ARMSTRONG ON THE ACTION OF SULPHURIC ACID V.-On the Action of XukT~uricAcid on the Natural Alkaloids By HENRYE. ARMSTRONG. THEexperiments the results of wliich are given in this paper form part of a series commenced in conjunction with the late lamented Dr. Matt hies s en. The action of sulphuric acid on the natural alkaloids was first made the subject of investigation by Arppe who in 1845 shortly described a body obtained from morphine to which he gave the for~i~la-4((C3,H,,N,06) + 550,. Ilaur ent and Gerhardt (Ann. Ch. Phys. [3] xxiv. 112) led by their researches on the Amides and Anilides to doubt the cor-rectness of this formula-which if true would imply that a reaction had taken place entirely without analogy-reinvesti- gated the matter further including both morphine and narcotine in their experiments.From the results obtained they con-sidered the cornposition of the bodies which they succeeded in preparing to be' represented by the f'ormuh C,,H,,N,O,S and c46H,,N,0,,S and termed them respectively suZp7iomorphide and suZphonamotide looking upon them as derived from the neutral sulphates of the bases by abstraction of the elements of 2 mols. OH, They considered them in fact to be true amides and as bearing the same relation to the neutral sulphates of the respective bases as sulphamide and snlphanilide bear to the neutral sulphates of ammonia and aniline. They were unable also to separate a base from the narcotine compound by the action of potash and this seems to have flirther strengthened them in t*heir opinion as to its constitution.By the action of sulpliuric acid on codeine they obtained a body identical with it in composition to which they gave the name of modified or amorphous codeine. The light thrown on the constitution of the organic bases by the researches of Dr. Matthiessen and others rendered it highly improbable that the above-mentioned bodies were con- stitutedas supposed by Laurent and Gerhardt and even allowed a speculation as to their probable nature for the dis- ON THE NATURAL ALKALOIDS. covery of apomorphine at once seemed to point to the possi- bility of the morphine product being none other than the sulphate of that base especially as the description of its pro- perties agreed moderately well with those characteristic of apomorphine.Their identity was proved by Drs. Matt hies sen and Wright! who succeeded in isolating npomorphine from the product of the reaction of dilute sulphuric acid on morpliiue in sealed tubes at looo. This rendered an analogous constitu- tion of the narcotine and codeine products probable a surmise which has however not been confirmed by experiment. On heating narcotine with an excess of acid of the strength of equal volumes of ordinary concentrated sulphuric acid and water in an open porcelain dish on a water-bath it is observed that the colour of the mass gradually darkens and that after a time and almost suddenly the whole becomes of a dark pink colour. It is t,lien thrown into a considerable quantity of warm water in which it entirely dissolves and a slight excess of ammonia is added which throws down the base as an amor-phous almost white precipitate.This was brought on a filter and washed with warm water. It was found to be very easily soluble in alcohol insoluble iii carbonate of soda readily soluble in potash and when heated under water to cake together in the form of a sticky semi-fluid mass. This behaviour at once Buggested the probable nature of the base; for all the above properties agree with those exhibited by the first product of the action on narcotine viz. dimethylnornarcotine ; and a further examination soon proved this to be really tlie case. The product was purified by solution in hydrochloric acid; addition of sodic carbonate in excess to remove any monome- thylnornarcotine which is soluble in the latter reagent ; redis-solution in hydrochloric acid ; and addition of potash in excess which dissolves dimethylnornarcotine leaving any nornarcotine or narcotine undissolved.The alkaline solution was then neutralised with hydrochloric acid and ammonia was added ; the base so obtained and dissolved in hydrochloric acid preci- pitated by sodic carbonate was extracted with ether ;the ether sliaken up with hydrochloric acid ; the hydrochloric acid solu- tion precipitated with ammonia; and the base after washing dried at 100' and analysed. 1. *4917 grm. gave -2332 grrn. OH and 1.1397 grm. CO,. 2. 0366grm. gave ,1724 grm. OH and 08455 grm. CO,. 58 ARMSTRONG ON THE ACTION OF SULPHURIC ACID Calculated.Found. v-/-I. 11. c, ..... 252 63.1G 63.2 63.1 H, ...... 21 5.26 5.2 5-24 N.. ...... 14 3-51 -0 ...... 