BENJAMIN S. SANDERSON AND GEORGE E. MACWOOD
314
Vol. GO
THERMODYNAMICS OF THE TITANIUM CHLORIDES. 111. THE SURIJMATION PRESSURE OF TITANIUM TRICHLORIDE1s2 BY BENJAMIN S. SANDERSON AND GEORGE E. MACWOOD Contribution from the McPherson Chemical Laboratory, The Ohio State University, Columbus, Ohio Received Auguet 16, 1966
Experimental measurements have been made of the sublimation pressure of TiCl3over the temperature range 628-823OK. hy the transpiration method. From these results, the heat and free-energy of sublimation of 298°K. are found to be 42.0 f 0.5 kcal./mole and 29.6 f 1.0 kcal./mole. The heat of formation at 298°K. of gaseous TiCll based on this work and the previously reported heat of formation of the solid is -130.2 f 1.4kcal./mole.
I. Introduction As part of a program on the thermodynamic properties of the chlorides of titanium, the sublimation pressure of Tic4 has been studied by the transpiration method. Originally, it had been planned t o use this experimental method to measure both the sublimation equilibrium +
TiC13(s) = TiC13(g)
equilibrium,6 gave very good results for the sublimation pressure of TiCl3. 11. Experimental Method The apparatus used was a modification of that of Treadwell and Werner.' The transpiration cell is Ahown in Fig. 1 and the condenser for Ticla and trap for TiCh are shown in Fig. 2.
(1)
and the disproportionation equilibrium 2TiCl,(s) = TiCl,(g)
9
+ TiCll(s)
boll joint
(2)
g$ inner joint
It was found that the transpiration method, though inadequate for the study of the disproportionation
i 16 .-
Liquid nitrogen level
Conde nser (For detoils see Fig. 2 )
Pyrex wool
Cell mode of
pyrex
I Fig. 2.-Trap Inert gas
I?-----
ff
outer joint-
ii u
Fig. 1.-Transpiration
- Manometer
c _
and condenser.
Helium was used as the carrier gas. It was purified by flow through a train containing Drierite, hot copper turnings and then activated charcoal held at liquid nitrogen temperatures. The purified helium entered the cell as indicated in Fig. 1, passed through the charge, and then the saturated gae assed out through the TiCla-condenser into the TiCL-trap . rom the trap, the helium passed through a drying tube (to prevent back-diffusion of water vapor) into a 5-gallon flask where the helium was collected over water. The volume of displaced water was measured by collecting it in a two-liter volumetric flask from which the volume of helium was determined. The furnace was of the nichrome resistance-type arranged so that it could be raised and lowered over the cell. This allowed the furnace t o be brought to temperature before
F
cell.
(1) This paper presents the results of one phase of a program sponsored by the Department of Navy, Office of Naval Research, under Contract No. Nonr-495(06). (2) Taken in part froin the dissertation submitted by Benjamin 8. Sanderson in partial fulfillment of the requirements for the Ph.D. degree a t the Ohio State University, March, 1955. (3) H. van Wsrtenburg. Z . Elektrocham., 19, 485 (1913). (4) E. Thiele, Ann. Physik. 14, 937 (1932). (5) K . Jellinek and C. A. Rosner, 2.physik. Cham., A148,51 (1929).
-
(6) B. S. Sanderson and G . E. MacWood, THISJOURNAL, 60, 316 (1956). (7) W. D. Treadwell and W. Werner, Hslu. Chim. Acta, 86, 1436 (1953).
SUBLIMATION PRESSURE OF TITANIUM TRICHLORIDE
Mar., 1956
lowering it into position over the cell. A Foxboro controller maintained the temperature of the charge within f0.5'. The cell temperature was measured with an ironconstantan thermocouple, calibrated in situ. The amount of TiCI3 condensed out was determined either by weighing or by spectrophotometric analysis using the titanium-thymol complex described by Griel and Robinson.* The amount of TiClr collected in the trap was determined by weighing.
05
-05
7-
-g
-10
n* -1.5
-2 c
-25
IV. Procedure
I
The TiCla was loaded into the transpiration cell with the aid of a polyethylene "dry bag." A current of dry argon was circulated through the bag which was tied over the standard taper of the cell. All of the equipment re uired for loading the cell, as well as the TiCls, was put ins%e the bag before closing it. After loading the cell, it was inverted and the Tic&manipulated into its proper position as shown in Fig. 1. The cell was weighed and the weight of sample determined. Then the cell was mounted in place and connected to the helium gas train. The weighed condenser and trap were put into place while a current of helium passed through the system. The furnace was brought to temperature in a position above the cell after which it was lowered around the cell. Thirty minutes was allowed for the sample t o reach temerature equilibrium before the helium flow waa started. g y proper adjustments, the total pressure and the flow rate were rapidly established at the desired values. At the end of a run, the furnace was lifted away from the cell. While the cell was cooling, the current through the heating tape around its lower end was continued in order to prevent TiCL from condensing in the TiCL-condenser When t8hecell had cooled, the condenser was removed and the amount of Tic18 collected determined either by weighing or spectrophotometric analysis. The amount of TiCl, collected in the trap was determined by weighing.
