m.“- p»;- 3.... ad. .u. 3 m .4. l.'.h an...“ ..a m cg W.- .1. 3.... a. mum» 3 37 ’9’... T'HESis \ . JUBRARY 3 ‘ h ‘ 2’ a ‘ 3:: 13 A'Iwhf'gfln Sta '2: 1 IV ‘ ‘43,, ,I. J‘ | ‘ u 1: \ ‘ ‘ I c L o - O 4 ' I ‘ .‘ i ‘ f ' V - __ ___ L . ‘4 A, .'A 4%-- I/ ‘ _AAL-a- &..1 ‘ r ‘ ‘ ‘ L- L. a o’- . c. I M.” V o‘ . 1 | - " : D ' I ' ‘ q l I W \ I ’ f I I _ 4 ‘ ‘ l J' , . ‘ l ‘- 0‘ . ‘ I Q u... ' ' - ‘g . \ - - ‘- I Q | I ‘ I ‘ l o a ‘ J I 0 ~ TITAI'IIUH HALI DES By Karl 3. Yarro- A THESIS albmitted to the School of Graduate Studies of Michigan State College of Agriculturo and Applied Selena. in partial mlfillment of the requirements for the dean. of MASTER OF SCIENCE Department of Chemistry 1953 ---’"-Cm"70.LEm:333~:T The author wishes to extend his sincere thanks to Dr. Frederic B. Button under whose inspiration, constant supervision and unfailing interest this work was carried out. He is also greatly indebted to Robert Vendor Vbnnen for his assistance in taking the magnetic susceptibility measurements. He also wishes to express his appreciation to the members of the faculty and staff who have given advice and encouragement tram time to time. II III Iv VII VIII TABLE 93 cormms Introduction Historical Known Compounds Which Could Be Used In The Preparation 0f Chlorofluorides Experimental Procedure A. Titanium.Trichloride B. Titanium.Trifluoride C. Titanium Trifluorochloride D. Titanium Tetrefluoride 3. Titanium Oxyfluoride ‘x-ray'Powder Diffraction Studies Magnetic Susceptibility Studies Discussion Summary Literature Cited PAGE 12 14 16 19 25 31 32 FIGURE I II III TABLE II III IV Capper Autoclave Titration Curve of Titanium.Trichloride Solution Manifold for Distillation of Anhydrous Hydrogen Fluoride LIST 3; TABLES eray Powder Diffraction Date for Titaniun.Trif1uoride X-ray Powder Diffraction Data for Titanium Oxyfluoride Structure rhetors for Titanium Oxyfluoride X-ray Powder Diffraction Data for Titaniukaetrefluoride Magnetic susceptibility Data 11 13 PAGE 21 23 24 27 irrnonucrron This study was concerned with certain chlorides and fluorides of titanium. The tetrafluoride is a solid under ordinary conditions, subliming at 284' C. The electronegetivity difference for the Ti-F bond is 2.4 which corresponds to about 55; ionic character." The thermal stability and non-solubility in organic solvents also indicate ionic character. The tetrachloride is a liquid, bop. 136' C. The olectronegativity difference for the TieCl bond is 1.4 which corresponds to about 30% ionic character hence titanium tetrachloride is commonly recognized as a covalent type of compound. This study was undertaken to synthesize a compound intermediate between these two types, noting its properties, and perhaps observing something of the nature of the transition from ionic to the cOValent type of compound. The possible intermediates are titanium trifluorochloride, titanium difluorodichloride, and titanium trichlorofluoride. A compound, approximately titanium trifluorochloride, has been reported by Hurt and Ipaan.2 As yet titanium difluorodichloride and titanium trichlorofluoride have not been reported. It was preposed to prepare titanium trifluorochloride by chlorinating titanium trifluoride. II " HI STORICAL Two different mixed halides, titanium dichlorodibrmide and, titanium trichlorobrauide were reported by Friedel and Guerin in 1890.3 They are liquids boiling at 176' C. and 154' C. respectively. Reader has shown by means of cryoscopic and ebullioscopic investigations that the properties of'mixtures of titanium tetrachloride and titanium totrabromioe correspond to the properties of titanium dichlorodibromide and titanium trichlorobromiCe reported by Friedel and Gusrin.‘ Therefore there is doubt that the mixed halides are individual compounds. KNOWN COMPOUNDS ’s‘v'HICH COULD E: U331) IN THE Z’REPARATION 93 CHLOROFLUORIDES III - KHOTN COfiPOUHDS WHICH COULD BE USED IN THE PREPARATION OF CHLCRO FLU 3111333 Titanium.trichloride has been prepared in solution or in the anhydrous room in a number of ways. Titenous chloride solutions have been prepared by the action of a suitable reducing agent on e solution of the tetrachloride. Linc has been used quite extensively for this purpose. The use or the zinc reductor in titanium analysis is an application of this reaction. Electrolytic reduction is another method of preparation.5 Two general methods have been used for the preparation of the anhydrous trichloride. Hydrogen gas will reduce titanium tetra- chloride Vapor in the mrnace.5 The second method employs metals such as antimony, aluminum, magnesium, zinc, arsenic, tin and silver which will reduce titanium tetrachloride at elevated temperatures to yield titanium trichloride and the chlorides of the metals used, a number of which are volatile.7 Titanium dichloride has been prepared by heating titanium trichloride in the furnace in a hydrogen atmosphere.6 Titanium trifluoride can be prepared by treating