: E . . 22 . ! 31 UNCLASSIFIED ORNL u 707 M . 2. W WP. A 1 2 11 N 14 . 1. ? MT I T . ( 11 " , " 7 1 ir " " .. ORNL-P-407 TE JAN 2 : 1965 MASTE To the Graduate Council: I am submitting herewith a thesis written by Arthur J. Shor entitled "A Study of the Pha se Behavior of the Aluminum Trichloride- Zirconium Tetrachloride System." I recommend that it be accepted for eighteen quarter hours of credit in partial fulfillment of the requirements for the degree of Master of Science, with a major in Chemistry. Major Professor We have read this thesis and recommend its acceptance: . - Accepted for the Council: . .... LEGAL NOTICE - TMI report me promene u Komut of domenu spunesond wort. Melchor da United Home, nor the Commission, nor w pornon setting a ball of the counterton: A. Makao muytuty or moperation, prend or lualle, mu roncoct to the acco- rosy. Our planeth, of world of labordou nontos report, or that the of my laboration, yuritu, method or process dirland la duke mport my not latring { print owned and or 1. 3. Assumse Bay liabilities with respect to the vas of, or for damages resulting from the on tuonuttag, uprantu, wodhod, os procedimelound is daaroport As wid to the whora, perna acting on ball of the corneraton" tncludere may be plogne of contractor of the Commission, or amphorn of wel contractar, to the use che much employs or contractor of the Conwasband, or sployme al mail contractor propurus, daromaster, or pronon noces bo, way worur me permetto Maplogut or contract nu the Conntastou, or Momploymat mah moal contrastor. . . ... Dean of the Graduate School II . . NI. Il . ts A STUDY OF THE PHASE BEHAVIOR OF THE ALUMINUM TRICHLORIDE- ZIRCONIUM TETRACHLORT.DE SYSTEM . t A A Thesis Presented to the Graduate Council of - - The University of Tennessee it .. 77 - . . - Y - 1 In Partial Fulfillment * of the Requi rements for the Degree Master of Science TV2 9 . , ' - - 1 - Arthur J. Shor December 1964 . . . 1 . ", + 4 y * wann wie 2 - . U 1* . ACKNOWLEDGMENT . . . - . A2: The author would like to express his sincere gratitude to Professor W. T. Smith of the l'niversity of Tennessee for constant encouragement and guidance during the entire course of this research; and to Mr. R. E. Thoma of the Oak Ridge National Laboratory for numerous helpful discussions and for the loan of the differential thermal analysis equipment. In particular, VE+ . S 1 i the consideration given the author by Mr. E. G. Bohlmann, Associate Director of the Reactor Chemistry Division, in allowing the use of laboratory time for part of the experimental work; and the encouragement, and many helpful . . . suggestions from Dr. E. L. Compere are gratefully acknowledged. The co-operation of a number of members of the staff of the Oak Ridge LP A National Laboratory was essential in the performance of this work. Special mention should be given to Mr. J. R. Hart for assistance in the construction and operation of part of the experimental equipnent; and to Mr. J. M. Baker and Mr. H. A. Friedman for valuable advice in design and operation. Mr. D. E. LaValle and Mr. R. B. Quincey, Jr. deserve special thanks for the preparation and purification of the salts used in this study. Chemical analyses by Mr. W. R. Laing and x-ray diffraction studies by Mr. .. R. M. Steele and Mr. R. L. Sherman, all of the Analytical Chemistry 2 Division, Oak Ridge National Laboratory, are gratefully acknowledged. . :37 MA Final thanks are due the Reactor Chemistry Division directed by Mr. Warren Grimes; and to the Oak Ridge National laboratory operated for the U. S. Atomic Energy Commission by the Union Carbide Corporation for support- ing this work. is UI Net haw 7 14 V 1 TABLE OF CONTENTS CHAPTER PAGE I. INTRODUCTION. ........................ i - A. Objectives of Study .................... i 1. General Pha se Behavior. ................ i 2. Friedel-Crafts Catalyste. ............... 2 A + A 3. Protactini unc. Recovery ................. k wn . . B. Previous Investigations of Similar Systems. ........ II. MATERIALS, EQUIPMENT, AND EXPERIMENTAL TECHNIQUES ....... - . T A. Materials Preparation ................... . . 1. Aluminum Trichloride. ........... . . . . • S IM . 2. Zirconium Tetrachloride ................. Experimental Techniques ................. cut . • R ŏ F F oo oo VPS . 1. Differential Thermal Analysis ............. . • 1. . 1 Experimental Equipment. ................ . . . • . D. Interpretation of Differential Thermal Analysis free . . . • _- TA E. Supplementary Experimental Techniques ..... . • 2S v 1. Thermal Analysis. ............. . • Direct Visual Methods . . . . . . . . . . . . . . 25 . • III. DISCUSSION OF EXPERIMENTAL RESULTS. ........ . • 14. A. General Phase Behavior. .................. Î Ô w W N N N B. Activity Coefficients of Zirconium Tetrachloride. . • C. The Aluminum Trichloride Rich Region. ...... . • D. The Eutectic Region .............. . • . E. Supercooling Effects. .............. . • iii * .. . is 2 . Wh o WTF." ***, P anasoni ::"77.2, 1", TFT win a PIT ' ' 11 = "" . C - M TV YA iv . . . * CHAPTER PAGE Sättin IV. SUMMARY AND CONCLUSIONS. ...................4 1. General Pha se Behavior ................ 42 n 2. Friedel-Crafts Catalysts ............... ENT 3. Protectinium Recovery. ................. BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Vai maitinimas ir .. .. 16 . " VE . . A . U 1 . ******w wmama liyompa UWA U TIN WWW 0 2N im WT . LIST OF TABLES E , TABLE PAGE I. Analysis of Chloride Salts. ................. 9 II. Spectrochemical Analysis of Chloride Salts. ..........10 III. Melting and Freezing Point Temperatures of the Zirconium Tetrachloride-Aluminum Trichloride Binary Pha se System. ... 32 IV. Activities of Zirconiu Tetrachloride . . . . . . . . . . . . . 1. His 257, . > . x . " " , 1 - 1 Wu . -- - MIST OF FIGURES FIGURE PAGE * 1. Differential Thermal Analysis Equi UudiIIUL TICIDI AIELYBL8 Muu punan. . . . . . . . . . . "! 2 2. Differential Thermal Analysis Capsules Holåer and Shell. ..... 3. Schematic of Differential Thermal Analysis Equipment ...... 16 4. Nickel and Quartz Specimen Capsules. .............. 17 5. Construction Drawing of Nickel Capsule ............. 18 6. Differential Thermal Analysis Heatup Curves. .......... 21 11 . 7. Binary Pha se Diagram, Aluminum Trichloride-Zirconium Tetra- SK . . . chloride Systenl. ....................... 28 . 8. Aluminum Trichloride Rich End Ol' the Aluminum Trichlorike- - . .. 1 Zirconium Tetrachloride Binary Pha se Diagram ......... 