OAK RIDGE NATIONAL LABORATORY operated by UNION CARBIDE CORPORATION for the U.S. ATOMIC ENERGY COMMISSION ORNL- TM- 1543 £y COPY NO. - DATE - June 1, 1966 REMOVAL OF PROTACTINIUM FROM MOLTEN FLUORIDE BREEDER BLANKET MIXTURES T . J‘R; -y Morde miaad &AL N0 Y “: C. J. Barton and H. H. Stone Iic,$’€:¢?;fifl€j§5?. ABSTRACT Three types of experiments involving removal of protactinium from molten fluoride breeder blanket mixtures were performed. The first showed that addition of thorium oxide to an LiF-BeF,-ThF, mixture containing a tracer concentration of 233Pa (less than 0.1 ppb) precipitated the protactinium and that treatment of the . mixture with HF redissolved the protactinium, In another experi- ment equilibration of molten LiF-ThF, containing 26 ppm of 231 Pa with a lead-thorium alloy resulted in removal of 99% of the pro- e tactinium from the salt phase but only a fraction of the reduced protactinium was found in the molten metal. A small amount of protactinium was transferred from the molten metal to a salt mixture by hydrofluorination treatment. 1In the third type of \) experiment, which was performed in both nickel and copper con- tainers, exposure of solid thorium metal to molten LiF-ThF, containing 20 to 30 ppm of 23!Pa precipitated 81 to 98% of the protactinium. More than half of the reduced protactinium was found in the unfiltered salt mixture, probably associated with small metal particles produced by the preliminary HF treatment of the mixture followed by hydrogen reduction. Again, hydro- fluorination of the mixture redissolved the protactinium. Efforts are continuing to find container materials to improve the reten- tion of protactinium in molten lead or bismuth. € RELEASED FOR ANNOUNCEMENT 1IN NUCLEAR SCIENCE ABSTRACTS NOTICE This document contains information of a preliminary nature oand was prepared primarily for internal use at the Ock Ridge Nafional Laboratory. It is subject to revision or correction and therefore does not represent a final report. The information is not to be abstracted, reprinted or otherwise given public dis- semination without the approval of the ORNL patent branch, Legal and Infor- mation Control Department, T e T - LEGAL NOTICE This report wos prepared as an account of Government sponsored worl-t; . Neither the United Stotes, _ . nor the Commission, nor any person acting on behalf of the Commission: B. Assumes any liobilities with relpec? to th- use of or for dnmugos uw!fing lrum the use of 7 A. Makes ony warranty or reprasentation, sxpressed or implisd, with respect to the accuracy, completeness, or usefulness of the informction contained in this report, or that the use of " any information, opparatus, methed, or proccn disclosed in this roport may not infringe privately owned rights; or any information, apparatus, method, or process disclosed in this report, ZA: vsed in the above, '‘person acting -on behalf of the Commission’ inc|udas any ornploy.c - contractor of the Commission, or .mployoo ‘of such controctor, to the extent that such nmp!oyc; ‘provides access to, any information pursuant to his employment or contract \vifh the Commission, or centractor of tho Commission, ‘or ompioyoc of such contractor prepares, dlss-minutos, or “or his employment with such con?rnctof. : . BkWMNARE INWVIILE This report was prepared a& an account of Gov ‘ erament sponsored work. Neithe: ; at_ate:. n’:r the Commission, nor any person acting on behalf of the Commission: " the Unlted - Makes any warranty or representation, expreased or implied, with respect to the acou- o RELEASED FOR ANNOUNCEMENT * such employeé or contractor of the Commiasion, or employee of IN NUCLEAR SC IENCE A BSTRACTS ' . disseminates, or provides 2ccess to, any information pursuant o his employment or contract ' with the qu_nmission. or his employment with such contractor, . A protactinium isotope, 233Pa (half-life 27 days) , ié an intermediate product in the conversion of 222Th to fission- able ?33U, It has been recognized for some time that it would be desirable to have a process for removing 233pa from fluoride breeder blanket mixtures used in a two-region molten fluoride 1 demonstrated that additions of BeO | reactor. Earlier Studies or ThO, to one molten fluoride breeder blanket mixture precipi- tated protactinium and that hydrofluorination of the