.
V. Results The results of the sublimation measurements are summarized in Table I. The values for the weight of Ticla collected reported t o 0.1 mg. were determined by weighing, while those reported to 0.1 p g. were determined by spectrophotometric analysis. The average Cl/Ti ratios given in the table were calculated from the amounts of titanium and chlorine which were lost in each run. The logarithms of the sublimation pressures are plotted against the reciprocal of the absolute temperature in Fig. 3. The straight line drawn through the points was determined by least squares and is given by the empirical equation logp(mm.) =
R.
- 8296 + 10.401
(8) J. V. Griel and J. Robinson, A n d . Chsrn., 8 8 , 1871 (1951). (9) J. M. Sherfey, J . Reeearch N a t l . BUT.Stands., 46, 299 (1951).
3
00
III. Material The Tic18 was made by an adaptation of the method of Sherfcyg developed in this Laboratory by Mr. John Reed. In this method, titanium tetrachloride was refluxed over a red-hot tungsten filament while hydrogen was passed through the apparatus. The tetrachloride was reduced to trichloride and condensed on the walls of the apparatus. The refluxing tetrachloride washed the solid trichloride into a vessel with a fritted-glass bottom. After most of the tetrachloride had been filtered off the solid, the trichloride was transferred to Pyrex tubes about two inches in diameter and ten inches long. These tubes, after filling with crude charge, were pumped down and sealed off. They were then heated in a resistance furnace to 500'. The TiClt sublimed to the cooler portion of the tube. Samples of both the crude and resublimed TiCh were analyzed at the National Bureau of Standards for heavy metals. Neither sample contained metal impurities in excess of 15 p.p.m. Titanium and chlorine analyses of the .trichloride gave an average empirical formula of T i C l s . ~ ~indicating , the presence of a small amount of TiCI,.
315
I
I
I
I50
140
Ix)
I
T
x 103
Fig. 3.-Sublimation
I
io
OK,',
pressure of TiCla.
The standard error of the slope is =k380,that of the intercept is h0.51. The average heat of sublimation for the temperature range 628 to 823"K., calculated from (3), is 38.0 1.7 kcal./mole. Heat and free energy of sublimation calculations were made, after the manner of Kelley,'O based on the present measurements and the following estimated heat capacity equations for solid and gaseous TiCla
*
+ +
c,(s) = 23 4 x 10-3 T - 1.7 c,,(g) = 19.1 0.18 x 103 T - 1.7
x x
105 ~
106
-
T-2
2
(4) (5)
The results of these calculations for the sublimation equilibrium are AC, = -3.9 - 3.8 X lO-*T (6) AHo = 43300 - 3.9T - 1.9 X (7) AP = 43300 3.9T In T 1.9 X 10-3T2 - 68.717' (8) AHOZOS= 42.0 kcal./mole AFOZSS= 29.6 kcal./mole ASOZSS= 41.6 cal./deg. mole
+
+
TABLE I SUBLIMATION PRESSURE OF Ticla T, OK.
628 686 699 738 780 788 789 800 823
Flow rate, GO../
mm. 11.0 13.0 19.6 15.0 13.0 27.0 13.6 14.0 14.0
Wt. TiCls,
Vol.
mg.
cc.
0.0128 0.0917 16.7 6.2 31.4 37.4 40.9 48.2 95.2
6007.0 6007.1 8008.3 6006.8 6010.7 5929.0 6006.9 5865.0 4030.6
He.
Moles TiClr X 104 4.7 4.3 17.9 33.9 32.1 44.7 33.3 45.5 28.4
P,T i C l ,
Cl/
mm.
2.58 1.85 2.67 1.23 6.18 7.14 8.13 9.75 2.78
X X X X X X X X
10-8
10-2 10-2
10-1 10-1
IO-' 10-1 10-1
2.996 2.267 2.735 2.644 2.546 2.939 2.413 2.779 2.283
(10) K. K. Kelley, United States Bureau of Mines, Bulletin, 476 (1949).
BENJAMIN S. SANDERSON AND GEORGE E. MACWOOD
316
The following thermodynamic constants for Tic13 obtained from the present results, heat of formation’ and disproportionation datas are considered as the most reliable values at present available AH&&) = -130.2 i 1.4 kcal./mole AHrop98(s)= -172.2 & 0.8 kcal./mole SoZas(g) = 73.8 i 2 cal./deg. mole S0ta8(s)= 33.3 2 cal./deg. mole
*
VI. Discussion Recently Farber and Darnell12 have reported values for the sublimation pressure of Tic13 determined by the Knudsen ‘method. Their values are considerably lower than those reported here. On the basis of work done in this Laboratory on the disproportionation equilibrium of TiCla6,it appears (11) D. G. Clifton and G. E. MacWood, THISJOURNAL,60, 309 (1956). (12) M. Farber and A.
