Aug. 20, 1960
1,5-DIKETONES WITH SULFURIC ACID
instrument employing sodium chloride optics and CC14 s o h tions of freshly purified materials in matched 1.0-mm. sodium chloride cells unless otherwise indicated. A study of the effect of dilution on the split carbonyl band in the infrared spectrum of a-morpholino-p-phenylacetophenone gave the results: 9.21 rng./rnl. CClr in a 1.0-mm.
[CONTRIBUTION FROM
THE
4245
cell, ylo=o, 1700/62, yzc=o, 1686/70 (az/al = 1.13); 3.T8 mg./ml. CClr in a 1.0 mm. cell, yIc=0 1702-38, y2c=o, 1686/47 (az/ol = 1.23); 0.75 mg./ml. CCl, in a 5.0-mm. cell, ylc=o, 1701/46, y ~ c = o ,1684/56 (az/al = 1.21). LINCOLN, NEBR.
DEPARTMENT OF CHEMISTRY, UNIVERSITY OF CALIFORSIA, BERKELEY ]
The Reaction of 1,s-Diketones with Sulfuric Acid BY WILLIAMG. DAUBEN AND
JAMES
W. MCFARLAND~
RECEIVEDDECEMBER 7, 1959
It has been found that certain 1,5-diketones upon treatment with sulfuric acid give rise to products with bridged bicyclic structures rather than the expected fused ring systems. The strain in certain 3,2,l-bicyclic systems is discussed.
There are two general synthetic routes which have been employed to fuse an additional carbocyclic ring to a cyclic ketone. The first procedure, t h a t of Robinson,2 is the cyclization of a 1,5-diketone, such as Ia, with dilute acid or base. The second procedure, discovered by Wichterle3 and widely employed by P r e l ~ g ,is~ the . ~ cyclization of a 2-(3'-chlorocroty1)-cyclonone, such as I I a , with concentrated sulfuric acid. Although the reaction conditions of these two processes are very different,
& a
0 Ia,R=H b, R = COOEt
R
dElCI
with concentrated sulfuric acid yields bicyclic ketone I V in good yield. In contrast to the excellent reaction of I I b with sulfuric acid, Prelog and Zimmermann5 found that ethyl 2- (3'-chlorocrotyl)-cyclopentanone-2-carboxylate (V) yielded only an intractable mixture of materials. In this reaction, the presumed intermediate would be ethyl 2-(3'-oxobuty1)-cyclopentanone-2-carboxylate (VI) whose reaction has now been studied. When the diketone V I was allowed to react with 9.5% sulfuric acid, a solid isomeric product (C12H1,04) was obtained in good yield. The material could be hydrolyzed in either dilute acid or base to a carboxylic acid which, in
IIa, R = H b, R = COOEt
COOEt VI
CO,C,H, IIIa, R = H b, R = COOEt
IV
in both reactions the octalone I I I a is obtained in IX VII, R = CZH5 VIII, R = H good yield. I t has been proposed t h a t the vinyl chloro group acts as a disguised ketone and upon treatment with strong acid gives rise to the same turn, upon treatment with diazoethane was reconverted to the original ester. The infrared 1,5-diketone employed in the Robinson reaction. One particular facet of these reactions which is of spectrum of the ester in carbon tetrachloride showed special interest arises when the potential angular a single strong band a t 1740 cm.-' and in chlorogrouping is carbethoxy. In this case, I b under form showed two bands a t 1732 and 1720 cm.-'. Robinson conditions2 yields the expected octalone The spectrum of the acid in chloroform possessed I I I b , while I I b under Wichterle conditions4 gives bands a t 1750 and 1719 crn.