mixture redissolved the precipitate. More recently, design and evalu- étion studies2 pointed out the strong economic incentive for development of a Pa-removal scheme that could be applied in large thermal molten salt breeder reactors. This document is a progress report of part of the continuing effort to develop improved Pa-removal methods. Results of a more extensive program that is being conducted by another ORNL group using tracer concentrations of 233Pa will be reported: elsewhere. Preliminary'fesults,bf this investigation have been published_.3 The presenf.report-deals primarily with studies conducted with Pa concentrations in the part per million range expected to be reached under reactor_operating conditions. A mixture of 2.-"IPa and'{3?Pa was.used in order to attain the required concentr#tion”withOUtlthe excessively high gamma activity levels associated with milligram quantities of 223Pa. Small amounts of this isotope (about 1 mc) were added to facilitate quick evaluation-of concentration changes occurring such contractor prepares, - during the experiments by use of gross gamma counting tech- niques. The actual concentration of 233Pa was five orders of magnitude less than the 231Pa concentration. Conclusions 1. A tracer level experiment (less than 0.1 ppb 233pa) showed that protactinium was precipitated from a molten LiF- BeF, -ThF, mixture by additions of ThO,, confirming the 1 and that the precipitated pro- results of earlier studies, tactinium was redissolved by a hydrofluorination treatment. 2. Equilibration of a molten fluoride mixture contain- ing 26 ppm 231pa with a lead-thorium alloy resulted in removal of 99% of the protactinium from the salt phase, but only é fraction of the reduced protactinium was found in the molten metal. A small amount (5% of the reduced protactinium was transferred from the molten lead to a fluoride salt mixture by a brief hydrofluorination treatment. It appears that better container materials are needed to assure the success of this Pa-removal technique. 3. Precipitation of ppm concentrations of 231pa in molten LiF-ThF, (73-27 mole %) can be accomplished by reduc- tion with thorium metal. The reduction rate was increased by an increase in surface-area. Hydrofluorination treatment of the reduced mixture redissolved the protactinium. 4. When the reduction of protactinium was effected by thorium turnings in nickel equipment, the largest fraction C , & -~ "o N - (more than.half) was found in the salt, probably associated with small nickel particles, and about 20% was associated with tne nickel-plated cepper screen used to support the'thorium in the melt. A comparatively small amount of the reduced protactinium was found on the nickel dip leg and vessel wall. 5. A similar distribution of reduced protactinium was found in copper apparatus. The results of the one experiment performed with tnis container material suggest that it is more difficult to dissolve Pa in copper containers than in nickel. EXperimental Facilities All the studies reported here were performed in the High- Alpha Molten-Salt Laboratory, which is.described in other reports,4 although the initial experiment was conducted with a small amount of 233Pa that does not require glove boxes for safe handling. Procedure for Thorium Oxide Precipitation Experiment A flanged nickel pot Fig. 1 was used for this experiu ment. A pretreated 516 g batch of LiF-BeF, -ThF, (73-2-25 - mole %D, to which 1 me of 233Pa had been added to 3 kg of | material,-was placed-in-a;z-in diameter nickel liner inside 'the,nickel pot. After}éeeling, the pot was evacuated and filled with helium. Heliem was introduced into the pot through the dlp leg at a rate of 350 cm3 /min while the contents of the pot were heated to a temperature of 575 Oc. © HOLES FOR Sin. 3-in. BOLTS ON {'¥gin. RADIUS, 60 deg APART\‘ 5/8 in. ORNL-DWG 66—-5396 3 ~in. SWAGELOK NUT ------------- j=—— KNURLED AND THREADED CAP Qy=e——3/g~ In. SWAGELOK UNION “~—— 3/, in. NICKEL TUBING — 4/,-in.NICKEL PIPE g~ in. SWAGELOK NUT 3~ in. COPPER GASKET vd N\ O 7 % NN A N DN 4.