ORGANIC CHEMISTRY.1. STEREOCHEMISTRY.ONE of the ambitions of all stereochemists has been rcalised byP. Maitland and W. H. Mills,l who have shown that an allene canexhibit the optical activity required of it by established theory.They dehydrated ay-diphenyl- ay-di- 1 -naphthylallyl alcohol (I)with d- and with Z-camphor-10-sulphonic acid and obtained d-and Z-diphenyldinaphthylallene (11), the specific rotations (A 5461)having the surprisinglyetc.etc.Annalen, 1935, 520, 11; A , , 1378TURNER : HETEROCYCLIC COMPOUNDS. 345(XLI) was obtained either from 5-ethoxytryptamine and p-toluene-sulphonyl chloride or by the action of alcoholic zinc chloride andp-ethoxyphenylhydr azine on y -p -toluenesulphonamido but aldehydediethyl acetal (XL). Methylation then gave the N-methyl deriv-ative, which was alternatively obtained from y-p-toluenesulphonyl-met h ylamido bu taldeh yde die thy1 ace tal and p -e thoxyphenyl-hydrazine. Removal of the toluenesulphonyl group by the use ofaniline and aniline hydrochloride gave 5-ethoxy-N-methyltryt-amine, which, with ethylmagnesium iodide, followed by methyliodide, gave isonoreserethole.The hydrochloride of this wasconverted by methyl iodide into dl-eserethole.The same authors have also carried out the synthesis of bufotenine(XLII) from 5-ethoxyindole-3-acetonitrile, the CH,*CN group beingsuccessively converted into CH,*CO,H, CH,*CO,Et, CH,*CH,*OH,CH,*CH,Br, and CH,*CH2*NMe, by usual methods. The finalproduct gave bufotenine when de-ethylated with aluminiumchloride.Ergot A ZEaZoids .-Our knowledge of these alkaloids has increasedconsiderably in the last few years.For an account of the olderwork, reference may be made to a paper by S. Smith and G. M.T i m m i ~ . ~ ~ The work of Barger, Dale, Cam, and Kraft showedthat ergotoxine and ergotinine were isomeric and interconvertible,boiling methyl alcohol changing ergotoxine into ergotinine, and hotdilute alcoholic phosphoric acid solution effecting the reverse change.A. Stoll56 isolated a third alkaloid, ergotamine, which underwentisomeric change into a fourth alkaloid, ergotaminine, under con-ditions similar to those .governing the ergotoxine-ergotinine con-versions. This suggested a close relationship between ergotoxineand ergotamine, which was supported by the observation 67 that thephysiological actions of these two bases were qualitatively andquantitatively identical ; according to other aufhors,5* ergotoxinehad a biological activity 1.66 times that of ergotamine.Smithand Timmis, in the above paper, described the isolation of allfour alkaloids in a state of high purity, and, by examining ergot5 5 J., 1930, 1390; A., 1930, 1050.66 Schweiz. Apoth.-Ztg., 1922, 60, 341 ; A., 1923, i, 127; see also K. SpiroRnd A. Stoll, Verh. Schweix. Nat. Gw., 1920; A., 1922, i, 47.57 H. H. Dale and K. Spiro, Arch. exp. Path. Phapm., 1922, 95, 337; A.,1923, i, 420.5 8 E. Lozinski, G. W. Holden, and (2. R. Diver, J . Pharm. Exp. Y'her.,1931, 42, 123; A,, 1931, 871346 OMANICJ CHEMISTRY.from seven different European sources, they concluded that ergotof rye contains only ergotoxine and ergotinine, though a NewZealand ergot growing on Pestuca contained ergotamine. In 1931,59they recorded the isolation of a new alkaloid, +-ergotinine, probablyisomeric with ergotoxine.A.Soltys60 showed that all four alkaloids on treatment withboiling alcoholic alkali gave one molecule of ammonia, and thatoxidation of any of the alkaloids by permanganate and by nitricacid produced benzoic and pnitrobenzoic acid, respectively.Smith and Timmis,G1 following up this work, isolated a new,optically active base, ergine, Cl6Hl,ON3, by heating any of thefour alkaloids with alcoholic potassium hydroxide.At this point, an outstanding series of papers by W.A. Jacobsbegan to appear. He found G2 that nitric acid oxidation of ergotininegave also a tribasic acid, possibly C,,H,08N, and, with L. C. Craig,G3showed that ergotinine and methyl-alcoholic potassium hydroxidegave (1) an acid, Cl,W12N(NMe)*C0,€€, named lysergic acid, (2)isobutyrylformic acid, and (3) ammonia, whereas, on similar treat-ment, ergine gave an acid which was at first thought not to belysergic acid but mas later identified as the acid sulphate of thissubstance. Smith and Timmis, however, showed 64 that erginewas the amide of an acid, C15Hl,N,*C0,H, probably lysergicacid.cJacobs and Craig found that lysergic acid was reducible withsodium and amyl alcohol to dihydrolysergic acid,65 and thatergotinine,G6 when reduced with sodium and n-butyl alcohol, gave,in addition to one molecule of ammonia, seven products : (1)a-dihydrolysergol, ~16H2,0N2, containing m e and forming amonoacetyl derivative ; (2) p-dihydrolysergol ; (3) cc-hydroxyiso-valeric acid (formed from isobutyrylformic acid) ; and four bases,numbered 11, IV, V, and VI.Since reduction of methyl lysergategave (1) and (2) but none of the other products, it follows that thefour bases arise from that part of the ergotinine molecule which isnot lysergic acid, the two portions of the molecule possibly beingjoined by means of the isobutyrylformic acid skeleton. Theformation of two dihydrolysergols is possibly due to the introductionof a new centre of asymmetry during the reduction of lysergic acid.68 J . , 1931, 1888; A., 1931, 1171.6O Ber., 1932, 65, [B], 553; A., 1932, 629.61 J., 1932, 763, 1543; A., 1932, 526, 759.62 J.Biol. Chem., 1932, 97,739; A., 1932, 1147.63 Ibid., 1934, 104, 547; A., 1934, 538.64 Nature, 1934,133,579; A., 1934, 667; J., 1934, 674; A., 1934, 787.6 5 J . Biol. Chern., 1934,106, 393; A., 1934, 1116.66 Ibid., 1935, 108, 595; A., 504TURNER : HETEROCYCIJC COMPOUNDS. 347Later work G7 on the above four bases showed that base VI wasa phenylpropanolamine, suggesting phenylaJanine as its precursorin ergotinine. The same authors also identified, as a componentof the above reduction mixture, proline methyl ester, which wasalso formed by the action of methyl-alcoholic hydrogen chlorideon ergotinine. Bases I1 and IV were probably piperazine deriv-atives formed by reduction of prolylphenylalanine anhydride andproline anhydride respectively, and base V, probably a-pyrrolidyl-carbinol, was derived from proline or its ester. This showed thatergotinine, and therefore ergotoxine, are built up of lysergic acid,Cl,Hl,OzN, (or its amide, ergine, C16H170N3), proline, phonyl-alanine, and isobutyrylformic acid.Addition of these fragmentswith removal of four molecules of water gives C35H3905N5, whichis the empirical formula of ergotinine. At this point, Jacobs andCraig thought that lysergic acid, the pharmacologically importantcomponent of the alkaloids, was joined to a dipeptide composedof proline and phenylalanine.These authors G8 also made a preliminary study of ergotarnine.This was shown to give rise to lysergic acid, ammonia, and phenyl-alanine, but proline was not definitely identified.One importantdifference between ergotarnine and ergotinins was discovered :none of the degradation processes used led to the formation ofisobutyrylformic acid or of its reduction product, a-hydroxyiso-valeric acid. Since in empirical formula the two alkaloids differby CzH4, it was thought that in ergotamine (and ergotaminine)pyruvic acid might take the place of isobutyrylformic acid inergotinine.G. Barger,cQ who, with A. J. Ewin~,~O first obtained isobutyryl-formamide by the thermal decomposition of ergotoxine and ergo-tinine, obtained this amide, and ergine, from the ergoclavinedescribed by W. Kiissner. 71 Jacobs and Craig 72 found that ergotinine,when heated with hydrochloric acid, gave Z-phenylalanine and prolinemethyl ester, whereas similar treatment of ergoclavine, alkalinehydrolysis of which gave ammonia, lysergic acid, and isobutyryl-formic acid, gave Z-leucine but no proline.Development of thework on the alkaline hydrolysis of ergotinine 63 led t o the isolation,in addition to lysergic acid and isobutyrylformic acid, of a mixture,acid hydrolysis of which gave almost inactive proline and phenyl-Evidence for this was actually found.6 7 J . Amer. @hem. SOC., 1936, 57, 383; A., 504.f18 Science, 1935, 81, 256 ; A., 764.O9 Merck's Jahresber., 1933, 47, 12.7o J., 1910, 97, 290.71 Merck's Jahraber., 1933, 47, 5.72 J . Amer. Chern. SOC., 1935, 5'7, 960; A,, 872.73 J .BioZ. Chem., 1935, 110, 521; A., 1137348 ORQAMC CHEMISTRY,alanine. Hydrolysis of ergotinine with hydrochloric acid, however,gave Z-phenylalanine, what appeared to be a proline-phenylalaninedipeptide, and also d-proline methyl ester. Whiht, therefore, thephenylalanine has the usual configuration, that of the proline isunique in this instance, and, of course, a Walden inversion toexplain it is probably out of the question. This confirmed theformer conclusion that ergotinine and ergotoxine are composed ofproline and phenylalanine combined in peptide linkage with lysergicacid, isobutyrylformic acid, and ammonia.In their most recent paper,'* Jacobs and Craig show that all theergot alkaloids are derivatives of lysergic acid, in which the latteris conjugated with a-amino-acids or with substances derived fromthem.A return was made to dihydrolysergic acid, which washeated with potassium hydroxide at 300" in an atmosphere ofhydrogen. The products were : (1) methylamine, (2) a substancewhich was probably 2-methyl-3-ethylindole, (3) propionic acid,and four other substances. It being known that by reductionwith sodium and butyl or amyl alcohol lysergic acid gives a dihydro-derivative, and that ergotinine or methyl lysergate gives thecorresponding alcohol (as a- and P-dihydrolysergols), and that it isM c u l t to reduce an indole structure, it becomes highly probablethat at least one ethylenic linkage is present in another part of themolecule. This suggests a 4-carboline structure, and, since methyl-amine is formed almost quantitatively in the above decomposition,the basic carboline nitrogen atom probably carries the methylgroup and is the salt-forming part of the molecule.This gives thepartial structure (XLIII). To make up the 16 C of lysergic acid,there are also required a propyl or propylene side chain and acarboxyl group, these presumably being attached to the reducedring. Further, the lysergic acid molecule has two double bonds.CH CH,CH:CMMe CHXHMe(XLIII. ) (XLIV.) (XLV.)One is readily reduced and is probably a t C5-C, (compare harmineand harmaline giving tetrahydroharmine). The inferred relation-ship to tryptophan places the carboxyl a t C5, so lysergic acid wouldbe (XLIV). This formula accounts for the optical activity, andexplains the formation of the epimeric dihydrolysergols (XLV)from either ergotinine or methyl lysergate, since a new centre of74 J .Biol. Chem., 1935, 111, 465; A . , 1512TURNER : HETEROCYCLIC COMPOUNDS. 349asymmetry appears at C5. The second double bond must be, asshown, in the side chain, but the latter may be an ally1 group,undergoing the common bond-shift during alkaIi fusion. If boththe carboxyl and the propenyl group are at C,, it is more difficultto explain the epimeric dihydrolysergols. It may be observed thata-dihydrolysergol was shown to give a true onium salt with methyliodide.A surprising new chapter in ergot chemistry was opened byH. W. Dudley and C. M0ir,7~ who showed that the characteristicoxytocic effects of ergot were due, not to ergotoxine or to ergotamine,as had been thought, but to a new alkaloid, ergometrine, the extrac-tion of which, by an improved method, was described by Dudley.76M.R. Thompson 77 also obtained evidence that there was a newalkaloid in ergot other than ergotoxine or ergotamine. Again,M. S. Kharasch and R. R. Legault 78 isolated “ ergotocine” fromergot, and showed that under condit’ions (hot alkali) such as gaveone molecule of ammonia with ergotoxine, ergotamine, and “sen-sibamine ” (see below), no ammonia was formed. Lysergic acidwas the only product identified. The “ ergobasine,” C,,H,,O,N,,isolated from ergot by A. Stoll and E. Burckhardt 7, was shownto be lysergic hydroxyisopropylamide.80 M.R. Thompson 81expressed the ‘‘ belief ” that ergostetrine (his new alkaloid) wasidentical with ergometrine, ergotocine, and ergobasine, and H. W.Dudley 82 put forward the same view more definitely.=CH/\-/ I It II ~~*CO*NH*CHMe*CH,*OH\/\A / m e (XLVI.1NH QHCHXHMeThe validity of the arguments of Jacobs and Craig being assumed,S. Smith and G. M. the constitution of ergometrine is (XLVI).7 5 Brit. Med. J., 1935, i, 520; A., 655.7 6 Pharm. J., 1935,154, 709; A , , 894.7 7 J . Amer. Pharm. ABSOC., 1935, 24, 24, 185; A., 894.7 8 J . Amer. Chem. SOC., 1935, 57, 956, 1140; A., 872, 995.79 Compt. rend., 1935, 200, 1680; A., 995.80 W. A. Jacobs and L. C . Craig, Science, 1935, 82, 16; A., 1137.81 J . Amer. Pharm. ASSOC., 1936, 24, 748; A., 894.62 Proc.Roy. SOC., 1935, [B], 118, 478; A., 1512.83 See also H. W. Dudley and C. Moir (Science, 1935, 81, 559; A., 1157);M. S. Kharasch and R. R. Legault (ibid., p. 614; A., 1157); M. R. Thompson(ibid., p. 636; ibid., 82, 62; A., 1157); N. L. Allport and S. K. Crews (Quart.J . Pharm., 1935,8,447 ; A., 1512) ; and E. C . Kleiderer ( J . Amer. Chem. SOC.,1935,57, 2007; A., 1512). As far as the writer is aware, the present accountof ergot chemistry includes all references since 1930, up to December, 1935360 ORQAXIC UHEMISTRY.Timmis 84 have isolated yet another ergot alkaloid, ergometrinine,isomeric with, and convertible into, ergometrine, these two alkaloidsbeing related in the same manner as ergotoxine-ergotinine andergotamine-ergotaminine.The identity of $-ergotinine 59 remainsunexplained.It is possible now to summarise the present position as follows :Active Alkaloids. Inactive Alkaloids. Components.Ergotoxine Ergotinine Lysergic acid}ErgineAmmoniaProlinePhen ylalanineisoButyrylformic acidC,,H,,O,N5Ergotarnine Ergotaminine Lysergic acid}ErgineAmmoniaProline ?