PLEASE DO NOT REMOVE FROM LIBRARY a R I 9 0 2 4 Q ^— 0 s * Bureau of Mines Report of Investigations/1986

Solubility of Anhydrous AICI3

in TiCI4 and VCI4

By Dennis A. Hansen and Davis E. Traut

UNITED STATES DEPARTMENT OF THE INTERIOR tteport of investigations 9 0 2 4

Solubility of Anhydrous AICI3

in TiCI4 and VCI4

By Dennis A. Hansen and Davis E. Traut

UNITED STATES DEPARTMENT OF THE INTERIOR Donald Paul Hodel, Secretary

BUREAU OF MINES Robert C. Horton, Director Library of Congress Cataloging in Publication Data:

Hansen, Dennis A Solubility of anhydrous3 in TiC^AICI and VCI4. (Report of investigations / Bureau of Mines ; 9024) Bibliography:8 . p. Supt. of Docs, no.: I 28.23: 9024. 1. Aluminum chloride—Solubility. 2. Titanium tetrachloride. 3. Vanadium tetrachloride. I. Traut, D, E. (Davis E.). II. Title. III. Series: Report of investigations (United States. Bureau of Mines) ; 9024.

TN23.U43 [TP245.A4] 622s [669’.7322] 86-600041 CONTENTS

A b s t r a c t ...... 1 Introduction...... 2 M a t e r i a l...... s ...... - ...... 3 E q u i p m e n...... t ...... 4 Experimental p r o c e d u r e ...... 5 Results and d i s c u s s i o n ...... 6 C o n c l u s i o n s ...... 8 R e f e r e n c e s...... 8

ILLUSTRATIONS

1. Autoclave a s s e m b l y ...... 4 2. Top half of a u t o c l a v e ...... 4 3. Solubility of AICI3 in Ti C l 4...... 6 4. Solubility of AICI3 in TiCl4—a comparison of results...... 6 5. Solubility of A1C13 in VC 1 4...... 6 6. Solubility of AICI3 in Ti C l 4- V C l 4 s y s t...... e m 6 7. Effect of V C4 I concentration on 3AICI solubility in TiCl4- V C l 4 s y s t...... e m 7

TABLES

1. Spectrographic impurity a n a l y s i s ...... 3 2. Equilibrium time for solubility 3of in AICI T i C l 4 and V C 1 4 at 75° C...... 5 3. Summary of solubility data for the4- A l CTiCl l 3-V C l 4 s y s t...... e m 7 UNIT OF MEASURE ABBREVIATIONS USED IN THIS REPORT

°c degree Celsius ym micrometer

h hour psi pound per square inch

K kelvin rpm revolution per minute

L liter wt pet weight percent mL m i lliliter SOLUBILITY OF ANHYDROUS AICI3 IN' TiCI4 AND VCl4

By Dennis A. Hansen1 and Davis E. Traut2

ABSTRACT

The solubility of anhydrous aluminum chloride in solutions of titanium tetrachloride and vanadium tetrachloride was determined as part of a re­ search project by the Bureau of Mines to make a homogeneous alloy sponge of titanium, vanadium, and aluminum. Solubility curves were established using an autoclave at 75°, 100°, 125°, and 150° C. The solubility of AICI 3 in T i C l 4 is 0.2 wt pet at 75° C; it increases to 18.3 wt pet at 150° C. The data concur with data from earlier researchers of this bi­ nary system for 75° to 125° C. For4-AICI the3 VCI binary system the solubility of AICI3 is 0.1 wt pet at 75° C, increasing to 1.6 wt pet at 150° C. For the TiCl4- A l C l 3- V C l 4 ternary system the solubility3 of AICI decreases from the values found with a pure4- A l C lTiCl3 system to the lower values found with a pure4-AICI VCI3 system at the respective temperatures.

