SYNTHESIS OF LACTONES
Aug. 5, 1955
repeated recrystallization from alcohol gave orange plates, melting a t 152-154' alone a i d on admixture with the cyclohexanone derivative. Anal. Calcd. for C12HldN40a: C, 51.79; H, 5.07; N, 20.13. Found: C, 51.88; H, 4.89; N, 20.23. The recrystallized distillation residue was again recrystallized from alcohol to give colorless prisms of dimethylamine hydroiodide, m.p. 147-150". Anal. Calcd. for C2Hyi'i.HI: C, 13.87; H, 4.62; N, 8.08. Found: C, 13.60; H , 4.41; N,8.21. (b) The Formation of VI1 from V.--\Yhen 130 mg. of the freshly prepared free base I' was heated in a sealed tube at loo", the material first fused and then decomposed after 30 seconds. The resulting solid mass was washed with dry ether and distilled with aqueous alcohol. The distillate gave a 2,4-dinitrophenylhydrazone (yield lOO7,, m.p. 155156') which mas identified as the cyclohexanone derivative by a mixed m.p. determination. (c) The Formation of VI1 and dl-cis-2-Chloro-N,N-dimethylcyclohexylamine (XII) from 11-Hydrochloride.Dried 11-hydrochloride (1.04 9.) in a sealed tube was heated in boiling xylene. It melted a t once and then decomposed after about 7 minutes to give a solid mass, which was treated as described under (b) . The 2,4-dinitrophenylhydrazone of cyclohexanone was obtained ; yield E167~, m.p. and mixed m.p. 150". The distillation residue was dissolved in water. The oily layer which apeared on the addition of alkali was extracted with ether. The ether layer yielded a picrate (yield 2470, m.p. 145-150') which was repeatedly recrystallized from ethyl acetate to give needles, m.p. 178', which were identified as dl-cis-2-chloro-N,X-dimethylcyclohexylamine ( X I I ) picrate by a mixed m.p. determination. Anal. Calcd. for C8H1&Cl.CcHSN307: N, 14.35. Found: N, 14.47. ( d j The Formation of VI1 and Crude dl-trans-2-ChloroN,N-dimethylcyclohexylamine from V-Hydrochloride.-VHydrochloride (40 mg.) was heated in a sealed tube a t 125". When the decomposed material was treated as described in ( b ) , the 2,4-dinitrophenylhydrazone of 2yclohexanone was obtained, m.p. and mixed m.p. 152-154 When the decomposed material was dissolved in water and extracted with ether after the addition of potassium carbonate, the extract gave a picrate which contained halogen; yield 45%, m:p. 138-142'. This picrate showed no depression of melting
.
[CONTRIBUTION FROM
THE
4079
point on admixture with an authentic sample of dl-trans-2chloro-N,N-dimetliylcyclohexylarniite picrate and a depression of melting point on admixture with its czs epimer. dl-cis-2-Chloro-N,N-dimethylcyclohexylarnine(XII) was prepared by the Mousseron procedurels; b.p. 100' (25 mm.). The hydroiodide, which melted at 168-169" after recrystallization from acetone-ether, was unchanged when heated in a sealed tube a t 140'. Anal. CaIcd. for CsHieNC1.HI: C, 33.16; H, 5.87; 5 , 4 3 3 . Found: C, 33.28; H, 5.90; N, 4.65. Intramolecular Interconversion of solution of 1 . O i g . of 11-hydrochloride in 10 nil. of absolute ethanol n a s refluxed for 5 minutes and then kept at room temperature for 3 days. Ether (15 ml.) waz added and the mixture kept for 50 days. During this interval colorless needles appeared (m.p. 150-155", yield 520 mg.) which after three recrystallizations from ethanol melted a t 156-157' alone and on admixture with the hydroiodide of 11. The picrate was also identical with the picrate of 11-HI. Anal. Calcd. for CRHIONI.HI: C, 25.17; H, 4.46. Found: C, 25.47; H, 4.58. After 50 days different needles appeared (yield 120 mg.) which were recrystallized from ethanol as cubes, m.p. 184185O. Anal. Calcd. for C s H d C l . H I (dl-trans-2-chloro-N,Xdimethylcyclohexylamine hydroiodide) : N, 5.84. Found: N, 5.78. The picrate (m.p. 166-168') gave a positive Beilstein test for halogen, but contained no iodine. Since the melting point of the picrate was depressed on admixture with authentic dl-cis-2-chloro-N,N-dimethylcyclohexylamine(XII) picrate (m.p. l W ) , it is probably the trans epimer. Anal. Calcd. for C8Hl&C1CeH3N40r(dl-trans-2-chloroN,N-dimethylcyclohexylamine picrate): C, 43.02; H , 4.86; N, 14.35. Found: C,43.25; H, 4.94; N, 13.95.
