876 ROBINSON A THEORY OF THE MECHANISM OF THE LXXV.-A Theory of the Mechanism of the Phyto-chemical Synthesis of certain Alkaloids. By ROBERT ROBINSON. ALTHOUGH in recent years largelly owing t o the investigations of Pictet and his collaborators there has been a due recognition of the importance of the r6le played by formaldehyde in t h e produc-tion of alkaloids in plants and although it is generally admitted (compare Winterstein and Trier " Die1 Alkaloide," pp. 263-317) that the amino-acids and carbohydrates are the most probable starting points f o r the) majority of phytochemical syntheses yet little pirogress has h e n madel in ascertaining the nature of these prwessm or even in the less ambitious task of formulating possible mechanisms based on laboratory analogies.The details of th PHYTOCHEMICAL SYNTHESIS OF CERTAIN ALKALOIDS. 877 schemes which have been suggested with but few exceptions, involve reactions for which little or no parallel exists in synthetical organic chemistry under conditions approximating to those obtain-ing in a plant. Thus Pictet’s view of the mechanism of the s p -thesis of nicotine was founded on observations of pyrogenic reac-tions of pyrrole derivatives and Winterstein and Trier seek to replace this hypothesis by another according to which the base results from the oxidation of a mixture of pyridine and N-methyl-pyrrolidine (Zoc. c i f . p. 298). Similar ideas have been advanced by Windaus and Knoop ( B e i f r . cliein. Pliysio!. Pnth. 1905 6 392) t o explain the formation of xanthine by the oxidation of a mixture of carhzmide and methylglyoxaline and of histidine by an analo-gous process applied to glycinel and methylglyoxaline.There has thus been a tendency to explain the results observed by the assumption that plants have a t their command enormously powesful reagents that are able to cause substances the properties of which have been investigated with considerable care to undergo transformations which cannot be induced in the laboratory. To a certain extent and especially in regard t o oxidation and reduction, this must be true but i t is probable that this aspect has been exaggerated and that an equally important cause of the variety and complexity of syntheses in plants resides in the highly reactive nature of the substances which function as intermediate products.The point of view reached in the present coinmunication is due t o a development of ideas which owed their inception to the hypo-thesis that the synthesis of tropinone recently described (this vol., p. 762) on account of its simplicity is probably the method em-ployed by the plant and confirmation of this theory was sought and found in the structures of hygrine and cuschygrine which stand t o one another in the same relation as styryl methyl ketone (benzyl-ideneacetone) to distyryl ketone (dibenzylideneacetone). I-Iaving found i t desirable to proceed irom ornithine in order to reach the bases of the pyrrolidine group it was obviously interesting to irquire as to whether similar methods applied t o the homologous lysine would lead to naturally occurring piperidine compounds.This proved to be the case and the investigation was then extended to include the more important of the alkaloids the constitutions of which have been determined. In the schemes given in the sequel, linking of carbon to carbon is traced to two processes only namely, the aldol condensation and the very similar condensatioii of carbinol-amines resulting from the combination of a11 aldehyde or ketone and ammonia or an amine and containing the group * ~ ( O H ) * ~ J * with substances coil taiiiing the group *~H*co*. The latter reaction has been investigated chiefly in connexio 878 ROBINSON A THEORY OF THE MECHANISM OF THE with cotarnine and similar pseudo-bases and the production of anhydrocotxmineacetone (I) may be cited as a typical example : $&H,O,-F H*OH C,H,O,-Y H*CH,*CO*CH, CH,*CH,*NMe CH CH,*NMe +CH,*CO*CII -.