No abstract

6-Carboline, together with 8-carboline, has been synthesized in reasonably good yield by the Fischer cyclization of cyclohexanone 3-pyridylhydrazone followed by dehydrogenation of the separated tetrahydro derivatives. The isomer ratios of products formed in this and similar reactions are discussed. Nitration of 6-carboline gives a mixture of the 6-and 8-nitro derivatives, the latter being the predominant product. The magnitude of the dipole moment of 6-carboline is appreciably higher than that predicted from n~olecular orbital calculations (15).The first unaillbiguous synthesis of 6-carboline'"1) was described in Part VIII of this series (I), in which the cyclizatioil of 3-azido-2-phenylpyridine was reported to give variable )~ields of (I). In order to study the substitution reactioils of this ring system a synthesis giving inore reproducible yields was required. An alternative method of obtaining the 6-carboline ring system was sought and found in the Fischer cyclization of cyclohexailone 3-pyridylhydrazone (11).The Fischer cyclization of pyridylhydrazones has been found by a nuinber of worlters to talte place with greater difficulty than the cyclization of the correspondiilg phenylhydrazones. Perkin and Robinsoil were unable to effect the cyclization of acetone 2-quinolylhydrazone (2), a i d Fargher and Furness (3) and Oltuda and Robisoil (4) obtained similar negative results using a number of 2-pyridylhydrazones. Cleino and Holt prepared 1-methyl-5,6,7,8-tetrahydro-p-carbolii in low ~~i e l d from cyclohexailoile 2-n1ethy1-3-pyridylh)~drazone and zinc chloride (5j. A moderate yield of 5,6,7,8-tetrahydro-a-carboline was obtained by heating cyclohexanone 2-pyridylhydrazone with polyphosphoric acid (4). Ficlten and Kendall have recently examined the cyclizatioil of the pyridylhydrazones of isopropylinethyl ltetone to give the corresponding azaindole derivatives (6). They found that heating with zinc chloride to a relatively high temperature (210-250") was necessary to effect ring closure. I t is interesting to note, as well, that isopropylmetl~yl ltetone 3-pyridylhydrazone was reported by these authors to give exclusively 2,3,3-trimethyl-313-4-azaindole; i.e. cyclization had occurred a t the a-position of the pyridine ring only. Similarly, Taltahashi, Sailtachi, Goto, and Shiinamura (7) obtained ~-chloro-2-inethyl-4-azaindole from acetone 2-chloro-5-pyridylhyclrazone.3-Hydrazinop3~ridine reacted with cyclohexanone to give the required cyclohexanone 3-pyridylhydrazone (11), which was somewhat unstable in air but could be kept for long periods in a closed container in the refrigerator. Cyclization was effected by fusion of (11) with zinc chloride a t 250" and gave a high yield (94%) of a mixture of 6,7,8,9-tetrahydro-6-carboline (111) (63%) and 5,6,7,8-tetrahydro-P-carboline (IV) (37%). The structures of (111) and (IV) were established by dehydrogenation to the corresponding fully aromatic structures. The dehydrogenations were carried out by heating the coinpound with 10% palladiuin-charcoal in boiling me...

(111). 'The evidence for the structure o f this compouncl is discussed. Similarly, heating 2-0-nitrophenylpyridi~~e methiodide and iV-oxide with f e r r o~~s oxalate gives (111) -o-azidophenylpyridine-iV-oxide gives a mixture o f 6-carboline and 6-carboline-py-A'oside in low yield. T h e mechanism o f the cyclization o f the azides and o f the reaction taking place on heating nitro-compounds with ferrous oxalate is discussed briefly; the formation o f a nitrene intermediate is favored. T h e catalytic reduction of 2-o-nitrophenylpyricline-iV-oxide giving rise t o the azoxy-, azo-, and hydrazo-derivatives is described and the ultraviolet absorption spectra o f thesc cornpounds are discussed. It is concluded the stel-ic inhibition o f coplanarity exists in the azosyand azo-compounds leading t o the lack o f elfective conjugation across the N=N bond.When this worlc was initiated no suitable nlethod for the synthesis of the 6-carboline ring system was available. Since that time, however, 6-carboline (I) has been obtaincd in this laboratory by the thermal cyclization of 3-azido-2-pl~engrlpyridiile (1). The yields in this reaction are variable, however, and a better method was sought. T o this end, the action of ferrous oxalate on 2-o-nitrophenylpyridine was investigated.Ferrous osalate has been used in a n~iinber of interesting cyclizations. Thus, heating it with 2-nitrodiphenyl gives rise to carbazole (2) ; phenazincs have also beell obtained from 2-nitrodipl~enylainiiles by this method (2, 3). On the other hand, 4-nitrocarbazole was not obtained froill 2,2'-dinitrodiphei~yl, b u t instead 3,4-bei1zocin1iolii~e was isolated (2). T h e action of ferrous osalate on 2-o-nitrophenylpyridillc (11) a t 300° has now been exanzined. I t was anticipated that the pyridine nuclear carbons might be relatively deactivated towards such a cyclization, b u t t h a t it could coilceivably occur a t the 3-position of the pyridine ring to give 6-carboline. Cyclization did, in fact, occur but onto the pyridille nitrogen atom to give pyriclo[l,2-blindazole (I I I ) , no 6-carboline being formed.

