The Reaction of Hypochlorous Acid with Some Quinones and Related

May 18, 1985 - react with hypochlorite ion1 to give the cor- responding epoxides. The epoxidation of these qui- ñones by means of hypochlorous acidha...
0 downloads 0 Views 3MB Size
Notes as well as acetone to which t-butyl hydroperoxide was added, resulted in the formation of I1 in yields ranging from 40 to 60%. It was apparent that in the presence of peroxides SOLOMON MARMOR the reaction of hypochlorous acid with quinones proChemistry Department, New Mexico Highlands University, ceeds via a free-radical mechanism. Precedence for Las Vegas, New Mexico the free-radical attack on a quinone with ultimate replacement of hydrogen a t the 2-position exists in the Received May 18, 1986 work of Fieser, et al. 2,3-Dimethyl-l,4-naphthoquinone was prepared from the monomethylquinone through l14-Naphthoquinone and 2-methyl-l14-naphthoqui- reaction with lead tetraacetate' or with acetyl peroxide.6 none react with hypochlorite ion1 to give the corThe reaction of 2-methyl-l14-naphthoquinonewith responding epoxides. The epoxidation of these quilead tetrapropionate to give 3-ethyl-2-methyl-l14nones by means of hypochlorous acid has been cited12 naphthoquinone was also r e p ~ r t e d . ~ The interaction but these claims were considered somewhat dubious of hypochlorous acid with toluene, under the influence in light of the established mechanism3 for the nucleoof light, to give benzyl chloride as the principal product philic epoxidation of a$-unsaturated carbonyl comis the only reported instances of a possible free-radical pounds. An investigation of the reaction of aqueous reaction of hypochlorous acid with an organic comhypochlorous acid (free of chlorine) with quinones, pound. The possibility that the hypochlorous acid described in this paper, revealed that the epoxide is simply decomposes to free chlorine, which is then innot a significant product, and that an unexpected revolved in the reaction with the quinone, is untenable. action does occur under appropriate conditions. Elemental chlorine does react with l14-naphthoquiStirring finely powdered l14-naphthoquinone (I) none7 to form the expected stable addition product, with aqueous hypochlorous acid for 20 hr. a t 25" 2,3-dichloro-2,3-dihydro-l,4-diketonaphthalene. Dehyresulted in no reaction, and the organic starting drochlorination of the latter occurs when it is treated material was recovered unchanged. The next logical with base, resulting in the formation of the monostep called for a reaction run under homogeneous chloroquinone. conditions. The major product isolated from the A preliminary experiment to determine the effectivereaction of I with aqueous hypochlorous acid in diness of an in situ reaction of HOCl with l14-naphthooxane proved to be 2-chloro-l,4-naphthoquinone(11), quinone in aqueous dioxane gave promising results. obtained in 60-i'070 yield after recrystallization. Chlorine was slowly bubbled into an aqueous dioxane was converted Similarly, 2-methy1-ll4-naphthoquinone solution of the quinone and a mixture of products was to 2-methyl-3-chloro-l,4-naphthoquinone(56% yield) , ultimately obtained. Through processes of column and 3-chloro-2,5-diphenylbenzoquinonewas obtained chromatography and fractional crystallization 2-chloroin excellent yield from the reaction of hypochlorous l14-naphthoquinone (42y0 of the total product) and acid with the quinone in dioxane. 2,3-dichloro-l,4-naphthoquinone (50.2%) were isolated. The dioxane used as the solvent was not specially A study of the reaction of hypochlorous acid with purified, except for a simple distillation. The results ,4-dione in dicis- and trans-1,4-diphenyl-2-butene-l of the experiments indicated that the presence of peroxane containing peroxides was undertaken to deteroxides in the solvent could be responsible for the obmine the application of the reaction to compounds strucserved substitution reactions. This was borne out turally similar to the quinones. The only product by further experiments in which other solvents were isolated from the reaction of the trans isomer was mesoused in place of the peroxide-containing dioxane. 2,3-dichloro-l,4-diphenyl-l ,4-butanedione (53.8% after The reaction of chlorine-free hypochlorous acid with recrystallization), The direct addition of chlorine to I in acetone or in peroxide-free dioxane under a nitrogen the trans-unsaturated dione was described by Conant atmosphere and in the dark led to the formation of and Lutz,8 who obtained only the meso addition prodnone or only small amounts (loyoor less) of the 2uct. The cis isomer reacted with HOCl in dioxane t o chloro derivative. Furthermore, reactions of HOCl give a mixture of dl-2,3-dichloro-l,Cdipheny1-1,4with I in other solvents known to contain peroxides, butanedione (68.23y0), the meso-dichloro compound such as tetrahydrofuran and isopropyl ether (used ,4dione (16.7%) , and trans-l,4-diphenyl-2-butene-l in combination with acetone to assure homogeneity), (15.0~0),which was separated by silica gel chromatography. The direct addition of chlorine to the (1) 9.Marmor, J. Ow. Chem., 28, 250 (1983). (2) American Cyanamid Co., Organic Chemicals Division, Technical cis-unsaturated dione in chloroform a t - 20" resulted The Reaction of Hypochlorous Acid with Some Quinones and Related Compounds

Bulletin, 1,4-Naphthoquinone, p. 13, cites a reaction of the quinone with 4% HOCl to give the epoxide. The literature sources referred to were T. Zincke and Wiegand, Ann. Chem., 286, 71 (1895); T. Zincke, Chem. Ber., 26, 3802 (1892); A. Mandinaveitia, et ai., Anales soc. espan. $8. quim., 27, 647 (1929); Chem. Abefr., 24, 359 (1930). These references actually described the reaction of aqueous calcium hypochlorite solutions, whose concentrations were equivalent to 4% HOCI. The abstract referred to contains the error of specifying HOCl as the reactant. (3) C. Bunton and G. Minkoff, J. Chem. Soc., 885 (1949).

3556

(4) L. F. Fieser and F. C. Chang, J. A m . Chem. Soc., 64, 2043 (1942). (5) L. F. Fieaer and A. E. Oxford, ibid., 64,2060 (1942). (6) B. F. Clark, Jr., Chem. News, 148,265 (1931); Chem. Abet?., 26, 1591 (1932). (7) L. F. Fieser, J . A m . Chem. Soc., 7 0 , 3170 (1948). (8) (a) J. B. Conant and R. E. Lutr, ibid., 4'7, 881 (1925); (b) R.E.L u t r , ibid., 49, 1108 (1927).

