SUBSTITUTED GLYC~DIC ESTERS
March 5 , 1034
phenyl diketone which was isolated as 2-phenyl-3-benzylquinoxaline, m.p. 96-97’ (lit. 97-98°,26 98-9Q08). The ultraviolet spectrum of the quinoxaline closely resembles
THE
the spectrum of 2-phenyl-3-benzhydrylquinoxaline,absorption peaks occurring a t 240 mp (e, 35,000) and 324 mw (emax 10,000). CAMBRIDGE 39, MASS.
( 2 5 ) T. L. Jacobs, THIS J O U R N A L , 58, 2272 (1936).
[COXTRIBLJTION FROM
1237
DEPARTMENT O F CHEMISTRY
O F THE
UXIVERSITY O F MAINE]
The Synthesis and Reactions of Certain a-Substituted Glycidic Esters B Y HORTON H.
LfORRIS AND hfARGARET
L. LUSTII’
RECEIVED SEPTEMBER 23, 1953 Several a-substituted glycidic esters, ethyl a-n-alkyl-a,@-epoxycyclohexylideneacetates, have been prepared. Saponification of these esters and decarboxylation of the resulting glycidic acids gave ketones as primary products, as proved by alternate synthesis, mixed melting points of derivatives and attempted oxidation. Contrary to an earlier report, ethyl a-%decyl-a,@-epoxycyclohexylideneaceticacid decarboxylated to yield I-cyclohexyl-1-undecanoneinstead of a-n-decylcyclohexanecarboxaldehyde. An important side reaction in the preparation of these glycidic esters was found to be the saponification of a portion of the glycidic ester during the course of the reaction, the resulting acid material accounting for over 40y0 of the yield in one case.
products in every case were ketones, as shown by alternate synthesis, mixed melting points of derivatives and oxidation studies. Evidently Darzens was misled by the fact that a small amount of an oxidizable contaminant present in the decarboxylation products gives aldehyde color tests. I n order to prove the structure of the decarboxylR’ H R‘ R”’ ation products, a series of cyclohexyl ketones were \ I KaOEt \ ’ C=O + R’”-C-COIR ‘2-C-COzR prepared by the following methods. The ethyl, npropyl and n-amyl cyclohexyl ketones were obxI tained by use of a modified Friedel-Crafts type acylation as adapted from the procedure reported by Nenitzescu and Cioranescu.* The n-heptyl, noctyl and n-decyl cyclohexyl ketones were prepared by means of a condensation of the appropriate din-alkylcadmium compound and cyclohexanecarboxyl chloride, following a procedure given by C a s ~ n . Semicarbazones ~ and 2,4-dinitrophenplRv H hydrazones were prepared from these ketones and I I1 Darzens has reported2 that glycidic acids with mixed melting points determined with the correhigher molecular weight a-substituents, in particu- sponding products obtained from the decarboxylalar n-decyl, rearrange on decarboxylation to form tion of the glycidic acids. No depression of the tertiary aldehydes of the type 11. It has been melting point was observed in any case. In order further to establish the structure of the s h o ~ n ~that ” ? if~ the a-substituent is methyl, ethyl, n-butyl or n-hexyl, a ketone of the type I results. decarboxylation product from a-n-decyl-a,P-epoxyIn order to determine a t what point between n- cyclohexylideneacetic acid and to attempt to find an hexyl and n-decyl the rearrangement to aldehyde explanation for the fact that the material gave altakes place, an homologous series of glycidic esters dehyde color tests, a sample was subjected to oxida(ethyl a-n-alkyl-alp-epoxycyclohexylideneacetates)tion with aqueous permanganate. Less than 15y0 with a-substituents of methyl through n-decyl was of the material could be oxidized. Repetition of prepared from cyclohexanone and a-bromo esters, the oxidation, using acetone as solvent, gave essenusing sodium ethoxide as the condensing agent. tially the same results. There seems to be no doubt that ketones will The glycidic esters with a-substituents of n-butyl result on the decarboxylation of a-substituted and n-hexyl were reported in a previous article3b from this Laboratory and are not included here. glycidic acids when the a-substituent is an alkyl The esters were saponified and the resulting acids group. Glycidic esters, however, have occasiondecarboxylated by heating for three or four hours ally been reported to rearrange to a-keto esters on a t atmospheric pressure. The decarboxylation heating, and such esters would be expected to yield aldehydes on decarboxylation. Such a rearrange(1) From the thesis submitted by M. L. Lusth in partial fulfillment ment may have occurred in the case reported by of the requirements of the degree of Master of Science in Chemistry, Darzens, and would account for the results he obJune, 1953. Presented before the Division of Organic Chemistry a t the 124th Meeting of the American Chemical Society, Chicago, Ill,, tained. Preliminary evidence indicates that this September 9. 1953. is not the case, but the possibility is being checked. (2) G. Darzens, Compf. rend., 195, 884 (1932). The preparation of the condensing agent in the (3) (a) M . Mousseron and R. Granger, ibid.. 218, 358 (1944); The condensation of an aldehyde or ketone with an a-halo ester, in the presence of a basic condensing agent, yields an a$-epoxyester (glycidic ester). The reaction is of interest since it provides a route to aldehydes or ketones of greater complexity than those used in the original condensation.
