¥ w L do fl.h“[ ey 3 _ liir o OAK RIDGE NATIONAL LABORATORY operated by UNION CARBIDE CORPORATION W for the U.S. ATOMIC ENERGY COMMISSION ORNL- TM- 1051 COPYNO. - (-8 DATE - March 11, 1965 " ~-© RECONSTITUTION OF MER FUEL BY REDUCING UF, GAS TO UF) IN A MOLTEN SAIT L. E. McNeese ¢. D. Scott / o 3 . oy 5 5 S B & ) o o 5 5154 8 . 'U: 3 g 2 & o — O o m“ B & 5 O L = =) it B g 3 o CZ> a O H o - 9 o=t H l—lw o HSae 9 = mm Z ‘o, o = g 21219 NG E'c'g g 9\o mfl)mtfi-—i ERINIE T .~ »p—d m g F o bfl-.E,’U - s 5| g = g ol 50fi CAUTION F mination of the preliminary informa BMtdined in this documeni is uvnavuthg d—wiHr Srr=apprayvg f the ORNL Pdicni anch-—o e egql g { ument req shts of EC McnaiCbcpier 3202. only to meet minlmum LEGAL NOTICE This report was prepared as an account of Government sponsored work. Neither the United States, nor the Commission, nor any person acting on behalf of the Commission: A. Makes any warranty or represenfation, expressed or implied, with respect to the accuracy, completenass, or usefuiness of the information contained in this report, or that the use of any information, apparatys, method, or process disclesed in this report may not infringe privately owned rights; or B. Assumes any liabilities with respect to the use of, or for damages resulting from the use of any information, apparatus, method, or process disclosed in this repory, As used in the above, “‘person acting on behclf of the Commission’ includes any employee or contractor of the Commission, or employese of such contractor, to the extent that such employee or contractor of the Commission, or employee of such contractor prepares, dissominafas, or provides access to, any information pursuant to his employment or controct with the Commission, or his employment with such contractor. CHEMICAI, TECHNOLOGY DIVISTION ING RECONSTITUT ION OF MSK FUEL BY REDUC EN SALT UF6 GAS TO UFh N A MOILT by L. BEe MclNeese ¢. D. Scott NOTICE ——————a This report was prepared as an account of work sponsored by the United States Government. Neither the United States nor the United States Fnergy Research and Development Adwinistration, nor any of their employees, nor any of their contractors, subcontractors, or their employees, makes any warranty, express or implied, or assumes any legaf liability or responsibility for the accuracy, completeness or usefulness of any information, apparatus, product or process disclosed, or tepresents that its use would not infringe privately owned rights, OAK RIDGE NATIONAL LABORATORY Ozk Ridge, Tennessee operated by UNION CARBIDE CORPORATION for the U.S. ATOMIC ENERGY COMMISSTON iii CONTENTS Abstract + ¢ ¢+ a & o o s v 4 s e 8 e e e e s s e Introduction + « . + .« . Proposed Process and Application to MSR Processing Experimental Equipment and Procedure Discussion of Experimental Results . . . . . . . Conclusion and Recommendations . . « . & References e e s e & e » Page 10 13 1h RECONSTITUTION OF MSR FUEL BY REDUCING UF6 GAS TO UFH IN A MOLTEN SALT L. E. McNeese ¢. D. Scott ABSTRACT The direct reduction of UF, to UF, in a molten salt ig proposed as a step in tge puri%ication of fuel salt from a molten salt reactor. This step would replace the conventional method of reduction in which UF, is reduced to UF, powder in a H, - F_ flame. RedlUction of the UF, In a molten sal% will result in a shorter and moré direct process for fuel salt purification. The reduction is to be effected in two steps which consist of absorption of UF6 into & molten salt containing UF, and of reduction of the resulting intermediate fluorides to UF, with hydro- gen. Experimental data on the absorption step are presented and information concerning the reduction of intermediate fluorides is considered. INTRODUCTION One proposed processing step for Molten Salt Reactor (MSR) fuel is the reduction of purified UF6 to UFh so that the UFu can be re- (1 turned to barren, purified fuel salt. ) The usuai method for re- ducing UF6 to UFLL is by use of H2 in a HE—F2 flame: (H., + I, ) UF6 + H2 2 2 - UFu + 2HF. This reduction is carried out in a tall column where the UF6 and H2 are introduced into a H,-F, flame and dry UF) powder (finely divided) 2 is collected. It is a routine production operation and there is much (2,3) available operating information. Such a process would not be desirsble for remote operation. It involves a solids handling problem which routinely requires equipment access and process control is some- times difficult. It would be desirable to reduce the UF6 to UFh in a molten salt environment, and thus circumvent the problems of solids handling and fuel make-up. Past experience of other workers has indicated the 2 feasibility of absorbing UF, into molten salt which contains UF, and | () 2 found that corrosion was not severe in absorption of F2 by molten salt reducing the absorbed UFé to UFM by sparging with H Kirslis containing U until the intermediate fluoride of uranium had a fluoride content greater than UF5. Lon%£5) found that H2 would reduce UFu to UF3 in & molten salt and Blood ~’ has reduced various metal fluorides in molten salts by use of HE' This report presents a proposed continuous processing method for the reduction of UF6 to UFu in a molten salt environment by absorption of UF, in the salt and reduction with H The results from a scouting test are analyzed to indicate process fgasibility. PROPOSED PROCESS AND APPLICATION TO MSR PROCESSING The current scheme for processing Molten Salt Reactor (MBR) fuel consists of removal of uranium as UF6 and volatile fission products (FP) from the salt by fluorination, separation of refractory FP from the salt by distillation, and recombination of the volatilized uranium and purified barren salt for return to the MSR (Fig. l)(l). During the fluorination step, both uranium and volatile fission products are re- moved from the salt bty the reactions: + I —"'"-‘UFéy UFh(in molten salt) 2 FP(in molten salt) + Fg-—-§-volatlle FP fluorides. The UF6 and volatile FP fluorides will be separated by sorption tech- nigues and the uranium will then be r=introduced as UFh to the purified barren salt to form the MSR fael. Thus, there must be a method for re- ducing UF6 to UFM' Since the end result of the UF6 reduction will be a2 solution of UFbr in molten salt rather than UFu as a dry powder, it is attractive to carry on the reduction in a molten salt environment and preferably in the purified barren salt. To achieve this requirement, UF6 can be con- tacted with a molten fluoride salt containing some uranium as UFh where it will be absorbed by reaction with the UFh to form an equivalent inter- mediate fluoride of uranium, such as UF in the salt: 5} ORNL DWG 65-1790 VOLATILE FP t SORPTION SEPARATION UF¢4 FP + UFg Ho Fo SORBED FP REDUCTION FLUORINATION F, BARREN SALT .._’ DISTILLATION FUEL MAKE-UP MSR l REACTOR CERAMIC FP PURIFIED FUEL SALT & FILTER Fig. 1. MSR Fuel Processing with Conventional UF, Reduction. UFh(salt) + Wy UFS(salt) This intermediate fluoride will then be reduced to UFh in the salt by means of ng + HF . UF + 1/21{2-——-@-”( 5(salt) salt) Such a process could be carried out continuously in a columm in which the barren salt and UF6 are introduced at the bottom of the column along with salt containing UFu which is recycled from the top of the column (Fig. 2). As the salt and UF6 progress up the column, the UF6 will be absorbed in the salt and subsequently reduced to UFM as it passes inte the H2 reduction section. The column off-gas will be a mix- ture of H2 and HF and a side stream of the overhead molten salt wiil be ready fTor return to the nuclear reactor core after filtration since the HF and H2 sparge usually given make-up salt will have been achieved in the reduction column. When this reduction step is incorporated into the flowsheet, the resulting process is more direct and shorter (Fig. 3). Tnitial tests (uext sect) indicate that the absorption step is very rapid, however, it will be desirable to keep the fluoride content of the intermediate fluoride below that equivalent