OCTOBER1965

NOTES

3557

of the phenyl hydrogens bands to quinoid hydrogen band is 9.8: 1.2, in good agreement with the theoretical value of 10: 1. Reaction of 1,4-Naphthoquinone with Chlorine in Aqueous Dioxane.-Chlorine ( 5 g., 0.07 mole) was slowly bubbled into a solution of 2.0 g. of 1,4-naphthoquinone in 150 ml. of dioxane Experimental Section (containing peroxides) and 10 ml. of water. The temperature All melting points are uncorrected. was maintained a t 15-20' during the addition of the chlorine, Materials.-Hypochlorous acid was prepared by the reaction which required 1 hr. The reaction mixture was stirred at 15" of chlorine with an aqueous suspension of mercuric oxide, followed for 30 min. more. The yellow solution was transferred to a by distillation under reduced p r e s s ~ r e . ~The resulting approxibeaker and allowed to evaporate in a hood overnight. The brown mately 1 M solutions were analyzed1° and found to be free of solid that separated was filtered, washed with water, and air chlorine. Solutions of HOCl were analyzed prior to each use dried. The crude product was recrystallized twice from ethanol to ascertain the absence of free chlorine. All other reagents giving 1 g. of light tan crystals, melting range 112-170". A were obtained from standard sources. portion (0.207 g.) of the mixture was chromatographed on a silReaction of Hypochlorous Acid with Quinones. General ica gel column, using benzene-cyclohexane (60:40 v./v.) as first Procedure.-To a solution of 0.01 mole of the quinone in 80 ml. eluting solvent. The bright yellow solid (0.192 8.) obtained of dioxane (containingperoxides) was added a solution of chlorinefrom the eluate was recrystallized twice from ethanol to give a free hypochlorous acid (0.04 mole)" over a period of 15 min. product, m.p. 190-192', whose infrared spectrum was the same The temperature of the stirred reaction mixture was maintained as that of an authentic sample of 2,3-dichloro-l,4-naphthoquinone a t 20-25' by means of an ice-water bath applied as needed. The (Aldrich Chemical Co.); no depression was observed in a mixture reaction mixture was then stirred at room temperature until melting point determination. analysis of 1-ml. samples of HOCl indicated that the reaction The mother liquors from the recrystallizations were combined was complete (usually about 2 hr.). and evaporated to dryness. The residual bright yellow solid The solution was evaporated to a t least half its original volume (0.087 9.) was recrystallized once from ethanol. The identity and then saturated with water. The solid that separated was of the product, m.p. 115-116", as 2-chloro-l,4naphthoquinone filtered and the filter cake was washed thoroughly with water. was verified from its infrared spectrum. After air drying, the product was recrystallized from a suitable Reaction of trans-l,4-Diphenyl-Z-butene-1,4-dione with Hyposolvent. chlorous Acid in Dioxane.-To a stirred solution of 1.0 g. of I n a typical run 2-chloro-1,4-naphthoquinone, m.p. 115trans-l,4diphenyl-2-butene-l ,4-dione (Eastman, recrystallized 115.5', was obtained in 63.8% yield, after recrystallization from from ethanol) in 50 ml. of dioxane (containing peroxides) was ethanol. The infrared spectrum of this product was identical added 25 ml. of 1.6 M HOCl over a period of 5 min. The temwith that of an authentic sample prepared according to Fieser'; perature of the reaction mixture was maintained a t 25' during the no depression was observed in a mixture melting point deaddition and for 4 hr. thereafter. The mixture was evaporated termination. A 56y0 yield of 3-chloro-2-methyl-l,4-naphtho- to half its volume and the white solid that separated was filtered. quinone (recrystallized from acetic acid-water, m.p. 152-153') Recrystallization of the crude product (1.3 9.) from acetone was obtained. The infrared spectra of this product and an afforded 0.7 g. (53.8%) of pure meso-1,4-diphenyl-2,3-dichloroauthentic sample7 were identical. l,Cbutanedione, m.p. 166-167" (lit.8m.p.167'). I n general, reactions of 1,4-naphthoquinone and hypochlorous Reaction of cis-l,4-Diphenyl-2-butene-1,4-dione with Hypoacid in other solvents were carried out under the same conditions chlorous Acid in Dioxane.-To a solution of 4.7 g. (0.02 mole) of as described above. The reaction in peroxide-free dioxane12 cis-l,4-diphenyl-2-butene-l ,4-dionela in 200 ml. of dioxane (conwas run in a nitrogen atmosphere. taining peroxides) was added 25 ml. of 1.6 M HOCl over a period I n those cases where only small amounts of 2-chloro-1,4of 10 min. The temperature was maintained a t 25-30' during naphthoquinone were formed, the following column chromathe addition and for 30 min. afterwards. Another 5 ml. of hypotography procedure was used to separate the product from unchlorous acid was then added and the mixture was stirred for reacted naphthoquinone. About 0.5 g. of the crude material 2.5 hr. more. The reaction mixture, which had become yellow, was deposited on a silica gel ( J . T. Baker 3405) column and then was evaporated to half its volume, and 50 ml. of water was eluted with benzene-cyclohexane (60:40 v./v.). A distinct added, causing the separation of a yellow oil. The organic separation of yellow 6ands appeared on the column. When the phase was extracted with 100 ml. of ethyl ether and the ether eluate became colorless, 100% benzene was used to elute the solution was washed once with 100 ml. of water. After drying next fraction. The chloroquinone was contained in the first over anhydrous magnesium sulfate, the ether solution was fraction, while the unreacted starting material was found in the filtered and the solvent was stripped on a rotating evaporator. second. A small amount of dark red material remained adThe residual yellow oil was chilled in an ice-water bath, wheresorbed a t the top of the column, but could be removed by eluupon it solidified. The solid was recrystallized once from ethtion with methanol. The identity of the latter material was not anol, but the product had a melting range of 83-140". determined, but appeared to be a mixture of several substances. The recrystallized solid and mother liquor residue were recomReaction of Z,5-Diphenyl-1 ,4-benzoquinone with Hypochlorous bined and the components of the mixture were separated by Acid.-Hypochlorous acid solution (25 ml., 1.6 M ) was added to column chromatography, using Baker silica gel as adsorbent. a stirred solution of 1.O g. of 2,5-diphenyl-l,4-benzoquinone The mixture (4.60 g.) was deposited on the column and then (0.0038 mole) in 150 ml. of dioxane containing peroxides. The eluted with 630 ml. of a benzene-hexane (25:75 v./v.) mixture. temperature was maintained a t 20-21' during the addition, The residue from the first eluates weighed 3.14 g. and melted which required 10 min. After the mixture was stirred for 1 hr. at 87-90'. After one recrystallization from ethanol, the prodat 20-25', another 15 ml. of HOCl solution was added. The uct, dl-2,3-dichloro-l,4-diphenyl-l,4-butanedione,melted a t mixture was stirred a t room temperature for 6 hr. and was 87-88'(lit.*m.p.86'). then evaporated to about one-fourth its original volume. Water The column was then eluted with 75y0 benzene-hexane (50 ml.) was added and the orange-yellow solid was filtered and (75:25 v./v.) to give 0.77 g. of colorless solid, m.p. 165-167'. washed thoroughly with water. The solid weighed 1.1 g. and The solid was recrystallized from acetone, and the purified melted a t 185-188'. The crude product was recrystallized product, m.p. 166-167', proved to be the meso-dichloro derivafrom acetone, and 0.95 g. (84.8% yield) of 3-chloro-2,5-diphenyltive, as determined by mixture melting point. Finally, elution 1,4-benzoquinone, m.p. 187-189', was obtained. of the column with 1OOyobenzene afforded 0.69 g. of yellow solid, Anal. Calcd. for C18Hllc1o2: C, 73.35; H, 3.76; C1, 12.03. which, after recrystallization from ethanol, melted at 1113-111'. Found: C,73.26; H,3.76; C1,11.96. A mixture melting point determination with an authentic sample The n.m.r. spectrum shows a multiplet at 7.21 p.p.m. (phenyl of the trans-unsaturated dione gave no depression. H), and a singlet a t 6.71 p.p.m. (quinoid H). The area ratio of

in the formation of a mixture of the d b and mesodichloro derivatives,8 with the former predominating.

(9) A. Chung and C. G.Israel, J . Chem. Soc., 2667 (1955). (10) "Scott's Standard Methods of Analysis." Vol. I, N. H. Furman, Ed., D. Van Nostrand Co., New York, N. Y . , 1939, p . 274. (11) When only small excesses of HOCl were used yields were considerably lower. (12) L. F. Fieser, "Experimenta in Organic Chemistry," 3rd Ed., D. C. Heath and Co., Boston, Mass., 1957, p. 284.

Acknowledgment.-The author is indebted to Dr. Audrey M. Small for the n.m.r. spectrum. Gifts of samples from American Cyanamid Company, Eastman Chemical Products, Inc., and Lucidol Division, Wallace and Tiernan, Inc., are gratefully acknowledged. (13) J. B. Conant and R. E. Lutz, J . Am. Chem. Soc., 46, 1304 (1923).

