1i:!
J.UfES
w. EDWARDS, HERRICKL. JOHNSTON .4ND PAUL E. BLACKBURN
imposing somewhat less stringent orientation requirements. It might also be predicted that the activation energy would be higher because of the greater strength of the S-0 bond than the S-S bond. If i t is assumed that the activation energy is proportional to the strength of the bond broken,31 a very rough estimate might be made from the relation
Vol. 73
water. It would be interesting t o investigate the exchange of aqueous sulfite with sulfate in sealed tubes at temperatures above 250'. Acknowledgments.-The authors wish to express their appreciation to Dr. E. L. King for helpful discussions during the course of these investigations. This work has been supported in part by the United States Atomic Energy Commission and in part by the University Research Committee with funds supplied by the Wisconsin Alumni Research Foundation.
by estimating the bond strengths from thermochemical values for the heat of formation of SO*and SzOj" from gaseous atoms (assuming hydration Summary effects cancel or are negligible). Taking the values 213.8, 145.5, 148.5, -66.3 and -59.1 kcal. mole-' The transfer of the outer sulfur atom of thioas the heats of formation of sulfate, thiosulfate, sul- sulfate ion to sulfite ion occurs in the temperature fite, sulfur atoms and oxygen atoms, respectively, 32 range of 60-100° a t a rate which is proportional to the estimated bond strengths are S-0 126.4 and the first power of the sulfite concentration and also S-S 63.3 kcal./mole, and the estimated E for the of the thiosulfate concentration. The rate and sulfate exchange is 28.7 kcal. mole-'. If the activation energy appear to be essentially infrequency factor is the same for the sulfate exchange dependent of PH in the range from PH 5 to 14. as for the thiosulfate exchange, this would result The second order velocity constant is given by the in a half-time for the reaction in 0.1 AT sulfite and relation k = 2.3 X lo6 e-14$500/RT liters mole-' 0.1 -11 sulfate of 3400 years a t loo', of 20 days sec.-'. a t 250' and of 1 hr. a t 400'. Voge3has shown that The evidence indicates that the equilibrium of no exchange occurs between sulfite and sulfate on thiosulfate with sulfur and sulfite is not a part of heating in 0.1 N acid or alkaline solution for 35 the mechanism of sulfur transfer. Other factors hours at 100' but has observed a slow exchange to be considered with regard to the mechanism of between sulfur dioxide and sulfur trioxide gases a t this reaction are discussed. 335', which was catalyzed by water. S 0 r 1 - i ~ ~ ~X preliminary value of 15 kcal. mole-' for the has observed an exchange between sulfur dioxide activation energy of the exchange of sulfide ion and concentrated sulfuric acid with a readily with thiosulfate has been determined. measurable rate in the temperature range of about Previous evidence indicating that the sulfur 200', and with a n apparent activation energy of atoms of thiosulfate are not chemically equivalent 28.8 kcal. mole-'. Huston and Norrisj3 have has been confirmed and extended to show that found that the rate of exchange between inten- exchange between them does not occur in 5 hours sively dried sulfur dioxide and sulfur trioxide gases a t 125' nor when the thiosulfate is precipitated as is conveniently measurable in the range of 400 triethylenediaminenickel thiosulfate and suhto 440' and, as indicated by Voge, that the re- seyuently dissolved in hydrochloric acid. action is heterogeneous and strongly catalyzed by -1 reproducible routine method of determining the relative specific activity of S35in barium sulfate (31) J. 0. Hirschfelder, J . Chem. Phys., 9, 645 (1941). (32) Bichowsky and Rossini, "Thermochemistry of Chemical Subprecipitates is described. stances," Reinhold Publishing Corp , New York, N \', (33) Noms, TFXIS JOURNAL, 72, 1220 (1950)
1936.
