NOTES
June 5, 1954
3041
Hydroxymethylene Ketones. 111. The Preparation of Methoxymethylene Ketones'
from catalytic amounts of sodium methoxide has shown this to be the case, elimination of methanol occurring readily with the formation of methoxyBY E. EARLROYALS AND KENTC. BRANNOCK~ methylene ketones in good yields. A similar reac10, 1954 RECEIVEDFEBRUARY tion has been described recently by Franke and The preparation of ethoxymethyleneacetone by co-workers4in which methanol was eliminated from the alkylation of sodium hydroxymethyleneacetone acetoacetaldehyde dimethylacetal by passing it in in ethanol with ethyl bromide has been describedaa the vapor phase over a basic catalyst. No effort has been made to determine optimum I n the present work, alkylation of crude sodium hydroxymethyleneacetone with ethyl iodide in conditions for the elimination reaction ; however, it ethanol gave a poor yield of an inseparable mixture was found advantageous to add methanol to the of ethoxymethyleneacetone and acetoacetaldehyde mixture of P-ketoacetal and sodium methoxide. diethylacetal. Though the difference in the pres- This effect is probably due to the increased solubilent results and those previously reported probaby ity of the sodium methoxide. While in most cases lies in the markedly different experimental proce- the elimination was accomplished at reasonably dures used, the refractive index of ethoxymethylene- low temperatures under vacuum, acetoacetaldeacetone reported by K a u ~ h a and l ~ ~the conclusions hyde dimethylacetal and the dimethylacetal of hydroxymethylene diisobutyl ketone invariably drawn therefrom are apparently in error. Alkylation of sodium hydroxymethyleneacetone gave incomplete conversions under these conditions; with ethyl iodide in acetone gave a fair yield of the use of higher temperatures a t atmospheric crude ethoxymethyleneacetone which could not be pressure was necessary with these two compounds. obtained in a satisfactory state of purity by frac- The products obtained are shown in Table I. During the preparation of hydroxymethylenepintional distillation. A similar alkylation with methyl iodide gave even poorer yields, no analytically pure acolone by the Claisen condensation of methyl formate products, and apparently carbon as well as oxygen with pinacolone in the presence of sodium methoxide alkylation, so that direct alkylation of sodium hy- the evolution of carbon monoxide was observed. droxymethylene acetone as a route to alkoxy- Though the fact that formate esters are decomposed to carbon monoxide and alcohol by sodium methyleneacetone was not pursued further. TABLE I METHOXYMETHYLENE KETONESPREPARED Yield,
Procedure
72.6 76.4 61.5 74.1 84.8
B A A
yo
Compound
CH3-CO-CH=CHOCH? CHaCHzCOC( CHs)=CHOCH, ( CH:,)~CH-CO-CH=CH-OCHI ( CH~)~C-CO-CH=CH-OCHS (C H ~ ) ~ C H - C H ~ C O - C = C H - O C H Z
I
A B
'C. 52-53 70 61-62 61 78-79
B.p.
FROM
@KETOACETALS n%
Mm.
4-5 4-45 3.5 3 1
1.4653 1.4752 1.4538 1.4536 1.4633
Carbon, %" Calcd. Found
I-Iydrogen, %" Calcd. Found
...
...
...
...
65.59 65.59 67.57 71.69
65.48 65.28 67.43 71.31
9.44 9.44 9.92 10.94
9.39 9.23 9.83
10.95
CH(CH3)z Microanalyses by Dr. G. Weiler and Dr. F. B. Strauss, Oxford, England. * Franke, et al. (ref. 4), report ~ Z O D 1.4699, p. 172" a t atm. pressure. Alkaline hydrolysis (ref. 5) gave equivalent weights of 100.8 and 101.1; theor., 100.11. 0
TABLE I1 Yield, &Ketoacetal
%
(CH3)3C-CO-CHrCH( OCH3)z ( C H ~ ) Z C I ~ - C H ~ C O - C H - C H (0CHa)z
54.6 56.8
I
B.& OC.
67-70 71-71.5
Mm.
4 1
11%
1.4202 1.4308
Car.bon, To" Calcd. Found
Hydrogen, 77" Calcd. Found
62.03 66.63
10.41 11.18
62.20 66.91
10.36 11.13
CH(CHI)Z Microanalyses by Dr. G. Weiler and Dr. F. B. Strauss, Oxford, England.
