March 1947
INDUSTRIAL AND ENGINEERING CHEMISTRY
42 1
vapors passed through the reactor at $00" C. Rectificntion of TABLE 11. PHYSICAL PROPERTIES OF ~ , ~ , ~ - T R I F L V O R Opu&edproducts gave CClF.&HBrCHl, CC1F?CHICH,13r, and DIHALOPROPAKES unreacted CC1F,CH,CH3. A residue of 17 grams was riot idenBoiling Pooint, Boiling Point, ' C. tified. The yields of CClF,CHBrCH, and CC1F2CHlCH2Br
Dihromopropanes CFICBrzCHa CFaCH2CHBn CFaCHBrCHzBr
Predicted
88.1
112 116.6
Found
Dichloropropanes
iii'
CFiCCliCHs CFICHICHCI~. CFzCHClCHzCl
ll5,8
C. 48.1 72 76,6
by a preponderance of the nionobromide in the reaction product and only small amounts of polybromides. l,l,l-Trifluoropropane (2.60 moles) x i s bubbled into liquid bromine a t the rate of 0.227 mole per hour, a total of 0.87 mole of bromine being vaporized. The temperature of tlie bromine T V ~ ? maintained bein-een 33' and 40' C. The mixed vapors nere passed through the reactor a t a temperature of 4.50' C.; the contact time \ v u about 20 seconds. Upon rectification the brominated product \vas found to contain CF3CHBrCR3 (0.22 mole), CF3CH?CIi?Br(0.38 mole), and CFaCH2CHBry (0.20 mole). Although it K ~ J&ion-ii that substitutive clilorination of CY,CII,CH, occurs almost exclusively on the omepn ciirlion atom(,'), considerable bromination occurs also on the middle cnibon atom. Identity of the broniofluoropropanes was establislied by fluorine and bromine aiialyses. boiling point analogy, and molar refractirity. There are three possible dibromo-l,l, 1-triiiuoropropanes obtainable by bromination of CF,CHzCH3without renrrangement : CFKHBiCHLRr, CFsCBrlCH,, and CF,CH?CHBr,. .is Table I1 shows, CF3CBr,CH3would boil about 25' C. loiver than the dibromotrifluoropropane (111' C.) obtained in this reaction. Consequently it \Vas excluded from further consideration. "3Dibromo-l,l,l-trifluoropropanewas prepared by adding bromine to CF,CH=CH2 obtained b y dehydrochlorination of CF3CH2CH2CI. The physical properties of CF3CHBrCH2Brobtained thus are different from those of the dibromo compound obtained by thermal bromination of CF,CH,CH, (Table I), Therefore the dibromo-1 ,l,l-trifluoropropane IVRP assigned the formula CFaCHZCHBr2. I-Chloro-1,l-difluoropropane (1.37 moles) lvas bubbled through liquid bromine at 2.5' C. over an 11-hour period. The mixed
were 376 and 1 2 5 , respectively.
2-Bromo-l-chloro-l,l-difluoropropane (CClF&HBrCH,) was also prepared by the fluorinatiou of C B r C l & H B C H , n i t h hydrogen fluoride and mercuric oxide. For emmple, i~ 730-ml. (0.83 nickel-lined autoclave {vas charged with CBrC12CHUK~€IJ mole), hydrogen fluoride (7.5 moles), and mercuric oxide (1.5 nio1es)in the manner previously reported ( 5 ) . Thca :intoelave i ~ a s then sealed and heated a t 88 C. for 72 hours. Iiectification of tlie products gave a 9 . 6 5 yield of CF3CHBrCH,:ind n 5 . 0 7 yield of CCIFzCHBrCH,. 2-Chloro-l,l,l-trifluoropropane (0.34 mule) a :i- huhbled tlirougli liquid bromine a t 45-30" C. The mixed vapclrs were passed at a uniform rate through a 120-cm. reactor of 12-mm. I'rrer t u b ing during a 2-hour period. Products were collected under ice water, separated, and dried. Rectification shoived t h a t t,he product coiitaiiietl 0.045 mole of unreacted CF3CHCICEI., and 0.037 mole of CF3CRrC'lCH.. So CF,CHCICI Il13i, from the product,. ACKNOWLEDGXIENT
The authors wish to express t,hrir appreciation t o the Mallirickrodt Chemical Works for the financial assistance which made this investigation possible. LITERATURE CITED
(1) Brice, T. J., Pearlson, W. H.. and Simons, J. H., J . Am. Chem. SOC.,68, 968-9 (1946).
