WILLIAM S. JOHNSON,JACOB
3726
SZMUSZXOV~CZ AND MAXMILLER
P-dimethylaminoacrylate to ethyl trimesate, which is the end-product of the self condensation of ethyl forrnylacetate. This appears to be characteristic of the structure R2NCH=CH--.' Experimental Raw Materials.-Alkyl j3-alkoxyacrylates were supplied through the courtesy of Dr.W. J. Croxall of this Laboratory, and were prepared by the methods described by h i m s The amines were commercial products used without purification. General Procedure.-The following preparations are representative. Ethyl P-Piperi+oacrylate.-To a mixture of ethyl 6ethoxyacrylate (57.6 g., 0.4 m.) and 20 g. of anhydrous potassium carbonate, heated on the steam-bath with stirring, piperidine (34 g., 0.4 m.) was added dropwise over a fifteen-minute period. The reaction mixture was heated on the steam-bath under reflux for four hours, filtered to remove potassium carbonate and distilled to yield 57.5 g. (7973 of ethyl j3-piperidinoacrylate. Ethyl B-Dimethylaminoacrylate.-To a mixture of ethyl j3-ethoxyacrylate (144 g., 1 m.) and 50 g. of anhydrous potassium carbonate, contained in a flask equipped with gas inlet tube, stirrer and acetone-Dry Ice condenser. was added dimethylamine gas (45 g., 1 m.) at 25'. The mixture was stirred for several hours a t room temperature, filtered t o remove potassium carbonate and distilled. (6) Mannich and Davidsen, Bcr., 69B,2106 (1936).
[CONTRIBUTION FROM THE
Vol. 72
After removal of alcohol and unreacted ethyl p-ethoxyacrylate (31 g.) there was recovered 92 g. (53%) of ethyl j3-dimethylaminoacrylate.
Hydrolysis of Ethyl fi-Dimethylaminoacrylate.-Thirty
grams of dilute hydrochloric acid (1:1) was added to ethyl Pdimethylaminoacrylate (21.5 g., 0.15 m.). The material formed a homogeneous solution which deposited an oily lower layer. After two weeks, crystals were present in this layer, and these were filtered off. The m. p. (132135") corresponded to that of triethyl trimesate and this wks c o n h e d by a mixed m. p. with an authentic sample. Only a few grams of crystals were obtained; the remainder of the oil may have represented intermediate stages of condensation.
Acknowledgments.-To Dr. E. L. Stanley who directed the analytical work, and to Mr. E. J. Smialkowski for assistance in the experimental work. Summary A number of p-dialkylaminoacrylic esters have been prepared by the interchange reaction of palkoxyacrylic esters with secondary amines. The ease with which these reactions occur indicates a surprising degree of labilization by the single ester PUP. RECEIVED FEBRUARY 6, 1950
PHILADELPHIA, PA.
LABORATORY OF ORGANIC CHEMISTRY OF
THE
U~IVERSITY OF WISCONSIN]
The Condensation of 1-Methyl-2-acetylcyclohexene with 1-Decalone, and with 1,s-Decalindione BY
WILLIAhl
s. JOHNSON,JACOB SZMUSZKOVICZ'
The extension of the Robinson-Rapson synthesis3-typified by the base-catalyzed reaction between l-acetylcyclohexene, I (R = H), and cyclohexanone to produce the ketododecahydrophenanthrene I1 (R = H)-to the production of fused ring systems containing the angular methyl group has been investigated by Huber,* who studied the condensation between 1-methyl-2-acetylcyclohexene, I (R = CH3), and cyclohexanone. The prod238 mp (in uct of this reaction was an oil, .,,A
I
A N D M A X MILLER^
ether) (E ca. 2060)' of expected composition which gave a semicarbazone in unspecified yield and on dehydrogenation with selenium afforded phenanthrene also in unspecified yield. From these observations Huber concluded that his product had the expected structure I1 (R = CH,). On the basis of Huber's report, Bagchi and Banerjee6 proposed the structure I11 for the product of the condensation of I (R = CH3) with 8methyl-4-hydrindanone, and Dimroths similarly assigned the structure IV to the condensation product of I (R = CH3) with 1-decalone. In the present paper we are reporting the results of some studies of condensations with 1-methyl-2-acetylcyclohexene, which force us to the conclusion that all of these structures I1 (R = CHI), 111 and IV are open to question. CHs
I1
(1) Sterling-Winthrop Research Institute, and Wisconsin Alumni Research Foundation Postdoctoral Fellow 1948-1949; U. S. National Institute of Health Postdoctoral Fellow 1949-1950. On leave of absence from the Weizmann Institute of Science, Rehovoth, Israel. (2) Wisconsin Alumni Research Foundation Research Assistant 1949-1950. (3) Rapson and Robinson, J . Chcm. Soc., 1285 (1935); Crowfoot, Rapson and Robinson, i b i d . , 757 (1936); Peak and Robinson, ibid., 759(1936). (4) Huber, Bcr., 71, 725 (1938).