%H3,O,%Pheny lalaninoPyruvic acidErgometrine Ergometrinine Lysergic acidp- Aminopropgl alcohol(Hydroxyisopropylamine) C19H2302N3The first four alkaloids give the same ergine, which is the amideof lysergic acid. Ergometrine, and presumably ergometrinine,give the same lysergic acid. It follows that the isomerism betweenthe active and the inactive alkaloids must reside in the lysergicacid portion of the molecule.If the Jacobs-Craig formula forlysergic acid is correct, it would seem to follow that the isomerismmust be due to cis- and trans-arrangements around the doublebond in the side chain, although this is not altogether satisfactory,considering the conditions under which the isomeric changes occur.(Sir) H. H. Dale pointed out s5 that the pharmacological activitiesof ergoclavine and sensibamine were identical with those of ergo-toxine and ergotamine.86 A. $toll 87 showed that sensibamine wasnot a definite C31E1&5N5y but was a molecular com-pound, C33H,505N,,C3,H,,05N5, of crgotamine and ergotaminine.Ergoclavine is also probably a mixture, but it should be recalledthat Z-leueine was obtained by Jacobs and Craig 72 as one of itsdegradation products , and it is therefore possible that ergoclavine,which has been given the formula C31H37Q5N5, is actually a mixture,one component of which is an undiscovered alkaloid, C31H3,03N5or C31H3502N5, built up from erginc, 2-leucine, and phenylalanine.Admitting that the constitution assigned to lysergic acid byJacobs and Craig was not intended by them to be taken as h a l ,it is not without interest to attempt to picture the possible structure84 Nature, 1935, 136, 259; A., 1256.8 6 Schweiz. med.Woch., 1935, 65, 585.A. Vartiainen, J . Pham. Exp. Ther., 1935, 54, 259.Schweix. med. Woch., 1936, 65, 1077TURXER : HETEROCYCLIC COMPOUNDS. 361of the alkaloids.corresponds to a structure such as (XLVPI) or (XLVIII).The known chemistry of ergotoxine and ergotinineIn ergotarnine and ergotaminine, the group CO*CO*CHMe, wouldbe replaced by CO-CO*CH, (XLVIII) ; in (XLVII) CO*C(OH)*CHMe,would be replaced by *CO*C(OH)Me.Curare Alkaloids (South American Arrow Poisons) .-An in-vestigation of more than usual interest is that of the curare alkaloidsby H.King,*g who has obtained Boehm’s tubocurarine in a,crystalline condition, and submitted it to degradation processes,Curine, the crystalline alkaloid accompanying tubocurarine, wasCH,:CH\/OMeshown by E. Spgth, W. Leithe, arid P. Ladeck 89 to be the lava-modification of d-bebeerine. H. King 90 suggested a probableformula, for curine, which was supported by the experimental88 Chem. and Ind., 1935, 739; A., 1138.Ber., 1928, 61, [B], 1705; A., 1928, 1264.Ann.Reports, 1933, 30, 249352 ORGANIC CHZMISTRY.work of E. Spath and F. Kuffner,gl who assigned to O-methylcurinethe structure (I). King has now 92 carried out the exhaustivemethylation of the accessible base, d-curine (d-bebeerine), obtainingin this way the known O-methylbebeerine methochloride. Hofmanndegradation of the latter gave three distinct methine bases, andthence three crystalline methiodides. Hofmann degradation ofthe mixture of the three methochlorides gave trimethylamine anda substance, C3GH320G. Similar degradation of d-O-methyltubo-curarine chloride led to a mixture of four methine methiodides,three of which were identical with the above three methiodides,the fourth being Z-0-methyltubocurarinemethine methiodide.Thesecond degradation stage gave trimethylamine and the substanceC,GH320,. If O-methylbebeerine is regarded as (I), the lattersubstance can be accounted for as (11). It is probable, however,that although O-methylbebeerine methochloride and O-methyl-tubocurarine methochloride are diastereoisomerides, the parentphenolic substances, bebeerine methochloride and tubocurarinemethochloride, have different arrangements of hydroxyl andmethoxyl groups.Dauricine.-The constitution of this alkaloid has been proved 93by a synthetic method, methyldauricine-a-methine (111) beingobtained as follows :Me C0,HM e o - 0 9 -+ CO,H(-?-O--IC) \Me I Me0CHO J COClCHO<>-O-C> c-- COCl@---Me0 I J.AzlactoneCH,*C02H-+ C0,M*C,H2CK2Dihomo- PCI,; H,O; Hal; H.Me0CH9 193Ber., 1934, 67, [B], 55; A., 1934, 312.H. Kondo, Z. Narita, and S. Uyeo, Ber., 1935, 68, [B], 619; A., 637.92 J., 1935, 1381353From this it follows that dauricine has the constitution (IV) sug-gested by F. Faltis and H. Frauendorfer.94TURNER : HETXROCYCLIC COMPOUNDS.Phaeanthine.-Whereas A. C. Saiitos 95 found this alkaloid tobe C3,H3806N2, H. Kondo andI. Keimatsu 96 find it to be C,8H420,N2,and a comparison of its physical properties with those of tetrandrinesuggested that these two alkaloids are optical antipodes. Theabsorption curves support this view.97 Probably tetrandrine,isotetrandrine, and phaeanthine represent three out of the fourpossible stereoisomerides corresponding to the two asymmetriccarbon atoms in (V) and (VI), which are the two possible structuralformula for the alkaloids : 9894 Ber., 1930, 63, [B], 809; A., 1930, 774.95 Ber., 1932, 65, [B], 472; A., 1932, 527.96 Ber., 1935, 68, [B], 1503.97 H.Kondo and I. Keimatsu, J. PhaTm. SOC. J a p n , 1935, 636.98 F. von Bruchhausen, H. Oberembt, and A. Feldhaus, Alzrzalen, 1933,507, 144; A., 1933, 1313.REP.-VOL. XXXII. &354 ORGANIC CHEMISTRY.Sophora Alkaloid.-H. Kondo, E. Ochiai, and K. Tsudag9 havealmost completed their investigation of the constitution of matrine,one of the principal alkaloids in the roots of Sopbra jlavescens.The alkaloid, which is not a stereoisomeride of lupanine, is probablyeither (VII) or (VIII).MeyH-yO Me? H-YH,CH, N CO CH(VIII.))6'bH2 'Nf )H2(VII.) CH CH bH CH/\/ 7% d p 2CH, N CH,& 2 ) G \ p 2\/\/ CH, CH, kH2'\CG2CH, N CH,LactoJlavin.Last year, R.Kuhn, R. Reinemund, and F. Weygand describedthe synthesis of lumilactoflavin (11) by a modification of the Kuhlingcondensation. P. Karrer, H. Salomon, K. Schopp, E. Schlittler,and H. Fritzsche 2 showed that, if lactoflavin (vitamin-B,) isirradiated, not in strongly alkaline solution (when lumilactoflavinis produced), but in dilute methyl-alcoholic solution, lumichrome(I) (6 : 7-dimethylalloxazine) is formed, and they suggested thatlactoflavin probably contained a pentose chain, CH,*C,H,O,, inposition 9. In a later paper 3 the two decompositions induced byirradiation were pictured thus :( b )N NH CH2-i-E CH*OH],*CH,* OH(a)---!--,(4 q)$>O~~ N CO f- ::QjvNH ,kA-\yo (111.)(1.1 N CO$(b)99 Ber., 1935, 68, [B], 1899; A., 1514.Kondo's papers on matrine may be found in A abstracts.been abstracted.Ber., 1934, 67, [BJ, 1460; A., 1934, 1114.Parts I to IV and XV to XVII ofThe rest have not2 Helv. Chim. Acta, 1934, 1'7, 1010; A., 1934, 1233.P. Karrer, H. Salomon, K. Schopp, and E. Schlittler, ibid., p. 1165; A.,1934, 1233TURNER HETEROCYCLIC COMPOUNDS. 355and the following synthesis of an analogous but less complexsubstance was described :CH,*CH(OH)CH2*oH ::ofo2 + NH,*CH,*CH(OH) -+ MeCH,*OH---+-alloxan J.I,1H,-CH( OH)*CH,*OHNSince the simpler product closely resembled lactoflavin, the aboveconstitution (111) for the latter was supported.At approximatelythe same time, R. Kuhn and F. Weygand,* starting with Z-arabinoseand d-xylose, synthesised in the following manner, two compounds(111), both of which resembled lactoflavin in solubility, absorptionspectrum, and dependence of fluorescence on pH, formed a redradical when reduced in mineral acid solution, and gave lumilacto-flavin when irradiated in alkaline solution : ZZGfo2 + NH,*CH,*[CH-OH],*CH,*OH -+ Nitro-compoundAlloxan I * (111) t- I [Amino-compound]P. Karrer and K. Schopp, who had already described the isolationof a flavin from green malt, later showed that irradiation ofcrystallised ovoflavin-e from dry technical ovalbumin gave 6 : 7-dimethylalloxazine. R. Kuhn and F. Weygand pointed outthat there were eight possible stereoisomerides corresponding toformula (111) above, and expressed the view that lactoflavin wasmost probably the Z-arabo- or the d-ribo-compounds (111).Thesame authors 9 had previously stated that the I-arabo-compoundhad the same growth-promoting effect as natural lactoflavin, a tthe same time showing that the corresponding compound devoidof the two methyl groups in one nucleus was inactive in that respect.Ber., 1934, 67, [B], 1939; A . , 1935, 224.Helv. Chim. Acta, 1934, 17, 1013; A., 1934, 1233.Ibid., p. 1557; A , , 1935, 95.Ber., 1935, 68, [B], 166; A., 358.Compare R. Kuhn and H. Rudy, ibid., p. 169; A., 369.Ber., 1934, 67, [B], 2084; A., 1935, 262356 ORGANI$ CHEMISTRY.A different type of synthesis was introduced by P.Karrer, K.Schopp, F. Benz, and K. Pfaehler,lo who found that, whereas thecondensation of o-chloronitrobenzenes with amino-sugars was notparticularly satisfactory, good results in the preliminary stages wereobtained by the reductive condensation of the monoacetyl deriv-ative of o-phenylenediamines with aldoses :qH,*OHCH,*[CH*OH],@2Ac + OCH*[CH*OH],*CH,*OH + ( 3 g A cBy this method, these authors synthesised Kuhn’s Z-arabo-com-pound (111), and they too thought that the latter might be identicalwith natural lactoflavin, the observed differences possibly beingdue to impurities in the natural product.At this stage, R. Kuhn and H. Kaltschmitt 11 showed that thelactoflavin isolable as the tetra-acetyl derivative from Californianlucerne was identical with milk lactoflavin. It may be notedthat, according to L.J. Harris,12 Kuhn’s lactoflavin had no anti-pellagra action, and in this view he was supported by P. GyOrgy.l3Yet B. C. P. Jansen 1 4 regarded the identity of lactoflavin andvitamin-B, as established.Shortly after this, P. Karrer, I<. Schopp, and F. Benz l5 describedthe synthesis of two compounds of formula (111), one being the d-xylo-compound obtained by Kuhn and WeygandY4 the other a newsubstance, 6 : 7-dimethyl-9-d-l’-ribitylisoalloxazine (I11 ; IV), whichresembled natural lactoflavin extremely closely, the tetra-acetylderivatives also being very similar. In a paper sent in five dayslater than that by Karrer and his co-workers, R. Kuhn, H. Rudy,and F.Weygand l6 definitely expressed the opinion that 6 : 7-dimethyl-9-Z-araboflavin could not be identical with lactoflavin,since it was at least three times less effective than the latter ingrowth-promoting activity. These authors further said thatlactoflavin was possibly the d-ribo-compound (I11 ; IV), andannounced a new type of synthesis of the d-arabo-compound,starting from 4 : 5-dinitro-o-xylene :lo Ber., 1935, 68, [B], 216; A., 359.l1 Ibid., p. 128; A., 415.l2 Biochem. J., 1935, 29, 776; A., 545.Ibid., p. 741 ; A,, 646.l4 Nature, 1935, 135, 267; A., 545.Helv. CZ~irn. Acta, 1935, 18, 426; A., 631.l8 Ber., 1935, 68, [B], 625; A., 760TURNER : HETEROCYCLIC COMPOUNDS. 357QH2*OHIH0.Q.HH0.C.HHO+H -5.(111)6 : 7-Dimethyl-9-Z-araboflavinalloxazine (Karrer) .(Kuhn) = 6 : 7-Dimethyl-9-l-l’-arabityliso-Other workers17 confirmed the identity of the riboflavin withlactoflavin, and showed that most other synthetic flavins (xylityl,rhamnityl, sorbityl, dulcityl, and mannityl) had no biologicalactivity, although the Z-araboflavin had some. They also pointedout that lactoflavin is the growth-promoting substance present invitamin-B,, but is not necessarily the anti-pellagra factor, and thiswould explain the results of C. A. Elvehjem and C. J. Koehn.lsAt the same time, Karrer and his co-workers17 described the re-ductive condensation of d-ribose with the mono-urethane derivedfrom 4 : 5-diamino-o-xylene as an alternative first stage in theflavin synthesis.In order further to remove doubt as to the identity of lactoflavin,the antipodal varieties of d-ribo- and Z-arsbo-flavins were prepared.l9It was found that Z-riboflavin bad no growth-promoting action.An interesting point was also discovered by the same authors,CH,*[CH~OH],*C€12*OHA Nnamely, that 7-methyl-9-d-l’-ribitylisoalloxazine (V) had strongvitamin-B2 action ; the o-xylene structure, therefore, is not essential1 7 H.von Euler, P. Karrer, M. Malmberg, K. Schopp, F. Benz, B. Becker,and P. Frei, Helv. Chim. Acta, 1935, 18, 522; A . , 760. See also P. Karrer,H. von Euler, M. Malmberg, K. Schopp, and F. Benz, Svensk Kern. Tidskr.,1935, 47, 99; A., 1286.18 J . Biol. Chern., 1935, 108, 709; A., 669.1@ P. Karrer, H. Salomon, K. Schopp, F. Benz, and B. Becker, Helv.Chim.Acta, 1936,18,908; A., 1134358 ORUANIC CHEMISTRY.HO--HHO--HHO--Hfor the latter. Later 2o they described the synthesis of the eighthand last possible compound of structure (111), uix., 6 : 7-dimethyl-9-d- 1’-lyxitylisoalloxazine, this establishing finally the configurationof the sugar residue in lactoflavin. Examination21 of the fluor-escence spectra of several synthetic flavins further confirmed thesame point.R. Kuhn and F. Weygand22 found that the yield of flavinsobtained by condensing sugar derivatives of o-diamines withalloxan is enormously improved if boric acid is used instead ofhydrochloric acid as the condensing agent, moderately anhydroussolvents being beneficial. For instance, when the N-arabitylderivative of diaminoxylene was treated with zinc dust and boricacid in boiling glacial acetic acid, followed by addition of alloxan,a 52% yield of the flavin was formed. Even with 2-aminodiphenyl-amine and alloxan under these conditions a 63% yield of 9-phenyl-isoalloxazine may be obtained.Another interesting flavin synthesis is due to P.Karrer andH. F. Meer~ein.~3 o-4-Xylidine was reductively condensed withd-ribose in presence of nickel or palladium, the product (VI) coupledwith diazotised p-nitroaniline, the azo-derivative catalyticallyreduced under high pressure in presence of nickel, and the resultingbase condensed with alloxan :CH2*[CH*OH],*CH2*OHCH2*[CH*OH]3*CH,*OHMe MeK:*c.,,,,,,IIn a summary of their work, Kuhn and his collaborators24 stressthe advantages of their synthetic methods, particularly (1) the use20 P.Karrer, H. Salomon, K. Schopp, and F. Benz, Naturzuiss., 1935, 23,21 P. Karrer and H. Fritzsche, Helv. Chim. Acta, 1935, 18, 911 ; A., 1134.2 2 Ber., 1935, 68, [B], 1282; A., 1134.23 HeEv. Chim. Acta, 1935, 18, 1130; A., 1510.24 It. Kuhn, I<. Reinemund, F. Weygand, and R. Strobele, Ber., 1935,355; A., 993.68, [B], 1765; A., 1382LINSTEAD : THE PORPHYRIN GROUP. 