1 Chemical engineer. Supervisory chemical engineer. Albany Research Center, Bureau of Mines, Albany, OR. 2

INTRODUCTION

The Bureau of Mines is conductingavailable were limited to the solubility studies on several approaches toof make A I C I3 in TiCl4 . high-quality titanium alloy powder. Sav­To determine the solubility of3 inAICI ings in cost and reduction of wasteTi Cmay l 4 , Eingorn (4^ used a thermal analy­ be realized in the production of titaniumsis method and investigated the system alloy products by the use of powderfrom met­ -24° to 120° C. Through a lateral allurgy techniques to produce high qual­extension in a glass vessel, just enough ity, near-net-shape parts. Several Tmeth­1C 14 was poured in to cover the thermom­ ods for producing spherical, high-purityeter bulb. The melting temperature of titanium alloy powder are being investi­the pure solvent was measured. Then a gated throughout the industry, butsmall only amount of AICI3 was added and the one method, the Plasma Rotating electrodetemperature of crystallization was again Process (PREP), has been proven to dpro­ etermined to estab l i s h a cooling curve,, duce a powder of high enough quality forMorozov (_5) used a saturation method use in aircraft manufacture_).3 (J Pro­ from 18° to 80° C and a thermal analysis duction of powder by the PREP processmethod is from 93° to 193° C. In both meth­ expensive because of the necessity ofods us­ a sealed vessel was used. For the ing electrodes accurately machined andsaturation method the chlorides were ground from billets forged from double-sealed in one arm of a two-armed glass or trlple-arc-melted alloy ingots.apparatus, Di­ and kept at the desired tem­ rect production of spherical alloyperature powder for 20 to 40 h. Then part of from titanium alloy sponge could elimi­the transparent supernatent was poured nate most of these costly meltinginto steps the unused arm. This arm was then and result In significant cost savings.cut off and the solution analyzed* For In the approach used by the Bureau the (2^) thermal analysis method a known to make alloy sponge, the chloridesweight of of each of the chlorides was titanium (Ti), vanadium (V), and aluminum placed in a seated apparatus, and the (Al) are simultaneously reduced by magne­apparatus was subjected to heating and sium (Mg) via a Kroll-type reaction.cooling. A The resulting temperature was homogeneous (as to Ti, Al, and V distri­measured by a pyrometer. Morozov's re­ bution) powder is considered essentialsults are currently used as a reference for final titanium alloy part integrity(6). (3). (Thus, both a homogeneous titaniumDruzhinina (_7) used a saturation method alloy sponge and a homogeneous metalfrom 55° to 150° C, very similar to that chloride feed are required to make used the by Morozov, As with Morozov, the powder.) At room temperature, titaniumclear supernatent was decanted from one tetrachloride (TiCl4) and vanadium tetra­test tube leg to another in a sealed sys­ chloride (VC14) are liquids, while alumi­tem. Mixing was done periodically by num chloride (AICI3) is a solid. A dis­shaking the test tube apparatus. From 4 persion of minute, solid3 particlesAICI to6 h was required to reach equilibrium in the liquid TiCl4”VC l 4 phase would notsaturation. Druzhinina found that the be desirable, as this would adverselythermal im­ analysis method for determining pact homogeneity (2^, p. 20). Therefore,solubility of this system was not satis­ the solubility relationship 3 of in AICI a factory owing to the supersaturation of T i C l 4- V C l 4 solution was required to AICIprop­3 in TiCl4 . erly prepare feed solutions for theEhrl Kroll i c h (8) used a saturation method work, but this relationship was notfrom found 25° to 125° C in which the chlorides in the literature. The solubility were data combined in a stirred container for an unspecified time to reach equilibrium. ■^Underlined numbers in parentheses Thenre­ a sample was taken. Ehrlich com­ fer to items in the list of referencesmented at that the data obtained by Morozov the end of this report. could not be confirmed and that the 3 solubility values of 3AICI in H C I 4 d e ­ researchers' methods and a duplication of termined by Morozov were about 10their times results in the 4- TiCl A l C l 3 system higher than his own solubility values.would confirm the Bureau's experimental Ruban (9) also used a saturation methodmethod and procedures as applied to not from 70° to 127° C, but with an unsealedonly the TiCl4- A l C l 3 system but also to reaction vessel. The equilibrium timethe T i C l4- A l C l 3- V C l 4 system. for each temperature was determined;Morozov it was the only researcher who ranged from 38 to 40 h at 70° C downused to the thermal analysis method to de­ 15 to 20 h at 127° C. Ruban expressedtermine solubility O ) . The data ob­ concern that the6-h equilibrium time tained in that study are not consistent used by Druzhinina was too short andwith thus four