—in. INCONEL—-SHEATHED { 8 ¢r-Al THERMOCOUPLE l Y- in. NICKEL TUBE (DIP LEG) =————— 2-in. SCHEDULE 40 R S e N N N e A e N N R A A N A A N N e SN RN o N N N e N N N N S N N N N s A N S A N BNANSANNNS RSN N als in. NICKEL PIPE \J \J NICKEL LINER— | NNNNNNNNNNNNNNNY Fig. 1. Flanged Nickel Pot & A _Hydrogenwas‘substitutedrfor helium at this point and then 15 min later,-at a'temperature of 590°, HF was mixed with " the hydrogen in the approximate volume ratio 10 H,:1 HF. After 40 min HF treatment (final temperature_éos ®) hydrogen fIOW-continued for 60 min,to effect reduction of NiF, intro- duced by the hydrofluorination’treatment._ Since the HF content of the effluent gas'streamlwas not monitored\in these experiments, there was no way to determine when reduction of rlNin was completed The sampling procedure consisted of connecting the Vcopper f11ter'units (see_Fig. 2)~to.the gas manifold inside ‘the glove box by means of flexible 1/4-in plastic tubing and Swageloklfittings._Helium-was used to flush'out the unit for several minutes and,then'it'was introduced into'the'pot, sealed by tightening a 3/8-in Swagelok nut around a Teflon gasket, and then-the_filter-wasimmersed in the melt with helium still flowing through it.. After allowing time for the filter to ! reach the temperature of the molten salt mixture, helium flow was terminated and a vacuum was applied to pull a sample of athe melt into the filter.~ After the full available vacuum “'-iwas reached (23 to 26 1n ),Vthe filter was raised to a point 'near the top of the pot and allowed to cool for a few minutes 'before(removing it_from_the,potf The_filter was_then sacri- lpficed_to recover thegfiltered{salt mixture. This'mas weighed " and then a 1 gm”portion'ofteach-sampleWas ueighed'fplaced in "a small vial, and bagged out in a plastic bag for gross gamma | counting and for 233Pa analysis by gamma spectrometry. PHOTO 82997 | Welded Nickel Pot and Filter Unit 2 Fig. The schedule followed in sampling and ThO, additions is given in Table 1. Procedure for Reduction by Molten Pb-Th Alloy One experiment (Run 1-12) was conducted in which pro- tactinium reduction was effected by metallic thorium in the presence of liquid lead using welded nickel pots of the type shown in Fig. 2. This experiment comprised a number of steps extending over a period of several days: 1' Mixed 1 ml of 17 M HF solution containing 9.0 mg 23!Pa with 4.0 gram irradiated ThF, containing 1.1 mc of 233pa evaporated to dryness, added to an unlined nickel pot containing 330 gran LiF-ThF,; (73-27 mole %), hydrofluori- nated the mixture at 600°C, and reduced at 625°C for 4 hrs 20 min, cooled to room temperature under helium. Repeated HF-H, treatment because 38 gram of the salt mixture had formed a plug near the top of the pot that prevented sampling of the melt; cooled to room tempera- ture under helium._ : Connected the nickel pot containing the treated salt to a tantalum-lined nickel pot containing 692 gram of Pb-Th alloy saturated with Th at 600°cC. Effected transfer of salt to the lined pot with the pots at 650 OC and the transfer line at 600 . Added about 5 gram ofi thorium turnings to the mixture after 141 min equilibration'of phases. Cooled to room temperature under helium after 434 min contact time, 10 5. Put 206 gram of uncontaminated LiF-ThF, mixture in the nickel pot used at the beginning of the expériment; after removal of residual salt, gaVe the melted mikture a 20 min HF-H, tfeatmént followed by 3 hrs H, treatment and cooled to rdpm temperature under helium. - 6. Connected the salt pot to the pot containing the lead” and Pa-contaminated salt. Heated to SOOOCVandapplied vacuum to the salt pot to effect'transfef of lead phaée. Cooled to room temperature under helium. 7. Disconnected pots and weighed to determine amount Qf materiél transferred. 8. Héated the pot containing the fresh éalt plus transferred lead to 625°C and treated the mixture for 60 min with HF-H, gas. Cooled to room temperature under helium. Sampling of the salt phase was performed with copper filters. Stainless steel filter units were used to take most of the samples of the lead_phase but, in two insténces, improper positioning of copper filter units resulted in removal of lead samples in these units.' The samplers were conneéted to the glove box manifold by means of flexible plastic tubing, and sampling was carried out in the manner described'in the previous section. | Procedure for Solid Thorium Reduction Experiments The procedure used in these experiments shou;d be evident from the tabulated data and fhe