A consideration of the probable mechanism of formation of 0-ketoacetals from chlorovinyl ketones and alcohols in the presence of bases, and of the cleavage of ,&ketoacetals by aqueous alkali led us to believe that the reaction
alkoxide has long been known, the evolution of carbon monoxide in the reactions of methyl formate with such ketones as acetone, methyl ethyl ketone and methyl isobutyl ketone is much less apparent. It was found that in the reaction of 0.5 mole of sodium methoxide, 0.5 mole of pinacolone and 0.75 0 Base /I H mole of methyl formate that roughly 1400 cc. of R-C-CHz-C(OCHa)i carbon monoxide was produced. It may therefore 0 be advantageous t o use an excess of the formate esI1 in condensations such as this. R-C-CH=CHOCHI + CHsOH terThe properties of the dimethylacetals derived is revei-sibk. Distiill-atim of s m a i j3-itetuzcetais from hydroxymethylenepinacolone and hydroxy(1) For t h e previous paper in this series see E. E. Royals and K. C. methylenediisobutyl ketone are shown in Table Brannock. THIS JOURNAL, 76, 1180 (1954). 11. (2) National Science Foundation Fellow, 1953-1954. (3) (a) R. Kaushal, J . I n d i a n Chcm. Soc., 20, 53 (1943); (b) L. Panizzi and M. Sbrillo-Siena, Gas% chim. ita!., '73, 335 (1943).
(4) W. Franke, R. K r a f t , D. Tietjen and H. Weber, Chcm. Bcr.. 86. 793 (1953).
3042
NOTES Experimental
VOl. 76
The Cyclization of a-Substituted Benzylsuccinic
Anhydrides The P-ketodimethylacetals were prepared as described previously.6 BY R. S. URBAN^ AND E. M. BEAVERS Alkylation of Sodium Hydroxymethyleneacetone in Ethanol.-The sodium salt was prepared in the usual manner JANUARY 20, 1954 RECEIVED from 23 g. (1 g. atom) of sodium, 58 g. (1mole) of acetone and 78 g. (1.05 moles) of ethyl formate in 1 1. of absolute The formation of cyclic ketones by the intramoether. iMost of the ether was removed under water-pump lecular acylation of benzylsuccinic acids (and anhyvacuum, and 181 g. (1.16 moles) of ethyl iodide in 750 ml. drides) can, hypothetically, lead to the generation of absolute ethanol was added. The mixture was stirred a t 50-55' for 10 hours after which time it was neutral to of both five- and six-membered rings, depending phenolphthalein. The mixture was concentrated by dis- upon which carboxyl group condenses a t the arotillation until prccipitated sodium iodide caused bumping, matic nucleus. It has been reported that benzylsuccooled, diluted with twice its volume of ether and the precipitated salts removed by filtration. This procedure was cinic acid,2a a-phenylbenzylsuccinic acid2b and repeated twice more, and finally distillation from a claiseri fluorenylsuccinic anhydride3 form six-membered flask gave 28.5 g. of product, b.p. 68-72' a t 5-6 mm., n Z 5 ~rings exclusively. 1.4236. The material was a t first insoluble in water but I n the course of another investigation in this dissolved on shaking, gave a positive fuchsin test, and on Laboratory, a-methylbenzylsuccinic anhydride4 standing in water gave crystals of triacetylbenzene. Disand a,a-dimethylbenzylsuccinic anhydride4 (11) tillation through a n efficient column gave no good fraction. A middle cut, b.p. 59.5-60" a t 5 mm., ~ 2 1.4194, 5 ~ gave on were prepared. Since these compounds have never alkaline hydrolysis an equivalent weight of 153.4 (theoretibeen subjected to cyclization, and since there cal for acetoacetaldehyde diethylacetal is 160 and for ethseemed to be a possibility that steric effects of the oxymethyleneacetone is 114). Reported for the diethylgem-dimethyl group in the one compound might acetal; b.p. 76-77' a t 10 mm., +OD 1.42266&; b.p. 82-87' a t 12 mm., n W 5 1.4189.6b ~ favor the formation of a planar five-membered ring, Alkylation of Sodium Hydroxymethyleneacetone in Ace- we investigated the cyclization of I and 11. tone.