359of palladium and not nickel for reductive condensations (Karrer’smethod necessitating the employment of high pressures), and (2)the use of boric acid in the final stage, yields of 90% being recordedwith d-riboflavin : condensation of N-d-ribityldiaminoxylene withalloxan was carried out in glacial acetic acid solution a t the ordinarytemperature.An account of some of the work on lactoflavin was given in alecture by R.K ~ h n . ~ ~ I n this it is suggested that the parentyellow plant enzyme corresponding to lactoflavin is (VII). Thisacid, in the form of its sodium salt, was prepared26 by treatinglactoflavin with phosphoryl chloride in presence of pyridine.E. E. T.8. THE PORPHYRIN GROUP.^General.Great advances have been made during the last few years inthe chemistry of this group, which constitutes probably the mostimportant class of natural organic colouring matters and is of thegreatest biological interest.2 Developments in the study of hzeminand associated compounds have already been r e p ~ r t e d . ~ Thechemistry of chlorophyll is reviewed below, together with that ofother natural porphyrins.The biological importance of the porphyriiis tends to obscuretheir great chemical interest.Before considering in detail thenatural pigments, we may briefly examine the peculiarity ofthe common structural unit. It is now generally believed that theporphyrins contain the basic unit porphin (A): as was first proposedby W. Kuster for the blood pigment. The synthesis of this parentsubstance in very poor yield was achieved by Hans Pischer, aftermuch negative workJ5 by heating pyrrole-2-aldehyde with alcoholin boiling formic acid. The substance is a typical porphyrin in itsgeneral properties, notably in its spectrum. The free compoundhas a five-banded spectrum with two sharp bands in the green region ;the copper compound has the usual two-banded spectrum of the25 Bull.SOC. Chirn. biol., 1935, 17, 905.26 R. Kuhn and H. Rudy, Ber., 1935,68, [B], 383; A., 545.1 The term should be read to include the free porphyrins, their metallic2 A. Treibs’s recent discoveries of porphyrins in petroleum and coal have3 Ann. Reports, 1932, %, 209.4 F. Hrturowitz and his associates have proposed a free-radical variant of5 H. Fisuher and W. Gluin, Annalen, 1936,521, 157.derivatives, and substances of similar cyclic structure.given the group a geochemical significance also (see p. 398).this (Ber., 1935, 68, [B], 1795; A., 1384)360 ORUANIC CHEMISTRY.metallic porphyrins, and the iron compound gives a typical haemo-chromogen spectrum on treatment with hydrazine.Simple deriv-atives such as octamethylporphin (with methyl groups on theP-positions of the pyrrole rings) have been known for some time.The Kiister formula was not a t first accepted, for, as Willstatterpointed out, it contained a 16-membered ring of a type withoutparallel. The evidence in its support comes mainly from thesynthetic work of H. Fischer.’ The syntheses of porphyrins bythe condensation of 3 : 4-dialkylpyrroles (such as opsopyrrole),with formaldehyde or formic acid,8 and by the fusion of 2 : 2‘-dibromodipyrrylmethenes with 2 : 2’-dimethyldipyrrylmethenes insuccinic are of particular value as indicating the structure.[The positions of the imino-hydrogen atoms and the double bonds in theseformulae are arbitrary.]All these synthetic methods, however, involve a loss of hydrogenill the final condensation, apart from the hydrogen bromide orwater, etc., eliminated between the condensing groups, and noneprovides a h a 1 proof of structure; nor, from the very complexityof the molecule, is this to be expected.On the other hand, thecumulative evidence is very strong, and the Kiister formula notonly accounts for all the reactions of porphyrins, but also providesa satisfactory basis for Fischer’s important investigations of theirisomerism .3The recently discovered phthalocyanines 10 throw some lighton the problem. The formula (B) given to phthalocyanine on6 H. Fischer and B. Walach, Annalen, 1926, 450, 164; A., 1926, 1261.7 H. Fischer and co-workers, ibid., 1926, 448, 178, 193, etc.; A., 1926, 962,963; particularly 1928, 462, 240; A., 1928, 902.Cf. Ann. Reporb, 1932,29, 209.* H. Fischer and A. Treibs, Annalen, 1926, 450, 132; A., 1926, 1266;Ber., 1927, 60, 379; A., 1927, 365.a H. Fischer, P. Halbig, and B. Walach, Annalen, 1927, 452, 268; A.,1927, 469.lo R. P. Linstead, J., 1934,1016; A., 1934, 1114LINSTEAD THE PORPHYRIN GROUP. 361chemical evidence l1 has been confirmed by J. M. Robertson’sX-ray examination.12 The molecular weight l3 and the dimen-sions l2 have been determined and the structure is certain (apartfrom detail of the fine structure). The entry of a metallic atom hashardly any effect on the size of the molecule.12 Phthalocyaninecontains a central 16-membered ring of a type essentially similarto that postulated for the porphyrins (but containing 4 -N= for4 -CH=) and resembles the latter in many physical and chemicalproperties.ll There is no doubt, therefore, that stable rings ofthis type can exist and the Kuster formula is strongly supported bythe resemblance .14When these large rings are present the chemistry of the moleculeis profoundly modified; e.g., phthalocyanine shows none of theordinary instability of an isoindole derivative, nor a porphyrinthat of a pyrrole.There is in fact an approach to an aromaticcharacter, presumably arising from the formation of the unbrokenconjugated ring. This is seen in the tendency to formation and inthe great stability of the type. The porphyrin nucleus survivesthe most diverse chemical treatment.Unlike almost all othercoloured organic compounds, porphyrins are not reduced by alkalinehyposulphite or by hydrogen over palladium in alkaline s01ution.l~They are reduced by more drastic reagents, but are readily regener-ated from the leuco-compounds (Fischer, R. Kuhn, Conant).Porphyrins can be halogenated,lG* l7 nitrated,l7. l8 and s~lphonated.~~The stereochemistry of both phthalocyanine and the porphyrinsis essentially aromatic ; the phthalocyanine ring has been provedexperimentally to be flat 12 and porphyrins show no optical activity.2011 C. E. Dent, R. P. Linstead,and A. R. Lowe, J . , 1934, 1033; A., 1934,1114.12 J., 1935, 615; A., 813. F. Haurowitz (ref. 4) has also found thatmetallic derivatives of porphyrins have substantially the same dimensionsas the parent substances.13 J.M. Robertson, R. P. Linstead, and C . E. Dent, Nature, 1935, 134,506; A., 689; cf. R. P. Linsteadand A. R. Lowe,J., 1934,1031; A., 1934,1114.14 In the last few months, H. Fischer, H. Haberland, and A. Muller(Annalen, 1936, 521, 122) have described pyrrole derivatives of a new type,called di-iminoporphyrins. These contain four pyrrole rings joined ontwo sides by the ordinary methenes and on the other two sides by (-NH-)or (-N=) groups. These substances to some extent bridge the gap betweenporphyrins and phthalocyanines.1 5 J. B. Conant and J. F. Hyde, J . Arner. Chem. Xoc., 1930, 52, 1233; A.,1930, 799.16 H. Fischer and G. Stangler, Annalsn, 1927, 459, 53; A., 1928, 76.1 7 H.Fischer and W. Neumann, ibid., 1932, 494, 228; A., 1932, 626.18 H. Fischer, M. Speitman, andH. Meth, ibid., 1934,508,154; A., 1934, 308.19 A. Treibs, ibid., 1933, 506, 196; A., 1933, 1309.20 H. Fischer and A. Stern,ibid., 1935,619,58; 520,88; A., 1135, 1383.M 362 ORGANIC CHEMISTRY.The absorption spectra of porphyrins in the visible region areremarkably like the ultra-violet spectrum of benzene.21 Finally,no isomerides are known corresponding with changes in the finestructure; e.g., no porphyrin exists in two forms, one with theimino-hydrogen atoms on adjacent pyrrole rings (as in A), theother with them on diagonally opposite rings. Neither are isomericmetallic derivatives of these forms known (Fischer, Conant, Dietz).The same appears to be true of phthalocyanines.This of coursecorresponds with the fact that! o-disubstituted benzenes do notexist in two forms.Chlorophyll has occupied the attention of chemists for nearly acentury, since Berzelius first attempted to isolate the pigment fromgreen leaves, but it is only in the last thirty years that real progresshas been made. To-day the problem of its constitution is nearlysolved, thanks mainly to the researches of Willstatter and HansFischer and their collaborators. Valuable contributions havealso been made by J. B. Conant, A. Stoll, and L. Marchlewski.Some of the earlier work has already been summarised in theseReports,Z3 but a short critical review of it in the light of the mostrecent developments is now given because the problem and thenomenclature are very complex and the early results have beencorrected in some respects.Chlorophyll occurs in the chloroplasts of green plants, mainlyin the leaves, accompanied by the yellow pigments carotin, xantho-phyll, and (in brown algze) fucoxanthin.G. G. Stokes24 andH. C. Sorby 25 provided the first evidence that chlorophyll is nothomogeneous and this was confirmed by M. Tswett,26 the discovererof the chromatographic method of analysis. At one time it wasthought that a large number of different chlorophylls existed,that of monocotyledons, for example, being different from that ofdicotyledons, but this was disproved by R. Will~tatter,~~ who21 J. B. Conant and S. E. Kamerling, J . Amer.Chem. SOC., 1931, 53, 3522;A., 1931, 1310; see p. 390.22 General accounts are given by R. Willstatter and A. Stoll (“Unter-suchungen uber Chlorophyll,” Berlin, 1913) and by H. Fischer (Oppenheimer’s“ Handbuch der Biochemie,” 1923,351; 1930,87; 1933, 262; J . , 1934, 245),and valuable short reviews by K. F. Armstrong (Chem. and Ind., 1933, 809)and E. M. Dietz ( J . Chem. Educ., 1935, 12, 208).23 Ann. Reports, 1911, 8, 144; 1912, 9, 169; 1913, 10, 151.?‘I Proc. Roy. SOC., 1864, 13, 144; J., 1864, 1’7, 304.Z 5 Proc. Roy. SOC., 1873, 21, 442.2c Ber., 1908, 41, 1352; Ber. deut. bot. Ces., 1906, 24, 316; A., 1908, i,27 R. Willstatt,er and A. Stoll, “ Chlorophyll,” p. 117.440 ; 1906, i, 973. LINSTEAD : TME PORPHYXCIN GROUP. 363showed that chlorophyll from a wide variety of sources is a mixtureof two very similar pigments, the blue-green chlorophyll a and theyellow-green chlorophyll b.The proportion of these forms in landplants is a,bout 3 of a t o 1 of b, but the amount of b is much smallerin brown algz. Dry leaves contain some 0.S70 of mixed chloro-phylls. One recent claim28 for the discovery of a third form ( c )has been disproved,29 but an independent line of evidence, althoughindirect, suggests that chlorophyll a may itself be a mixture of twoi~omerides.~~Chlorophyll is a waxy solid which has not yet been obtainedcrystalline. When green leaves are extracted with alcohol, acrystalline green pigment is produced,31 but this involves a secondarydecomposition, the formation of a chlorophyllide.Chlorophyll isbest isolated by extracting dried green leaves with acetone con-taining 15-20y0 of water; the solution is then diluted with lightpetroleum, and the acetone removed with water.32 About 80% ofthe total chlorophyll can be extracted in this way. Tswett 36separated crude chlorophyll in small amounts into the a and the bcomponent (and the carotenoids) by passing a benzene solutionthrough a column of calcium carbonate or sugar. R. WillstBtterand M. Isler removed the chlorophyll b by extracting the petroleumsolution of the crude pigment with 90% methyl Tswett’smethod has been improved by A. Winteratein and G. Steinf4 andWillstatter’s by A. Stoll and E. Wiedemann.35 Absorption spectraindicate that the older preparations of chlorophyll b containedsome 10% of a.For work on the degradation products it is usual,not to separate the two chlorophylls, but to hydrolyse the mixedpigment to pheophorbide (a + b ) and separate this into its com-ponents by acid fractionation (p. 366).Chlorophyll and closely related substances retain solvents(particularly water) tenaciously and their analysis is difficult.The formule originally given to the two components by Willstatter,36and accepted to-day, are : a, C,6H,205N4Mg ; 6 , C,,H,,06N4Mg.2 8 F. P. Zscheile, Bot. Gaz., 1934, 95, 529; A., 1934, 1115.Z g A. Winterstein and K. Schon, 2. physiol. Chem., 1934, 230, 139; A.,3Q J. B. Conant and E. M. Dietz, Nature, 1933, 131, 131; A., 1933, 287;31 J. Borodin, Bot. Ztg., 1882, 40, 608.32 R.Willstiitter and E. Hug, Annulen, 1910, 318, 21 ; A., 1911, i, 393.33 Ibid., 1912, 390, 269; A., 1912, i, 710.34 2. physiol. Chem., 1933, 220, 263; A,, 1934, 91; compare ref. 29.35 Helv. Chim. Acta, 1933, 16, 739; A., 1933, 838.36 “ Chlorophyll,” p. 144.1935, 362.see p. 388.The analyses for chlorophyll a actually agreedwith the values required for a semihydrate, C6,H,B0,N,Mg,&H,0364 ORGANIC CHEMZSTRY.Earlier workers had believed that chlorophyll contained iron,phosphorus, and other elements. The quantitative basis forthese formula is mainly the analysis of degradation products,particularly the chlorophyllides and phaeophorbides.ChZoro~hyZZi&es.