discussion of results in the following section. iy f 1 | o ik o ;11 Results and Discussion Oxide Precipitation A summary of the oply compieté experiment on precipita- tion of protactinium by oxide addition is given in Table 1. This test was conducted with only a tracer concentration of 233pa present in the melt in order to check out the glove box equipment before the system became contaminated with a high activity isotope such as 231pa., The 233Pa activity in the melt (1 millicurie in 3 kg of melt) is equivalent to a con- centration of approximately 0.02 parts per billion. The data in Table 1 show that ThO, effects the precipi- tation of 23°Pa from molten LiF-BeF,-ThF, and that the pro- tactinium can be rediésolved by hydrofluorinating the mixture long enough to convert the ThO, to ThF,. It appears that, under the conditions used in this experiment, precipitation of ?33pa occurs rather slowly and additions of ThO, were made too rapidly to permit determination of the amount required for compiete precipitatifin-at equilibrium. The re-precipix= tation of partially fedissOlvedz33Pé shown by Sample #13 was due to water vapor from the dry iéé—cooled hydrogen trap that_warmed upduring_qyernight_operation. The reason for the low value in Sample #15 is not clear. When the top ofthejflanged pot was removed at the con- clusibn of the experiment,it was found that a large part of the fused salt mixture had frozen on the liner wall approxi- mately two inches above the surface of the melt. A large 12 Table 1. Precipitation of Tracer 233Pa From Molten LiF-BeF,-ThF, (73-2-25 mole %) At 630°C By ThO, Additions (Run 9-22) Sagg}é Treatment of Melt Prior to Sampling | 233P?XC{$5; gm. - salt as received | 3.34 1 95 min H,-HF 3.26 2 73 min He 7 3.32 3 50 min after addition of 0.809 g ThoO, 3.12 4 110 min after addition of 2.000 g ThO, = 2.54 5 45 min after addition of 1.989 g ThO, - 1.93 6 10 hrs helium | . ~0.09 7 90 min after addition of 2.068 g ThO, <0.01 8 68 min helium | - <0.01 9 85 min after 4.807 g ThO, <0.01 10 63 hrs helium <0.01 11 12 hrs helium | ~0. 06 12 4 hrs 20 min H, -HF 0.68 13 12 hrs H,, 2 hrs H, -HF . <0.01 14 2 hrs H,-HF 1.05 15 2 hrs H,-HF | ~0.07 16 5 hrs H, -HF 2.12 17 12 hrs H,-BF (?)} 4.39 - 202 g wall material : 2.32 145 g bottom material 5.54 # o 13 temperature gradient exists in the furnace well due to the necessity of maintaining a cool glove box floor. Part of the mixture obviously splashed up the wall to a point below the freezing point of the melt. This material was removed; weighed, ground and analyzed for 233Pa content along with a sample of the material in the bottom of the liner. The results are recorded at the bottom of Table 1. This finding, together with the above-mentioned failure to reach equilibrium after each ThO, addition, robs the experiment of any quanti- tative significance. However, the previously stated quali- tative conclusions are not affected. Reduction by Molten Pb-Th Alloy The data in Table 2 confirm the results of tracer-level experiments conducted earlier by other investigators® at this laboratory. In brief this experiment demonstrated that pro- tactinium can be effectively removed from a molten salt mixture by contacting it with molten lead containing thorium but it also showed that only part of the protactinium content - of the salt can be found 1n the molten lead. Protactinlum that does stay in the lead long enough to permit transfer to ~a new pot can be recovered in a molten fluoride by hydro- lfluorlnating the mlxture., The exact amount of protactlnlum tpresent in the molten lead seems to be open to quest1on. Most of the samples of.this_phase were removed by use of.stainless steel filter units. The'exterior wall of each unit was scraped to remove any salt adhering to it and then it was treated with Table 2. Reduction of 26 ppm ?3!Pa Dissolved in 330 g Molten LiF-ThF, (73-27 mole %) by - Equilibration With Pb-Th Alloy at 630°C. (Run 1-12) Salt Phase Metal Phase Sample 2-Phase Contact 23!'Pa C/m Total Pa % of Total Z3TpPa C/m Wt Total Pa % of Total No. ¥1me : Pa Sample Sample - Pa ' min) 1 gm (mg) gm (mg) 1 - 1.29x108 8.5 100 2 - 1.18x108 7.8 92 3 34 1.6x105 '5.605(ss) 0.39 4.6 4 57 6.68x10% 4.4 52 | 5 88 3.7x10% 6.017(ss) 0.09 1.0 6 110 6.29x105 4,2 49 Added Th metal after 141 min contact B 144 5.6X10% 3.7 44 7 216 1.38x10* 0.091 1.1 | 8 251 2.8x10% 5.882(ss) 0.07 0.8 9 341 1.1x106 10.350(Cu) 2.7 32 10 359 3.4x104 5.559(ss) 0.08 1.0 11 404 1.25x10% 0.083 1.0 o Transferred 412 g Pb phase to unlined pot containing 206 g LiF-ThF, 12 40 min HF-H, 1.0x10% 0.40 4.7 | | A 13 40 min HF-H, 3.8x105 10.4(Cu) 0.30 3.5 14 60 min HF-H, 0.19 . 