0
I11
0
IV
(6) Bagchi and Banerjer, J. Ind. Chon. SOL.,28, 397 (1946). (6) Dimroth, Angcw. Chcnt., 69, 215 (1947).
Aug., 1950
l-METHYL-2-ACETYLCYCLOHEXENE WITH
When 1-methyl-2-acetylcyclohexene and 1decalone were allowed to condense in the presence of a mixture of potassium and aluminum t-butoxide, a high-boiling oil was produced in 37% yi.eld. Although this material gave no ketone derivatives, the presence of the a,&unsaturated carbonyl system was indicated by the ultraviolet absorption spectrum Amax. 248.5 mp ( E 6982). When this product was treated with lithium aluminum hydride, then dehydrated over potassium acid sulfate, and finally dehydrogenated over palladiumon-charcoal, there was no evidence of the presence of chrysene which would be the expected product from IV. Instead there was produced a crystalline hydrocarbon, Cl9H18 (isolated in 28% yield as the di-picrate) which was shown to be 1-(anaphthyl)-2-(o-tolyl)-ethane (VI) by comparison with an authentic specimen prepared by the addition of di- (o-tolyl)-cadmium to 1-naphthylacetyl chloride followed by WOE-Kishner reduction of the resulting ketone V, as depicted in the accompanying flow sheet.
CHtCOCl
D DECAL ONE
3727
methyl group attached to the ,&carbon of the CY,,% unsaturated ketone system apparently inhibits the Michael addition so that an intermolecular aldol condensation occurs thus precluding ~yclization.~ The possibility that VI1 was formed from IV by a reversal of the Michael reaction seemed unlikely when it was found that 1-methyl-2-acetylcyclohexane condensed with 1-decalone just like the unsaturated ketone I (R = CHs) giving what was undoubtedly VIII, as shown by the dehydrogenation sequence to give VI. The a,&unsaturated carbonyl system was apparent from the absorption, Amax. 247 mp ( E 6800), which was similar to that of VILE The conclusion that Dimroth’s product was indeed VI1 is valid provided the likely assumption is made that the different conditions employed by us (potassium t-butoxide instead of amyloxide) did not alter the course of the reaction. When aluminum t-butoxide, which is a good catalyst for the aldol condensation,g was also added to the reaction mixture, the condensation proceeded more cleanly, fewer tarry by-products being formed, The course of the reaction, however, did not appear to be altered by this modification as shown below in the decalindione study. The isolation of phenanthrene by Huber on dehydrogenation of his cyclohexanone ccmdensation product suggests the presence of a t least a small amount of cyclized material, but unfortunately no yield or experimental details were reported for this degradati~n.~a
n V
VI
These observations suggest that the original condensation proceeded by an aldol condensation between the carbonyl group of 1-decalone and the reactive hydrogens of the acetyl group of 1 (R = CHI) to produce the structure VII. C o n m a tion of this structure was afforded by ozonization experiments which yielded &acetylvaleric acid, CH3COCH2CH2CH2CH2COOH(by cleavage a t points a and b) and 1-decalone (by cleavage a t point e).
U\J II
0
VI I
VI11
It seems probable that the mechanism of the Robinson-Rapson reaction involves a Michael addition of the cyclic ketone to the orlj3-unsaturated ketone followed by an intramolecular aldol condensation. In the present case the additional
8 IX
X
The mole for mole condensation of I (R=CHs) with 1,5-decalindione (IX) was also studied and found to give (with or without added aluminum tbutoxide) an oily product in better than 60% yield,10about half of which was obtained crystalline. This material, m. p. 66-67’, was shown to (7) I t seems unlikely that a system VI1 would undergo an intramolecular Michael type of addition to give IV, because the formation of the necessary resonating anion involving the 2-position of the decalin nucleus, would concomitantly inhibit the polarizing effect of the carbonyl group on the double bond in the cyclohexane nucleus. (8) The principal absorption of this type of dienone may be close to that of the eneone. Cf.for example the spectrum of 1,4-cholestadiene-3-one,. ,A 236 mp ( E 15,800),with that of 4-cholestene-3-one, Am. 234 mp ( E 16,400); Dannenberg, Abhandl. preuss. Akad. Wiss., 81, 26 (1939). Cf.however, phorone and mesityl oxide. (9) Wayne and Adkins, THIS JOURNAL, 62, 3401 (1940). (9a) ADDEDIN PRo0a.-Dr. Richard B. Turner has kindly informed us that attempts to repeat the work of Huber have given a product which proved to consist largely of uncyclized material having a structure analogous to VII. (10) The yield dropped to about 17% when the condearation W M esrried aut by m prosetlure iisoilar t o HUbW’b (ref, 4),
37%
WILLIAM S.JOHNSON,
JACOB
SZMUSZKOVICZ AND MAXMILLRR
have the structure XI1 as follows. The high extinction coefficient ( E 19,950) of the ultraviolet absorption maximum a t 249.5 mp is approximately that expected from the sum of the E (cu. 13,820) for tbe octalone system as in compound XI (see below), and E (