112 2S.07 -C,,H,,N@ 399 100*00 The above aiinlyses prove beyond doubt the identity of the base with dime tli yhor :I m a 0 tine. The conversioii of the nnrcotine was pcrfect and if the reaction M as arrested immediately 011 the appearance of the pink coiour throughout tlie ~lde aii almost pure product iiic~xs was obtained as is shown by tlie following aiialj-sis of the crude product obtained by rlii*ectly precipitating tlie base with ammonia and washing ivith warm water :--54 grrn. gave -265 grm. OH and 1.247 grm. CO,; whence C = 63.0 per cent. H = 5.4 per cent. Tlie reaction has therefore taken place according to the equation :-C,,H,,NO + H,SO = C,,H,,NO + CH,HSO,.The methylsulphoviiiic acid formed is again decomposed into sulphuric acid and methyl alcohol for on adding bark carbonate in cxcesx filtering and evaporating to dryness no baryta salt wits obtaincd. For the complete conversion of 50 grms. narcotiue I have found it necessary to heat between two and three hours. If the heating be continned beyond the above-mentioned point a second atom of methyl can be removed arid probably a third; but a very impure product is obtained sulphurous anhydride is given off and partial carbonisation takes place. By the analysis of a crude product obtained by further heat- ing &c. the following results were obtained :-,4125 grrn.gave -1945 grm. OH aid *937 grm. CO,; there-fore C = 61.9 per cent.; H = 5.2 per cent. It is probable that the substance analysed by Laurent and Gerhardt was a mixture of the sulpliates of di- and mono-metliylnornarcotine. That they should have been led to consider it an arnide and not the sulpliate of a base evi- ON THE NATURAL ALKALOIDS. 59 dently arises fi-om the fact that dinietlilvliioriiarcotine is soluble in potash and that on the addition of an acid the respective salt is precipitated and only dissolved again on addition of a large quantity of water. On treating their product with nitric acid they obtained a yellow substance soluble in ammonia andbf course were able to detect sulphiiric acid in the solu- tion whereas if the pure substance prepared as above is similarly treated the yellow body is obtained which is per- haps a nitro substitution product but not a trace of sulphuric acid can be detectcd.On treating codeine in a similar manner and heating until the precipitate produced by sodic carbonate did not visibly increase; then dissolving the product in water) precipitating with sodic carbonate redissolving in hydrochloric acid and repre- cipitating with sodic carbonate which operation was twice re- peated in order to remove any codeine,-then extracting with ether and shaking up the ether with concentrated hydrochloric acid a crystalline hydrochlorate was obtained which on analysis gave numbers agreeing very well with those calcu- lated for hydrochlorate of codeirie.I. *517grm. gave 0309 grm. OH + 1.2162 grm. CO,. 11. -5725 grrn. , *3423grm. OH + 1.3558 grm. CO,. 111. -2257 grm. , -0905 grm. AgC1. Calculated. Found. I 7-I. 11. 111. c, .... 216 64-38 64-15 64-58 -H, .... 22 6.26 6.64 6-64 -NO .... 62 --I c1 ... . 35.5 10.5 -10.4 A platinum salt was prepared and analysed with like result. 0552grm. gave *lo67Pt = 19-32 per cent. Pt. BC,,H,,NO,HCl,PtCl requires 19.5 per cent. The first product of the action of sulp1iur.k acid on codeine is therefore a body isomeric therewith. 11s properties are as follows :-The base fhlls down as a snow-white amorphous precipitate on the addition of sodic carbonate. (Codeine is only precipitated after some time from concentrated solutions and always crystalline.) The hydrochlorate crystallises in groups 60 ARJTSTRONG ON THE ACTION OF SULPHURIC ACID ETC.of apparently hexagonal pyramids radiating from a common centre and differs from the corresponding codeine hydrochlo- rate by losing both its molecules of water of crystallisation at loo",whereas the former loses only 4 at loo" the remaining only being expelled at 120". The platinum salt is quite amorphous and may be obtained anhydrous by drying at loo" whereas the platinum salt of codeine is crystalline and does not lose the whole of its water of crystallisation below 100". The amor- phous platinum salt contains 1mol. water of crystallisation which it loses even by prolonged exposure over sulphuric acid. Gerhardt's description of the base &c.differs from mine iriasmuch as he says that not only the base but all its salts are amorphous. By the further action of sulphuric acid I have evidence that first 1mol. OH is removed from '2 mols. of codeine an inter- mediate compound between codeine and apocodeine being produced a sort of anhydride in fact; and then 1mol. OH from 1mol. codeine; and I believe that by its flirther action on this apocodeine apomorphine is formed i.e. CH is removed. I am however not yet certain as to the exact conditions under which the first of these changes takes place and the products being amorphous it is extremely difficult to obtain a satisfactory separation. With regard to morphine I have as yet unfortunately not had sufEcient quantities of material to work on to enable me to speak with certainty but I think it probable that it is first converted into such an isomeric modification as codeine and that also intermediate compound exists between it and apomorphine and derived by abstraction of OH from 2 mols.of mor-phine. By the action of sulphuric acid on other bases I have ob-tained entirely negative results ; neither strychnine nor Papa-verine is 111 the least affected by it even after many hours' digestion. It will be interesting to extend the investigation to those alkaloids known to contain the methyl group such as brucine and caffeine ;and with these I have already commenced experi- ments. PAGES ARE MISSING FROM 61 TO 84