2738 EWINS: THE ALKALOIDS OF QUEBRACHO BARK. PART I.CCLV.-The All~aloids of Quebracho Bark. Palst I.The Constitution of Aspidospewnine.By ARTHUR JAMES EWINS.ASPIDOSPERMINE, C,H,O,N,, is the most readily obtained of thealkaloids of quebracho bark. It was first isolated by Fraude (Ber.,1878, 11, 2189) from the bark of Aspidosperma Quebracho, the" quebracho hlanco " of the Argentine, where it a t one time foundemployment in the treatment of fever. A further investigation oEWINS: THE ALKALOIDS OF QUEBRACHO BARK. PART I. 2739the alkaloids of this bark was carried out by Hesse (Annulen, 1882,211, 2491, who claimed to have isolated no less than six different,but closely related bases. Of these, however, only two, namely,aspidospesmine and quebrachine, may be said to have been wellcharacterised, and of these the latter has recently been shown tobe identical with yohimbine (Fourneau and Page, Bull.Sci.Pharmacol., 1914, 21, 7).* According to Hesse, quebrachine onlyoccurs in some specimens of the bark, and in those examined by thepresent author, at most only traces of this alkaloid have beenfound. The present paper is concerned mainly with some resultsobtained in experiments on the constitution of aspidospermine. Atthe same time the author has not so far been able t o confirm theexistence of the various bases described by Hesse with the exceptionof aspidospermine and qtiebrachine. Further, Hesse employedwarm dilute sulphuric acid €or the extraction of the alkaloids fromthe bark, and under these conditions, as is shown by the experi-ments recorded in this paper, aspidospermine is hydrolysed, givingrise to a new base, the properties of which make i t appear possiblethat the bases aspidosamine and hypoquebrachine described byHesse may have been impure forms of this decomposition productof aspidospermine.It may be mentioned, however, that in the course of the presentwork two new well-defined crystalline alkaloids were obtained insmall quantity. One, characterised by its sparing solubility inchloroform and by its failure to give colour reactions with oxidisingagents, crystallised from ethyl acetate in well-formed octahedramelting a t 176-177O.The other is very sparingly soluble inether, and crystallised from light petroleum in stout prisms meltinga t 149-150°.The latter base gives colour reactions whichresemble those given by the base obtained by the hydrolysis ofaspidospermine, but are less intense. It is hoped that these basesmay form the subject of a future investigation.Aspidospsrmine is a feeble base, which does not yield crystalline* By fractional precipitation of a portion of the crude alkaloid as tartrate asmall quantity of a crystalline salt was obtaincd from which a crystalline basemelting a t 225-226" was isolated. This base corresponded in general with thatdescribed by Hesse as quebrachine, and the recent statement by Fourneau and Page(Zoc. cit.), that the latter base is identical with yohimbine, was confirmed by thefollowing observations. A specimen of yohimbine (for which I am indebted toDr.G. Barger) melted a t 226", and a mixture of equal parts of the two bases melteda t the same temperature. Both bases melt to a red liquid and sublime a t200-210"/5 or 6 mm., forming clusters of prismatic needles. Further, Barger andMiss Field have found (private communication) that yohimbine gives a character-istic aulphonic acid derivative. The formation of this acid from the base(quebrachine) mentioned above was carried out, and its identity with that obtainedfrom yohimbine readily established2740 EWINS: THE ALKALOIDS OF QUERRACHO BARK. PART I.salts. It' does not react with methyl iodideexcept after prolonged heating a t 100°, and then yields a mixtureof products which have not yet been investigated.Aspidosperminecontains one methoxyl group, and does not' contain an AT-methylgroup, but has an acetyl group attached t o one nitrogen atom.This is shown by the following facts.The hydriodic acid solution obtained after treating aspidosper-mine according t o Zeisel's