halosilanes with the values for the chlorosilanes reveals that the ethyl group is much more positive with respect to silicon than is hydrogen. This is interpreted as a hyperconjugation effect. Replacement of an ethyl by a phenyl group results
JC:O\TRIHUTIOU FROM I"F
in a n increase in moment for all compounds studied. The alkyl- and phenylsilyl groups are more polarizable than the corresponding hydrocarbon radicals. NOTREDAME,INDIANA RECEIVEDJANUARY 19,1950
1)BPARTMEN C O F
CHEMISTRY. DUKE UNIVERSITY]
The Electric Moments of Five of the Isomeric Hexachlorocyclohexanes 1 3 Ti'nw.mr> ~ L. Lrn-I),' \l.mcr:s E. H O B B.IND ~ PAULIT.GROSS
Recent interest in the y-isomer of hexachlorocyclohexane as an insecticide has given renewed impetus2 to investigations of the structure of cyclohexane derivatives. The hexachlorocyclohexane molecule affords an unusual opportunity in this connection and, a t the same time, structural characterization of its geometric isomers is of considerable interest for the general problem of relating molecular structure and insecticide toxicity. 'The present investigation is concerned with the determination of the electric moments of the a, j3, 7 , 6 and B isomers of hexachlorocyclohexane. The values obtained resolve some conflicting statements in the literature and provide information which, along with other types of data, help in the elucidation of the structure of the isomers studied.
Experimental Solvents.-Thiophene-free benzene (Jones and Laughlin reagent grade) which had been stored over sodium was refluxed and distilled over fresh sodium through a 2-meter Dufton column immediately before use. The middle fraction, distilling over a range of 0.02" or less, was collected with precaution to prevent orption nf atmospheric moisture. 1,4-Dioxane (Carbide and Carbon Chemical Corp., c. P. grade) was refluxcd over solid potassium hydroxide until a brown precipitate no longer separated on addition of fresh potassium hydroxide. After distillation from the potassium hydroxide, this aldehyde-free dioxane was refluxed with hydrochloric acid as described by Fieser.a This was followed by distillation from potassium hydroxide and two distillations from sodium. The second distillation from sodium was a fractionation in a 2-meter Dufton column. The middle fraction with a distilling range of 0.06" or less was collected with precautions t o prevent absorption of atmospheric moisture. Solutes.-The samples of a, @, y and b-hexachlorocyclohexane were separated from the crude mixture of isomers using a procedure developed from the method outlined by Slade. e from the comHooker Electr lized four tim ct was recrystallized once from, chloroform and 4 times from benzene, 111. p. 156.5-1.57.0 . -l_l_
I 1) Part of the thesis of Edward L Lind submitted in partial fulfillment of the requirements of the M A. degree in Chemistry at Duke University, June, 1948 (2) R. E Slade, Chcm and Ind., 314 (1945). (3) L.F.Fieser, "Experiments in Organic Chemistry," D C. Heath and Company, N e w York, N.Y., 1941, p. 369. (4) The ratio of P and p solubilitie$ in acetone is unfavorable for the formation of the eutectic of P and p isomers shown to exist bp van der Linden, Bcr , 411, 231 j1912).
The sample of @-isomerwas separated from the mixture of isomers prepared by the chlorination of benzene as described by van der Linden.' After one recrystallization from methyl acetate, two from 1,4-dioxane and two from benzene a product with a m. p. of 30&309° was obtained. The product showed no detectable pleochroism when examined with a polarizing microscope. The ,%isomer is the only reported isomer of hexachlorocyclohexane with a cubic crystal structure.6 Only a very faint trace of chloride ion was observed in an alcoholic potassium hydroxide solution of the @-isomerafter eight hours a t 25". The @-isomeris the only isomer of hexachlorocyclohexane reported to be stable under these conditions.6 A sample of y-isomer' was separated from commercial grade hexachlorocyclohexane wpplied by the Hooker Electrochemical Company. This was combined with a sample of y-isomer obtained from E. I. du Pont de Nemours Co. which. had been recrystallized from chloroform to the same melting point. The combined sample was recrystallized twice from chloroform, once from carbon tetrachloride and twice from benzene, m. p. 111 .5-112.0°. &Hexachlorocyclohexane was separated from the commercial grade mixture of isomers supplied by the Westvaco Chlorine Products Corp ,, Commercial Solvents Corporation and by E. 1. du Pont de Nemours and Co. This was recrystallized twice from carbon tetrachloride and 3 times from benzene, m. p. 136-137". The sample used for the measurements in benzene was recovered for the measurements in l,.l-dioxane. Disagreement with previous measurements warranted characterization of the sample used in this investigation. Mixed melting points over a range of composition with samples of &isomer obtained from the Dow Chemical Co. and the U. S. Department of Agriculture showed no melting point depression. Each of the three 8-isomer samples was found to give the same meltihg point depression for the same composition of a mixture of a- and &isomers. A different but constant melting point depression mas observed for the same composition of a mixture of a- and 8-isomers. The refractive indices of the three Gsamples a t 145' mere found t o be approximately 1.513. The solubility of the &isomer used in this investigation was determined in ethanol and foiind to be in reasouable agreement with the high solubility of the &isomer reported by Slade.L The sample of &-isomerobtained from the U.S. Department of Agriculture was used t o check the measurements in this investigation since the sharp melting paint indicated high purity. It was used with no treatment other than drying i n vacuo over phosphorus peutoxide, m. p. 137.5-138'. A small amount of e-hexachlorocyclohexane was borrowed from the Dow Chemical Co. and recrystallized twice from benzene, m. p. 216.5-21S0. The €-isomer (8) L W. Daasch, Anal. Chcm , 19, 779 (1947). (6) S. 5. Cristol. Tats JOURNAL, 69, 338 (1947). (7) Obtained from a methanol solution as described by Slade.: h considerable purification was effected by dissolving the crude 7-isomer in chloroform, filtering from the small amount of 8-isomer and adding an equal volume of isopropyl alcohol. The purified isom om or was found t o precipitste from tbie solutiw OD rtandiag.