Acknowledgment.-The authors are indebted t o the Service Center of Microanalyses of the Kyushu University and also to the Microanalytical Section of this Institute for the microanalyses. (15) M. Mousseron, R. Jacquier, R. Henry and M. Mousseron Carnet, B d . soc. chim. France, 628 (1953).
KATAKASU, FUKUOKA, JAPAN
RESEARCH DEPARTMENT, UNIONCARBIDE CHEMICALS Co.]
Synthesis of Lactones1 BY PAULS.STARCHER AND BENJAMIN PHILLIPS RECEIVED FEBRUARY 21, 1958 The use of peracetic acid in an inert solvent mas investigated in the Baeyer-Villiger synthesis of lactones from cyclic ketones. Side reactions were minor with cyclopentanone and various cyclohexanones, and it was possible to isolate by distillation the corresponding 6-valerolactone and c-caprolactones in high yield as stable monomeric compounds. Several previously unreported alkyl-substituted e-caprolactones were prepared by this method. Extension of the reaction to cycloheptanone and cyclooctanone resulted in extensive ring cleavage to the corresponding dibasic acids.
A previous contribution2 from this Laboratory described a new synthesis of peracetic acid-one which gave the product as a solution in an inert solvent. The present paper deals with the use of this reagent for the preparation of lactones by reaction with cyclic ketones. The limitations which have made the Baeyer(1) Presented in part at the Atlantic City Meeting of the American Chemical Society, September 1950. (2) B . Phillips, F. C.Frostick, Jr., and P. S. Starcher, THISJ O U R N A L , 79, 5982 (1957).
Villiger reaction3 and its various modifications unsatisfactory in the past for the preparation of many acid-sensitive products, such as b-valerolactone and e-caprolactone, have been noted.4 High yields of these products were obtained in the present study, and this was attributed to the type of peracetic acid solution used. The absence of inor(3) A. Baeyer and V. Villiger, Ber., 32, 3625 (1899); 38, 858 (1900). (4) W. F. Sager and Alan Duckworth, THIS JOURNAL, 77, 188 (19.55).
TABLE I n OXIDATION OF KETONES, (CH2)nC=0, Molar Keratio tone, peracid/ n ketone
5 6
0.33 0.33 1.23
7
1.25
4
Yol.
PAULS.STARCHER A N D BENJAMIV PIIILLIPS
4080
IYITH
so
PERACETIC XCIU
W Reaction Time, Temp., hr. OC .
8 6.25 10 8.5
40 40 70
170
Products
6-Valerolactone E-Caprolactone r-Enantholactone Pimelic acid p-Caprylolactonea Suberic acid
Yield, M.p. or b.p., % O C . (mm.)
n30~
84 85 29 33 6
83 (4) 108 (10) 70 (5) 102 , ,
....
59
137
,...
1.4540 1.4605 1.4688 , . ,.
Sapon. or neut. equiv. Calcd. Found
100 114 128 80.1
87.1
101 113 130 80.8
Hydrazide, m.p., 'C. Bound Lit.
105 116-116.5 123.5124
105-10610 1147115% 123'
87.5
S o t isolated.