29 9. Activities of Zirconium Tetrachloride in Monomeric and DlueilC ALULAW Dimeric Aluminum Trichloride .. ITICALOTlae . . . . . . . . . . . . . . . . n. & . . . a vi . ! 7 1 . TEN . X - immune CHAPTER I INTRODUCTICN A. Objectives of Study 1. General Pha se Behavior site damianinowe The phase behavior of molten fluoride salts has received consider- merawatan mentioning met able attention in recent years as a result of their promising physical mmista internet dating and chemical properties as nuclear reactor coolants. Chloride salts have received somewhat less attention. However, with the increasing interest awi in the latter as fast reactor coolants, as solvents in the reprocessing de grande ristmi in hindi hindi mov in the onder mirtinnere m of spent reactor fuel elements by the chloride volatility technique, and as components of nuclear reactor blarikets, research on chloride salt systems sti seisationen is accelerating. Part of this work involves the study of binary mixtures in which one or both of the components is an alkali or alkaline earth nis tritic chloride. Relatively little effort has been allotted to mixtures of Group m i IIIA and Group IVB metal chlorides such as binary combinations of aluminum s trichloride (A1C13) with titanium tetrachloride (T1C14), zirconium tetra- *** . . - chloride (ZrC14), or thorium tetrachloride (ThC14). These binary systems mostaten te moeilik te site are especially interesting since the components do not exhibit the exten- 12 sive ionic bonding of the alkali and alkaline earth halides, nor are they ther DU entirely covalent. They range from the rather predominantly ionic struc- rin ns tures exhibited by thorium tetrachloride to the more covalent type struc- aard.com loi imaginati tures of titanium tetrachloride and aluminum trichloride. In addition, the chloride anion in these melts is more polarizable than the fluoride ion w 2 adding to the complexities of the systems interactions. It is likely that * la it is n 23 the cations do not have a simple structure in solution, but consist of associations or complexes of various kinds. There is considerable evidence for the existence of complex solid state crystalline structures for the pure salts. Aluminum trichloride is particularly interesting in that conductivity and x-ray measurements Indicate that at the melting point a change from an ionic to a covalent structure occurs together with a large increase in the molar volume. tags Zirconium tetrachloride is believed by Blumenthal to exist, as a hexe chlorozirconate ion (Zrc16-),“ while Olah states that the structure 18 a molecular crystal lattice of the tin tetraiodide (SnId) type.' Ittle is known of the liquid structure of ZrC14. Thus a phase study of the al.uminum trichloride-zirconium tetrachloride binary could be significant in providing some insight into solid liquid structure and thermodynamic activities of the components. 2. Friedel-Crafts Catalysts The importance of aluminum trichloride as a Friedel-Crafts catalyst is well known. Among the aluminum halides, it is outranked in catalytic activity by aluminum tribromide (AlBr3) and even more so by aluminum tri- iodide (A1I3). This effect is attributed to the decreasing energy require- ments in obtaining the monomeric form which is believed t:o be the active form of the catalyst. However, the relative level of catalytic activity of aluminum trichloride, although significant in these reactions, is of less Weinheimisni commercial importance than the excessive activity sometimes exhibited by aluminum trichloride which tends to catalyze undersirable side reactions. water putine teith There is also an advantage in industrial processes in using a catalyst of nceitarinnar not only, uniform activity but of constant physical state, preferably the liquid state, since in this way the probability of poisoning or coating e t id in Holms the catalyst is aiminished. Mixtures of aluminum trichloride and metal chlorides such as antimony trichloride (SbC13), titanium tetrachloride, or tin tetrachloride are active liquid catalysts which have beer. used in such applications. Solutions of alwninum trichloride in ethers, or ketones, or alcoholic solvents do not have sufficient catalytic activity. Since zir- conium tetrachloride is considered a fairly good Friedel-Crafts catalyst, suitable combinations of aluminum trichloride and zirconium tetrachloride may oifer useful properties in relation to catalytic activity and to the chemical and physical state of the catalyst. Furthermore, certain combina- tions of the two may increase the versatility of either catalyst so that a "tailor-inade" mixed catalyst may be feasible. 3. Protactinium Recovery A technique for the recovery of protactinium-233 (Fa 23.), which is one of the products of neutron irradiated thorium tetrachloride, has been proposed which depends on the volatility and extractability of the protac- tinium-233 by aluminum trichloride. In this process a mixture of thorium tetrachloride and aluminum trichloride is irradiated, or a fluid stream of aluminum trichloride is passed over thorium tetrachloride irradiated in the bianket of a nuclear reactor. The aluminum trichloride leaves the melt, carrying with it protactinium-233 from the finely divided thorium tetrachloride. It is surmised that the effectiveness of the removal of protactinium depends on the solubility of thorium tetrachloride in aluminum trichloride. Since a saturated solution of thorium tetrachloride in alu- minium trichloride in the temperature range of 200-400° contains less than 1 per cent thorium tetrachloride, methods of enhancing its solubility have been sought. Several physical and chemical properties of zirconium MHC tetrachloride, such as melting point and ionic nature on the Pauling electronegativity scale, are intermediate between thorium tetrachloride and aluminum trichloride. In addition, zirconium tetrachloride has a low cross section for the absorption of thermal neutrons. Thus the study of the binary mixtures of aluminum trichloride-thorium tetrachloride, thorium tetrachloride-zirconium tetrachloride, and aluminum trichloride-zirconium tetrachloride are indicated in that a ternary mixture may be suggested which would improve the practicability of the protactinium recovery pro- Ce88.? It was for these objectives that the present study of the aluminum trichloride-zirconium tetrachlorida system was