other studies. Ruban suggested produced low AICI3 solubility. that the thermal analysis method may not Martynov (10) used a saturation methodallow the system to reach equilibrium from -20° to 50° C in a stirred quartzsaturation owing to insufficient time vessel (11). Martynov compared the atdata a specific temperature. In addition, of Ehrlich, Druzhinina, Ruban, and Mor­the method would require the determina­ ozov with his own and found the datation of of solution composition by the input Morozov to be significantly differentof known weights and volumes of the mate­ from data from all other investigators.rials rather than actual analysis. Both The emphasis of this study was thethe sol­ possible hydrolysis of the materials ubility of AICI3 in a T i C l 4- V C l 4 system. used and the instability of4 might VC1 It was not the initial intent of preventthis the accurate determination of study to duplicate the previous studiesthese weights and volumes. For these on the TiCl4- A l C l 3 system as they seemedreasons the saturation method was chosen adequately done. However, the materi­to determine the solubility relation­ als in both systems are difficultship to in the TiCl4- A l C l 3, V C 1 4-A1C13, and work with since they hydrolyze T easily. i C l 4- A l C l 3~ V C l4 systems in this study. Therefore, a careful review of previous MATERIALS The spectrographic analyses of VCI com­3 and C I 2 gas. This decomposition pounds used as starting materialsis are increased by temperature but hin­ shown in table1. dered by pressure of the product2 g a s , Cl Anhydrous AICI3 was purified by sublim­and usually must be photochemically or ing technical-grade AICI3 in a glass re­ actor inside an electric tube furnaceTABLE 1. - Spectrographic impurity that was held at a temperature above theanalysis,' weight percent s u blimation temperature of 182.7° C (12). The sublimed AICI3 was swept away with Ion AICI 3 TiCl 4 VCI 4 argon carrier gas to another chamber, >10 <0.01 ND where it was condensed as a white powder <.003 <.003 ND using water-cooled coils. <.01 <.01 <0.01 The T i C l 4 was distilled in a 5-L round- <.01 NDND bottomed flask containing 2 L of TiCl4, <.03 <„1 <•1 and the condensed vapor collected in a <.01 <.01 <.01 2-L flask. The system was purged with <.03 >10 <.3 argon prior to use. The distillate was <.3 ND >10 then redistilled using the same methods.Zn...... <.03 ND ND The double-distilled TiCl4 was colorless.ND Not detected. The V C 1 4 was triple distilled by the'The materials were also analyzed for same methods used for the TiCl4J Thisthe following elements, but none was de­ V C 1 4 distillate was dark brown.4 is VC1 tected: Ag, B, Be, Ca, Cb, Cd, Co, Cr, known to be unstable and decomposesHf, very Mn, Mo, Na, P, Pb, Pt, Sb, Sn, Ta, W, slowly at room temperature intoZr. solid 4 catalytically initiated (14-15). Howev­insoluble VCI3 and C l 2 gas should not in­ er, V C I3 is insoluble in TiCl4 (8^) and fluence the solubility of 3 AICI in the in nonpolar solvents such 4 as (16). VC1 ternary or binary systems. The presence of VCI3 in this study of All of the above compounds hydrolyze the ternary system of4- A l TiCl C l 3- V C l 4 when contacted with water. Exposure to and the binary VCI4-AICI 3 system wasair and the contained humidity will eas­ not desired but was not preventableily nor hydrolyze 4 TiCl and V C I 4. Care in measurable. However, this presencehandling of was therefore exercised. EQUIPMENT The apparatus used to contain andbefore mix reaching the sample bottle. A the metal chlorides during the solubilitysampling valve was located on the sam­ study was a 300-mL, stirred, stainlesspling line just above the sampling port. steel autoclave, shown in figures1 and The mixture was stirred at 100 rpm with 2. Figure 1 includes the support stand,an impeller turned by an adjustable-speed stirring assembly, and the heater. electric Fig­ motor, drive belt, and pulley ure 2 shows the top half of the reactor.system. A series of packing cones and The reactor material was 316 stain­spacers around the shaft and inside the less steel with the main vessel havingair-cooled a packing gland kept all vapors T e f l o n4 fluorocarbon polymer liner.inside A the reactor. A safety rupture fritted stainless steel10-pm ( pores)disc, rated at 2,000 psi, was used. A filter prevented solid3 AICIand any VCI3 type J (iron-constantan) thermocouple in­ present from entering the sampling sidetube. the reactor was connected to a po­ The sampling line outside the reactortentiometer, was using an ice bath as a preheated using heat tape to preventreference the temperature. This thermocouple AICI 3 from precipitating out of solutionindicated the actual process temperature. Another type J thermocouple was located ^Reference to specific products between does the heating mantle and the reac­ not imply endorsement by the Bureautor. of This thermocouple was connected to M i n e s . the autoclave temperature controller and