Oct., 1950
ELECTRIC MOMENTS OF FIVEOF THE ISOMERIC HEXACHLOROCYCLOHBXANES 4475 TABLEI Solvent
Number of solutions
CeH6 1,4-C1HsOz CeH6 1,4-C4Hs02
C6& 1,4-C4Hs02 CeHe 1,4-C4Hs02 CsH6
1,4-C,H80? 0
8 4 4 4 4 4 J
4
4 4
AE/faE (av.)
Ad/ftb (av.)
6.72 * 0 . 0 2 8 . 0 6 * 0.15 0.62 * 0.02 0.81 * 0.Oi 11.39 * 0.04 12.93 * 0.08 7.33 * 0 . 0 i 7 . 9 4 * 0.06 0.78 * 0.04 0.86 * 0 02
1.57 * 0.01 1 . 3 5 * 0.01 1.61 * 0.01 1.33 * 0 . 0 1 1.60 * 0.01 1 . 3 8 * 0.01 1 . 5 9 * 0.01 1.36 * 0.01 1 . 6 4 * 0.01 1.37 * 0.01
The deviations shown here are the average deviations of the mean value, oiz.,
P? cc. 151 167 59.2 61.2 220 237 158 165 60.6 60.9 av. dev. of a
FD
2.16 (2.14) 2.34 (2.30) 0.34 0 0.46 0 2.84 (2.83) 3.00 (2.96) 2.24 12.22) 2.32 (2.28) 0.43 (0.26) (-0) 0.46 single observation where n
dn
is the number of independent measurements. * The deviations shown here are estimates of the error in the measurement. for p in CeHs) for C ~ Has B solvent and PD= The values in parentheses in this column are calculated using PD= 59.2 (Pz 61.2 (Pz for /3 in 1,4-C4H802)for 1,4-c&s02 as solvent. This measurement was checked by a single independent measurement of the dielectric constant of a 0.006 mole fraction solution in CsHe of a sample of 6 isomer obtained from the U.S. Department of Agriculture, The AE/f2 for this solution was found to be 6.95. The moment calculated assuming Ad/fi = 1.59 as shown in the above table is 2.13 (2.11)D. used for the measurements in benzene was recovered for the measurements in 1,4-dioxane. Apparatus.-The heterodyne heat method of measuring capacity change was used for the measurement of dielectric constant. The oscillator system has been previously described.* The liquid dielectric measuring cell used in this investigation was of the same design as that used by de Bruyne, Davis and Gross.8 The capacitance change of the cell due to the solvent was measured with a large variable capacitance. The increment in capacitance due to the solute mas measured with a small calibrated Wulf (E. Leybold, h’achfolger) precision condenser with a variable capxitance of 30 rnmfd. graduated in 5000 scale divisions. All dielectric constants were measuzed relative to the dielectric constant of benzenelo a t 30.0 The density determinations were made with a conventional pycnorneter.l1
.
Results The molar polarization a t infinite dilution was calculated from the Hedestrand12 equation since both density and dielectric constant increments were found to be linear with respect to mole fraction. The form of the equation used was
where PF is the riiolar polarization of solute a t infinite dilution, E the dielectric constant, Af the molecular weight, (1 the density aiid f the iiiole fraction. The subscripts 1 and 2 refer to solvent and solute, respectively, while Ad and AE are the differences in density and dielectric constant of the solution and the solvent. The values of Ad/fi and AE/fi used in the above expression were the averages for each series of solutions. All graphs of AE against fi gave a 0,O intercept within experimental error. (8)Weith, Hobbs and Gross, THISJOURNAL, 70, 805 (1948). (9) de Bruyne, Davis and Gross, ibid., 66, 3936 (1933). (10) Hartshorn and Oliver, Proc. Roy. SOC.(London), 8113, 68-1 (1029). , (1!-):3:4). (11) Connell, Vosburgh and Butler, J . Chcm. S O L .993 (12) Hedestrand. Z.physik. Chcm., 84, 428 (1929)
The electric moment in debye units was calculated from the expression ( 2j
where the molar refraction, RD = 56.9, was calculated from the refraction equivalents given by Fuchs and Wolf. l 3 The data and results are summarized in Table I. I n general the mole fractions of the solutions measured were within the limit 1 X IOw3and 1.5 x All measurements were made a t 30.00 * 0.02”.