-'. These data indirise to the bicyclic ketone IV. If the vinyl chloro cate the presence of a cyclopentanone and a cargroup is acting as a disguised carbonyl group, then boxyl function. No direct chemical evidence for the strongly acidic conditions used in the Wichterle the carbonyl group could be obtained since the reaction must be a controlling factor directing product would not react with 2,4-dinitrophenylring closure to a bicyclic system. Such now has hydrazine, semicarbazide or thiosemicarbazide, been found to be the case since I b upon reaction even under forcing conditions. The fourth and remaining oxygen function was shown to be a (1) United States Rubber Co. Fellow in Chemistry, 1955-1956; hydroxy group by the presence of a weak but Dow Chemical Co. Fellow in Chemistry, 1956-1957. definite band a t 3450 ern.-' in the infrared spec(2) E. C. DuFeu, F. J. McQuillin and R . Robinson, J. Chem. Soc., 53 trum, and by the presence of one active hydrogen (1937). (3) 0. Wichterle, J. Prochazka and J. Hofman, Col?. Czech. Chcm. as determined by the Zerewitinoff reaction. C o m m . , 13, 300 (1948). Based upon these data and assuming no deep (1) V . Prelog, P. Barman and hI. Zimmermann, Helw. Chim. Acta. seated rearrangement, there are two possible formsa, 1284 09491 ( 5 ) 1'. Prelog a n d M.Zimmermann, rbzd., 34, 2360 1 ~ 4 9 ) . ulations for the reaction product, VI1 and IX.
WILLIAMG. DAUBEN AND JAMES W. MCFARLAND
4246
However, the chemical reactivity of the material clearly establishes the correctness of structure VII. It was found t h a t upon oxidation by the KuhnRoth procedure the ester possessed two C-methyl groups, and the parent acid VI11 possessed one such function. Further, the stability of the tertiary hydroxy function to the strong dehydrating conditions would not be expected of structure IX, whereas the bicyclic nature of VI1 could endow such stability t o an alcoholic grouping. The stability of the hydroxyl function, and the unreactivity of the carbonyl group will be discussed in more detail later. Since the reaction of VI with sulfuric acid led to a single product in high yield, the Wichterle reaction of V was reinvestigated. A mixture of neutral and acidic products was obtained. Fractional distillation of the neutral portion gave two distinct fractions, a lower boiling one in 2.6yG yield, and a higher boiling one in 32y0 yield. A solid compound was obtained from the acidic fraction in 6.875 yield. The lower boiling neutral fraction had the composition CllH1802, and possessed two C-methyl groups (Kuhn-Roth) , one double bond (microhydrogenation) and one ethoxyl group (Zeisel). The infrared spectrum of the material exhibited a band a t 1732 cm.-' which is characteristic of a non-conjugated ester. From these data and the plausible reaction mechanism shown in the diagram, the product was assigned the structure of ethyl 4-methyl-~4-cycloheptenecarboxylate(X) . 6
O=c'
Le) K5
F02r2H5
-- H20
- 11s'
?-
x
Vol. S%
e1 XI
XI1
The solid acidic product had the composition C ~ O H ~ ~ and O ~ Cpossessed ~, one C-inethyl group (Kuhn-Roth). No double bond was present because the acid would not reduce solutions of bromine or potassium permanganate. The infrared spectrum of the material exhibited bands a t 1760 (cyclopentanone) and at 1704 cm.