(2) Henne, A. L., and Whaley, A. XI., Ibid., 64, 1157 (1942). (3) McBee, E. T., Henne, A. L., Hasa, H. B., and Elmore, N.,Ibid., 62, 3340-1 (1940). (4) McBee, E. T., et al., ISD.EXG.CKEX.,39,409 (1947). ( 5 ) McBee, E. T., et al., J . Am. Chem. SOC.,submitted for publication. ( 6 ) Robbins, B. H., J. Pharmacol., 86, 197-204 (1946). AssraacTED from a doctoral thesis of W . G. Toland, J r faculty of Purdue University.
, submitted to the
Preparation and properties of
SULFUR HEXAFLUORIDE Walter C. Schumb MASSACHUSETTS INSTITUTE OF TECHNOLOGY, CAMBRIDGE, MASS.
LTHOUGH the chemical inertness and thermal stability of certain nonmetal fluorides, such as sulfur hexafluoride, carbon tetrafluoride, and nitrogen trifluoride, have been known for many years, practical applications of these substances have been s l o in ~ developing. Thus in 1900 the discoverers of sulfur hexafluoride, Noissan and Lebeau ( 6 ) , described it as a colorless. tasteless, incombustible gas, with a n inertness resembling nitrogen, unchanged even at the softening temperature of hard glass and unaffected b>- water or caustic potash. Some of its physical constants were reported by them and were later somewhat modified by Prideaux (8) and by Schumb and Gamble (9). The excellent properties of this gas, as well as Freon 12 (CClCFJ for high voltage insulation were observed by the General Electric Company some years ago ( 2 ) , as ne11 as in tlie high voltage laboratory of this institute, where the gas was employed in tests as an insulating medium for an electrostatic generator Ix-ith applied potentials as high as 5.6 million volti direct current across the gap between terminal and tank ( 1 ) .
I n order to be able to carry out larger scale iests n-ith larger generators than were employed in these earlier tebts, some 1500 pounds of sulfur hexafluoride w r e prepared for the high voltage project in 1942 by the Hooker Electrochemical Company by the method described belon-; with a procedure worked out in this e laboratory, approximately 130 pounds more of the nnbst ~ n r were prepared and stored as liquid in steel cylinders. PREPARATION OF SULFUR HEXAFLUORIDE
The method employed in the laboratory con4sted uf burning coarsely crushed roll (lump) sulfur in a current of fluorine gas in a steel box, provided with staggered horizontal shelves, each bearing about 8 pounds of sulfur. The top of the box was securely bolted down and provided with an asbestos gasket to make i t gastight. (The method employed a t the Hooker Electrochemical Company was to burn molten sulfur, contained in a horizontal eteel pipe, in a stream of fluorine.) The fluorine passed in a cir-
INDUSTRIAL AND ENGINEERING CHEMISTRY
422
TAR1.E
I.
Vo!. 39, No. 3
PEO13ERTIES OF S U L F U R H E X A F L U O R I D E
REFERENCE 5.106 (air = 1) at 20.0' C , i 5 3 . 5 m m . 2.51 a t m.p. (-50' C.) 2 683 a t - 196' C. 2 . i 4 a t -273' C. (estd.) Liquid: 1.787 a t -39' C . 1819 at -450 1 0 1 at -503
DESSITYG a s :
Solid:
c. c.
-50.83 -50.80
JIELTINQ POIST
SCBLI\IATIOI
TEMPERATURE -63.i'
CRITICAL 'rSa1PERATVRE HEATOF F n s ~ o s
WEAT O F VAPORIZATION HEATOF SCBLIMATION
TROUTON CONSTAST '~
* 0.2' c.
c.
*
0.5' C . C. -63.5' C. -63.8' C . 540 c. 1390 cal. 1100 cal 4500 cal. at b.1, 5640 cal. 5570 cnl. 22 -50'
(G.48)
SURFACE TEXSION A T B P.