NOTES

3558

A Novel Condensation during the Low-Pressure Hydrogenation of 5-Hydroxyisoquinoline IANW. MATHISON Department of Pharmaceutical and Medicinal Chemistry, University of Tennessee College of Pharmacy, Memphis, Tennessee 58105 Received February 1, 1966

During the synthesis of some compounds designed to affect the cardiovascular system, it became necessary to examine hydrogenation procedures for 5hydroxyisoquinoline. Two series of compounds were envisaged : first, 1,2,3,4-tetrahydroisoquinoline analogs and, second, the completely reduced decahydroisoquinoline derivatives. Hydrogenation of the isoquinoline nucleus to the tetrahydro stage from 5hydroxy-2-ethylisoquinolinium bromide (VI) proceeded smoothly at low pressure using Adams platinum oxide as catalyst. It was conclusively shown that the hydrogenation of the salt (VI) proceeded more rapidly a t 50 p.s.i. over platinum oxide (100% of total uptake in 0.5 hr.) than reduction of the free base, 5-hydroxyisoquinoline (22% of total uptake in 0.5 hr.), under identical conditions. I n each case, the corresponding tetrahydro material was obtained from the reaction mixture in quantitative yield. This variation in the rate of hydrogenation is in agreement with the work of Lasslo and co-workers1who similarly found that various substituted pyridinium salts were hydrogenated more rapidly than the corresponding free bases, although emphasis was not placed on this feature in the accounts of their work. For the complete saturation of the substituted isoquinolines discussed in this paper, the method outlined by Witkop2 for the decahydrogenation of isoquinoline, per se, was chosen. It appeared to be the only direct method available, though several other indirect reductions to the decahydro stage have been rep~rted.~-ll Accordingly, 5-hydroxy-2-ethylisoquinolinium bromide (VI) was hydrogenated in glacial acetic acid containing a small amount of sulfuric acid using Adams platinum oxide catalyst. The reaction mixture yielded a bis(2-ethyldecahydroisoquinoline) ether (I); no other products were isolated by the above-described procedure. Compound I was identified by its infrared absorption; the ether peak was clearly identified at 1110 cm.-l and there was no indication of the presence of a free hydroxyl function. The ultraviolet spectrum of the new ether did not show any absorptions, indicating complete saturation at the decahydro level. The postulated formula was further confirmed by preparing the corresponding quaternary cliethiodide derivative (1) A. Lasslo, W. M. Marine, and P. D. Waller, J . Org. Chem., 21, 958 (1956). (2) B. Witkop, J . Am. Chem. Soc., 70,2617 (1948). (3) A. Marohant and A. R.Pinder, Chem. Ind. (London), 1366 (1953). (4) B. Witkop, J . Am. Chem. Soc., 71, 2559 (1949). (5) R. B. Woodward and W. E. Doering, ibid., 67, 860 (1945). (6) N.Sugimoto end S. Oshiro, Pharm. Bull. (Tokyo), 6, 316 (1957). (7) A. P.Gray and D. E. Heitmeier, J . Am. Chem. Soc., 60,6274 (1958). (8) E. Ochiai and M. Ikehara, Pharm. BUZZ. (Tokyo), 8, 454 (1955). (9) F. E. King and H. Booth, J . Chem. Soc., 3798 (1954). (10) K.Miyaki and H. Kataoka, J . Pharm. SOC.Japan, 59, 222 (1939). (11) T. Miyamoto and A. Kataoka, ibid., 59, 478 (1939).

VOL.30

(11); again spectrophotometric data was consistent with the proposed moiety. Considering the prevailing reaction conditions, the spectroscopic and all other confirmatory evidence, it becomes apparent that we are dealing with an ether of 5-hydroxy-2-ethyldecahydroisoquinoline;the electrons of the other bond of the oxygen atom are shared with one of the carbon atoms of the alicyclic ring component of the second isoquinoline moiety. While it is probable that C-5 of the second isoquinoline nucleus is involved in the ether linkage, there is no direct evidence on this point. WSJCzH5

WJ

0 /

cc

NCzHs

I

+

+

I- ( C Z H S ) Z N C ~ H ~ S O C ~ H , ~ N ( C ~ H S ) ~ I I1

No evidence has been found in the literature of this type of reductive condensation having been observed before. The formation of ethers under catalytically reductive conditions has been reported for some aliphatic and alicyclic ketones,12 and reports have been noted of the formation of ethers by the reduction of acetals in acidic media.18-15 One could not consider, however, the reduction of a heterocyclic phenol, wherein two independent molecules combine to yield an ether, comparable. Further evidence that I is formed by reductive condensation was obtained by subjecting 5-hydroxy-2ethyldecahydroisoquinoline (V) to the same acidic reaction conditions, including the presence of bromide ions (the hydrobromide salt of V was used), as those described for the preparation of I, but without the platinum oxide catalyst or hydrogen, for 48 hr. Careful isolation procedure yielded only the original starting material V. Two factors may have contributed to the formation of I during the hydrogenation: (a) the presence of the alkyl substituent on the nitrogen atom, and (b) the nature of the anionic group. Attempts to evaluate the relative importance of these factors prompted us to investigate this matter further. First, the hydrogenation of the free base, 5-hydroxyisoquinoline, under identical conditions with those outlined for the preparation of I (see Experimental Section) resulted in 5-hydroxydecahydroisoquinoline (IV), and second, the identical hydrogenation of 5-hydroxy-2-ethylisoquinolinium hydroxide (see Experimental Section) yielded 5-hydroxy-2-ethyldecahydroisoquinoline (V). The results obtained from these additional experiments suggest that it is indeed the anionic group which influences the course of the reaction observed in the formation of ether I. The ionic moieties involved are shown in Figure 1. Only with the N+C2H5Brmoiety was the anomalous ether formation observed. The insolubility of corresponding iodides and chlorides ( i e . , 5-hydroxy-2-ethylisoquinoliniumiodide and 5-hydroxyisoquinoline hydrochloride) in glacial (12) M.Verzele, M. Aoke and M. A n t e u h , J . Chem. Soc., 5598 (1963). (13) W. L. Howard and J. H. Brown, J . Org. Chem., 26, 1026 (1961). (14) F. Sigmund and G. Marohart, Monatsh., 46, 267 (1927). (15) F.Sigmund and R. Unohann, i b i d . , 61, 234 (1929).

