ORNL 1030 Reactors-Resea 3 4456 D28y831 5 M and Powe [y B (NN cory B g!?r“ - AN INVESTIGATION OF @ ThF4> Composition = 1000 g/L Th Th = 1 atom Vapor Pressure Melting Point £ 760 mm at 500- 800°¢ £1300°C U-233 = 0.02 atom Moderator = 100 atoms {760 m at 500-800°C < 300°¢ ‘ -8- ORNL-1030 Thorium Reactor Solutions {comtinued) specifications required; they are intended to give merely a rough picture of vhat, in the light of current thinking, would be desirable characteristics for an eventual thorium reactor solution to possess. In both applications, the nuclear considerations are of parsmount im- portance. If the specifications of Table I are accepted, it becomes apparent that one is limited to only & small number of fluorides having sufficlently low capture cross sections and favorable vepor pressures to enable them to be used in a reactor solution with thorium. These salts are listed with their neutron capture cross sections and melting points in Table II. It 'is realized that of the compounds listed in Table II, ZrF) might prove unusable because of its tendency to sublime at temperatures of the order of gseversl hundred degrees and BiF3 might eventually prove troublesome because of difficulty in preventing the displacement of Bi**from the melt by manmy of the structural metals of which the container walls would normally be made. Other f;uorides of slightly higher cross section than those listed in Table II might be considered for use but could, of course, be present only in relatively smaller concentrations. A Literature Survey A survey of the project and open literature was made to supplement the 1) search made by Grimes and Hill.( Special emphasis wes placed on phase studies * of the particular salts listed in Table II. The results of this survey which - -9~ ORNL-1030 Table II Cross Sections and Melting Points of Several Imorganic Fluorides Melting Polmt Compound ¢ Barng BeF, 800 0.03 11'F gl 0,03k BiFs 727 0.045 MgF» 1263 0.08 PbF, 813% 0.21 ZxF), 872 0.22 AlFq 1040 0.23 ThF), 1080% T %% The crogs section of F in this teble has been taken as 0.0l barns. ' * The figures so marked are experimental determimations made as part of the present work. aln. -10- ORNL-1030 A Literature Survey (continued) were adjudged to be of the most interest to this problem are noted in Table III. The ThF} -KF and ThF}-RbF reference was included in Table III despite the unfavorable nuclear characteristics of K and Rb because it represented the only specific reference to ThF), eutectic mixtures found in the literature. Zachariasen(1l) has studied double salt formation in the systems NeF-ThF, and KFwThFh by the X-ray diffraction method but he apparently did not investigate the complete phase diagrams of the?e systems. ThF) has also been found to form a very stable complex with Rb, having the formmula, Rb3ThF7.(3) Experimental Apperatus The equipment used for this work was the same as that used by other work- ers for thermal analysis studies of UF), salt mixtures.(le) In essence, it con- sisted of a 5-inch chromel-wound pot furnace capable of operation at tempera- tures up to 1100°C. The tempersture of the furnace was controlled by means of a variable transformer connected in series with the A. C. supply. The salt mixtures were heated in graphite crucibles which fit into the furnace in such a manner that an atmosphere of No could be malntained over the melt during heating and cooling periods. Temperatures were measured by means of a chromel- alumel thermocouple situated on the ingide of a graphite stirrer which extended to the bottom of the crucible. The temperatures were measured and recorded by " a Brown "Electronik" potentiometer and tests indicated that the temperatures . Table III ORNL~1030 Some Published Phase Relationships of the Fluorides of Table IT Eutectic Composition Eutectic Temp. System Mole Percent oC Reference ThF), -KF 17 ThF), 664 (3) 33 ThF) 750 57 ThF), 878 80 ThF), 954 ThF),RbF 15 ThFy 664 (3) 37 ThF), 762 80 ThFh_ 1000 LiF-A1Fy 1.5 AlFq 706 () 37 AJI‘3 691 LiF-AlF4 36 AlFg 710 (5) LiF-MgF, 33 MgF, Th2 (6) LiF-MgF, 53 MgF, 