titanium trichloride with hydrogen fluoride, or by the reduction or potassium hexafluotitenete end leaching potassium fluoride from.potaesium penta- fluotitunite.8 Titenium.difluoride has been reported,9 but this preparation has been refuted by further studies. errzninnnrli Pgocsnuan ‘ \_‘_ Iv - rocpmr'mm. rrzoczamro: The general method chosen for the preparation of the chloro- fluorides of titanium was as follows: 1. Reduce titanium tetrachloride to titanium trichloride 2. Replace chloride ions with fluoride ions to obtain titanium trifluoride 3. Chlorinato titanium trirluoride with chlorine Titanium tetrachloride obtained Iran the Stautter Company was redistillcd over copper turnings and calcium hydride to remove any excess chlorine, water and other impurities. Titanium tetrachloride could not be used for reduction in pure water since it hydrolyzes readily. It can be dissolved in cold concentrated hydrochloric acid and then diluted to obtain a solution of titanium tetrachloride in hydrochloric acid and water. A -‘PRSPAT’TION 0F TITANIUE TRICELORIDE The solution of titanium tetrachloride may be reduced by: (l) the use of zinc and hydrochloric acid; (2) a zinc reduetor; (3) electrolytic reduction. The use or granulated zinc and hydrochloric acid tended to giro excessive amounts of zinc salts as impurities in the titanium.trichloride solution. A sine reductor was then constructed following directions in the text by Cumming and Kaytlo A reducing column about one inch in diameter and ten inches high was used in some of the earlier preparations with considerable success. It had two drawbacks: (1) some zinc salts which would have to be removed later were always introduced into the solution; (2) the reduction was not complete unless the titanium tetrachloride solution was allowed to stand for quite some time in the reductor. I The last method used was electrolytic reduction, accomplished by an adaptation of the method given by halton.5 A.two liter resin Jar, with a four hole ton was fitted with a large lead cathode, which had been amalgamated with mercuric chloride. The anode, a graphite electrode small enough to permit chlorine gas to escape around it easily, was placed inside a piece of 28 mm. Pyrex tubing. The end of the tubing was fitted to a coarse tritted glass filter funnel which had the lower portion of the apron removed. This prevented the chlorine gas liberated during the electrolysis from coming in contact with and re-oxidizing the titanium trichloride solution. The other two holes in the cover were utilized to provide an inert gas inlet, and an outlet tube for removing the titanium trichloride solution when the reduction was completed. Two liters of 1.6 molar titanium tetrachloride solution were prepared by slowly adding 570 grams of titanium.tetrachloride to 700 mls. of cold, (0 to 5' 0.), concentrated hydrochloric acid. Cold water was slowly added to reach thedesired volume. The clear solution was placed into the resin Jar, and enough dilute (4:melar) hydro- chloric acid was placed in the anode comparunent to obtain good current flow. A.currcnt of 1 to 1 1/2 amperes was passed through the cell. The duration of the reduction was calculated by allowing 96,500 ampere-seconds for each mole of titanium tetrachloride to be reduced. Bubbles of hydrogen form at the cathode when the reduction is complete. This method was considered to yiold a superior product since there was no contwiinntion with zinc or other reducing substances. The reduction could be carried to completion with a.nininum of attention since the cell needed only periodic checking once the reduction was started. The titanium trichloride prepared in this way aervcd v.11 for precipitation reactions but would not serve for the preparation of anhydrous titaniun.trichloride. Evaporation or vacuum distillation results in hydrolysis. Anhydrous titanium trichloride was prepared by slowly passing e.mixture of titanium tetrachloride vapor and hydrogen through a furnnce at 660' C. The product in condensed on a cold finger. This method is very slow, gives small yields and involves a problem of protection from oxygen and moisture in transfer.‘ The method of Billy and Brasseur11 for the preoaration of anhydrous titenium.trichloride was then tried in which finely divided antimony is used as a reduotsnt. The antimony is obtained from.a solution of antimony trichloride by displacement with zinc metal. After purification and drying 4 grams or the antimony was placed into a Carina tube containing 30 grams of titanium tetrachloride. After being eVscuated the tube was sealed and heated in a Carina furnace to 840' C. for several hours. Crystals of antimony trichloride indicated the reduction of titanium tetrachloride. Excess titanium tetrachloride was removed with anhydrous carbon tetrachloride and antimony trichloride was removed with anhydrous other. .9. This method of puririCation is tedious, and difficult. It