-l (acid). The ultraviolet spectrum showed a maximum only a t 290 mp ( E 22) which is indicative of a non-conjugated ketone. However, the acid did not react with 2,4dinitrophenylhydrazine, semicarbazide or thiosemicarbazide. The reaction of this acid with 6 N hydrochloric acid to give VIII, however, permits one to formulate its structure as 4-chloro-4methylbicyclo [3 : 2 : 11octan-8-on-1-carboxylic acid (XII). These studies afford some interesting observations on the chemistry of bicyclo [3 : 2 : 11octane systems. Of primary importance in determining the structure of the ketol-ester VI1 was the question of the surprising stability of the tertiary hydroxyl group under conditions known to cause elimination in the bicyclo [3: 3 : 11nonane system. The ease with which X I was converted to VI1 by OOyo sulfuric acid, however, would seem to indicate t h a t a double bond in such a bicyclic system introduces considerable strain, a strain released by hydration of the double bond. I n line with this hydration of a double bond are the results of Newman and Yu6 who found that by treating A2-bicyc!o [ 2: 2 : 21 octene with 80% sulfuric acid, 2-bicyclo [3 :2 : 11octanol was obtained. It is interesting to note t h a t while the ketone function of the ketol-ester VI1 is inert to a reagent such as thiosemicarbazide, X I reacts with no apparent difficulty. The difference in reactivity must be due to the additional substituent a t the C-4 position. It is evident t h a t in forming a thiosemicarbazone the C-8 carbon atom must expand its sp2 orbitals to sp3 at some point during the course of reaction. Consideration of the most probable conformation of the ketol-ester XI11
.lnalyscs of the higher boiling neutral fraction established a composition of C1?Hl6Oi,and, in addition the material possessed two methyl groups (Kuhn-Roth). one double bond (microhydrogenation) and one ethoxy group (Zeisel). Bands of OH 1760 and 1727 cm.-' in the infrared spectrum indicate cyclopentanone and ester features in the structure. The ultraviolet spectrum displayed a maximum only a t 294 r n p ( e 76) which demonstrates t h a t there is no conjugated system. The compound XI11 reacted readily with thiosemicarbazide to give a shows that the addition of any reagent to the carcharacteristic derivative. These data indicate bonyl group creates a 1,3-diaxial repulsive interthat the product is ethyl 4-methyl-A3-bicyclo- action between the substituents on C-4 and C-h. [ 3 : 2: 11octene-8-on-1-carboxylate ( X I ) , the five- Apparently, in this case the interaction is strong membered ring analog to the series of compounds enough to prevent reaction under normal condistudied by Prelog and his c o - ~ o r k e r s . ~ When tions. X I was treated with 90% sulfuric acid, the Experimental7 ketol-ester VI1 was obtained, thus demonstratEthyl 2-(3'-oxobutyl)-cyclohexanon-2-carboxylate ing the relationship between the two compounds. (Ib) prepared in the manner deqrribed by Drciding :und In additioii, the reaction of X I with G N hydro- Tornnsewski .x chloric acid qave VI11 as the product. Saponi(6) 11. S. Newman a n d Y . T. Yu, THISJ O U R N A L 7 4 , ,507 (1!352) fication of X I , however, gave an oily material (7) All melting p r i n t s a n d boiling points reported here are u n c o r ~ which could not be characterized. rected. T h e ultraviolet spectra were all taken in 95% etlianul Only ~
Aug. 20, 1960
1.5-DIKETONESWITH SULFURIC ACID
4247