J
&I. Mousseron, el ai., Bull. S O C . chim. France, 598 (1947); (b) N. K. Nelson and H. H. Morris, THIS J O U R N A L , 76, 3337 (1953).
(4) C. D. Nenitzescu and E. Cioranescu, Bcr., 69, 1820 (19313). ( 5 ) J. Cason, THIS JOURNAL, 68, 2078 (1946).
1238
HORTON H. MORRISAND MARGARET L. LUSTH
Vol. 76
glycidic ester condensation is a time-consuming Technical grade reagents were used and the esters purified fractionation through a suitable column. task, and therefore the use of finely dispersed so- bySince i t did not prove practicable t o use the sodium disdium (particle size 1-10 p ) in xylene was attempted. persion as condensing agent for the reaction, the dispersion The results of this work indicate that sodium dis- was used t o prepare sodium ethoxide and was very satisfacpersions can be used as condensing agent in this re- tory for this purpose, since i t eliminated the time-consumand somewhat dangerous method previously reported action only if there is good cooling. The sodium ing The condensing agent, sodium ethoxide, was prepared by dispersion was used quite successfully, however, to dropping the stoichiometric amount of absolute ethanol into prepare anhydrous sodium ethoxide-the ethoxide a vigorously stirred sodium-xylene dispersion. A dry nioften being formed in a finely divided form, which trogen atmosphere was maintained throughout the reaction. product was usually a dispersion of sodium ethoxide in remained in suspension in the xylene and could be The xylene and could be added from the flask t o the mixture of poured into the mixture of the other reagents. cyclohexanone and a-bromo ester by means of a suitable From each glycidic ester preparation a certain addition tube or a separatory funnel having a wide bore amount of acid material was recovered. With one stopcock. of Glycidic Esters.-To a mixture of one mole exception, the acid materials were found to con- of Preparation the a-bromo ester and 1.5 moles (147 g.) of cyclohexanone tain the corresponding glycidic acid, as shown by in a 2-liter, 3-necked flask partially immersed in an ice-saltthe fact that they decarboxylated to give the ex- bath, and equipped with a Trubore (Ace Glass, Inc.) stirrer pected ketones in yields comparable to those ob- fitted with a Hirschfelder type blade, a reflux condenser with nitrogen gas inlet and outlet, is added 1.5 tained from the saponification and subsequent equipped moles of freshly prepared sodium ethoxide dispersed in decarboxylation of the pure glycidic esters.6 -41- xylene, by means of a suitable addition tube or separatory though the acid material recovered in the prepara- funnel. The condensing agent should be added at such a tion of ethyl a-n-heptyl-a,P-epoxycyclohexylidene-rate t h a t the mixture does not heat up. Throughout the addition the mixture is vigorously stirred. The addition, acetate seemed to decarboxylate normally, the subsequent stirring, and steam-bath treatment should take decarboxylation product was not a ketone or alde- place in a n atmosphere of dry nitrogen. When the addition hyde and may have been an unsaturated acid, since is complete, the mixture is stirred for three hours, and alit was soluble in base and decolorized bromine in lowed to stand under nitrogen for 18 hours. The mixture is then heated on a steam-bath for one hour, with stirring, carbon tetrachloride. Yields of the acid material during which time the mixture usually changes from a dark in the various preparations ranged from 20% to as yellow t o a pale yellow or white color. The cooled mixture high as 4670, making the over-all yield of condensa- is then treated with 100-150 ml. of 5% acetic acid solution, tion product in the neighborhood of 70-8070, al- the mater layer separated, and 500 ml. of ether added to the layer. The ether-xylene solution is washed with two though only 40-45% yields of the glycidic esters xylene 100-ml. portions of saturated sodium bicarbonate solution, were obtained. No satisfactory explanation for two 100-ml. portions of water, 100 ml. of saturated sodium the formation of the acid material has yet been chloride solution, and dried over anhydrous sodium sulfate found, although it is suspected that saponification for several hours. The ether is then removed on the steambath, the xylene removed by flask distillation a t reduced may be taking place during the washing procedure. pressure and the crude