method yielded a new base, aspido-s i m , C,,H,,ON,. This base therefore differed in composition fromaspidospermine by the complex C,H,O, of which the hydrolysis ofone methoxyl group accounts for CH,. The residual C,H,O pointedt o the presence, in aspidospermine, of an acetyl group, which,from the constitutbn of the base, must be attached to a nitrogenatom. The presence of an acetyl group was confirmed by thefollowing results.Aspidospermine on hydrolysis with boiling dilute hydrochloricacid gives a new crystalline base, deacetylaspidbspermi,ae,C,,H,,ON,, which is readily reconverted into aspidospermine onacetylation.The new base on boiling with concentrated hydriodicacid (D 1.7) as would be expected yields aspidosine, the relation-ship of these' bases being shown as follows:Aspidospermine, C22H300JST2It is Izvorotatory.BoilingHI (1.7)Deace t y laspidospermine, C,H,,0N2 --+A s pi dosine, C l&&N2Nitronitrosodeacetylaspidospermine, C20H,,0,N,.Deacetylaspidospermine forms a characteristic, sparingly soluble,crystalline dihgdriodide ; the corresponding hydrochloride andhydrobromide are very readily soluble, and have not been obtainedcrystalline. It combines with methyl iodide, forming a crystallinederivative having the1 composition C,,H2,0N,,2CH,I. It yields amonoacetyl derivative (aspidospermine) and a crystalline mono-benzoyl derivative.It reacts with nitrous acid, and under suitableconditions yields a crystalline derivative, which appears t o be anitronitroso-compound. Deacetylaspidospermine thus appears to bea secondary tertiary base. The formation of a nitronitroso-derivEWINS: THE ALKALOIDS OF QUEBRACHO BARK. PART I. 2741ative, together with the colour reactions of the base, probablyindicate the presence in the molecule of a reduced quinolinenucleus. On boiling with hydriodic acid, nitronitrosodeacetyl-aspidospermine yields aspidosine. Deacetylaspidospermine isoptically active, but of opposite sign (dextrorotatory) to that ofaspidospermine.9 spw3osine is a crystalline lzvorotatory base, which, like deacetyl-aspidospermine, gives a characteristic, sparingly soluble, crystallineIiydriodide of the composition C,,H,,ON,,HI. It is less basic thandeaeetylaspidospermine, probably owing to the presence of aphenolic hydroxyl group.Probably also for this reason the baseis somewhat unstable in solution.On oxidation with chromic acid, aspidospermine yields a newcrystalline base, which forms a sparingly soluble hydrochloride.Owing t o want of material t'lie base has not' yet been fully investi-gated. It gives none of the colour reactions characteristic ofaspidospermine and its immediate derivatives.EXPERIMENTAL.Preparation of A spidospermine.The finely ground bark was completely extracted with hot alcohol(about 95 per cent.), and the alcohol removed by distillation.Thedark-coloured, viscous residue was extracted with a 20 per cent.solution of acetic acid until the extracts gave only very feeblealkaloidal reactions. The extract was diluted, which caused theprecipitation of a certain amount' of resin, and, wit,hout filtration,treated with a saturated aqueous solution of normal lead acetate,until t'he filtrate no longer gave a precipitate either on dilution oron addition of lead acetate. Thc precipitate was collected, the leadremoved from the filtrate as sulphide, and the solution madealkaline with ammonium hydroxide. A voluminous precipitate wasproduced, which was collected and dried. The filtrate still con-tained some alkaloid in solution, which was readily removed byextraction with chloroform.The residue obtained on distilling offthe chloroform was added t o the crude alkaloid first precipitated.The mixture of bases so obtained was dissolved in a small amountof absolute alcohol, and on keeping, aspidospermine crystallised out.This was collected and purified by recrystallisation from methylalcohol, or, f o r analysis, by sublimation in a vacuum.According t o Hesse (Zoc. c i t . ) quebrachine if present crystalliseswith aspidospermine under conditions similar t o those