T YT el TR L TN AT G Ay T T [ T W g R UNCLASSIFIED OAK RIDGE NATIONAL LABORATORY * EXTERNAL TRANSMITTAL : Operated By L AUTHORIZED UNION CARBIDE NUCLEAR COMPANY B 0 RN |_ ~ POST OFFICE BOX X . . CENTRAL F".ES NUMBER OAK RIDGE, TENNESSEE | | 59-2-61 ; | - ~ Second Issue' . DATE: - April 1, 1959 L - copyNno, #O SUBJECT: n Processing of lfblten Salt Power 'Reactor Fuel TO: o Distribution | - FROM: - Do Campbell and G. I. Cathers ,ABS‘J!RACT o -Fuel reprocessing methods are being investiga.ted for molten salt nuclear . reactors which use LiF-BeFp salt as & solvent for UFy and ThF). A liquid HF - dissolution- procednr_e coupled with fluorination has been developed for re- covery of the uranium and IiF-Be %&:ggent salt whieh 1s highly enriched in Li-T. The recovered selt is decon ted in the process from the major . reactor poisons; namely, rare earths and neptunium.” A brief investigation ‘of elternate methods, including oxide. precipita.tion, partisl freezing, and metal reduction, indicated that such methods may give some geparation of the solvent salt from reactor poisons » but they do.not appear to be suffi- ,ciently quantitative for a simple processing @eretion. ' o Solubilities of IiF and seF2 in eq_ueaus 70-100% EF are presented. The _ 'BeF2 solubility is appreciebly increased- in.the presence of water and large ‘emounts of IiF, Salt solubilitiés of 150 g/liter ere attainsble. = Tracer experiments indicate that rsre earth solubilitie -fi’ ‘relative to LiF-BeFa . solvent salt solubility, increase from {e.“bout 10 I " mole % in BO% HF. le%ingafisfi'tooooj Fluorins.tion of uranium i‘rom LiF-BeF se.lt has ‘been d.emonstrated. '.Ehis | 'eppesrs feasible also for the recovery of the relatively swall concentration - of uranium pro@ced in the IiF*BeFQ-ThFl, ‘blanket. R A proposed chemical flowsheet is presented on the ’oasis of this explora.tory N _work as e.pplied to the semicontinwl&'ocessing of 8 600 Mir power reactor. This document contains !nformofion of a prelummoty n* i&'ermetion is not to be abstracted, nature and was prepared primarily for intemo! use teprintad or otherwise given public disceminatinn -at the Ock Ridge National Laboratory. It is subject githout the approva! of the ORNL patent branch, to revision or -correction and therefore dou not 'Qpre“n” fina! rep m - Legal aps 1nfermahon Control Department. uncmssmw . LEGAL NOTICE This report was prepared as an gccount 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 respsct 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 respact 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 smployes of such contractor, to the extent that such smployee or - contractor of the Commission, or employee of such contractor prepares, disseminates, or provides access to, any informahon -pursuvant to his employment or contract with the Commission, or his omploym-nt with luch contractor., ® " type’reactor would be 1deally & Ii . economy. This materiel would act as a solvent or cerrier for the fluorides of the .;2... IN‘]IRODUCTION - High temperature fluid fuel reac‘bors Esing molten fluorides ‘have been pro-. posed for the production of nucleefi The core or blanket salt in this F-BeFo mixture for optimum neutron moderation and fissile or fertile elements, uranium, plutonium and thorium. The feasibility and ~ economic Justification of such & reactor system. depends on fuel processing, i.e., a processing method that will"maintain the degired neutron economy -and reactor opera= bility at a reasona'ble cost. This paper presents & description of a new chemical process for this type of fluoride fuel based on two principles, namely, volatiliza- tdon of UF; from the reactor salt by fludrination and recovery of the IiTF-BeFp solvent salt for reuse by a HF dissolution process. fi_ MOLTEN SALT REACT(B DESCRIPTION The Afrcraft Resctor Experiment, in vhich MaF-ZrF PR, fuel circulated through inconel tubing in BeO moderator , demenstrated. ’bhe basie feasibllity of a - high temperature molten salt reactor. Detailed calculetions for & power reactor ' ha. e been published on 600 Mw heat two-region homogenéous machines using 63 mole % 11!