606
CHARLES R. HAUSER, PASSIE 0. SAPERSTEIN [CONTRIBUTION FROM THE
DEPARTMENT OF
AND JOSEPII
CHEMISTRY,
c. SHIVERS
Vol. 70
DUKE UNIVERSITY]
Types of Reaction of Carboxylic Esters with Grignard Reagents’ BY CHARLES R. HAUSER,PASSIE 0. SAPERSTEIN AND JOSEPH C. SHIVERS A strong base such as the algide ion2 is capable of reacting with carboxylic esters a t any of the four positions indicated by asterisks in the following general formula: a t the carbonyl carbon, the a-hydrogen in the acyl portion, the a-carbon or the @-hydrogenin the alkoxy portion.
u\.3
R-C
47
H’ IY
-c-c
/d
OR
f
I H *-C-C*=O
R-Mg-X
R--hIg--X
R-C
0 -+ Mg-X
J“R
‘A$
0 X-Mg-R
II
RC-0
/”
H
qrkI
I
It is well known that ethyl acetate and higher Since the strongly basic alkyl or aryl anion is aliphatic homologs as well as ordinary aromatic potentially present in Grignard reagents these esters such as ethyl benzoate exhibit mainly the reagents should be capable of reacting a t any of carbonyl carbon reaction with most Grignard rethe four vulnerable positions of esters. Actually, agents. However, t-butyl acetate,6 which has a reactions a t three of these positions have been ob- relatively unreactive carbonyl carbon, and espeserved previously; reaction a t the fourth position, cially ethyl phenylacetate,B which has a relatively the @-hydrogen,has been realized in this investi- reactive a-hydrogen, are self-condensed by isogation. propylmagnesium bromide. With the more comThe reactions of the potential alkyl or aryl an- plex Grignard reagent, t-butylmagnesium chloions of Grignard reagents a t the four vulnerable ride,? ethyl propionate mainly self-condenses, positions of esters are of different types. Addition while with mesitylmagnesium bromideEeven ethyl occurs a t the carbonyl carbon to form a ketone or isovalerate and ethyl isobutyrate self-condense. more often a tertiary alcohol3; the a-hydrogen is In this investigation ethylmagnesium bromide removed to form the magnesium halide derivative was found to effect the self-condensation of ethyl (enolate) of the ester which condenses with un- phenylacetate to the extent of only 8% under the changed ester to form a @-ketoester (acetoacetic conditions that have produced a 9470 yield of the ester condensation) ; substitution takes place a t condensation product in the presence of the more the a-carbon to form a hydrocarbon; and the 8- complex isopropylmagnesium bromide.B Ethyl hydrogen is removed, accompanied by the elinii- diphenylacetate was found to react to some extent nation of the carboxylate group, to form an olefin with isopropylmagensium bromide in refluxing (8-elimination). Because the magnesium atoms ethyl ether but neither the self-condensation of Grignard reagents probably first coordinate product of the ester nor the carbonyl carbon prodwith the carbonyl or alkoxy oxygen atoms of es- uct was obtained. Since treatment of the reacters it is possible that these four types of reaction tion mixture with benzoyl chloride gave some take place within the resulting coordination com- ethyl benzoyldiphenylacetate whereas none was plexes. Indeed, the formation of the coordination obtained from the ester alone with benzoyl chlocomplex should activate both the potential alkyl ride, part of the ester must have been converted or aryl anion of the Grignard reagent and the elec- to its magnesium derivative by the Grignard retron accepting atom of the ester; moreover, these agent, thus reactive centers should be brought in close con(CH&CHMgBr tact within the coordination ~ o m p l e x . The ~ elec- (CeHb)EHCOzCnHs tronic changes that may occur within the coordination complexes leading to the four types of reaction may be indicated as This appears to be the first case in which the mag(1) This work was supported in part by a grant from the Duke nesium derivative of an ester has been condensed University Research Council. with a component other than the original esterg; (2) Hauser, Shivers and Skell, THIS JOURNAL, 67,409 (1945).
-
(3) Obviously, the tertiary alcohol results from reaction with a second molecule of the Grignard but i t is not clear whether the ketone is formed as an intermediate; see Gilman, “Organic Chemistry,” Vol. I, 2nd ed., John Wilep and Sons, Inc., New York, N. Y., 1943, pp. 582-504. ,(4) See Johnson, rbid., Vol. 11, pp. 1866, 1880, Young and 68, 649 (1946), Whitmore and George, Roberts, THIS JOURNAL, Abstracts, Division of Organic Chemistry, Meeting of the American Chemical Society, Atlantic City, September, 1941,
( 5 ) Hudson, Shivers and Hauser, THIS JOURNAL, 6S, 2051 (1943). (6) Conant and Blatt, ibid., 61, 1227 (1929); see also Ivanov and Spassov, Bull. doc. chim., [4] 49,375 (1931). (7) Zook, McAleer and Horwin. THIS JOURNAL, 68,2404 (1946). ( 8 ) Spielman and Schmidt, ibid.. 69, 2009 (1937). (9) An unsuccessful attempt to benzoylate the magnesium derivative of ethyl isobutyrate, prepared by means of mesitylmagncsium bromide. was made previously in this laboratory; Hudson and Hauser, ibid., 68, 3157 (1941), note 8.
Feb., 1948
REACTION OF CARBOXYLIC ESTERSWITH GRIGNARDREAGENTS
judging from previous results with diethylaminomagnesium bromidelo the magnesium derivative of most esters condense very rapidly with the original ester. An attempt to carbonate the: magnesium derivative of ethyl diphenylacetate failed, presumably because of the instability of the carbonation product.” Substitution a t the a-carbon in the ;alkoxy portion of esters has been observed with certain esters in which the carbonyl carbon is hindered and the a-carbon is attached to relatively strong electron releasing groups. Arnold and co-workerd* have realized this type of reaction with certain allyl esters, for example allyl mesitoate, with which phenylmagnesium bromide gave a 7oy0 yield of ally1ben~ene.l~Fieser and Heymann“ found that triphenylmethyl acetate reacts with methylmagnesium iodide to form a,a,a-triphenylethane in good yield. In the present investigation triphenylmethyl benzoate has been found to react similarly. C&CO&(CsHs)o f CHaMgI + (CsHs)&CHa
4-CsHs(:OzMgI
&Elimination of esters with Grignards has apparently not been reported previously.16 This type of reaction has beenrealized in the present investigation with a,P-diphenylethyl me.sitoate in which the carbonyl carbon is hindered and the j3-hydrogen is activated. With phenylmagnesium bromide in refluxing n-butyl ether this ester yielded the elimination products stilbene and mesitoic acid in yields of SO-SO%; this was shown not to be merely a thermal elimination.
+
Mes-C02CH(CBHS)CH&sH, C&MgBr + Cd&.CH=CHC& f Mes-COzMgBr
With methylmagnesium iodide or ethylmagnesium bromide under similar conditions the yield of stilbene was only 30-40% and another product, 1,2,:
