table 1.
91
3. Experimental setup An experimental setup was designed and fabricated for this study. The setup consisted a recrystallized alumina tube of 80 mm ID and 90 mm OD 800 mm long, closed at one end was used as a muffle. Viton O-ring sealed water cooled stainless steel flanges were used on the open end of the alumina tube. The cover flange had openings for gas inlet and outlet connections for high purity argon with <0.1 ppm of oxygen. Other feed throughs on the flange were used for the pellet holder and the thermocouple. Heating of the alumina tube was done by a 6 kW resistive heating furnace controlled by a programmable PID controller with an accuracy of ±3 K. A pack of radiation shields of 78 mm diameter, supported from the flange was used to prevent heat losses and to maintain a uniform temperature zone. A high density graphite crucible of ~50 mm diameter was used to contain the salt mixture. The crucible containing the electrolyte rested on an alumina support tube of 45 mm outer dimension in the uniform temperature zone, which in turn rested on the floor of the alumina tube. The UO2 pellet holder was made out of high density graphite, which had a provision to hold one end of a UO2 pellet of ~13 mm diameter and ~19 mm height. The pellet holder had a provision to thread to a 8 mm diameter molybdenum rod. The high temperature structural members were made of either graphite or molybdenum. The furnace was housed inside a fumehood and the purge gas released into the fumehood was passed through a series of HEPA filters to separate the fine radioactive particulate mass and a water scrubber to remove fluoride vapour before exhausting to the environment. The schematic diagram of the experimental set-up is shown in Fig. 1 and the photograph in Fig. 2a. Fig. 2b shows the photographs of uranium oxide pellet and its holder. Due safety clearance from the regulatory body was obtained to carry out experiments.
4. Salt mixture preparation and purification All the salts used in the experiment were dried individually in vacuum at 1073 K for 3 h in a graphite crucible. They were retained as stock in desiccators. Around 30 g batch fluoride salt mixture was
Table 1 List of impurities in sintered UO2 pellet. Sl. No.
Element
Value, ppm
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.
Ag Al B Ca Cd Ce Co Cr Cu Dy EBC Er Eu Fe Gd Mg Mn Mo Na Ni Si Sm Zn
<1.0 <25.0 0.12 13.0 <0.1 <0.5 <2.0 <10.0 <6.0 <0.1 ~0.71 <0.1 <0.1 47.0 ~0.08 9.0 2.0 5.0 35.0 1.0 <30.0 <0.1 <1.0
EBC-Equivalent boron content. SSA-Specific surface area of UO2 powder 2.55 m2/g in case where UO2 powder was used to prepare UO2þC pellets.
Fig. 1. Schematic of the experimental setup UO2 for solubility study in fluoride melts (a) Immersion with pellet holder (b) Full immersion of the pellet in the electrolyte.
92
N. Alangi et al. / Journal of Nuclear Materials 470 (2016) 90e96
3.
4.
5.
6.
7.
Fig. 2. a Photograph of the experimental setup for UO2 solubility inside fumehood. b Uranium oxide pellet and pellet holder used in the solubility experiments.
prepared by weighing the individual salts of LiF, BaF2 and LaF3 in the desired proportions and mixing by mortar and pestle. The pretreatment involved first equilibrating the electrolyte mixture in a graphite crucible at 673 K for 24 h and then at 1473 K for another 24 h in a flowing pure argon atmosphere. The composition of the salt mixtures thus prepared and characterized consisted of equimolar LiF: BaF2(1:1) base salt with 10, 20, 30 and 50 wt% LaF3. The mixed electrolyte of LaF3 and UF4 will be produced during the course of electrolysis. Hence, in a few experiments 8.1 wt%LaF31.9 wt% UF4 and 28.1 wt%LaF3-1.9 wt% UF4 mixtures were used instead of 10 wt% LaF3 and 30 wt% LaF3, respectively.