NOTES

OCTOBER1965

3559

m

tracted with dilute hydrochloric acid. This acid extract, made alkaline with sodium hydroxide, was extracted with ether and dried. The ether was removed to yield 3.6 g. of a pale yellow oil. The hydrobromide salt of this base was recrystallized from during the hydrogenation of during the hydrogenation ethanol-ether t o yield 3.7 g. (32.3%) as short white needles, m.p. of 5-hydroxyisoquinoline 5-hydroxy-2-ethylisoquinolinium 213.2-213.4'. hydroxide Anal. Calcd. for C21H26BrN05:C, 55.75; H , 5.75; Br, 17.70; N, 3.10. Found: C, 55.59; H, 5.71; Br, 17.80; N, 3.18. -&CZH5Br5-Hydroxydecahydroisoqdnohe (IV).-5-Hydroxyisoquinoline (5 g.) was dissolved in 50 ml. of glacial acetic acid and 0.5 ml. of concentrated sulfuric acid was added. The resulting solution during the hydrogenation of 5-hydroxy-2-ethylisoquinolinium was hydrogenated over 5 g. of A d a m platinum oxide a t 50 p.s.i. bromide (VI) for 36 hr. a t room temperature. The exhausted catalyst was filtered from the hydrogenated solution, and the filtrate was Figure 1 diluted with approximately 50 ml. of water, made alkaline by the addition of sodium hydroxide pellets, and extracted with ether. acetic acid did not permit an immediate study of the The dried ethereal extract was evaporated carefully to yield a yellow oily residue (2.7 9.) which was treated with dry hydrogen effect of other halide ions upon the reaction taking place chloride in ether to yield 2.9 g. (51%) of the base hydrochloride, during this hydrogenation. Further discussion of recrystallized from ethanol-ether, m.p. 191.0-193.0". The this subject matter is reserved for a future occasion. ultraviolet spectrum of this compound was taken in water and showed no absorption throughout the range 220-340 mp. Anal. Calcd. for C9H&lNO: C, 56.25; H, 9.38; C1, 18.75; Experimental Section16 N,7.29. Found: C,56.33; H,9.43; C1,18.80; N, 7.22. 5-Hydroxy-2-ethyldecahydroisoquinoline (V) .-5-Hydroxy-25-Hydroxy-2-ethylisoquinoliniumBromide (VI) .-5-Hydroxyethylisoquinolinium bromide (VI) (8 g., 0.031 mole) was treated isoquinoline (40 g., 0.28 mole) was dissolved in 100 ml. of abwith moist silver oxide (13 g., prepared from the action of sodium solute ethanol and refluxed for 8 hr. on a steam bath with 50% hydroxide on silver nitrate) in 100 ml. of a 50% aqueous methanol excess ethyl bromide (40 g., 0.37 mole), The ethanol and excess solution for 24 hr. The silver bromide formed was then filtered ethyl bromide was then evaporated off and the resulting brown from the solution using Celite-charcoal mixture as a filter aid. solid was recrystallized from ethanol to yield 57.47 g. (82%) The solvent was evaporated under reduced pressure a t as low a of light brown needles, m.p. 209.4-210.6'. temperature as possible to yield 5.7 g. (96%) of 5-hypxy-2Anal. Calcd. for CllHlzBrNO: C, 51.97; H, 4.73; Br, ethylisoquinolinium hydroxide. The latter was hydrogenated as 31.50; N, 5.51. Found: C, 51.77; H , 4.68; Br, 31.50; N, described in the preparation of IV. The base was converted to 5.13. the hydrobromide, 5.62 g. (81%), which was recrystallized from 5-(3,4,5-Trimethoxybenzoyloxy)isoquinolie (VlI) .-3,4,5-Triethanol-ether to yield fine white needles, m.p. 218.4-219.0'. methoxybenzoyl chloride (13 g., 0.057 mole), prepared according The ultraviolet spectrum of this compound indicated no abto the method of Lasslo and Jordan," was dissolved in 200 ml. of sorption in the range 220-310 mp. sodium-dried benzene and 42 g. (0.29 mole) of 5-hydroxyisoAnal. Calcd. for CllHzzBrNO: C, Fj0.00; H, 8.33; Br, quinoline, suspended in dry benzene, was added. The reaction 30.31; N , 5.30. Found: C, 49.80; H, 8.31; Fir, 30.38; N , mixture was then refluxed for 8 hr. The solid was filtered off; 5.52. the benzene filtrate was evaporated to dryness; and the re5-(3,4,5-Trimethoxyoenzoyloxy)-2-ethyldecahydroisoquinoline sulting solid was recrystallized from ethanol to yield 15.5 g. Hydrobromide (X).-5-Hydroxy-2-ethyldecahydroisoquinoline (87%) as pale yellow prisms, m.p. 198.5-199.0'. (V, 5 g., 0.027 mole) was dissolved in 20 ml. of sodium-dried Anal. Calcd. for C18H1TNO5: C, 67.24; H, 5.02; N, 4.13. toluene and added to a solution of 15 g. (0.065 mole) of 3,4,5Found: C, 67.39, H, 5.12; N , 3.90. trimethoxybenzoyl chloride and refluxed for 48 hr. The re~-Hydroxy-2-ethyl-l,2,3,4-tetrahydroisoquinoheHydrobrosulting suspension was filtered, and the toluene filtrate was exmide (VIII) .-5-Hydroxy-2-ethylisoquinolinium bromide (VI) tracted with dilute hydrochloric acid. Neutralization of this (5 g., 0.0196 mole) was dissolved in 250 ml. of absolute ethanol acid extract was followed by extraction with ether. The ether and hydrogenated (Parr hydrogenation apparatus) over 300 mg. of Adams platinum oxide a t 40 p.s.i. a t room temperature. A extract was dried and evaporated to yield a brown viscous oil. This was purified by chromatography on a Florisil column and white crystalline solid separated out when the hydrogenation was terminated (5 hr.). The hydrogenated suspension was eluted with petroleum ether (b.p. 60-90°)-ether (10: 1). Large heated on a steam bath until the precipitated solid had redisprisms separated from the fractions on standing overnight. This material waa crystallized from petroleum ether (b.p. 30-60') solved; the exhausted catalyst was then filtered off. The filtrate to yield 4.52 g. (52%) of large prisms, m.p. 99.8-100.3'. was concentrated, from which 4.5 g. (90%) of colorless plates The ultraviolet crystallized on cooling, m.p. 223.8-224.6'. Anal. Calcd. for C21H31N05: C, 66.82; H, 8.28; N, 3.71. Found: C , 67.06; H,8.32; N , 3.64. spectrum of this compound had, ,X 272 and 277 mfi (log- 6 3.24 One gram of the base was converted to the hydrobromide and a i d 3.23, respectively). recrystallized from ethanol-ether to yield 1.06 g. (88%) of Anal. Calcd. for Cl1H1J3rNO: C, 51.17; H , 6.20; Br, 31.01; short white needles, m.p. 197.8-198.6'. N, 5.43. Found: C,51.17; H , 6.06; Br, 31.28; N, 5.42. 5-(3,4,5-Trimethoxybenzoyloxy)-2-ethyl-l,2,3,4-tetrahydroAnal. Calcd. for C*lH3zBrN05: C, 55.02; H, 7.04; Br, 17.43; isoquinoline Hybrobromide (IX).-5-Hydroxy-2-ethyl-1,2,3,4N,3.06. Found: C,55.12; H,7.15; Br, 17.50,N,3.11. tetrahydroisoquinoline hydrobromide (VIII) (4.3 g., 0.0166 The Bis(2-ethyldecahydroisoquinoline) Ether 1.-5-Hydroxymole) was dissolved in 100 ml. of water and sodium hydroxide 2-ethylisoquinolinium bromide (VI, 5 g.) was dissolved in 50 ml. of glacial acetic acid with the aid of gentle heating on a steam solution was added until no further precipitation was evident. bath. Concentrated sulfuric acid (0.5 ml.) was added and The resulting suspension was extracted with ether and dried, the mixture was hydrogenated over 5 g. of Adams platinum and the solvent was removed to yield 3.0 g . (100%) of white solid. This base was dissolved in 100 ml. of dry benzene and added to oxide a t 50 p.s.i. for a period of 36 hr. a t room temperature. a solution of 4.0 g. (0.0173 mole) of 3,4,5-trimethoxybenzoyl The exhausted catalyst was filtered, and the acidic filtrate was chloride in 100 ml. of dry benzene. Two grams of dry sodium diluted with water and made strongly alkaline by the addition of bicarbonate was added to this mixture, and the whole was resodium hydroxide solution. The base was then extracted with fluxed for 8 hr. on a steam bath. The gelatinous precipitate ether and converted to the hydrobromide salt. Upon recrystallization from ethanol-ether, 2.14 g. (42%) of fine needles were which separated out was filtered (3.3 g.), and the filtrate was exobtained, melting at 196.6-197.1'. The ultraviolet spectrum (16) Melting points were determined using a Swiasco melting point of this compound showed no absorption through the range 220apparatus and are corrected. Infrared spectra were recorded on a Perkin310 mp. The infrared spectrum (KBr) showed a medium abElmer Model 137B Infracord spectrophotometer. Ultraviolet data were sorption band a t 1110 em.-' in accordance with the absorption recorded in aqueous solution using a Beokman ratio recording spectrophoreported for an ether linkage.I8 tometer,.Model DK-2. Elemental analyses were carried out by Dre. G. -!LAC

-

- CzHSAc-

Weiler and F. B. Strauss, Oxford, England. (17) A. Lasslo and W. D. Jordan, J . O w . Chem., 21, 805 (1956).

(18) L. J. BeUamy, "The Infrared Spectra of Complex Molecules," John Wiley and Sons, Inc., New York, N.Y., 1958, p. 115.