19.40; mol. wt. (Kast), 240; C-inethyl (Kuhn-Roth): Ethyl 4-Methyl-A3-bicyclo j3 :3 : 1Inonen-9-on-I-car2.0 per mole; active hydrogen (Zerewitinoff): 0.95 per boxylate (IV) and the Corresponding Acid .--Sulfuric acid (95%, 5.0 cc.) was added dropwise with swirling and mole. cooling to 4.51 g. (0.0188 mole) of I b . T h e resulting redCompound VI1 does not discolor solutions of either brobrown solution was allowed t o stand for 36 hours at room mine or potassium permanganate under the usual conditions temperature after which time i t was poured cautiously into of testing unsaturation. All attempts to make a 2,4-dinitro200 cc. of saturated sodium bicarbonate solution. T h e basic phenylhydrazone, semicarbazone or thiosemicarbazone aqueous solution was extracted three times with 25-cc. derivative failed. portions of ether. The combined ethereal extracts were 4-Hydroxy-4-methylbicyclo[3:2:1]octan-8-on-l-carboxylic dried and the solvent was removed. The residual oil was Acid IVIII). (a).-With heating on a steam-bath for 30 fractionally distilled to yield 2.43 g. (58%) of I\', b.p. minutes, 505 mg. (2.23 rnmolesj of I'II dissolved in 10 cc. 96-98' (0.3 mm.), i t z 6 1.4921, ~ Am,, 287 rnp ( e 55); vmax of 6 N hydrochloric acid. Upon cooling, the solution gave 2720 , 1730 cm.-' llit.4 b.p. 82-84' (0.1 mm.), Amax 292 270 mg. (61y0) of VIII. The material was twice recrystalmf*( e 50) I. lized from chloroform-carbon tetrachloride; m .p. 176The basic aqueous solution from which the keto-ester l77", A,,,211 mp ( E 150); vmgX1750-1719 cm.-I. had been extracted was acidified with sulfuric acid, and then Anal. Calcd. for CloH~rOa(198.24): C , 60.59; H, 7.12. was extracted three times with 25-cc. portions of ether. Found: C, 60.45; H, 6.98; neut. equiv., 198; C-methyl The combined extracts were dried and the solvent was (Kuhn-Roth), 0.78 per mole. removed to yield 425 mg. (12%) of crude 4-methyl-A3Compound VI11 did not discolor either bromine or perbicyclo [3:3 : 1]nonen-9-one-l-carboxylic acid. This prodmanganate solutions. All attempts to make a 2,4-dinitrouct was twice recrystallized from hexane-chloroform; m.p. 140-141°, , , ,A 290 mg ( e 49) [lit.4 m.p. 139', Amax 293 phenylhydrazone, semicarbazone or thiosemicarbazone failed. mp ( € 3 2 )I . (b).-A solution of 243 mg. (1.07 nimole) of \TI and 7.0 Anal. Calcd. for CllHj403 (194.22): C, 68.02; H , 7.27. cc. of 2 N sodium hydroxide was stirred at room tempcraFound: C, 67.87; H , 7.18; neut. equiv., 191; C-methyl ture for 24 hours. A small amount of amorphous, insoluble (Kuhn-Roth), 0.45 per mole. material was filtered, and the filtrate was acidified. The The acid also was obtained from I V in the following aqueous solution was extracted with ether three times, and manner. A solution of 525 mg. (2.36 mmoles) of IV, 0.65 the combined extracts were dried. T h e solvent was reg. of potassium hydroxide and 6 cc. of methanol was heated moved to give a waxy solid which was twice recrystallized under reflux for 4 hours. The solution was poured into 50 from chloroform-carbon tetrachloride; m.p. 175-176'. cc. of water, and was acidified with hydrochloric acid. The ?;a depression of the melting point was observed when this aqueous solution was extracted three times with 20-cc. acid was mixed with t h a t obtained in ( a ) . portions of ether. The combined extracts were dried, and The Reconversion of VI11 t o VI1.