2596
COMMUNICATIONS TO THE EDITOR
VOl. 81
on the hydroxydimethylcyclohexanone obtained exothermic reaction of 3-bromopropylamine hyby alkaline degradation of streptovitacin A drobromide with 1-(1-cyclohexeny1)-piperidine in ( C I S H ~ ~ N suggested O~) the 3-position for the hy- dimethylformamide. After dilution with water droxyl group,’ (i) Wolff-Kishner reduction of the and basification, the product is extracted with hydroxyketone to 1,3-dimethylcyclohexanol,and ether, dried and distilled. The infrared absorption (ii) resistance of streptovitacin X to oxidation by and the melting point oi its picrate salt are identical periodate, demonstrated that the hydroxyketone with those of a sample prepared by an earlier route‘ is 4-hydroxy-2,4-dimethylcyclohexanone. Strepto- involving three steps with an over-all yield of 44%. vitacin A therefore is 3- [a-(5-hydroxy-3,3-dimethyl- The alkylation of enamines with alkyl halides 2 - oxocyclohexyl) - 2 - hydroxyethyl] - glutarimide is well known.2 However, N,N-diethyl-3-bromo(14.9 propylamine hydrobromide does not react with Alkaline degradation of streptovitacin B, an 1- (1-cyclohexenyl)-piperidine under these condiisomeric ring-hydroxylated cycloheximide, gave tions to produce isolable amounts of 2-(3-diethylpredominantly 2,4 - dimethyl - 2 - cycl~hexenone,~aminopropy1)-cyclohexanone, indicating that the thereby suggesting the 3-hpdroxy-2,4-dimethyl- reaction of 3-bromopropylamine hydrobromide is cyclohexanone structure for the hydroxyketone probably not a simple alkylation of the enamine moiety of streptovitacin B. Also, while acetyla- followed by ring closure. It is postulated that the tion of streptovitacin A under mild conditions reaction occurs through loss of a proton from the 3gave a monoacetate, m.p. 16E~-168~ (calcd for one bromopropylanimonium ion to the more strongly acetyl, 12.7; found, acetyl, 11.5) similar treatment basic enamine, addition of 3-bromopropylamine to of streptovitacin B gave a diacetate, m.p. 155-158’ the resulting enamine salt, and elimination of (calcd. for two acetyls, 22.6; found, acetyl, 21.6). piperidinium ion to forni N -cyclohexylidene-3We therefore propose the structure of strepto- bromopropylamine. This imine is in equilibrium vitacin B to be 3- [2-(4-hydroxy-3,5-dimethyl-2- with its enamine form, and, a s the enamine, it oxocyclohexyl)-2-hydroxyethyl]-glutarimide (Ib). cyclizes to A1(9)-octahydroquinoline hydrobromide. Elemental analyses of streptovitacins Cz, 1n.p. 91-96’ (C, 60.53; H I 7.98; N, 4.67; 0, 27.71; C-CH3, 8.2) and D, n1.p. 6’7-69’ (C, 60.42; H, 7.89; N, 4.83; 0, 27.00; C-CHa, 9.0) showed these compounds also to have the empirical formula C16H23N06 (calcd. C, 60.58; H, 7.69; N, 4.71; 0 , 26.91; C-CH3, 10.1). Acid-catalyzed dehydration of streptovitacin C2 gave the phenolic derivative (11),while alkaline degradation provided a hydroxyketone which appeared from infrared spectra to be similar but not identical to that obtained from streptovitacin A. Periodate oxidation indicated one mole of oxidant consumed per mole of CZ. From these facts, we H propose the structure of streptovitacin CZ to be (CH,)0Br KH (CH2)3Br I 3-[2-(3-hydroxy-3,5-dimethyl-2-oxocyclohexyl) -2hydroxyethyll-glutarimide(IC). Acid degradation of streptovitacin D gave the phenolic derivative (11) showing this compound The conversion of an enamine to an imine under also to be a ring-hydroxylated cycloheximide. these conditions is illustrated by warming equivaAt the present time the position pf the hydroxyl lent amounts of 1-(1-cyclohexenyl)-piperidine and group in streptovitacin D is not knoTrn. butylamine hydrobromide in dimethylformamide; The author wishes to thank Mr. W. -1.Struck dilution with ether precipitates an S4Y0 yield of and associates for microanalyses, l l r . AI. F. piperidine hydrobromide. Evaporation of the filGrostic for infrared spectra, Dr. G. Slomp for trate, and then distillation, produces a moderate nuclear magnetic