3560

Noms

VOL. 30

Anal. Calcd. for C&iBrNzO: C, 52.52; H, 8.29; Br, 31.31; light. It gives a picrate, m.p. 204") and a benzoyl N , 5.49. Found: C, 52.57; H, 8.48; Br, 31.20; N , 5.41. derivative, m.p. 142O, in contrast to LaCoste and In each of the three instances in which I waa prepared, using Sorger's values of 165 and 137") respectively. There identical reaction conditions, yields of 42%, or in excess thereof, s e e m little doubt that these workers had, at best, were obtained. Diethiodide of the Bis(2-ethyldecahydroisoquinoline) Ether a very impure sample of the original compound. @).-The diethiodide derivative (11) of the above basic ether (I) The previously unreported 9-phenyljulolidine (11) W M prepared by refluxing the free base, obtained from the neuwas also obtained in the decomposition as e x p e ~ t e d . ~ tralization of 1.0 g. of the bis(2-ethyldecahydroisoquinoline) ether dihydrobromide, dissolved in 25 ml. of dry benzene, for 4 hr. with excess ethyl iodide. The precipitated quaternary iodide Experimental Sections was filtered and recrystallized from an alcohol-ether mixture to N,N'-Bis(p-biphenyl)-1,3-diaminopropane.-A solution of p yield 0.93 g. (72%) of fine white needles, m.p. 245.0-245.4'. The infrared spectrum (KBr) showed a medium intensity absorpaminobiphenyl' (47.5 g.) and 1,3-dibromopropane (14.0g.) in tion band a t 1110 cm.-l. 70 ml. of benzene was refluxed for a total of 20 hr. The precipiAnal. Calcd. for CiE3wIeNzO: C, 47.28; H, 7.63; I, 38.43; tated amine hydrobromides were twice removed during this time N, 4.24. Found: C, 47.28; H, 7.40; I , 38.20; N 4.27. by adding 400ml. of benzene and then washing with 10% aqueous sodium hydroxide solution before reducing the volume by Acknowledgment.-The author is indebted to Marion distillation. The mixture wm finally washed with aqueous akali Laboratories, Inc., Kansas City, Missouri, for their and with water and dried (KzCOS), and the excess p-aminobiphenyl (23.2 g.) was removed by distillation, b.p. 152-154" (1 financial assistance in support of this project and also mm.). The required compound was obtained as a yellow solid to Dr. A. Lasslo for his useful discussions and interest (10.2g., 39%), b.p. 300-310" (0.5 mm.). It was washed with in the work. a little ether and then recrystallized from ethanol t o give 7.5 g.of materia1,m.p. 111". A n d . Calcd. for C ~ ~ H Z ~ C, N Z85.68; : H, 6.92; N , 7.40. Found: G85.62; H, 7.06; N , 7.41. 6-Phen yltetrahydroquiuinoline 6-Phenyltetrahy droquinoline . A .-The diaminopropane above (6.0 g.) was decomposed at 260-270" in the presence of 0.2 ml. of 48% hydrobromic acid by the method previously R. D. TOPS OM^ described.' An ethereal solution of the resultant distillate (5.0g.) was washed with aqueous alkali and with water and the primary Department of Chemistry, University of Canterbury, aromatic aminw then removed by treatments with 50% aqueous Christchurch, Neur Zealand zinc chloride solution (2.52g. of p-aminobiphenyl was recovered from the complex obtained). Distillation of the ethereal solution gave 1.0 g. (31%) of 6-phenyltetrahydroquinoline (b.p. Received May 18,1966 210-220' a t 20mm.). B .-In a second preparation the intermediate diaminopropane I n 1885, LaCoste and Sorger2 claimed to have prewas not isolated. p-Aminobiphenyl (60 g.), 1,3-dibromopropared 6-phenyltetrahydroquinoline (I) by the reducpane (18 g.), and anhydrous potassium carbonate (18 g.) were tion of 6-phenylquinoline with zinc and hydrochloric heated together at 160" for 5 min. The mixture was then cooled and extracted by shaking with ether and water. The acid. They reported that the compound was unstable excess p-aminobiphenyl (32.5 g. recovered) was removed from and that it was soluble in hot water but insoluble in the ether layer by shaking with aqueous zinc chloride solution. chloroform and benzene. The only other reference to The crude diamine obtained after evaporation of the ether was this compound appears to be in 1957 when Avramoff decomposed as above and the p-aminobiphenyl formed was removed. Distillation then afforded the crude 6-phenyltetraand Sprinzaka suggested that the instability claimed hydroquinohe (4.1 g.), b.p. 150-189" (1 mm.). [The next above might explain their failure to isolate the comfraction (3.0 g.), b.p. 190-192" at 1 mm., was 9-phenyljulolipound, as its picrate, on reduction of 6-phenylquinoline dine.] Redistillation gave the tetrahydroquinoline (2.2 g.) as under conditions that gave 8-phenyltetrahydroquinoan oil, b.p. 154-164" (1mm.), which gradually solidified. The 6-phenyltetrahydroquinohe was recrystallized from pzline from 8-phenylquinoline. troleum ether (b.p. 6CrSO")t o give colorless crystals, m.p. 79.5 The infrared spectrum (melt) showed peaks at 3415 (N-E), H 892 (isolated C w H ) , 825 (two adjacent CAI-H bonds), 761 and 698 cm.-l (monosubstituted benzene). And. Calcd. for C&t,~N: C, 86.08; H, 7.22; N, 6.69. Found: C,85.90; H, 6.98; N,6.55. The compound gave a benzoyl derivative, m.p. 142" from aqueous ethanol. I 11 A d . Calcd. for CBHlONO: C, 84.31; H, 6.11; N, 4.47. Found: C, 84.21; H, 6.02; N,4.57. We have now prepared an authentic sample of the It also gave a picrate, m.p. 203-204". compound by the acid-catalyzed thermal decomposiAnal. Calcd. for GTHI~NAO,: __ __ - . N. 12.78. Found: N,12.85. tion416 of N,N'-(p-biphenyl)-l,3diaminopropane. It The 9-phenyljulolidine was recrystallized from petroleum is a colorless crystalline solid, m.p. 79.5", insoluble in ether to give colorless crystals, m.p. 69" (sharply depressed by addition of 6-phenyltetrahydroquinoline). The infrared spechot water and freely soluble in benzene and chloroform. trum had no peaks in the ranges 3600-3400 and 85C-800 The infrared spectrum and elemental analyses are in em.-'. accord with the structure. The compound seems as Anal. Calcd. for CIBIBN: C, 86.70; H, 7.68; N, 5.62. stable as similar tetrahydroquinolines, showing only Found: C, 86.511 H,7.68; N, 5.41. It gave a methiodide, m.p. 210". a slight tendency to darken on exposure to air and

.

~

(1) Sahool of Chemical Saienoea, University of East Anglia, Norwich, England. (2) W. LaCoste and C. Sorger, Ann., 580, 1 (1885). (3) M.Avramoff and Y . Sprinzak, J . Oro. Chem., 35,571 (1957). (4) G. B. Russell, G. J. Sutherland, R. D. Topsom, and J. Vaughan, dbid.,37,4375 (1962). (5) I. K. Lewis, G. P. Russell, R. D. Topsom, and J. Vaughan, ibid., 59,1183 (1964).

(6) Analyses were by the mioroanalytical laboratories at the University of Otago (Dr. A. D. Campbell) and Drs. G . Weiller and F. B. Strauss, Oxford. Infrared spectra were determined with a Perkin-Elmer 237 spectrometer. (7) This oompound is considered to be a dangerous carcinogen: private communication from Dr. R. A. M. Case, Chester Beatty Research Institute, London. (8) See footnote 35 of ref. 5.

OCTOBER1965

NOTES

Anal. Calcd. for CUHZIN: C, 58.32; H, 5.67; N, 3.58. Found: C, 58.40; H,5.67; N, 3.64.

Irisolidone dimethyl ether was identical with tri0-methyltectorigenin. This along with the formation of antiarol and anisic acid on alkali fusion of the dimethyl ether established the substitution pattern for irisolidone which was confirmed by the alkaline degradation of diethyl ether first to the deoxybenzoin (11) and then to 3,5-diethoxy-4-methoxyhenol and anisic acid.

Acknowledgment.-The author is indebted to the New Zealand Universities Research Committee for financial assistance and to the Nuffield Foundation for a Traveling Fellowship during the tenure of which this work was completed.