is hard to remove all impurities. The method was improved by evacuating the tube and heating to remove the impurities, because all of them were mmch more volatile than the titaniun.trichloride. Antimony reduction in this manner gave smell yields, and necessitated further transfer for fluorination. Lhen the titanium trichloride was removed from the Carina tube for transfer it immediately began to fume. lithin one minute it was red hot and oxidation nus soon complete. Since the finely divided trichlorids is oxidized very quickly on exposure to oxygen this method was discarded. The use of powdered alwminmm and aluminum chloride as a catalyst will give anhydrous titanium trichloride at a lower temperature than antimony.5 Huff's method was modified by using a cooper auto- clave for the reaction chamber, (Figure I). Sixty grams of titanium tetrachloride were placed in the bottom.of the autoclave. Two grams of aluminum were ground with two grams of aluminum chloride in a mortar. The mixture was added to the titanium tetrachloride, after which the autoclave was assembled, evacuated and the Valve was closed. East was applied so that the pressure rose to 120 pounds, and was maintained at that point for two hours by regulating the temperature. A large dry ice trap was attached to the sidearm of the autoclave. Suction was applied, and the trap was cooled in a dry ice-isOpropyl alcohol mixture. The autoclave valve was opened so that the aluminum trichloride and excess titanium tetrachloride could be distilled out. The flame was maintained on the bottom or the autoclave for two hours ~0- COFPER AUTOCLAVE I 1 VALVE BITE TEFLON PACKING WATER CONDENSER UNION NUT {inn ....___J FIGURE I to effect as complete a removal of aluminum trichloride as possible. The autoclave valve was then closed, and titanium.trichloride remained, in the container ready for further treatment. B - P2‘L’"’.—‘.ii;.'l"lol‘-3 0F TI‘Z‘JTJT: TRIE'JIOLIIJE The fluorination of titanium trichloride to produce titanium trifluoride can be carried out in several ways. For the titanium trichloride solution it was desirable to precipitate titanium tri- fluoride or some compound which could be easily converted to titanium trifluoride. For anhydrous titanium trichloride it would be desirable to treat with some reagent to leave pure anhydrous titanium trifluoride. If a solution of titanium trichloride is treated with hydrogen fluoride no precipitate is formed. Evidently titanium trifluoride is quite soluble under these acid conditions. A.precipitate will be formed on the addition of a solution of an alkali fluoride. Ammonium fluotitanites seem to be the least soluble. If ammonium fluoride solution is added to titanimn trichloride solution a precipitate of swuonium.pentafluotitanite is obtained.12 If the order of addition is reversed ammonium hexafluotitanite is obtained.12 It was hoped that ammonium fluotitanite Could be heated in a combustion tube to remove ammonium fluoride by sublimation and leave titanium trifluoride. A‘pH titration was carried out to learn more about the reaction of titanium trichloride with ammonium fluoride. The electrolytically prepared titanium trichloride solution was found to be 1.5 molar by titration with standard carats. Concentrated ammonium fluoride (13 molar) was prepared by dissolving 117 grams in enough distilled water to'mnke ~10- 250 mls. of solution. The pfl.of the titanium trichloride solution was 0.0 or less as shown on e Beckman pH meter. while the pH of the emmonium.fluoride solution was 7.5. Seventy mle. of titanium tri~ chloride solution were placed in.e beaker, and several drape of toluene were placed on top of the solution to prevent sir oxidation. The ammoniun.tluoride was placed in s burette and added in 1.001mb'incremente. A glass electrode and s calomel electrode were used with the pH meter, and the solution was agitated with a magnetic stirrer. The pH become stable after lb to 30 seconds of stirring and readings were then taken. The date is plotted in Figure II. The color of the solution changed during the titration so that certain parts of the curve are characterized by definite solution colors. At the beginning the trichloride solution was deep purple and slowly changed to a dark brown at the first peak. In the vicinity of the trough the solution changed from an almost opaque brown td’s bright clear green color. No precipitate was visible up to this point, but upon the addition of more fluoride the color changed from green to a violet, and a precipitate formed. After purification in s Soxhlet extractor with ethanol the ammonium.fluotitanite was dried in en Abderhalden dryer st 100' C. for several hours. The material was then placed into a large tube and evacuated. Hest'wse applied in an attempt to drive off ammonium fluoride and leave titanium trifluoride. However the ammonium fluotitanite eublimed without decomposition to the trifluoride. Since the trifluoride could not be prepared in this way this method was discarded. -11.. II SKLJ V1711“ 7:11- I" ,L KI L L ‘n7 (3? L—J g'I‘I ON CT”?