undertaken. B. Previous Investigations of Similar Systems The system aluminum trichloride-titanium tetrachloride has been studied by I. S. Morozov and D. Y. Toptygen while B. G. Korshunov and co- workers have reported on the aluminum trichloride-thorium tetrachloride binary system. The solubility of aluminum trichloride in the low melting titanium tetrachloride is shown to be very low from -30°, the melting point of titanium tetrachloride, to approximately room temperature. Similar be- havior is reported for the ferric trichloride (FeC13)-titanium tetrachloride system. The everse situation is true for the aluminum trichloride-thorium tetrachioride system, in that very little thorium tetrachloride dissolves in the temperature range of the aluminum trichloride melting point, i.e., 193° up to 400° and higher. Solubility of aluminum trichloride in thorium tetrachloride at the melting point temperature of 760° is shown to be very small. Reports of the ferric trichloride-thorium tetrachloride system . . w 1 indicate very similar behavior.° Work on the aluminum trichloride-zirconium tetrachloride system, albo by B. G. Korshunov, " led them to postulate a broad liquid-liquid immiscibility region above 127° based on some rather sketchy and questionable data of melting of eutectic and monotectic mixtures. Li quidus data were not reported, apparently because of experimental diffi- culties related to the high vapor pressures developed by the binary at elevated temperatures. Thus their phase diagram 18 at least incomplete. On the other hand, I. S. Morozov and L. Tsegledi?? report that the ferric trichloride-zirconium tetrachloride system gives a simple binary diagram with a eutectic at 14 mole per cent zirconium tetrachloride and 300°. The superficially similar aluminum trichloride (A1C13)- selenium tetrachloride (SeC14) system (melting point of selenium tetrachloride is about 305°, vapor pressure is one atmosphere at 170°) has been reported by H. Houtgraaf et alla A 1:1 compound melting at 163.8º is shown with a eutectic at about 30 per cent aluminum trichloride. No region of liquid immiscibility was found. The selenium tetrachloride rich portion of the diagram is not reported, presumably due to the experimental difficulties with high vapor pressures. The same paper reports on the aluminum trichloride (A1C13)-tellurium tetra- chloride (TeC1y) system (melting point of tellurium tetrachloride is 227.9°) showing a 1:1 compound melting at 149.3°. The nitrosyl chloride (NOCI). aluminum trichloride (A1C13) system was also studied by Houtgraaf. This the more system is particularly interesting because it shows a liquid-liquid immi s- cibility gap in the high AlCl3 concentration region which extends from about *. that it is Ummi7 0,5 to about 10 per cent nitrosyl chloride. The immiscible phases were dis- tinguishable in this system by the rather dark yellow-brown color of one .. Imidt liquid phase. Is irus... -.. 12. W 1 i . The binaries of boron tribromide (BBr3) with tin tetralodide and tin tetrabromide (SnBrų) might be mentioned, 14 since boron 18 a homolog of aluminum, and the tetrahedral molecular crystal structure of tin tetraiodide 18 considered to be similar to that of zirconium tetrachloride. The se binaries ehow eutectics close to the lower melting boron tribromide par- ticularly with tin tetralodide where the eutectic composition is only several mole per cent boron tribromide. I, N. Belyaev, in his extensive review of binaries forming two liquid phases, lists a single Group III and IV binary, that of antimony trichloride-tin tetrachloride. It is interest- ing that this system possesses a wide immiscibility region similar to that postulated by B. G. Kor shinov et al. for the aluminum trichloride-zirconium tetrachloride system. In addition, a large number of systems involving aluminum trichloride and aluminum tribromide (A1Br 3) in association with Group I and II halides showing liquid-liquid immiscibility gaps of various widths are quoted in this review. The se data are of particular interest in terms of the discussion of the aluminum trichloride-zirconium tetra- chloride system which follows in that a nunber of properties of the se systems such as liquid-liquid immiscibility and eutectic composition are similar to the results reported in this study. CHAPTER II MATERIALS, EQUIPMENT, AND EXPERIMENTAL TECHNIQUES A. Materials Preparation 1. Aluminum Trichloride Aluminum trichloride was prepared from the elements since efforts to achieve adequate purity by resublimation of reagent grade material from commercial sources were not successful. Procedures used were classical, in that high purity (99.99 per cent) aluminum metal powder was charged to a multisection quartz vessel and was reacted with carefully dried chlcrine gas. The drying agent consisted of C.P. grade aluminum trichloride - the novel element in this preparation - thus minimizing contamiration from the drying agent. The composition of the aluminum trichloride obtained by this method is shown in Table I. It should be empha sized that extreme precau- tions are necessary in this procedure due to the high vapor pressure of aluminum trichloride at elevated temperature and due to the exothermic for- mation reaction. No difficulty was experienced when using adequately de- signed vessels and maintaining the reaction te perature at about 200° (the triple point of aluminum trichloride is at 192.7º and a pressure of 171.5 cm. Hg as dimeric aluminum chloride). Since aluminum trichloride is ex- tremely reactive, all handling was performed in a helium atmosphere dry box where the moisture was maintained at a level of 20-30 parts per million by EN ! volume. Further checks on the purity of aluminum trichloride were made with 2 the aid of emission spectrographic analysis, x-ray diffraction, and direct . - - e. --- thermal analyses. Table II shows a summary of the results of the spectro- graphic analyses. Since x-ray diffraction powder pattern techniques will only show gross impurities, such application was principally made later in the course of the research to determine possible compound formation of A1C13. ZrC14. Resublimed reagent grade aluminum trichloride showed evidence of the presence of the A1C13•6H20. Alminum trichloride derived from high purity aluminum metal showed a normal x-ray diffraction pattern. Melting points obtained by thermal analyses of the fresh material using calibrated chromel-alumel thermocouples and a Rubicon Type B High Precision Potentio- meter were within <£1° of the triple point temperature. (Freezing point temperatures were depressed due to subcooling). Fresh material was pre- pared from time to time since the aluminum trichloride slowly deteriorated even when exposed to the dry box atmosphere. 