ester fractionated at reduced presDarzens also reported that a-n-decyl-p,P-di- sure. A 40-45010 yield of glycidic ester is obtained. The methylglycidic acid and a-n-decyl-P-phenyl-p-methsodium bicarbonate, water and sodium chloride washes are ylglycidic acid decarboxylated to give tertiary alde- combined with the original water layer, which is then poured slowly and with constant stirring into a slurry of concenhydes. Preliminary work has indicated that ke- trated hydrochloric acid and ice. The mixture is immeditones are obtained from these acids also. ately extracted with ether and the ether solution washed As expected, the derivatives of the asymmetrical with water, after which i t is dried over anhydrous sodium ketones reported here are low-melting-the tnelting sulfate. The ether is removed on the steam-bath 1e:tving highly viscous, crude glycidic acid behind. The acid3 points decreasing with increasing molecular weight the are probably solids, but are very difficult t o crystallize. of the ketones. The crude acid is subjected to the decarboxylation procc.3‘1
Experimental Preparation of Intermediates.-The cyclohexanone used was Matheson C.P. grade and was redistilled through a 10 bubbler Clarke-Rahrs fractionation column. The xylene was Will Corporation “Water and Acid Free” xylene. Both the xylene and cyclohexanone were dried over anhydrous sodium sulfate before use. Absolute ethanol was prepared from commercial absolute ethanol by distilling from lump sodium. The sodium dispersion was a 50% by weight dispersion in xylene and was kindly donated for this work by National Distillers Corporation and by E. I. du Pont de Nemours and Company. The particle size of the sodium was stated as being from 1 to 10 p . The a-bromo esters were prepared by the Hell-Yolhard-Zelinsky reaction. (6) W. S. Johnson, el al , TKISJ O U R K A L , 7 6 , 4995 (l953), in a n excellent study of t h e glycidic ester condensation, published after this paper was written, have shown t h a t t h e method of saponifying t h e glycidic esters used in this work often gives cyclohexenylglycolic acids as a product. T h e formation of such acids is entirely possible in t h e present work and might account for t h e results reported concerning the acid material recovered in t h e preparation of ethyl a-n-heptyla,@-epoxycyclohexylideneacetate. We have assumed, for purposes of yield calculation, t h a t the acid materials were entirely glycidic acid (our hasis for this is mentioned), h u t the presence of appreciable amounts of t h e cyclohexenylglycolic acids is certainly not prohibited in view of t h e rclativeiy low yields o f ketone ohtained from the d e carlx,xslation < > [the glycidic acids.
dure described later. Saponification of the Esters.-A sample of the ester i5 added t o 1.5 times the equivalent amount of sodium hydroxide dissolved in diethylene glycol. A small amount of water may be added t o effect the solution of the sodium hydroxide in the glycol. The mixture is refluxed for one hour or until a homogeneous mixture is obtained. The cooled mixture is then poured into a slurry of concentrated hydrochloric acid and ice and the mixture immediately extracted three times with ether. The ether layer is separated and washed twice with water and once with a saturated sodium chloride solution, after which i t is dried over anhydrous sodium sulfate. The ether is removed by means of a steanibath, leaving a 90-95yo yield of crude acid. Decarboxylation of the Glycidic Acids.-The decarboxylation of the glycidic acids was effected b y heating the crude acids a t atmospheric pressure, under a fractionating column a t such a temperature t h a t they bubbled vigorously, but did not distil. Evidence of decarboxylation was obtained by testing the evolved gas with lime water. After three hour\, or after bubbling ceased, the decarboxylated product W L S distilled a t reduced pressure. A 40-60% yield (based on glycidic ester) of ketone was obtained. The results of some preliminary work on the use of sodium dispersion as condensing agent in the glycidic ester condensation indicate that the sodium dispersion is satisf;ictc)r!. only when there is good cooling. Of four attempted p r e p t rations of ethyl ~-n-butyl-a,~-epoxycyclohcxylidciicacct~ite, using :t sodium-xylene dispersion as condensing Ligcnt, only
CY-SUBSTITUTED GLYCIDIC ESTERS
March 5 , 1954
1239
TABLE I
R