- - ‘ r .m.. m1. I v-v --'- o 1‘ oi- 1 "-' it. TIT" ’ - 21:1; 1:?- fail-h -1}: -. :L fl: .0 Ru 41F— ._*-aanw*-.d- O O O .. - __QL1___..__...H--.__._.L ___ o .,.oo ._..._,._1 _ ._ -W p. r. 3 2 l O C" 4 ‘ u} flu r..( '70 60 40 flv 7p 10 The copper autoclave was utilized for the preparation of anhydroue titanium trifluoride. Titanimm trichloride was prepared (page 7) and the autoclave was then attached to the manifold, (Figure III) and an exceee of twice the neeeeaary.amount of anhydroue hydrogen fluoride nae condensed on top of the titanimm trichloride. After sealing the autoclave and removing it frm the manifold, it wee heated to about 100' C. to complete the reaction. Hydrogen chloride gas nae allowed to eeeape elowly at 80 to 100 p.e.i. This method eeene to leave some aluminum trifluoride in the titenium.trifluoride. Another method for producing titanium trifluoride 1e to.make potassium hexatluotitanate, reduce it to potassium.pentarluotitanite in the furnace at 600’ C..Iith hydrogen and finally leaCh out potaeeium fluoride with water to leave the lean soluble titanium trifluoride.6 This method leavee some potassium fluoride in the titanium trifluoride. C - PRi-IPARATIOI‘I OF TITAN I815 TRIF'LIJOROC‘LHJORIDE The last etep in the production or titanium tritluoroohloride ie the chlorination of titanium.trifluoride. Thie hae been done in the furnace and in tent tubee heated, in an oil bath or with a buneen burner. Shall amounts, about one gram, of titanium tritluoride were placed in tent tubee arranged in eeriee. Chlorine gae wee paeaed through them.and they were heated in an oil bath. At 220 to 225' C. a yellow eublimate covered the walle of the teat tube above the titanium trifluoride. '43.- r130: o "“.‘ ? 2:17- wriuupul’a- . -"-‘\va‘u ' ._) (I. L J1 FLITCFL‘IIT? ‘E‘I'F'Ifi AID " t ‘ ‘1‘~|A' /.l ‘Ii ”XXL 3*»Jle FIT}? TO C CPPEH ATTOCLAVE ‘ i \_ \ \ .\ \- \ \I r“ in. \ O ’ '30 WATER A13 PIRATOR FIGURE III TO ATHEYDR CUS HYI‘)R~OC?EJ I? FLE 'OR I IE CYL I 1‘? DEC R ‘TJF‘ADUA TED ALT F CHILE -1‘. Qualitative tests indicated the presence of titanium). chlorine, and fluorine. unnatitetive analysis gives the following date: Per Cent Arerage Theoretical Ti 35.4 34.8 34.4 34.9 34.1 P 36.4 34.8 34.9 35.4 40.6 Cl 26.8 28.7 27.2 27.6 25.3 Upon standing in the atmosphere titanium trifluorochloride changes from a yellow to a white solid. xbray powder diffraction patterns indicate the product is titanium oxyfluoride. mien weighed amounts of titanium trifluorochloride are allowed to stand in the air, and then are dried overnight the weight changes correspond to the change from titanium trifluorochloride to titanium exyfluoride. Wt. of T1339; at. of T1033 Theoretical I .2120 .1549 .1539 II .3406 .2270 .2473 Then titanium trifluorochloride was placed in water it hydrolyzed vigorously, and dissolved completely leaving an acid solution. D - PREPhflhTIOH 0F TITANIUM TETRAFLUORIEE The preparation of titanium tetrafluoride has been described . by several workers. 'Usually titanium tetrachloride is treated with anhydrous hydrogen fluoride. The methods of treatment vary frrxn one worker to the next. The method used for this work involved the use of the copper autoclave as a reaction chamber, (rigors 1). After cleaning and drying, the autoclave was attached to a manifold which was connected to a manometer, a vacuum.pump, a cylinder of anhydrous hydrogen fluoride, and a semi-transparent graduated fluorethene ampoule, (Figure 111). -15. The system was evacuated using the vacuum pump and then sealed off. The graduated ampoule was cooled in a mixture of dry ice and isopropyl alcohol. Anhydrous hydrogen fluoride was distilled into the ampoule until the desired quantity, about 50 mle., was obtained. The autoclave was then cooled, while the ampoule was warmed. In this way the quantity of hydrogen fluoride distilled into the autoclave was determined. The autoclave was evacuated after the transfer of hydrogen fluoride was complete. About 60 grams of titanium tetrachloride was added on top of the hydrogen fluoride. The autoclave was removed from the manifold, and attached to the container of titanium tetrachloride. The lower part of the autoclave was kept cold in the dry ice and alcohol mixture. to retard the reaction. The vacuum inside the autoclave, and gravity were used to transfer the titanium tetrachloride. Hydrogen chloride gas which came from the reaction was allowed to bubble through the titanium tetrachloride until the transfer was complete. The autoclave valve was closed and the titanium tetrachloride transfer apparatus was removed. Then the autoclave was allowed to warm slowly. Eben the pressure reached 100 pounds the hydrogen chloride gas was allowed to escape slowly through the valve. Finally cold water was circulated through the condenser while the lower part of the autoclave was heated in a gas flame to complete the reaction. The autoclave was cooled, evacuated and filled with dry nitrogen to a pressure of lO p.e.i. The union not was loosened just enough to permit removal with the hands. The positive pressure of the nitrogen .16.