ORNL-3544 UC-4 — Chemistry TID-4500 (25th ed.) REDUCTION OF URANIUM HEXAFLUORIDE 2N / ” [ - o ol iR / Cind i : \s - - S P — RETENTION ON BEDS OF MAGNESIUM " 1L 2} FLUORIDE USED FOR REMOVAL OF TECHNETIUM HEXAFLUORIDE Sidney Katz OAK RIDGE NATIONAL LABORATORY operated by UNION CARBIDE CORPORATION for the U.S. ATOMIC ENERGY COMMISSION DISCLAIMER This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency Thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof. DISCLAIMER Portions of this document may be illegible In electronic image products. Images are produced from the best available original document. Printed in USA. Price: $0.50 Available from the Office of Technical Services U. S. Department of Commerce Washington 25, D. C. 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 representation, expressed or implied, with respect to the accuracy, completeness, or usefulness of the information contained in this report, or that the use of any information, apparatus, method, or process disclosed 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 report. As used in the above, ''person acting on behalf of the Commission'' includes any employee or contractor of the Commission, or employee of such contractor, to the extent that such employee or contractor of the Cemmission, or employee of such contractor prepares, disseminates, ur provides access to, any information pursuant to his employment or contract with the Commission, or his employment with such contractor. ORNL- 3544 Contract No. W-7405-eng-26 CHEMICAL TECHNOLOGY DIVISION Chemical Development Section B REDUCTION OF URANIUM HEXAFLUORIDE RETENTION ON BEDS OF MAGNESIUM FLUORIDE USED FOR REMOVAL OF TECHNETIUM HEXAFLUORIDE Sidney Katz DATE ISSUED JAN 3 1 1964 OAK RIDGE NATIONAL LABORATORY Qak Ridge, Tennessee operated by UNION CARBIDE CORPORATION for the U.S. ATOMIC ENMERGY COMMISSION THIS PAGE WAS INTENTIONALLY ~ LEFT BLANK CONTENTS Abstract I»nfroducfion Materials Experimental Work Test 1: Deleterious Effect of Grossly Inadequate Pretreatment of Magnesium Fluoride Pellets Test 2: Favorable Effect of Extensive Pretreatment of Magnesium Fluoride Pellets Test 3:. Importance of the Prefluorination Step in the Pretreatment of Magnesium Fluoride Pellets Test 4: Desorption of Uranium Hexafluoride from Well-Stabilized Magnesium Fluoride Test 5: Lack of Effect of Hydrogen Fluoride on Well-Stabilized Magnesium Fluoride Pellets Discussion References REDUCTION OF URANIUM HEXAFLUORIDE RETENTION ON BEDS OF MAGNESIUM FLUORIDE USED FOR REMOVAL OF TECHNETIUM HEXAFLUORIDE Sidney Katz ABSTRACT The excessive loss of uranium incurred when discarding magnesium fluoride, (the adsorber used to selectively remove technetium hexa- fluoride from uranium hexafluoride streams) is a problem common to all volatility processes for recovering enriched uranium fuels. As a result of the work described, two schemes for the release of the uranium hexa- fluoride from the magnesium fluoride and its separation from the tech- netium hexafluoride are proposed. One scheme depends on preferential thermal desorption of the uranium hexafluoride at 350°C and the other on selective adsorption of the uranium hexafluoride on sodium fluoride pellets following the codesorption of the two hexafluorides with fluorine at 500°C from the magnesium fluoride pellets. These proposals are aim- ed at reducing the amount of retained uranium to less than 1 g per 1000 g of discardable magnesium fluoride. ! In the work reported here, the deposition of uranium on magnesium fluoride as a function of heating, fluorination, and hydrogen fluoride pretreatment of the magnesium fluoride pellets prior to exposure to uranium hexafluoride was characterized in a series of gasometric studies. The dependence of the quantity of uranium hexafluoride adsorbed on pressure and temperature was also determined. The data show that physical adsorption is the mechanism far the deposition of most of the uranium hexafluoride on well-stabilized magnesium fluoride pellets. More than 90% of the adsorbate can be removed by heating to 350°C. Chemisorption (formation of a double salt) is probably not involved because of the small 0.05) mole ratio of UFg/MgF7 observed. . INTRODUCTION This report describes a gasometric study of