5. Methods of determination of saturation solubility The method of determination of saturation solubility of the oxide in an electrolyte bath in this study was selected from among several known methods described below. 1. In the fluoride melts, oxide is added to the salt bath in a transparent quartz container and laser based visual techniques are used to determine the point beyond which oxide is not soluble. Since the fluoride melt attacks quartz this method was not chosen. 2. The second method involves equilibrating the salt bath with different amounts of oxide, quickly quenching and observing the undissolved oxide metallographically. Quenching the salt
mixtures was avoided as it involves potential release of radioactive powders and vapors. Potentiometric measurement of the electrode potentials with respect to a reference electrode with addition of different amounts of oxide and the point beyond which the potential remains constant may be considered as the limiting solubility. The electrode potential measurement is possible in fluoride media. However, reliable high temperature reference electrode was not available to adopt this method. The potential at which the anode effect occurs depends on the concentration of dissolved uranium oxide. In this technique, a known quantity of uranium oxide powder is dissolved and the point beyond which the anode effect potential (determined by linear sweep voltammetry) does not vary is considered the limit of solubility. Here too, the non availability of a high temperature reference electrode was a handicap. The fifth method involves, collecting a sample of the fluoride melt that has been saturated with oxide and conducting the chemical analysis to determine the extent of solubility after filtering out the suspended undissolved oxide fully. Several attempts of collecting a sample through a porous graphite filter from the surface of the fluoride bath again did not yield reliable results. In the sixth method, a solid shape of the oxide is immersed in the fluoride bath and after sufficient time has elapsed, the weight change is measured. A sintered UO2 pellet was immersed in the salt bath. However, the mechanical operation to remove the adhering salt led to chipping of the pellets, which introduced errors. Similar problem of fluoride salt adhesion to solid rods also have been reported by Greenfield and Hyde [7]. A variant of the last mentioned is to measure the concentration of uranium in the electrolyte instead of measuring weight change in the UO2 pellet. The major mass of the solidified electrolyte was easily separable from the pellet. The minor quantity of salt adhering to UO2 pellet, if any, was removed separately and was not mixed with the bath sample. The uranium concentration in the sample was measured and the total uranium in the fluoride bath was estimated. The procedure for determination of uranium in fluoride salt mixture was also developed and described in the experimental method section. It was ensured that the pellet did not have any sharp edges, which could chip off during the experiment. Around fifteen trials were conducted to qualify this method. In all the trials the integrity of the sintered pellet remained intact. This method which was free of the problems of the other methods, was adopted
It is to be noted that uranium concentration can be unambiguously related to the dissolution of UO2. However, oxygen concentration will not be exactly equivalent to the oxygen from UO2 dissolution, as fluoride salts invariably pickup oxygen from the moisture and oxygen in the environment [12]. During this work, authors have observed presence of substantial quantities of complex oxy-fluorides while preconditioning. 6. Experimental methods Two experimental procedures were used in the solubility studies. In the first, a pre-weighed sintered UO2 pellet was held in a graphite pellet holder, and after preconditioning of the fluoride bath the pellet was immersed in the molten salt and removed before cooling of the salt (as shown in Fig. 1a). In the second procedure the pellet was immersed fully without a holder in the molten salt and was allowed to cool down along with the bath (Fig. 1b) after equilibration. Since the total surface area available was higher with the UO2 pellet fully immersed in the salt bath, this