+ I (1.1 The condensing agent employed by Lieberniaiiii and Kropf (Bey., 1904 37 211) in this case is unnecessary and the reaction proceeds t o completion in aqueous solution and a t the ordina’ry temperature.111 fact these pseudo-bases are substances which enter into a variety of condensations with the greatest facility and a largo number of substances have been prepared in this way and in most cases in good yield. Employing these admissible methods i t is possible in each ir;stance to obtain the alkaloid skeleton and the further modifica-ticns are usually made by means of oxidatioiis or reductions and by elimination of water with the formation of an aromatic nucleus or occasionally of an ethylene derivative.The more important star t-ing points employed are ammonia and formaldehyde ornithine (arginine) and lysine and degradation products of carbohydrates. Of the latter citric acid is suggested as the source of acetone residues which i t supplies in the form of acetonedicarboxylic acid as the result of oxidation. Seekamp has observed the formation of acetme ( A m a l e ? ~ 1893 278 374) by the photochemical decom-position of a 5 per cent. aqueous solution of citric acid containing 1 per cent. of uranium oxide’ and it can scarcely be doubted that the dicarboxylic acid was an intesmediate product. The author does not however wish t o eniphasise unduly this theory of the source of the acetone residues and i t is interesting to note that Lippmann (Bey.1893 26 3057) found that acetonedicarboxylic acid wm obtained during the spontaneous decomposition of calciunl txisawharate. Further. a reactive acetone derivative may be found in diacetylacetone or other ‘polyketen,’ the formation of which iii plants has been discussed by Collie (T. 1893 63 329; 1907 91, 1806) and in that case the acetyl groups would be removed by hydrolysis subsequent to the condensations. However the occur-rence of the carboxyl group in ecgonine suggests t%hat in the synthesis of cocaine i t is a carboxylated acetone derivative which is the forerunner of the alkaloid. Except in the cases of hygrine and tropinone the carboxyl groups have been omitted from the acetone rests for the sake of simplicity in representation.As a starting point for both quinoline and isoquinoline bases i t has been found convenient t o assume the intervention of acetylglycoll-aldehyde (11) or ha-butene-ap-diol-y-one a substance which has not yet been isolated. It may readily be derived from a pentose o PHYTOCHEMICAL SYNTHESIS OF CERTAIN ALKALOIDS. 879 methylpentose by loss of water and oxidation as shown below and i t is perhaps significant that a methylpelztose quinovose occurs in a state of combination as quinovin in cinchona-bark. OH*CH,-CH(OH)*CH(OH)*CH(OH)*CHO 1 CH,*CH(OH>*CH(OH)*CH(OH)*CH(OH)*CIIO - HzO 1 C H *C 0 C H (0 H ) - C H (0 33 ) C H 0 d. jlOxidat ion I \/Oxidatjon The Pyrrolidiiz e Group. It has recently been demonstrated by Hess (Bey.1913 46, 4104*) that the methylation of an amine with the aid of form-aldehyde is accompanied by oxidation amino-alcohols yielding rnethylaminoketonea The methylating and oxidising action of formaldehyde on ornithine might theref ore yield a carbinol-amine of the pyrrolidine series in accordance with the equation NII,*CH,*CH,*CH,*CH(NH,)*CO,H + CH,O = NHMe*CH,*CH,*CHO -+ ~ ~ ~ * C H ( ~ ~ ~ > N M e ] + NH + GO,. [ Naturally the possibility is not excluded that the oxidation of the amineacid .f- is preceded by the formation of a hydroxy-acid, and ammonia and alcohols may in all cases be intermediate between amines and the aldehydes obtained by oxidation. Further oxidation accompanying methylation might attack both ++ I n which a complete list of references to earlier work by Eschweiler and others will be found.t Dakin ( J . Bid. Chern. 1906 1 171) has applied Fenton's method of oxidation by means of hydrogen peroxide and a trace of ferrous sulphate to the ammonium salts of the amino-acids and has obtained aldehydes, ammbnia and carbon dioxide. Hurtley and Wootton (T. 1911 99 288) find that alloxan effects the same change which had already been observed in the particular instances of the conversion of alanine into acetaldehyde and of leucine into isovaleraldehyde by Strecker (Annalen 1862 123, 363). Quite recently Schweitzer (Biochem. Zeitsch. 