to UF5 in order to minimize corrosion. The rate of the hydrogen reduction reaction is not known, although the limited data available loocks favoravle. In studying the reduction of UFM to UF in molten salt by H,. by the reaction: 3 2 : 1 v UF), (ga1t) + L/ 2Hy——=UF tHE 3(salt) observed that equilibrium was established between a H2=HF stream (5) and molten salt containing uranium fiuvorides after the gas bubbles had Long rigen a few inches through the salt. His data alsc indicate only 1% reduction of UF), to UF, by a gas stream cortaining 1% HF in H, at a faaid 3 pressure of 1 atm at £00°C. EXPERIMENTAL EQUIPMENT AND PROCEDURE The experimental equipment was assenbled from existing equipment available as a result of work in support of the Molten Salt Fluoride Volatility Process. Means were provided for contacting UFM’ disscived in moliten salt, with UF6 in the first step of the reduction process aud BARREN SALT Fig. 2. SALT + UFy " PUMP SALT ———» ORNL DWG 65-1791 H2+HF ~~~{—————p FUEL SALT e UFy Continuous Reduction of UFg by Hy in a Molten Salt. ORNL DWG 65-1792 VOLATILE FP 1 ) SORPTION SEPARATION UFg FP + UFé ‘ — — SORBED FP X RED TION wlHy —p U . > U — F, 2 v UCTION UFg FP + SALT I DISTILLAT|ON MSR REACTOR CERAMIC FP PURIFIED FUEL SALT FILTER Fig. 3. MSR Fuel Processing with Continuous UF, Reduction in Molten Salt. 7 for contacting the resulting uranium fluoride with hydrogen in the second step. Details of the experimental equipment and of the procedure for testing the reductlon process are discussed in the following sections. EQUIPMENT AND MATERTALS The reduction test was carried out in the vessel shown schematic- ally in Fig. 4. The vessel was constructed from 4-in.-diam Sch LO nickel pipe and was 26 inches in length. A 3/8-in. nickel inlet line was located in the center of the vessel and terminated 1/4-in. from the bottom of the vessel. A 3/h=in., fitting on the top flange allowed the insertion of a cold, 3/8-in. nickel rod which was used for sampling the salt., The vessel was heated by two nichrome wire resistance furnaces., A flow diagram for the equipment used in the test is shown in Fig. 5. The egquipment included a UF6 metering system, a hydrogen metering syse tem, means for purging both the UF6 system and HE system with NE’ the reduction vessel, and a NaF trap downstream from the vessel for absorb- ing UF6 or HF from the off=-gas of the reduction vessel. The salt charge was prepared by mixing LiF, Zth and UFu’ The LiF was reagent grade and contained<0.23 wt % impurities (mostly NaF). The zirconium content of the Zth was found by analysis to be Sh.78% which compares favorsbly with the stoichiometrical value of SL.6h; the uranium content of the UFu was found to be 76.9% which also compares favorably with the stoichiometrical value of 75.8%; and the uranium hexafluoride contained less than 200 ppm impurities. Hydrogen that was used contained less than 0.005 vol fraction impurities. FXPERIMENTAL PROCEDURE A salt charge consisting of 5320 g Zth, 863 g LiF, and 61.8 g UF), (0.197 gmole UFu) was placed in the reduction vessel and heated to 6OOOC. At this temperature the depth of molten salt was 12 inches. The salt mixture had a UF) concentration of" 1wt % and a melting point of approx- imately 51000. A salt sample (UR-1) was taken by insertion of a cold 3/8-in.-diam nickel rod into the molten salt. A UF, flow of 1.5 g/min wag then fed through the by-pass around the reduction vessel for 16 ORNL DWG 65-1793 /3/8” NICKEL GAS INLET 3/8"-~diam. OFFGAS LINE SR R AR S COPPER GASKET 4-in. SCHEDULE 40 NICKEL PIPE R R et R e ///,’* (A% /?