1890
COMMUNICATIONS TO THE EDITOR
Vol. 85
the equatorially oriented bromide of I11 by pyridine displacement reactions a t the anomeric center of sugar with inversion of the anomeric center. If so, this restructures. reaction would have to be very rapid as compared to The occurrence of the above migration without doubt the formation of I1 from I since the results require the explains the stereochemical routes of reaction observed formation of I11 to be rate-controlling in the sequence in the synthesis of 3-carboxamidopyridinium nucleoof reactions leading t o VII. Also, the concentration of s i d e ~ . ~ , Also, ~ ~ , the ~ * discovery likely has an important I11 must be very small throughout the course of the bearing on the formation of both the anomeric forms in reaction since, although it is the anomer with the the syntheses of other types of nucleosides by way of bromine in equatorial orientation, it is the thermody0-acylated glycosyl halides. l5 --I7 namically less stable a n ~ m e r . ~ The J driving force for (13) N. A. Hughes, G. W. Kenner and A. R . Todd, J . Chem. SGC.,3733 this condition has been termed the “anomeric e i T e ~ t . ” ~ . ~(1957). I t seemed more likely that the rapid formation of (14) M. Viscontini and E. Hiirzeler-Jucker, Hclu. Chint. Acta, 89, 1620 (1956). VI1 by way of I11 would involve participation of the (15) M.Hoffer, Chcm. Bcr., 9S, 2777 (1960). 2-acetoxy group in the reaction of 111 to lead to a 1,2(16) J. J. Fox, N. C. Young, 1 . Wemper and M. Hoffer, J . A m . Chcm. acetoxonium ion. In fact, the &bromide (111) has Soc., 83, 4066 (1961). recently been isolateds and its reaction properties cor(17) F. J. McEvoy, B. R . Baker and M . J. Weiss, ;bid,, 82, 206(1960). respond to those previously established for tetra-0DEPARTMENT OF CHEMISTRY acetyl-P-D-glucopyranosyl ~ h l o r i d e . ~Since the reR. u. LEMIEUX UNIVERSITY OF ALBERTA A. R. MORGAN EDMONTON, ALBERTA action of the latter compound in pyridine containing an CANADA alcohol leads t o the formation of a-D-glucopyranose RECEIVED MARCH15, 1963 1,2-(alkyl orthoacetate) t r i a ~ e t a t e , ~these . ’ ~ observations related to the formation of VI1 led to the prediction that the addition of the a-bromide (I) to pyridine conElectrophilic and Nucleophilic Substitution of Allylic taining methanol would result in the formation of the Mercurials’~2 1,2-methyl orthoacetate (VI).*l In fact, the reaction of I in pyridine containing 3 moles of methanol per Sir : mole of I , under conditions (30% of I) which in the Organomercurials have been studied extensively as absence of the methanol provided the a- and p-pyridinsubstrates for electrophilic substitution a t saturated ium glucosides in the ratio 3:2, gave as the product carbon,a and more recently for generation of carbonium a mixture of the methyl orthoacetate (VI) and pions4 by so-called “demercuration.” The correspondpyridinium glucoside (11) in the ratio 3 : 2 . That ing carbon-mercury bond cleavages are depicted by is, the presence of the methanol blocked the route to R:l Hg and R /:Hg, respectively. Both substitutions, the formation of the a-pyridinium compound (VII), electrophilic as well as nucleophilic, are especially precisely the result expected should the latter compound interesting with allylic mercurials, and we report on arise from the 1,2-acetoxonium ion intermediate (IV). this matter in the present communication. The The n.m.r. spectrum of the 1,2-orthoacetate (VI) in present obsexations on the behavior of crotyl- and chloroform showed the presence of the two possible cinnamylmercuric derivatives furnish considerable indiastereoisomers arising from a change in the consight into the competition between the two kinds of figuration of the new asymmetric center in the dioxosubstitution and the mechanistic preferences displayed lane ring.12 The isomer, which produced signals for by each of them. the methoxy and orthoacetyl groups a t 3.30 and 1.72 Crotylmercuric bromide,6 prepared from the butenyl p.p.m. (trimethylsilane), respectively, and believed Grignard reagent and mercuric bromide, was recrystalon the basis of n.m.r. to have structure VIII, was formed lized from pentane-acetone; m.p. 90.8-91.2’ dec. in approximately 6.2 times greater amount than that The trans-crotyl structure was confirmed by the preswith the corresponding signals a t 3.46 and 1.57 p.p.m. ence of only two vinyl protons as indicated by the In order to confirm these notions, the reaction of n.m.r. spectrum in chloroform solvent, the absence of tetra-0-acetyl-P-D-glucopyranosyl chloride with pyriinfrared bands at 905-915 cm.