1916,78,37) has shown that the tyrosinase from potatoes can oxidise glycine with the formation of form-aldehyde carbon dioxide and ammonia. The ieaction takes place in the presence of an alkali preferably calcium hydroxide and appears t o be accelerated by chlorophyll 880 ROBINSON A THEORY OF THE MECHANISM OF THE ends of the molecule with the production of succindialdehyde -E and methylamine as shown below : NH2*CH2*CH2*CH,*CH(NH2)*C02H + 2CH,O = CHO*CH,*CH,*CHO + CH2*CH(oH) bH2.oH)>h TMe ] + 2NH,hle + CO,.After condensation with acetonedicarboxylic acid and elimina-tion of carbon dioxide hygrine (111) cuschygrine (IV) and tropinone (V) are obtained. The synthesis of the latter base (this vol. p. 762) by means of the reaction here assumed t o occur in nature was accomplished in dilute aqueous solution a t the ordinary temperature. Willstatter (Bey. 1900 33 1161) has already given reasons for supposing that the synthesis of atropine and its con-geners and of cocaine is preceded by that of tropinone and i t is now possible t o add that the carboxyl of cocaine f- is the result of partial decomposition of tropinonedicarboxylic acid possibly occasioned by the formation of the methyl ester or of an acid salt.I n the latter case the process resembles thht which Bandrowski employed for the preparation of propiolic acid from acetylenedi-carboxylic acid (Bey. 1880 13 2340). The question of the synthesis of benzoic and tropic acids is not examined here since there is no evidence from accompanying hydroaromatic compounds as to how these may be derived from the carbohydrates. * Succindialdehyde would be more readily obtained from a diaminoadipic acid which does not however appear to have been isolated.t The simplest substance produced on the cocaine model is probably arecoline a comparison of the formula of which with that of cocaine shows that the only divergences which it is necessary to postulate in the synthesis of the former from that of the latter are the employment of two molecules of formaldehyde instead of one of succindialdehyde and loss of water instead of benzoylation. CH*OBz CH,-CH, Cocaine. Rrecoline. Trigonelline. Arecoline also closely resembles a hypothetical intermediate in the synthesis of nicotine. If it is oxidised as was assumed for tha.t more complex tetra,hydropyridine then the bet.aine t,rigonelline would be obtained PHYTOCHEMICAL SYNTHESIS OF CERTAIN ALKALOIDS. 881 NMe 70 H CH,. CO*CH,*CO,H -+ /\ TH2 YH*OH CH,-CH, NMe NMe /\ 7OZH /\ $!H2 QH*CH*CO*CH,*CO,H -+ QH QH CH,*COMe CH,-CH2 CH2--CH2 (VI.) (111.) NMe 70,H FOpH NMe CH,*CO*CH + HO-$X€ QH2 YHz QH-OH CH,-CH CH,-CH, /\ NMe NMe I i Reduction i J.J. C H-CH* CO,K (Ale) Tropine $-Tropine j 4 ' I Ecgonine I Hyoscyamine Tropa-UH2 I I atropine etc. cocaine. CH2 I I I NMeC'O Renzoyl ecgonine Reduction .1 t-\ i Cocaine .i. \&f+-&f, The condensation product (VI) which forms the source of hygrine may also be the progenitor of nicotine (VIII) and the reactions necessary involve condensation with formaldehyde and ammonia t o a piperidone (VII) containing the nicotine skeleton, after which there are alternative ways of expressing the remaining stages. M M 582 ROBINSON A THEORY CO CO,H Nhle /\/ /\ CO,H*CH C'H-VH QH, CH,O CH,O CH,-CH2 NH, OF THE MECHANISM OF THE -t N Me CH-OH (VTI.) I -H*0 CH 4 (VIII.) The Piperidiize Group.Starting from lysine * (IX) and proceeding as illustrated above in the case of the lower homologue ornithine the formation of the bases X and XI would be anticipated. The latter is the alkaloid $-pelletierine isolated by Ciamician and Silber and by Piccinini from the root-bark of the pomegranate tree and since tropinone has not yet been obtained from natural sources the occurrence of its ring homologue is particularly valuable evidence. The base X has been synthesised by Hess Merck and Uibrig (Ber. 1915 48, 1886) and later identified by Hess and Eichel (Ber. 1917 50, 380) with a methylpelletierine occurring in small relative amount in the alkaloids of the pomegranate tree.CH, / \ O NH,*CH,bCH,*CH,*CH,*CH(NH,)*CO,H y H 2 YH2 (IX.1 OH CH*CH,*COAle \/ * Clearly the source appears in piperine. & Me (X.1 CH,-CI-X-CH, b H 2 kMe 30 UH,-UH-CH I I I (XI.) through cadaverine of the piperidine rest whic PHYTOCHEMICAL SYNTHESIS OB CERTAIN ALKALOIDS. 