-W~lstatter found that, when chlorophyll ingreen leaves is extracted with ethyl alcohol, a phytyl group (C2oH39)is replaced by ethyl, the reaction being catalysed by an enzyme,chlorophyllase, present in the leaf .37 The bluish-green product(“ crystalline chlorophyll ”) is a mixture of ethyl chlorophyllides(a + b ) , which form mixed crystals.These may be separated bypartition between aqueous methyl alcohol and a mixture of etherand light petroleum. Ethyl chlorophyllide a has the formulaC3,H380,N4Mg and the change in the a series is interpreted byWillstatter as follows :These formulze embody the fact (proved later) that chlorophyll andethyl chlorophyllide contain a carbomethoxy-group. This made theexamination of the structure of ethyl chlorophyllide dScult,because Zeisel determinations yielded a mixture of methyl andethyl iodides.Methyl chlorophyllide u (111, R = CH3) can bemade similarly, and a free chlorophyllide a (111, R = H) by hydro-lysing leaf chlorophyll with acetone containing 33% of water andseparating the a and the b component by partition.38CO OGH, CHMe<[ CH,],*CMe:CH *CH2-OHC32H300N4Mg<C0.0R [CH2]3*CHMefCH2]3*CHMe,(111.) VV.1I n the chlorophyllase reaction the phytyl group is liberated asphytyl alcohol (IV), the structure of which has been determined byF. Gottwalt Fischer and K. Lo~enberg,~~ who synthesised it from+-ionone. It contains four isoprene units. The chlorophyllfrom about 200 different plants yields the same phytol in some30% ~ield.~O’ The phytyl group is responsible for the wax-likeproperties of chlorophyll and its power of forming colloidalsolutions.37 R.WillstBtter and M. Utzinger, Annalen, 1911, 382, 129; A., 1921,38 R. WillstBtter, “ Chlorophyll,” p. 216.39 Ann. Reports, 1929, 26, 220.40 R. Willstlitter, F. Hocheder, and E. Hug, Annalen, 1909, 371, 1; A.,i, 659.1910, ii, 150LINSTEAD : THE PORPHYRIN GROUP. 365Phceophorbi&es.-In 1907, R. Willsttitter and 3’. Hochedersucceeded in removing the magnesium from chlorophyll and thechlorophyllides without secondary decomposition. The olive-brown solids obtained were called phaeophorbides. Leaf chloro-phyll on treatment with hydrated oxalic acid in alcohol yieldedphytyl phaeophorbide or phsophytin (V),* and methyl chloro-phyUide on treatment with 17% hydrochloric acid in ether wasconverted into methyl phEophorbide (VI).The action of strongeracid hydrolysed the phytyl group from phaeophytin and one methylgroup from methyl chlorophyllide to yield phsophorbide (VII),which contained a free carboxy-group. Methyl phaophorbidecould be obtained by hydrolysis of phzophytin with methyl-alcoholic hydrogen chloride.C02Me oxalic acidMeOH,enzymecx2H300N4Mg<(J0. O.C,,~,, -These substances were obtained as mixtures of a and b compoundswhich could be separated by partition between ether and hydro-chloric acid. The separation of phEophorbide a and b is veryimportant in practice, as it is more easily carried out than that of themagnesium- containing compounds.These reactions reveal a considerable difference in ease of hydro-lysis of the two ester groups.Esters of the carboxy-group whichcarries the phytyl group in chlorophyll are readily hydrolysed,whereas the original carbomethoxy-group present in the naturalpigments remains intact.Although the phytol-free compounds can easily be obtainedcrystalline, the determination of their molecular formulae hasgiven some difficulty. The formulae shown above all contain5 atoms of oxygen, following Willstatter. When the study ofchlorophyll was renewed by H. E’ischer and J. B. Conant, bothinvestigators at first preferred formula containing 6 atoms ofa1 AnnaEen, 1907, 354, 205; d., 1907, i, 959; “Chlorophyll,” Chap.* The fornuke (V-VII) are for the CG series.xv366 ORGANIC CHEMISTRY.oxygen.42 In 1932, however, A. Stoll and E.Wiedemann reaffirmedthe 0, formulae43 and these were also adopted independently byboth the other schools.44Acid Xeparation (Willstiitter) .45-Similar substances of thechlorophyll or of the porphyrin series are separated by means oftheir difference in basicity, an ethereal solution of the materialbeing extracted successively with hydrochloric acid of increasingstrength. This procedure cannot be used with compounds con-taining labile metals, such as chlorophyll itself, unless the eliminationof metal is immaterial. The acid number of a substance is definedas the percentage concentration of hydrochloric acid which extractstwo-thirds of it from an equal volume of an ethereal solution :methyl phEophorbide a has an acid number of 16 and is completelyextracted by 18% acid, whereas methyl phaeophorbide b, with anacid number of 21, requires 23% acid for complete removal.Extraction of an ethereal solution of methyl phaeophorbide a + bwith 17% acid therefore leads to almost complete separation.Solutions of the sodium phosphates also are used for suchseparations.Chlorin e.-Willstafter found that, when either phzeophytin orphzophorbide is hydrolysed by boiling methyl-alcoholic causticpotash for 30 seconds, phytochlorin e (a series) and phytorhodin g( b series) are formed.46 These can be separated by acid fraction-ation or the individual substances can be prepared from theseparated phzeophorbides.From this point it is convenient to concentrate on the compoundsof the a series.These have received much fuller study because theyare more accessible and crystallise better.The conversion of phzophorbide a into chlorin e (VIII) involvesthe hydrolysis of the stable methyl ester group and the addition ofit molecule of water :42 H.Fischer and R. Biiumler, Annalen, 1929, 474, 69; A., 1929, 1185;J. B. Conant and J. F. Hyde, J . Amer. Chem. SOC., 1929,51, 3668; A., 1930,225.43 On the occasion of the celebration of Willstiitter’s 60th birthday; Natur-wiss., 1932, 20, 628. The analytical figures were published later (Helv. Chim.A&, 1933, 18, 202, 739; A., 1933, 287, 838).44 H. Fischer and H. Siebel, Annalen, 1932, 499, 84; A., 1932, 1263;compare H. Fischer, 0. Moldenhauer, and 0. Sus, ibid., 1931, 486, 158; A . ,1931, 744; J.B. Conant and C. F. Bailey, J . Amer. Chem. SOC., 1933, 55,797; J. B. Conant and E. M. Dietz, ibid., p. 839; A., 1933, 1403.45 “ Chlorophyll,” p. 262.46 These substances are now generally called chlorin e and rhodin g re-spectively. The suffixes date from the time when series of chlorins andrhodins (of doubtful homogeneity) were obtained by acid fractionation.Only chlorin e need be considered hereLINSTEAD : THE PORPHYRIN UROUP. 3672H,O 40,H-3 C,lHBN,@O,H + MeOHC0,H(VIII.)Chlorin e is a tribasic acid and yields the same characteristic tri-methyl ester by the action of methyl sulphate on the potassiumsalt4' and by treatment with dia~omethane.~~ This shows thatthe fifth oxygen atom in phzophorbide is part of a +acidic group.Willstatter interpreted the reaction as the opening of a lactam ring,a suggestion later adopted by J.B. Conant and E. M. diet^,^^whereas Fischer regards it as the fission of a cyclic P-ketonic ester.These suggestions will be considered later.The three carboxy-groups of chlorin e are dissimilar, and twomonomethyl and two dimethyl esters have been obtained bypartial esterification of the free acid and partial hydrolysis ofthe trimethyl ester.50 Fischer has recently found that phaeo-phorbide and methyl phaeophorbide both yield the trimethyl esterof chlorin e on standing with diazomethane in methyl alcohol.51This remarkable reaction involves a methanolysis catalysed by thediazomethane ; when ethyl alcohol is used as solvent, the +acidicoxygen emerges as a carbethoxy-group.Chlorin e and rhodin g are only obtained by the allraline hydro-lysis of phaeobhorbides under conditions which prevent (or delay)atmospheric oxidation (see p.385). In the presence of oxygenthese and other derivatives of chlorophyll undergo the " phasetest." This is a colour reaction which occurs when an etherealsolution of chlorophyll, a phorbide, a phyllide, or chlorin e (tri-ester)is shaken with methyl-alcoholic potash. The colour changes toa bright yellow-brown and then back to green (Molisch). Chloro-phyll derivatives which show this change are said to be phase-positive.A reaction analogous to the formation of chlorin e is observedwith chlorophyll itself. Hot potash under suitable conditions con-verts this into isochlorophyllins as tripotassium salts.The phytyl47 R. Willsthtter and M. Utzinger, A?malen, 1911, 382, 171; A., 1911, i,659; compare H. Fischer and 0. Moldenliauer, ibid., 1930, 478, 54; A . , 1930,482.48 A. Treibs and E. Wiedemann, ibid., 1928, 466, 264; 1929, 471, 146;A., 1928, 1383; 1929, 941.49 J . Amer. Chern. SOC., 1933, 55, 839; A., 1933, 403.50 H. Fischer and H. Siebel, Annalen, 1932, 499, 84; A . , 1932, 1263; J. B.Conant and K. F. Armstrong, J . Amer. Chem. SOC., 1933, 55, 829; A . , 1933,403.61 H. Fischer, W. Gottschaldt, and G. Klebs, Annalen, 1932, 498, 194;A., 1932, 1263; H. Fischer and J. Riedmair, ibid., 1933, 506, 107; A., 1933,1308368 ORGANIC CHEMISTRY,and methyl groups me hydrolysed and the #-acidic group fixes thethird potassium atom.The magnesium is not affected. iso-Chlorophyllin a (IX) yields chlorin e on treatment with acids.Vigorous Degradation by AZkaZi.-The conversion of chlorophyllderivatives into porphyrins by the action of alkali at high ternper-atures was first achieved by E. H~ppe-Seyler,~~ E. S ~ h u n c k , ~ ~ andE. Schunck and L. Mar~hlewski.~~ Willstiitter and his collabor-ators 56 found that the action of methyl-alcoholic potash and pyridineon isochlorophyllins under pressure at 140-190" led to gradeddecarboxylation and the formation of a series of magnesium-containing phyllins. These could be separated through theirpotassium or ammonium salts. Acid eliminated the magnesiumfrom the phyllins with the formation of the corresponding porphyrins(H, for Mg).The third carboxy-group could not be eliminatedin this way and the final product was phyllophyllin,C3,H,N4MgG0,H , the porphyrin corresponding to which (phyllo-porphyrin) had already been isolated by the 'earlier workers.53* 55The alkaline degradation of the corresponding magnesium-freecompounds, particularly of chlorin e, yielded a series of porphyrinswhich could be separated by the Willstktter method.57* 58* 59 Thereis some difference between the results of the various schoolsregarding the intermediate substances isolated in this reaction, butgeneral agreement on the main final products. These are threeporphyrins of known constitution : two are monocarboxylic acids,phylloporphyrin, C3,H3,N,*C02H (already mentioned), and pyrro-porphyrin, C30H33N,*C0,H, and one a dicarboxylic acid, rhodo-porphyrin, C,oH3,N4(C02H)2.It was first thought that the twomonocarboxylic acids were isomeric, but it is now known thatphylloporphyrin is a methyl pyrroporphyrin. Rhodoporphyrin isa carboxypyrroporphyrin and can readily be decmboxylated to62 Probable constitution; cf. " Chlorophyll," Chap. XVIII.53 Z.physio1. Chern., 1879, 3, 339; 1880, 4, 193; A., 1880, 53, 894.54 Proc. Roy. SOC., 1891, 50, 302.55 Ibid., 1895, 57, 314; Annalen, 1894, 284, 81; A., 1894, i, 341; 1895,i, 296.5~3 R. WillstSitter and A. Pfannenstiel, ibid., 1907, 358, 205; R. Willstiitterand H. Fritzsche, ibid., 1909, 371, 33; A., 1908, i, 198; 1910, i, 126. Will-stiltter also studied the corresponding degradations of allomerised chlorophyll.57 R.Willsthtter, " Chlorophyll," p. 353.68 H. Fischer and A. Treibs, Annalen, 1928, 466, 188; A., 1928, 1382.5@ A. Treibs and E. Wiedemann, ibid., p. 264; A., 1928, 1383LINSTEAD : THE PORPHYRIN GROUP. 369pyrroporphyrin. Phylloporph_yrin can be converted into pyrro-porphyrin by drastic treatment with alkali.The last carboxy-group in pyrro- and phyllo-porphyrins (andin the corresponding phyllins) is very stable. Decarboxylationis best effected by dry distillation with soda-lime 56# 6o or by pyrolysisof the free acids alone or in high-boiling solvents.68 I n this waythe oxygen-free parent substances of the group are obtained ;phylloporphyrin yields phylloetioporphyrin, C,,H,,N,, and pyrro-porphyrin yields pyrroaetioporphyrin, C,,H,4N,.Phylloaetiopor-phyrin is degraded to pyrroztioporphyrin by sodium ethoxide ata high temperat~re.~~ The corresponding magnesium compound,(pyrro)ztiophyllin, has been obtained by decarboxylating rhodo-phyllin . 61-+ Phylloporphyrin --% Phyllozetioporphyrin (- clod - a0alkali I (-cE,> NaOEt (- CHJI-- Chlorin eI(- CO,) - GO, + Rhodoporphyrin -3 Pyrroporphyrin -+ PyrroetioporphyrinIt was a t first thought that the a?tioporphyrins from chloro-phyll were identical with that derived from tho blood pigment(mesoaetioporphyrin). This has been shown to be incorrect byHans Fischer. Nevertheless these chlorophyll porphyrins resemblethe porphyrins of the hzemin series in general properties andare red, crystalline solids with low acid numbers.They aresimply inter-related, but the reaction by which they are formedfrom chlorin e is comparatively complex, as will be seen later.We have now left the exact chlorophyll type, but are approachingsubstances of known constitution.Degradation to PyrroZes.-The drastic oxidation of intact chloro-phyll has given no useful results, probably on account of compli-cations arising from the presence of the phytyl group. FollowingW. Kiister’s classical work on the oxidation of hemin,62 L.Marchlewski oxidised phylloporphyrin with chromic acid to thenitrogen-free anhydride, C,H,O,, corresponding to hematic acid(X, 0 for NH).63 R. Willstatter and Y. Asahina64 found thatrhodo-, phyllo-, and pyrro-porphyrins and chlorin e gave the samemixture of methylethylmaleinimide (XI) and hzmatic acid (X)60 R.Willstatter and Max Fischer, 2. pkysiol. Chem., 1913, 87, 430; A , ,1913, i, 1251.6 1 Idem, Annalen, 1913, 400, 182; A., 1913, i, 1218.62 2. p?qsiol. Chetn., 1899, 28, 1, etc.; A., 1900, i, 68.I33 J . p r . Chem., 1902, [ii], 65, 161; A., 1902, i, 387.6.1 Annalen, 2910, 373, 227; A . , 1910, i, 499370 ORU ANIC CHEMISTRY.when oxidised by chromic acid, Caro’s acid, or lead peroxide.structures of these were already known.TheNH NH: NH NH NHCOACO C O A C O M e A /\Me M e 0 , N eMe!==dEt MeLl*qH, M e l u l E t Me!