Noms
3938
Vol. 70
Grignard Cleavage of Allylic Esters. IV. Steric Hindrance in the Grignard Reagent
and allyl benzoate, the neutral material obtained from the ether solution was saponified by refluxing four hours with alcoholic potassium hydroxide. Extraction of the mixture B Y RICHARD T. ARNOLD, R. WINSTONLIGGETT~ AND SCOTT with water and ether and acidification of the alkaline extracts gave benzoic acid (3.2g., corresponding to 26% reSEARLES, JR.~ covery of unchanged allyl benzoate). Distillation of the A Grignard reagent may react with an allylic ether layer gave allylmesitylene (6.7 g., 360/0), b. p. 215222°(737mm.),ns~1.5149,and2,4,6-trimethylbenzophenester of a tertiary carboxylic acid in two ways: one (2.4g., l l % ) , b . p . 174°(12mm.),n21~1.5738,aswell (1)by addition to form a ketone or an alcohol and as some mesitylene (5.5 g.), as expected from the hydroly(2) by cleavage to form an olefinic hydrocarbon sis of unused Grignard reagent. and a salt of the acid, The two processes gener- SCHOOL OF CHEMISTRY ally take place simultaneously and are the results UNIVERSITY OF MINNESOTA MINNESOTA RECEIVED JULY 12, 1948 of different modes of rearrangement of the coordi- MINNEAPOLIS,
nation complex which is formed as the first step in the Grignard, r e a c t i ~ n . ~ . ~ R-
//O ?O--CE[-CH=CH,
The Osmium Tetroxide Oxidation of Some Long-chain Unsaturated Fatty Acids
f R’MgX + 0
BY ALFREDR. BADER
A s a means of extending the evidence of the configurations of the products of oxidation of oleic, elaidic, linoleic, erucic, and brassidic acid with permanganate and peracids, a study has been made of their oxidation by osmium tetroxide. R’CH-CH=CHz Since this reacts through a cyclic ester intermediate,’ a cis-ethylene must give an erythro diol Unless retarded by steric hindrance the addition and a trans ethylene a threo diol. In each case the process (1) usually predominates over the cleav- oxidation obtained were the ones obage process, as in the reaction of allyl benzoate tained alsoproducts in the alkaline permanganate oxidation with phenylmagnesium bromide. With allylic esters of sterically hindered acids, however, it has of the same acids. Hence the 9,lO-dihydroxybeen shown in previous c~mmunications,~~~ that stearic acid melting a t 132’ and the 13,14-diacid melting a t 132’ are the the phenyl Grignard reagent reacts chiefly by the hydroxybehenic erythro-diols, the two isomers melting a t 95” and cleavage process. lolo, respectively, are the threo-diols and the eight It would be anticipated that the Grignard re- diastereoisomers of 9,10,12,13-tetrahydroxystearic agent may have the degree of steric hindrance reacid have the configurations assigned them by quired to retard the addition reaction and permit McKay and Bader.2 cleavage to predominate with allylic esters of unThis is in accord with the work of Wittcoff and hindered acids. Actually it was found that in the c o - w o r k e r ~ who ~ > ~ pointed out that the criteria reaction of rnesitylmagnesium bromide with allyl (viz., formation of boric acid complexes, rate of oxibenzoate, the cleavage process occurred to the ex- dation with lead tetraacetate and periodate, and tent of a t least 57%,’ while the normal addition rate of formation isopropylidene compounds) reaction to form 2,4,6-trimethylbenzophenonepro- which characterize of cis-glycols in cyclic structures ceeded to the extent of about 15%,’ The Grig- also characterize threorather than erythro-diols in nard cleavage of allylic esters due to steric hind- straight chain compounds. rance in the Grignard reagent has not been reSwern5 has criticized McKay and Bade9 for ported previously. their supposed assumption that alkaline permanExperimental ganate oxidation proceeds by “trans hydroxylaA Grignard soiution prepared from bromomesitylene tion”; this criticism is due to the use2 of the con(25 g.), magnesium (3.2 g.) and ether (50 ml.) was allowed to react for eighteen hours at 25-35’ with a solution of allyl fusing designation “trans” glycol to describe a benzoate (16.2 9.) in ether (50 ml.). The reaction mix- compound of the erythro series. “trans” was ture was then decomposed with slightly acidified ammo- used2 for erythro and “cis” for threo, because the nium chloride solution, and the ether solution was ex- vicinal hydroxyl groups in threo diols possess a tracted thoroughly with sodium carbonate solution (10%). closer spatial relationship than those in erythro Acidification of the alkaline solution gave boenzoic acid diols.4 The present work supports the actual con(5.1 g., 42% of theoretical) ; m. p. 121-121.3 . which ~ . ~ *are ~ the opposite of Because of the similar boiling points of allylmesitylene clusions r e a ~ h e d ~ * those of Hilditch and c o - ~ o r k e r s ~and ~ ~ of * *Dorde (1) Du Pont Post-doctorate Fellow, 1941-1942. (2) Du Pont l’redoctoral Fellow, 1946-1947. Present address: Department of Chemistry, University of Illinois. (3) Johnson, Gilman’s “Organic Chemistry,” 2nd ed., John Wiley and Sons,Inc., New York, N. Y., 1943, p. 1880. (4) Arnold, Bank and Liggett, TEISJOURNAL, 63,3444 (1941). (5) Gilman and Robinson, Bull. SOC. chim., 45, 640 (1929). ( 6 ) Arnold and Liggett, THIS JOURNAL, 64, 2875 (1942). (7) Actual yield adjusted for recovery of unchanged allyl benzoate.
(1) (2) (3) (4) (5) (6) (7) (8)
Criegee. Ann., 528, 75 (1936). McKay and Bader, J . Org. Ckcm., 13, 75 (1948). Wittcoff and Miller, THIS JOURNAL, 69, 3138 (1947) Wittcoff, Moe and Iwen, ibid.. TO, 742 (1948). Swern, ibid., TO, 1239 (1948). Hilditch, J . Chcm. Soc., 1828 (1926). Hilditch and Lea, ibid., 1676 (1928). Atherton and Hilditch, ibid., 204 (1943).