883 The frequent concurrence of closely related alkaloids such as the pelletierines++ or coniine and its associates is probably t o be explained by variations of sane primary product due t o alternate hydration and dehydration o r by oxidation and reduction. I n this connexion a highly significant discovery has been reported by Hess Eichel and Uibrig (Ber. 1917 50 351) who find that the reaction between amino-alcohols and formaldehyde is reversible.The methylaminoketone (X) for example is decomposed in alcoholic soIution by semicarbazide acetate with the formation of ths demethylated amino-alcohol (XII) and f ormaldehydesemi-CH CH2 /\ /\ \/ \/ ' (X) + p 2 QH2 - - + I CH2 p 2 CH CH*CH,CH(OR)*CH CH CH*CK,*CH,*CH, NIT (XII.) NH j (XIIL) CH2 /\ \/ p 2 I C Li c! H* C H2*UH,* C €3, I I NMe (XIV.) J. CIT, /\ \/ 7% ?"a ,,p2 FH2 -3 CH CH*CH:CH*CH CH CH*CH(OH)*CH,*CH, N €I \/ NH WV.1 CH, /\ \/ _ Y H 2 EH /F2 ( p 2 QHz CH C:CH*CH,*CH CH C*CH,*CH,*CH, NH \-/ N €1 (XVI.) * The imino-ketone pelletierine containing one reactive methylene-group (Hess and Eichel Zoc. cit.) has not yet been completely investigated but on the evidence so far recorded and from analogy i t should be the ketone corresponding with conhydrine or more probably the 3-piperidone obtained ky addition of water to coniceine followed by oxidation.In either case i t will fit into the lysine-acetone scheme. M M" 884 ROBINSON A THEORY OF THE MECHANISM OF THE carbazone. This suggests not only that oxidation accompanies N-methylation and 0-methylation but also that reduction is associated with N-demethylation. Coniine (XIII) N-methyl-coniine (XIV) conhydrine (XV) and coniceine (XVI) represent on this view modifications of the amino-alcohol (XII) and therefore of the normal product (X) from lysine formaldehyde and a reactive acetone derivative. I n the above and also in some of the sections which follow the reduction of alkylamines has been assumed and it may be pointed out that there is evidence that these substances are reducible the reaction probably depending 'on the intermediate formation of an unsaturated or cyclic ammonium hydroxide.Condensations between formaldehyde ammonia and a reactive acetone derivative such as acetonedicarboxylic acid can explain in an astonishingly simple manner the formation of alkaloids con-taining the curious quinuclidine ring system. Ammonia three molecules of formaldehyde and the acetone derivative lead to the formation of the piperidone (XVII) two molecules of which enter another molecule of acetone derivative producing XVIII. This, owing to the stability of the ring system produced undergoes internal aldol condensation and complete reduction then results in sparteine (XIX).00 /\ CO,H*CH CH,*CO,H CH,O CH,O -+ NH8 OH,O co /c* co / I I \ I l l I 1 I CH CH CH,*G'O*CH CH CH, CH OH CH, '\/ (XVIII.) I . PHYTOCHEMICAL SYNTHESIS OF CERTAIN ALKALOIDS. 885 \A/ N (XIX. 886 ROBINSON A THEORY OF THE MECHANISM OF THE CH /'\ / i \ CH CH CH*CH(OH)*CH, 5 2 $ z I CH CH CH*CH:CH, I l l CH(OH)*CH CH CH, \/\/ + N Ccl CO*CH C (0 €I ) - C (0 H ) C H , / / I i / ' I l l - - - ? I 1 I CH CH CH l~edilctiolt CH CH (311, +-CH(OH)*CH CE CH +-CH(OH)*CH C'H CH, N (XXII, CH-C( 0 TI)* C H C(OH)*CH.CH, / CH C'H CH /' I l l o r l l / CH CH C'H - H20 +-CH(OH)*CH CH CH +CH(OH).CH CH CH, \ ! ' \/ N PRYTOCHEMICAL SYNTHESIS OF CERTAIN ALKALOIDS.887 the origin of cinchonine and quinine. The quinoline derivative (XXI) enters the molecule in the manner assumed throughout that is by a carbinol-amine condensation and the remainder of the reactions are of the type already formulated in the case of sparteine. An alternative for the later stages is presented in the scheme which is given below. It is possible that reduction of the diketone (XXII) produces first a n internal pinacone (compare Kipping and Perkin T. 1891 59 214) which after further reduction, undergoes a transformation of the ring system accompanied by elimination of water. With regard to the genesis of the quinoline derivatives required for cinchonine and quinine i t is natural to seek an explanation from the constitution of quinic acid (XXIII) which accompanies these alkaloids in the plant and is readily converted into quiaol derivatives.The