2. Zirconium Tetrachloride Material obtained by resublimation of reagent grade zirconium tetra- chloride was generally found to be inadequate. Starting with so-called -- - - reactor grade material* and employing repeated sublimations, a material was obtained which was considered adequately free of sulfur, carbon com- pounds, and oxygen impurities. Table I and II list results from chemical and spectrochemical analyses. X-ray diffraction analysis did not show any additional lines in the powder pattern. Melting points were sharp both on heating and cooling and gave 437º and 438°, respectively, compared to a . S 17 reported melting point temperature of 438 1º.'. *Supplied through the courtesy of U. S. Industrial Chemicals Campany. T . TABLE I ANALYSIS OF CHLORIDE SALTS Alci, 2rCl. . . 'W ' Melting Point, °C Observed Literature 192 192.7 437 438 Chemical Analysis, Wt % Cation Observed Theoretical Chloride Observed Theoretical 20.31 20.23 39.16 39.14 79.60 79.77 58.3 60.86 H20 Observed 0.07 -0.27 A1C12.98 (H20).02 Empirical Formula Zrc13.83 (H20)..01 - - - - ES- . 10 TABLE II SPECTROCHEMICAL ANALYSIS OF CHLORIDE SAITS Limit of Detection Limit of Detection Wt % ALC13 Wt % Zrcle Wt % Element Wt % Al 10-100 Ag 0.002 0.02 0.1 0.005 0.01 0.002 0.02 0.08 0.004 0.005 0.2 0.0001 0.001 0.0001 0.009 0.0004 0.0002 0.2 0.00006 ANA 8888 888 0.0003 0.09 0.006 0.005 0.00001-0.0001 0.0001 -0.001 0.000.1-0.001 0.008 0.003 0.003 0.008 0.04 0.007 0.18 0.14 1.5 0.01 0.5 0,0001 -0.001 Мо 0.007 0.02 0.2 Nb 0.0007 0.002 0.2 0.04 0.0005 0.2 0.01 0.0003 III I II III IIIT Na 0.02 0.44 0.05 0.02 0.003 Ru SO 0.03 0.4 0.02 0.01 0.001 -0.01 0.001 -0.01 0.02 0.2 0.1 EBS SS 0.003 0.08 0.09 0.002 0.12 0.001 0.09 0.02 0.007 11 111 0.02 0.7 . ? - 10-100 diante sott NN . 12 . B. Experimental Techniques Many types of experimental equipment and techniques have been used for the study of the phase behavior of fused salts. Several of these were applied in this study – the principal one being differential thermal analy- sis (DTA). This method, which may be considered a refinement of thermal - analysis, consists of comparing the temperature of the specimen under study with that of a suitably selected reference material of similar physical properties while both are heated simultaneously at a controlled and con- stant rate. For the purpose of determining melting points for the charac- terization of the pure components, the somewhat less sensitive but probably more accurate measurements by conventional thermal analysis were favored. Where necessary, the study of some regions of the pha se diagram wak supplemented by direct visual observation of the di sappearance of the solid phase. Due to the volatile nature of the salts, mixtures were sealed in quartz tubes of suitable design to withstand the vapor pressure generated. These were immersed in a transparent bath of a molten salt eutectic mix- ture. Direct viewing of liquidus points was particuarly useful in the low to intermediate concentration region of zirconium tetrachloride. To explore the possibility of compound formation, x-ray diffraction analysis was applied to a number of aluminum trichloride-zirconium tetra- chloride mixtures at several elevated temperatures to detect any changes in phase or composition. Conventional x-ray powder patterns were obtained and analyzed. 1. Differential Thermal Analysis The technique applied in this study was differential thermal analy- sis. This method has been used for a number of years in the earth sciences . 1 and recently has received wider application in the study of fused salts. Review articles and books by Smothers and Chiang 18 and W. Wendlandt19 are excellent sources. Generally the technique 18 to encapsulate a specimen and a physically similar and chemically inert reference substance. Both are placed in a relatively large heat conducting cylinder of stainless steel. The subassembly is symmetrically located inside a furnace which is heated at a constant rate. Any changes in phase causes a change in the heating rate of the specimen generating a temperature difference against the reference capsule. The differential temperature 18 recorded on an x-y plotter against either the temperature of the reference or the tem- perature of the specimen capsule. Whereas, curves obtained in conventional thermal analysis show the transition directly as a temperature halt, the analysis of curves obtained from differential thermal analysis in relation to the initiation or termination temperature of a transicion is less direct. Since peaks have finite temperature widths, the help of the pure substances as references are useful in the determination of transition temperatures. C. Experimental Equipment The experimental equipment used in these studies was originally de- signed by T. B. Rhinehammer and co-workers.cc Figure 1 is a photograph of the assembly. The furnace shown to the right is a 3 in. diameter, 12 in. long resistance heated, 1400 watt unit manufactured by the Hevi-Duty Electric Company. A model 9835-13 Leeds and Northrup D.C. microvolt ampli- fier is shown at the top of the instrument rack. This instrument is used to amplify the differential thermocouple signal providing increased sen- sitivity, and to prevent interaction in the x-y instrument by isolating the v M -- 13 — .- 14 & - .244 im . . ATA . monum nin n i . # 'S . > 27 7 oti . cias...adies Nit mediante i Ah nviartieto . i i : *7 f , . A 7 .. . ALT . 13 1 2. in the . . A . UU . " . > Figure 1. Differential Thermal Analysis Equipment. .:. 