resonance measurements, and amount of N-cyclohexylidenebutylamine. That Dr. E. E. van Tamelen for helpful discussions and these imines exist partially in the enamine form is suggestions. suggested by the fact that N-cyclohexylidene(9) Th14 qtructure is the same as t h a t shown by K V Rao for a ethylamine and -isopropylamine react similarly degradation product “desacetrl E-71” (Abstract, 134th Am Chem with 3-bromopropylamine hydrobromide to proSOCMeeting Chxago, Sept 1‘1~8) The stereochemical relationship duce A1(g)-octahydroquinolinein 7 5 4 2 % yields. of the two compounds is n o t known This equilibrium also is suggested by the failure of RESEARCH LABORATORIES THEUPJOHNCOMPANY Ross R. HERR 3-bromopropylamine hydrobromide to react with KALAMAZOO, MICHIGAY 1-(1-cyclopenteny1)-piperidinet o produce 2,3,4,RECEIVED MARCH27, 1959 4a,6,7-hexahydro-5H-1-pyrindine, due to a lesser tendency of the imine double bond to migrate into the five-membered ring. However, N-methyl-3A NEW REACTION OF ENAMINES bromopropylamine hydrobromide reacts with 1-
Sir:
L. A. Cohen and B. Witkop, THIS JOURNAL, 77, 6595 (1955). It has been found that A1(9)-octahydroquinoline (1) (2) J. N. Collie, Ann., 446,316 (1884); R. Robinson, J . Chem. Soc., (2,3,4,4a,5,6,7,8-octahydroquinoline) may be pre- 109, 1038 (1916); G. Stork, R. Terrell and J. Szmuszkovicz, THIS pared in 8@-84% yields in one step by the JOURNAL, 76, 2029 (1954).
May 20, 1959
COMMUNICATIONS TO THE EDITOR
2597
(1-cyclopenteny1)-piperidine to produce a 74% dichloride and cleavage of the B-C bond with hyyield of mixed 2,3,4,5,6,7- and 2,3,4,4a,5,6,- drogen peroxide gave p- and m-cresol in a 3 :2 molar hexahydro-1-methyl-1H-1-pyrindine,b.p. 65-66' ratio (infrared determination) ; no o-cresol was (6 mm.), confirmed by analysis (calcd. for CgH15N: present in detectable quantities. When prepared C, 78.77; H, 11.02. Found: C, 78.71; H, 11.19) a t 35") the ratio of pura to meta isomers in tolyland by infrared absorption of the free base and of dichloroborane was about 4.6:l. The aryldithe perchlorate salt, m.p. 215-217'. Similarly, chloroboranes formed from the isomeric xylenes N-methyl-3-bromopropylamine hydrobromide re- a t 35" were: meta, 3,5-xylyl- with trace amounts of acts with 1-(1-cyclohexeny1)-piperidineto produce a 2,5-xylyl-; ortho, 3,4-xylyl-; and para, 2,5-xylyl-. 73% yield of mixed 1,2,3,4,5,6,7,8- and 1,2,3,4,4a, Mesitylene at 140O gave largely 2,5-xylyldichloro5,G,7-octahydro-1-methylquinoline, b.p. 82-84 borane (-15% yield) with a small amount of the (6 mm.), confirmed by infrared studies and by 2,4- isomer. At 35O, mesitylene reacted to form catalytic reduction to cis-l-methyldecahydroquino- traces of mesityldichloroborane. Durene was inline (picrate salt, m.p. 199-201 O, reported m.p. active a t 140'. At 30") naphthalene and also 199-200°,3anal. calcd. for C ~ O H I ~ N . ( N O Z ) $: ~ ~biphenyl H ~ ~ H appeared to give more than one type of C,50.25; H,5.80. Found: C, 50.35; H,5.71). arylboron derivative. Further studies are under way to determine the The distribution of isomers in this synthesis of scope of this reaction. arylboranes is comparable to that in FriedelCraft reactions that are run in the presence of alu(3) M. Ehrenstein and W. Bunge, Be*., 67, 1715 (1934). minum chloride? Accordingly, it is suggested that P A R K E , DAVISAND COMPANY DETROIT32, MICHIGAN ROBERT F. PARCELLthe active species in this synthesis may be BCIz+ RECEIVED FEBRUARY 16, 1959 or ArH.BCl2+. Such species would be stabilized by the formation of the A1C14- anion, and reaction of the cation complex with the active aluminum SYNTHESIS OF ARYLDICHLOROBORANES surface would then produce the ArBCI2 compound. Sir: The aspect of reaction mechanism is being investiWe have found a simple and direct synthesis of gated. aryldichloroboranes from boron trichloride and (5) Alumimm chloride is a co-product in this synthesis of arylboron aromatic hydrocarbons. chlorides. These