the mechanisms of the undesirable deposition of uranium hexafluoride on magnesium fluoride and suggests two methods to reduce to acceptable amiounts the uranium loss on the discarded magnesium fluoride. The codeposition of uranium on magnesium fluoride beds that are used to selectively remove technetium hexafluoride from uranium hexafluoride streams is a problem common to all volatility processes for recovering enriched uranium from spent fuel elements. The magnitude of this codeposition is indicated from the ex~ perience in the Oak Ridge National Laboratory (ORNL) Volatility Pilot Plan’r,] in which 14 g of uranium was deposited on 1000 g of magnesium fluoride out of the 600 g of uranium passed through the bed as uranium hexafluoride. The extent of codeposition was somewhat less in a large-scale operation at the Paducah Gaseous Diffusion Plant,2 where massive quantities of uranium hexafluoride are passed through magnesium fluoride beds; 3.25 kg of uranium was recovered from 500 kg of the used magnesium fluoride. In the previous application of magnesium fluoride beds for the separation of technetium from uranium hexafluoride at the Paducah Caseous Diffusion Plant, the codeposition of uranium on the bed was of small concern because (1) the uranium was of low isotopic enrichment and represented only a small fraction of that which passed through the bed, and (2) the technetium recovery process dlso provided eco- nomical uranium recovery. However, in the ORNL volatility application, the iso- topic enrichment is high; the fraction of the throughput codeposited is greater; and the reprocessing costs are higher because of the fission product activity involved. Since in volatility applications, it is desirable to merely discard the used magnesium fluoride, the uranium that accompanies it must be held to an economic maximum (less than 1 g of uranium per 1000 g of magnesium fluoride). In the work reported here, the quantity and form of uranium deposited was studied as a function of a variety of pretreatments of the magnesium fluoride pellets. The pressure and temperature dependence of the amount of adsorbed uranium hexafluoride was also observed. The data showed that the uranium hexafluorlde is physically ad sorbed when well-stabilized magnesium fluoride is used. Also, the uranium hexa- fluoride can be desorbed to such an extent that the used magnesium fluoride can be economically discarded. MATERIALS Magnesium Flucride Pellets The "as-received" pellets, taken from the same batch used in the ORNL Volatility Pilot Plant, contained 10.7% water. They had been manufactured at the Paducah Gaseous Diffusion Plant to meet the requirements of their technetium trapping program. Similar pellets were reported to have a surface area of 111 m2/g after heating and purg- ing with fluorine.2 : In a preliminary examination of the pellets, the weight loss and surface area were determined for a number of possible pretreatments. The effect of heating the pellets for half an hour was tested at four temperatures until only 0.07% water remained. The data follows: Temperature (°C) Cumulative Wt Loss (%) Surface Area (m2/g) 160 10.0 102 260 13.2 80 360 16.6 35 460 - 17,5 20 From the original water content (10.7%) and the cumulative weight loss (17.5%), « calculation indicates that 52.3% of the water was converted to hydrogen fluoride during the heat treatment, “ The effect of a combination of heating at 160°C for a half hour followed by treating with fluorine at atmospheric pressure for 2 hr at 100°C resulted in a cumu- lative weight loss of 11.1% and a surface area of 89 m2/g. These data permit an estimate of the physical and chemical properties of the magnesium fluoride pellets as used in the tests that follow. EXPERIMENTAL WORK A gasometric sysfem3 was used in a series of five tests to determine (1) if in- adequate pretreatment of the magnesium fluoride could result in gasometrically measurable adsorption of uranium hexafluoride, (2) how much uranium hexafluoride would be adsorbed on well-stabilized magnesium fluoride, (3) the importance of the fluorination step in the pretreatment of magnesium fluoride, {4) the temperature de- pendence of the desorption of uranium hexafluoride from magnesium fluoride, and (5) whether