3561

The Chemical Examination of Iris nepalensis. 111. Isolation and Structure of Irisolidone LALITPRAKASH, ASIF ZAMAN, AND A. R. KIDWAI

II Department of Research in Unani Medicine, Tibbiya College, and Department of Chemistry, Aligarh Muslim University, Aligarh, U.P., India Received April ,99, 1964

Isolation of irisolonel and irigeninz from the rhizomes of Iris nepalensis D. Don (Iridaceae)has been reported earlier. Petroleum ether extracts of plants do not usually contain flavonoids3 but here an examination of this extract indicated the presence of a new isoflavone in 0.01% yield. This has been characterized as 5,7dihydroxy-6,4’-dimethoxyisoflavone(I, R = H) and named irisolidone.

Comparison of authentic 5,7-dihydroxy-6,4’-dime thoxyisoflavone8 (mixture melting point and infrared spectrum) with irisolidone showed the two to be identical. A number of species belonging to Iridaceae have been examined for isoflavonoids and as a result of this it is possible to make some observations regarding their biogenesis and chemical taxonomy which may be of significance. Thus it is interesting that all the isoflavonoids so far reported from this family have tectorigenin substitution pattern in ring A, as indicated in Table I. TABLEI Plant

dR

6 I

Irisolidone gave the characteristic color reactions of isoflavones. Its infrared spectrum showed strong absorption in the 6-6.6-p region usually associated with flavonoid systems14and the ultraviolet absorption spectrum was similar to that of irisolone and tri-0methyltectorigenin. Irisolidone showed correct analysis for C17HI4O8and two methoxyl groups. Color with ferric chloride and formation of diacetate, dimethyl ether, and a diethyl ether (I, R = Ac, CHI, and CZ&, respectively) showed the presence of two free hydroxyl groups in the molecule. Specific color reactions further indicated that one of these must be located in the 5-position and that vicinal hydroxyl groups were not present. A bathochromic shift of 10 mp in the ultraviolet spectrum characteristic of 5-hydroxyl was observed on addition of aqueous aluminum chloride5p6and a similar shift of 5 mp on addition of fused sodium acetate suggested a 7-hydroxyl. Though a shift of 10 mp7 is usually associated with the presence of 7-hydroxyl, Gottlieb and Magalhaes have recently reported a shift of only 8 m p for caviunin. (1) K. W. Gopinath, A. R. Kidwai, and L. Prakash, Tetrahedron, 16, 201 (1961). (2) K. W.Gopinath, L. Prakash, and A. R. Kidwai, Indian J . Chem., 1, 187 (1963). (3) T. R. Seshadri, “The Chemistry of Flavonoid Compounds,” T. A. Geiasman, Ed., Pergamon Press LM.,London, 1962, p. 8. (4) 0. R.Gottlieb and M. T. Magalhaes, J . 01.9. Chem., 16, 2449 (1961). (5) T. Swain, Chem. Ind. (London), 1480 (1954). (6) L. H.Briggs and T. P. Cebalo, Tetrahedron, 6, 145 (1959). (7) L. Jurd, ref. 3, p. 147.

Isoflavonoid

I . jlorentina I . kumuonensis I. germanica

Irigenin

I . tectorum

I . nepalensis

Substitution in ring A

Substitution in ring B

5-OH 6OCHa 7-OH

3‘-OH 4’-0C& 5’-OCHs

Tectorigenin

5-OH 6-OCHa 7-OH

4’-OH

Irisolone

5-OCHs

4’-OH

t:>CH* Irisolidone

SOH

4‘-OCHa

Irigenin

6-OCHs 7-OH SOH 60CHZ 7-OH

3’-OH 4’-OCHs 5’-OCHs

With respect to the substitution pattern of ring B the above members of Iridaceae, with the exception of I . tectorum, elaborate isoflavones with oxygen functions at the 3‘-, 4‘-,and 5‘-positions. I . kctor“, on the other hand, produces tectorigenin which carries only a 4‘-hydroxyl. I . nepalensis contains both types of isoflavonoids and thus appears genetically related to the two. Further, all isoflavonoids from Iridaceae have a 6-methoxyl, with the exception of irisolone which has a G17-methylenedioxy group. Formation of this may involve an oxidative cyclization of the methoxyl as suggested by Bartons in connection with the biogenetic origin of the methylenedioxy group in alkaloids, etc. (8) L. Farkaa, J. Varady, and A. Gottaegen, Acta Chim. Acad. Sci. Hung., 33, 339 (1962). (9) D. H.R. Barton, G. W. Kirby, and J. B. Taylor, Proc. Chem. Soe., 340 (1962).

NOTES

3562 Experimental Section

All ultraviolet spectra were measured with a Beckman Model DC instrument in 9570 alcohol. Infrared spectra were measured in a Perkin-Elmer Infracord either in chloroform or Nujol mulls. Isolation of Irisolidone (I, R = H) .-Air-dried powdered rhizomefi (50 kg.) of Iris nepalensis were extracted four times with hot petroleum ether (b.p. 60-80') and Lhen with chloroform. The chloroform extract yielded irisolonel and irigenin,Z whereas the petroleum ether extract, on concentration to about 2 1. and standing in refrigerator for a few days, deposited a solid. This was dissoli.ed in methanol and the solution on concentration yielded a yellow compound which on further crystallization gave irisolidone (5 g.), m.p. 195-196', lit.8 m.p. 191-192'. Irisolidone gave a blue color with ferric chloride, positive tests with boric acid,1° boric acid in acetic anhydride (Dimroth reagent), and Gibb reagent,ll and negative tests with chloropentaminocobaltic chloride12 and sodium amalgam.18 Infrared bands (in chloroform) occurred, inter alia, a t 2.85, 6.05, 6.15, 6.25, 6.35, 6.65, 6.90, 9.40, 9.65, 10.10, and 11.98 p. The ultraviolet spectrum in 9570 ethanol showed a single maximum a t 270 mp (log E 4.65). On the addition of 3 drops of 10% aqueous aluminum chloride the maximum shifteci to 230 mp (log a 4.72). Addition of saturated alcoholic fused sodium acetate solution also shifted the maximum to 275 mp (log E 4.65). Anal. Calcd. for C17Hlr06: C, 64.96; H , 4.49; 2-OMe, 19.17. Found: C, 64.76; H,4.56; OMe, 19.14. Irisolidone Diacetate (I, R = Ac) .-The acetate was prepared from irisolidone (0.1 g.), acetic anhydride (1 ml.), and pyridine (2 drops) by allowing to stand a t room temperature for 24 hr.; i t crystallized from methanol in colorless needles, m.p. 162-163'. Anal. Calcd. for CZ1H18O8: C, 63.31; H, 4.55. Found: C, 63.45; H, 4.60. Irisolidone Methyl Ether (I, R = CH3).-A mixture of irisolidone (1.5 g.), freshly distilled dimethyl sulfate (6.4 ml.), anhydrous potassium carbonate (8.4 g.), and dry acetone (120 ml.) was refluxed for 36 hr. On working up in the usual manner, colorless plates of irisolidone methyl ether (1.35 g.), m.p. 181°, were obtained, identical with tri-0-methyltectorigenin (mixture melting point and infrared spectrum). Anal. Calcd. for C1&806: C, 66.66; H, 5.30. Found: C, 66.52; H, 5.26. Irisolidone Ethyl Ether (I, R = C2Hs).-Irisolidone (2 g.), potassium carbonate (4 g.), ethyl iodide (20 ml.), and :wetone (200 ml.) were refluxed together for 40 hr. Work-up in the usual manner gave irisolidone ethyl ether, m.p. 113-114'; infrared bands (in chloroform) occurred, inter alia, a t 6.05, 6.25, 6.35, 6.75, 6.90,9.30,9.45,9.70,10.00,and12.00p. Anal. Calcd. for CZ1Hz2O6:C, 68.09; H, 5.99. Found: C, 68.00; H, 6.30. 4'-Methoxybenzyl 2-Hydroxy-4,6-diethoxy-S-methoxyphenyl Ketone (11).-hisolidone ethyl ether (450 mg.) was suspended in water (30 ml.) and a, slow current of nitrogen free from oxygen was passed through the suspension. After 5 min. aqueous sodium hydroxide (15 ml., 10%) was added, and the mixture was refluxed for 2 hr., cooled to room temperature, and extracted with ether. The ether extract was dried (NazSOd) and evaporated, and the residue was crystallized from methanol: m.p. 102-103". It gave a brown color with alcoholic ferric chloride. Anal. Calcd. for C20Hz406: C, 66.65; H, 6.71. Found: C, 66.42, H, 6.82. Oxidation of Irisolidone Ethyl Ether.-The ethyl ether (0.5 g.) in acetone (50 ml.) was heated on a water bath and treated with powdered potassium permanganate in small lots until the pink color persisted. The solution was filtered and the solid was washed with acetone and then extracted with hot water. Acidification of the acqueous solution gave a solid, m.p. 186187" (from hot water), lit." m.p. 184". It gave no depression with an authentic sample of anisic acid. Alkali Fusion of Irisolidone Ethyl Ether .-Irisolidone ethyl ether (1 g.), sodium hydroxide (2.5 g.), and water (2 ml.) were heated in a copper tube a t 220' for 45 min. Work-up in the usual manner yielded 3,5-diethoxy-4-methoxyphenol, m.p. 72', (10) C. W. Wilson, J . Am. Chem. Soc., 61, 2303 (1939). (11) F. E.King, T. J. King, and E. Sellars, J . Chem. Soc., 563 (1957). (12) Y. Asahina, J. Asano, and U. Ueno, Bull. Chem. S O C .Japan, 17, 104 (1942). (13) G.Bargellini, Gazz. chim. ital., 49, 47 (1919). (14) I. Heilbron and H. M. Bunbury, "Dictionary of Organic Compounds," Vol. I , Eyre Spottiswoode, London, 1953,p. 178.