N. Alangi et al. / Journal of Nuclear Materials 470 (2016) 90e96
procedure was adopted for all the experiments. All the experiments, particularly the safety related ones, were checked before commencing each experiment. The sintered UO2 pellet was weighed and placed in the high density graphite crucible. The fluoride salt mixture was added to the annular gap between the crucible and the pellet. Before commencing the flow of argon purge gas, the complete system was evacuated to 1E-2 mbar and back filled with argon at least 8e10 times to remove the trapped air. The next step involved equilibrating the bath and the sintered UO2 pellets for 48 h at the temperature of study. Initially, the time required for reaching saturation was determined by conducting total 8 experiments with immersion duration of 12, 24, 36 and 48 h. It was observed that the concentration reached saturation after 24 h of immersion of the pellet. In all the reported experiments in this study the sintered UO2 pellets were equilibrated with the salt bath for 48 h, which ensured saturation even with the experimental uncertainties. In an attempt to study the effect of increase in the surface area available for dissolution, in one of the experiments equilibrating a UO2þC pellet with the salt bath was tried. It was observed that the UO2þC pellet had disintegrated and no meaningful determination of solubility limit was possible. Hence using UO2þC pellet was not continued. The equilibration temperatures chosen were 1473 K, 1523 K and 1573 K. In the experiments the argon purge was continued throughout till the system was cooled and ready for dismantling. The oxygen concentration in the inlet gas was continuously monitored by an online oxygen gas analyzer (Systech 800 Series analyzer). After cooling to room temperature the frozen salt mixture was removed. The photograph in Fig. 3 shows the colors developed in various samples. The graphite crucible had a draft angle of 3 to remove the solidified salt mixture easily. The salt mixture was separated from the pellet and the entire sample was ground for homogenizing and collecting a sample for analysis of the total uranium concentration. All the uranium in the sample was assumed to be obtained from the dissolved UO2. Analysis of uranium in the fluoride salt posed a challenge. The standard method of preparing sample by dissolution in HNO3 (15.8 M) or HCl (11.6 M) always led to an undissolved residue. Hence, microwave dissolution technique was developed.
93
heating were used. A standard microwave oven of 1600 W in continuous mode supplied by M/s Milestone was used for heating. The acids tried for dissolution were pure HNO3(15.8 M), pure HCl (11.6 M) and combination HNO3 and HCl, and HCl with different proportions of boric acid. The microwave heating was carried out for 1 h. A summary of the dissolution studies is given in Table 2. The combination of 6 ml Conc HCl, 6 ml H3BO3 solution (0.7 M) was selected. The time was then optimized to 20 min, which ensured complete dissolution of the sample. Inductively coupled plasma mass spectrometry (ICP-MS) Model JY238 Ultrace was used for the determination of the total uranium present in the dissolved sample. All the frozen samples were also characterized by X-ray powder diffraction (XRD), to identity the phases of LiF, BaF2, UF4, UO2, LaF3 present. A Rigaku Miniflex II diffractometer with Cu Ka1 radiation (l ¼ 1.5405 Å) was used. The XRD data for indexing and cell-parameter calculations were collected in the scanning mode with a step of 0.02 in the 2q range from 20 to 80 and a scanning rate of 2.0 min1 with silicon as the internal standard. 