describedabove. I n the present investigation, however, quebrachine wasnever found t o be present.The alcoholic mother liquors still contained a very considerabl2742 EWINS: THE ALKALOIDS OF QUEBRACHO BARK.PART I.amount of crude alkaloid. Attempts have been made t o work outa satisfactory method of separation of the bases present in thismaterial, butl so far without very much success. Evidence has beenobtained, however, of the presence of a t least two hitherto un-described crystlalline bases, but the amounts so far obtained weretoo small t o permit of further investigation.The separation of aspidospermine from the crude alkaloids bythe method described above is by no means sharp. As alreadystated, however, aspidospermine on hydrolysis with dilute mineralacids givee a new base, deacetylaspidospermine, which can bedistilled under greatly diminished pressure, and is very readilysoluble in light petroleum.The crude alkaloids obtained from the alcoholic mother liquorsafter separation of aspidospermine were therefore boiled with 10 percent.hydrochloric acid for tlwo hours. The solution was thenfiltered and the filtrate made alkaline with ammonium hydroxide.The precipitated bases were collected and dried, and both the driedmaterial and the alkaline filtrate then completely extracted withlight petroleum.The extracts were combined, the solvent removed by distillation,and the residue distilled under 1-2 mm. pressure. The distillate,which collected between 210° and 220°, was dissolved in hot diluteacetone, and on cooling crystals of almost pure deacetylaspidosper-mine melting a t 109O were obtained. The increased amount of baseisolated by this procedure corresponded in some instances with asmuch as 40 per cent.of the amount of aspidospermine originallyobtained.The amount of aspidosgermine present in quebracho bark appearsto vary considerably. From one batch of bark (50 kilos.) a yieldof only 0.06 per cent. of aspidospermine was obtained, althoughfrom smaller bstches yields up to 0.2 per cent. were obtained, Thetotal alkaloid of the bark varies considerably with age. Accordingt o Hestxi (Zoc. cit.), the young bark contains up to 1.4 per cent. ofalkaloid, whilst old bark may contain as little as 0.3 per cent.Aspidospermine, C22H3,0,N2, crystallises from alcohol or lightpetsoleum in needles melting a t 208O. It sublimes under dimin-ished pressure a t about 180°, and can bs distilled under 1-2 mm.pressure a t about 220O.It is fairly readily soluble in most organicsolvents, but almost insoluble in water. It is precipitated by alkalisfrom acid solution as a white, amorphous solid, which becomescrystalline after keeping for a few minutes. Aspidospermine isonly feebly basic, and dow not form crystalline salts. It dissolvesin concentrated sulphuric acid, forming a colourless solution, fromwhich it may be recovered unchanged after prolonged keepingEWINS: THE ALKALOIDS OF QUEBRACHO BARK. PART I. 2743Addition of a crystal of potassium dichromate to this solution givesa brown colorztion, which becomes olivegreen after some time. Onwarming with perchloric acid, aspidospermine gives a rose-redcolour.Aspidospermine is laevorotatory.Determinations of its specificrot’atory power gave the following results :( 1 ) in alcoholic solution at 1B0, a, -1*79O, c=1*81, Z=1 dcm.;(2) in chloroform solution a t 18O, a, -1.6So, c = l * S l , Z=1 dcm.;The corresponding values obtained by Hesse (Zoc. cit.) were- 100.2O and - 83.6O respectively.The formula C,H,O,N,, originally due to Fraude and later con-firmed by Hesse, was confirmed by analysis (Found, C=74*2;H=8*6; N=8*3. Calc., C=:74-5; H=8*4; N=8*0 per cent.), andstill further by the constitution of the bases derived from it,as will be seen below.A determination of the molecular weight by Barger’s microscopicmethod (T., 1904, 85, 286) gave the following result:0.056 in 0.549 pyridine = 0.25 mole. M.W. = 408.A determination of &he meth6xyl groups present in the base was0.2131 gave 0.1406 AgI.OMe=8*7.Aspidospermine theref ore contains one methoxy-group.