ganic impurities, notably water, hydrogen peroxide, benzoic acid but only after reaction times of 17 and mineral acids and salts, reduced polymerization to 42 days, respectively, a t room temperature. We a minimum during the reaction step and avoided found that a t 7'0" both cycloheptanone and cyclothe formation of many of the by-products which octanone reacted reasonably rapidly with peracetic plagued previous investigator^.^-'* Co-product acid, but the major products were not monomeric acetic acid was the only significant initiator of by- lactones. Cycloheptanone gave approximately product formation in the present synthesis. The equal amounts of l-enantholactone and pimelic acid problem of an acidic co-product is always en- while cyclooctanone gave principally suberic acid. countered when any peracid is used in the Baeyer- Cyclohexanone is much more resistant to ring Villiger reaction. Acetic acid is one of the least cleavage than either cycloheptanone or cycloocobjectionable not only because i t is a weak acid but tanone. Oxidation of cyclohexanone with peracealso because its relatively low boiling point makes tic acid a t 80" gave only a small amount (170) of for its easier removal and facilitates subsequent re- adipic acid." Table I summarizes the data or1 the covery of the lactones by distillation under re- oxidation of these unsubstituted cyclic ketones. duced pressure. The use of an excess of ketone or Although no kinetic data are available on the rcan azeotropic agent such as ethylbenzene also helps action rates of various alkyl-substituted cycloin the rapid removal of acetic acid. hexanones with peracetic acid, a rough approxiThe lactones obtained in this study are stable. mation of their relative reactivities may be obSamples which were stored for three years a t room tained from Table I1 where the times and temtemperature showed only slight changes in refrac- peratures for a given yield are indicated. Cyclotive index. Previous ~ n e t h o d s , ~ -based '~ on a hexanone reacted more rapidly with peracetic acid variety of peracids, employed neutralization of the than alkyl-substituted cyclohexanones. Increasing reaction mixture and extraction of the co-product the number of substituents appeared to decrease acid and, in most cases, except for €-substituted t- the reaction rate: note methyl-, dimethyl- and tric a p r o l a ~ t o n e s , ~ ~the ' ~ ~ products '~ were primarily rnethylcyclohexanones. -in increase in the size of polymeric or polymerized during or after distil- a substituent, particularly one in the Zposition, lation. also appeared to decrease the rate. The position In contrast to cyclopentanone and cyclohexa- of substituents on the ring also appeared to affect tione, both of which reac trelatively rapidly with the reaction rate. For example, 2,G-dimethylperacetic acid to give good yields of lactones, cyclo- cyclohexanone reacts more slowly than the 3,sheptanone and cyclooctanone reacted very slowly isomer, probably because the methyl groups in the with peracetic acid a t moderate temperatures. 2- and 6-positions tend to mask the carbonyl group. This was in line with the earlier work of Friess,',I6 Except for 3- and 4-methylcyclohexanones, the who obtained lactone polymers from cyclohepta- above conclusions are in accord with studies'?made none and cycloactanone after treatment with per- on the stabilities of the cyanohydrins of alkyl( 5 ) G H. Sjorklund and &', F i . Hatcher, 7 ' , a n e . I ( o y d .Sur ( : a n . , substituted cyclohexanones, where the steric con4 4 , Sec. 111, 2 5 (1950). cepts of "axial crowding" and "equatorial inter(6) W. Dilthey, 11. Inckel and H Stephan, . I . f i r u k l . V/fe!?L., 1 6 4 , ference" were used to explain the relative stabilities 219 (1940). of highly alkylated cyclohexanone cyanohydrins. (7) S . L. Friess, THISJ O U R N A L , 71, 2571 (1949). Several cyclohexanones substituted in the 2(6) M. Hudlicky, Colleclion Czcchoslw. Chem. Commrrns., 1 6 , 283 (1951). position have been treated with peracetic acid and (9) R P. Linstead and H . N . Rydon, J . Chem. SOL.,1999 (1934). the resulting lactones obtained in very high yield (10) R . Robinson and L. 1%. Smith, i b i d . , 371 (1937). (see Table 11). Although two products are pos(11) M. Stoll and W . Sherrer, Helv. Chipn. Acta, 13, 14%(1930). sible (equation, R' = alkyl, R = H), only one (12) N. Tokura and R. Oda, Bull. Insl. Phys. Chem. Reseavch (Tokyo), 22, 835 (1943). product (I) was obtained. We were unable to (13) P. Karrer a n d 0 . H a a b , Heiw. C h i m . Acta, 32, 973, 111 (1949). find evidence of the alternate product 11. Our re(14) R . J . Taylor and W.T. Weller, British Patent 748,801. sults agree with those of Baeyer and Villiger3 and (15) H . X I . Walton, J . O r g . C h e m . , 2 2 , 1161 (1967). with the generalization made by Doeringlg that (16) S. L. Friess and P. E. Frankenberg, THISJOURN.41., 7'4, 2679 (1952). T h e reactivities of peracids with cyclic ketones of different ring secondary and tertiary groups migrate more
size have been compared with reacti\ities of other reagents which add t o carbonyl groups, and the low reactivity of the medium size rings has been explained on t h e basis of I-strain and ring conformation involving a n "0-inside'' carbonyl group. See V. Prelog, J . Chem. Soc., 4'23 (1950), and H. C. Brown, R R Fletcher and R . R. Johannessn, Tars J O U R N A L , 78, 212 (19611,
(17) B. Phillips and P. S. Starcher, unpublished results ( 1 8 ) 0 . H. Wheeler and J Z. Zabicky, Ch,~!ws/vy& Z m f u W v IXXX (1950).
(19) W .von E . Doering and Louise Spews, 'I'ms (1.060)
8
Tc>iruv,\i.,
74,551s
ilug. 5 , 193s
SYNTHESIS OF LACTONES
40s 1
readily to oxygen in this reactionIgathan do primary groups. 0 Peracetic
,/'..,/