/F-=37 mole % BeFp with up to 1 mole % UF) end Th¥) 85 & core salt and 71 mole % I-i7F--l6 mole % BeFp=-13 mole % ThF) as a blanket £81t2. The reactor design used as reference in this paper has a uranium (U233 or 0235) inventory of 600-1000 kg (verying with time) with en 8-ft-dis care snd & total fuel volume of 530 £t3 (334 f£t3 external to the core) The core salt weight is approximately 75 ,000 1b, the blanket 150,000 1b. | - In the reference rea.ctor et 1east 90% of the power is produced. in the core, yield- ing about 180 kg of fission products per year. After operation for one year without ~ processing (exce t inert gas removal), the fission products ebsorb sbout 3.8% of all neutrons and UP3° and 1233 absorb sbout’ 3. 9%, If the reactor fuel is processed for fission product removal et the rate of one -fuel volume Eer %r » 05 years, the fission products would ebsorb 2. % of the neutrons; 23 uge 3% yould ab- sorb 10.4% (mostly 1P: 36); ; end Np 37T about-0.9%. Without fuel processing neutron absorption by fission products would continue. to increase elmost linearly with time, exceeding 10% after 10 years; Np237 would build up at an accelerating rate, absorbing - about 3% of -the neutrons after 10 yea.rs ; the amount of fissionable materisl required to keep the reactor criticsl would. increa.se 'by a.‘bout 200 kg/yea.r 3 and the conversion ‘ratio would deereaee ma.rkedly. : Even-num‘bered urenium ieotopes j pe.rticula.rly 023_6, are the worst poisons y but thelr removal is beyond the scope of chemical reprocessing. Of the 180 kg of fission prod.ucts per yeer 22 atom % with he.lf-lives of more than 78 min ere 7_ lR c. Briant, A. M. Wein‘berg, et a.l, ."'.'l’.he Aircreft Eeactor Experiment " Nucl. Sci. and Eng., , 797-853 (1957) EJ. A. Isne, H. G. MacPuerson and F. Maslan, "Fluid Fuel Reactors,” - Addison-Wesley Publishing Co., Inc., 1958 -3 - sub,jeét to'removal from the reacter es rare gases; these would contribute 26% ~of the fission product pdisoning for 100 ev neutrons. About 26 atom % of the long-lived fission products are rare earths, which contribute 40% of the total fission product poisoning. The rest of the fission products consist of a wide variety of elements, no one of which is outstan from the nuclesr poisoning point of view. After reasonably long operation Np 37 is the worst individual poison other than the rare earths. - FLU%IDE VOLATILITI PROCESS FOR MSR FUEL - Fluoride volatilization processing for uranium recovery eppears feasible for molten salt reactor (MSR) fuel on the basis of lsboratory studies. It is based on direct fluorination,of the fuel salt to convert UF) to UFg with atten- dant volatilization and recovery. Similar volatility processes have been pro- posed and d.gve%oged for zirconium alloy reactor fuel elements after dissolution in fused salt. One of these, the ORNL Volatility Process, was successfully used for recovery and decontamination of uranium from the NaF-Zth-UFh salt fuel of the Aircraft Reactor Experiment.’ The MSR volatilization process would differ, however; from other volatility processes in that complete decontamina~ tion of the product UF. would not be essential, since it could be remotely re- duced to E’I*"l+ and reconstituted :l.nto reactor salt. A series of smallsscale fluorinations was carried out with a 48 mole % LiF--52 mole % BeF, eutectic mixture conta.ining about 0.8 mole % UF),. (MSR fuel would contsin 0.25 to 1.0 mole % UF, , depending on the operating time. ) The eutectic salt was used instead of the fuel salt In order. to invsstigate lower temperature operation. In fluorinations at 450, 500, and 550°C, the rate of uranium removal increased with the -temperature ('l'able 1). The thorimn-containing 'bla.nket salt cannot be processed for uranium re- covery at as low a temperature as that used to process the fuel salt. 3e. I. cathers, "Urantun Recovery