--!!Et Me-EtCH2*C02H(XI.) (X.) (XII.) (XIII.) (XIV.)One molecule of the chlorophyll derivative yielded rather lessthan one molecule of the acid (X) but considerably more thanone molecule of the imide (XI).Following preliminary work by M. Nencki and L.Mar~hlewski,~~R. Willstatter and Y. Asahina reduced phylloporphyrin with hydro-gen iodide and acetic acid under drastic conditions to hzmopyrrole(XII), kryptopyrrole (XIII), and phyllopyrrole (XIV).66 Ethylchlorophyllide was reduced similarly to it mixture of dimethylethyl-pyrroles, and chlorin e to crude hzemopyrrole and phyllopyrrole.The yields in all these degradations were noticeably worse than thoseencountered in the hEmin series. It was subsequently observedby H. Fischer, A. Merka, and E. Plotz that chlorophyll, unlikehzemin, gave on reduction with hydriodic acid about one molecularproportion of carbon dioxide, corresponding to the presence of alabile carboxyl group.The general conclusions from these results are that at leastthree pyrrole rings are present in the molecule of chlorophyll andits derivatives, that one pyrrole carries a methyl group and a propionicacid group in the P-position, that two pyrroles carry methyl andethyl groups in the @-position, and that the pyrrole rings are joinedthrough carbon atoms (such as methene groups) in the a-position.Structure of Rhodo-, Phyllo-, and Pyrro-porphyrins.-We nowturn to Fischer’s brilliant synthetic work in the field.The existenceof four stable nitrogen atoms in the molecule and the drastic degrad-ation to pyrroles suggested that phylloporphyrin and its associatesbelonged to the porphyrin series. Moreover, the absorptionspectrum of pyrroporphyrin, and to a less extent of phylloporphyrin,was close to that of the porphyrins of the blood series, such asmesoporphyrin.If the parent stioporphyrin of the chlorophyllseries were identical or isomeric with that of the hzmin series,it seemed possible that pyrroporphyrin (and phylloporphyrin,which was then thought to be isomeric with it) might be a simplemonocarboxylated derivative, namely, a tetramethyltriethylporphin-6 5 Ber., 1901, 34, 1687; A., 1901, i, 554.6 6 Annalen, 1911, 385, 188; A., 1912, i, 41.6 7 irbid., 1930, 478, 283; A., 1930, 620LINSTEAD : THE PORPHYRIN GROUP. 37 1propionic acid 68 such as (XV). (It is true that this would meanthat pyrroporphyrin would contain 33 instead of 31 carbon atoms,but analysis could not decide definitely on this point.) The eightpossible isomeric acids of this type were synthesised by Fischerand his co-workers 69p70 and characterised by the melting points oftheir methyl esters and in other ways.None was identical witheither pyrro- or phyllo-porphyrin.Et Me(XV.) Me/' 'Et (p = CH2*CH2*C0,H here and subsequently.)Me\ /MeP EtAt about the same time it was found by A. Treibs and E. Wiede-mann 59 that pyrroporphyrin could be brominated to a mono-bromo-compound which yielded bromocitraconimide on oxidation.Hence pyrroporphyrin contained at least one ring of the typeN<'*TH and was probably a tetramethyldiethylporphinpropionicCCMeacid. It was possible to test this by subjecting the free methenegroup to a series of reactions resembling that used in the synthesisof h ~ m i n , ~ l namely, R-H __p1 R*COMe -+ R*CH(OH)Me ->R*CH:CH, I_, R*Et (where R = tetramethyldiethylporphin-propionic acid, less H).By this process pyrroporphyrin wasconverted into one of the tetramethyltriethylporphinpropionicacids already synthesised. The acid actually obtained 70 (methyl6* The methene groups and pyrrole rings of porphin are numbered andlettered as shown in (XVI). The placing of the hydrogens on the nitrogenatoms, and the consequent position of the double bonds, is arbitrary. Forthe purpose of indicating the position of substituents, the formula is mostconveniently abbreviated t o (XVIa), where the lines represent the #?B-sidesof the four pyrrole rings.669 H.Fischer, H. Grosselfinger, and G. Stangler, Anlzalen, 1928, 461, 221 ;70 H. Fischer, H. K. Weichmann, and K. Zeile, ibid., 1929, 475, 241; A.,7 1 See Ann. Reports, 1932, 29, 209.A , , 1928, 651.1929, 1465372 ORUANIC CHEMISTRY.ester, m. p. 271") was identical with the synthetic acid of formula(XV), in which the orientation of the groups followed from themethod of synthesis. This a t once proved that pyrroporphyrinwas a tetramethyldiethylporphinpropionic acid. It was uncertainwhich of the ethyl groups of the triethyl compound had beenintroduced by the acetylation and subsequent treatment and whichwere the two already present in the pyrroporphyrin. The fkeemethene group of the latter could, however, only be at C,, C,, or C,.Ashas been stated, this substance contains two carboxy-groups,one of which is easily removed with the formation of pyrroporphyrin.It was known from work in the hamin series that a-carboxylicacids of the porphyrin series (e.g., the large class which containsthe carboxy-group at the end of a nuclear ethyl group, as--CH,*CH,*CO,H) are only decarboxylated with great difiiculty.As the labile carboxy-group of rhodoporphyrin was not substitutedin an alkyl group, it was presumably nuclear, being placed on thefree methene group of pyrroporphyrin.Rhodoporphyrin wastherefore either (XVII), (XVIII), or (XIX).This was settled by an examination of rhodoporphyrin.Et Me Et Me C0,H Me Et EtM/ \Et Me/ \CO,H Me/ \Et Me/ \MeP\ /C02HMe MeFX.1'p IdMe MeMe\ /Me /Me(XIX.)P CO2H P Et(XVII.) (XVIII.)It was known from the study of natural and synthetic uroporphyrinsthat the absorption spectra of compounds with adjacent carboxy-groups differed slightly from those in which the carboxy-groupswere separated by allryls.Comparison of the spectrum of" natural " rhodoporphyrin with that of a synthetic isomeride (XX)showed the same difference and indicated that in rhodoporphyrinthe nuclear carboxyl was on C,, next to the propionic acid group.72Synthetic 1 : 3 : 5 : $-tetra-met hyl-2 : 4 -diethyl-6 - carboxyporphin -7 -propionic (XXI ;A = CO,H, B = CH2*CH,-C02H) was identical with rhodopor-phyrin. The corresponding compound without the 6-carboxy-group (XXI; A = H, B = CH2*C€€,*C0,H) was also synthesised,and proved to be identical with pyrr~porphyrin.~~ It followed thatthe completely decarboxylated pyrroetioporphyrin was 1 : 3 : 5 : 8-This was confirmed by synthesis.acid72 H.Fischer and A. Schormiiller, Annalen, 1929, 473, 211; A,, 1929,73 H. Bischer, H. Berg, and A. Schormiiller, ibicl., 1930, 480, 109, 189;1184.482, 232; A., 1930, 931; 1931, 101LTNSTEAD : THE PORPHYXIN GROUP. 373tetramethyl-2 : 4 : 7-triethylporphin (XXI; A = H, B = Et)*This also was proved by synthesis.73Et MeThe very important fact emerges that the parent porphyrinsof the hsmin and chlorophyll series are closely related in structure.74Hemin is the ferric (chloride) derivative of protoporphyrin (XXII).71On reduction the latter gives mesoporphyrin (XXIII), which ondecarboxylation yields aetioporphyrin-I11 (XXIV).This hasexactly the same substituents, except on C,, as pyrroaetioporphyrinand is in fact 6-ethylpyrroaetioporphyrin. This relationship isof considerable biological significance.Cl3,:CH Me Et Me Et Me nne/ \/CH:CH, Me/ \Et Me/ \EtP P P P Et Et(XXII.) (XXIII.) (XXIV.)The simiIarity between the two series suggested the possibility ofinterconversion and this was realised by Fischer in both directions.Pyrroporphyrin was converted into mesoporphyrin, identicalwith that obtained from h ~ e m i n , ~ ~ by a complicated series ofreactions, in which the essential changes on C, were the following :Me, ,Ne Me,, ,Me Me\, 6,M"Mesoporphyrin dimethyl ester has been degraded to a pyrropor-~ h y r i n , ' ~ isomeric with the '' natural " compound, with the C, andC, groups reversed.Phylloporphyrin, which is richer by CH, than pyrroporphyrin,must contain a similar disposition of nuclear substituents, as ityields the latter with sodium ethoxide.The most probable structure7 4 Verdeil's claim (Compt. rend., 1851, 33, 689) that the pigments of theleaf and the blood were related was thus prophetic, although based on invalidexperimental work.75 H. Fiacher and H. J. Riedt, Annulen, 1931, 486, 178; A., 1931, 744.76 H. Fischer and J. Ebersberger, ibid., 1934, 509, 19; A., 1934, 421374 ORGANIC CHEMISTRYwas that having two of the pyrrole rings joined, not by an a-methenegroup as in the simple porphyrins, but by the groupThere are four dissimilar methene bridges (or, p, y , 6 in formulaXVI) in pyrroporphyrin, so four bridge-methyl derivatives arepossible.These were all synthesised; the substance with themethyl group on C,, (XXV) was identical with phylloporphyrinfrom natural sources. 78 It followed that phylloEtioporphyrin hadthe analogous structure with an ethyl group for --CH2*CH2-C02Hon C7.If the structure of the porphyrins be held to be proved bythese syntheses,79 the question arises as to how far we can applythis knowledge to the structure of chlorophyll itself. Willstiitterpointed out the need for caution here, because the vigorous alkalinedegradation may cause some deep-seated change in the molecule.Nevertheless several facts suggest that the relationship is fairlyclose : (i) The same porphyrins and not isomerides are alwaysobtained as final products in the alkaline degradation ; the order ofsubstituents remains the same, and hence it is unlikely that there hasbeen a ring fission and resynthesis.(ii) Some of the reagents whichconvert chlorin e into phylloporphyrin are comparatively mild,e.g., boiling quinoline, phosphoric acid at 140°.80(XXV.)On the hypotheBis of a fairly close structural resemblance, wearrive a t the skeleton (XXVI) for chlorophyll a,81 in which thedouble bonds are neglected. The magnesium atom is placed on twopyrrole nitrogen atoms by analogy with the iron of hzemin and77 A-CH=CH- bridge was also considered by Fischer, but was abandonedin view of the synthetic results.78 H.Fischer and H. Helberger, Annalen, 1930, 480, 235; A., 1930, 932;€1. Fischer, W. Siedel, and L. Le Thierry d'Ennequin, ibid., 1933, 500, 137;A.., 1933, 286.The evidence is strong but not final (see p. 360).R. Willstatter and M. Utzinger, Annalen, 1911, 382, 171; A., 1911, i,This should be compared with the formuke advanced by Fischer, Conant,659.and Stoll (p. 379)LINSTEAD : THE PORPHYRM GROUP. 375because the existence of stiophyllin shows that the metal can becombined independently of oxygen. If to the formula of pyrro-porphyrin the methyl and phytyl groups and the magnesium atomare added and the four hydrogens which these groups replace aresubtracted, a formula is obtained which differs from that of chloro-phyll a (C55H7205N4Mg) by (c303) ’[C31H3402N4 + CH3 + c2(jH3, + Mg - 4H = C52H,202NaMg]’From the formation of phyllo- and rhodo-porphyrins we may placea carbon atom on Cy and an oxygen-carrying carbon atom on C,.If allowance is made for the hydrogen atoms replaced, the unlocatedsubstituents become (CH,O,), which may be presumed t o beattached to the substituents on c6 and C,.Purther, chloro-phyll a contains two carboxy-groups, carrying the phytyl and amethyl group. One of these is located in the propionic acid groupon C, ; the other must presumably form part of the “ 6ysystem.”Actually the structural position is more complicated than is indicatedby this summary, because the fundamental ring system of chloro-phyll, the phorbides, and chlorins differs slightly from that of thederived porphyrins.The next stage in the elucidation of the structure was theestablishing by less drastic reagents of the connection betweenchlorophyll and the porphyrins.A suitable reagent was foundby Pischer 82 in a mixture of hydriodic and acetic acids. At about60” this reduced phaeophorbide a to a leuco-compound, aerialoxidation of which gave it series of phEoporphyrins, which con-tained the same number of carbon atoms as the initial material(34, neglecting the ester group). The exact products formedvaried with the conditions of reaction. Similar treatment ofchlorin e or its trimethyl ester gave another series of C,, and C,,porphyrins, the chloroporphyrins. The most important compoundsof these two series are shown below; the formuh are stripped ofester groups to facilitate comparison.83From phsophorbide a, (C,4H,405N4) :Oxyphsoporphyrin a5, 84 C34H3406N4 \ Isolated as monomethylPhsoporphyrin a5, C3&3,05N4 I esters.82 H.Fischer and R. Biiumler, Annalen, 1929, 474, 65; 1930, 480, 197;A., 1929, 1185; 1930, 932; H. Fischer and 0. Moldenhauer, ibid., 1930,478, 54; H. Fischer, A. Merka, and E. Plotz, ibid., p. 284; A., 1930, 482,620; K. Noack and W. Kiessling, 2. physiol. Chem., 1929, 183, 36; A., 1929,727.83 A useful summary of their formation and properties is given by Fischerin Oppenheimer’s “ Handbuch der Biochemie,” ErgBnzungswerk I, 1933,The corresponding ethoxy-266-2 67.84 Originally called neophaeoporphyrin a,.phseoporphyrin a, was originally called phseoporphyrin a8376 ORGAXJC CHEMZSTRY.Prom chlorin e, (C34H&$,) :Isolated as monomethyl { ester.Chloroporphyrin eG, c34153606N4Ghloroporphyrin e6, C33H3405N4Chloroporphyrin e4, Cs3H3604N4 ) Isolated as free acids.[The suffixes a and e denote the derivation from phaeophorbide aand chlorin e respectively. The numbers which follow denote thenumber of atoms of oxygen in the molecule.]These compounds crystallise well and yield polymethyl esterswhen treated with diazomethane. The two series differ in absorp-tion spectra. It should be noted that the stable methyl ester groupof chlorophyll a persists in three of the compounds named in the list.The most important from the structural point of view is phaeopor-phyrin u5, which is isomeric with phzophorbide a and, like it, is amonomethyl ester, but differs from it in absorption spectrum andother properties.85 It was also formed fromphaeophorbides by catalytic hydrogenation over platinum, followedby aerial oxidation of the leuco-compounds formed.s6 J.B. Conantand J. 3’. Hyde also showed that the near derivatives of chlorophyllcould be converted into porphyrins in this way.87Phzeoporphyrin u5 readily loses its carbomethoxy-group andyields phylloerythrin, C33H3403N4, a substance of great importance,first discovered by L. Marchlewski.88 This may be obtained directfrom chlorophyllide and the phzeophorbides by the prolongedaction of boiling 20% hydrochloric acid and also by the biologicaldegradation of chlorophyll. It has been found in ox-bile, thefaeces of ruminants, elephants and other herbivora, and in cattlegall-~tones.