7. Results and discussions In the high temperature molten salt based experiments several unavoidable errors occur in the measurements. Some of them are due to not so clean separation of the frozen electrolyte from the pellet, liquid phase segregation, inhomogeneous precipitation of the dissolved UO2, incomplete dissolution and error in the ICP-MS. Therefore, each experiment was carried out 3 times for obtaining reliable data. Thus, the individual solubility data presented here is the mean of at least 3 observations. The uranium oxide solubility data reproducibility is around ±5% of the value reported. This limit includes the errors enumerated above along with preparing the samples for analysis. At first the experiments were done with 10 wt % LaF3 and 50 wt% LaF3. Reporting significantly different result with ±5% reproducibility, the number of data points in the range of interest, that is 10 wt% to 30 wt% LaF3 for electrolysis was limited. It was then decided to study the linearity of the solubility between 10 wt% and 30 wt% LaF3. As the results were reassuredly linear, 20 wt% LaF3 was studied to cover the range of electrolysis. 7.1. Effect of quantity of LaF3
6.1. Development of microwave dissolution technique for preparing sample for ICPMS For complete dissolution of the solid sample we tried a noninterfering complexing agent such as boric acid (H3BO3). High pressure dissolution and high temperature operation is necessary to dissolve fluoride salt. High pressure TFM-fluropolymer based proprietary containers (M/s Milestone) of capacity 100 ml, designed for 100 bar and 573 K and compatible with microwave
Fig. 3. Colors developed due to uranium concentration in different salt baths. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Studies were conducted with 10 wt%, 20 wt%, 30 wt% and 50 wt% LaF3 additions in LiFeBaF2(1:1) base salt. The solubility of UO2 in equimolar LiFeBaF2(1:1) at 1473 K without any solvent is reported as 0.08% by weight [12]. In our experiments of measurement of solubility in equimolar LiFeBaF2(1:1) at 1473 K, the dissolved uranium was found to be below detection limit (BDL) (<20 mg/1.5 g sample) when measured by Energy Dispersive X-ray Fluorescence Spectrometer (EDXRF). The dissolved uranium in LiFeBaF2(1:1)LaF3 melts was determined by ICP-MS following the procedure outlined above. The results of solubility at different LaF3 concentrations are presented in Table 3 and graphically in Fig. 4. The solubility of UO2 was higher with increase in LaF3 in the melt at all the temperatures of the experiments, namely 1423 K, 1473 K and 1523 K. The solubility variation between 30 wt% LaF3 and 50 wt% LaF3 was within the standard deviation of the analysis. The reason for not having a large variation of UO2 solubility in melts containing 30 wt% LaF3 and 50 wt% LaF3 may be owing to the reduction in the total availability of free LaF3, which could form Ba2LaF7 compound with BaF2 of the base salt mixture. At 1523 K probably the compound is less stable thereby freeing some LaF3 which increases the solubility of UO2 at that temperature. In XRD investigations BaF2 and Ba2LaF7 phases were seen in XRD of frozen salt mixture. Probably, all the dissolved uranium was present as solid solution.
94
N. Alangi et al. / Journal of Nuclear Materials 470 (2016) 90e96
Table 2 Parametric study for microwave dissolution of fluoride samples. Salt and solvent
A
D
E
Quantity, mg
50
100
50
B 100
50
C 100
100
50
100
200
>300
LiF CaF2 LiFeCaF2 LiFeBaF2 LiFeBaF2-10 wt%UF4 LiFeBaF2-25 wt%UF4 LiFeBaF2-10wt%LaF3 LiFeBaF2-20wt%LaF3 UO2
√ √ √ X X X X X e
√ √ √ X X X X X e
√ √ √ X X X X X e
√ √ √ X X X X X e
X X X X X X X X e
X X X X X X X X e
X X X X X X X X e
e e √ √ √ √ √ √ √
e e √ √ √ √ √ √ √
e e √ √ √ √ √ √ √
e e √# √# √# √# √# √# √#
A: 7 ml Conc HCl, 1 ml Conc HNO3 solvent, for 20 min at 220 C. B: 7 ml Conc HCl, 1 ml Conc HNO3 solvent, for 30 min at 220 C. C: 5 ml Conc HCl, 5 ml Conc HNO3 solvent, for 30 min at 220 C. D: 8.7 ml Conc HCl, 1.3 ml Conc HNO3, 0.5 gm H3BO3 solvent, for 40 min at 220 C. E: 6 ml Conc HCl, 6 ml H3BO3 solution (0.7 M), for 20 min at 220 C. √ - Complete dissolution. X - Incomplete dissolution. #-with proportionate increase of solution required for dissolution of 200 mg.