for’ Spent Fuel, "Nucl Sei. a.nd Eng 168-777 (1957). o, "H. H. ‘Hymen, R. C. Vogel end J. J: Ka:bz, "Proceed.ings of Internat:!.onal Con- ‘ference on Peaceful Uses of Atmnic Energy, _ Vol_ 9, pp., 613-626 _United. Nations ’ Ncw Yerk (1956) | - o 5G. I. Cathers et al., "Reccvery of Ureniun from H:lghly Irradiated Fuel," United Nations Paper 535, 2nd. International Cmference on Peaceful Uses of Atomic Energy (1958) | Table 1. Effect of Fluorination Temperature on the Fluorination of Urenium from LiF-BeF, (40-52 mole %) - Fluorination Uranium in Salt after Treatment, wt % Time, hr At 450 At 50006 At 550°C 0 3,39 5.0 b.91 0.5 ) 1.96 0.20 - 0.55 1.0 | 0.39 0.17 0.20 1.5 0.2 0.12 0.06 2.5 0.32 0.11 0.05 No induction period before uranium evolution. 5 wt$ added; some of the uranium probably precipitated as oxide. o/ - aqueous HF solutions indicated thet both materials are solub extent in solutions containing 70 to 90 wt % HF (Ta'bles 2 ana. 3). In general, -5 . The uranium concentration in the blanket, however ’ 15 very low; it has been esti- mated that with continuous processing at the rate of one blanket volume per year, the blanket salt (IiF-BeF. ~ThF) , T1<16-13 mole %) will contain approximately 0.004 mole % UF, (140 ppm) after one year and 0.01k mole % UF) after 20 years. Fluorinations o% +wo such mixtures at 600°C for 99 min -gave uranium concentra- tions in the salt of 1-2 ppm, the lowest urentum concentrations ever obtained “in fused salt laboratory fluorinstions. Over 90% of the uranium was removed in 15 min. It is concluded, ‘therefore, that fluorination of uranium from blanket salt can be accomplished. , The behavior -of protactinimn in the ‘bla.nke'b galt during rluorination is of | interest, although the protactinium is not lost, in any case, since the salt is returned “to the reactor. A IAF-BeF,-ThF %a 71-16~13 mole %) mixture containing sufficient irradiated thorium to give # FaR33 concentration of 5.5 x 10-9 g per gran of salt was fluorinated for 150 min at 600°C, there was no measurable de- crease in protactinium activity in the salt. Protactinium volatilization in the process-seems to be unlikely. ~ However, the protactinium concentration in the blanket of the referenee d.es:lgn reactor is higher (~10 & per gram of salt). 14TF-BeF,, SALT RECOVERY WITH EF Experimental work has démonstra'hed that the LiF-BeFa salt can be pro=- cessed by dissolution in aphydrous or nearly anhydrous liquid hydrogen fluoride. ' Decomtemination of the IiF-BeF, salt from the major neutron poisons, rare earths ~ and neptunium, is schieved due“to the relative insolubility of the fluorides of these elements In such solutlons. The LiF-BeF2 salt is recoverable from the HF solution by evaporation. It _is well krviown that BeF, is very scluble and IiF is rather inéoluble in ‘water in contrast to liquid HF where the reverse is true. In drous HF the polyvalent element fluorides generally exhibit low solubilities.® Initial con- sideration of the problem suggested thet use of aqueous HF (greater then 80% HF) would give sufftcient solubility of both I4F and BeFo to meet process objectives. The solubllity studies were-therefore carried out over the range of 70-100% HF. However, in addition to the mixed solvent -effect there existed also the possi- bility that enhydrous HF would be suiteble as a solvent for the LiF-BeF, salt .complex in the shalogous,_sense that cryolite y Na.A1F., is quite soluble whereas is insoluble in HF. 7 Some emphasis was thefefore pleced on using "anhydrous™ - %fi:h the material being obtained by -vepor transfer from & cmnmercia.l tank | (nominally conteining less then 0. 