- prevented moisture from diffusing into the autoclave. The lower part of the autoclave was inserted in the dry box. A large one hole rubber stopper was placed around the mtoclave Just above the union mt. The outside edge of the rubber stopper fit tightly in a port in the dry box so that the lower part of the autoclave could be sealed in the dry box. After drying the autoclave was opened. At this stage the titanium tetrafluoride had a light tan color. Since the pure compound is white, this indicated impurities, and the product had to be renblimed. The first attcnpte at resublimation were carried out in evacuated glass tubes. This gave a white product which would not resubline further. This material was titanium cryfluoride. Evidently the titanium tetrafluoride attacked the glass and titanium oxyfluoride was obtained, rather than titanium tetrafluoride. A nonel tube was substituted for the glass tube with satisfactory results. A white predict which was solid along the walls of the tube, but similar to glass wool at the center, was obtained. E - PREPARATION OF TITJ-JIIUM OKYTUJOMDE This material was obtained when an attupt was ads to resublime titanium tetrafluoride in a glass tube. Ruff and been reported that titanium tetrafluoride dihydrate can be obtained by dissolving titanium tetrafluoride in water and evaporating the solution. Solutions of titanium tetrafluoride could be produced by dissolving titanium dioxide in squeals hydrogen fluoride. If the observations of fluff and Ipsen were correct evaporation of these solutions should also yield titanium tetrafluoride di hydrate. -17- In an effort to determine what actually does happen two types of experiments were carried out. (1) The hydrolysis of a weighed amount of titanium tetrafluoride; (2) the reaction of a weighed amount of titanium dioxide with hydrogen fluoride. (1) Several grams of titanium tetrafluoride were placed in a platinum crucible. Distilled water was added dropwise until the titanium tetrafluoride Just dissolved. The material was placed in a covered container and allowed to dry in the oven at 120 to 140' C. ‘A crust formed at the surface, while a pasty'material remained below. The crust was periodically broken and the material allowed to dry. at the end of three weeks, the material was thoroughly dried. Ht.ofATiF; wt. of TiOFz Theo. Wt. Difference 4e617‘ gme. 3e8040 gins. 3e7975 mle 0e0065 Edna. Difference I 0.17% (2) Titanium dioxide was dissolved in aqueous (50%) hydrofluoric acid and eVaporated. The material was similar to that obtained by treating titanium tetrafluoride with water. It took two to three weeks to dry this material. Xeray powder diffraction patterns indicated that titanium oxyfluoride had been produced. Finally titanium dioxide was placed in the copper autoclave. The autoclave was attached to the manifold used for the production of titanium tetrafluoride. The system was evacuated and an excess of anhydrous hydrogen fluoride was distilled into the autoclave. The autoclave was sealed off, removed from the manifold, and the lower portion heated in a steam bath. at 80 to 100 p.s.i. the excess hydrogen -13- fluoride was allowed to escape slowly. Finally the too of the auto- clave was wrapped with asbestos tape to provide insulation. The autoclave was evacuated with an aspirator and the lower part was then heated in a gas flame to drive off the last of the constant boiling hydrogen fluoride. EE:_2£_I£92 3" 0’ Ti9?b Theo. 2t. Difference 4000 8138s figs-25 We 51.0 8:35. 1085 813's Difference - 2.5; The material obtained nave powder diffraction patterns corresponding to titanium oxyfluoride. The color was not shite but e light tan. Chlorine gas passed over the tan material in en eight inch.Pyrex test tube with heating, gave a white product, and carried off volatile impurities. Quantitative analysis yielded the following results: Per Cent Average Theoretical Ti 47.10 47.18 47.14 47.01 F 36.11 36.33 36.22 37.29 The material is not very soluble in water, dilute acids or bases. A sodium carbonate fusion was used to put the material into solution. The methods used in the analyses are given in the appendix. K-‘Zl POWDER DID‘IRACTION STUD£3§ -19- V " X'Q'tY PO'SDER DIFFRlCTIDIG S'FJIJI ‘23 Irray powder diffraction patterns were used extensively to check on the identity of the compounds prepared in this work. Data were not available for all of the compounds, and that which was published for titanium tetrarluoride was even to some queetion.13 Exposures were made with a North American Phillips er y unit. The cameras had a diameter of 11.459 cm., and the capillaries had a diameter of 0.5 mm. The