14 temperature thermocouples thus permitting the use of only two capsule thermocouples. See Figure 3 for the therwocouple circuit. The x-y plotter which records differential temperatures (horizontal) versus temperature (vertical) and the furnace temperature recorder are showrs mounted beneath the amplifier. The programmer furnace controller 18 located in the lower- most section. Figure 2 shows the shell of the chamber which contains the experimental capsules. The top of this chamber can be seen just protruding from the furnace in Figure 1. On the left hand side of Figure 2 18 shown the sealing flange which supports the 1.75 in. diameter cylindrical capsule block. A capsule centered with a magnesium silicate ring 18 shown inserted in a 0.5 in. hole in the block. When used with reactor materials of low volatility, the block is placed into the shell and the chamber may be either evacuated or supplied with a special atmosphere. Since the aluminum trichloride and zirconium tetrachloride salts are chemically active and very volatile at the lting points, evacuated and sealed nickel capsules designed to withstand very high pressures were used. A closeup photograph of these capsules as well as a quartz capsule used in direct visual studies is shown in Figure 4. A construction drawing for the nickel capsule show- ing welding details is shown in Figure 5. A diagrammatic representation of the entire apparatus is shown in Figure 3. Nickel was selected for a capsule material as a result of tests which indicate.!. good resistance to corrosion to the chloride salts at ele- vated temperatures. In addition, nickel possessed the strength to with- stand the high pressures developed in these experiments. The capsules were de igned for a working pressure of 2070 pounds per square inch a : 427°C. However, it should be noted that the allowable pressure diminishes rapidly TIT . 15 - M E !-s .. 1 ' - C 1. . - - - 14 Y I. . " S: . P " Prinsen < i TY . €, * . * 1. ! ! s - . . . Y ' . T . TP. . . , W '. - .. - TW - . L . ., : . .. . wami 1 2. . . " V # -.. *** . - 3 R 1 :. VM - 1 !* ' . . ". - isi .. . * .: * *** * 1 . 1 ** X * 1 Y . .1 2 . I i 1 . I .1 M . + . A . . . - ** T . .' . . ' . 1 , . 22 _ . 4 ' ' . no. 1 '. . 2 . , 1.1 . . .. : ) ' j y ST SU ! - 2 - - . K ... I ... $ WMV WYN - TV * ! . T w . STTAL " . TAYO 10 . 1 CV.UZ . ! PASS 2. . . * 1 . LAULT . *** . , . . " 2 ?, . KY ? The human 1 . . e Ti ty: A .. . . . La , . . . CA . .. : : . VE . WA TY . **: : RE 4 . 7 vy . . 1 . . - . X 4 . - - . 312 . . . = , " . - Figure 2. Differential Thermal Analysis Capsules Holder and Shell. 00 : :: RECORDER c UZ a O . 16 . DC-MICROVOLT AMPLIFIER (AT) 0 2 FURNACE HEATING PROGRAMMER Figure 3. Schematic of Differential Thermal Analysis Equipment. I 1 ,- ' 2. Pa - 17 ' . . - L " - - . . . - * .. - . in. : 1.1. 1. - - - - - - - . P - - . ; 4 ." ta EX K T 1 . AY, .. . .1 A Wh 4 TN + . A - 6 . UN ) * * .! .. - ::. · . .!" . (* . . . . . . - . IV. . ! . 7 * LEW 1. ! pl . " . . . C v + . i SO . . . s', ! w . AAC . " * . .. , 7 E 12. WA AT . C ." . * S 1 1 TA W . 2 W ! . . ** . NAK - AU . A 14 :: * . 1 / . 1 P I Y " ... . ' . ' Figure 4. Nickel and Quartz Specimen Capsules. ODNOTE-1 NOTE-1) 1 Q IT NOTE-1) IQ 9d; K'NOTE 4 NOI, NOTE -1 CAPSULE ASSEMBLY NOTE - 1: PS8, CLASS 3, SCHED C 2/16 in. 300 TRE 5/8 in. 3/8-in. * 0.035 - in.- WALL NICKEL TUBE (1 REQD) 0.093 $0.002 in. 0.035 in. le 0.030 in. 77 13732 in. 1 300 ندددددددددددددلداسكتدبير Vo in. تتتتتتتتتتتتتتتيننتشلنلنلندا WELL, Va-in. NICKEL. TUBING (1 REQD) 1.0.080 in. NICXEL PLUG (1 REQD) SLIP FIT WITH TUBE Ile kos in. NICKEL END PIECE (2 REQD) . Figure 5. Construction Drawing of Nickel Capsule. -------... ......we se sentiam 19 - - 3 to about 850 pounds per square inch at 538° due to the rapid decrease in yield strength of nickel at elevated temperature. Allowance in the design was made for the presence and the nature of the welds. The thermocouple well was located along the axis of the cylinder and extended to about; one- half the length of the salt column. Sufficient material was charged to a capsule to minimize internal gas volume after melting so that composition changes were negligible. This required considerable free volume for aluminum trichloride rich mix- tures due to the change in density of aluminum trichloride on melting from about 2.4 to 1.3 8./cm3. All transfers of material were performed in a special atmosphere dry box in a helium (He) atmosphere with moisture main- tained between 20-30 parts per million by volume. The capsule filling tube was heliarc welded while mechanically sealed insuring the integrity of the sample. Similar treatment was afforded the aluminum oxide (A1203) reference capsule. In performing an experiment, the usual procedure was to place the reference capsule and specimen capsule in the cylindrical block. Cali- brated stainless steel sheathed, chrome).-alumel thermocouples were then - inserted and secured in the capsule wells and the block was placed into " the furnace chamber. The thermocouple network was plugged into the ampli- more word-wrapsis fier-plotter units, and the furnace programmer was adjusted for the rate of heatup and cooldown desired, generally, 1 to 2°/min., the former being the lowest rate available. Additional adjustments for maximum operating temperature and for length of time at maximum temperature or "soak" time were also made at this time. The "soak" temperature which was held for ministre, mwanamke min mente DRY . While the four hours was selected to exceed the liquidus temperature by about 50°. someti at OT 7. 7*** ***** 11. 1. If the r'un represented an original heatup of the mixture, an eight-hour halt at maximum temperature was programmed. Invariably, the first run was discarded. A number of hea tups were made, both for equilibration purposes and to determine reproducibility for any given cycle. Prior to startup the "y" or temperature axis of the x-y plotter was calibrated with a Precision Leeds and Northrup Potentiometer. The "x" axis on which the differential signal was plotted was checked less frequently. A second check on the "y" calibration was often made following a run to ensure freedom from drift over the approximately twenty-four hour running period. D. Interpretation of Differential Thermal Analysis Curves Figure 6 shows a representative selection of differential thermal analysis heating curves recorded at several different compositions of aluminum trichloride-zirconium tetrachloride mixtures. Differential tem- perature is shown plotted in the vertical direction versus the temperature -1.- god of the winn of the aluminum oxide reference. The practice of plotting the reference in temperature rather than the specimen temperature along one axis was pre- : I m ferred in these studies since comparisons with other work in the litera- ture is facilitated, and since it was found that more consistent and contrariisi interpretable data were obtained. When the melting of a pure substance occurs, the differential thermal annan decipria analysis curve starts to deviate from the flat baseline while the difference i en in temperature of the specimen increases as in Curve a in Figure 6. The og Lime tus graph rises at a rate related to the rate of hearing and the physical pro- entendirawatan mana, perties of the material. When the pure sample is finally melted, the Watu haw.com Videoer . " Line . . . .. Totantotan katossats 2 Trend, 1 "TWT 1 . } _ _ . 