G. Damell, ibid.,
probable that these low values are due to a low accommodation coefficient for gaseous TiC13. The agreement between the values of the heat of sublimation obtained by the two methods may imply that the temperature coefficient of the accommodation coefficient is very small as has been observed for gaseous TiCL in study of the disproportionation of Tic13 by the Knudsen method.12 Skinner and Ruehrwein,I3 have reported measurements using the transpiration method. Their experimental values, although admittedly low, are of the same order of magnitude as those reported here. Their pressures, obtained by estimating the degree of saturation of the carrier gas, appear to be high. This results in a slightly higher entropy of sublimation at 29SoK., 43.3 compared with 41.6, since the heats of sublimation at 298°K. agree within the precision of measurements. (13)
69, 156 (1955).
Vol. 60
G.B. Skinner and R. A. Ruehrwein, ibid..
69, 113 (1955).
THERMODYNAMICS OF THE TITANIUM CHLORIDES. IV. THE DISPROPORTIONATION OF TITANIUM TRICHLORIDES1*2 BY BENJAMIN S. SANDERSON AND GEORGE E. MACWOOD Contribution from the McPherson Chemical Laboratory, The Ohio Stale University, Columbus, Ohio Raceivsd Awust 16, 1866
The disproportionation equilibrium of TiC13 was studied in the temperature range 593-821°K. by an effusion and static method. The two sets of measurements do not agree and indicate that the vaporization coefficient of TiCl,(g) is of the order of 10-4. On th.e basis of the static measurements and previous work, a consistent set of thermodynamic properties of TiClp and TiCl, are given.
I. Introduction This is the fourth paper on the thermodynamic properties of titanium chlorides being investigated in this L a b o r a t ~ r y ~and - ~ gives the results of the investigation of the disproportionation equilibrium of TiC13 2TiCla(s) = TiClz(s)
+ TiCL(g)
(1)
The course of the equilibrium as a function of temperature was studied by following the pressure of the titanium tetrachloride. The usual methods for measuring vapor pressures may be employed and, in this investigation, an effusion and static method were used. 11. Experimental 1. Materials.-The TiClr and TiCh were prepared as previously described.’-6 2. Apparatus. (a) Knudsen Method.-The apparatus is shown in Fig. 1. The body of the cell, shown in Fig. 2, was machined from stainless steel. The diaphragm, which contained the orifice, was made of 3-mil nickel sheet. It was held firmly in place by a screw-collar, a stainless steel washer being used to prevent tearing of the thin diaphragm. The weight loss of the cell was determined by following the (1) This paper presents the results of one phase of a program wonaored by the Department of Navy, Office of Naval Research, under Contract No. Nonr-495(06). (2) Taken in part from the dissertation submitted b y B. S. S. in partial fulfillment of the requirements for the Ph.D. degree a t The Ohio State University, Marcb, 1955. (3) I. D. G. Clifton and G. E. MacWood, THIB JOURNAL. 60, 309 (1956). (4) 11, D. C. Clifton and 0. E. MaoWood, ibid., 60. 311 (1956). (5) 111, B. S. Sanderaon and G . E. MacWood, ibid., 80, 314 (1956).
change in length of a quartz helical spring, by comparison with a standard meter bar and cathetometer. The cathetometer has a Filar-micrometer eyepiece which could be read to 0.008 mm. The spring, obtained from the Houston Technical Laboratories, had a sensitivity of 5.092 cm./g. It had a maximum load capacity of ten grams. The cell was suspended from the spring by a I-mil molybdenum mire. The portion of the apparatus housing the helix was protected on three sides to minimize random temperature fluctuations. The cell was heated in a vertical resistance furnace, controlled by means of a Thyratron regulator. The sensing element for the regulator was B and S No. 30 platinum wire wound around the central portion of a nichrome tube which fitted between the glass tube containing the cell and the inside wall of the furnace. The wire was wound non-inductively and had a resistance of 35 ohms at room temperature. The resulting temperature regulation was within r t O . 5 O . The cell temperature was determined by means of a chromel-alumel thermocouple, placed about 1 cm. below the cell. From this measurement and preliminary calibration of the furnace temperature gradient as a function of position and temperature, a corrected value of the cell temperature was obtained. The system was connected to a conventional high-vacuum system for maintaining low pressures during measurement. (b) Static Method.-The Pyrex apparatus used for the static pressure determinations is shown in Fig. 3. The stopcocks were lubricated with a special fluorinated hydrocarbon grease. The tube for measuring the liquid T i c & was calibrated against water with the aid of a weight-buret. The gage used t o measure pressures inside the system was a “Spoon-gage.’’ I n order to increase its sensitivity, while still maintaining mechanical strength, three “spoons” in series were used. A number of these gages were used during the course of the invc tigations. Their average sensitivity was about 0.03 mm. The outside jacket of the gage was connected through a two-way stopcock to vacuum pump or a supply of dry argon. The balancing argon pressure was measured with a tipping McLeod gage in the range