-Using the method dethe solvent was removed to give 374 mg. of a crystalline scribed by McKay and his co-workers,'O diazoethane was red-brown residue. The material was recrystallized from passed through a solution of 194 mg. (1.00 mmol.) of VI11 hexane-chloroform to give 166 mg. (36%) of 4-rneth~1-A~- and 10 cc. of ether. After standing a t room temperature for bicyclo [3 : 3 : 1]nonen-9-one-l-carboxplic acid, m.p. 14015 minutes, the solution was treated with a few drops of 142'; niixed m.p. with acid obtained previously, 140.5glacial acetic acid to destroy the excess diazoethane. The 142'. ether was distilled, and the oily residue was then recrystalThe thiosemicarbazone of IV was obtained by dissolving lized from methanol-water to give 89 mg. (39%) of V I I , 100 mg. (1.10 rnmoles) of thiosemicarbazide and 22 mg. m.p. 61-62', mixedm.p. 6142'. (1.00 mmole) of I V in 5 cc. of hot 80% methanol, and alEthyl 2-(3'-Chlorocrotyl)-cyclopentanone-2-carboxylate lowing the resulting solution to stand at room temperaV) was prepared by the procedure of Prelog and Zirnture for 18 hours. The crystalline thiosemicarbazone was mermann.j filtered; yield256 mg. (87%), m.p. 209-210Odec. The Reaction of V with Sulfuric Acid.-With cooling and Anal. Calcd. for C14H2101N3S (295.40): C, 56.92; H , swirling, 14 cc. of 95% sulfuric acid was added dropwise to 13.1 g. (0.0533 mole) of V. Caution had to be observed 7.17; 3, 14.23; S, 10.86. Found: C, 57.20; H , 6.88; about half-way through the addition of the acid because the N, 14.40; S,10.70. Ethyl 2-(3'-Oxobutyl)-cyclopentanone-2-carboxylate (VI). evolution of hydrogen chloride became so vigorous that the --A solution of 49.6 g. (0.318 mole) of ethyl cpclopenta- reactants were in danger of foaming over the walls of the non-2-carboxylate, 30.0 cc. (25.5 g., 0.354 mole) of anhp- reaction flask. When the addition was complete, the redrous methyl vinyl ketone, 12.0 cc. of triethylamine and sulting solution was allowed to stand at room temperature for 7 days. The solution was poured into 700 cc. of ice and 200 cc. of dry benzene was allowed to stand at room temwater, and was then neutralized with sodium bicarbonate. perature for 7 days. The volatile components were then The neutral organic was extracted from the aqueous soluremoved and the product YI was fractionally distilled, tion using three 80-cc. portions of ether. The combined ~ yield 66.3 g. (927,) b.p. 110-142" a t 2.5 mm., n Z 51.4641, extracts were dried, and the solvent was removed to give an (lit.9 b.p. 139' at 3 mm.). residue which was fractionally distilled into X and X I . Anal. Calcd. for C12H1804(226.26): C, 63.70; H , 8.02; oily Compound X was presumed to be ethyl 4-methyl-A4OCZH5, 19.92. Found: C, 63.84; H, 8.14; OCZH,, cycloheptencarboxylate plus about 5% of an n,B-unsaturated 20.28. isomeric ester, yield 250 mg. (2.6%), b.p. 78-80' (2.0 inm.), The disemicarbazone uas made in the usual manner, ~ 2 1.4640, ~ 5, , ,A ~ 232 mp ( e 510), ymav1732 ern.-'. m.p. 178-179" (lit.$ 178-179'). Anal. Calcd. for CllHI8O2(182.23): C, 72.49; H , 9.95; Ethyl 4-Hydroxy-4-methylbicyclo [3:2:1]octan-8-on-l-carOC2H6,24.78. Found: C, 72.29; H , 9.80; OC2Hj, 24.22; boxylate (VII).-While cooling in an ice-bath, 1.0 g. of bonds (microhydrogenation, Pd-C in 95T0 ethanol), 95% sulfuric acid was added dropwise with swirling to 500 double 1.04 per mole; C-methyl (Kuhn-Roth), 1.64 per mole. mg. (2.21 mmole) of 1'1. The resulting solution was alCompound XI was ethyl 4-rnethyl-~~-bicyclo [3 : 2 : 1 ]lowed to stand a t room temperature overnight, and then was poured into 10 cc. of ice and water. The aqueous solu- octene-8-on-1-carboxylate; yield 3.53 g. ( 3 2 % ) , b.p. 111294 In@ ( e 76); 115' (1.2-1.6 m m . ) , n Z 5 D 1.4826, A,, tion was neutralized with sodium bicarbonate, and the ymax1760, 1727cm.