coincidence that quinine is also in a sense a derivative of p-aminophenol strengthens the conviction that quinic acid is the source of the quinoline half of the molecule. I n any case it' would seem necessary that the acid should become oxidised t o the ketone (XXIV) so as t o provide the six-carbon system and admit the entry of the nitrogen atom. Further a reactive CH(OH)*CHO I CH*OH ClI-OH CO /\ /\ \ \ CH, /' \ / \ i I I I \ / \\ / CH-OH CH CH-OH UH, CII C(OH)*CO,H CH CO -+ CH,O \/ CH-OH NET, \/ CH*OH (XXIII. ) (XXIV. ) CH(OH)*CHO 1 CH t CH, \ / / CH*OH 888 ROBINSON A THEORY OF THE MECHANISM OF THE CH( OH)*CHO I c1 I c I I I I VHZ0 CH (OH) * CHO I /\A I 1 \A/ N inethylene group is so CH (OH).CHO I CH ir HO*CH,*OCH C CH I I I + CHzO -+ CH N I j -CH(OH)*CHO I /\/\ \/\/ Meal I I N produced enabling the occurrence of the condensation which closes the quinoline ring.This process is formulated below and the heterocyclic nucleus is obtained by a reaction similar to that by means of which quinaldine may be syiithesised from o-aminobenzaldehyde and acetone. In the majority of the complex examples discussed it is obvious that the order in which the reactions may be supposed to occur can be considerably varied without fundamentally altering the character of the suggested processes and in the case of quinine there are several plausible variations of this kind the most important of which is perhaps that the quinoline ring may be closed immediately after the appearance of an acetyl group in a modified carbohydrate such as quinovose and the product inust then suffer further degradation by oxidation of the side-chain.The isoQzcinoline Group. I n the discussion of this large and important group it will clear the ground t o construct’ in the first place a table which it is con-sidered represents the genetic relationships of the more important members PHYTOCHEMICAL SYNTHESIS OF CERTAIN ALKALOJDS. 889 Carbohydrate degradation products + ctnimonia 4 .1 4 Hydroaromatic base (HB) + Hydroaromatic aldehyde (HA) Aromatic base (AB) -++ Aromatic aldehyde (AA) (HB+HA-H,O) Morphine Bulbocapnine is o T h e bai n e 1 Codeine Corytuberine etc.+Thebaine (HB + AA - H20) 4 Glaucine dicentrine (AB + AA - H20) + Norcorydaline + Norlaudanosine -+ Laudanine laudanosine, I I papaverine xanthaline 3. 3. hydrastine Corydaline Norcanadine 4 Csnadine berherine protopine cryptopine. (HI3 + C!H,O) + Pyrogallol derivatire + *4A - H,O 4 Narcotine narceine. [?( AB + HA) -+ Hydrastine norcanadine etc. corydaline.] A. The Hydroaromatic base (XXV) and Aldehyde (XXVI).-The clue to the nature of the hydroaromatic substances from which the isoquinoline bases are derived is given by the constitution of morphine and it is regrettable that general agreement on this subject has not yet been reached. However it is only the “Pschorr formula’’ which it has been found possible to dissect in such a manner as to show a relation with the other opium alkaloids and this formula has been adopted especially since the formuls which it is suggested should replace it cannot without hesitation be accepted as superior expressions of the properties of the substance.By working back from morphine a scheme for the synthesis of the isoquinoline alkaloids by means of aldol con-densations has been deduced and the preliminary stages startin 890 ROBINSON A THEORY OF THE MECHANISM OF THE from ammonia formaldehyde,* a reactive acetone derivative and acetylglycollaldehyde are represented below. CH,*CO*CH(OH)*CHO + CO,H*CO*CI-f,.CO.CO,)-E CH,*CO*CH(OH)*CH(0H)*CH2*CO.CH + CH,O + "I, 4 CO /\ HO*qH CH, H 0' 6H CO e C H CH N H, \/ CH2 co 4 - HzO co co /\ \/ Oxidation HO*S)H EH /\ \/ HO*FH YHz HO-CH C(OH)*CH,*CHO f- CH C~CH,,.CH,*NH, CH CH2 (XXVI.) (HA) 4 - IJpO OH I 3.O H HO/\ HO/\ I ~H,GH,*NH \/ I 'CH,*CHO (XXVII.) (AB) \/ (XXVIII.) (AA) * In all cases where formaldehyde is supposed to enter into a carbinol-amine condensation it is clearly equivalent to assume the intervention of glyoxylio acid or even of an aldehyde R'CHO in which tho group R is readily oxidisable to carboxyl after the condensation. (a) *CO*CH + CH,O + NH -+ *CO*CH,*CH,*NH,. (b) *CO*CH + CHOoC0,H -I- NH + *CO*CH,-CH(NH,)*CO,H. (c) *CO*CH,+CHO*CH,*OH (e.9.) +NH + The amino-acids formed in such reactions must be a-amino-acids and the elimination of carbon dioxide will therefore always be explicable.The ,+ ? *CO*CH,*CH (NH,)*CH,*O PHYTOCHEMICAL SYNTHESIS OF CERTAIN ALKALOIDS. 