1.3x105 5.8(ss) 71 o\ 15 acid before dissolving. It was found possible to dissolve the jackets with only a small part‘of the lead sample while the bulk of the lead was dissolved and analyzed separately. The results are shown in Table 3. It was not feasible to make a similar separation in the case of the samples obtained with copper samplers. The data in Table 3 demonstrate that very little of the protactinium associated with the samples was in the filtered lead. It seems likely that protactinium was removed from the'lead while passing through the sintered stainless steel filter media but, since this material was dissolved along with the side wall, evidence in support of this belief cannot be given at present. The higher pro- tactinium content of the samples obtained with the copper filter units, as compared to the stainless steel samples, likewise cannot be explained on the basis of presently avail- able information. Reduction With Solid Thorium " The first experiment”in'which a solid thorium rod was exposed to a molten fluoride mixture contalning protactinium (Run 2-22) gave somewhat anomalous results, probably due to electrochemical effects._hThe_results can be summarized as follows: Exposure of the 3/8-in diameter rod to 240 gram “of LiF-ThF, (73-27 mole %) at 625°C for 65 min resulted in a reduction in 23!pa content of the melt from 11.1 mg to 0.087 mg (0.8% of starting conoentration). A further 5-hr exposure resulted in an increase in 23! Pa content of the Table 3. Analysis of Stainless Steel Sampler Jackets and | Their Contained Lead Samples (Run 1-12) 231pa Content (C/min) Lead Distribution (Gm) Sample Jacket Lead % 231 pa Jacket Core Total No. Solution Core in Core Solution Solution 3 1.6 x 105 3.8 x 103 2.4 0.355 5.250 5,605 5 3.7 x 104 <50 <0.1 0.417 5.600 6.017 8 2.4 x 10% 4.0 x 103 14 0.432 5.450 5.882 10 3.4 x 104 80 0.2 0.409 5.150 5.559 91 -l o\ 17 melt, based on analysis of the filtered sample, to 0.54 mg (4.9%0; Samples obtained during 63 hrs treatment of the melt with a H, -HF mixture gavé protactinium contents varying from 3 to 16% of the initial concentration. It was found that 70% of the bottom l%g-inch of the rod (28 g) had been eroded by the exposure. The recovered salt had a large amount Qf black material in it, some of which was magnetic and analyzed 45% nickel and 30% thorium. It contained 0.15 mg %231 Pa per gram. The non-magnetic material contained 22% nickel, 49.5% thorium, and 0.27 mg 231Pa per gram. It seems likely that the thorium rod was in contact with the bottom of the nickel pot during this experiment, causing a current flow that eroded the rod. The chunks of black material were rather brittle and it seems probable that they were compacts of finely divided thorium and nickel particles rather than alloys, together with a small amount of the LiF-ThF, salt. | Another experiment (Run 3-2) was performed in which a section of the same 3/8f1nrth9rium_rod used in Run 2-22 was exposed to a fresh'24sgrgm batch of LiF-ThF, under the same cénditions used in the_pfevioué expériment except that care waé_exerted to preventicontadt.of the rod.with.thewbottOm of the"piCRel_pot. _Measuremehts.fiere'also made of the.potentiai difference between the_fiickel_rod which supported the thorium (insulated from the body of the pot by a Teflon bushing) and the grounded shell of the pot. During the first immersion of the rod, a value of 0.161 volt (grounded side negative) 18 was measured. During the second immersion, the value:varied’ 3 from 0;242 to 0.231 volt. Data obtained in this experiment are sumnarized in Table 4. Reduction of Pa occurred at a much slower rate and there fias no noticeable erosion.of the thorium rod. The drop in 23‘Pa concentration that occurred on remelting the' mixture after the initial HF and hydrogen reduction freéf-» ment was completed and the melt had cooled overnight under rhelium probably indicates that some oxide contamination of the melt occurred. There is no obvious explanation for the drop in 