VOL. 30

lit.lb m.p. 74", and anisic acid. No depression in mixture melting points was obtained with authentic samples.

Acknowledgment.-The authors wish to thank Professor L. Farkas, Technical University, Budapest, Hungary, for kindly supplying a sample of 5,7-dihydroxy-6,4'-dimethoxyisoflavone. L. Prakash is grateful to the Council of Scientific and Industrial Research (India) for the award of a Senior Research Fellowship. (15) M. Krishnamurti and T. R. Seshadri, Proc. Indian Acad. Sei., 39A, 144 (1954).

Abnormal Products during Isolation of Isonicotinic Acid Hydrazide UMAPRASANNA BASUAND SAKTIPADA DUTTA

Bengal Immunity Research Institute, Calcutta 16, India Received February 9, 1966

I n the course of isolation of isonicotinic acid hydrazide (I) from the reaction mixture of ethyl isonicotinate and hydrazine hydrate (free from ammonia) three basic by-products, namely, symmetrical diisonicotiny1 hydrazine (11), hydrazodicarbonamide (111), and 3,5-bis(4-pyridyl)-1,2,4-triazole-l-carboxhydrazide (IV), have been isolated in small quantities as artifacts arising from the large-scale operation, Their separation is to be specially considered when isonicotinic acid hydrazide is required in high purity. The solubility of I1 and I11 in dilute alkali and the solubility of I1 in ethanol have rendered separation possible, and the products could be finally purified by crystallization from water. The product IV affords benzylidene and cinnamylidene derivatives. Oxidation under diverse conditions such as moderately dilute nitric acid, acid potassium permanganate, or sodium hypobromite IV gives rise to the formation of 3 :5-bis(4-pyridyl)-1,2,4-triazole (V),carbon dioxide, and nitrogen. The identity of V has been confirmed by synthesis from I and isonicotinamide. Compound IV in solution of 10% sulfuric acid or in glacial acetic acid, when treated with bromine, gives a perbromide which by the action of alkali changes to V. A solution of I in hydrazine hydrate saturated with carbon dioxide or a mixture of I, isonicotinamide, and carbohydrazide in aqueous medium when heated to reflux for some time leads to the formation of IV. Similarly, a solution of hydrazine hydrate saturated with carbon dioxide admixed with ammonia affords 111. I n the formation of the by-products I11 and IV, the mechanism of reaction involves the participation of atmospheric carbon dioxide, hydrazine, and ammonia (decomposition product of hydrazinez). Hydrazine and carbon dioxide form carbazic acid3 which with ammonia and hydrazine lead to the formation of semi(1) 0. Silberrsd, J . Chem. Sac., 77, 1185 (1900). (2) M. Berthelot and C. Matignon, d m p t . rend., 126, 1042 (1896). (3) R. Stolle and K. A. Hofmann, Be?., 37, 4523 (1904).