891 B. Formation of the Aromatic Conzpou~zds.-Simple dehydra-tion of XXV and XXVI leads t o 3 4-dihydroxyphenylethyl-amine (XXVII) and 3 4-dihydroxyphenylacetaldehyde (XXVIII) , and attention is directed to the peculiarity of the constitution of the hydroaromatic substances which renders it extremely improb-able that an orientation in the benzene derivatives other than that observed could result from the process. Oxidation of XXV would however produce a pyrogallol derivative and this oxida-tion might be the result of methylenation of the two adjacent hydroxyl groups preventing the removal of oxygen from the ring. The dihydrobenzene so obtained would then be oxidised to XXIX.co - HZO /'\ \/' CH, (XXV.) O * y c gH -+ ' H2<0 C €3 C C H C H N H C H 0 H OH (XXIX.) C. Elaboration of the Af orpiline Sub-yroup.-The hydro-aromatic aldehyde1 (XXVI) condenses with the hydroaromatic base (XXV) producing XXX,. XXXI and XXXII by a carbinol-amine condensation an internal aldol condensation and by elimination of water. A t this stage, in order to obtain mor-phine it is necessary to assume the! only reduction which is encoun-tered in the whole group. The! hydroxyl group marked ( a ) is in a position with respect to the nitrogen atom which renders plausible syntheses of histidine tryptophane and tyrosine probably involve reactions of this type. In the case of hiatidine the complete representation would be the following : NH QHO NH H'Co21CHoN€€3 CO*CH CHO*CO,€€ -+ In general the group 'CO'CH becomes *CO*CH;CH(N~)CO,H and the carbonyl group can then take part in aldol condensations leading to cyclic structures whilst the carboxyl group may be eliminated or the amino-acid be oxidised to an aldehyde.The adoption of such a view in the case of the forerunner of the isoquinoline alkaloids is favoured by the author but the simpler suggestion is retained in the text in order to avoid undue complexity of the formulae ('IXXX) 'H3 'H3 / \/\ ('XXX) 'HH3 'HD / \/\ "H? H0.b H?*OH HN HA H3*OH \ /\/ H 3 03 'H3 'H3 /\ /\ \/ \/ 'HQ H? H?*OH eJ" H 3 H3*OH ('IIXXX) 'H3 'HH3 - / \ / \ 'H? (")HO*? H?*OH H N H3 H3*OH \ / \ / \/\ 1 /OH \/OH ( 'AXX ) 'H3 "3 / \/\ 'HN 'HD HD-OH \/ 'H? HO*? H()*OH HO*H3 HO*H PliYTOCHEMICAL SYNT-HESIS 0 F CERTAIN ALKALOIDS.893 the1 assumption of the formation of a cyclic ammonium hydroxide and after the reduction morphine (XXXIII) is obtained by N-methylation and elimination of two molecules of water. 0-Methylation in addition provides codeine and thebaine is the result af still further methylation accompanied by loss of two atoms of hydrogen. D. The d rottintic I'hennnthrem ICi(b-,qr~trp.-The coinpound XXSII by symmetrical eliiniiiation of three molecules of water yields the phenol derivative (XXXIV) which by N-metliyla-tion and methylation of two phenolic liydroxyl groups be-comes isotliebaine (XXXV) which Klee (,trclr. Ylumn. 1914, 252 21 1) has sliown accompanies thebaine in Yopacer orien tale.HO(\ I f the hydroxomatic ring of XXXII becomes oxidised as may well happen simultaneously with the 11'-methylation of formalde-hyde one of the *CH(OH)- groups becoming *GO* then elimina-tion of water leads t o XXXVI which by methylation and methylenation provides numerous alkaloids of the corydalis-phen-antlirene type such as bulbocapnine (XXXVII) (Gadamer and Kuntze Arch. Pharm. 1911 249 598). I n glaucine (XXXVIII) and dicentrine (Asahina A mlr . Phcirm., 1909 247 201; Gadaiiier ibid. 1911 249 680) the orientation of the catecliol ether groups silggests that these bases are the resul 894 ROBINSON A THEORY OF THE MECHANISM OF THE of condensation of the aromatic aldehyde (XXVIII) with the hydroaromatic base (XXV) and the condensation which closes the phenanthrene ring may then produce a 4 5-substituted catechol derivati 76.Apart from this the reactions necessary will entirely resemble those required f o r bulbocapnine. E. NorlazLdaizosine and tJte Products obtaiued froin it by the A ction of 3'ornznldekyde.