SOMEHALOBENZENE PHOSPHONIC ACIDS
Oct., 1948
3465
PbC16, by addition of ammonium chloride. This salt, when treated with concentrated sulfuric acid a t O o , separated out lead tetrachloride as a dense oily, yellow liquid, which was purified by washing with concentrated sulfuric acid. The liquid was kept at 0 ” while the Raman spectrum was being taken. Germanium tetrachloride was made by boiling germanium dioxide with concentrated hydrochloric acid in a stream of hydrogen chloride gas.8 The escaping hydrogen chloride gas carried the germanium tetrachloride through a water condenser into an acetone-Dry Ice-bath where germanium tetrachloride was frozen out of the gaseous mixture. The product was purified by distillation and treatment with sodium carbonate.
The Raman shifts obtained for germanium tetrachloride were found to be essentially the same as those found by the two previous investigators (Table I). With long overexposure, however, no trace of a line could be found in the region between 500 ern.-' and 600 cm.-l, so i t was concluded that the value of 452 cm.-l for v3 was the correct one rather than the possible alternate value listed by Haun and Harkins. The deviation of v3 may be explained by consideration of the masses of the atoms involved. VI (A1) and v2 (E) represent, respectively, a “breathing” and a bending freDiscussion quency. Both are independent of the mass of the Considerable difficulty was encountered in lo- central atom, while v3(F2) and vq(F2) represent vicating the Raman frequencies v2 and v4 in the case brations in which the central atom as well as the of lead tetrachloride since they lay close to the ex- chlorine atoms are displaced. It would, thereciting line and were obscured by the background fore, be expected that the deviations in the latter two frequencies might occur, as shown by the v3 from it. Plate tracings showed a peak a t 90 cm.-l, but on only one plate were two separate lines re- frequency of germanium tetrachloride. I n an attempt to eliminate the effect of the mass of the solved, these appearing at 85 cm.-’ and 93 cm.-l. The latter values are inclosed in brackets in Table central atom, force constants calculated by the I. The frequencies obtained for lead tetrachloride method of RosenthallO rather than frequencies are in satisfactory agreement with those which were plotted against the central atom-chlorine would have been predicted by extrapolating the distance as shown in Fig. 2. It will be noted that in this plot the germanium tetrachloride shows no curves in Fig. I,. marked deviations. TABLE I A more complete discussion of the correlation of the Raman spectra with metal-halogen disRAMANFREQUENCIES OF GERMANIUM TETRACHLORIDE tance is given by Hildebrand.2 Investigator VI VI
Haun and Harkins? Schneiderg The authors
Y4
V2
451(1) 171(6) or 508 396 134 453 172 397(10) P 132(2) D 452(1) D 171(3) D 397(10)
132(6)
Summary
For germanium tetrachloride, the correct value of v3 is 452 ern.-', and the deviation from the value predicted by plotting the metal-halide distance RAMAX FREQUENCIES O F LEADTETRACHLORIDE against Raman shift may be explained by considThe authors 227(10) P 90(3) D 348(3) D 90(3) D eration of the mass of the germanium atom. The [or 851 [or 931 Raman frequencies of lead tetrachloride are essentially what would have been predicted. ( 8 ) Foster, Drennan and Williston, THIS JOURNAL, 64, 3042 (1942). (9) K. W. F. Kohlrausch, “Der Smekal-Raman-Effect, Erganzungsband,” 1981-1937, J. Springer, Berlin, 1938.
[CONTRIBUTION FROM
THE
(10) J. E. Rosenthal, Phys. Rev., 46, 538 (1934); 46, 730 (1931).