718 (7) LiF-BeFo 52 BeFp 360 (8) LiF~MgF,~NaF, 10 MgF,, 43 NaF 630 (6) 29 MSFE 3 12 NaF 68!4- MgF,=BeF, Complete Miscibility (9) BiF4-FbFp Complete Miscibility (10) t 1. ORNL-1030 so -recorded were accurate to + 5°C. Materials IhE), The thorium fluoride used was obtained from the Iowa State College. The thorium analyzed, gravimetrically, 75.1% and the fluoride, 25.5%. The the- oretical Th content is 75.3%. A spsctrographic analysis indicated the sample wvas essentially free of rare earths. The melting point of this ThFh was found by experiment to be 1080 + 5°C. No reference could be found to a previous melt- ing point determination for this compound. ALF3 The A1F3 was prepared from a stock of Baker and Adamson.AIF3-XH20 by heat- ing the hydrated material in an atmosphere of HF to about 600°C over a pericd of 3 to 5 hours. The vendor reported impurities in the A1F3-xHéO amounting to less than 0.014%. The dehyrated product amalyzed 32.5% Al and 67.9% F.(the- oretical Al = 32.1%). The high volatility of AlF; in the neighborhood of 1000°C prevented an experimental determination of its melting point with the equipment on hand. MgFo The MgFs used in this work was purchased from Eimer and Amend and was re- ported by them to be 99% pure. A spectrographic analysis indicated the major impurities to be Ca, Na, Cr, Fe and Ta. PbF2 The FbF» used was Baker and Adamson "Purified” material and was not . 2 -13- ORNL-1030 PbF, (continued) analyzed chemically. The melting point was found experimentally to be 813 + o o (13) 5°C, which may be compared with a literature value of 822°C, Lir The LiF was material purchased from the Maywcod Chemical Works. No analysis of the LiF was carried out but an experimental determination of its melting point (845°C) agreed exactly with the most relisble value obtained from the literature.(lh) On the basis of this agrsement and its cleen appear- ance, it 1s belleved that this was materlal of high purity. U The UF} used was obtained from K-25 through the ORNL SF Accountability Office. Prior to their use in this work, all of the chemicals were dried by heat- ing in an oven at 110-115°C for 24 hours and were stored in dessicators upon removal from the oven. Results bRy, -LiF The phase dlagram for this system 1s given in Figure 1. A eutectic con- taining about 26 mole % ThFj is formed which melte at 550°C. Some difficulty wag experienced in obtalning liquidus points from cooling curves of mixtures in the vicinity of the eutectic becausze of very pronounced super cooling. A compound with an incongruent melting point at about 925° 13 formed at T5 mole % ThF). _— -2 ORI 2030 ThF), -MgFp The ThF)-rich side of this system up to 60 mole % MgF, was investigated and the results are shown in Figure 2. Two eutectics were found which melted at 915° and 925°C corresponding to compositions of 25 mole % and 40 mole % ThFy, , respectively. A compound with a congruent melting point of 937° was indicated at 33 mole % MgF, and can be represented by the formula, MgThoFyg. The investigation of this binary system was not carried further than 60% MgFo because of the temperature limitatioms of the equipment but it seems probable that no further eutectics of immediste interest to this problem would be found at higher MgF2 concentrations. Th¥), -PbFo The proposed phase dlaegram for this system is given in Figure 3. Two eutectics were obtained, one at 35 mole % ThF)y , melting at 9250 and a second at 62 mole % ThFl which melted at 880°C. Some indications of a third eutectic melting at about 760° and containing less than 2 mole % ThF), were obtained but its presence was _not definitely established. Two compounds with congruent melting points at about 950° and 942°C were indicated, corresponding to the formlas Pb17Th3F46 and PbThFg, respectively. A third compound having en incongruent melting point of sbout 1045°C was Indicated with a formuls, PbThgF3g. Measurements of this binary proved unsatisfactory in a sense because of a reduction of the Pbtt to elemental Pb by the graphite crucible and stirrer. This reduction was not observed to occur