hydrogen fluoride pretreatment of the magnesium fluoride influenced subsequent uranium hexafluoride adsorption. In each of the tests, after some specific pretreatment of the magnesium fluoride pellets, a gasometric measurement of uranium hexafluoride adsorption was made under the following conditions: 200 mm Hg pressure of uranium hexafluoride with the mag- nesium fluoride pellets at 100°C (deviations from these conditions are noted in specific cases). After the adsorption, the chemical form of the retained uranium was determined by chemical analysis and by gas evolution methods. The definitive chemical makeup of the magnesium fluoride pellet, itself, was deduced from chemical analysis and gasometric measurements. The data are presented with the description of each of the five tests and are sum- marized in Table 1. Test 1: Deleterious Effect of Grossly Inadequate Pretreatment of Magnesium Fluoride Pellets Part A: Pretreatment by Heating at 150°C The conditions and observations are listed below: Table 1. Adsorption of Uranium Hexafluoride on Magnesium Fluoride: Effects of Various Pretreatments Magnesium Fluoride Pellets Uranium Hexafluoride Magaesium Fluoride Pellet Residue Pretreatment Retained (millimoles) Wt % Uranium Final Ny Surface Test Wt (g) Heat Fo HF Gasometric® Anal.¢ Total U(Vl) Wt (g) Area (m2/g) 1A 0.631 150°C No No 0.1 125 to 2 hr 25°C 1B No No Yes 0.7 at 25°C 18.7 18.4 0.728 2 12.526 400°C 300°C No 0.95 0.64 1.40 1.37 10.877 16.5 reached 1 atm slowly 18 hr 3A 12.594 500°C No No 10.564 reached slowly 3B 400°C No No 1.23 0.68 1.44 1.01 10.805 15.2 1/2 hr 4 42,651 450°C 350°C No 3.92 0.12 0.05 35.625 17.0 ' 2 hr 1. atm 2 hr 5 253159 No 350°C Yes 2.20 0.23 2.05 25.320 17.6 1 atm 1 hr a . . . . Hydrogen fluoride treatment as used to activate sodium fluoride.3 Gasometric measurement with pressure of 250 mm Hg UFg in reactor; at 100°C unless nated otherwise. c » . s . M - . Remaining on the pellet residue after evacucting reactor at 100°C; calculated from urarium analysis. d_, . . I . This starting material is part of the pellet residue from run 4, Magnesium fluoride: 0.631 g of "as-received" pellets Pretreatment: Heated at 150°C for 2 hr, with pumping to about 1 mm Hg UF¢ adsorption:’ None detected gasometrically at 125°C to 25°C It was concluded that the limit of detection for the gasometric system (0.1 millimole) was too large to permit the measurement of the adsorption of uranium hexafluoride on a small sample to magnesium fluoride (10 millimoles) under these conditions. Part B: Effect of Excess Hydrogen Fluoride on Adsorption by Inadequately Pretreated Magnesium Fluoride Pellets The conditions and observations follow: Magnesium fluoride: Residue from part A Pretreatment: Exposed to hydrogen fluoride at atmospheric pressure at room temperature; removed excess gases by pumping to less than 1 mm Hg UF, adsorption: 0.7 millimole at 25°C, by gasometric measurement Desorption: Heated the pellets to 320°C, resulting in evolution of 1.2 millimoles of gases which were not UFy, as determined from condensation characteristics Solid residue: 0.728 g containing 18.7 wt % total U [18.4 wt % U(VI)]} The implications are that the adsorbed uranium hexafluoride had been converted to a nonvolatile oxyfluoride by reaction with water. Also, treating magnesium fluoride that contains water with hydrogen fluoride makes the water more readily available for reaction with adsorbed uranium hexafluoride. (It will be shown in test 5 that excess hydrogen fluoride does nol similarly offect.adsorption of uranium hexafluoride on well- stabilized magnesium fluoride.) Test 2: Favorable Effect of Extensive Pretreatment of Magnesium Fluoride Pellets Conditions and obscrvations were: Magnesium fluoride: 12.526 g of "as-received" pellets; larger sample taken to improve gasometric sensitivity Pretreatment: Heated slowly to 400°C; copious qudntifies of gas evolved,mostly below 200°C: fluorination for 18 hr at 300°C; fluorine pressure, 1 atm Solid residue: 10.877"g containing 1.40 wt % U [1.37 wt % U(VI)]; surface area, 16.5 m%/g Converting the results to a weight basis, about 14 g of uranium was retained as hexa- valent uranium per 1000 g of magnesium fluoride. Another 7 g uranium per 1000 g of magnesium fluoride