MADISOX 6, WIS. I_.__--_-
[ C O N T R I R U T I O L FROM THE CRYOGENIC
LABORATORS AND
RECEIVED JUNE 26, 1950
__-
THE D E P A R T X E N T OF CHEMISTRY,
THEO H I O S T A T E UNIVERSITY]
Vapor Pressure of Inorganic Substances. IV. Tantalum between 2624 and 2943 BY JAMES It-.EDWARDS, HERRICK I; TOHNSTON AND PAULE. BLACKBURN Although there have been two previous investigations of the vapor pressure of the results were in considerable disagreement. We have, therefore, redetermined the vapor pressure of tantalum, in this Laboratory, using the method of I. Langmuir. The method and techniques used by us have been described in earlier papers4 from this Laboratory. (1) This work was supported in part by The Officeof Naval Research under contract with The Ohio State University Research Foundation. (2) D. B. Langmuir and L. Malter, Phys. Rev., 66, 743 (1949). (3) M. D. Fiske, ibid., 61, 513 (1942). (4) (a) H.L. Johnston and A. L. Marshall, THISJOURNAL, 62, 1382 (1940); (b) R. B. Holden, R. Speiser and H. L. Johnston, ibid., 70, 3897 (1948); (c) R. Speiser and H. L. Johnston, Preprint No 1 1 Thirty-First annual Convention of the American Society for hfclai.. Cleveland, Ohio, r k t . 1; 2:. :%ii
Procedure.-A sample of better than 99.9% pure taiitalum was obtained from the Fansteel Metallurgical Corp. An analysis, conducted by the manufacturer, showed that the principal impurities were carbon and iron and that neither of these was present to more than 0.03%. The rate a t which the surface of a sample of tantalum evaporated in wucuowas determined. The sample was in the form of a hollow cylinder one inch high by one inch diameter with a L/* inch wall thickness. The ends of the cylinder were each covered with several split radiation shields made of tantalum sheet. This arrangement practically eliminated evaporation from the ends since the split shields did not heat efficiently. Three longitudinal holes drilled into the cylinder wall and one hole through the upper set of radiation shields served as black-body holes. The sample was placed in an evacuated metal chamber and heated by radio frequency induction. Temperatures werc measured by means of it carefully calibrated disappearingfilaincnt optical pyronicter sighted on the four black-body
holes. The maximum uncertainty of the temperature scale in the range of the investigation was about 8 degrees, based on a calibration by the National Bureau of Standards. The probable error of determinations, however, was 3°K. with an additional error of 2’K. introduced by the window correction. The surface temperature of the radiation shields was determined by correcting observed brightness temperatures for spectral emissivity.
The sample was weighed to within *0.00014 gram before and after each run. During the investigation, the sample was found to increase in size from run t o run, probably as a result of rapid heating and cooling. I n calculating the effective surface area, i t was assumed that thermal expansion was uniform so that the area a t high temperature, A T ,was given by =
AT
A , (1
+2 F )
where the subscript Y refers to room temperature. The proportional linear thermal expansion, AL/Lo, was determined earlier a t this Laboratory5 as 6.080 X 1 W 6 ( T
ALjLo
173
VAPORPRESSUREOF TANTALUM
Jan., 1951
-
291)
+ 7.50 X
lO-’”(T 291)* (2)
-
where T i s the absolute temperature. T o account for small temperature fluctuations during each run and for evaporation during initial heating and final cooling, an averaging process was used t o calculate the effective time, t , of each run.4a I t was assumed that the rate of evaporation, m ( T ) , as a function of temperature, T, could be expressed by the equation log m ( T ) = a / T
+c
J v z ( T ) dt/m(T.,)
(4)
where integration is performed graphically. The effective total weight loss, We, was calculated from the total weight loss W t by the relation
we= wt(loss
-7.0 a; M
!A -7.5
-8.0 3.40
3.50
3.60 lo4//.
loss by cylinder by cylinder loss by ends
+
3.90
3.80
3.70
--, Langmuir
Fig. 1.-Vapor pressure of tantalum: and Malter; @, Fiske; 0, this research; search, eqn. 9.