1. 12 → be O PURE AICIZ DIFFERENCE TEMPERATURE 6 25 % ZrCla 21 Ky C 67 % Zr614 1 1 . - INCREASING REFERENCE TEMPERATURE : Figure 6. Differential Thermal Analysis Heatup Curves. O , 17 . . * * WWW . M . . Y - ., . 1 . --* 22 - -*: . - * - . - '. . specimen capsule temperature "catches up" to the reference tem nrature and - - . . * . .'.* . ' ' the curve returns to the horizontal base.line. - - ' ' -'9 -'1 .- ... . FI In the case of simple mixtures, a peak 18 first obtained which rep- resents the melting of the eutectic mixture. Since melting continues to occur after pacsing the eutectic temperature, the curve does not return to the baseline until all solids has melted. This may be seen as a second peak or as an abrupt change in slope of the differential thermal analysis curve as in Curves b and < in Figure 6. Smyth" has attempted a theoretical analysis of a differential thermal analysis system to determine the significance of the various portions of the differential temperature versus temperature curves. He computed the shape of the peaks by analyzing a mathematical model consisting of a thin -- . LI slab of material subjected to a linear heating cycle while undergoing an ----...---.-~*" endothermic reaction. For thermocouples symmetrically located in the mate- * rial, the temperature at the apex of the differential temperature versus . temperature curve indicated the transition temperature of the specimen if the differential temperature was plotted against the specimen temperature. nara If the differential temperature was plotted against the reference tempera- ture, the transition temperature for the endothermal peak occurred on the curve, between the horizontal baseline, which is traced during the steady -- state period of the run, and the apex. :- . It was further suggested by Schraid" that if the heating rate is sufficiently slow to produce a straight-line portion during an endothermic ina sainted stickatkaradadadadadamonhan setantanetNSARKLAG lastnina tidak selalu me reaction, the extrapolation of this line to the baseline will give a good mea sure of the transition temperature. It was found in this study that . TE such a procedure gives values for the melting point of the pure substances M . - . . . - ---- - - - - - . which most closely and consistently correspond to accepted values. However, . . - ' 7 5 - - small corrections to the transition temperature were necessary in practice * * . .'. .* -* ' -'? since situations which only approximate reality are used in developing the .:. V V mathematical models. National Bureau of Standards lead was used as a fi - - .. . standardization material in addition to lithium chloride (101) and sodium chloride (Naci). These samples gave temperature halts wheil using conven- tional thermal analysis, which were within +1° of the accepted values. - - - , . . As seen in Figure 4, some curvature was encountered on the rise of . the curve obtained in the analysis for mixtures. In this case the slope of the leading edge of the curve at its inflection point was used for pur- poses of back extrapolation. Curve c in Figure 6 shows a liquidus transi- tion fol.lowing a eutectic melting. In this case, the differential tempera- ture does not return to the baseline after the initial melting, as with a pure substance, but continues a gradual return depending on the shape of the liquidus line of the pha se diagram. The curve shows a discontinuity in slope when all the solid is consumed. The liquidus temperature is taken close to this break. This interpretation was confirmed by comparing the more abrupt changes on freezing in cases where supercooling did not occur. However, it should be emphasized that generally cooling curves were not useful in this study for purposes of discovering transition temperatures, .هههه مد ممسلوقمتممنننهمين particularly in the aluminum trichloride rich section of the phase diagram. At high zirconium tetrachloride concentrations and for pure zirconium tetra- chloride, liquidus transitions for both heating and cooling were reproducible within 1 to 2 degrees. The supercooling occurring in the aluminum trichlo- ride rich section is discussed in more detail below in relation to the in- terpretation of studies by Korshunov." on the aluminum trichloride-zirconium نحمننهمننننننننملنمنننهنممنهد:تابع ننش tetrachloride system. WY V KA. UN Curve b in Figure 6 shows a differential thermal analysis curve with a double peak. Such shapes were obtained in a part of the aluminum trichlo- ride rich section of the phase diagram. The second peak represented, as will be discussed below in conjunction with the phase diagram, a change from pure zirconium tetrachloride in equilibrium with liquid solution to an equilibri.um with a two phase liquid mixture. Thus the second peak rep- resented an abrupt change between a rather low rate of energy absorption to a high rate, noth resulting from the melting of' zirconium tetrachloride. The apex of the second peak was interpreted as the temperature of the mono- tectic transition. Attempts to observe these transitions by direct visual methixds generally gave scattered and rather high results averaging about 50 above the temperatures represented by the second peak. These are discussed below in the section on experimental observations. E. Supplementary Experimental Techniques 1. Thermal Analysis In order to standarize observations of transition temperatures made by differential thermal analysis, it was necessary to determine melting points of the pure components by an alternate method. Although differential thermal analysis is a more sensitive means of detecting transition tempera- tures, the relative lack of ambiguity in conventional thermal analysis was a factor in using such procedures for standardization. The pure components were loaded into nickel capsules identical to those described for differential thermal analysis observations and the se were placed into a cylindrical furnace similar to that described above. Calibrated chromel-alume I thermocouples were