231Ppa concentration in the 6th sample. It may indicate non-equilibrium conditions in the mixture. fhe ~ high concentration in the ground (-80 mesh) sample of recovered salt seems to indicate that the initial HF treat- . ment of the melt was inadequate for complete dissolution of the added 23! pa. The slow reduction of dissolved 23!Pa observed in Run 3-2 encouraged efforts to increase the reduction rate by increasing the surface area of the solid thorium metal in Run 3-10. A nickel-plated copper screen was used to supportr 3.6 gram of thorium metal turnings in the melt which con—' -sisted of the salt recovered from the previous experiment. The data obtained are shown in Table 5. The data indicate that a larger thorium surface area did indeed increase the rate of protactinium reduction, as expected. The treatment of the'melt with thorium was not extended because the basket &s& n 3 »h ¥ .y ¥ (_ Table 4. Exposure of Thorium Rod to 243 g Molten LiF-ThF, (73-27 mole %)Containing R | 32 ppm 231pa (Run 3-2) Sample 231 pg 231pgy % of Total® 231pa No. C/m - 1 gm. Concentration Initial Comments | ‘mg/gm Conc. (mg) 1 1.59 x 10® 3,18 x 107 100 7.72 50 min H,-HF and zyoffiifi H, 2 f1.33x§io§'7-';iz;66 x 10~% - 84 6.30 Remelted under helium 3 4.82 x 105 9.64 x 1073 30 2.22 60 min Th rod exposure 4 3.04'x:105 o '.'6,01 x 1073 19 1.35 120 min Th rod exposure 5 1.35 x 106 2.7 x 1072 85 5.91 65 min H, -HF 6 1.19 x 10¢ 2.38 x 1072 75 5.07 135 min H, -HF 7 1.54 x 106 3,08 x 1072 97 6.37 75 min H, -HF A 1.91 x 106 3.82 x 1072 120 7.69 Ground sample of recovered salt a Calculated on the basis that 6 gms of salt were removed in each sample. 61 Table 5. Exposure of Thorium Turnings to 204 g Molten LiF-ThF, (73-27 mole %) Containing 27 ppm 23'Pa (Run 3-10) 231 2 Shaple PR Cone e ot Connents 1 1.54 x 108 3.08 x 1072 | 100 Short HF-H, and H, treatment | 2 1.37 x 105 2.74 x 1073 9 65 min thorium exposure 3 1.36 x 108 2.72 x 1072 88 95 min HF 4 1.42 x 108 2.84 x 102 92 75 min HF A 1.48 x 106 2.96 x 102 96 43 min HF-ground recovered salt (unfiltered) 0 -3 21 came loose from its support rod the first time that an attempt was made to remove it from the pot. The thorium was apparently dissolved by the hydrofluorination treatment as the bésket was completely empty, except for some solidi- fied salt, when it was removed from the melt. In this case, the ground (-80 mesh) recovered salt showed a ?3!Pa concen- tration slightly lower than the initial value. The 23!pa content of the first sample was exactly the same as that of the last filtered sample (No. 7) in Table 4 and this agreement casts doubt on the validity of the high 231pa content of sample A, Table 4. Run 3-31 was essentially a repeat of Run 3-2 except that no effort was made to return the reduced protactinium to solution., The épparatus was disassembled and all parts that had been in contact with the melt were analyzed for 231pa content in order to determine the distribution of reduced prétactinium in the nickel equipment. The results given in Table 6 show that moreftpaprhali_oflthe reduced protactinium was'in_the‘unfiiteredsalt:mixture; probably attachéd to small metal particlesdispersed in the salt mixture, about 1/5'was adsorbed on (Or'ailbyéd with) the nickel-plated copper basket, and only 7% reached the wall of the nickel .~ pot. As is indicated in Table 6, some magnetic particles were removed from the recovered salt. There is no good explanation at present for the missing 20% of the protactinium, ~ An experiment similér to Run 3-31 was conducted with the Table 6. Reduction of 19 ppm 231pa Dissolved in 280 g LiF-ThF, (73-27 mole %) by Exposure to Thorium Turnings and the Distribution of ' Reduced Protactinium in Nickel Apparatus (Run 3-31) Sample 231 pa Conc. Total % of Total No. C/m - 1 gm. 231 pg 231 pg Comments or Sample (mg) 1 9.29 x 10° 5.20 100 43 min HF and 60 min H, 2 9.29 x 105 5.20 100 65 min H, 3 3.08 x 10% 0.13 2.5 115 min thorium exposure A 5.30 x 105 2.65 51.0 Recovered salt ground to -80 mesh Dip leg 2.21 x 108 0.044 0.9 Part of nickel dip leg exposed to melt Basket 5.22 x 107 1.04 20.0 Includes part of support rod - screen very brittle Pot 1.80 x 107 0.36 6.9 Pot was cut above the salt line Magnetic 2.94 x 106 0.06 1.1 Mostly nickel particles from sawing pot Matl. but includes some particles from salt n ~