eoyper X04, radiation was passed through a nickel filter for all exposures. The orposuree sere made at 55 kilovolts and 16 milliamperes for 4 to 6 hours. Relative line intensities were measured with a film strig made from a step tablet of the ratio one to the square root of two. Some titanium trifluoride nae carefully prepared to obtain high purity, and a diffraction pattern was obtained. The data is given in Table I. The structure was complex and no attempt at structural analysis was made. Samples of titanium trifluorochloride were loaded into capillaries in the dry box. No patterns were obtained from.tour different samples which indicates that the particles are too fine. Samples of titanium oxyfluoride gave patterns which indicated a cubic structure with a I 3.790 1 .005 2. The density obtained using toluene in a pyenometer at 25' C. was 2.82. The calculated density is 3.10 which is in fair agreement since oxperhnental densities are usually lower than those calculated from eray data. The diffraction data is given in Table II. TABLE I X-RAY'POUDBR DIFFRACTION DATA FOR [O- 4.23 3.84 3.57 8.11 2.71 2.31 2.06 1.92 . 1.80 1.73 1.68 1.57 1.36 1.28 1.089 1.042 .9033 .8287 .8074 Diffuse Diffuse Diffuse Diffuse Diffuse TITAEIUE TRIFZUCRIDE 1 100 100 85 .1130 1 100 46 10 25 6 100 85 45 88...»... 15 TABLE II I-IL".Y P07337911! DI FRET"! 3N DATA FOR TITJ'iNIUI-‘E OXYFIUOdIDE 3.75 2.66 2.18 1.89 1.69 1.54 1.34 1.26 1.20 1.09 1.048 1.010 .947 .919 .893 .848 .827 .807 _I_ III 50 100 35 1 9‘5 95 50 100 35 50 55 15 16 15 1? 12 18 Plenshk1 100 110 111 200 210 211 310 222 320 321 400 410 411 420 421 332 k 300 &}322 & 330 ~22- atructure factors were calculated using the equations given in the text by Eunn.14 Values for each individual plans, (100, 010, 001) etc.. were obtained and summed for the appropiate series of planes, (100, 010, and 001 Calculated individually and owned for 100 series) etc. The data is giten in Table III. it appears that the atoms are arranged with titenium.et the corners of the cube. Eight fluorine atoms are located on the left and right sides halfway between the titanium atoms. Four oxygen atoms are located in the remaining spaces halfway between the titanium atoms. 17-— 0——'T,- r /' / ./ F /F 77 --é (D -- 71' i: I I “,0” 7., FT ./ ’. fr ,l’ l , F T7;0—-Tf / A sample of titanium oxychlorids was prepared by evaporating an aqueous hydrochloric acid solution of titanium tetrachloride, grinding the residue and drying in the oven at 120' G. for two days. This material gave no diffraction pattern. Therefore no comparisons could be made between the structure of titanium oxyfluorids and titanium oxychloride. Diffraction patterns from titanium tetrafluoride which had been reeublimed in e monel tube were measured. The comparison between the values obtained and those in the literature is given in Table IV. -23- TABLE III SRRUCTURE FACTORS FOR TITANIUQ UXYILUORILE (Timporature factors and absorption coefficients omitted.) 2 mmd lean. $.10 Jobs. 100 1305 100 110 48 35 111 1 1 200 224 95 210 118 100 211 9 35 220 35 60 221 9 300 39 55 310 8 6 311 2 O 222 9 15 320 30 16 321 6 15 400 20 2 ‘10 § 322 50 17 £11 0 330 7 6 331 1 O 420 49 12 421 28 18 332 6 1 TABLE IV -24- X-R-ZY P031333 DIFFRACTION Uzi-HA FOR TIT. .13 UL: was: 2. so. 55 3.5 6.0. 4.00 3.79 3.68 3.09 2.68 2.35 2.22 2.01 1.89 1.72 1.69 1.63 1.50 1.39 1.34 1.28 1.25 1.04 1.00 42 50 35 T79? ilki’L'JC III .7313 12 8 85 1 100 35 10 to 10 70 20 20 1? 17 20 6.50 5.45 4.75 4.35 3.80 3.50 3.40 3.22 2.68 2.30 2.18 1.90 1.70 1.61 1.55 1.485 1.345 1.266 1.098 1.054 1.015 .921 [LSTM Data I 40 £0. 10 3 Oi at ..88 12.5 10 ~25- VI " MAGNETIC SUWL‘IPTIBILITY STUDI'PS Hognetic susceptibility data on titanium compounds is limited. Measurements on the compounds used in this work were made to add to the existing data and to distinguish between possible typoe of bonding in titanium tritluorochloride. .A Gouy type magnetic susceptibility apparatue with e semi-micro balance one used. The 7 mm. Pyrex tube was calibrated at 1.80 c.c, with a mark which was 97 mm. from the septum. The elctromagnete were need at the following amperagee which produced the fields given: Amperee 3 8 8 ll 15 Fielde 2,950 4,780 7,360 9,060 10,.70 (aerated) Klemmgb‘meaeured the magnetic eueceptibility of titanium.trichloride and round it to'be paramagnetic with e specific eueceptibility of 7.13 x 10'". Measurements on titanium trifluoride indicated a emell amount of ferromagnetic impurity which did not permit an accurate value for the susceptibility but did show the trifluoride to he paramagnetic with e aueceptibility o? the order correeponding to one unpaired electron. A plot of 1/H versueQ/a was made to obtain the paramagnetic susceptibility without the effect of the ferromagnetic impurities. roan l/H me extrapolated to infinite field strength, (l/H - o), the Value 7.5 x 10" was obtained for x , which gave on effective Bohr magneton number of 1.40. The