halides were prepared by charging a stain- CONTRIBUTION No. 530 FROM THE less steel-lined pressure vessel (400 ml. internal CENTRAL RESEARCH DEPARTMEST, EXPERIMENTAL STATION capacity) with 100-125 g. of aromatic hydrocarbon, E. I. DU PONTDE NEMOURS AXD COMPASY, DELAWARE E . L. MUETTERTIES 2-30 g. of aluminum powder, 0.1 g. of aluminum WILMINGTON, RECEIVED FEBRUARY 17, 1959 chloride, iodine, or methyl iodide, and 60 g. of boron trichloride2 and heating the vessel, under autogenous pressure and with agitation to 120- O N THE MECHANISM OF FATTY ACID SYNTHESIS' 150" for 5-60 min. or to 30-50°3 for 24-48 hours. Sir : The product, usually a liquid slurry, was filtered that the first step Recently we have and the filtrate was distilled. In the case of in fatty acid synthesis is the carboxylation of benzene, the conversion to purified C8H5BCl2 acetyl c o i l to malonyl CoA catalyzed by the biotin ranged from 60 to 72%, b.p. 95" (48 mm.). Anal. containing Rlgo fraction in the presence of ATP Calcd. for Ca&BC12: C, 45.38; H, 3.17; B, 6.79; and Mn++. C1, 44.66. Found: C, 45.87; H, 3.54; B, 7.39; Malonyl Co-4 readily is converted to palmitate C1, 44.31. Variations in the C,H, to BC1, ratio in the presence of R2g3c and TPNH.2 This conversion had no significant effect on the nature of the prod- can be followed spectrophotometrically or isoucts, and there was no evidence for the formation of topically. The addition of acetyl Co-4 will signifi(C6&)2BC1 or C&(BC&)z. Polysubstitution in cantly increase the rate and extent of synthesis of the aromatic nucleus undoubtedly is unfavorable palmitate. Furthermore, a significant amount of because the strongly electronegative BCl2 group C14-acetyl CoA is incorporated into palmitate when deactivates the ring..' unlabeled malonyl CoA is used (Table I). The Tolyldichloroborane was obtained from toluene amount of label introduced into palmitate corin 60% conversions a t 140'. Hydrolysis of the responds to about one eighth of the total amount of (1) The classical methods for the preparation of aryldihaloboranes "CZ units" converted to palmitate (measured by have been discussed in a recent review by hl. F. Lappert, Chem. Rev.., TPNH oxidation). Unlabeled acetaldehyde does 66, 1049 (1956). E. Pace (Atli A c c a d . Lincei, 10, 193 (1929)) reported the synthesis of C6HaBCIz from benzene and boron trichloride over not reduce the amount of CI4-acetyl CoA incorpalladium biack a t 5OD-1500~. The Pace synthesis has been studied porated into palmitate and acetaldehyde is not by W. L. Ruigh, el nl. (WADC Technical Report 55-26, Parts 111-IV formed by the enzymic reduction of acetyl CoA (1956), P.B. Nos. 121,374 and 121,718. U. S. Dept. of Commerce, by TPNH.4 Not only acetyl CoA but also C14Washington, D. (2.). They found that, with charcoal-supported palladium catalysts, the yields were variable, probably due to sensitivity butyryl coL4 and C14-octanoyl CoA can be inof the catalyst t o poisons. corporated into palmitate in presence of malonyl (2) Boron tribromide and triiodide also proved operable, but boron CoA. trifluoride did not react under these conditions. (3) In one experiment with benzene, an exothermic reaction set in a t 3' and the internal temperature flashed t o 120'. A 6G'30 conversion to CsHbBCIz was obtained. Variations in reaction rate are attributed to variations in the activity of the aluminum surface. (4) The BClr group should be a t least as effective as CI in deactivating the ring, and chlorobenzene itself was found to be inert to the BClrAl reagents a t 50°.
(1) This work was supported in part by several grants: 20-88 and RG-5873. Division of Research Grants ( N I H ) ; H-2236(C3), National Heart Institute ( N I H ) ; G-3227, National Science Foundation; and AEC Contract AT(l1-1)-64, Project 4. (2) S. J Wakil, TEISJOURNAL, 80, 6465 (1958). (3) S. J. Wakil and J. Ganguly, fled. Proc., 18, 346 (1959). (4) R. 0. Brady, P?oc.N o t . A c a d . Sci., 44, 993 (1958).