-This sodium salt prepared from 11.5 g. of sodium, 55.5 g. of ethyl formate and 34.8 g. of acetone in 250 ml. of ether, was freed of ether under water-pump vacuum. I t was then stirred a t 50' with 93.6 g. of ethyl iodide and 400 ml. of acetone for 24 hours a t which time the mixture was neutral t o phenolphthalein. When worked up as described above there mas obtained 19.8 g. of crude product, b.p. 55-70' a t ca. 5 mm., n 2 5 ~1.4602. Distillation through an efficient column gave 13.3 g. of impure ethoxymethylene acetone, b.p. 68-71' at 7-7.5 nim., n Z 51.4618 ~ (Kaushal, ref. 3a, reports ?z3% 1.6075). CHI CHI CH3 Anal. Calcd. for C&@!: C, 63.13; €1, 8.83. Fouiitl: I \/ C, 64.28; H, 9.46. This impure product gave triacctylbenzene on standing in water, a purple color with ferric chioride, and a posirive fuchsin test. Using similar amounts of reactants and similar conditions but substituting methyl iodide for ethyl iodide there mas ob0 0 tained 12.2 g. of crude product, b.p. 60-95" a t 25 mm., I11 IV nZ5D 1.4526. This product darkened rapidly on standing. On attempted fractional distillation crystals of what is preWhen a-methylbenzylsuccinic anhydride (I) was sumably the carbon-alkylation product (the solid isomer of hydrox~'methylenemethy1ethyl ketone) formed in the contreated with concentrated sulfuric acid, two cyclic denser while most of the material resinified. isomers were i ~ o l a t e d . ~These were shown to be Elimination of Methanol from 8-Ketodimethylacetals. Procedure A.-The P-ketodimethylacetal was dissolved in cis and trans isomers of 4-methyl-3-carboxy-1-tetraan equal volume of methanol and 1-2% by weight of solid lone (111) by Clemmensen reduction,6 followed by sodium methoxide was added t o the solution. Methanol dehydrogenation to the same 1-methyl-2-carboxywas then removed by distillation a t ca. 150 mm. until the naphthalene. pot temperature reached 60-75", and the pressure was then a, a-Dimethylbenzylsuccinic anhydride (11), reduced slowly maintaining the same temperature range. (1) Deceased. After loss of methanol had stopped the product was distilled ( 2 ) (a) J. von Braun, Bcr., 61B,441 (1928); E. C. Horning and in vacuo. The analytical samples were obtained by distillation through an efficieut column, and iri all cases the G. N. Walker, THISJOURNAL,74, 5147 (1952); A. J Attwood, A. Stevenson and J. F. Thorpe, J. Chem. Soc., 1755 (1923). (I)) C. L. crude product was found t o be quite pure. Hemett, ibid., 506 (1936); R. D. Haworth and G. Sheldrick. i b i d . , Procedure B .-The P-ketodimethylacetal was heated slow1~-in a claisen flask with ca. 1% by weight of sodium 636(3)(1935). E. Bergmann and M . Orchin, THIS J O U R N A L , 71, 1917 (1949). niethoxide to a temperature of 140-170°, while methanol mas (4) E. hl. Beavers, Brit. P a t e n t 668,574 (Mar. 19, 1952); W. G. removed by distillation at atmospheric pressure. The Bickford, e l al., J . A m . Oil Chem. SOL.,25, 251 (1948). pressure was then lowered while the remaining methanol (5) J. W. Cook and A. M. Robinson ( J . Chem. Soc., 505 (1938)) mas removed, and finally the crude product was distilled report t h e cyclization of a phenanthrene analog t o yield the two in vacuo. The analytical samples were obtained as in proisomers. cedure A . DEPARTMENT OF CHEMISTRY EMORY UNIVERSITY EMORY UNIVERSITY, GEORGIA ( 5 ) E. E. Royals and K. C. Brannock, THISJOURNAL, 75, 2050 (1953). (6) (a) A. N. Kesrneyanov. N. K. Kochetkov a n d A t . I. Rybinskaya, I z u e s f . N a u k . A k a d . S.S.S.R.. Oldel K h i m . Nauk, 395 (1951) [C. A , . 46, 3007 (195211: (b) Henry 1. Richmond, U. S.Patent 2,570,713 (1951).
0 (6) One isomcr was identical t o the known I-methyl-2-carhoxytetralin.