-l and 985-995 cm.-‘ for dine was examined. As expected, N - (tetra-o-acetylterminal methylene, and the presence of an infrared a-D-glucopyranosyl)-pyridinium chloride was formed band a t 965 cm.-’ for a trans-olefinic group. Crotylbut none of the @-isomer. I t is therefore concluded mercuric and cinnamylmercuric bromide,6 on treatthat the driving force for the ready formation of VI1 ment with silver acetate in acetone, gave rise to the from I11 is derived from the anchimeric assistance corresponding crotylmercuric acetate,6 m.p. 76-77’, provided by the participation of the 2-acetoxy group and cinnamylmercuric a ~ e t a t e m.p. , ~ 98.0-98.5’. and that, consequently, the last stage of the reaction The predominant result from treatment of the allylic involved an intramolecular rearrangement of a tranmercurials with acidic media depends on the nature of the anionic ligand in RHgX. Halide as the ligand is sient 1,P-arthoacetyl pyridinium bromide (Vb). The plausibility of such a migration is well supported relatively unfavorable to nucleophilic substitution through the consideration of a molecular model and and, therefore, electrophilic cleavage of the mercurial the fact that the rupture of the C1-to-oxygen bond in is predominant. Thus, at room temperature crotylthe dioxolane ring of Vb must involve participation by mercuric bromide is converted essentially quantitatively the oxygen of the pyranose ring. The latter type of to olefin by excess hydrogen chloride in ethyl acetate participation is clearly involved in the first stage of all (1) Research supported in part by the National Science Foundation. (4) 0.Hassel and B. Ottar, A c f a Chcm. Scand , 1 , 929 (1947). (5) R. U.Lemieux and C. Brice, Can. J . Chcm., 30,295 (1952). (6)R. U. lnmieur and N . J. Chii, Abstracts of Papers, 133rd National Meeting of the American Chemical Society, 1958, p. 31N. (7) J. T. Edward, P . F . Morand and I. Puskas, C a n . J . Chcm., 39, 2069 (1981). (8) F. Weygand, H. Ziemann and H. J. Bestmann. Chcm. Bcr., 91, 2534 (1958). (9) R. U.Lemieuxand C. Brice, C a n . J . Chcm., 33, 109 (1955). (10) R . U. Lemieux and J. D. T. Cipera, ibid., S4, 906 (l95G). (11) B.Helferich and K . Weis, Chcm. B o . , 89, 314 (1956). (12) A. S Perlin, Can. J . Chcm., 41, 399 (1963).
(2) Research was supported in part by a grant from the Petroleum Research Fund administered by the American Chemical Society. Grateful acknowledgment is hereby made to the donors of this fund. (3) E.g., 0. A. Reutov, Rcc. Chcm. Progr., 22, 1 (1961). (4) (a) F . R . Jensen and R . J. Ouellette, J . A m Chcm. SGC.,83, 4477, 4478 (1961); 86, 363 (1963); (b) S. Winstein, E.Vogelfanger, K . C . Pande and H. F . Ebel, ibid., 84, 4994 (1962); (c) S. Winstein, E.Vogelfanger and K . C . Pande, Chcm. I n d . (London), 2061 (1962). (5) Correct elementary analyses were obtained for the indicated compounds. (6) A. N. Nesmeyanov and 0.A. Reutov, Im.A k a d . Nauk S S S R , O l d . K h i m . Nauk, 855 (1953); Chcm. Abstr., 48, 12692g (19541.
COMMUNICATIONS TO THE EDITOR
June 20, 1963
in a few minutes, or by 0.02 M perchloric acid in acetic acid solvent in 0.5 hr. Olefin is also the product from crotylmercuric acetate and hydrogen chloride in ethyl acetate, ligand exchange presumably being sufficiently rapid so that crotylmercuric chloride is the species cleaved to yield olefin. On the other hand, treatment of crotylmercuric acetate in acetic acid with dilute perchloric acid gives rise essentially exclusively to solvolysis products, the butenyl acetates, by way of the crotylmercuric ion. Two noteworthy features of the acid cleavages of the allylic mercurials are the high rates of reaction compared to saturated mercurials and the relatively complete allylic rearrangement which accompanies them. Thus, crotylmercuric bromide reacts with hydrogen chloride more rapidly than does the n-butyl analog' by a factor which is roughly 10.' From this cleavage, as well as from the crotylmercuric acetate cleavage with hydrogen chloride, the butene obtained is very nearly pure 1-butene (Table I). This is true also of the buTABLEI OLEFINSFROM RCH=CHCH*HgX
__--
K
CH3 CHI CH3 CdHa CsH5
X
OAC Br Br OAC Br
Solvent
EtOAc EtOAc AcOH Et20 Et20
% Olefins-----
Reagent KCHzCH=CHz
HCl HC1 HCIOl HCl HC1
>99.2 >99.3 >98.9 98.4 98.4
RCHXCHCHa Lrans CiS
0.3 .2 .6 .9 1.1