~~ The best source is sheep dung.It is extremelystable and its chemical properties and absorption spectrum show it tobe a porphyrin. The formation of a porphyrin by a simple bio-logical process shows independently that chlorophyll contains abasic structure very like $hat of the porphyrin~.~~The actual process in the digestive tract is one of isomerisation85 H. Fischer and 0. Sus, Annalen, 1930, 482, 225; A., 1931, 102.8 6 H. Fischer and H. Helberger, ibid., 1930, 480, 260; A,, 1930, 932;M. Fisher and E. Lakatos, ibid., 1933, 506, 123; A., 1933, 1308; compareE. M. Dietz and T. H. Werner, J . Amer. Chem. SOC., 1934,56, 2180; A , , 1934,1371.It yields a monoxime.87 Ibid., 1930, 52, 1233; A., 1930, 799.8 8 2.physiol. Chem., 1929, 185, 8; A., 1929, 1468.8@ H. Fischer and R. Hess, ibid., 1930, 18’9, 133; A., 1930, 634.So Early evidence of the close connection between the green derivativesof chlorophyll and the porphyrins comes also from J. B. Conant and J. F.Hyde’s demonstration that porphyrins were produced by the pyrolysis ofchlorin e ( J . Amer. Chem. SOC., 1929, 51, 3668; A., 1930, 225)LINSTEAD : THE PORPHYRIN QROUP. 377and decarboxylation, and appears to be confined to the higheranimals. A number of other degradation products of chlorophyllin vivo have been discovered, e.g., probophorbides (sheep dung)and phyllobombycin (fsces of silk-worms). These contain amodified porphyrin system.The structure of phylloerythrin has been established by Fischerboth analytically 91*92 and by synthesis.One of the oxygen atomsis ketonic; fhe oxime on reduction by the Kishner-Wolff methodgives deoxophylloerythrin ( C33'H3602N4), in which the two remainingoxygens are present as carboxyl. Deoxophylloerythrin can beobtained directly from phaeophorbide a by treatment with hydrogenbromide in acetic acid a t 180". It is not identical with any of theisomeric te tr ame t h yltr ie t h y lporp hinpr opionic acids, which onlydiffer from it in formula by two hydrogen atoms. Phylloerythrinon treatment with sodium ethoxide in presence of air yields phyllo-,pymo-, and rhodo-porphyrins.Fischer's interpretation of these facts, which is generallyaccepted, is that in both phylloerythrin (XXVII) and its deoxo-compound (XXVII, CH2 for CO), the C, and C, carbon atomsform part of a, five-membered carbon ring.Et Me(XXVII.)The presence of this was shown by the synthesis of deoxophyllo-e r ~ t h r i n .~ ~ The relationship between the two substances wasconfirmed by the oxidation of deoxophylloerythrin to phylloerythrinby oleum containing ~u1phu.r.~~ The yield was poor and somechloroporphyrin e5 was formed by further oxidation. The positionof the carbonyl group at C, in phylloerythrin follows from its fissionby alkali into phylloporphyrin and rhodoporphyrin, which recallsthat of deoxybenzoin into benzoic acid and toluene : R*CH,*COR'_I, R*CH3 + HO*OC-R'.1931, 496.81 H.Fischer, 0. Moldenhauer, and 0. Sus, Annalen, 1931, 485, 1 ; A.,O2 Idem, ibi&., 1931, 486, 107; A., 1931, 744.93 H. Fischer and J. Riedmair, ibid., 1931, 490, 91; 1932, 49'7, 181; A,,94 H. Fischer, J. Heckmaier, and J. Riedmair, ibid., 1932, 404, 86; A.,1931, 1431; 1932, 1045.1932,625378 ORGANIC CHEMISTRY.The structure of deoxophylloerythrin has been confirmed by itssynthesis from phyll~porphyrin.~~ The latter was converted intoits hEmin (ferric compound), and a CH,*OMe group introduced on c6 by means of chloromethyl ether and stannic chloride. Theproduct was freed from iron and purified by successive treatmentwith hydrogen bromide and alkali. The oxymethyl porphyrin wasfinally converted into deoxophylloerythrin by fusion with succinicacid, which eliminated methyl alcohol and formed the 9 : 10-bond.Phzoporphyrin a5 (XXVIII) was recognised by Pischer as the@-ketonic ester corresponding to phyll~erythrin,~~ a view acceptedby Conant and Stoll.When hydrolysed with acid or cold alcoholicpotash, it underwent ring fission of the usual cyclopentanonecarb-oxylate type and yielded chloroporphyrin e6, the methyl hydrogenester of the corresponding open-chain (adipic) acid (XXIX) :C0,Me Ph ylloporphyrin(XXVIII.) C0,Me (XXIX). / (XXX.) \(XXXII. )II IIOHCH CO H0,C C0,H--+ yy y-y\/ 0(XXXIIa.) (XXXI.)[These partial formulae show the portion of the molecule below the dottedThe methyl hydrogen ester (XXIX) can be cyclised back to(XXVIII) by rather unusual reagents: pyridine and sodiumcarbonate, or a mixture of hydriodjc and acetic acids.E’ormic acidremoves the carbomethoxy-group from chloroporphyrin e6 to givechloroporphyrin e4 (XXX), which can be further decarboxylatedto phylloporphyrin. Chloroporphyrin e4 is therefore y-methyl-rhodoporphyrin (or phylloporphyriii-6-carboxylic acid) and chloro-porphyrin e6 is the ester of rhodoporphyrin-y-acetic acid. Thepresence of the 6-carboxyl makes the methyl group of y-methyl-rhodoporphyrin very easily oxidisable (contrast phylloporphyrin) ;according to the conditions it yields chloroporphyrin e5 (y-formyl-9 j H. Fischer, M. Speitmann, and H. Meth, Annalen, 1934, 508, 154; A.,1934, 308.line in (XXVII), the rest of the molecule being the same.LMSTEAD : THE PORPHYRM GROUP.379rhodoporphyrin, XXXIP ; which appears to exist normally in thecyclic form XXXIIa) or rhodoporphyrin-y-carboxylic acid (XXXI).The latter is also formed by the action of oxygen and alkali onphylloerythrin. A number of other transformations of theseporphyrins have been studied by Fi~cher.8~~ 9~ 92* 94Fischer’s view that phEophorbide a and phzeoporphyrin cc5(monoester) on the one hand, and chlorin e and chloroporphyrin e6on the other, are pairs of isomerides has not been accepted byJ. B. Conant, who regards the chlorophyll derivatives as beingricher in hydrogen than the porphyrins. This point cannot besettled by analysis, but calorimetric determinations by A. Sternand G. Klebs96 support the idea, of isomerism. In any case, it isCIQH2 C02MeC0,PhytylQH2 C0,MeC 0,Phy t yl(XXXIII.) (XXXIV.)H. Fischer.97 J. B. C ~ n a n t . ~ ~Et CH Me(7H2 TH CH*OHYHz C02MeC0,PhytylA. st011.9996 Annalen, 1933, 505, 295; A., 1933, 1173.97 Ibid., 1933,502, 175; A., 1933,617. Revised from several earlier formuls.98 J. B. Conant and E. M. Dietz, J . Amer. Chem. SOC., 1933, 55, 839; A.,1933, 403. Revised from a, previous formula (ibid., 1931, 53, 2382 ; A., 1931,1075). The y-side chain wits believed to be of the type >C:CH*CO,Me+>CH.CH( OH)*CO,Me.Experimentalresults appeared later (HeZv. Chim. Acta, 1933, 16, 183; A , , 1933, 287, etc.).gg A. Stoll and E. Wiedemann, Natwwiss., 1932, 20, 706380 ORGANIC CHEMISTRY.now clear that chlorophyll and its near derivatives contain a ringsystem similar to but not identical with that of the porphyrins.This is referred to as the isoporphyrin (isoporphin) system.In 1932-1933, the three formulze for chlorophyll on p.379 wereproposed and it is convenient to consider further investigations inrelation to them.These forrnulz have certain common features, based mainly onFischer’s synthetic work, namely, a modified porphin ring, virtuallythe same arrangement of alkyl substituents in the p-positions ofthe pyrrole rings, and carbon substituents on C, and C,,. Thephytyl group is placed on the propionic acid side chain at C,, and themethyl group on the C,, carboxyl, as first proved by C0nant.lPischer proposed the reverse arrangement,92 but later corrected it .2The evidence on this point is briefly as follows : (i) Phzophorbide acontains the original carbomethoxy-group of chlorophyll intactand has one free carboxy-group which originally carried the phytyl.On pyrolysis it yields pyrophzophorbide a (the phorbide corre-sponding to phylloery thrin) with loss of the carbomethoxy-group,but the free carboxyl remains intact.Only a p-propionic acidgroup could survive pyrolysis in this manner. (ii) Ethyl chloro-phyllide also contains the original carbomethoxy-group and hasethyl in place of phytyl. Ilydriodic acid converts it into the ethylester of phzoporphyrin a5, which on pyrolysis yields phylloerythrinethyl ester. Hence the ethyl group in ethyl chlorophyllide andthe phytyl group in chlorophyll arc carried by the propionic acidside chain.Stoll accepted Fischer’s carbocyclic ring, but Conant 98 did notconsider that the presence of this in chlorophyll a or the phgophor-bides was proved by their transformation into phzoporphyrin a5and phylloerythrin.He suggested that the ring might be formedduring the reaction just as it is in the formation of phylloerythrinfrom the chloroporphyrins.All three formulz differ as to the state of reduction of the mole-cule and the mode of combination of the fifth oxygen atom.Fischer placed this in a carbonyl group a t C,, Stoll in a secondaryalcohol group, and Conant in a lactam ring. Early attempts toprepare ketonic derivatives from chlorophyll a and the phorbidesof the a series had failed, though H. Fischer and J.Riedmair hadobtained indirect evidence for the keto-group in the formation oftri-acid derivatives with diazomethane in an atmosphere of nitrogen 31 J. B. Conant and J. F. Hyde, J . Amer. Chem. Soc., 1929, 51, 3668; A.,1930, 225; J . B. Conant, E. M. Dietz, C. 3’. Bailey, and S. E. Kamerling,ibid., 1931, 53, 2382; A., 1931, 1075.H. Fischer, 0. Siis, and G. Klebs, AnnuZen, 1031, 490, 38; A., 1931, 1431.3 Ibid., 1933, 506, 107; A , , 1933, 1308LINSTEAD : THE PORPHYRIN GROUP. 381(which would prevent the oxidation of a secondary alcohol of thetype suggested by Stoll). Apparent support for the alcohol formulacame from A. Stoll and E. Wiedemann’s preparation of a benzoylderivative from phsophorbide a and methyl phzeophorbide a.4The presence of a keto-group was proved independently andconclusively by Fischer 5 and Stoll,6 who prepared a phase-positiveoxime from methyl phzophorbide a.This could be converted intothe oxime of phsophorbide, phzophorbide itself, or the oxime ofphaoporphyrin a5 according to the conditions. This result ledStoll to abandon the alcohol grouping on C, in favour of carbonyl,but he retained the dihydroporphin arrangement of the nucleus.The benzoylation was interpreted as a reaction of the enolic formof the phorbides. The presence of a, keto-group in phEophorbide ais incompatible with Conant’s lactam formula, which is also opento objection on stereochemical grounds.Fischer’s 1933 formula (XXXIII) contained two free imino-groups and three of the methyl groups converted into methylenes(arbitrarily the ones on C,, C,, and C,).Experimental support forthe free imino-groups seemed to be provided by Zerewitinoffdeterminations of active hydrogen, but the method is of doubtfulvalue in this series and in view of subsequent developments theresults probably are not significant. Strong evidence has recentlybeen found by Fischer for the presence of unsaturated side chainsin chlorophyll and the isoporphin derivatives generally, by the“ oxo-reaction ” and the addition of diazoacetic ester.H. Fischer and J . Riedmairs found that when phaophorbideor methyl phaeophorbide was treated with cold hydriodic andacetic acids, best in a steam of oxygen, a crystalline porphyrin,oxophEoporphyrin a5, containing 6 atoms of oxygen, was formed.In nitrogen the reaction gave the usual phsoporphyrin a5.Theoxo-porphyrin yielded a dioxime and hence contained a secondcarbonyl group : the reaction was t’herefore an oxidation in additionto the usual isoporphin --+ porphin change. The oxo-reaction wasalso given by chlorin e, chlorin e4 (formed by decarboxylationof chlorin e in pyridine), and pyrophaophorbide, but not by thederived porphyrins. It was first thought that these oxo-compoundscontained an aldehydo-group in place of an unsaturated sideHelv. Chim. Acta, 1933, 18, 739; A., 1933, 838:H. Fischer, J. Riedmair, and J. Hasenkamp, Annalen, 1934, 508, 224;A. Stoll and E. Wiedemann, Helv. China. Acta, 1934, 17, 163; A., 1934,There was a polemic as t o priority : both the papers cited 6.6 are datedA., 1934, 420.308.December, 1933.7 H.Fischer and P. Rothemund, Ber., 1931, 64, [B], 201; A., 1931, 497.* Annulen, 1933, 505, 87; A., 1933, 959382 ORGANIC CHEMISTRY.hai in,^ but a comparison with synthetic formylporphyrins showedthis to be incorrect. Fischer accordingly abandoned the hypothesisthat chlorophyll a and its near derivatives contained methylenegroups (XXXIII) and considered that the oxoporphyrins weremonoacetyl compounds in which the acetyl group was derivedfrom a single ethylidene group on C, or Cp.9 The position of thissubstituent was found as follows: Under the conditions of theoxo-reaction, pyrophEophorbide a yielded oxophylloerythrin. Whenthis was heated with hydrochloric acid in a sealed tube, the carbonyl-containing side chain was eliminated and two new porphyrins wereformed.These resembled pyrroporphyrin and phylloerythrinrespectively but were not identical with them. The first containedtwo free pyrrole methene groups and was considered to be a de-ethylpyrroporphyrin (such as XXXVI) ; the second, with onefree methene group, was a de-ethylphylloerythrin (such asXXXVII) which could be reduced to the corresponding deoxo-compound, with CH, for CO.(XXXVI.)P(XXXVII.)The structure of these was proved by their synthesis by Fischer’sgeneral methods. 