had been adsorbed at 200 mm Hg pressure and desorbed upon pump- ing down to about 1 mm Hg pressure. Test 3: Importance of the Prefluorination Step in the Pretreatment of Magnesium Fluoride Pellets Conditions and observations for this test are shown below. Magnesium fluoride: 12,594 g of "as-received" pellets Pretreatment: ' Heated to 500°C slowly; 105 millimoles of gas evolved; the 105 millimoles of gas are estimated to weigh 2.03 g, assuming 52.3% of held water was converted to hydrogen fluoride; that weight agrees well with a measured weight loss of 2.03 g during pretreatment; sample was removed for that weight measurement UF¢ adsorption: Reheated to 400°C for half an hour, starting part B _ of this test; 1.23 millimoles measured gasometri- cally; after removing uranium hexafluoride in gas phase from reactor by pumping, only 0.68 millimole remained, as moasured by analysis of residue Residue: . 10.805 g containing 1.44 wt % total ° U[1.01 wt % U(VI)] surface area, 15.2 m2/g Only 4 g of uranium per 1000 g of magnesium fluoride was retained in a chemically re- duced form when prefluorination was omitted, that quantity may be lower if the adsorp- tion is performed in the presence of fluorine, as is done in the Volatility Pilot Plant at ORNL. This suggests that prefluorination of the magnesium fluoride may not be necessary. Test 4: Desorption of Uranium Hexafluoride from Well-=Stabilized Magnesium Fluoride In the desorption test, the conditions and cbservations were: ng.nesium fluoride: 42,651 g of "as-received" pellets Pretreatment: Heated at 400°C for 2 hr followed by fluorination for 2 hr at 350°C under fluorine at 1 atm UF ¢ adsorption: 3.92 millimoles by gasometric measurement; 2,25 millimoles estimated to have remained after removing uranium hexafluoride in gas phase from reactor by pumping UF¢ desorption: " The temperature was raised stepwise, holding each new temperature for half an hour Cumulative Desorption Temperature (°C) (millimoles) 160 0.40 220 1.36 345 2.89 420 3.28 480 >4.28 Residue: 35.625 g containing 0.12 wt % U[0.05 wt % U(VI)]; surface area, 17.0 m2/g It is significant that, of the uranium adsorbed on well-stabilized magnesium fluoride, most of the hexavalent uranium is readily desorbed; the chemically reduced uranium remaining as a residue represents less than 1 g of uranium per 1000 g of magnesium fluoride. Assuming that uranium hexafluoride was desorbed first in this test, a temper- ature of less than 350°C should be adequate for removing adsorbed uranium hexa- fluoride down to acceptable concentrations. The volatiles desorbed in excess of the uranium hexafluoride must have been residual compounds not previously removed, for example, water. . Test 5: Lack of Effect of Hydrogen Fluoride on Well~Stabilized Magnesium Fluoride Pellets The conditions and remarks are listed below. Magnesium fluoride: 25.315 g or residue from previous test Pretreatment: Refluorination for 1 hr at 350°C under 1 atm of Fgp; expesing to 1 atm of HF followed by pumping off excess, all at room temperature UF 4 adsorption: 2.20 millimoles, measured gasometrically UF, desorption: Residue raised to 350°C and evolved gases removed : by pumping Residue: 25.320 g containing 0.23 wt % U,[0.05 wt % U6+] and measuring 17.6 m2/g No appreciable retention of uranium was noted when well-stabilized magnesium fluo- ride was pretreated with excess hydrogen fluoride, in contrast to the results obtained in test 2b. DISCUSSION The uranium adsorbed after the exposure of rigorously pretreated magnesium fluoride to uranium hexafluoride at 100°C is largely hexavalent and can be removed by heating or pumping (see tests 2, 3, 4, and 5 in Table 1); therefore, the adsorbed uranium must be present as the hexafluoride, either adsorbed physically or in the form of a complex. Physical adsorption is the most probable mechanism, since the maximum quantity of uranium held is insufficient to yield a reasonable complex with the mag- nesium fluoride. Significantly, at 350°C, less than 1 g of the uranium per 1000 g of magnesium fluoride remains adsorbed. The drastic loss of surface area of the magnesium fluoride pellets (down to 15.2 ta 17.6 m2/g for the pellets in tests 2, 3, 4, and 5) represents primarily the cumulative sintering effects of exposure to heat. The quantities of uranium hexafluoride adsorbed or recovered in these tests and in ORNL pilot plant run R-8 and at Paducah3 are in sufficiently good agreement to indicate that the magnesium fluoride in the larger=- scale operations also undergo surface area reductions. Some of the volatile material associated with the pellets remains trapped even after heating them to over 400°C and after extensive fluorine treatment at 300°C (see test 4). The occluded volatile material, presumably a mixture of hydrogen fluoride and water, must be unavailable to the uranium hexafluoride since otherwise the water would react with the hexafluoride and prevent subsequent desorption of the uranium. ‘ Little uranium in a reduced valence state was found on the magnesium fluoride residues except where prefluorination had been omitted; in each case (tests 1 and 3) about 0.3 to 0.4% quadrivalent uranium waos present. This reduction may be accounted for by an equivalent fluorination of the nickel reactor or the tray upon which the pellets rested. CONCLUSIONS AND RECOMMENDATIONS Physical adsorption is responsible for most of the uranium adsorbed on well- stabilized magnesium fluoride pellets, and the uranium hexafluoride can be removed down to less than 1 g of uranium per 1000 g of magnesium fluoride by heating to 350°C. These two facts lead to two schemes for the release of the physically adsorbed uranium hexaflucride and its separation from technetium hexafluoride and provide a means of economically discarding used magnesium fluoride pellets. " The first scheme, which appears simplest to try and put into pilot-plant practice, is to heat the loaded pellet bed to about 350°C in order to preferentially release the uranium hexafluoride. According to the data of Golliher and co-workers,2 the technetium compound is poorly desorbed (18% at 1000°F in nitrogen). The alternative scheme is to release both the uranium and technetium hexa- fluorides from the loaded pellet bed by heating to 500°C in fluorine and thén to selectively adsorb the uranium hexafluoride on sodium fluoride at 100°C; Golliher ~ and co-workers2 found that only 4% of the technetium that passed through a sodium fluoride trap at 200°F was retained. Simplifying the pretreatment of the magnesium fluoride pellets might be considered also. A more rigorous preheating treatment may permit omission of the fluorination step. REFERENCES 1. Chemical Technology Division, Annual Progress Report, Period Ending May 31, 1963, ORNL-3452, p 26-50 (Sept. 20, 1963). 2. W. R. Golliher, R. A. LeDoux, S. Bernstein, and V. A. Smith, Separation of Technetium=99 from Uranium Hexafluoride, TID-18290 (1960). 3. S. Katz, A Gasometric Study of Solid-Gas Reactions, Sodium Fluoride with Hydrogen Fluoride and Uranium Hexafluoride, ORNL=-3497 (Oct. 15, 1963). THIS PAGE WAS INTENTIONALLY LEFT BLANK 1 ORNL-3544 UC-4 — Chemistry TID-4500 (25th ed.) INTERNAL DISTRIBUTION 1. Biology Library 51-52, Sidney Katz 2-4. Central Research Library 53. L. J. King 5. Reactor Division Library 54. C. E. Larson 6-7. ORNL — Y=12 Technical Library 55. R, B. Lindaver Document Reference Section 56, M. J. Skinner 8-42. Laboratory Records Department 57. S. H. Smiley (K=25) 43. Laboratory Records, ORNL R.C. 58. J. A. Swartout 44, R, E. Blanco 59. A. M. Weinberg 45. G. E. Boyd 60. M. E. Whatley 46. J. C. Bresee 61. P. H. Emmett (consultant) 47. W. H. Carr 62. J. J. Katz (consultant) 48. F. L. Culler 63. T. H. Pigford (consultant) 49. C. E. Guthrie 64, C. E. Winters (consultant) 50. H. L. Hemphill 65. 66. 67. 68, 69. /0. 71, 72, 73. 74, 75. 76. 77-665, EXTERNAL DISTRIBUTION E. L. Anderson, Atomic Energy Commission, Washington, D.C. H. Schneider, Atomic Energy Commission, Washington, D.C. H. M. Roth, Atomic Energy Commission, ORO L. P. Hatch, Brookhaven National Laboratory G. Strickland, Brookhaven National Laboratory O. E. Dwyer, Brookhaven National Laboratory R. H. Wiswall, Brookhaven National Laboratory R. C. Vogel, Argonne National Laboratory A. Jonke, Argonne National Laboratory J. Fischer, Argonne National Laboratory J. Schmets, CEN, Belgium Research and Development Division, AEC, ORO Given distribution as shown in TID-4500 (25th ed.) under Chemistry category (75 copies — OTY)