-, this re-
The reliability of the data was checked by computing values of LL@,the heat of sublimation a t the absolute zero, from the several vapor pressure points by the equation
in which the parenthetical terms are the free energy functions for solid tantalum and for tantalum vapor, respectively. Values of the solid free energy function were obtained from the equation ( p- H ; ) / T = 559/T - 9.773 X T - 9.6591 log T
(3)
where a and c are constants obtained from the actual periods a t high temperatures and from the weight losses of the various runs. Equation 3 was then used to calculate m ( T ) and m(Tav) in the equation t =
-6.5
+ 17.26
(7)
based on calorimetric data between 1100 and 2500°K.,6 obtained in this Laboratory. Free energy functions for the vapor were computed statistically from the relationship
(F” - H ; ) / T = RIn[RT(2nmkT)8/~/Nit3]+ R In Qe (8) where Q e is the electronic partition function7 and the other symbols have their usual significance.8 The results of these calculations are presented in Table 11. The average value of AH:thus obtained is 185.5 kcal., with a probable experimental error of 0.3 kcal. Values of were also calculated from
a
wherein the parenthetical quantities were calculzted from equation 3.
TABLE I1 HEATOF EVAPORATION OF TANTALUM
Discussion of Results The experimental results are summarized in Table I, and plotted in Fig. 1 as log p u s . 1/T, along with the data of Langmuir2 and of F i ~ k e . The ~ solid curve is computed from equation 9. TABLE I EVAPORATION OF TANTALUM
Run
OK.
Effective time, sec.
16 23
2624 2638 2760 2839 2888 2925 2948
15860 17962 3784 2066 3222 2246 2050
Temp.,
20 19 17 18 21
Effecnz(T) tive g.-cm.-Z area, sec.3 Weight loss, 6 . Total Effective sq. cm. X 108 0.02484 0.0243 21.17 .03359 .0328 21.31 .03724 ,0363 21.32 .04892 ,0476 21.35 ,13329 ,1295 21.34 ,13074 ,1269 21.29 ,18158 ,1762 21.41
7.234 8.579 45.01 108.4 188.4 265.4 401.4
Pressure atm. X 109
6.216 7.390 39.66 96.92 169.8 240.8 366.5
( 5 ) James W. Edwards, Rudolph Speiser and Herrick L. Johnston, J . A p p l i e d Phys.
t o he published soon in
-RInfi, Run Temp., cal. mole-1 no. OK. deg.-1
16 23 20
2624 2638 2760
19
2839 2888
17
15
21
2925 2948
___-
37.55 37.21 33.87 32.09 30.98 30.28 29.45
cal. mole-’ deg. -1
cal. mole -1 deg. -1
18.12 18.15 18.47 18.67 18.80 18.89 18.94
51.379 51.415 51.722 51.916 52.034 52.121 52.175
Av. Deviation from mean
AH:, kcal.
mole -1
185.8 185.9 185.2 185.5 185.5 185.8 184.8 185.5 0.3
(6) G. W. Zeigler, J r . , R. Speiser and H. L. Johnston, to be published. (7) Term designations and values taken from Charlotte E. Moore, “Term Designations for Excitation Potentials,” The Princeton Observatory, Princeton, N. J., 1934. (8) Values of physical constants used for all calculations are from R. P. 1634, National Bureau of Standards, February 1948.
GORDON B. SKINNER, JAMES W. EDWARDS AND HERRICK L. JOHNSTON
174
VOl. 73
TABLE 111 the rates of evaporation published by Langmuir , ~ these results are HEAT OF EVAPORATION OF TASTALUM USIYC FISKE'S and Malter2 and by F i ~ k e and summarized in Tables I11 and IV. As is seen, the DATAFOR RATEOF EVAPORATION
2633 2649 2700 2737 2804 2807 2850
34.80 34.85 33.94 33.12 31.33 31.90 30.59
28.8 28.0 43.8 65.9 160 120 230
33.26 33.26 33.26 33 . 2d 33 .2.5 38.25 3 3 , 24
179.2 180.4 181.4 181.7 181.1 182.9 181.9 Av. 181.2 Mean deviation 0.9
data of Langmuir and Malter are in exact accord with our own while Fiske's vapor pressures are somewhat high. By use of equation 6, in conjunction with equations 7 and 8 and the average value of 185.5 kcal. for AIG, we obtain the vapor pressure equation K In p = - (185.5 X 103)/T 3.7 X 10-jT 8.4 X 10-8T24- 32.87 (9) where p is pressure in atmospheres and R is the gas constant in cal./mole/deg.