1 : 3 : 5 : S-Tetramethyl-4-ethylporphin-7-prop-ionic acid (XXXVI) gave an ester identical with that of the de-ethyl-pyrroporphyrin from oxophylloerythrin.lO Both the possiblede-ethyldeoxophylloerythrins were synthesised by H.Fischer andW. Rose; the 2-de-ethyl compound (XXXVII; CH, for CO)was identical with that prepared from natural sources.12 It followsthat the oxo-group in the oxoporphyrins, and hence the unsaturatedside chain in the phorbides, is in the 2-position. The nature of theoxo-group was proved as follows. The ester of 2-de-ethylpyrro-porphyrin (XXXVI) was converted into the hamin (ferric) com-pound and acetylated to a diacetyl derivative (CH,*CO for H a tC, and C,). The same compound was formed by the acetylationH. Fischer and J. Hasenkamp, Annalen, 1934, 513, 107; A., 1934, 1370.lo H. Fischer and S. Bockh, ibid., 1935, 516, 177; A., 633.11 Ibid., 1938,519,l; A., 1134.l2 The nomenclature is rather confusing : “ Deoxo ” here implies thereduction of the keto-group at C, ; ‘‘ 0x0 ” in oxophylloerphrin.that an acetylgroup is present instead of ethyl at C,LINSTEAD : THE PORPHYRIN CROUP. 383(at C,) of ox~pyrroporphyrin.~~present at C2 is acetyl.Hence the substituent alreadyThe formula of the oxoporphyrins is typified by (XXXVII1) :(XXXVIII. ) (XXXIX.) (XL.)R = CO,Me, Oxophaoporphyrin a5.R = H, Oxophylloerythrin.The unsaturated side chain which gave rise to the acetyl groupmight be ethylidene or vinyl. This was settled by H. Fischer andH. Medick by use of the diazoacetic ester reaction.14 Protopor-phyrin, the metal-free porphyrin Corresponding t o hamin, is knownto contain two vinyl groups.71 It reacted with diazoacetic ester at100" with elimination of nitrogen to yield the corresponding dicyclo-propane compound (XXXIX) .14 The absorption bands wereshifted towards the blue region and the product no longer reactedwith hydrobromic acid. Oxidation of (XXXIX) with chromic andsulphuric acids gave a mixture of hzematic acid and the cyclopropanemaleinimide (XL).Methyl phaeophorbide u also reacted with diazoacetic ester.The absorption bands were shifted slightly towards the blue region,but without alteration of the characteristic phorbide type.Theproduct on isomerisation with hydriodic and acetic acids gavethe corresponding porphyrin, the analysis of which agreed with theaddition of a >CH*CO,Me group. Pyrophsophorbide gave asimilar compound, which could also be obtained by the pyrolysisof the methylphaeophorbide product.Catalytic reduction andreoxidation of this yielded the porphyrin addition compoundcorresponding to phylloerythrin. Vigorous oxidation of this gavemethylethylmaleinimide and the cyclopropane compound (XL)identical with that obtained from the addition compound of proto-porphyrin. The same substance was obtained in bad yield bythe direct oxidation of the diazoacetic ester adduct of pyrophaeophorbide. Chlorin e trimethyl ester also added diazoacetic ester13 H. Fischer and J. Hasenkamp, Annulen, 1935, 519, 42; A., 1134.1* Ibid., 1936, 517, 245; A., 871384 ORGANIC CHEMTSTRY.The significance of these results is increased by the fact thatporphyrins with saturated side chains such as the haemato- andmeso-porphyrin of the haemin series, and the phaeoporphyrin a5and phylloerythrin of the chlorophyll series, do not react withdiazoacetic ester.The important fact therefore emerges that the nearest derivatives ofchlorophyll contain the vinyl group in the same position as one of thevinyl groups of hcernin and the leaf and blood pigments thereforecome into even closer relationship.The recognition of the unsaturated side chain as a vinyl grouphas the advantage that its conversion into the acetyl group bythe “ oxo-reaction ” can be interpreted simply as follows : l4CHXH, --> \ CHI*CH, 5% CH(OH)*CH, >O*CH,/- /- /-The mechanism which had previously been advanced to explain theforitlation of a carbonyl-containing side chain from an ethylideneor methylene group had been comparatively complicated.A difficulty which remained was that, though the vinyl groupsin haemin add the elements of hydrogen bromide or water readily,this had not been observed in the chlorophyll series.H. Fischerand J. Hasenkamp have, however, found that the vinyl groupsof phzophorbide a add hydrogen bromide under tho conditionswhich convert haemin into haematoporphyrin. The hydrogenbromide adduct is unstable, but its methanolysis products havebeen isolated.The assignment of one double bond in the side chains in placeof the three of the earlier Pischer formula (XXXIII) was in harmonywith the formation and properties of dihydrophaeophorbide. Thishad been prepared by the hydrogenation of phaeophorbide a informic acid with one mole of hydrogen over Adams’s ~ata1yst.l~It is phase-positive and yields an oxime; hence the carbonyl grouphas not been reduced. Preliminary experiments indicated that thisand other dihydro-compounds gave a poor yield of “ oxo-com-pounds.” Re-examination of dihydrophaeophorbide, carefully puri-fied by the chromatographic method, showed that only phaeo-porphyrin a5 and no oxoporphyrin was formed under the conditionsof the oxo-reaction.l3 Hence the dihydro-compound carries anethyl group in place of vinyl at C,.It seems probable that theearly discovery by 5. B. Conant and J. F. Hyde l6 that phaeophor-bide a and chlorin e, but not porphyrins, could be reduced in aqueoussolution by hydrogen over palladised asbestos, can also be explainedas a, reduction of the vinyl group.l6 J .Amer. Chem. SOC., 1930, 52, 233; A., 1930, 799.H. Fischer and E. Lakatos, Annalen, 1933, 506, 123; A., 1933, 1308LMSTEAD : THE PORPHYRIN GROUP. 385These results led Fischer to propose formulae for chlorophyll abased on the following for phaeophorbide a.(XLIa.)CH:CH,I Me$?H COC0,Meor (XLIb.)More recent modifications of these are dealt with later.Allomerisation and the Phase Test.-When an alcoholic solutionof chlorophyll is evaporated to dryness or allowed to stand, it isallomerised and loses its power of crys ta1li~ation.l~ This processwas recognised as an oxidation by J. B. Conant,ls who is responsiblefor much of our knowledge in this field. He showed that theoxidising agent was aerial oxygen and that two equivalents of 0were utilised.lg This was confirmed by H.Fischer,20 who foundthat benzoquinone could bring about a similar change. Anotherdehydrogenating agent is potassium molybdicyanide.21green colourchange brought about in chlorophyll derivatives by the additionof methyl-alcoholic potash (p. 367). This also was shown to involvean oxidation by two equivalents of aerial oxygen.22 As has beenstated, rapid saponification of phaeophorbide a with boiling methyl-alcoholic potash yields chlorin e without oxidation, because thealcohol vapour protects the phorbide from the air. In phase-testsaponification, the hydrolysis is slower and oxidation can proceed.Similar results were obtained by H. Fischer, 0.Sus, and G. Klebs,20who found that the first colour change (green to brown) does notinvolve oxidation, which only occurs during the change back togreen. Some chlorin e is normally formed during phase-test17 R. Willstlitter and M. Utzinger, Annalen, 1911, 382, 129; A., 1911, i,18 J. B. Conant, J. F. Hyde, W. W. Moyer, and E. M. Dietz, J. Amer.19 J. B. Conant, S. E. Kamerling, and C. C. Steele, ibid., p. 1615; A.,20 Annalen, 1931, 490, 84; A., 1931, 1431.21 J. €3. Conant, E. M. Dietz, C. F. Bailey, and S. E. Kamcrling, J. Amer.22 C. C. Steele, ibid., p. 3171; A., 1931, 1169.The “ phase test ” is the green + brown659.Chem. SOC., 1931, 53, 359; A., 1931, 368.1931, 745.Chem,. SOC., 1931, 53, 2382; A., 1931, 1075.Compare ref. 19.REP.-VOL.XXXII. 386 ORQANIC UHE116ISTRY.saponification, but it was found by J. B. Conant and W. W. Moyer 23that this could be entirely prevented if ethyl or, better, n-propylalcohol was used in place of methyl alcohol as a solvent. Thisis due to the ready absorption of oxygen by the solutions of alkaliin the higher alcohols. Allomerised chlorophyll does not show abrown phase, but gives a green solution a t once under phase-testconditions. These results show the correctness of Willstatter’sview that the changes brought about by the phase test and allomer-isation are essentially the same.had shown that the principal product of phase-tes tsaponification of phzeophytin was an unstable chlorin, phytochloring. Conant and Moyer 23 found that the same mixture of unstablechlorins was formed by treatment of the phorbides and of chlorin etri-ester under the conditions of the phase test.On standing inether, these changed into two purplish-brown crystalline solidscalled phaeopurpurin 7 and 18.24 Phaopurpurin 7 was a mono-methyl ester and yielded with diazomethane a trimethyl esterwhich was also formed directly from the unstable chlorins by thesame reagent. When hydrolysed with hot methyl-alcoholic alkali,it lost the group -CO*CO,Me, yielding chlorin f and potassiumoxalate. This indicated that it contained an a-ketonic acid group.Chlorin f 2 5 is a dibasic acid which on treatment with hydrogeniodide and reoxidation yielded rhodoporphyrin, which was alsoformed directly from phzeopurpurin 7 and alcoholic potash a t150°.23 Pyrolysis of chlorin f yielded pyrroporphyrin (see schemebelow).PhEopurpurin 18, which contains one free and two maskedcarboxyl groups, is formed from the unstable chlorins by fissionof carbon dioxide on standing in ether.26 Its hydrolysis yielded achlorin (tri-acid) apparently identical with Willstatter’s chlorin a andcalled chlorin p 6 by Fischer.This regenerated phzeopurpurin 18 onstanding or by the action of heat. When heated alone or in solvents,phEopurpurin 18 yielded the curious green anhydride of rhodo-porphyrin-y-carboxylic acid,27 and with alcoholic potash at 150”23 J . Amer. Chem. SOC., 1930, 52, 3013; A., 1930, 1299.24 These suffixes correspond with the acid numbers of the compounds.(See p.366.)25 Probably identical with Willstatter’s phytochlorin f (“ Chlorophyll,”p. 303), and named rhodochlorin by Fischer.26 E. M. Dietz and W. F. Ross, J. Amer. Chem. SOC., 1934, 56, 159; A.,1934, 308. These authors point out that chlorin a and phEopurpurin 18had been described by H. Malarski and L. Marchlewski in 1912 (Biochem. Z.,42, 219; A., 1912, i, 641) under the names p-phyllotaonin and anhydro-P-phyllo t aon in respec t ively .27 H. Fischer, W. Gottschaldt, and G. Klebs, Annden, 1932, 498, 194;A., 1932, 1263.WillstatteLINSTEAD : TH.E POXPHYRIN GROUP. 387rhodoporphyrin itself.23 From these results, Conant concludedthat both allomerisation and the phase test involve a change fromthe potential phylloporphyrin t o the potential rhodoporphyrinstructure.18Fischer and Conant are in substantial agreement on the structuresof the products of these reactions, though they differ as to thenature of the isoporphin-porphin change and the mechanism ofallomerisation.It will be convenient to neglect the first differencefor the moment and to write P for the main porphin unit of thechlorophyll porphyrins and iso-P for the corresponding unit of thephorbides, chlorins, and purpurins. [On this basis chlorin e(according to Fischer) is iso-PLCH2*C0,H ( y ) and rhodo-porphyrin is P-H ( y )C0,H (6)\CH2*CH,*C02H (7)/CO,H ( 6 )\CH2*CH2*C02H (7) .]Dimethylphzeopurpurin 7 is recognised as the tri-ester of they-glyoxylic acid corresponding to chlorin e, and its transformationsare illustrated below : 28tion methane diazo- /CO,Me ( 6 ) allomeriea-Phorbides -3 Unstable chlorins --+ iso-P--CO*CO,Me ( )phase test \CH2-CH,-C8,Me (7)Dimethyl phEopurpurin 7.A t Ipyrolysis C0,HP-H /H +- ; s o - d H\CH,*CH,*CO,H \CH,*CH,*CO,HPyrroporphyrin.Chlorin f (Rhodochlorin).alkaliat 160°-~Rhodoporphyrin.The reactions of phaeopurpurin 1.8, which contains one carbonatom less than phzeopurpurin 7, are interpreted on p. 388.18* 231 2 6 s 27The purpurins and the chlorins derived from them are here repre-sented as containing the isoporphin structure, and hence, accordingto E’ischer, a vinyl group on C,. This is not fully proved, but issupported by a recent statement 29 that phaeopurpurin 7, rhodo-chlorin, and chlorin p 6 all react with diazoacetic ester.28 Based on the scheme given in ref.18.20 H. Fischer and A. Stern, Annalen, 1935, 520, 88; A., 1383388 ORGANIC CHEMISTRY.The study of phzopurpurin 7 has opened up a new field forinvestigation. When the unstable chlorins prepared by the phasetest or allomerisation are allowed to stand, phaeopurpurin 7 isobtained as a monomethyl ester. The methyl group correspondsto that in phzophorbide a and chlorophyll a. When the compoundis pyrolysed in boiling diphenyl, it yields, not chlorin f, but itsmonomethyl ester,18 i.e., the methyl group is not lost with the10-carboxyl. This suggests that phzopurpurin 7 contains acarbomethoxy- (or similar) group at C, ; the suggestion, however,conflicts with the weight of evidence which places this group inother chlorophyll derivatives at Clo.Conant and Dietz 3o findthat phzopurpurin 7 originates from an isomeric impurity inphaeophorbide. The purest phaeophorbides by the same treatmentyield no phzopurpurin 7, but a monomethyl compound in which thecarbomethoxy-group is at C,, as in the usual formulae. Theseexperiments suggest that the isomeric impurity with the abnormalposition of the carbomethoxy-group corresponds to a second formof chlorophyll a. If this can be substantiated, it opens up interestingpossibilities.on standing, - CO /">o Unstable chlorins ---- -$ iso-P-co\CH,*CH,*CO,H/ ,Phaeopurpurin 18. (7)P > , + P-co /CO&P -33/ p , Hiso-P- C0,H\CH,*CH,*CO,H \CK,*CH,*CO,H \CH,*CH,*CO,HChlorin a (Chlorin p 6).Rhodoporphyrin. Rhodoporphyrin -y-carboxylic acidanhydride.Fischer and his collaborators have made a thorough study ofthe allomerisation of chlorophyll derivatives by benzoquinoneand by iodine, the general conclusion being that oxidation occursat Clo. The following results are some of the most significant.If ethyl chlorophyllide is allomerised with benzoquinone in ethylalcohol, and the product treated with hydrochloric acid in ether,crystalline ethyl 10-ethoxyphaeophorbide is 0btained.