‘Z,’/;f ,'/’//2‘/4' L N R R N 26 tH i //.;//-;“A e N Ak Fig. 4. Vessel Used for Reduction of UF, to UF4 in a Molten Salt. PRESSURE CONTROL VALVE H, SUPPLY UF4 SUPPLY CALIBRATED CAPILLARY ORNL DWG 65-1794 REDUCTION VESSEL NaF TRAP I————D OFFGAS Fig. 5. Flow Diagram for Equipment Used in Reduction of UFé to UF4 in a Molten Salt, 10 minutes in order to free the system of N The UF6 flow was then divert- ed into the dip line of the reduction veisel and was continued for 25 minutes. At this time, the UF, flow was étopped and the system was purged with N, for 5 minutes a%ter which a galt sample (UR-2) was taken. The guantity Sf UFQ fed to the system during this step was 38.2 g (0.108 gmoles). Tfie salt was then purged with H, at the rate of 95 st. cm3/min for 25 minutes. A total of 0.107 gmoles H, was added during 2 this step. After the system was purged with N, for 10 min, a salt e sample (UR-3) was taken. The system was then allowed to cool down over- night. The following day the system was heated to 600°C and a salt sam- ple was taken (UR-4). ‘The salt was then sparged with H. at a rate of 85 2 st. cms/min for 20 min during which a total of 0.076 gmoles H2 was fed to the system. The system was then purged with N2 for 10 min and a salt sample (UR-5) was taken. The system was then cooled and the test con- cluded. DISCUSSION OF EXPERIMENTAL RESULTS Two questions related to the experimental work are of primary inter- est. These are (1) the fraction of UFQ which was absorbed by the molten salt and (2) the valence of the uraniu; in the resulting mixture. Also, of interest are the concentration level of trace impurities such as Ni and 02- The composition of the various salt samples is given in Table 1. The uranium concentration in the initial salt (UR-1) was found by analy- sis to be 0.6h6 wt % which is 11.2% lower than the calculated uranium concentration of 0.75 wt %$. The calculated uranium concentration for complete absorption of the UF, bubbled through the salt during the 25 min addition period was 1.15 ;t %. The average uranium concentration after the UF; addition was found by analysis to be 1.07 wt % which is ™ lower than the calculated value. It was concluded that, within the accuracy of the experimental data, complete absorption of the UF6 had occurred. Tt is believed that the addition of UF, to a salt containing UF) results in the formation of dissolved fluorides of uranium with a va- lence intermediate between +4 and +£. This behavior is indicated by Table 1. Composition of Salt Samples Taken During Uranium Reduction Experiment Sample U +h U+6 Zr Li Ni 02 Remarks wt % wt % wt % wt % wt % ppm ppm UR-1 0.666 L6 .65 3.6h 87h LOL5 Tnitial salt melt UR-2 1.05 .95k < .01 933 Le9s After UF, addition UR-3 1.01 1.07k <.01 1002 Loko After 1st H, asddition [ UR-4 1.07 0.990 < .01 1007 3060 After cooling over- night amd remelting UR-5 1.1h 0.922 < .01 862 3245 After 2nd H, sparge 12 the fact that quantities of F2 sufficient for the formation of UF5 can be absorbed by molten salt containing UFu without the evolution of UF6. A similar behavior is also noted in reactions between UFh and UF6 in the absence of molten salt to yield intermediate fluorides such as UfiFlT' Tt is assumed that uranium present in a moliten salt as an intermediate b +6 fluoride will appear as U* and U = after dissolution in phosphoric acid (7) The concentration of U in the sample after UF6 addition was below the limit of detection of 0.01 wt % and the U‘+lL concentration was 0.95 wt % (Table 1). After the first and second hydrogen sparges, the U+h in preparation for analysis. concentration was found to be 1.07 wt % and 0.922 wt %, respectively. Although differences in U+ concentration were oObserved, it is felt that these are within the limits of analytical error and are not meaningful. Reduction of the Ufi6 to U'—I_LL probably occurred during the addition of UF6 by reaction of the intermediate fluoride with nickel from the vessel wall or with reduced fluorides of nickel, chromium, or iron initially present in the salt. The nickel concentration increased from 874 ppm initially to approximately 1000 ppm during the UF6 addition and the initial hydro=- gen treatment. This increase in Ni concentration of approximately 130 ppn is sufficient for the reduction of approximately 15% of the UF6 added . The concentration of oxide in the salt during the test was approximately 4000 ppm. In the absence of