[CONTFUBUTION FROM THE DEPARTMENT OF CHEMISTRY AND CHEMICAL ENGINEERING AT TFIE UNIVERSITY OF PENNSYLVANIA, AND THE BAKERLABORATORY OF CHEMISTRY AT CORNELL UNIVERSITY]
T H E MICHAEL C0NDENSATION.l VI. THE INSTABILITY OF SOME ADDITION PRODUCTS2 PETER L. DEBENNEVILLE, DOROTHY DYOTT CLAGETT? RALPH CONNOR
AND
Received April 4, lS4l
In discussions (1) of the influence of structure upon reactivity in the Michael condensation it has been shown that, except in rare cases, the yields of addition products decrease as either the addenda or the acceptors become more highly substituted. In some cases there are unusual differences in reactivity between a substituted active methylene compound and its next higher homolog. For example, benzalacetophenone gave an 80% yield of addition product with ethyl methylmalonate (2) but none with ethyl ethylmalonate (la). If these differences are a result of spatial CsHb CH=CHC 0 C6 Hs
+
RCH( COO CSH,),
NaO C2HL CsH5 CHCH, COCaH6 I R C ( C 00 C2 Hs)2
I, R CHa--, 80% 11, R = C2H5-, 0% interference with addition by the larger substituent, it might be expected that 11, once prepared, would be stable. This paper describes experiments which might be expected to give 114but which gave other results. These data show that it is sterically possible for ethyl ethylmalonate to add to benzalacetophenone, but that I1 is unstable in the presence of sodium ethoxide and undergoes retrogression. The literature contains abundant examples of the alkylation of addition products from the Michael condensation and it was hoped that this might be a useful method for the synthesis of I1 and related compounds. The For the fifth paper in this series, see Connor and McClellan (1). Presented before the Organic Division of the American Chemical Society a t the Baltimore meeting, April 3,1939. a A portion of this communication is constructed from the thesis submitted by Miss Dorothy Dyott in partial fulfilment of the requirements for the degree of Master of Science a t Cornell University in June, 1935. 'The discussion deals with I1 aa an example; data which are described in the experimental part and which might lead t o other compounds not obtained from the Michael condensation, might be interpreted similarly. 690 1
THE MICHAEL CONDENSATION.
VI
691
general results can be shown by considering the alkylation of ethyl acarbethoxy-b-phenyl-y-benzoylbutyrate (111); other examples are described in the experimental part. Methylation of I11 gave ethyl a-carbet€ioxy-cu-methyl-~-phenyl-y-benzoylbutyrate (I), as well &s some cleavage (25%) to ethyl methylmalonate and benzalacetophenone. Alkylation of I11 with other alkyl halides gave only cleavage products; ethylation, for example, gave benealacetophenone and ethyl ethylmalonate.