appreciably at temperatufes less than 800° but became a greater problem with increasing temperature. A -15- ORNL-1030 " Drawing # 11461 1100 ] I [ l I 6 J— FIGURE | _— PHASE DIAGRAM OF ThF, - LiF BINARY SYSTEM " o 1000 900 s—— ) \O N 800 AN S q " N\ S 70 O ]— < as \\ s / = = 600 - o S00 400 0 10 20 30 40 50 60 70 80 90 100 MOLE PERCENT ThF, 3 CHEM-TECH-DIV---LAB-SEC-~JEF--8-1[-51 - | °C TEMPERATURE -16- ORNL-1030 Drawing i 11462 1200 10O 0,/ 0 \ / .0 1000 5 0/ \ /O’bo/. S _¢—0 - ® ®—0 900 800 ) el FIGURE 2 | ( ThF4~MgF, BINARY SYSTEM MthzFlo 700 600 0o 10 20 30 40 50 60 70 80 MOLE PERCENT MgF, CHEM-TECH-DIV--LAB-SEC--JEF~ 8-11-8I- 2 °C TEMPERATURE 1200 1100 1000 900 800 700 600 -17 - ORNL-1030 Drawing # 11443 PHASE DIAGRAM OF ThF,— PbF, BINARY SYSTEM FIGURE 3 §=—o—0— \ /° T T ° l OI | \ 1 S | l | 1 I \ ‘ FoThs ol 0 10 20 30 40 50 60 70 80 90 100 MOLE PERCENT Th F4 o CHEM-TECH-DIV-LAB-SEC~JEF-- 5-11°5Il- 3 s ~18- ORNL-1030 ThF), -AJ.F3 Considerable effort was expended on this system with only moderate suc- cess. The AJ.F3 sublimed at temperatures from 900 to 1100°¢ to such en extent that reliable results could not be obtained with mixtures containing more than 20 mole % AIF3. A reproducible eutectic halt was cbtained at about 950°C in the cooling curves of this binary system but the composition of the eutectic mixture could not be determined. It is estimated that it lay somewhere in the vicinity of 25 mole % AlF3. THF), -UF) The isomorphism of ThF) and UF;L(]'T) would lead one to expect them to be miscible in both the solid and liquid states. Although supercooling prevented the accurate determination by thermal analysis of a phase diagram for this system, indications were that liquidus and solidus curves existed which ex- hibited neither a maximum nor minimm point. X-ray anslyses of several of the gsolidified melts showed that solid solutions were formed. ThF}, -LiF-MgF o This was the only ternary system investigated to any sppreciable extent and it was studied in only an exploratory mammer. An indication of a ternary eutectic of high ThFj content melting at 690° was obtained but since 'bh:fs tem- perature was greatly in excess of the desired melting point, no formal attempt was made to establish the composition. ThF), -L1F-NaF-MgF, A reference found in the literaature to the NaF-LiF-MgF, eutectics(3) g ~19- ORNEL-1030 ThF) -LiF-NaF-MgFo (continued) seemed worthy of further investigation, especially the secondary eutectic con- taining only 124 NaF. Experiments were carried out which coasisted, at Pirst, of adding ThFy to the 29% Mg, 12% NeF, 59% LiF eutectic mixture and finally, cf changing the concentrations of the other components empirically. In this way, & quaternary eutectic was obtained which melted at 530°C. The approximste composition of this eutectic, expressed in mole percentages was: 20.0% TLFy, 62,3 LiF 12,7 Na¥F 2.0 MgFo UF), -LiF-NaF-MeF Proceeding along the same lines as described above for the case of the ThFh quaternary mixture, a eutectic was found in this system which melted at about 460°C. Tts approximate composition was: 25.0% UF), 58,2 LiF 11.8 NaF 5.0 MgF, Discussion Of the systems studied, two eutectic mixtures stand out in the respect that their melting points are both more than 100D under the next lowest melt- ing mixture found. The eutectics referred to are the TXF)-LiF binary eutectic, - containing 26 mole % ThF) and melting at 550° and the ThF)-LiF-NeF-MgF, quater- nary, containing 20 mole % ThF) and melting at 53006. The concentration of Th -~ . 4 -20~- CRNL-1030 Discussion (continued) in these eutectics corresponds to about 2000 g/L in the binary and 1500 g/L in the quaternary. The neutron capture cross section ratios (Table I) are 73 and 16, respectively end the vapor pressures at 800°C were,by cbservation, much less than 760 mm. On the other hand, the melting points are still considerably higher than the 300°C taken as the maximum desirable melting point. The ThF)-FPbF,, ThF) -MgF, and ThFh-A1F3 binary systems show no eutectic of sufficiently low melting point to be of immediate interest to this prcblem. Thelr principle value lies in the contribution they would meke to & study of ternary systems contalning these components. From the same viewpoint, binary systems of ThF) with BeFp, BiF3 and ZrFy should be investigated. The UF)y-LiF-NaF-MgF, quaternary eutectic is similar im both meliing point and ursnium concentration to a UF)-LiF-NaF ternary eutectic reported previously as being under consideration as a fuel for the ANP reactor.