1% water ). Initia.l mcasmements of the solu'bilities of I.:LF a.nd BeF. E; separe.tely, to an apprecisble \6A W. Jache and G. w. Ca.dy, J. Phye cmm., 56, 1106-1109 (195a) 71.. F. Mudrieth and J. K?.einberg ’ "Non-Aqueous Solven'bs ’ Chap'ber 10 ’ John Wiley and Sons, Inc., New York (1953). Fa { LiF-BéE‘_, SALT A ————————— U,Th,FP FROM REACTOR FLUORINATOR et JF. SOLVENT © a UNCLASSIFIED ORNL-LR-DWG 28749A CONDENSER ~ 450°C " SALT HF, H,0 VAPORS Th,FP | - ¥ DISSOLVER | > 90% HF,< 10% H,0 32°¢C ~10% SALT _ SOLUTION IN HF - H20 _+Th,FPSOLIDS CONTINUOUS BATCH SALT SOLUTION SOLID - LIQuID SEPARATOR TH, FP SOLIDS SOME SALT AND SOLVENT ' WASTE EVAPORATOR HF Ha0 Th RECOVERY fig. 2. Tenkotive Flowshest for Fluoride Volatility and HF Disolution Procensing of Molten=Salt Reactor Fuel. ~ EVAPORATOR FLASH 100-400°C " | MOLTEN .} SALT | FUEL MAKEUP AND PURIFICATION e — e — | UE,ThE a4 LiF - BeF,=Thf,-UF, TO REACTOR "G Table 2, Solubility of I4F in Aqueous EF Solutions LiF 1n Solution, mg/g of solution Tempg:(;ame, -0 6 | 96.2 o 12 ' 110.7 31.7 58.h 7h.8 88.2 62x 20.6 | Ly * | 41.0 37* | 62.8 80.4 32.5% : , | | - 9.4 *: Reflux temperature. Teble 3. Solubility of BeFp in Agueous EF Solution ‘BeF, in Solution, mg/g of solution Temperature, 2.8. 8.3 0.8 95.0 100 o wiE wt3EF w$EF wHE wt b HF 12 5.8 . 26.3 9.2 2.8 o012 ..7 - IAF vas more soluble at higher temperatures if water was present in the solvent, ‘but the effect of temperature on BeFp solubility was not definitely established. The 14F solubility decreased rapidly as water was added to anhydrous HF, end the BeFa solubility increased from near zero; the golubllities were roughly the geme 1n°80 wt % HF, 25 to 30 g/kg. The BeFp was glassy in pature and digsolved glowly. The solubility values reported for IiF in Taeble 2, except at 12°¢C, were obtained after refluxing HF over the salt for 3 hr. No further messurements were made with I4F or BeF, alone, since the solubilities of the two -components together were of primary ortence. - _ < T The solubility of ‘BeF,, was also measured ot verious HF concentrations snd . temperatures in the presence of IiF. The resulte generally confirmed that the presence of IiF as well as water increases the BeF, solubility. ‘A plot of solubility velues at one temperature (12°C) illustrates both effects (Fig. 1). These end data at other temperatures are presented in Webles L4, 5, and 6. The favoreble effect of water on BeF, solubility is shown particularly in 70 and 80% HF solutions (Teble L4). BeF, solubllity was cnhanced further when 70, 80, and 90 wt % HF solutlons were safurated with IiF (Teble 5). Because of the iicrease in BeF, solubility at higher HF concentrations vhen IiF is present, the 90-100% BF range was studied further (Teble 6), The solubilities in epproxi- mately 100% EF eppear sufficiently high for proeess purposes. | The resulte presented here must be considered preliminary. Beryllium golubilities, in particular, may be generally higher then indicated because of analyticel problems. Difficulty wes also encountered in determining accu- rately the water content in highly concentrated HF, This was partly e problem ‘ of sempling the volatile solutions end partly the result of determining the emall emount of water by the differemp of the sample weight and the total ‘weight of I4F, BeFp, and HF. = | - | 86lubility measurements were made also on 63-37 mple $ LiF-BeFo slt (48-52 wt %) in the 80 to 98 wi & HF range {fable 7). This salt contained “sbout 0.1 mole % ZrFly 0.2 mole % mixed yave earth fluorides (Lindsay Code 370), end trace fission products (see following section). The salt was crushed but not ground -to & powder; average particles were flakes about 10 mils thick and 50 to 100 mils across. The solutions were sampled 15 and 60 min after salt addition to permit an estimste of the rate of dissolution. The results indi- " cated that an apprecisble concentretion is reached repidly, ‘but that the solu- tions ere not saturated, especielly with respect to BeFo, in less then several @ays. They slso demonmstrate that the fused salt behaves similarly to the two ~ components added individually. - | ‘ | o : . FISSION FPRODUCT DECONTAMINATION IN THE HF SOLUTION FROCESS The sgueous HF solutions-:of'fihe aai?b' from the preceding experiment were f1ltered through sintered nickel, end rediochemical enalyses were made to de- termine the fission product solubilities or decontamination effect. The resulte (Teble 8) show that rere earths, as represented by cerium, were relatively in- goluble in the HF solution and were therefore effectively separated from the salt. {10 | l | ® © LiF / 100 —— o BeF, % — _ ‘ .l - LUF _7" SOLUBILITY | 80 Q\\\ - 7 20 __ SATURATED | / \ IN LiF / 60 \\ . 