-
C ~ H S C H C H ~ C O C ~BH ~CaHsCHCHzCOC6H6
I
CH( C 00 C2Hs)Z
CzK6 (CO 0 CzHs )z
I11 CeH5 CH=CHCOCsHs
+
CHz (Co0CzHs)z
I1
D
C B HCH=CHCOCaHb ~
+
.__$
C2Hs CH( COOCzH6)z
These results indicate that if ethylation of I11 occurs (step B), the product (11) is unstable in the presence of sodium ethoxide and undergoes retrogression (step C ) . The results could also be explained by assuming that the initial reaction is the retrogression of I11 (step A) followed by alkylation (step D) to give ethyl ethylmalonate; if this is the case, the failure of ethyl ethylmalonate to condense with benzalacetophenone might be due to steric hindrance or to the instability of 11. Explanation based on steric hindrance seems improbable in view of the results obtained (la) by the reaction of the sodium derivative of ethyl ethylmalonate with benzalacetophenone in dry benzene. Under these conditions, which are favorable for bringing about further reaction of the initial addition product, the products are ethyl a-ethylcinnamate and ethyl benzoylacetate. These products must result from a Michael condensation, regardless of whether the subsequent reaction is the formation of a fourmembered ring (IV) as proposed by Holden and Lapworth (3) or a sixmembered ring (V) as proposed by Michael and Ross (2). It is therefore CsHg CH-CHCOCeHg
'I A
C2H5 c-
0
CO 0 Ca Hb
IV
CeHsCOCH-CHCBH5
1 CHCOCs& 1
CaH6 CH C2Hs
b-bo I
00CzH6 V
692
DE BESNEVILLE, CLAGETT, AND CONKOR
sterically possible for ethyl ethylmalonate to add to benzalacetophenone, (step C') ; even ethyl benzylmalonate (la) undergoes some addition. It then follows that some addition products, such as 11, undergo retrogression so readily that they may not be isolated. The difference between the results of methylation and ethylation of 111 probably depends upon the relative rates of the alkylation reactions (step B) and the retrogression reactions (step C). Methylation occurs rapidly enough to remove the alkoside before I undergoes complete retrogression, although I is completely cleaved (2) with an equivalent of catalyst. Ethylation, on the other hand, does not occur rapidly enough to permit I1 to be isolated. Since it has been shown (-1) that retrogression is favored a t higher temperatures, an investigation of the reaction at -78" was carried out in the hopes that the temperature coefficients (of steps C and C') might be different enough to permit the isolation of 11. However, no addition products could be isolated from the reaction a t -78" of benzalacetophenone with ethyl ethylmalonate, ethyl a-phenylbutyrate, or ethyl pheny-lmalonate. The retrogression reaction, therefore, occurs readily over a wide temperature range. It was also found that I11 undergoes retrogression more readily than might be expected from the high yields obtained in its preparation. Under conditions similar to those used for the alkylation. except that ethyl phenylacetate was added instead of the alkyl halide, over .!io?& of I11 was converted to ethyl a ,P-diphenyl-ybenzoylbutyrate (VI).
I
CeHgCHCHzCOCeH6 I
C6HgcHCOOC~Hg
VI 1
The total cleavdge of '111 must have been considerably more than 50%, since step E is irdeversibk (as shown by the failure of VI to react with ethyl malonate under these conditions) and since the reaction of benzalacetophenone with e t h d malonate is more rapid than with ethyl phenylacetate. (The last conclusion is reached because the same results were obtained when ethyl phenylacetate reacted with I11 as when it reacted with a mixture of benzalacetophenone and ethyl malonate. In both cases,
THE MICHAEL CONDENSATION.
VI
693
therefore, it appears that the retrogression of I11 governs the rate of the reactions.) These data show that in some cases the addition products that might be expected from the Michael condensation are eo readily cleaved by sodium ethoxide that they cannot be isolated and that steric hindrance does not prevent addition. It cannot be decided whether the results obtained by the alkylation of addition products such as I11 are due to retrogression of I11 followed by alkylation (path AD), or to alkylation followed by retrogression (path BC), or to both reactions. Ac,knowZedgment. The authors are grateful to the Faculty Research Cominittee of the University of Pennsylvania for a grant to aid this investigation and to Dr. V. H. Wallingford of the Mallinckrodt Chemical Works for liberal supplies of ethyl a-phenylbutyrate and ethyl phenylmalonate. EXPERIMENTAL PART
1. Alkylation Studies
