1726
EDWARD G. LOCKE,WALLACE R. BRODEAND ALBERTL. HENNE [CONTRIBUTION FROM
THE
Vol. 56
CHEMISTRY DEPARTMENT AT THEOHIO STATE UNIVERSITY ]
Fluorochloroethanes and Fluorochloroethylenes BY EDWARD G. LOCKE.WALLACE R. BRODE AND ALBERT L. HENNE During the last four years, a large number of fractional distillation, an operation which is fluorinated derivatives of ethane and of ethylene extremely easy to perform because each replacewere prepared, but their complete presentation' ment of a chlorine atom by a fluorine atom in was delayed. Some of these derivatives were re- GC16 lowers the boiling point of the resulting ported recently by Booth and his students,2 and, chlorofluoride by about 45'. in general, both sets of data agree well. The reDetermination of the Physical Properties.sults presented here are more detailed and com- The methods used have been described beplete; they embrace all the possible derivatives of fore.5 Analysis.-The liquid compounds were burned ethane and of ethylene which have all the hydroin a Parr bomb, and after combustion the bomb gen atoms replaced by chlorine or fluorine. content was dissolved in perchloric acid. The Synthesis of the Saturated Derivatives.-The method of synthesis is based on the classical chlorine was titrated by means of silver nitrate, procedure of Swarts3which consists in heating to- as usual, while the fluorine was titrated with gether a mixture of hexachloroethane and anti- cerium nitrate.6 Gaseous compounds were anamony trifluoride, in the presence of about 10% of lyzed by combustion over heated silica.' The antimony pentachloride. However, in order to molecular weight of liquids was computed from accelerate the operation, to be able to replace a the observed freezing point depression in benzene, larger number of chlorine atoms in the hexachloro- and the molecular weight of the gases was derived ethane molecule and to obtain better yields, three from the gravimetric analysis of a known volume, modifications of the process were adopted, namely: a procedure which gives correct results because (1) the mixture of antimony trifluoride and anti- gaseous chlorofluoro derivatives of ethane and mony pentachloride was replaced by antimony ethylene are almost perfect gases. fluorochloride, SbFaClz; (2) steel equipment was Results used; (3) the fluorination was performed under The physical properties and analyses appear in pressure and the desired fluorinated derivatives Table I. were removed from the reaction field as soon as Determination of the Configurations.-Only formed. These improvements have been prethe compounds numbered from 1 to 6 (inclusive) viously described in detaiL4 They make it posare found in the fluorination of C2C16. After obsible to obtain from hexachloroethane almost taining their formula by analysis and molecular quantitative yields of CC&CC12F, CC12FCClF2, weight determination, it remained to prove their CCIF2CCI2F,CClF2CClFz or CF3CCIF2. spatial configuration, which was done as follows. The compounds CCLCClF2 and CC13CF3 which (1) Monofluoropentachloroethane.-Only one formula were needed for identification purposes were syn- is possible, CClSCClZF. This compound, treated with thesized from CHClzCHFz and from CH2CICF3, zinc in alcohol, gives a quantitative yield of CCl,=CCIF by chlorination in sunlight. Their properties are (described below) and zinc chloride. Only a trace of zinc listed here, but their synthesis is the subject of an- fluoride was detected. (2) Difluorotetrachloroethane.-Two formulas are posother paper. sible, CClSCClFz (1) and CClzFCCllF ( 2 ) . The f i s t Purification.-The crude products were formula is improbable, because such a compound is rewashed with water, then with a weak caustic ported to melt at 52". Moreover, a treatment with solution, and dried by means of calcium chloride, zinc and alcohol gives a quantitative yield of dichlorodior sulfuric acid. They were finally purified by fluoroethylene, which is separable into two isomers. Since (1) Presented in part at the Buffalo meeting of the American Chemical Society, September, 1931. (2) Booth and his collaborators, I n d . Eng. Chem., 24, 328-31 (1932), THIS JOURNAL, 66, 223 (1933). (3) Swarts, Bull. A c a d . Roy. Bclg., [ 3 ] 24, 474 (1892); and [3] as, 874 (1896). (4) Midpley and Henne, Ind. Eng. Chem., 22, 542 (1930); also U.S. Patent 1,930,129, October 10, 1933.
formula (1) would not permit the formation of geometrical isomerism in its ethylene derivative, it follows that formula (2) is correct and that the ethylenic compound is CClF=CClF. ( 5 ) Shepard, Henne and Midgley, THISJOURNAL, 68,1948 (1931). (6) Ratchelder and Meloche, ibid., 63, 2131 (1931). (7) Hubbard and Henne, ibrd., 66, 1078 (1934).