50 - .\‘7’\ _ | ( 30 o - )< | N - .. ) J) A \ _ 20_ | | / - - \ N LiF(_ng’/kg) 10 NO LiF | ol L] 65. 70 . 75 80 SALT SOLUBILITY (mg /g of solvent) UNCLASSIFIED ORNL—-LR—DWG 34548 HF CONCENTRATION IN 'SOLVENT (wt %) Fig. 1. Effect of LIF on BeFy Solubility in HF-H,O Solutions at 12°C. -7a- - B Teble 4. Solubility of BeFp in Aqueous HF Containing IAF BeF,_in Solution, mg/g of solution -—-268 . Tempera.ture s - I4F Added, .6 " T19.5 L 090.0 °c ng/g of solvent fi vt $EF wt % HF 2 7.0 . | 12 | 15.0 | - 2.8 15.6 «60 . T.0 65 .8 -60 15.0 . 38.2 16.5 ’ ~ Table 5. Solubility of I4F and BeF» in Aqueous EF -Saturated-with Both Szalts | | _ __Salt in Solution, ng/g of snlution o Temperature, 65.6 wt ¢ HF 79.5 wt ¢ HF '90.0 wt % HF % __ . LF BeF, = LIF EeR, LF BeF, 12 129 82 29.9 53.8 6h2 u8.2 %60 . 85 6.8 226 5k2 k.0 58.2 Table 6. Tgmpsra.turg ’ C o =60 k3 Lox 36% 33*% * -@- Solu'bfl:ity o:t’ L:I.F and BeF,, in Solvents COnta.in:Lng @95t0100vt"%HF Salt in Solution, me /g of solution 555wt b EF 5wt GEF OBwh 4 BF 100 wb & BF LF BeF, LF BeF, WF peF, DF . _BeR, 60 I 78 27 92 26 107 28 50 2k, 60 33 62 3 90 30 69 e o 87 66 | 98 46 105 Lo *Reflux ‘i:‘empgrature . | Té,ble T Time 15 min 1w st 5hr ,aohr’ | Solubility in Aqueous HF Solution of IiF and BeF from Salt Mixture COnt&:[ning 63 mole % I4F and 37 mole 9 BeF, Salt in Solution, ng/e of selution ,.'Tempgmture, zggwt@m* gzwtfifli‘ 05 vt 5.EF OB vE BT . - Og BeE, LIF I4F BeF, LF 12 -~ 60 2 28 . 33 50 ;un % 17 28 29 ..~ 3 51 50 .43 17 22 3 58 6 wm 32 68% T9 B2x 63 28 Tox 22 21 65 Th 102 70 36 T 22 | ,3u; 100‘7 % % 88* 53 j - 6ox 22 BeF, -k 3 Essentially all the cmponent mdicate& had dissolve& a.nd therefore the ‘80lubility may be higher than the velue given. Insufficient salt was - added to some of the golutions to ensure an excess of both components. mu 8. Fission Product Solubilitiea in Aqueous HF Solutions L:lF-BeF (63-37 mole- %) + ~0.2 mole % rare e&rth fluarides + trace fission products between 1 and. 2 years old “Activity in Qrigina.l Salt and in EF Solution, : counts /min of. selt “Fission 01‘181118-1 79.5 9.5 95 , - 98 Product* Salt ot fimv ) wtfiBI' wfifim‘ . wt % HF Gross B 75 x 10° 230 x 10" 200 x 10" 225 x 10 250 x 10 Grossy 208 293 213 ol 282 Csy - 174 251 196 223 " 258 Srp 104 73 67 72 75 IRE £ 650 97 2, oL 101 " Cé P 510 7.9 3.6 1.3 0.28 - ¥Pp 105 73 66 74 81.5 Zr and ¥b precipitated from molten salt 'before this experiment end ‘were not present- in smifica.nt concentration. -3 The solubllity decreased as the .HF concentration was increased. The total rare earth (TRE) and trivalent rare earth (except ceiium) analyses do not show this separation because of the presence of the y'btrium daughter of strontium. No rare earth activity other than cerium (and yttrium) was detected in the HF solutions. The slight apparent decontamination from strontium is not understood; strontium is expected to be fairly soluble in these solutions. Cesium is known to be soluble ’ end all cesium in the sa.lt added apparently dissolved. Thus the rare earths y &5 represented. by cerium, ere removed from I.:!.