Thie alkylations were carried out by two general methods. A. To a solution of sodium ethoxide prepared by dissolving 1.3 g. (0.057 gram atom) of sodium in 100 ml. of absolute ethanol was added 0.06 mole of the compound to be alkylated and 0.06 mole of alkyl halide. The mixture was refluxed on the steam-bath for forty-two hours, cooled, diluted with water, and acidified with acetic acid. The product was extracted with ether, dried over magnesium sulfate, and distilled under reduced pressure. B. The procedure was like that described above except that the reaction period was three weeks a t room temperature. Reaction of ethyl a-carbethoxy-8-phenyl-ybenzoylbut yrate (ZIZ) with methyl iodide. Method A gave 2.4 g. (26%) of ethyl methylmalonate, b.p. 82" (12 mm.) The residue was difficult to purify (an indication of the presence of benzalacetophenone) but finally gave from ether-ligroin crystallization ethyl a-methyl-a-carbethoxy-8phenyl-ybenzoylbutyrate, (I) m.p. 90-93" (2). Method B gave practically identical resulte. Reuction of ethyl a-carbethoxy-p-phenyl-ybenzoylbutyrate(ZZZ) with ethyl iodide. Method A gave 7.6 g. (75%) of ethyl ethylmalonate, b.p. 115-123' (30 mm.) which waa identified by conversion to 5-ethylbarbituric acid. The residue was 11.0 g. Method B gave 7.6 g. (65%) of ethyl ethylmalonate, b.p. 115-123" (30 mm.). The absence of any rearrangement-retrogression products (1) was shown by the negative reaction with ferric chloride. The distillation residue weighed 12.5 g. (theory for benaalacetophenone, 10.8 g.) but the absence of I1 or of unreacted I11 was shown by hydrolysis with dilute sulfuric acid. The hydrolysis mixture was made alkaline, the benzalacetophenone removed by ether extraction and the aqueous layer acidified. S o insoluble acid was present. Reaction of ethyl a-carbethoxy-p-phenyl-y-benzoylbutyrate (ZZI) with benzyl chloride. Method A gave 3.1 g. (44%) of benzyl ethyl ether, b.p. 60" (9 mm.) and 1.1 g. (8%) of ethyl benzylmalonate, b.p. 75" (9 mm.). The residue was refluxed with '20% sodium hydroxide solution, extracted with ether, and acidified. There was obtained
694
DE BENNEVILLE, CLAGETT, AND CONNOR
5.2 g. (250Jo) of benzylmalonic acid, which after recrystallization melted at 116-117'
[the reported (5) melting point is 117'1. Reaction of ethyl a ,y-dibenzoyl-p-phenylbutyrate ( 1 ) with ethyl iodide. Method A gave 2.4 g. (30%) of ethyl benzoate, b.p. 89-99' (9 mm.), and a residue of 14.8 g. from which benzalacetophenone crystallized after chilling in an ice-salt-bath. The benealacetophenone, after repeated recrystallizations, melted a t 54". The 9.6 g. remaining was shown to contain ethyl a-ethyl-a-benzoylacetate by refluxing for three hours with 20% sodium hydroxide solution, extraction with ether, and acidification of the aqueous layer. Benzoic acid precipitated and butyric acid was indentified by distillation of the filtrate and conversion to the p-bromophenacyl ester m.p. 59-60'. The reaction products presumably are formed as follows:
CaHaCOOC2H5
CeHbCH=CHCOCsH6
+
+ CH, CH2 CH2COOCzH5
,/alcoholysis
CsH6COCHCOOCaH6
1
/"
C2 H6
\
\conc. NaOH
L
CaHaCOONa
+ CHsCH2CHsCOOXa
8. Attempted Condewations at Variow, Temperatures
A solution of 20.8 g. (0.1 mole) of benaalacetophenone and 0.1 mole of the active methylene compound in 100 ml. of dry ether was added t o a solution of sodium ethoxide prepared by dissolving 2.3 g. (0.1 gram atom) of sodium in the minimum amount of absolute alcohol. The mixture was allowed t o stand for seventy-two-hours in an insulated box containing solid carbon dioxide. At the end of this period an excess of glacial acetic acid was added, the mixture then allowed to come to room temperature, poured into ice-water, separated, and the ethereal layer dried over anhydrous magnesium sulfate. The ether was removed by distillation and unchanged active methylene compound distilled under reduced pressure. With ethyl ethylmalonate, ethyl phenylmalonate, and ethyl a-phenylbutyrate the recovery of starting material was good and there were found no indications of condensation except traces of very high-melting solids which may have been trimolecular compounds (2) ; the latter were not obtained in quantities large enough for examination. Ethyl a-phenylbutyrate was also used with one-sixth equivalent (0.4 9.) of sodium but the results were similar to those described above. Attempts to obtain condensation at higher temperatures and the experiments in which the sodio derivatives were used in dry benzene have been reported earlier (la). 5. Retrogression Studies Ethyl a-carbethoxy-8-phenyl-ybenzoylbutrate ( 2 ) (III) with ethyl phenylacetate. A sodium ethoxide solution was prepared by dissolving 1.04 g. (0.045 gram atom) of sodium in 100 ml. of absolute alcohol. To this was added 16.6 g. (0.045 mole) of 111 and 7.38 g. (0.045 mole) of ethyl phenylacetate and the mixture allowed t o
THE MICHAEL CONDENSATION.