Table 3 Solubility of UO2 in LiFeBaF2(1:1)-LaF3 melts. S.no
Sample No and weight % LaF3
UO2 solubility at 1423 K, weight %
UO2 solubility at 1473 K, weight %
UO2 solubility at 1523 K, weight %
1 2 3 4
10 20 30 50
0.11 e 0.2 0.19
0.415 0.482 0.57 0.56
0.76 e 0.8 0.984
wt% wt% wt% wt%
LaF3 LaF3 LaF3 LaF3
substantially when the temperature was raised from 1423 K. In the LiFeBaF2(1:1)- 10 wt% LaF3 electrolyte the solubility of UO2 was 0.11 wt% at 1423 K, which increased to 0.145 wt% at 1473 K and further increased to 0.76 wt% at 1523 K. Similar increasing trend with temperature was observed for LiFeBaF2(1:1)- 30 wt% LaF3 and LiFeBaF2(1:1)- 50 wt% LaF3 electrolytes. The trend can be clearly seen in Fig. 6. The maximum solubility of UO2 in LaF3 containing electrolytes observed at 1523 K in LiFeBaF2(1:1)- 50 wt% LaF3 electrolyte, was 0.984 wt%. 7.3. Effect of addition of UF4 to LaF3eLiFeBaF2(1:1) electrolyte (mixed electrolyte)
Fig. 4. Variation of solubility of UO2 in LaF3 solvent based molten fluoride salt.
Perhaps, for the same reason, UO2 or UF4 were also not seen in the XRD plot. As the dissolved uranium was less than 1 wt%, the uranium bearing phase could not be seen. The XRD plot of LiFeBaF2(1:1)-10 wt% LaF3 is shown in Fig. 5.
1.9 wt% of UF4 was added as a replacement of LaF3. Experiments were conducted with LiFeBaF2(1:1)- 8.1 wt%LaF3-1.9 wt% UF4 and LiFeBaF2(1:1)-28.1 wt%LaF3-1.9 wt% UF4 at 1473 K and results are shown in Table 4. With the addition of 1.9 wt% UF4 to the LaF3, the combined effect of the mixed electrolyte resulted in an increase in UO2 solubility in comparison to the UF4 free electrolyte. The total uranium was analyzed and the contribution from UF4 was accounted for to arrive at the solubility of UO2. The UO2 solubility at 1473 K in LiFeBaF2(1:1)- 8.1 wt%LaF31.9 wt% UF4 and LiFeBaF2(1:1)-28.1 wt%LaF3-1.9 wt% UF4 electrolyte was 1.04 wt% and 0.92 wt% respectively, corresponding solubilities in UF4 free electrolyte were 0.415 wt% and 0.57 wt %. The higher dissolution due to the addition of UF4 is shown in Fig. 7. This could be due to greater probability of complex formation between UO2 and UF4.
7.2. Effect of temperature
8. Conclusions
Data of UO2 solubility as a function of temperature in different electrolytes is important as the temperature of electrolysis plays an important role. As anticipated intuitively, at a given percentage of LaF3 in the molten fluoride mixture the UO2 solubility increased
The solubility of UO2 in ternary mixtures of LiFeBaF2(1:1)-LaF3 was determined over a range of composition in the temperature range 1423 Ke1523 K. The measurement confirmed that the UO2 dissolves in LaF3 and its solubility increases with increase in the
95
Ba2 La F7
N. Alangi et al. / Journal of Nuclear Materials 470 (2016) 90e96
5.0e+003
40
50
60
Ba2 La F7 Li Ba F3
Li Ba F3 Ba2 La F7
Ba2 La F7
Ba2 Ba2 La La F7 F7
70
LiLiBa Ba F3 F3 Ba2 La F7
Ba2 La F7 Li Ba F3 Ba2 La F7
Ba2 La F7
Ba2 La F7
Li Ba Ba2 LaF3 F7
Ba2 La F7
Ba2 La F7
Li Ba F3 Ba2 La F7
Ba2 La F7
Ba2 La F7
30
Ba2 La F7 Li Ba F3
0.0e+000 20
Li Ba F3 Ba2 La F7
1.0e+003
Li Ba F3
2.0e+003
Ba2 La F7
Ba2 La F7
3.0e+003
Ba2 La F7 Li Ba F3
Intensity (cps)
4.0e+003
80
Barium Lanthanum Fluoride Lithium Barium Fluoride
Fig. 5. Typical XRD plot of sample from frozen LiFeBaF2 1:1e10% LaF3 salt mixture.
Fig. 6. Variation of solubility of uranium oxide with temperature in LaF3 based fluoride bath.