value or 1.73 would correspond to one unpaired electron. The tetrafluoride contained a Very minute trace of impurity but did give a specific diamagnetic susceptibility of -0.037 x 1.0"6 at 25' C. The trirluorochloride gave e epeciric diamegnetic eueceptibility or (—0.060 g 0.020) x 10“ at 25° 0. Within the limite or experimental error the Value: for the tetrerluoride and the trifluorochloride are about the lame. Measurements on titanium oxyfluoride indicated the preeence of ferromagnetic impuritiee. A plot of UK vereue 24 givee}4 :- 4 (0.42 g 0.02) r 10‘5 at 22- c. which 1. slightly paramagnetic. Data for magnetic eueceptibility'measuremente is given on the touching page e TABLE V ILAGHETIC SHEIZEE’TIBILITY STUDIES Titaniuw trifluoride 27‘ C. Amperea :14 3P 5 . 17.1 x 10"6 11 10.7 15 10.1 Titanium tetrafluorochloride 23" C. Amperee 2/, .P O - 11 no weight change detected 16 -0.060 r 10"6 Titanium tetrafluoride 25‘ 0. Meme 3%,, -5 25 -0.006 x 10 11 ~0.025 15 -0.037 Titanium oxyfluoride (prepared in the autoclave from C. P. titanium oxide and anhydroue hydrogen fluoride) 22' C. Amperee )5. up 5 . 0.476 x 10'5 a 0.471 ll 0.464 15 0.d60 «29.80 c. 5 0 0.623 x 10"6 6 0.596 11 0.592 15 0.576 ~77.4° c. 5 . 0.760 x 10"6 a 0.759 11 0.747 15 0.738 DISC'JSSION -23- VII " DISCUSSION In the preparation of titanium trifluoride the main problem wee that of protection from the moieture and oxygen of the atmoephere. Of the eeverel methods tried the copper autoclave eliminated meet of the probleme eiece the reduction and fluorination could both be accoupliehed without trenefer of the titanium canpound. Neither X-ray powder diffraction petterne nor magnetic eueceptibility date wee available for the trifluoride. The data obtained indicated that the etructure 1e canplex. Single - cryetal X-ray enclyeie for further etructural etudy wae not poeeible eince the product obtained was alwaye a fine powder. Magnetic eueceptibility data indicated the preeence of a ferro- magnetic impurity, preeumably from the iron in the preeeure suege, valve, or pipe leading to the valve. A plot ofx vern- l/H extra- poleted to infinite field etrength gave an approximate value for . Since the pointe plotted did not lie directly on e etreight line, an exact value cannot be claimed. kbwever the value ie cloee to that expected for e canpound with one unpaired electron, and to relatively eloee to that obtained for the trichloride. Chlorination of the trifluoride gave a yellow compound which eublimed at a lower temperature than the titanimn tetrcfluoride. The etructure could not be determined eince no x-ray powder diffraction petterne were obtained from the material. The diamegnetic eueceptibility wae cloee to that of the tetrafluoride. The triflucrbchloride ie insoluble in benzene and carbon tetrachloride, but it ie eamewhet eoluble in chloroform. It reacte readily with water dissolving completely and leaving an acid eolution. The addition of dilute eodium.hydroxide yields a precipitate of titanium hydroxide. From this the following overall reaction ie postulated: r1230]. ., 4021' --r Ti(OH)‘0 3F.* 01" If the trifluorochloride ie expoeed to the atmosphere the yellow color elowly dieappeare and a white eolid remains. The white eolid gives a.powder diffraction pattern for titaniwm oxyfluoride. From thie the following equation ie postulated: Tirbcl e :23 —7' T1015 0 H! e HCl The color of the trifluorochloride could not be explained on the baeie of an unpaired electron, but may be ascribed to the polarization of the molecule. The difference in electronegativity of the fluoride and chloride in relatively large, and polarization effecte would be encountered with a chlorofluoride that would not be found in either binary halide. The titanium tetrafluoride which had been purified for.X-ray analyeie gave data which was ehnilar to that in the literature. There nae a difference for eeveral linee indicating that eome impurity poeeibly the oxyfluoride may have been preeent in the material for the original work. Titanium.oxyfluoride was obtained fram the sublimation of titanium tetrafluoride in a glass tube, and from.the reaction of titanium dioxide with aqueoue or anhydroue hydrogen fluoride. In the evaporation of a eolution of titanium tetrafluorida crystals of material corresponding -30- to the formula TiF4 reactione are poetulatedi ' 2320 were not found. From this the following Ti!