FLUOROCHLOROETHANESAND FLUOROCHLOROETHYLENES
Aug., 1934
1727
TABLE I PROPERTIES AND
1 2
Formula CClaCClzF CClrFCClzF
3
CClrFCClFz
No.
4
CCIF~CCIFI
CClnFCFs CCIFrCF8 CFaCFa 8 CClaCClF; 9 CClr.CFx
B. p. 760 mm. corr., OC. 137.9 92.8
M. p . , OC. 101.3 24.65
-36.4
Glass
6 6 7
to
CONSTANTS OF ETHANEDERIVATIVES Analyses, Chlorine % Mol. wt. Fluorine Calcd. Found Calcd. Found Calcd. Found 220.3 217.9 8.62 8 . 8 2 80.48 79.96 1.41297 1.41823 2 0 3 . 8 2 0 2 . 0 1 8 . 6 4 1 8 . 7 3 6 9 . 5 6 6 9 . 2 3 1.40831 1.41368 187.4 186.2 30.42 30.64 56.77 56.29 1.35572 1.35981 1.33124 1.36488 1,3092 1 7 0 . 8 168 44.49 44.39 41.45 41.52
Density and refractive indices at t o
25.0 35.0
47.7
0.0
3.8
25.0 36.0 0.0 25.0 35.0
- 2 (about) 38 -78.3 91.5 45.8 20.0
d4
nu
1.64470 1.62316 1,6300 1.56354 1.63982 1.5312 1.455 1.422
1.41100 1.40356 1.35113 1.34933 1,3073
nD
ng
-
40.6 13.2
2 0 3 . 8 198 1 8 7 . 4 184
1.5702
69.6 56.8
70.1 57.1
Compounds 5 and 6 have been obtained only as by-products in the industrial preparation of compound 4. Compound 7 has not been obtained by the present method. It can be made by electrolysis of CFaCOzH [Swarts, Bull. soc. chim. Belg., 42, 102 (1933) I or by passing CFI through a carbon arc [Ruff and Bretscheider, 2.anorg. allgem. Chem., 210, 173 (1933)l. Compounds 8 and 9 were obtained independently, by chlorination of CHClzCHFzand CHzC1CF3,respectively, as indicated in paragraph (2) of page 1726. The melting point of compound 8 is reported by Swarts, M6moires couronnds Acad. Roy. Belg., 61, 68 (1901), as 52', but this could not be duplicated. (3) Trich1orotrifluoroethane.-As this compound is prepared by the fluorination of CClZFCClzF, its formula is probably CC12FCC1Fz. The configuration is proved by treatment with zinc and alcohol, which yields CClF=CFZ and zinc chloride, but no zinc fluoride. This rules out the only other possible formula, CFaCCla. Moreover, this latter compound was prepared by chlorination of CFsCHZCl in sunlight (to be reported later); it melts a t 13" and boils a t 45'. (4) Dich1orotetrafluoroethane.-The formula, CClF2CClF2, of this compound was proved by the formation of zinc chloride and CFz=CF2 in the treatment with zinc and alcohol. Had the formula been CClzFCF3, the only other possibility, zinc would have removed one atom of
chlorine and one atom of fluorine, to yield CClF=CR, zinc fluoride and zinc chloride. Moreover, in the industrial preparation, a limited amount of CClzFCFs was obtained; it has a boiling point of about -2'.
Ethylene Derivatives.-These compounds were all obtained from the corresponding sathrated products by a treatment with zinc in absolute alcohol. As highly fluorinated materials have a low boiling point, it is impractical to carry the operation a t atmospheric pressure, and the elimination of 2 chlorine atoms from C2C13F3 and from CzClzF4 was performed under pressure (10 atm.
TABLE I1 DERIVATIVES OF ETHYLENE AND No.
Formula
Mdz.,
Bd p., C.
11 CClz=CCIF 72.1 12 CClZBrCClFBr 122.5 13 CCl-CFz 15 14 CC12BrCFzBr 46 117.1 15 CCIF=CFz - 23 16 CClFBrCF2Br 92.9 17 CFz=CFz -78.4 18 CFZBrCF2Br - 112 46.4 19 CClF=CClF -130.5 21.1 cis isomer 20
21
CClF=CClF trans isomer
-110.3
Density and refractive ind. d4 n t
1.5541 n, 1.4360 20
THEIR
DIBROMIDES
Analyses % Mol. wt. Chlorine Bromin; Fluorine Calcd. Found Calcd. Found Calcd. Found Calcd. Found
149.4 148 7 1 . 2 7 1 . 2
293
285 2 4 . 2 2 4 . 0 54.6 54.4
2.2318 n, 1.4272 20 276.5 268
2.149 25 260 1.4950 nar1.3752 0 n, 1.3777 ng 1.3825 2 2 . 0 1.4936 nu 1.3764 0 n, 1.3798 ng 1.3850
12.87 12.5 57.85 5 7 . 9
255
61.5 61.3 53.36 53.6
28.6
28.7
53.36 5 3 . 4
28.6
28.6
CClFBrCClFBr 3 2 . 5 139.7 293 287 2 4 . 2 2 4 . 2 54.6 55.5 12.98 12.94 Compound 13 was also obtained from CHC12CClFz and alcoholic KOH. Compounds 19 apd 20 were obtained as a mixture and were separated by fractional distillation.