-1. solid product was filtered and washed with water; yield Anal. Calcd. for C12H160S(208.25): C, 69.20; H, 7.75; Two recrystallizations from 375 mg. (75%), m.p. 53-56', methanol-water gave pure VII, m.p. 61-62', , , ,A 211 mp OCZH5, 21.64. Found: C, 69.42; H , 8.02; OCzH6, 21.69; double bonds (microhydrogenation, Pd-C in 95y0 i s 170); 3450, 1732, 1720 cm.-'. A m / . Calcd. for C12Hr804 (226.26): C, 63.70; H , 8.02; ethanol), 0.91 per mole; C-methyl (Kuhn-Roth), 1.74 per mole. OCzH5, 19.92. Found: C, 63.42; H. 8.19; O C Z H ~ , The aqueous solutioii from which the neutral organic the definitive peak? of the infrarrd ipectra are recorded. T h e analyses products had been evtracted W-:E ucitlifietl with sulfuric acid, ere performcd hy t h e Microanalytical Laboratory, Collefie of Chemisand was cxtractcd four times with 70-cc. portions uf ether. t r y , the University of California. (8) A . S. Dreiding and A . J Tomaseaski, THIS J O U R N A L , 77, 168 (1955) ( Y ) J R . Nunn and W. S. Kapson. J C - l l e i ~ t Soc.. , 825 (1940)
( I O ) 8.F. M c K a y , TIXIS J O U R N A I , , 71, 1908 (1949); A . F hlcKay, W L O t t . G U'. Taylor. M. S Buchannn and J . F. Crooker. CU?L.J . Res , Bas, 683 (11150).
4248
WILLIAMG. DAUBEN AND
T h e combined extracts were dried, and the solvent was removed to give 3.38 g. of a white powder, m.p. 130-155'. Two recrystallizations from chloroform-carbon tetrachloride gave 788 mg. (6.870) of 4-chloro-4-methylbicyclo[3:2: 1]octan-~-one-l-carboxylicacid ( X I I ) , m a p . 180182". One more recrystallization gave a n analytical sample, m.p. 182-183', Xma, 290 mp ( e 22); vmax 1760,1704 ern.-'. Anal. Calcd. for CloHlaOaCl (216.66). C, 55.43; H, 6.05; C1, 16.37. Found: C, 55.31; H, 5.90; C1, 16.14; C-methyl (Kuhn-Roth), 0.76 per mole. Attempts to make a 2,4-dinitrophenylhydrazone, semicarbazone or thiosemicarbazone derivative of X I 1 all failed. The Thiosemicarbazone of XI.-A solution of 265 mg. (1.27mmoles) of XI, 116 mg. (1.27 mmoles) of thiosemicarbazide and 7 cc. of 95% ethanol was heated under retlux for 4 hours. Rapid cooling of the reaction solution gave 182 mg. (:lye) of the thiosemicarbazone as leaflets, m.p. 191193.5 . (When the cooling occurred more slowly, a dimorphic form of the product was obtained: prisms, m.p. 185-188" .) Two recrystallizations from 9570 ethanol gave a n analytic sample, m.p. 193-195'. Anal. Calcd. for C13H1902NaS (281.37): C, 55.49; H, 6.81; K, 14.93; S, 11.40. Found: C, 55.58; H, 6.63; S , 14.86; S, 11.20. The Conversion of XI to VI1.-With swirling and cooling in an ice-bath 536 mg. (2.58 mmoles) of X I was treated with 1.0 cc. of chilled 90% sulfuric acid (dropwise addition). T h e resulting solution was allowed to stand at room temperature overnight, and was then poured into 100 cc. of saturated sodium bicarbonate solution cautiously. The solid organic material was filtered t o give 189 mg. of a product, melting 45-51'. Recrystallization from methanolwater gave 58 mg. (10%) of VII, m.p. 58-60', mixed m.p. 60.5-62 '. The aqueous solution was extracted three times with 20-cc. portions of chloroform. The combined extracts were dried, and the solvent was removed to give 313 mg. of a n oil which would not crystallize from various solvents The infrared spectrum of the material, however, showed it to be largely VII. Hydrolysis of XI by 6 iV Hydrochloric Acid.-A mixture of 270 mg. (1.30 mmoles) of XI and 5.5 cc. of 6 N hydrochloric ac