read with a Rubicon Type B 12 . W 25 High Precision Potentiometer. Readings of temperature versus time were also recorded continuously wilth a suppressed zero recorder having a 3 mv. range over a 10 in. scale. Heatup rates were about 1°/min. The potentio- meter was read every minute when going through a transition. These points were subsequently graphed manually. The latter method gave very satisfac- tory temperature halts which compared well with literature values. In thium chloride, sodium chloride, as well as aluminum trichloride, zirconium tetra- chloride, and thorium tetrachloride, were studied in this manner. A stan- dard sample of lead from the National Bureau of Standards, melting point 327.40°, was also used for the calibration of the differential thermal analysis apparatus. Direct Visual Methods Since very high vapor pressures developed in the se salt mixtures, the application of direct visual methods were necessarily limited. The quartz capsules shown in Figure 4 which were similar in size to the nickel capsules used for the differential thermal analysis tests were capable of withstanding temperatures and pressures corresponding to aluminum trichlo- ride heated to 250° or about 10 atmospheres. Capillary tubes measuring * 5 mm. outside diameter and I mm.. wall thickness easily withstood pressures in the region of the aluminum trichloride critical temperature or about 29 atmospheres at 350° and were thus useful for the higher temperature ranges. The tubes were immersed in a lithium nitrate (LiNO3), sodium Lia nitrate (NaNO3), potassium nitrate (KNO3) eutectic contained in a 1 liter * beaker. The melt was very transparent and practically colorless, remaining Convenient quantities for a one liter bath are: 404 g. LINO3, 189 g. NaNO3, and 752 8. KNO3. The density of the melt ranges from 2.4 to 2.25 8./cm3.23 26 molten to below 150°. The bath was stirred with a stainless steel stirrer and heated with a hotplate and an immersion heater. Glass wool insulation surrounding the beaker was required to reach higher temperatures. After some experience with this arrangement, it was possible using variable transformers to adjust the power input to give a reasonably slow and con- stant temperature rise. The capsules were tied to glass rods which were supported overhead. Chrome.l-alumel thermocouples enclosed in 310 stainless steel sheaths were used to measure temperature. Mercury thermometers were not useful since attack by the salt was sufficient to quickly alter the calibration. Temperature gradients, with even a low rate of stirring, at reasonable distances from the heaters were negligible. The capsules were shaken frequently and occasionally removed from the bath and quickly up- ended and returned. Very good agitation was thus attained. The highest temperature attempted in this setup was 380° which is the critical tempera- ture of a 14 mole per cent mixture of zirconium tetrachloride and aluminum trichloride. .ver... oni standarta sme miento en met versteenvender com a D de a Sal i miner T . . .! CHAPTER III DISCUSSION OF EXPERIMENTAL RESULTS Figure 7 shows the solid-liquid binary pha se diagram of the aluminum trichloride-zirconium tetrachloride system throughout the range of composi- - en v tion. A magnified portion of the system in the low zirconium tetrachloride * . - . concentration range is shown in Figure 8. The data are plotted as tempera- . . . ture in °C versus mole percentage of zirconium tetrachloride. For the sake . -.,.. - - - of simplicity, monomeric aluminum trichloride is used as a basis for graph- ing. The open circles on the diagrams represent information derived from ,-Liniers. - es differential thermal analysis experiments while the solid circles are data r obtained by the direct viewing of melting in quartz capsules. It was wie cor necessary to use visual observations in the 10 to 25 per cent zirconium tetrachloride region of the diagram because the very small changes in curva- ture of the differential thermal analysis trace due to the steepness of the liquidus curve in this range rendered precise interpretation difficult. The smooth merging of the liquidus data from the differential thermal analysis observations and those by visual observations may be noted on the diagram. Direct visual observations of eutectic and monotectic temperatures on heat- - .: معموه سمنهدينهم سوفت ing were obscure and difficult to reproduce. Differential thermal analysis data was shown for those transitions. Data obtained on cooling are shown as solid squares. Supercooling effects on the freezing point of the eutectic mixture are readily apparent. At zirconium tetrachloride concentrations تتحد مختممننلهلننغدوتح greater than ~33 per cent, supercooling of the liquidus transition was . 27 ت د --- ... - OOTA - MELTING • VISUAL - MELTING U OTA-FREEZING .............. TEMPERATURE (°C) COL 28 La + Zrcla ................ “2 f... .......... .. .. ...... iLothayo ...... ... ...**** . Lo + ZrCla Lola _ Li+AICIZ 125 Lt AICIE 10 20 30 40 50 60 ZrClA COMPOSITION (mole %) 70 80 90 Zrcia Figure 7. Binary Phase Diagram, Aluminum Trichloride-Zirconium Tetrachlcride System. X X O DTA - MELTING • VISUAL - MELTING • DTA-FREEZING Lit L - -*- Was L2 + ZrCla TEMPERATURE (°C) A=DATA FROM B.G. KORSHUNOV 10 10 25 30 35 15 20 Zrcla COMPOSITION (mole %) - - . . . Figure 8. Aluminum Trichloride Rich End of the Aluminum Trichloride- Zirconium Tetrachloride Binary Phase Diagram. . . .. . . . . ..... ........- . JJ N T " 1 . LL MV WWII WT . 7 . 30. relatively small. Throughout the entire range of concentrations super- Data obtained by B. G. Korchunov et, a1.10 for the aluminum trichlo- ride-zirconium tetrachloride system are shown in Figure 8 nb crosses. These which represent supercooled freezing temperatures of the binary liquid mix- tures. In Figure 7 in the region of ~10 to 100 mole per cent zirconium tetrachloride, complete miscibility of the liquid mixture 18 postulated, while in the region of ~3 to 10 mole per cent zirconium tetrachloride a narrow two phase section is shown with a monotectic caposition of 10 per cent zirconium tetrachloride at 192º. An aluminum trichloride-zirconium tetrachloride eutectic is evident at 2.5 per cent zirconium tetrachloride and 170°. The occurrence of a monotectic and eutectic separated by only ~22° results in a narrow temperature range where pure zirconium tetrachlo- ride is in equilibrium with liquid l. At temperatures directly above this --- - band, zirconium tetrachloride is in equilibrium with liquid 2. Attempts -- - . . . to determine the limits of the region of immiscibility as a function of . . temperature or the upper critical solution temperature by differential -- thermal analysis and direct observation were unsuccessful. Resolution of . the eutectic and liquidus transitions was not possible on the aluminum tri- : chloride rich side of the eutectic composition either by differential thermal 46 . 