(l5) Although indi- cations are that the unavailability of L17 isotope will preclude, for the time being, the use of a LiF constituted fuel in this reactor, it should be pointed out that, other factors being equal, the better neutroen economy of this guater- nary would likely make it more acceptable than the termary eutectic. The similarity between the diagram found for the ThF), -LiF system and that obtained by Grimes, et. al.(12) for UFh-LiF is rather remerkable. A UF), ~-LiF eutectic melting at 480°C was found at about 26 mole % UF), and likewise a com- pound, LIU3F13, with an incongruent melting point was obtained. The same in- veatigatérs(ls) have found & eutectic in the UF,-FbF, system at 62 mole % UF), a_— i -21- ORBL-1030 Discussion (comtinued) which melts at about 730° and have obtained a diagram for the UF,-PbF, system vhich conforms in most principle respects to the character of the diagram pre- sented in Figure 3. No attempt was made to rigorously defime the extemt of the similarity between ThF) end UF, systems but from indications obtained in this investigation, a close coordination between the work done on these two general systems should be maintained. From the results obtained thus far in this work, the possibllity seems rather remote that a mixture meoeting all the specifications set for this prob- lem will be found. On the other hand, there appears to be a very real 1ikli- hood that mixtures which come nearer meeting them than anything so far found can be obtained and such mixtures might eventually prove as satisfactory for reactor use as the "ideal" one considered here. The work done to date can be considered as no more than a start on such a search and whenever the importence of this problem is adjudged more immediate, a thorough and more nearly complete solution to the problem should be allowed. %“’A / ! J. 0. Blomeke A -22~ CRNL-1030 (1) (2) (3) (%) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) Bibliography Grimes, W. R. and D. G. Hill, "High Temperature Fuel Systems - A Literature Survey,” Y-657, (July 20, 1950). Chase, L. Hey Ro C. Crooks, J. N. Pattison, J. J. Ward and J. W. Clegg, "Chemist:;'y of Liguid Fuels for Nuclear Reactors," BMI-T-53, (January 15, 1951 Dergunov, E. P. and A. G. Bergman, Doklady Akad. Nauk. S.S.S.R. 60, 391-4 (1948). Puschin, N. A. and A. V. Baskoff, Z. anorg. Chem. 81, 347 (1913). Fedotiev, P. P. and K. Timofeev, Z. anorg. allgem. Chem. 206, 263 (1932). Bergman, A. G. and E. P. Dergtmov, Compt. rend. a.cad. gci. U.R.S.S. 31, 755 (1941). Hynieki, V. P. and P. F. Antipine, Chinie et industrie 17, 601 (1927). I({oy, 1))e1.'l.a M., Rustum Roy and E. E. Osborn, J. Am. Ceram. Soc. 33, 85 1950 Venturello, Giovanni, Atti. acad. sc¢i. Torino, Classe se¢i. fis., mat. nat. 76 I., 556-63 (1941); Chem. Zentr. 1942 I., 111k, Croatto, Ugo, Gazz. chim. ital. 7k, 20 (1944). Zacheriasen, W. H., J. Am., Chem. Soc. 70, 2147 (1948). See also CC-3401, (Jan. 10, 1946), and CC-3426, (Feb. 9, 1946). "Chemistry of Liquid Fuel Systems,” The Aircraft Nuclear Propulsion Project Quarterly Progress Report for Period Ending August 31, 1950, ORNL-858, p. 104 £f (December 4, 1950). Tables of Selected Values of Chemical Thermodynamic Properties, Series 2, UcSo N&‘t"l.BUI- Stand&rds, 191"7, P- 27"10 2 Pc 91"'1. "Chemistry of Liquid Fuels,” The Aircraft Nuclear Propulsion Project Quarterly Progress Report for Period Ending December 10, 1950, ORNL-919 P. 234 £f (February 26, 1951). — -23- ORNL-1030 (16) Grimes, W. R., et al., Unpublished results. (17) zachariasen, W. H., "The Crystal Structure of Fluorides of Th, U, Np and Pu," MDDC-1151, (Jenuary 11, 1947).