Hydroxy-des-N-methyl-a-obscurine, C I~H~~O ?~\ T ? , has been isolated froin L . fEabelliforme and a structure assigned to it on the basis of physical evidence. A s t r i l c t~~r e is assigned t o Lycopodium allraloid L.5, which we have named flabellidine, on the basis of physical evidence and its chemical derivation from lycodine. The mass spectra of hydroxy-des-N-methy1-aobscurine, flabellidine, and flabelline are discussed.The alltaloids of L. flabelliforme were first investigated by i\/Ianslte and Marion ( I ) , who isolated lycopodine (the major alkaloid), nicotine, and six bases, which they designated by number only. Alkaloid L.1 is now ltnown t o be dihydrolycopodine (2), L.2 t o be acetyldihydrolycopodine (3), and L.6 t o be a mixture of a-and P-obscurine (4). The structures of lycopodine and its derivatives (2) and of a-and p-obscurine (5) are now known. The structure of alkaloid L.5, t o which we have assigned the trivial name flabellidine, is discussed in this paper, leaving alkaloids L.3 and L.4 of the original six still uninvestigated.By using chromatographic and countercurrent distribution methods, an additional seven alltaloids have been isolated from L.flabell.iforme in this laboratory. Thus Young (6) has reported the isolation of two new alltaloids, flabelliformine and flabelline, t o each of which a structure has been assigned (7,8). Alam (9) has found four alkaloids of ltnown structure (clavolonine, lycodine, annotinine, and des-N-methyl-a-obscurine) and a new base hydroxy-des-N-methyl-a-obscurine, the structure of which is discussed in this paper.The structure assigned t o hydroxy-des-N-methyl-a-obscurine I rests on physical evidence as insufficient material was available t o investigate its chemistry.The presence of a hydroxyl group in I is inferred from its infrared spectrum, which, in Nujol and in chloroform, shows a pealt a t 3510 cm-l. The position of this peak in chloroform solution is not shifted upon dilution, thereby indicating that the hydroxyl group is intrainolecularly hydrogen bonded. Aside from the pealt in the hydroxyl region, the spectrum of this compound is remarkably similar t o that of a-obscurine (11) and des-Nmethyl-a-obscurine (111). The spectra of the three are reproduced in Fig. 1 for the region 3600-1500 cm-I.The ultraviolet spectrum of I (l:f:;OH 255 mp, log E = 3.78) is almost identical with that of I1 (~::i~~ 255 mp, log E = 3.75) and I11 (l:f:{OH 255 mp, log E = 3.81). From the spectral evidence i t appears that I must have the same chromophoric system as I1 and 111.The mass spectrum of the hydroxy compound shows a molecular weight of 276, corresponding t o the molecular formula C16H2402b?2. This formula differs in colnposition from I11 (C16H240N2) by one oxygen atom, suggesting t h a t the new base may be a simple hydroxy derivative of the latter. T h e fragmentation pattern of I , discussed later, is in agreement with this proposition.The nuclear magnetic resonance (n.m.r.) spectrum of I shows a three proton doublet For personal use only.

I t may be concluded that all three coupling constants in the heterocyclic ring of quinoline have the same sign, in agreement with the results of Rao and Baldeschwieler (4) for 2-amino-3-picoline. E X P E R I M E S T A LThe n.m.r. spectra were obtained with a Varian high-resolution spectrometer operating a t 100 iCIc 1s. The spectra were calibrated by the usual side-band method (7), using a Hewlett-Paclrard AIodel 20OAB audiooscillator and a Hewlett-Paclcard Model 521C frequency counter. The double-resonance experiments were carried out as described by Johnson (8). With the center-band sigrial phased out, the first modulation side band H I of the V3521 Integrator is used to record the n.m.r. spectrum while a second, stronger side band, Hz, provided bq an audio-oscillator, is dpplied to the system uilder investigation such that lo?-oil is equal to the separation necessary to do the double-resonance experiment. The output of the oscillator is quite critical (3). The ratios of the chemical shifts to the coupling constants in the present instances n ere sufficiently large that corrections to the modulation frequency were unnecessary (9). The frequency of the first tnodulation side band of the Integrator varied slightly from day to day and \+as checked before each experiment.As a check oil the method, the results reported by Freeman (3) for t~ans-crotoilaldehyde \\ere repeated and duplicated. _ Quinoline (synthetic grade) was obtained from Matheson, Coleman, and Bell and degassed by the nitrogeil bubbling technique prior to the n.m.r. measuren~ents. ACKSOWLEDGMEKTThe financial support of the National Research Council is gratefully acknowledged. SELF-CONDENSATION OF METHYL PROPIOLATEWhen during the course of a study of allialoid synthesis methyl propiolate was observed to undergo a chemical change in the presence of a tertiary amine, this unusual reaction was investigated. Exposure of a benzene solution of the ester to triethylamine, N-methylpyrrolidine, or N-methylpiperidine for a short period of time gave good yields of a diiner. Elemental, infrared, ultraviolet, and proton magnetic resonance analysis of the product as well as its conversion to dimethyl adipate on catalytic hydrogenation revealed the compound to be dimethyl trans-but-1-ene-3-ine-1,4-dicarboxylate (I) (1, 2). In some reactions a side product was formed. Its analyses as well as its transformatioil to dimethyl P-oxy-bis-propionate (3) on catalytic hydrogenation showed it to be dimethyl trans-poxy-bis-acrllate (11). Since this product reflected intervention of water in the reaction, its yield \\-as expected, and then shown to increase on deliberate addition of water to the reaction medium. (6) processes responsible for the formation of I have known precedents, the present case represents an interesting example of a combination of these reactions of the same propiolyl system and appears to be a high-yielding variation of the dimerization of acetylenes to en-ines under the influence of cuprous salts and acid (2, 7). Unfortunately there seems to be a sub...