BERKELEY 4, CALIFORNIA
RECEIVED JUXE 5 , 1948
CENTRAL RESEARCH DEPARTMENT, M o z r s ~ ~ CHEMICAL ro Cu.1
Synthesis of Aromatic Phosphonic Acids and their Derivatives. 11. Some Halobenzene Derivatives BY GENNADY M. KOSOLAPOFF Since the scientific literature contains the syntheses of only the p-chloro- and the p-bromo-benzenephosphonic acids, i t was felt advisable to prepare the possible isomers of halo-benzenephosphonic acids containing the principal halogens. This paper describes the syntheses of the m- and the p-isomers of the chloro-, bromo- and iodobenzenephosphonic acids, and includes the preparation of the related (p-chlorobenzene)-(9-aminobenzene)-phosphinic acid. The 9-chloro derivative was prepared in the ( 1 ) Michaelis. A n x . , 293. 193 (1896)
manner already described2 and was used for the preparation of phosphanilic acid according to B a ~ e r which ,~ was readily converted into the piodo-acid by the diazo-reaction. The preparation of the $-bromo-acid was conducted by our modified Michaelis reaction,2 in the course of which i t was found that considerable debromination took place, apparently through the action of the aluminum chloride catalyst. The m-isomers were prepared from benzene(2) Kosolapoff, THIS JOURNAL, 69. 2020 (1947) (3) R a n e r 1.4900),estimat- that the dimeric methyl esters had the normal structure, and that the P - 7 to a-/3 shift of the ing the percentages of A and B to be 13 and l27', respectively, of the total dimer. By saponifica- double bond occurred during the alkaline saponition of A they obtained acids from which, by frac- fication of the ester. The present work indicates that the esters comtional crystallization, they isolated four different crystalline disorhic acids. One of these acids, m. p. posing mixture A have the normal structure, and 216O,isolated in 2-4.5y0 yield (0.3 to 0.6% of total that mixture B is composed largely of dimers with dimer), was definitely shown to have structure I, a double bond in the a-p position to a carbometh1-rnethyl-2-propenyl-4-cyclohexene - 3,4- dicarbox- oxy group. The conjugated acids isolated by the ylic acid. A second acid, m. p. 191') isolated in a English workers from mixture A of the esters must similarly small yield, was shown to have structure therefore have been formed by a P-7 to a-/3 shift 11, l-methyl-3-propenyl-4-cyclohexene-2,4-dicar-of the cyclic double bond during the alkaline s;iboxylic acid. A third acid, m. p. ZOO', apparently ponification. These conclusions are based largely was a cis-trans isomer of 11, and a fourth acid, on considerations of the ultraviolet absorption m. p. 164-169', was not characterized. They characteristics of the dimeric methyl esters, and oi could isolate no crystalline dimeric acid from mix- the acids obtained from them by alkaline saponiture B, except traces of the same acids as from A fication. Iodine values of the esters and acids were in acand B was not investigated further. As was pointed out by these workers, the cord with the indicated structures, but molecular structures shown for these acids differ from those refractions of the conjugated forms did not show expected as a result of the usual Diels-Alder re- as much exaltation as was expected. The high yield of crude dimeric methyl esters CHs reported by Farmer was confirmed. However, a single efficient fractional distillation afforded 43% -COOH c&=CH-CHa COOH of dimeric ester A, and 32,5Y0of dimeric ester B, corresponding in nD to the fractions separated by Farmer in 13% and 12% yield, respectively. The 4 boiling point of fraction B was 16-18' above that COOH LOOH mi. p. 216' of fraction A. Refractionation of A afforded 47% I1 I of a lower boiling fraction, A-1,with a lower nD CH: CH: I and 27% of a higher boiling fraction, A-2, with a higher nD. Fraction A-2, when crystallized from -COOH petroleum ether a t -40°, yielded 25% of a crysCH=CH-CHs talline dimeric ester, A-2-p, m. p. 64.5-65.5'. No crystalline methyl ester could be isolated from A-1 COOH COOH or B by similar treatment. Alkaline saponificaIV I11 tion of A-2-p afforded the dimer of sorbic acid rn. I
(1) Farmer and
(1940).
Morrison-Jones. J . Chem. Soc..
1339-1346
(2) Bradley, Symposium on Vegetable Drying Oils, Univ Minnesota. March 28, 1947
of