‘r“-:BeF2 salts by dissolution of the salt in an aqueous HF golution, and strontium and cesium are not. The rare earth solubtlity ih these HF solutions saturated with IAF and BeF,, increased from about 107* mole % in 98 wt % HF to 0.003 mole % in 80 wt % HF,abaaed en the amount of IiF + BeF, diesolved. Reactor fuel with = l-year fuel cycle will contain ra_;'e earths % & concentration of sbout 0.05 moled. SOLUBILITIES OF HEAVY ELEMENTS IN EF | Keptunium is the most 'serigg qficlear poison other then the rare eartbs in e molten salt reactor, burning U= ‘_ The neptunium solubility in 80 to 100% HF saturated with LiF and BeF, waes found sufficiently low to permit its removal-along with the rare eflths (Teble 9). For these determinations smell quantities of neptunium (equeous Fp™* nitrate solu- tion) were added incrementally to HF solutions saturated with IiF end BeF,. The concentration of nitrate so added was ~0.02% of the fluoride concentration. Solu- bilities are reported in mg Np per g solution end in mole % of neptunium relative to dissolved ]'..i‘fi‘-BeF2 salt. The neptunium was determined 'by alpha counting wi'bh pulse enalysis to dis- tinguish neptunium from plutonium activity present as an impurity. The plutonium appeered to be cerried with the neptunium to a considerable extent; the ratio of the amount of plutonium in solution to the emount undissolved was within a fec- _tor of 2 of the ratio for neptunium a.lthougb*the plutonium concentration was ' sma.ller by a factor of 500. Addition of 1ron e.nd nickel metal to the HF solutions resulted in a signi- : ficant rednction in the gross o ectivity, probably as the result of reduction to Kp(III) “end Pu(III) ‘This has not yet been verified by pulse analysis. The: trivalent state might be expected to behave in somewhat the seme way &s ‘the rare earths; the observed solubilities are in the same renge &s those re< portéd previously for rare earths. It is expected that the rare earths, nep< " tunium, plutonium, end possibly uranium will behave as a single group end therefore exhibit lower solubilities when present toge'bher than the values reported here for’ the separate componen'bs . - Mea.surements of the solubilities of thoritm a.nd m.'a.nium fluorides in 90 to 100% HF indicate .that urenium and thorium ere relatively insoluble. All | ThF), d.e';eminations showed less than 0.03 mg of thorium per grem of solution, u«., T e e, Table 9. Np(IV) Solubility in I4F-BeF, Baturated Aqueous HF Solubility in Eolvent, mg/g 80 ~50 ~of 100 14F 2 & %6 12 BeF,, 11 70. 60 Lo I@ ‘ 00026 : otoll . 000086 0.0029 Np solubility relative to E . : - ' salt, mole % 0.0031 0.0012 0.00072 0.00024 gt S -13- the 1limit of detection. Better analyticel methods are availsble for UF), and solubilities were generally in the range 0.005~0.010 mg of uranium per gram of solution. In the presence of fresh iron or nickel metal solu- bilities were somewhat lower, 0.002~0.005 mg/g, indiceting perhaps & higher golubility for higher oxidation states. Relative to dissolved salt these solubilities are the order of 0.001 mole %. | SOLUBILITY OF CORROSION PRODUCT FLUGRIDES IN HF SOLVENT Measurements were made of the solubilities of corrosion products fluorides (iron, chromium, and nickel) in HF solutions saturated with 1iF. Chromium fluoride was relatively soluble, with values of about 8, 12, and 18 mg of chramium per gram of solution in 100, 95, and 90 wt % HF, respectively. Iron and nickel fluorides were less soluble. Measurements of the iron fluoride golubility varied from 0.08 to 2 mg iron per grem of solution; the average and average deviation neglecting extreme values:was 0.1% + 0.10. Measurements of the nickel fluoride solubility varied from. 0.01.to 1.3 mg of nickel per gram of solution, the everage being 0.15 + 0.06. The results were so incensistent that no trend with HF concentra- tion could be established. When iron, nickel and chromium fluorides were - present together, the solubilities of 2ll appeared to be somewhat lower. ' Similer measurements in90 and 100% EF (no dissolved salt) fell in the same range, but in 80 wt % HF the iron end nickel solubilities appeared to be significantly higher. - _ : _ RECOVERY OF SALT FROM SOLUTION The IiF end BeF, salt can be recovered from the HF solutions by evaporation. There gs some question of hydrolysis of the salt in re- covery, the result particularly of the existence of a high-boiling azeotrope at 38 wt % HF, which would be obtained if somé water was present initielly end if flash evaporation was not used. A°90% HF solution.satureted