3. Protactinium Recovery '7r. - The use of zirconiun tetrachloride as a "bridging solvent," between the aluminum trichloride and thorium tetrachloride was investigated by studying the three binaries involved. In the temperature range of interest, 200-350°, which is considered suitable for a process for the removal of - 1 .3 protactinium from neutron irradiated thorium tetrachloride, the solubility * - : - of thorium tetrachloride in aluminum trichloride was found to be less than . L ** 1 per cent while the solubility of zirconium tetrachloride in alminum tri- chloride was found in the present study to range from 10 to 40 mole per cent. However, the investigation of several zirconium tetrachloride- W thorium tetachloride mixtures indicated very little change in the melting nej , ja.« Widfimis. temperature of zirconium tetrachloride. The evidence from the study of the three simple binaries of the aluminum trichloride-zirconium tetrachloride- thorium tetrachloride system suggests that the possibility of finding a homogeneous ternary composition of appreciable thorium tetrachloride con- ::.-..*nusten dir centration is remote. in - * 1. . -' .*. . . A L- .. . BIBLIOGRAPHY BIBIIOGRAPHY . - - - 1. W. Biltz and A. Voigt, 2. Anorg. Allgem. Chem., 126, 39 (1923). 2. W. Blitz and W. Klemm, 2. Anorg. Allgem. Chem., 152, 267 (1926). . -.- *-...- . -- sports. .. 3. K. Sesuari, Acta Phys. Acad. Sci. Hung., 9, 195 (1958). 4. W. B. Blumenthal, "The Chemical Behavior of Zirconium," D. Van Nostrand Company, Inc., New York, N. Y., 1958, p. 108. ....no . . --***:*::* George A. Olah, "Friedel-Crafts and Related Reactions," Vol. I, Interscience Publishers, New York, N. Y., 1963, p. 263. L...i lonte .-** 6. George A. Olah, "Friedel-Crafts and Related Reactions," Vol. I, Interscience Publishers, New York, N. Y., 1963, p. 244. . *,*!.. - ..imagine a '. 7. A. J. Shor and E. L. Compere, "Salt Separation Chemistry of Protactium From Irradiated Thor: um," Reactor Chem. Div. Ann. Progr. Rept. Jan. 31, 1964, Oak Ridge National Laboratory, ORNL-3591, p. 56. -- * na... 8. I. S. Morozov and D. Ya. Toptygin, Zhur. Neorg. Khim., 2, 1917 (1957). - * -- 7 * 9. B. G. Korshunov and v. I. Ionor, Izvest. Vysshikh Ucheb. Zavadenii, Tsvetnaya Met., 6, 114 (1960). . * L * * 10. B. G. Korshunov, A. M. Reznik, and I. S. Morozov, Trudy Moskov. Inst. Tonkoi Khim. Tekhnol. im. M. V. Lomonosova, 7, 127 (1958). . ll. I. S. Morozov and L. Tsegiedi, Russian J. of Inorg. Chem., 12, 1397 (1961). .:-, mint 12. H. Houtgraaf, H. J. Rang, and L. Vollbracht, Rec. Trav. Chim., 72, 983 (1953). 13. H. Houtgraaf and A. M. De Roos, Rec. Trav. Chim., 72, 963 (1953). . a**. Imo.. mari..nn.-:.. 14. R. F. Adamsky and C. M. Wheeler, Jr., J. Phys. Chem., 58, 225 (1954). 15. I. N. Belyaev, Russian Chem. Rev., 29, 428 (1960). **, prinsipi.md .. 16. R. E. Kirk and D. F. Othmer, "Encyclopedia of Chemical Technology," Vol. I, The Interscience Encylopedia, Inc., New York, N. Y., 1952, p. 632. 17. "Gmelins Handbuch der Anorganicher Chemie," Achte Auflage, Zirconium, System - Nummer 42, Verlag Chemie, Gmbh., Weinheim/Bergstrasse, 1958, p. 285. 18. W. J. Smothers and Y. Chiange, "Differential Thermal Analysis, Theory and Practice," Chemical Publishing Company, New York, N. Y., 1958. re mo ity invisibi 48 ASR 49 19. W. Wendlandt, "Technique of Inorganic Chemistry," Vol. I, edited by H. B. Jonassen end A. Weissberger, Interscience Publishers, New York, N. Y., 1963, p. 209. T. B. Rhinehamer, D. E. Etter, and L. V. Jones, Paper No. 27, Inter- national Conference on Plutonium Metallurgy, Grenoble, France, April 19-22, 1960. 21. H. T. Smyth, J. Am. Cer. Soc., 34, 221 (1951). 22. W. J. Smothers and Y. Chiangc, "Differential Therinal Analysis, Theory and Practice," Chemical Publishing Company, New York, N. Y., 1958, p. 106. 23. Personal communication with W. L. Marshall, Oak Ridge National J. H. Hildebrand and R. L. Scott, "Solubility of Non-electrolytes," 3rd ed., Reinhold Publishing Corp., New York, N. Y., 1950, p. 13. 25. A. A. Palko, A. D. Ryon, and D. W. Kuhn, J. Phys. Chem., 62, 319 (1958). 25. .. J. H. Hildebrand and R. L. Scott, "Solubility of Non-electrolytes," 3rd ed., Reinhold Publishing Corp., New York, N. Y., 1950, p. 121. * - 27. George A. Olah, "Friedel-Crafts and Related Reactions," Vol. I, Interscience Publishers, New York, N. Y., 1963, p. 242f. 28. J. Kendall, E. D. Crittenden, and H. K. Milier, J. Am. Chem. Soc., 45, 963 (1923). 29. J. H. Hildebrand and R. L. Scott, "Regular Solutions," Prentice-Hall, Inc., New Jersey, 1962, p. 143. 30. N. K. Druzhinina, Tr. Vses. Nauch.-Issled. Alyumin.-Magnievyi. Inst., 50, 143 (1963); C.A. 59: 9382h. 31. W. M. Smit, "Purity Control by Thermal Analysis," Elsevier Publishing Company, Amsterdam, 1957. . VITA r.- The author was born on June 10, 1923, in New York City, New York. He received his early education in the public schools of that city and was graduated from the College of the City of New York with a Bachelor of Chemical Engineering degree. After employment at the Argorine National Laboratory, he transferred to the Oak Ridge National Laboratory in 1958. He is married to the former Roberta Sue Wagner and has two sons, Andrew Benjamin and Joel Thomas. t " . Air I- mitment Store pika wanithan Aitorejas tormentera 50 BE DATE FILMED 2 / 24 /65 1 . . ! wiwiteite r LEGAL NOTICE 7 TE . . . Es . 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 accu- racy, completeness, or usefulness of the information contained !n 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 am- ployee or contractor of the Commission, or employee of such contractor, to the extent that such employee or contractor of the Commission, or employee of such contractor prepares, disseminates, or provides access to, any information pursuant to his employment or contract with the Commission, or his employment with such contractor. - . I is 7 .' Llimianie . . PU . . . 1 C. TO L . . ic. . 1 9 SN' t. 1 U . 23 2 END comme