-Norlaudanosine (XXXIX) is the result; of condensation of XXVII and XXVIII as illustrated below (com-pare Wintersteia and Trier Zoc. cit. p. 307) : OH OH I 1 (XXVIII.) HO{) \/ CH2 CH2 I \/ I dH HO/\ I 1 (XXVIII.) CH2 dHO HO/) I y 3 2 CH (XXVII. ) (XXXIX.) Norlaudanosine is ?ub,ject to attack by formaldehyde a t six points and the1 nature of the substances produced largely depends on the amount of oxidation which acoompanies the methylations.( a ) Laudanosine, Com~lete methylation of phenolic hydroxyls and of the imino-CI6Hl7O4N 4- 5CHz0 = Cz1H270,N + 50. group unaccompanied by internal oxidation. (6) Papveri?ie, Metliylation of phenolic liydroxyl and oxidation of the tetra-C,,H,,O,N + 4CH,O = C,,H,,04N -t 2 0 + 2H,O. hydroisoquinoline ring. (c) X a n t 11 alin e (pa pi u e ra ldin e ) , Cl6HI70,N + 4CHz0 = C2,HI9O5N + 3HZO. (d) Hydrastiite, CI6Hl7O4N + 5cHzo = Cz,Hz106N + 3HzO. The essential reaction here is a " Lederer-Manasse " syntliesis, introducing ths group *CH,*OH into the benzene ring. The product then becomes siniultaneously metliylated and oxidised as indicated below PHYTOCHEMICAL SYNTHESIS O F CERTAIN ALKALOIDS. 895 ( e ) The Berb e r k e Sub-group.The establiskxnent of a bridge between the nitrogen and the benzene ring produces norcanadine (XL) and the reaction is of course that which Pictet and Gams (Co~npt. rend. 1911 153 386; Ber. 191 1 44 2480) devised for the synthesis of tetrahydrober-berine : From norcanndine the following are obtained by the action of f ormddehyde : ( a ) Canadine, Methylation of two phenolic hydroxyls and methylenation of C,7H,;O,N -t 3C'HZO = CzOH2I0,N -1- 2 0 + H2O. two more 806 ROBINSON A THEORY OF THE MECHANISM OF THE (8) Berbem'ne, C17H1704N + 3cH20 = C2,-,H1905N + 2H20. The sime processes as with canadine using the oxygen for the convesrsion of the tetrahydroisoquinoline derivative into a true alkylisoq iinolini1.m hydroxide. (y) Protopine, Complete methylenation of norcanadine yields XLI and the action of formaldehyde1 on this substance may produce protopine (XLII) (Perkin T.1916 109 875). C,;H,,O4N + 3CH20 = C,,H,,O,N -t 2H20. This is a reaction of a more hypothetical character than any other introduced in this communication but since a ring-scission in this direction of tetrahydroberberine methohydroxide has been estab-lished by Pyman (T. 1913 103 828) and since these results have been paralleled in the isotetrahydroberberine Eeries in connexion with derivatives of cryptopine (Perkin 7oc. cit. p. 841) there appears to be little improbability in the assumption and the details of the process may follow the partial scheme: I I I CH UH UH bH2 C!?H dH2 \/ \/ \/\' CH,*OH I NH 1 :<OH + \/\/' I N ' I * ' /\OH CH2 / CH, / UH, / (6) C'ryptopi?i.e (Perkin loc.cit. p. 831), This would result from norcanadinel by methylation and methyl-CI7Hl7O,N + 4CH20 = C2,HZ7O,N + 20 + HzO PHYTOCHEMICAL SYNTHESIS OE' CERTAIN ALKALOIDS. 897 enation to forin an isotetrahydroberberine which might yield the bzse by the fcrtlier action of formaldehyde as with protopine. F. Narcotine arid iliarceine.-3 4-Dihydroxyphenylacetaldehyde (XXVIII) and the base XXIX are supposed to condense with formatiov- of the methylene ether of hydroxy-fl-methylnorlaudano-sine (XLIII5 which is conveirked into narcotinel (XLIV) in pre-cisely the manner employed above in the case of hydrastine. The ring-scission of narcotine with the formation of iiarceiiie is similar to the production of protopine and cryptopine except that the further methylation of thel nitrogen is not accompanied by internal oxidation : OH ()OH \/ I C*, OH bH 0.CH,*OH /\O*CH,* OH ,/CH2 I *OH 4C!H2O I --+ HO*CH CH - H,O \ i -+ CH (XLIII.) OMe /\-OM9 1 I-CO \f I CH-o (XLIV.) G. Co3daline.-Notwithstanding the observation of Gadamex (L41.cl~. Phartti. 