Stirring motor, drive belt, and pulley system Autoclave Sample valve and port

mantle

Thermocouple

Fritted sample filter

Impeller'

FIGURE 1. - Autocfave assembly. FIGURE 2. - Top half of autoclave. 5 was used to control the processprevent tempera­ wide temperature swings due to ture. A calibration curve relating the the thermal gradient of the Teflon liner. temperature inside the reactor The to apparatusthe was placed inside a glove temperature outside the reactor wasbox, used which was filled with argon while to set the controller temperature.the reactor This was loaded and when the ma­ double thermocouple arrangement wasterial rec­ was sampled. The glove box was ommended by the autoclave manufacturervented to to an exhaust hood.

EXPERIMENTAL PROCEDURE

An experiment was performed to run,deter­ an amount of 4 VC1 in TiCl 4 was added mine the amount of contact time forto the the vessel. This transfer was done system to attain equilibrium. Rubanin (9)an argon atmosphere within a glove had stated that at lower temperatures,box. An excess amount of3 AICI powder greater time is necessary for equilibr­was added to the vessel. The apparatus ium. For the present study, excess3 was AICI assembled and the temperature con­ was contacted with T-iCl4- at 75° C troller for was adjusted to an internal tem­ times up to 144 h, and a sample was takenperature of 75±0.5° C. The contents were as soon as the system reached temperatureleft at each temperature for at2 h least and at various intervals after that to time. reach equilibrium, as determined by

A similar equilibrium study using4 VC1the earlier study. Then under argon, ap­ at 75° C was also completed for proximatelytimes 5 mL of fluid was withdrawn up to 144 h. No significant differenceand then discarded in order to purge the was found in the analysis of samplessample for line. Two samples were then taken times of 2 h or more. (See table and2.) quickly capped. The temperature con­ Therefore, at least 2 h of contacttroller time was adjusted to the next higher at the desired temperature was allowedtemperature (100°, 125°, and 150° C) and before sampling during the solubilitythe above procedure was repeated. program. For analysis, each sample was added to Measurements of solubilities werea made1-L graduated cylinder containing 400 at 75°, 100°, 125°, and 150° C using the mL of ice and water. The sample bottle purified chlorides. The temperaturewas was rinsed three times with more ice wa­ fixed but the pressure was not measured,ter, and this rinse water was added to as the effect of pressure on solubilitythe graduated cylinder. After the ice is considered very small (16). For melted,each the mixture was stirred and about 25 mL of liquid was taken for analysis. TABLE 2. - Equilibrium time for solubil­About 5 mL of concentrated HC1 was added

ity of AICI3 in TiCl 4 and VC1 4 at 75° C to each sample to prevent precipitation of solids.