[a], -99O.[alD -93O.C,,H,O,N, requires M.W. = 356.made according to Perkin’s modification of Zeisel’s method :C,,H,,O,N, requires OMe = 8.7 per cent.Furtherheating with hydriodic acid up t o 300° or rather higher accordingt o Herzig and &!eyer’s method showed that the base contained noN-methyl group.The ,4 c t i o n of Boiling Hydriodic Acid 011, Aspidospermine.Formation of a New Base, Aspidosine, C,,H,,ON,.It was observed that, after treatment of aspidospermine withhydriodic acid for t8he determination of the methoxyl group, theresidue which remained after removal of thO bulk of the hydriodicacid by distillation consisted of needles with a metallic lustre,obviously the periodide of a base.The pure base was obtained inlarger quantity as follows. One gram of aspidospermine was boiledfor one and a-half hours under reflux with 20 C.C. of hydriodic acid(D 1-7). A t the end of this time the excess of hydriodic acid wasremoved by distillation under diminished pressure, the residuesuspended in water and decolorised by sulphur dioxide. The result-ing solution was filtered from a small amount of flocculent material2744 EWINS: THE ALKALOIDS OF QUEBRACHO BARK. PART I.and made alkaline by the addition of ammonia. The white,amorphous precipitate rapidly became crystalline, and was thencollected and dried. The product was almost pure, and weighed0.81 gram.F o r purification the base was recrystallised from alcohol or fromxylene, when it was obtained in well-formed, rectangular prisms orplat'es, which melted a t 244-245O after sintering from about 236O.Repeated recrystallisation failed to effect any change of meltingpoint :0.1260 gave 0.3526 CO, and 0.0994 H,O.0.1634 ,, 13.2 C.C.N, (moist) a t 20° and 763 mm. N=9.3.C,,H,,ON, requires C = 76.5 ; H = 8.7 ; N = 9.4 per cent.Aspidosine is fairly readily soluble in alcohol, ethyl acetate, orxylene; very sparingly so in chloroform or light petroleum, andalmost insoluble in water. I t s solutions in organic solvents becomecoloured on keeping, and the base itself was invariably slightlycoloured. For these reasons accurate determinations of the rotatorypower of the base could not be made'.It is levorotatory, and hasabout [aID -16O in alcoholic solution. I n the presence of alkalithe base gradually dissolves, forming a greenish-blue solution.It dissolves inconcentrated sulphuric acid, forming a pale rose-red solution.AdditIon of oxidising agents, such as potassium dichromate, leadoxide, or nitrous acid produces a reddish-violet coloration. A dropof nitric acid added t o a few drops of the sulphuric acid solutiongives a deep orange-red colour. Crystals of aspidosine moistenedwith ferric chloride are coloured greenish-blue, gradually changingto reddish-brown. I n dilute acid solution the base gives with ferricchloride a reddish-brown colour, which passes through brownish-purple to deep red.Aspidosine Hydriodide, C,gH,,ON,,HI.-This salt was firstobtained during the preparation of the base described above.If t othe solution which has been treated with sulphur dioxide, ammoniais added drop by drop, a crystalline solid separates even while thesolution remains distinctly acid. This proved to be the hydriodideof aspidosine, the free base being obtained from it on furthertreatment with ammonia. The salt is very sparingly soluble incold, but very readily 60 in hot water, from which it crystalliseson cooling in regular octahedra and cubes. Its melting point liesabove 280O.C = 76.3 ; H = 8.8.Aspidosine gives very intense colour reactions.The salt is anhydrous:0.1048 gave 0.0588 AgI. 1=30.3.C,,H,,ON,,HI requires I = 29.8 per centEWINS: THE ALKALOIDS OF QUEBRACHO BARK.PART I. 2745The Action of Dilute Hydrochloric Acid on Aspidospermiue.Formation of Deacetylns~'dosperrnilLe, C,,H,,ON,.One gram of aspidospermine was heated for three hours a t looowith 10 C.C. of 10 per cent. aqueous hydrochloric acid. The result-ing solution was rendered alkaline whea an amorphous base wasprecipitated, which on keeping for a short time became