1915 253 274) that corydaline loses its optical ac-tivity when it is oxidised to didehydrocorydaliiie and that therefore the two asymmetric carbon atoms should be contiguous an alteration in the accepted formula of the alkaloid cannot be considered. Especi-ally is th3 position of the methyl group clearly proved by thel argu-ments of Dobbie and Lauder (T. 1902 81 154) relating to the methylpyridinetricarboxylic acid obtained by them from corydic acid, although a slight modification of the constitution of the latter sub-stance would now be made so as t.0 represent it as a true pyridine derivative and as a betaine (compare Perkin Zoc.cit. p. 836) 898 ROBINSON A THE!ORY O F THE MECHANISM OF THE Corydaline accordingly appears as a methylated norlaudanosine with an ethylidene bridge between the nitrogen atom and the aromatic nucleus a structure produced by the interven-tion of acetaldehyde instead of formaldehyde. Attention may be again directed to the possibility that what may be termed the hemipinic orientation may result from aldol condensation with the hydroaromatia progenitors of 1 liese alkaloids whilst the metahemi-pinic orientation is t o be expected if the condensations are subse-quent to the conversion of the hydroaromatic compounds into catecliol derivatives by loss of water.H. The Conditions affectirig t h e ProductioIL of the isoQuinoline Hnses.-The' problem may be put in the form of a question Why do not alkaloids assumed as above to have a common origin always occur together in the plant? Why for example does not morphine occur i n Hydrnstis caizudensis? It is reasonable to expect some answer to this but not in the circumstances a complete one. I n the parti-cular case mentioned the scheme shows that the molecule must suffer reduction before morphine can be obtained. That there are reducing conditions is clear also from the formation of laudanosine. On the other hand the formation of hydrastine and of berberine follows such a course that all the oxygen available from the methyla-tions is used internally and powerful reducing agents would there-fore appear to bs absent during the formation of t'hese bases.This would be sufti'cient to account for the non-production of morphine. Among other circumstancels that may affect the nature of the end-products are the stereochemical relations of the hydroaromatic sub-stances the rapidity with which these lose water with the forma-tion of aromatic compounds the stage a t which N-methylation occurs the concentration of formaldehyde and other reagents the presence o r absence of enzymes and other catalysts all of which are without doubt connected with the needs of the plant itself. If these varying conditions are taken into account the objection that the siniplicity of the schemes proves too much cannot well be sus-tained.Some of the more remarkable' coincidences encountered in the development of thel foregoing suggestions may now be noted in su-mmary and conclusion. In the pyrrolidine group hygriiie, cuschygrine and tropinone are related by their method of forma-tion and by introducing a piperidone sing into hygrine by means of a reaction entirely similar t o that required for sparteine i t is seen that ;ti1 explanation is provided of the &position of the pyrrol-idine complex in the pyridine nucleus of nicotine. In the piper-idine group the methods are homologous with those1 of the pyrrol-idine group and the nature of the reduced pyridine derivative REDUCTION OF ALIPHATIC NITRITES TO AMINES. 899 which have been obtained from plants is quite in harmony with ths origin from lysine. I n particular the a-propyl chain of coniine is explicable orL the assumption that it represents a reduced acetone residue. In the quinuclidine group the bridged rings of sparteine and quinine are produced by similar methods the application of which fixes the position of the connecting methylene group of sparteine and the vinyl group of quinine. Substances accompany-ing quinine anci cinchonine in cinchona-bark have been found to b2 suitable starting points f o r a synthesis of the quinoline-half of the molecules of these alkaloids. I n the isoquinoline group it is noteworthy that one and the same hypothetical hydroaromatic substance derived in part from a carbohydrate degradation product also required for quinine suffices f o r the production of the whole of the bases of the group the constitutions of which have been elucidated. It is hoped that it will prove possible to employ thiv theory as a working hypothesis in several directions. THE UNIVERSITY, LIVERPOOL. [Received July 23rd 19 17.