+
Summary
Iv The vapor pressure of tantaluni was determined over the temperature 262.1 to 2948°K. by measuring HEATO F EVAPORATION O F TANTALUM USING D. B. LAS(;the rate a t which a metal surface evaporates in MUIR AND L. MALTER'S DATAFOR RATEOF EVAPORATION TABLE
FQ-&
- - K i n p,
OK.
m(T), cm.-1 sec.-l x 10s
2000 2200 2400 2600 2800 3000 3200
0.000163 ,00978 .304 5.54 66.1 579 3820
'Temp.,
g.
-'(L-)*
cal. mole-1 deg.-1
CXI.
mole-' deg.-'
33 28 33.28 33.27 33.27 33.25 33.22 33.19 Av. AX can devi,ition
59.08 50.85 43.94 38.09 3.3.09 36 i t 24.90
[CONTRIBUTION FROM I l i h CRrOGI3hIC
AH:?
kcal. mole-'
184 7 185 1 185 3 185.5 185 7 185.8 185.9 185.4 0.3
LABORATORY A N D THE
a
vacuum. Values of were calculated from the individual vapor pressures and showed no appreciable trend, the average value being 185.5 * 0.3 kcal. By combining this value with free energy functions for solid and gaseous tantalum, the following vapor pressure equation was obtained I< I n p = - (185.5 X 10"j/T 3.7 X 10-4T -
+
8.4
x
10-8P
+ 32.87
Our experimental results agree accurately with the earlier work of Langmuir and Malter, but disagree with those of Fiske. COLUMBUS 10, OHIO
RECEIVED JULY 5, 1950
D E P A R T M E X T O F CHEMISTRY, T H E OHIO S r A l E ~ ; h I V C R 5 I T Y ]
The Vapor Pressure of Inorganic Substances. V. 2054°K.'
Zirconium between 1949 and
BY GORDON B. SKINNER, JAMES W. EDWARDS AND HERRICK L. JOHNSTON Introduction The vapor pressure of zirconium has been measured a t this Laboratory by the vacuum evaporation method of Langmuir2 as modified by Marshall, Dornte and Norton.* The method consists of measuring the rate a t which a metal evaporates into a vacuum from a ring supported on a wire tripod and heated inductively by high frequency current. Zirconium is difficult to volatilize apparently due to tightly adhering film that covers its surface even under relatively high vacuum conditions. We have overcome this obstacle by working a t pressures of less than 2 X lo-' mm. Sample Purity.-A sample of zirconium was vacuum cast from crystalline zirconium which was prepared by the iodide method and furnislied through the courtesy of the Philips Research Laboratory of Eindehoven, Holland. A spectrographic analysis4 showed that the final sample con(1) This work was supported in part by the Office of Naval Research under contract with the Ohio State University Cryogenic Laboratory (2) I. Langmnir, Phrs. Reo., 9, 329 (1913). (3) Marshall, Dornte and Norton, THISJOURNAL, 69, 1161 (1937). (4) Analyses were made by Mr. John Center, Chief Analyst of the Hattelle Memorial Institute, Columbus, Ohio.
tained 0.99 atom per cent. of hafnium, 0.05 atom per cent. wolfram, and 0.37 atom per cent. of all other impurities including silicon and aluminum Wolfram is known to have a lower vapor pressure than that measured for zirconium. Vapor pressures are not known for hafnium, but general chemical considerations indicate that its vapor pressures should be a t most one-tenth those of zirconium a t any given temperature. The presence of these two metals should make the measured vapor pressures of the zirconium sample approximately lG& low, but since this value lies within the experimental error of our data and since i t would be partially conipensated by the presence of low percentages of the more volatile impurities, such as aluminum, we have applied no correction. Experimental Apparatus and Procedure The sample was machined t o the form of an annular ring of about 1-in. 0.d. by 3/8-in. i.d. '/2-in. high, dnd was supported on three wolfram rods iii the Pyrex apparatus shown in Fig. 1. With the exception of the liquid air trap 0, the system mas baked out before rach ruii. The charrod1 trap was baked out for about 86 hoiiri d t 450' a n d thc remainder f l , t i n c d for 1 hour ,it .iImut 100". The app.ir.itui v a 5 then