2049
DIELECTRIC DEVIATIOX P m m . a t 0' C . 708 Dielectric deviation
2026
2.68 3.47 5.41
25
i.09
35 40 45 50 55
-235,000 c i l . at 25" C . 263.000 c a l . a t 26' C ., cnni, t a n t pressure 13 7 8 dynes 'en1
M E A XDIELECTRIC COEFFICIENT AID E:'
13 15
30
FREEENERQYO F FORMATION HEATOF FORYATIOS (GAS)
YO
60 65 TO
75
80
I'OL.ARIZA nou
E-" "
p"
2018
16 51
(12, 1 3 ,
85 n0 91 3
8.41 9.42
10 26 10.93 11.52 12.03 l2,5l 12 94 13.39 13.90 14.40 15.22 16.15 Tran.ition point
95 100 110 120 130 140
I50 160 170 180 190 200
210 220 222 5 225 '30
13 63 13.94 1.1 65 I 5 40 16.20 li.06 17.96 18.96 20.11 21.4:1 22.98 24.88 27.8.; 44. 31.1'. 26.50 28.56
(.L
704 590 4,57 360 '716 131 1 9 1 1.90 1 5 8 1 2:? 0 0 8 0.57 0 32
itlc on g l i ~ 3 2 \voi,l. l ' h t recuitoue path over the sliclvc, THE reniarhablc propc'rtie- of c i r l f u r he\afluoritle. SFs, maining step was the tranafer in succession, entering at the for high toltage insulation has a$+akened the general inof thegas to steel cylinder? in top shelf. The sulfur ignited terest in the production of this gas, eqieciall) since the which i t could bi; stored as and produced a crude gas aiailability of industrial fluorine, required for its largeliquid. (The gage pressure of consisting largely of SFe hut wale preparation, seems now to be assured. The combusthese cylinders n-as about 300 also some hydrogen fluoride tion of sulfur in fluorine gas leads to a mixture containing to 350 pounde per square inch (with possibly some oxygen principally SFs, together with some lover fluorides of sula t ordinary rooni tcniperaor oxygen fluoride) from the fur, such as SzFz. SFd, and S2F1,: some air. oxygen, or ture.) This transfer was acgenerator, some air, which oxygen fluoride (OF,), and hydrogen fluoride from the complished in one of two ways. thorough washmay not have been comfluorine generator ( 1 0 ) . P?rolqsis of SZFLO, A steel comprcssor was used ing and alkali scrubbing, dr?ing, and condensation of the pletely excluded from the in one case t o bPing the gas gas is followed b\ transfer to steel cjlinders, in which the system, and variable quantibetween 100 and 200 pounds liquid is stored under about 300-350 pounds gage pressure. ties of other, l o m r fluorides pressure, and, follon.ing the of sulfur, such as S,F?, SF,, Production of oier 1600 pounds of SFs by this process in compressor, a coolcr chilled and S2Fl+. Some of these 1912 is described. A brief summary of the chief physical the gas sufficirntly t o liquc,fy lower f l u o r i d e s a r e q u i t e properties of this gas is included. poisonous and corrosive and, it; the liquid ~v:ib run into the steel cylinder, ivhic~lirras until purified, the gas mixture of :t tF1w pruvidcd \\-it11 an inlet and an outlet, so that air or other must be secured against leakage into the air. permanent gases could he vented from the cylinder as the liquid Purification of the crude gas coming from the sulfur burner entered. Other portions of the gas were collected in 20-litcr glass was carried out in a n all-metal apparatus until i t was safe to usc glass. The gas mixture was first passed through a Monel or nickcl carboys by displacement of a solution of caustic soda and were later transferred to the steel storage cylinders, as fo1lon.s: The tube heated electrically t o about 400" C. in order to pyrolyze the gas in the carboy was displaced with water, dried thoroughly, and S,Fla into SF4 and SFs. The efficiency of this process had been condensed to the solid state by cooling t o -78" C. in a roomy Pypreviously checked by independent experiments. Then the gas rex trap. The trap containing the solid was now connected by it passed through copper tubing to the bottom of a packed copper ground joint to a metal system for transferring the gas t o stccl