‘ 0 H20 -' 'I'iOE‘3 + 2H? Tioz t 2H? -ev TiOFz e 820 Apparently the reaction: 'i'io2 e 4111'” -r' TiF‘ 4' 21120 proceeds only when the eyeten ie euept continuously with hydrogen fluoride. If hydrogen fluoride is added to titanium dioxide as in the autoelave and heated the oxyfluoride will be produced. Evaporating an aqueous solution of titanium and fluoride loan will give the oxyfluoride, which suggeste the exietence of the titanyl ion rather than the nonetomie tetravalent ion. The magnetic susceptibility data indicates a slight amount of paramagnetiam. This may be due to the different orbitals need for bonding, and to the different anions in the compound. The structure of titanium oxyfluoride in apnarently cubic. Since the ionic radii of the oxide and fluoride ions are almost the same this may be expected. VIII " WELAKY Titanium trifluoride was prepared by the reduction of titanium tetrachloride and subsequent fluorination. Xeray'powder diffraction data, which was not previously available was obtained. Titanium trifluorochloride, not previously characterised, was prepared and some of its properties were observed. The properties of this substance are intermediate between those of the tetrafluoride and the tetrachloride. Attempts to prepare titanium tetrafluoride dihydrate as reported by others always yielded the oxyfluoride. This compound results from either the hydrolysis of the tetrafluoride or the treatment of titanium dioxide with aqueous or anhydrous hydrogen fluoride, unless provision is made for the removal of water. The oxyfluoride has a cubic structure, with a - 3.790 2. Magnetic susceptibility.measurements were made on the tri- and tetrafluoride, the trifluorochloride and the oxyfluoride. The tri- fluorochlorids and tatrafluoride are diamagnetic with similar values. The oxyfluoride is slightly paramagnetic and the trifluorida is paramagnetic with a value corresponding to approximately one unpaired electron. LITERATURE CITED -32- LITERATURE CITED 1. 2. 3. d. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. nannay, n. B. and anyth, c. P., J. Am. Chem. Soc., _6_§_, m (1946). Ruff, 0e and 1138631, Re. Rte. 519-. 1777 (19%)0 ”106.1, Ce, and Guerln, Jo. COMPte “Dds. gb 889 (1875)e Raeder, n. 0., K31. Noreke Videnskab. Selskabs Skrifter, g, Walton, H. 1.,“Inorgamo Preparations: Prentice-Hall Inc., hoyer, F., Bauer, A. and Schmidt, 3., Ber..§§_§, 1908 (1925). Ruff, O. and Reumann, F., Z. anorg. Chen., lgg, 81 (1923). Weber, H., Journ. Prakt. Chem., 29, 214 (1863). Iiautefeuille, P., oanpt. rend., 93, 148 (186:5). Cumming, A. and.Kny, 8.,‘A.Text-Book of Quantitative Analysisf“ D. Van Nostrand Co., Inc., New York, 1th.ed, 1939, pp. 121. Piccini, A. 1b,, 31, 1054 (ms). Rinn, Eh W., oral communication. Bunn, C. 3.,”Chemical Crystallography: Oxford University Press, London, 1946, pp. 218. Klemm, fl. and Krose, 3., Z. anorg. Fhem., 253, 809 (1947). APPENDIX The titanium oxyfluoride was fused with sodium carbonate, and dissolved in dilute nitric acid. Titanium was determined by precipitating the hydroxide with sulfur dioxide. fluoride was determined as lead chlororluoride. Titanium triflnorochloride completely dissolved in water. Titanium hydroxide was precipitated upon the addition of sodium hydroxide, filtered at pH 5, washed and ignited. Chloride was determined from the filtrate using the gravimetric silver chloride method. Fluoride was then determined as lead chlorofluoride. TITANIUM HALI DES Hy Karl 8. Vbrres AN ABSTRACT ‘ Submitted to the School or Graduate Studies at Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree or MASTER OF SCIENCE Department of Chemistry Year 1953 ,/\ ' /Z’ , I ' ’1 . . ,4. __‘ ADDI‘OVOd Ii ”afi’L’JVK (C: AL 4. 1., "D (PM... Karl S. vcrree Titanium»trifluorochlcride was prepared in an attempt to observe the transition.frcm.an ionic type of compound, characterized by titanium tetrafluoride, b.p. 884' C. to a covalent type of I compound, titanium tetrachloride, b.p. 136' C. The trifluoroe chloride was prepared by reducing the tetrachloride with aluminum and aluminum.chloride in a capper autoclave, fluorinating with hydrogen fluoride, and finally chlorinating with chlorine. The yellow compound sublimes readily at 220 to 225' C. The magnetic susceptibility is close to that obtained for titanium tetrafluoride. No X-ray powder diffraction patterns were obtainable. Titanium oxyfluoride was prepared from the reaction of aqueous or anhydrous hydrogen fluoride with titanium dioxide. The cxyfluoride was also obtained from evaporating an aqueous solution of titanium tetrafluoride, and from the effect of atmospheric moisture on the trifluorcchlcride. x~ray powder diffraction patterns indicate a cubic structure for the oxyfluoride with a I 3.79 X. magnetic susceptibility measurements indicate a slight paramagnetism. Titanium trifluoride gave complex X-ray powder diffraction patterns. The magnetic susceptibility corresponds to approximately one unpaired electron. Titanium tetrafluoride was prepared in the copper autoclave by adding titanium tetrachloride to anhydrous hydrogen fluoride. Resub- limation in a glass tube yielded the oxyfluoride. A.monel tube eliminated the difficulty. The tetrafluoride obtained was used for Xkray'pouder diffraction data which indicated that the American Society for Testing Materials data for titanium tetrafluoride corresponds to a mixture of the tetrafluoride and oxyfluoride. TY ”'TITJ'MHLRMLIEIflifljrijfy't‘flflmuflflfgfiflfi“