Contact time, h AICI 3 , wt pet' Titanium concentration was determined

In TiCl 4 In VC1 4 volumetrically using a solution of ferric

0 . 2 1 0.08 ion as the titrant and thiocyanate ion as 0 . 5 ...... ND .08 the endpoint indicator. Aluminum concen­ .19 .08 tration was determined by atomic absorp­ . 24 .09 tion spectrophotometry. Vanadium concen­ 4 ...... 25 ND tration was determined volumetrically 2 4 ...... 24 .09 using a solution of ferrous ion as the ND .08 titrant and diphenylamine sulfonate as 72...... 23 ND the endpoint indicator. Exact dilution

. 2 2 ND and sampling quantities were not re­

1 2 0 ...... 23 ND quired, as only relative cations were 144...... 23 .08 analyzed. A calculation using these ND Not determined. analyses for Ti, Al, and V was used to

'Values are the average of the samplesobtain the actual concentration 4 of, TiCl taken at each time. AICI 3 , and VC1 4 in the original sample. 6

RESULTS AND DISCUSSION

The results are shown in figuressaturated 3-7, solution. Systems below the and summarized in table 3. The resultscurve contain unsaturated solution. The for the 1 C T1 4 -A 1 C 1 3 binary system arecurve represents systems that are satu­ shown in figure 3. Systems containingrated and contain no solid3 .AICI compositions that are above the solubil­In figure 4, a comparison of existing ity curve contain both solid3 and AICI data (4-5, 7-10) and the results obtained

TEMPERATURE, °C

FIGURE 5. - Solubility of AICI3 in VCI4.

TEMPERATURE, °C

FIGURE 3. - Solubility of A3 1Cin 1T i Cl 4 .

TEMPERATURE, °C 100 75 50 25 0 TT ~ T NS A % * z o ▼ * KEY & < _▼ This investigation tr • Eingorn (1) A Morozov (£) A Druzhinina (7) □ Ehrlich (8) • Ruban 0 O Martynov (J_0) o

1 J _____ L 20 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 I0VT, K' 1 VCU, wt pci

FIGURE 4. - Solubility of AIC13 in T i C14 —a com­ FIGURE 6 . - Solubility of AIC13 in T i C l4 - VC 14 parison of results. system. 7

wt pet, compared with 7.2 wt pet as reported by Druzhinina and with 41.1 wt pet as interpolated from Morozov's data. Figure 5 depicts results for the VC14- AlCl^ binary system. The solubility of

AICI 3 in VC1 4 increased from 0.1 wt pet at 75° C to 1.6 wt pet at 150° C.

Figure 6 shows the solubility curves at 75°, 100°, 125°, and 150° C for the VCI„, wt pci TiCl 4 -A 1 C 1 3 -VC 1 4 system. Generally, the solubility of 3 AICI at 75° and 100° C FIGURE 7. - Effect of VCI4 concentration on3 A 1C 1 is less than1 wt pet for any mixture solubility in 4FiCl-V C 14 system. of TiCl4 and VC14. At 150° C the solu­ bility was found to be between 1 to 3

in this study for the4 -AlCl TiCl3 binary wt pet except when the4 is VC1 less than system are shown. The log mol fraction1 2 wt pet. versus 103/T plot shown as figure 4 Figure has 7 shows the effect of4 con VC1

been used by other researchers tocentration pre­ on the solubility3 of in AICI sent their data and obtain thermodynamicthe TiCl 4 -AICI 3 -VC1 4 system. The great­ values (_9, 10). The data found in theest solubility of3 AICI occurs in pure present study were in good agreementTiCl 4 (0 wt pet VC14) for each of the with data of Druzhinina(7), Ehrlich (JO, two temperatures shown. As4 isany VC1 Ruban (9), and Martynov (10) for the 75°introduced into the system, the solu­ to 125° C region. However, at 150°bility C of AICI3 decreases. Temperature the solubility of 3 AICI in TiCl 4 was 18.3