crystalline.It was collected and recrystallised from dilute acetone, when i tformed long, prismatic needles, melting a t l l O - l l l o :0.1356 gave 0.3803 CO, and 0.1092 H,O. C = 76.5 ; H = 8.9.0.1464 ,, 11.4 C.C. N, (moist) a t loo and 767 mm. N=9.36.0.1862 ,, 0.1362 AgI. OMe=9.7.C,,H,ON, requires C = 76.9 ; H = 9.0 ; N = 9.0 ; OMe = 9.9 per cent.I)ecccetylasln'dospermi,Le is readily soluble in most organic sol-vents, but very sparingly so in water. It distils unchanged a t about210° under 1-2 xm.pressure. It dissolves in sulphuric acid t o acolourless solution, which on the addition of a drop of nitric acidgives a violet, or of potassium dichromate a deep brownish-purplecolour. With ferric chloride a magenta colour is produced. Withweaker oxidising agents, such as mercuric acetate, a rose-red colouris obtained, which slowly changes t o violet.Deacetylaspidospermine is feebly dextrorotatory ; a 2.5 per cent.solution in absolute alcohol has [a]= +2*S0.Deacetylaspidosperrnine hcydriodide is obtained when deacetyl-aspidospermins is dissolved in a small quantity of hot dilutehydriodic acid.On cooling the hydriodide separates in stout,rect'angular prisms, melting a t about 243O aft.er sintering from235O. The salt is very sparingly soluble in cold, but fairly readilyso in hot water, or alcohol:0.1134 gave 0.0920 AgI. I=43*8.C20H280N2,2HI requires I = 44.7 per cent.Benzoyldeacetylas~"dosperlnilze is obtained by benzoylating thebase either by Einhorn's method in pyridine solution or by heatinga t looo for one hour with benzoic anhydride. It crystallises fromdilute alcohol in stout rhonibs melting a t 186-187O:0.1017 gave 0.2904 CO, and 0.0720 q0. C = 77.9 ; H= 7.9.0.1246 ,, 7.5 C.C. N, (moist) a t 19O and 764 mm. N=7.0.C,,H,,O,N, requires C = 77.9 ; H = 7.7 ; N = 6.7 per cent.If deacetylaspidospermine is dissolved in a small quantity ofmethyl iodide, the solution warmed for a few moments and thenallowed to remain, a crystalline solid separates, which, when re-crystallised from methyl alcohol, forms well-defined octahedramelting a t 176-177O.It has the composition C,,H2,ON,,2CH,I 2746 EWINS: THE ALKALOIDS OF QUEBRACHO BARK. PART I.0.1046 gave 0*0804 AgI. I=41*5.C,,EI,,0N,,2CHyI requires I = 42.6 per cent'.Formation of Asp'dosper,mine by Acetylation of Deacetyl-asp'dospermirze.That deacetylaspidospermine is derived from aspidospermine bythe removal of an acetyl group is further confirmed by the follow-ing experiment, in which aspidospermine was formed by acetylatingdeacetylaspidospermine. 0*2 Gram of deacetylaspidospermine wasdissolved in 1 C.C.of acetic anhydride, one drop of sulphuric acidadded and the mixture boiled for one minute. The solution wascooled, diluted with water, and rendered alkaline with ammoniumhydroxide. The precipitated amorphous base was collected andcrystallised from methyl alcohol. It separated in needles meltinga t 206-207", and wlien mixed with an equal weight of aspido-spermine showed no depression of melting point. I n all otherrespects also it was identical with aspidospermine. The yield waspractically quantitative.Action of Nitrous Acid 0.12 Deacety1as~'~ospermiize : J'ormationof a LVitronitroso-derivative ( 1).If deacetylaspidospermine is dissolved in dilute hydrochloric acidand treated with sodium nitrite' solution a crystalline solid sepa-rates, which melts indefinitely at about 160--170°, and on recrystal-lisation from a mixture of pyridine and alcohol gives a productmelting a t 220-230° after sintering from about 200O.The sub-stance is obviously a mixture; it contains chlorine, and attemptsto obtain a pure compound were unsuccessful. If the reaction iscarried out in the presence of concentrated hydrochloric acid adeep magenta solution is obtained, from which no crystallineproduct could be isolated. If, however, the reaction is carried outin acetic acid solution a pure substance is obtained.0.5 Gram of deacetylaspidospermine was dissolved in 5 C.C. of a10 per cent. solution of acetic acid. To the cooled solution satur-ated aqueous