conclusive information on the reduction of uranium fluorides intermediate between UFh and UF6, reference can be made to data on materials having similar characteristics. The eguilibrium between UFM and UF, In molten mixtures of LiF and BeF2 has been studied by Long(5) 3 by oObserving the concentration of HF in H, in equilibrium with the salt. 2 Equilibrium was oObserved to have been established during the rise of H2 bubbles through a few inches of molten salt. The data indicated that reduction of UFh to UF, could be achleved over a wide range of operating 3 conditions. The reduction of NiFQ, CrFE, FeF_ to the metals by hydrogen is utile 2 ized for removal of these contaminants from molten salt. Tt was also observed that molten salts containing uranium fluoride with a valence> 5 13 (%) are gquite corrosive toward nickel metal. Since the reactions: 2UF. + Ni———“"QUFu + NiF > 2 I\I:'LF2 + HE‘——.'Ni + 2HF are known to occur in molten salts, it is felt that the reaction: UF5 + 1/2H2—-—'-UFh + HF will also occur. CONCLUSIONS AND RECOMMENDAT IONS The following conclusions have been made on the basis of the mater- ial presented in the preceding sections. 1. Urahium hexafluori&e is absorbed rapidly by molten fluoride salt containing approx. 1 wt % UF) at 600°C¢. Tt is velieved that the absorption results in the formation of an intermedi- ate fluoride of uranium. 2. It is likely that reduction of the intermediate fluoride to UFM can be accomplished by contact with hydrogen. | 3. TIncorporation of the proposed reduction step into the flowsheet for MSR processing results in a shorter and more direct process. It is recommended that experimental work on the reduction of UF6 to UFM in a molten salt environment be continued with emphasis in the follow- ing areas: 1. rate of reduction of intermediate uranium fluoride to UFu>by hydrogen. 2. corrosivity of molten fluoride melts containing intermediate uranium fluorides. 3. adaptation of the reduction process to continuous operation. 1L REFERENCES W. L. Carter and C. D. Scott, MSBR Fuel and Fertile Stream Pro- cessing. Preliminary Design and Evaluation of a Reactor Integrated Plant, ORNL~-3791 (in press ). H. C. Francke, Y-12 Plant, Union Carbide Corp., Oak Ridge, Tenn., Personal communications (1965). S. H. Smiley, D. C. Brater and R. H. Nemmo, Development of the Continuous Method for the Reduction of Uranium Hexafluoride with Hydrogen - Process Development - Cold Wall Reactor, K-1248(Del) (1959). S. S. Kirslis, personal communications (1965). G. Long, Stability of UF.,, (in preparation). 3) C. M. Blood, Solubility and Stability of Structural Metal Di- fluorides in Molten Fluoride Mixtures, ORNL CF 61-5-4 (1961). W. R. Laing, personal communication (1965). - o D H OCW -1 OV Fw - 13. 1Lk, 15, 16. 17. 18. 19. 20, o1, PO, o3, 2L . 5. 26, o7. QNS dOdrdg oS mddd YT EZ - . - F. Baumann E. Beall R. Bennett S. Bettis E. Blanco F. Blankenship . B. Briggs . BE. Brooksbank . B. Brown . H. Carr . L. Carter T. Cathers L. Culler, Jr. . E. Ferguson . M. Perris . C. Francke E. Goeller . R. Grimes E. Guthrie W. Horton L. Jolley R. Kasten S. Kirslis G. Kitts J. Kelly B. Lindauer Long 60. E. 61. D. 62. 0. 63. L. 6L . g, 65 . 0. 66 . R. 15 Internal Distribution 28. W. B. McDonald 29. H. F, McDuffie 30. L. E. McNeese 31. H. G. McPherson 32. S. Mamn 33. F. W. Miles 34. R. P. Milford 35. W. R. Musick 36. A. M. Perry 37. W. W. Pitt, Jr. 38. M. W. Rosenthal 39. A. D. Ryon 4LO. J. B. Ruch 41. C. E. Schilling 42. C. D. Scott 43. J. H. Shaffer hi. S. H. Smiley L5. R. E. Thoma L6, J. W. Ullman 47, ¢. M. Watson 48. M. E. Whatley Lo, R. G. Wymer 50. E. L. Youngblood 51. Patent Office H2-5%. Document Reference Section 5L-55., Central Research Library 56-57-58. Laboratory Records Department 59. Laborat?ry Records Department (R.C. External Distribution L. Anderson, Jr. (AEC, Washington) C. Davis (AEC-0RO) E. Dwyer (BNL) P. Hatch (BNL) Lawraski (ANL) Roth (AFC, Washington) C. Vogel (ANL) 67. Research and Development Division (AEC-ORQ) 6£8-82. DTIE