TABLE 3. - Summary of solubility1 datafor the TiCl4 -AICI 3 -VC1 4 system, 2 weight percent At 75° At 100°C At 125° C At 150° C

TiCl 4 AICI 3 VC1 4 TiCl 4 A1C1, VC1 4 TiCl 4 , VC1 4 TiCl 4 A1C1, VC1 4

99.76 0.24 - 0 - 99.36 0.64 - 0 - 97.18 2.82 - 0 - 80.75 19.25 - 0 -

99.76 .23 - 0 -* 99.33 .67 - 0 -* 97.16 2.84 - 0 -* 81.72 18.28 - 0 -*

99.77 .23 - 0 - 99.32 . 6 8 - 0 - 97.88 2 . 1 2 - 0 - 82.52 17.48 - 0 -

99.77 .23 - 0 - 99.36 .64 - 0 - 96.83 3.17 - 0 - 82.65 17.35 - 0 -

97.17 .18 2.65 96.88 .61 2.52 96.63 2.07 1.30 88.47 11.53 0 . 2 1 97.28 .13 2.59* 96.93 .64 2.43* 96.57 2.18 1.26* 87.08 11.92 .18*

91.85 .14 8 . 0 1 93.58 .35 6.07 92.00 1.14 6 . 8 6 85.70 9.46 4.83 91.95 .16 7.89* 92.39 .30 7.31* 91.49 1.08 7.43* 85.77 9.78 4.45*

83.10 .17 16.73 83.01 . 2 1 16.78 82.56 .47 16.97 83.62 3.24 13.15

82.57 . 1 1 17.32* 82.71 .18 17.10* 82.78 .49 16.73* 84.96 2.87 12.18* 65.48 .08 33.75 67 = 43 .14 32.43 67.95 .42 31.64 67.72 2.35 29.93

66.72 .14 33.14* 67.20 .13 32.66* 6 6 . 8 6 .43 32.71 69.89 2.19 27.92*

43.46 .06 45.63 .08 54.28 47.51 . 2 1 .99 56.47 52.28_ . _ _ * 46.96 52.05 43.40 .06 56.54 - 46.04 .09 ' 53.87* 44.86 • 2 7 54.88 47.32 1.27 51.41

0 . 1 1 99.89 0 .16 99.84 0 .60 99.40 0 1.62 98.38

0 . 1 0 99.90* 0 .18 99.82 0 .52 99.48 0 1.55 98.45*

'in addition, 1 analysis at 135° C showed 87.10 wt4 pet, 12.90 TiCl wt pet A1C15 , and 0 wt pet VC14. 2Values marked with an asterisk (*) correspond to values referred to in text that are usually the compositions2 d of of 2 the samples taken at a given time and temper­ ature. It was felt in these cases that the 2d sample better represented the vessel's contents, as 1 thest sample taken after the purge would have further purged the sam­ ple line and valve. However, very little difference was seen between the 2 sample analyses in any case. increases the solubility of3 A1C1 greatest effect when less than 20 wt pet

combinations of solutions, but it has4 the is present. VCI

CONCLUSIONS

The solubility of anhydrous3 in AICI 2. The AICI 3 is more soluble in TiCl4

TiCl 4 and VC1 4 was studied by the Bureauthan in VCI4 at any of the temperatures of Mines as part of a research projectstudied. to

make a homogeneous alloy sponge of 3.Ti, In the TiCl4 -AlCl 3 -VCI 4 ternary

A1, and V. From this effort the followsystem, the solubility 3of decreased A1C1

ing general conclusions may be drawn. from the values with a pure4 -A TiCl lCl 3

1. Data on the TiCl4 -AlCl 3 binary sys­system to the lower values found with a

tem are in good agreement with datapure of V C4 I-AICI 3 system at the respective previous investigators for the rangetemperatures. 75° to 125° C. REFERENCES