sodium nitrite solution was added drop by drop untilthe separation of a crystalline solid appeared to be complete. Thesolution was allowed to remain for about an hour, when the crystal-line solid was collected, washed with water, and recrystallised fromdilute acetone, from which i t separated in pale yellow prisms, melt-ing and decomposing a t 155-156O.Recrystallisation from amixture of pyridine and alcohol gave a similar product, having thesame melting pointEWINS: THE ALKALOIDS OF QUEBRACHO BARK. PART I. 21410.1179 gave 0.2668 CO, and 0.0758 H,O. C = 61-7 ; H = 7.1.0.1182 ,, 14.6 C.C. N, (moist) a t 20° and 753 mm. N=14*1.0.11794 ,, 0.1084 AgI. OMe = 8.0.C2UH2,04N4 requires C = 62.2 ; H = 6.7 ; N = 14.5 ;OMe = 8.0 per cent.Analysis of this product gave some trouble, since i t was difficultwithout extraordinary precautions t o bring about complete reduc-tion of the easily liberated nitric oxide.The results appear t o indicate that the substance is in all proba-bility a nitronitrosodeacetylaspidospermine, and its formationperhaps points to the presence in the nlolecule of a reduced quino-line nucleus.The substance undergoes decomposition by boilingwith alcohol, but the isolation of the corresponding nitro-derivativehas not so far been accomplished. I n the presence of acids morecomplex decomposition appears t o take place.Nitroiaitrosodeacetylaspidospermilze ( 1 ) forms pale yellow prisms,melting and decomposing a t 155-156O. It is very readily solublein acetone o r pyridine, less readily so in alcuhd, and very sparinglyso in water.When dis-solved in concentrated sulphuric acid or warmed with dilutemineral acids a brilliant reddish-purple solution is produced. Withphenol and sulphuric acid it gives a dark green solution, which ondilution becomes deep red, With ferric chloride solution no colouris produced.Boiling hydriodic acid (D 1.7) converts nitronitrosodeacetyl-aspidospermine into aspidosine.Its aqueous solution reacts faintly acid.Oxidution of Aspidospermine with Chromic Acid.Two grams of aspidospermine were dissolved in 40 C.C. of dilutesulphuric acid (25 per cent. by weight). Three grams of chromicacid were gradually added in small quantities to the boiling solu-tion, and the mixture was then boiled for five hours.The resultingsolution was treated with hot saturated barium hydroxide solutionuntil distinctly alkaline. The hot solution was filtered underpressure, the precipitate repeatedly extracted with boiling wateruntil the extracts no longer gave alkaloidal reactions, and the' com-bined filtrates were completely freed from sulphuric acid andbarium. The aqueous liquid was then conceatrated to small bulk,made faintly acid towards Congo-red with hydrochloric acid, andfurther concentrated until crystallisation commenced. The solutionwas aet aside, and after some time the crystalline product wascollected and recrystallised from absolute alcohol. The productproved t o be the hydrochloride of a base which was set free' b974.8 BRADBURY AND WEIZMANN :ammonia and crystallised from ethyl acetate in stout prisms, melt-ing a t 192-193O. The base is very sparingly soluble in water, butfairly readily so in alcohol, ethyl acetate, or chloroform. It givesvery marked i.eactions with alkaloidal reagents, but gives none ofthe colour reactions characteristic of aspidospermine or its deriv-atives already described. The' yield is only about 5 per cent. ofthe weight of aspidospermine employed, arid on this account a fullinvestigation of the base has not so far been possible.The hydrochloride forms plates which melt a t 286--287O, and issparingly soluble in water or in cold alcohol :09956 gave 0.2104 CO, and 0.0706 H,O. C = 60.0 ; H = 8.2.0.1322 ?, 0.0638 AgC1. Cl=ll.S.C,,H,,O,N,,HCl requires C = 59.9 ; H = 8.3 ; C1= 11.8 per cent.From these results the most probable1 formula for this basewould appear t o be C,,H2,0,N,, but this cannot be said t o beestablished with any degree of certainty a t present.WELLCONE PHYSIOLOGICAL HRSP,.~ I:CII I,ABORATORI IV,HERNE HILL, S E