Fig. 7. Solubility of uranium oxide in mixed fluoride melt containing LaF3/UF4 in LiF:BaF2 base salt.
Table 4 Solubility of UO2 in mixed electrolytes.
Division, Bhabha Atomic Research Centre for providing facilities and permission to publish this work, and specially thank Dr A K Suri incisive discussions. Authors thank Mr Sunil Dehade and Mr Mukesh Verma for their support in carrying out the experiments and acknowledge the help of Mr Hareendran and Dr Sujoy Biswas in the analysis by ICP-MS.
Electrolyte
Weight % UO2 solubility at 1473 K
LiFeBaF2(1:1)- 9.4 wt%LaF3-0.6 wt% UF4 LiFeBaF2(1:1)- 8.1 wt%LaF3-1.9 wt% UF4 LiFeBaF2(1:1)-28.1 wt%LaF3-1.9 wt% UF4
0.46 1.04 0.92
References LaF3 quantity. The UO2 solubility is also found to increase at higher temperature for a given concentration of LaF3. Replacing a small amount of LaF3 with UF4 has a net positive effect and it enhances the solubility. The exact reason for the increased solubility needs separate investigation. These results provide guidance for the selection of uranium free solvent for the electrolytic cells to produce liquid uranium having any isotopic composition without the need of a separate synthesis step for isotopically tailored UF4. Acknowledgment Authors are grateful to Head, Laser and Plasma Technology
[1] R.S. Hargrove, J.B. Knighton, R.S. Eby, J.H. Pashley, R.E. Norman, Integration of the avlis process into the nuclear fuel cycle, in: Summer Meeting of the AIChE, August 24-27, 1986. Boston, Massachusetts. [2] C.D. Harrington, A.E. Ruehle, Uranium Production Technology, 1959. [3] L.W. Niedarch, B.E. Dearing, The preparation of uranium metal by the electrolytic reduction of its oxides, J. Electrochem. Soc. 105 (6) (1958) 353e358. [4] R.D. Piper, R.F. Leifield, Electrolytic reduction of uranium metal from uranium oxides, I & EC Process Des. Dev. 1 (3) (1962) 208e212. [5] T.A. Henrie, Electrowinning rare-earth and uranium metals from their oxides, J. Metals (1964) 978e981. [6] P.A. Haas, P.W. Adcock, A.C. Coroneos, D.E. Hendrix, Small cell experiments for electrolytic reduction of uranium oxides to uranium metal using fluoride salts, Metall. Mater Trans. B 25 (1994) 505e518. [7] B.F. Greenfield, K.R. Hyde, The Solubility of Uranium Dioxide in Fluoride Melts, United Kingdom Atomic Energy Authority Report AERE R-6463, 1983.
96
N. Alangi et al. / Journal of Nuclear Materials 470 (2016) 90e96
[8] L.W. Niedrach, B.E. Dearing, Electrowining of Uranium from its Oxides e I. Laboratory Studies, Knolls Atomic Power Laboratory, Rep No KAPL-1761, 1957. [9] P.A. Haas, P. Cooper, Solubility of Uranium Oxide in Fluoride Salt at 1200 C, U.S. Department of Energy Report K/AIS-243, 1989. [10] J.W. Varwig, Solubility of UO2 in Molten Salt, Mallinckrodt Chemical Works, Rep No. MCW-1490, 1964.
[11] B. Porter, E.A. Brown, Determination of Oxide Solubility in the Molten Fluorides, Bureau of Mines Report of investigation 5878, 1961. [12] P.A. Haas, Stanley P. Cooper, Solubility of Uranium Oxides in Fluoride Salts at 1200 C, Am. Chem. Soc. 38 (No. 1) (1993) 26e30. [13] R.D. Piper, R.F. Leifield, Electroreduction of Uranium Oxides to Massive Uranium Metal, Mallinckrodt Chemical Works, Rep No. MCW-1447, 1960.