Discussion Table I shows the moments of the b- and c-isomers to be quite small compared to the moments of the CY-,y- and &isomers. The small moment observed here for the P-isomer is certainly indistinguishable from zero within experimental error when the uncertainties attributable to solvent effect, and the neglect of atomic polarization, are considered. The @isomer has been shown to be centrosymmetric, l4 thus one would expect the electric moment to be zero. For purposes of comparison all moments have been calculated using the polarization values of the &isomer in benzine and 1,il-dioxane as the distortion polarization of the other isomers in the respective solvents. This is seen to have little effect on the riioinents of the a-, y - and 6-isoinersI but the moment of the e-isomer becomes zero within experimental error. The electric moment of the @-isomerwas first found to be zero in benzene solution by Hassel and NaeshagenI5 and later verifiedl6 by these authors after Williams and Fogelberg” reported the finite value 0.7 D in the same solvent. Williams and (13) Puchs and Wolf, “Hand- und Jahrbuch der chemischen Physik.”Bd. VI, Abschnitt 1 B, Akad. Verlags., Leipzig, 1935,p. 258. (14) (a) Hendricks and Bilicke, THIS JOURNAL, 48, 3007 (1926); (h) Dickiason and Bilicke, ibid., 60, 764 (1928). (15) Hassel and Naeshagen, Tids. Kcmi Bergvcsen, 10, 126 (1930). (16) Hassel and Naeshagen, 2. physik. Chcm., B16, 373 (1932). (17) Williams and Fogelberg, THISJ O U R ~ ~ L68, , 2096 (1931).
4476
EDWARD L. LIND,MARCUSE.HOBBS AND PAULSI. GRUSS
1-01.7 2
Fogelberg also found the ariotiialous \-slue 2.U D ane ring since these f o r m involve the least strain. The thirteen different geometric chair forms and for the 13 isomer in l,i-diosc~tlr.The slight increase in the polarization ot the &isoilier observed twenty-four different geometric boat forms of the in the present iiivestigation on going from beIlLene hexachlorocyclohexaries are tabulated in Table II. t o 1,-&-dioxanea s wlveiit is probably a t tributnble The six numbers lor each form refer t o the position of substitution for chlorine atoms as given in tu wlvent effect. f'c, ify the value of the 1110Figs. l a and lb. . i n inspection of the Fishernient in 1,+dioxme observed by Williain\ 'mcl Fogelberg ;I polarization of I 1 0 cc. would ha\*?h ~ d Hirshfelder models for the chair and boat forms in to be found instead d the I) I cc. actually obscr\-eri. Table IT enabled us to list the various structures in what appears to he the order of increasing strain. 1lie electric niotnetit of the +isoinvr 1131 no1 beeti prekiously reportect. l-.ii$ll I I CIi.ur formp , ctllcd. liomer The electric i~ioitieutoi the CY-isoiiier111 b e n ~ e i ~ e i eportrd in T:il)le I is in ,~greeiilentwith the value 2 , *j,4,5, 7 , 12 If s', e > 2.1,; D found by \Villiariis m d Fogelberg" sild 2 A , 3, 5 , 6 , 8 , 12 ;j 2 n, 6 2 2.20 D by Hassel and !i;aeshagen. I' 'l'he moment .; 2, 3,4, 5,0,12 3.2 fY, ti? of this isomer in I , LdioxCLtie has not been preI 1,%,3,4,5,6 0 B \-iously reported. ;i 3 , 3 , 5, 6, 10, 1% 5.8 Only the y-isomer of thesc isoniers thus far re'i 1,2,4,7,9,11 3.2 CY, &' ported in the literature possesses riiarked insecti7 I, 4,5 , 7 , 10, 11 4.6 r? cidal properties.? it is seen that the electric moY 1, ti, 8 , 9 , 10, 11 0 i, e? ment of this isomer is the largest of the ;i isomers (1 1 , 1 , i,9, 10, 11 3 2 c y , 6, studied in this investigatioti ; however, it is C c m IO 1,2,3,7,8,9 8. 0 siderably lower than the value 3.6(5) D reported 1L I, 2,7,8,9, 11 3.6 1~ \IelantlerL4for