with IiF and BeF, was slowly eveporated to dryness end heeted to 450°C. X-ray d1ffraction indicated thet the resulting salt wvas about 90% 2IiFeBeF and 10% BeF,, and petrographic examination - indicated Yess than 3% of-materisl other than these fluorides. A se= . cond sample was eveporated after_eddition ef & little NELF (which would decompose and perhaps hydreflnorinste the selt at elevated tempéretures) .and fused at T00°C. This s<_sappeared to be entirely the binary com- - pound, and the x-ray pettern ¥as somevhat cleaner than in the first case, * Howevery hydrolysis did not eppeer to be a significent factor in any event. : e e - - PROCESS FLOWSHEET FOR REACTCR PROCESSING A tentative flowsheet for spplicetion of the fluoride volatility ~ and HF dissolution processes to molten galt reactor fluids has been prepared (Fig. 2). In this system the uranium is separated from the - o L g T '8 -1h- -salt a8 UF, befpre mv d.issolution of the salt, although the reverse might be feasible in some circunstences.. The liF«BeF, galt is then dissolved in _concentrated EF (>90% HF) for separation prifisrily from the rare earth and neptunium neutron poisons. - The sa.lt is re~formed in an evaporation , step, from which it would proceed to a final mekeup snd pm'if:l.cation step. The latter would perbaps involve the H, end HF treatment now believed necessery for all galt used in & reactor. The UFg prodnced in the vola- t1lity procese would be converted to UF, for reuse in salt by hydrogen reduction, or alternately it could possibly be reduced in situ in the ~ fused salt by essentially the same methiod. Although the volatility process achieves high decontamination, and the dissolution process leads ‘to thec@limination of the rare earths and neptunium activities, the salt recyecle would require shielding and xemote operation. ‘.me scale of the process vith a_l-year cycle, relative to the reactor, is indicated in Fig. 3. The d=21ly core salt processing rate would be about 125 kg containing ebout 4 kg of uranium., Assuming a 10% solubility for LiF-BeF, in the HF solution the solution processing rate would be less than one-liter per te for the reference reactor (260 Mir electric). The d=ily blanket proceseing rate would be about 250 kg containing only about 0.1 kg of UP33 vhich would be recovered by the fluo- ride volatility process. The blanket salt would then be returned to the reactor, after treatment-to remove corrosion products, if necessary. In- - frequent processing of the blenket salt to remove fission products would be required since at most a2 few percent of the fissions occur in the blanket, Although more develppment of some of the process chemistry is certainly needed, the principal features of the procese sppear to be sulted to the objectives for processing molten salt reactor fuel, Mich more work will be needed for developing chemical flowsheets for s;pecific reactor sys'bems | EIGH-m{PERAME mocn:ssmc- | ' \ _ The direct ranoval of rare earths rrom mwolten fluorides was in~ vestigated triefly using two methods=--oxide precipitation and partial - salt crystallization. 'These results were geperslly not fevorable end were not pursued further. -An alternative approach which sppears to have considerable premise consists of direct equilibration of g salt xwith & bed of ce‘f‘s ‘bo remnrve the wm'se po:lson&, the ra.re earths © Oxide precipitation, achieved by small additions to the fused salt . of water or Ca0, was tried with 5050 mole % NaF-ZrFj, and 11.5-k2-46.5 - mole % MaF-KF«LiF gince it had not been determined et the time that ~ I4F-BeF. salt would definitely be preferreds Trace fission products - were edded, Oxide addition results in precipitetion of primarily - zirconim in the case of zirconium-bearing selt without too mch effect J. H. Sehaeffer, N. V.&ni'fih, R. A. Sfi'ehlow, v. P, Ward ana. G. M. Watson, "High Temperature Processing of Molten' Fluoride Rea.ctor Fuels.' Im press = Chemical Engineering. : -e- | VOLATILITY]| PROCESS 1—4 kg U/DAY