torver (filled with pieces of copper scrap). A spray of mater storage cylinders. T h i j apparatus consisted of a high vacuum passed down through the to\ver and left a t the bottom through line through which the SFs trap, cooled by liquid nitrogcn, was a n S-shaped exit tube. The gas leaving the top of the serutilwr pumped down t o remove air or other inert gases. The v a c u u n ~ passed next to the bottom of a steel column packed with piecca of carbon (or scrap iron or other suitable material), through line \\-as also connected by metal tubing to a metal trap and from that t o the storage cylinder, which was also provided with a pre,?which a spray of 1 Ar sodium hydroxide Tras continuously circusure gage. The metal trap could be s l u t off from the glass porlated by means of a small motor-driven, circulatory pump; the tions of the apparatus by means of a valve. TVhen the vacuuni latter raised the alkali solution from a supply carboy, situated line and the steel cylinder had been well pumped out, the mctal below the tower, up t o the top of the alkali tower. Hydrolysahle trap was cooled with liquid nitrogen, and the SFa from the &AT fluorides, such as SjFz and SF,, together with hydrogen fluoride, trap n-ascondensed in it. When the metal trap was then a l l o w d Irere largely removed, leaving SF6, moisture, and some air t o to warm up and the valves properly adjusted, the SFs was drivc'n pass on. over into the storage cylinder. After the addition of SF6 in this The gas next passed through a drying tower containing solid way had raised the tank pressure to ahout 300-3jO poundj, 1iqut.caustic soda (or other mitable material) and then through similar faction of the gaS hrgan n-ithin the cylinder, arid thereafter a eonabsorbers filled with porous barium oxide or phosphorus pentox-
March 1947
INDUSTRIAL AND ENGINEERING CHEMISTRY
stltnt prcssurib ri'ading (at room tenipc,rature'l \vas registered tJy the gage on the cylinder. The conversion of fluorine into SFs by the above process yielded not less than 8 7 5 of the amount theoretically possible. PROPERTIES O F SULFUR H E X i F L U O R I D E
Table I summarizcs sonie of the propdrties of SF,as found i n the literature, together with data obtained in this laboratory The total gage pressure (meawrcd by F. S.Preston) shown by liquid SFs Contained in stec.1 cylinders was 189 pounds at 0" c'. and 330 poundi. at 25' C.
( 5 ) Kleiiim and Henkel, Z. anorg. iiiliiem. Chem., 207, 73 (1932). (6) Moissan and Lebeau, C'ompt. rend., 130, 865 (1900); Ann. chim. p h ! p ~[7] , 26, 1 4 i (1902). (7) Pearson and Robinson, J . Chem. S O C . , 1933, 1427. (8)Prideaux, l b i t l . , 89, 323 (19@6), Schumb and Gamble, J . Am. Chem. SOC.,52,4302 (1930). Schurnb, P h y s . Rea., 69, 692 (1946). Traute and Ehrmann. J . prubt. Chem., 142, 70 (1935). (12) Watson and Ramasn-amy, Proc. Roc. SOC.(London), A156, 144
(1'336).
(13) F a t s o n , Rao, and Ramaswarn>., Ibid., A132, 569 (1931); 431, 558 (1934). (1.1) T o s t and Russell, "Systematic Inorganic Chemistry", 11. 299. Xew York, Prentice-Hall, Inc., 1944. PRESEXTED :it the nieering of the l m e r i c a n Physical Sorie-y at Carribridre. ~ I L I $ < ,3Iay , 1916. T h e x o r k described in this paper constituted a small
LITER t T C R E C X E D
Ruerhner, Tan de Graaff, Sperduto. Burrill, IIcIntosh. LTrquhart. r h u s . Reo.. 69, 692 (1946). ('7) Cherlron and Cooper, Gen. Elec. Rea., 40, 438 (1937), (3) Eucken and Schroder, 2 . physik. Chem.. B41, 30i-19 (1938'. (4) Fuoss, J . A m . Chem. SOC.,60, 1633 (1938). (1)
423
nnil
!jor:iun of the 1 l . I . T . project in high voltage rsdiography, u n d e r t h e aiiqpices o f the Office u f Scientific Research a n d De\.elopment (1.1) R c'. C ontract OEI\Isr-294:, a n d :xi15 under the general superrision of I