NOTES

OCTOBER1965

3563

I V (10 g., 0.034 mole), and acetic acid (50 ml.) was heated to reflux for 4 hr., cooled, and diluted with ice-cold water (200 ml.). The aqueous part, after extraction of unreacted benzaldehyde with ether, was basified with 10% caustic soda solution in the cold and the crystalline solid that separated was crystallized from ethanol in brilliant cream-colored needles to afford the I benzilidene derivative, 6.8 g., m.p. 196-198'. Anal. Calcd. for Cz&6N70: N , 26.55. Found: N , 26.7. carbazide and carbohydrazide, respectively, through The hydrochloride crystallized from ethanol in light brown the corresponding ammonium and hydrazinium salts needles, m.p. 316-318'. Anal. Calcd. for C&16N70*2HCl: HC1, 16.5. Found: of carbazic acid. Additional ammonia reacts with I HC1,16.3. to form is~nicotinamide~ which again reacts with I I n a similar way the cinnamyledene derivative of IV was obto give diisonic~tinylamide.~Semicarbazidea affords tained in brown needles, m.p. 206-208'. 111, whereas diisonicotinylamide and carbohydrazide Anal. Calcd. for CzzHL7N70:N, 24.8. Found: N , 24.28. Oxidation of 3,5-Bis(4-pyridyl)-l,2,4-triazole-l-carboxhydrareact to furnish IV.7 zide (IV) to 3,5-Bis(4-pyridyl)-l,2,4-triazole (V) .-The triazole I V (15 g., 0.051 mole) dissolved in 10% sulfuric acid (200 ml.) was treated with potassium permanganate (30 g., 0.19 mole) added gradually in small installments a t room temperature. 111 When the color of one addition was discharged, another addition 11 followed. The gas evolved during oxidation was introduced to freshly prepared calcium hydroxide solution (100 ml., 5%) protected from atmospheric carbon dioxide. As the oxidation progressed, the calcium hydroxide solution turned milky indicating the formation of carbon dioxide. After oxidation, the reaction mixture was treated with 48% (w./w.) caustic soda solution to make the solution alkaline. The manganese dioxide which separated was filtered off and washed with water (100 ml.). The combined filtrate and washing were brought to p H IV,R = CONHNHP 6-7 with hydrochloric acid and the product that separated waa purified by crystallization from a water-ethanol mixture (1:1) V, R = H to afford V in white flaky crystals, 8.2 g., m.p. 28C-282'. It is amphoteric in nature being soluble both in acid and alkali, inExperimental Section8 cluding ammonia and hydrazine. Anal. Calcd. for CI2H9N6: C, 64.5; H, 4.0; N, 31.3. Found: Isolation of Symmetrical Diisonicotinylhydrazine (11) , HyC,64.3; H,4.1; N,31.1. drazodicarbonamide (111), and 3,5-Bis(4-pyridyl)-1,2,4triazoleThe picrate crystallized from ethanol in yellow needles, m.p. 1-carboxhydrazide (IV).-The separation of first crop of I (m.p. 258-260' dec. 171-173", purity 99.5%) from the reaction mixture of ethyl Anal. Calcd. for ClzHsN6.2C8H3N307: N, 22.5. Found: isonicotinate (b.p. 218", sp. gr. 1.1003 a t 25', 25 kg., 165 moles) N , 22.3. and hydrazine hydrate (free from ammonia, 35% w./v., 175 1.) The hydrochloride crystallized from water-ethanol mixture afforded a mother liquor which on concentration, cooling, and (1:1)in white needles, m.p. 338-340" dec. filtration gave a solid cake consisting of the mixture of I, 11,111, Anal. Calcd. for C12H9N6.2HC1: N, 23.61; HC1, 24.6. and IV. This cake, after trituration with cold water (18 1.) Found: N, 23.8; HC1,24.2. and jiltration, afforded insoluble material (0.55 kg., a mixture of On oxidation of I V with moderately dilute nitric acid the 11,111,and IV), and the filtrate gave isonicotinic acid hydrazide triazole V waa isolated as its nitrate, which on crystallization from on concentration. Caustic soda solution (lo%, 200 ml.) diswater in white needles showed m.p. 244-246' dec. solved I1 and I11 leaving behind I V as an insoluble material, Anal. Calcd. for CIZHBNB.~HNO~: N, 28.1. Found: N , which on crystallization from water, ethanol, or pyridine gave 28.6. IV (9.8 g.) in fine white needles, showing m.p. 338-340' dec. From the above nitrate V was isolated a t p H 6-7. Anal. Calcd. for C13HllN70: C, 55.5; H , 3.9; N, 34.8. Formation of V from IV by Bromine and 10% Caustic Soda.Found: C,55.42; H,3.76; N,34.6. To the triazole IV (20 g., 0.068 mole), suspended in 10% caustic The picrate crystallized from ethanol in yellow needles and soda solution (250 ml.), bromine (10 ml.) was slowly added with showed m.p. 238-240" dec. stirring and ice cooling; IV underwent decomposition with evoluAnal. Calcd. for C13HllN70.2C~H8N307:N , 24.56. Found: tion of gas (nitrogen) and gradually goes into solution. The N, 23.8. straw-colored solution (pH 1C-11) obtained after complete addiThe hydrochloride crystallized from water in beautiful yellow tion of the bromine was brought to p H 6-7 with hydrochloric shining needles showing m.p. 308-310" dec. acid with evolution of carbon dioxide and separation of the Anal. Calcd. for C13HllN70.2HCl: HC1 20.6; N, 27.7. triazole V, 12 g., after crystallization. It showed no depression in Found: HCl, 20.4; N, 27.5. melting point when mixed with a sample of V obtained from the The alkaline solution on adjustment of p H to 6-7 gave a previous experiment. mixture of I1 and I11 from which I11 was obtained as an insoluble Perbromide from 1V.-To the triazole IV (25 g., 0.085 mole) material by stirring with ethanol a t room temperature. I11 dissolved in 10% sulfuric acid (250 ml.), bromine (10 d.) waa was crystallized from water in beautiful shining needles (9.1 g.) added slowly under external ice cooling. The semisolid mass showing m.p. 244246' dec.6 and no depression when adthat separated solidifies to crystalline material (orange-yellow mixed with an authentic sample of 111. Removal of ethanol needles) on stirring. The material was filtered, washed with gave I1 which Crystallized from water in white needles (500 g.) water thoroughly until sulfate free, and finally washed with showing m.p. 262-264' (lit.g m.p. 262-264' and no depression alcohol and ether to afford the perbromide, 48 g. It did not when admixed with an authentic sample of 11. l-Benzylidenecarboxhydrazino-3,5-bis-(4-pyridyl)-l,2,4-tri- melt evenat 340'. Anal. Calcd. for C18HllBr4NT0: Br, 53.2. Found: Br, azole.-A clear solution of benzaldehyde (10 g., 0.94 mole), 52.7. (4) T. Curtius and C. Sturve, J . pro&. Chem., [ii] 60, 295 (1895). The perbromide, on refluxing with acetone, afforded mono( 5 ) G. Pelliazari, Uazz. chim. {tal., [ii] a4, 222 (1894). bromoacetone, b.p. 64' (50 mm.), and the hydrobromide of IV (6) R. Stolle, Ber., 46,260 (1913). which crystallized from ethanol in light yellow needles showing (7) K. Brunner. Monatsh., 36,509 (1915). m.B. 262-264" dec. ( 8 ) Melting points were determined in capillary tubes using an eleotrical Anal. Calcd.. for C I Z H I I N ~ O - ~ HN, B ~22.10; : HBr, 36.65. air bath or a silicone oil bath and are uncorrected. Microanalyses were Found: N , 21.85; HBr 36.4. performed in the microanalytical laboratory of Bengal Immunity Research The perbromide underwent instantaneous decomposition in Institute, Calcutta 16,India. (9) H. H. Fox and J. T. Gibaa, J. 078. Chem., 17, 1653 (1952). contact with 10% caustic sods, solution a t room temperature with

COIWNH, I

3564

NOTES

evolution of nitrogen and went into solution, which a t p H 6-7 with evolution of carbon dioxide afforded the triazole V showing no depression in melting point when admixed with a sample of V obtained from the previous experiment. Synthesis of 3,5-Bis(4-pyridyl)-l,2,4-triazoie (V) .-An intimate mixture of isonicotinic scid hydrazide (13.7 g., 0.1 mole) and isonicotinamide (0.1 mole) waa heated in an oil bath for 16 hr. a t 170-220' but mainly in the vicinity of 200° for last 4 hr. Water formed in the reaction escaped as there was no condenser attached to the flask in which the experiment waa being done. The reaction mixture on being cooled, solidified and dissolved on treatment with water (150 ml.) and ammonia (10 ml.). This solution, being adjusted to pH 6-7 with hydrochloric acid, afforded V, 10.5 g.; the mixture melting point with a sample of V obtained by oxidation of IV waa not depressed. Synthesis of 3,5-Bis(4-pyridy!)-l,2,4-triazole-l-carboxhydrazide (IV) from Isonicotinic Acid Hydrazide, Carbon Dioxide, and Hydrazine 1Iydrate.-Isonicotinic acid hydrazide (25 g., 0.18 mole) suspended in hydrazine hydrate [150 ml., 50% (w./v.), free from ammonia] was saturated with carbon dioxide under external ice cooling and then heated to reflux with a small flame. The reaction mixture became clear on heating and after 8-10 hr. of refluxing, separation of white crystalline material began. The heating to reflux was continued for 18 hr. The reaction mixture was cooled and the white crystalline material that separated was filtered, washed with water, and crystallized from water, in which it is very sparingly soluble, in white needles to afford IV, 4.2 g., m.p. 338-340" dec., no depression when admixed with a sample of IV isolated as by-product. Synthesis of IV from Isonicotinic Acid Hydrazide, Isonicotinamide, and Carb0hydrazide.-A mixture of isonkotinic acid hydrazide (13.7 g., 0.3 mole), isonicotinsmide (12.2 g., 0.1 mole), and carbohydrazide (9 g., 0.1 mole) in water (100 ml.) waa heated to reflux. During heating the solution became clear and after 4-5 hr. white crystalline material began to separate. After refluxing for 12 hr. the reaction mixture was cooled; the white crystalline solid that separated was isolated and finally crystallized from water to afford IV, 6 g. The mixture melting point with a sample of IV from the previous experiment showed no depression.

VOL.30 CHART 1

la,R=CH3 b, R = H

2

1

5a, R = CH3 b,R=H

(i-Bu)ZAlH

6a, R = CK3 b,R=H CI43

7a,X=O

Lactols Derived from Steroidal 17a-Oxa-D-homo Lactones

H

b, X'CoH H c,x=< *oCH3

JOHNS. BARAN

7d,X=
M< CBr4

R ~ C E June ~ V E8,~1966

Olefin 1,5-Cyclooctadiene l,&Cyclooctadiene 1,5-Cyclooctadiene l,&Cyolooctadiene l,5-Cycloootadiene cia-Cyclooctene cia-Cyclooctene

VOL. 30

+ CHaLi + Br [:CBrz] + >M< + C''

Br

/ \

(1)

>c---C< Br C ''

Br

+ CHaLi +>C=C=C