VI
695
stand three weeks at room temperature. A t first a homogeneous solution was formed, but within two days a solid appeared. The reaction mixture was poured into cold water containing an excess of acetic acid, filtered, and dried. The weight of crude material was 15.4 g. This was crystallized from alcohol and gave 7.72 g. of ethyl a,@-diphenyl-r-benzoylbutyrate,m.p. 148-150". Concentration of the filtrate gave 0.58g., m.p. 140-146", and 0.30 g., m.p. 135-145". Calculated as pure VI, the total of 8.6 g. represents a 51% conversion. A check run gave 8.0 g. (48%). Concentration of the mother liquors gave oils which did not crystallize. Ethyl a,@-diphenyl-7-benzoylbutyrate(6) ( V I )with ethyl malonate. A sodium ethoxide solution waa prepared by dissolving 0.69 g. (0.03 gram atom) of sodium in 100 ml. of absolute alcohol. To this was added 11.1 g. (0.03mole) of VI and 4.8 g. (0.03 mole) of ethyl malonate and the reaction mixture treated as described in the preceding paragraph. Recrystallization of the product gave 9.40 g., m.p. 152-152.7" and 0.95 The total (10.35 g.) represents a recovery of VI of 93.570. The g., n1.p. 151-152'. additional solid formed by dilution of the mother liquor with water appeared to be entirely VI but was not pure. Benzalacetophenone m'th ethyl malonate and ethyl phenylacetate. A sodium ethoxide solution was prepared by dissolving 1.04 g. (0.045 gram atom) of sodium in 100 ml. of absolute alcohol. To this was added 9.36 g. (0.045 mole) of benzalacetophenone, 7.38 g. (0.045mole) of ethyl phenylacetate, and 7.2 g. (0.045mole) of ethyl malonate. The reaction mixture was treated as previously described. Recrystallization of the product gave 6.47g., m.p. 141-144". Concentration of the filtrate gave 0.52 g. of a sticky material with no definite melting point, Calculated as pure VI, the total of 6.99 g. represents a yield of 41.7%. Further concentration of the mother liquor gave an oil which would not crystallize. SUMMARY
It has been shown that in some cases the addition products which might be expected from the Michael condensation are so readily cleaved that they cannot be isolated in the presence of sodium ethoxide; failure to isolate the addition products cannot be attributed solely to steric hindrance. PHILADEPHILA, PENNA: REFERENCES
(1) (a) CONNOR AND ANDREWS, J. A m . Chem. SOC.,66,2713 (1934). (b) ANDREWSAND CONNOR, J. A m . Chem. Sac., 67, 895 (1935). (c) CONNOR AND MCCLELLAN, J. Org. Chem., 3,570 (1939). (2) MICHAEL AND Ross, J. A m . Chem. Sac., 66,1632 (1933). AND LAPWORTH, J. Chem. Sac., 1931,2368. (3) H:OLDEN AND POWELL, J. Chem. Sac., 119, 1976 (1921). (4) INQOLD (5) CLAISENAND CRISMER,Ann., 218,139 (1883). Ber., 43, 4497 (1909). (6) BORSCHE,