FRANCIS F. HEYROTH AND JOHN R. LOOFBOWROW
1728
and 60 atm., respectively) in a steel container. Under those circumstances, a considerable amount of ethylene appears, and the reaction stops short of completion. Table I1 gives the results. Differences in the Boiling Points.-A tabulation of the boiling points makes apparent the regularity of the boiling point depression caused by replacing chlorine by fluorine. €3. p.. OC.
cc13cc13 CClsCC12F FFCClzF
186 138 92 or 91 47 or 45 3 or - 2 -38 - 78 120 71 22 or 15 23 -78.4 139.6 92.8 46.4
t
CClaCClFz J CClzFCClFz or CC18CF* CClFz(!ClFZ or CCl*F(!Fz CClFzCF, CFsCFa cc12=cc12 CClz=CClF CClF==CCIF or
CCI,=CFz CCIF==CFe CF2=CFz CClFBrCClFBr CClFBrCFzBr CFzBrCFzBr
-
[CONTRIBUTION FROM
THE
Molecular Refractions.-The Lorentz-Lorenz formula has been used to calculate the molecular refractions, and from the experimental results the atomic refraction of fluorine has been computed by subtracting the sum of the atomic refractions of carbon, chlorine, double bonds and bromine (“International Critical Tables” values were used).
Depression
10
CCIzFCCIzF
48 46 or 47 45 or 46 46 or 47 41 40
VOl. 56
CCIzFCClFz CClFnCClFi CC~SCFJ CClFBrCFlBr CClz;.CClF CClF=CClF Cisand trans
25 35 25 35 0
20 20 20 0 0
Molecular refraction D B
30.768 30.896 31.242 30.841 30.959 31.316 26.065 26.170 26.440 26.142 26.273 26.516 21.35 21.473 . .. . . .. 26.405 . . , , 28.533 . . , .. 25.15 . , . . 20.36 20.48 20.71 20.44 20.60 20.85
. . .
. .. ..
F. Atomic refraction D B
1.105 1.096 1.097 1.142 1.128 1.134 1.147 1.144 1.145 1.173 1.179 1.170 1.17 1.176 .. . . , 1.223 . . 1.09 . . . . . 0.68 . . . 0.99 0.99 0.96 1.03 1.05 1.03
.
.. .
..
The values found for the atomic refractions of fluorine agree well with the values given by Swartss for a large variety of substances; they vary with the type of compound considered.
45 55
Summary All the possible derivatives of ethane and of ethylene (together with their dibromides) which have all their hydrogen atoms replaced by chlorine or fluorine are described.
46.8 46.8
THEMIDGLEY FOUNDATION RECEIVED MARCH5, 1934 COLUMBUS, OHIO
49 49 or 58
(8) Swarts, J . Chim. Phys., 20,30 (1923).
BASICSCIENCE RESEARCH LABORATORY, UNIVERSITY
OF
CINCINNATI]
Correlation of Ultraviolet Absorption and Chemical Constitution in Various Pyrimidines and Purines BY FRANCIS F. HEYROTH AND JOHN R. LOOFBOUROW
It was shown previously that the ultraviolet irradiation’ of dilute solutions of uracil and other pyrimidines and purines induces marked changes in their ultraviolet absorption spectra indicative of constitutional changes, and that, with sufficiently long continued irradiation, all selective absorption could be made to disappear. To afford a basis for the interpretation of these changes, information as to the absorption spectra of a series of related compounds was desirable. This paper presents the ultraviolet absorption spectra of eighteen such compounds, and permits comparisons to be made between their spectra and constitutions, so that conclusions may be drawn as (1) Heyroth and Loofbourow, THISJOURNAL, 68, 3441 (1931).
to the manner in which various structural modifications are responsible for, or affect, the ultraviolet absorption of the compounds of this series.
Experimental Method.-The technique used in determining the absorption spectra has been described previously. As there noted,’,* the measurements have not been corrected for certain deviations due to failure of the reciprocity law, but with the method employed these deviations are negligible, amounting to less than the experimental error a t most wave lengths. In all cases the solvent was redistilled water. The concentrations were chosen by trial so as to result in the greatest accuracy in the absorption determinations. Values of PH were determined by the quinhydrone elec(2) Loofbourow, Bull. Bas. Sci. Rn.,in publication.