1. Krishnamurthy, S., R. G. Vogt, 9. Ruban, N. N . , V. D. Ponomarev, and F. H. Froes, and D. Eylon. DevelopmentsK. A. Vinogradova. (Solubility of Alu­ in Titanium Powder Metallurgy. U.S. Airminum Chloride in Titanium Tetrachlo­ Force AFWAL/MLLS, Memo. Rep. 83-1, 1983, ride.) Izv. Akad. Nauk Kaz. SSR, Ser. p . 5. Metall., Obogashch. Ogneuporov, No. 1, 2. Henry, J. L., W. E. A n a b l e , and 1961, pp. 33-40. J. L. Schaller. Pioneering Experiments10. Martynov, Yu. M., I C. Syrkina., on the Coreduction of Titanium, Aluminum,and E. A. Kreingol'd. Behavior of Solu­ and Vanadium Chlorides To Produce Titani­tions of Aluminum Trichloride in Titanium um Alloy Sponge. BuMines RI 7034, 1974,Tetrachloride. Russ. J. Phys. Chem., v.

2 0 p p . 42, 1968, pp. 999-1001. 3. Manlabs Inc. Report of the "Tita­11. Martynov, Yu. M., and I. G. Syr- nium Workshop." Cambridge, MA, 1983, pp.kina. Solubility of Iron (III) Chloride

60-69; available upon request from D.and A. Antimony (III) Chloride in 4 CCI,

Hansen, BuMines, Albany, OR. SiCl4, and SnCl 4 . Russ. J. Inorg. Chem., 4. Eingorn, L. N. (Titanium Halidesv. 10, 1965, pp. 510-512. as Solvents.) Ukr. Khim. Zh. (Russ.12. Liley, P. E., and W. R. Gambill. Ed.), v. 16, No. 4, 1950, pp. 404-413. Physical and Chemical Data. Section 3 5. Morozov, I. S., and D. Ya. Topty-in Chemical Engineers' Handbook, ed. by gin. Reactions of Titanium TetrachlorideR. H. Perry and C. H. Chilton. McGraw- With Tantalum, Niobium, and Aluminum Hill, 5th ed., 1973, table 3-1, pp. 3-6. Chlorides. J. Inorg. Chem., USSR, (Engl. 13. Clark, R. J. H. The Chemistry of

Transl.), v. 2, No.8 , 1957, pp. 322­ Titanium and Vanadium. An Introduction 333. to the Chemistry of the Early Transition

6 . Levin, E. M., C. R. Robbins, and Elements. Elsevier, 1968, p. 31. H. F. McMurdie. Phase Diagrams for Ce­14. Ivan, L., A. Imre, I. Imre-Baan, ramists. The American Ceramic Society,and T. Lakatos. (Producing Pure Vanadium Inc. , 1964, p. 401. Tetrachloride and Vanadium Trichloride.) 7. Druzhinina, N. K., and N. M. Femip. K u t . Intez. K ozlem., v. 10, 1971, Rusakova. (Solubility of Aluminum andpp. 171-185. Iron Chlorides in Titanium Tetrachlo­15. Kirk, R. E., and D. F. Othmer. ride.) Tr. Vses. Nauch.-Issled. , Alyum.-Encyclopedia of Chemical Technology. Wi­ Magnievyi Inst., No. 46, 1960, pp. 122—ley, 3d ed., 1983, v. 23, p. 693. 126. 16. Hildebrand, J. H., and R. L.

8 . Ehrlich, P., and G. Dietz. (The Scott. The Solubility of Nonelectro­ Solvent Powder of Titanium Tetrachloridelytes. Dover Publ., 1964, p. 295. for Solid Chlorides.) Z. Anorg.u. Allg. Chem., v. 305, 1960, pp. 158-168. IN T.- BU.O F MIN ES,PGH.,P A. 28255 U.S. GOVERNMENT PRINTING OFFICE: 1986 -605-017/40,035