~1 Thisyields a dihydro-compound (presumably with reduction of the30 Nature, 1933, 131, 131; J. Amer. Chem. Xoc., 1933, 55, 839; A., 1933,31 H. Fischer and J. Riedmair, Annalen, 1933, 508, 107; A., 1933, 1308;287, 403.cf. H. Fischer, L.Filser, and E. Plotz, ibid., 1932, 495, 1 ; A., 1932, 756LTNSTEAD : THE PORPHYRIN GROUP. 3892-vinyl group) and, when it is fully reduced with hydrogen iodide 31or catalytically 32 and reoxidised, gives ethoxyphaeoporphyrin a5(phaeoporphyrin as). The structure of this was known from thework of H. Fischer and J. He~kmaier,~~ who had obtained it by theoxidation of phaeoporphyrin a5 ((timethyl ester) with iodine inethyl alcohol in the presence of sodium carbonate. On the basisof Fischer's formulae, these results can be summarised by thescheme :d x k E t )*CO,Me/QO (6) allomerisation . isO-P-CH.CO,*Me ( y ) ---+ $80-\(-JH,.CH,.CO,Et (7) (quinone* EtoH) \CH,*CH,*CO,EtEthyl phsophorbide a. 10-E thoxyphaoph orbide.1 I HI reduction,.L reoxidationHI reduction, 1.L reoxidationI,EtOH,NaPCO yo -- 4 ~ c ( ~ E t ) * C O , M e\CH2*CH,*C02€€ \CH2*CH2.C02H/ y o P-CH*CO,MePhsoporphyrin u5.E thovyphsoporphyrin a5.Methyl phaeophorbide a can be oxidised similarly with iodinein the presence of sodium acetate to methyl lo-hydroxyphaeo-p h ~ r b i d e . ~ ~ The acetyl compound of this contains the phorbide(isoporphin) ring structure intact and giws the " oxo-reaction."When the free 10-hydroxy-compound is hydrolysed with causticsoda, the isocyclic ring is broken and unstable chlorins are formed,which with diazomethane yield dimethylphaeopurpurin 7.34Difl'erent mechanisms of allomerisation have been proposed byXischer and by Conant. The former suggested that the first processwas dehydrogenation with the formation of a double bond betweenC, and Cl0, this being followed by the addition of water or alcoh01.3~Conant's view was that allomerisation consisted in the dehydro-genation of a secondary alcohol group at Clo.Further comment isunnecessary in view of the unsettled. state of the subject.According to Fisbher, the brown phase of chlorophyll derivatives,the formation of which does not involve oxidation, is connected10 9 - 7-with the formation of an enol at C, : 36 ?--'(OH)- R. KuhnCOzMe32 H. Fischer, E. Lakatos, and J. Schnsll, Annulen, 1934,509,201 ; A., 1934,33 Ibid., 1934, 588, 250; A., 1934, 420.34 H. Fischer, J. Heckmaier, and T. Scherer, ibid., 1934, 510, 169; A.,1934, 785.35 This has been altered in a recent paper by H.Fischer and A. Stern(ibid., 1935, 519, 63; A., 1134) to explain measurements of optical activity,but it is doubtful if the alteration is necessary.666.36 H. Fischer and H. Siebel, ibid., 1932, 499, 84; A., 1932, 1263has suggested that an enolisation of the 10-carbomethoxy-groupoccurs.37 The lengthening of the conjugated chain so obtainedcauses the red absorption band to move into the infra-red region,causing a lightening of the visible colour (green --+ yellow-brown).Fine Xtructure of ChZorophyZZ a.-Considerable light has beenthrown on this difficult subject by the application of three physicalmethods. It has long been known that porphyrins have multi-banded absorption spectra which may be used for their identification.J. B.Conant and S. E. Kamerling 38 have now found that a t thetemperature of liquid air the porphyrins show a unique spectrum ofnarrow bands in the visible region. The green derivatives ofchlorophyll (phorbides, chlorins) have wider bands in the absorptionspectra and are intermediate between most organic colouringmatters and the porphyrins. The relationship between the visiblespectra of porphyrins and the phorbides (or chlorins) resemblesthat between the ultra-violet spectra of benzene and cyclohexadiene.This suggests strongly that chlorophyll derivatives contain thedihydroporphin ring.Evidence pointing in the same direction has come from a studyof the basicity of the four pyrrole-like rings in phorbides, chlorins,and por~hyrins.~~ Willstatter's method of acid fractionation wasbased on the variation in basicity of chlorophyll derivatives and hehad obtained evidence of the greater basicity of two of the nitrogenatoms by the isolation of dihydrochlorides. Conant and hiscollaborators measured the basicity by potentiometric titrationswith perchloric acid in glacial acetic acid, using a chloranil elec-trode.4O The derivatives of chlorophyll were compared withsubstituted pyrroles of known constitution.The porphyrins werefound to contain two relatively strongly basic groups, presumablyof the pyrrolenine type (XLII), and one very weakly basic groupof the pyrrole (XLIII) or isopyrrole (XLIV) type. The secondweakly basic group could not be detected by the method. Thephorbides contained one relatively strong and one very weak basicgroup and a group intermediate in strength.The chlorins resembledthe phorbides except that the intermediate group was rather morebasic. These facts are compatible with the green chlorophyll37 It. Kuhn, P. J. Drumm, M. Hoffer, and E. F. Moller, Ber., 1932, 65,38 J . Amer. Chem. Soc., 1931, 53, 3522; A., 1931, 1310.39 J. B. Conant, B. F. Chow, and E. M. Dietz, ibid., 1934, 56, 2185; A , ,4" Cf. J. B. Conant) and T. H. Werner, &d., 1930, 52, 449; A., 1931, 40;pi, 1785; A., 1933, 52.1934, 1371.J. B. Conant and B. F. Chow, $bid., 1933,455, 3745; A., 1933, 1121LINSTEAD : THE PORPHYRIN GROUP. 391derivatives containing one pyrrolerrine ring, two pyrrole or iso-pyrrole rings, and a dihydropyrrole nucleus of type (XLV) ,N NH NIE NHA-11 11-(XLII.) (XLIII.) (XLIV. ) (XLV.)It had long been thought that chlorophyll and its derivativeswere optically inactive ; 41 the leuco-compounds showed no activityand the intense colour had prevented an examination of the pig-ments themselves.I n 1933, A. Stoll and E. Wiedemann42 over-came the experimental difficulties and showed that chlorophyll aand b were both lmorotatory with [a];!& = about - 265". Theactivity might have come from the phytyl group, as phytyl alcoholis optically active before di~tillation,~~ but this was shown to beincorrect, since the phytyl- and magnesium-free phorbides of thea and b series also were active.42 Stoll and Wiedemann state thatchlorophylls, methyl phaeophorbides, and phaeophorbides (a and b )all racemise in acetone or methyl-alcoholic solutlion and that oldsamples of crystalline methyl phrffophorbides are inactive.H.Fischer and A. Stern, using white light, have confirmed the opticalactivity but not the ease of racemisation. They state that bothsolutions and solids are optically table.^^^^^ The following sub-stances, inter alia, were found to be active : pyrophaeophorbide a,chlorophyllides, 10-ethoxyphaeophorbide, rhodochlorin, chlorin p 6(tri-ester), and dimethylphaeopurpixrin 7, all being laevorotatoryexcept the last two. The activity survived allomerisation, thephase test, and conversion of the active compounds into theirmetallic derivatives. On the other hand, all the porphyrinsderived from these substances anti those from hzmin (and alsoblood hemin itself) are optically inactive.The only known activeporphyrin is uroporphyrin from mussel-shells, and in this theactive centres are very probably difrerent from those of the chloro-phyll derivatives, being located in succinic acid side chains.45As far as the chlorophyll series is concerned, the activity isundoubtedly associated with the isoporphin structure. The factthat phaeophorbide was not racernised on conversion into pyro-phzophorbide showed that C,, was not the active centre (or not4 1 See, e.g., H. Fischer and H. Siebel, Annalen, 1932, 499, 94; A., 1932,42 Helv. Chim. Acta, 1933, 16, 307; A., 1933, 515.48 R. WillstStter and F. Hccheder, AnnaEen, 1907, 354, 248 ; R. Willstlitter,E. W.Mayer, and E. Huni, ibid., 1911,378, 84; A., 1907, i, 784; 1911, i, 144.44 Ibid., 1935, 519, 58; A., 1134.46 Ibid., 520, 88; A., 1383.1263392 ORCrANIC CHEMISTRY.the only active centre). As this was the only asymmetric carbonatom in Fischer's current formuke (XLIa and b), these becameinvalid. Fischer and Stern first suggested44 that the y-carbonatom carried a hydrogen atom and was an active centre, butabandoned this idea in view of the activity of rhodochlorin (chlorinf),in which this carbon atom must be of the type -CH= or -CH,-.They point out that asymmetric carbon atoms can only appear bythe formation of a dihydroporphin system and that reductionmust occur on one of the pyrrole rings and not a t the methenebridges or by addition of hydrogen to two nitrogen atoms.45 Ofthe various possibilities, Fischer now prefers formula (XLVI) forchlorophyll a.I n this, ring I11 is tentatively selected as the reducedring and hence carbon atoms 5 and 6 as the asymmetric centres.45This formula (or some near modification) 46 can explain all the morerecent facts and is by far the most satisfactory of those in the field.CHICH,I CH Me(XLVI.)4 6 It appaars to the Reporter that a slight modification of (XLVI) in whichthe two extra hydrogens are placed on a (C=N) bond of one of the pyrrolerings (probably ring 111) would accommodate the physical results better. Thepartial formula is shown in (XLVIa), the resL of the molecule being the sameYH-COC0,Meas in (XLVI). Formula (XLVIa) contains an interrupted conjugated chain,iq place of the continuous conjugation of the porphyrins and (XLVI), toaccount for the considerable change in absorption spectrum; and also thedihydropyrrole ring, as suggested by Conant.Apart from C1, it containsonIy one asymmetric carbon atom (starred). The substance (XLVI) cangive rise to two inactive and resolvable (cis- and truns-) forms according t othe arrangement of t'he substituents a t C, and C,. No such isomerides appearto be knownLINSTEAD : THE FOItPHYRIN UROUP. 393The formulz of Conant and Stoll are not very different from theabove with regard to the isoporphin system. If it be accepted thatthe true porphyrins from chlorophyll contain an ethyl group at C,,and the isoporphyrins (phorbides and chlorins) a vinyl group, thenwe can reconcile the view that the porphyrins and isoporphyrins areisomeric and at the same level of reduction (Fischer) with theview, advanced by Conant from the physical results and also heldby Stoll, that the isoporphyrins conhain a dihydroporphin structure.I n this connection it is of interest that Conant 47 was able to dehydro-genate chlorin f to it porphyrin, isorhodoporphyrin (called +verdo-porphyrin by Fischer).This could be converted into rhodopor-phyrin by reduction with hydrogen iodide and reoxidation, also bymineral acids under certain conditions, but not by mild treatment .47,48Rhodoporphyrin is known to carry an ethyl group at C2 and thelatest work shows that chlorin f carries a vinyl group.On thisbasis isorhodoporphyrin also should have a vinyl group at C, andshould contain two hydrogen atoms less than rhodoporphyrin. 5OThis is in keeping with the results of catalytic hydrogenation 49, 48and with other experiments briefly mentioned in Fischer’s recentpaper.45 If this can be confirmed, one of the remaining difficultiesof the subject will disappear and the changes can be representedas follows : 51CH2*C0,H (7)Chlorin f (C3$&04N4).Rhodoporphyrin ( C32H3404N4).4 7 J. B. Conant and C. F. Bailey, J. Amer. Chem. SOC., 1933, 55, 795; R.,48 E. M. Dietz and T. H. Werner, ibid., 1934, 56, 2180; A , , 1934, 1371.49 H. Fischer and E. Lakatos, AnnaZerL, 1933, 506, 123; A., 1933, 1308.50 Comparison of the absorption spectra of these substances, accordingto A.Stern and H. Wenderlein, supports the idea that they are not isomeric(2. physikal. Chern., 1934,170, 337; A., 1935, 10).5 1 Dietz and Werner (ref. 48) summarise the position from the other pointof view (namely, that the true and the iso-porphyrins are isomeric and thatthe nucleus of chlorophyll is a dihydroisoporphyrin), which does not takeinto account the unsaturated side chain at C,.1933, 403.N 394 ORGANIC CHEMISTRY.On this basis the term isoporphyrin ceases to have any specidsignificance, beyond implying a 2-vinyldihydroporphyrin.Partial Synthesis of Chlorophyll a.-The synthesis of phyllo-erythrin by Fischer leaves only four main gaps in the completesynthesis of chlorophyll a. These are (i) the introduction of acarbomethoxy-group, (ii) the conversion of the porphin into theisoporphin system, (iii) the introduction of the phytyl group, and(iv) of the magnesium atom.The last two stages have beenachieved by Pischer.R. WTillstatter and L. Forsen 52 introduced magnesium intophzophytin and other chlorophyll derivatives by the action of theGrignard reagent, a method also used by A. Stoll and E. Wiede-mann.53 Pischer and his collaborators used the product(RO-MgBr) of the decomposition of the Grignard reagent with analcohol, in order to avoid secondary reactions with carboxy- orcarbomethoxy-groups,5* but his latest paper on the subject 55makes it doubtful if any of these processes are satisfactory for thepreparation of phase-positive material.The method finally foundsatisfactory is the interaction of the phorbide with a compound ofthe RO*MgBr type in the presence of pyridine and an excess ofmagnesium. 55 By this process pure phase-positive chlorophyllide acan be prepared from methyl phzophorbide a more readily than bythe chlorophyllase reaction. This confirms the relationship betweenthe two substances.The esterification of the propionic acid group by phytyl alcoholwas achieved by It. WillstBtter and A. Stoll 56 by a biologicalmethod. H. Pischer and W. Schmidt 57 have successfully appliedEinhorn’s method.58 A solution of phzophorbide a in pyridine ontreatment with phytyl alcohol and carbonyl chloride gave phzo-phytin a indistinguishable from the natural product. Whenhydrolysed in the presence of chlorophyllase, it regenerated phEo-phorbide a ; hence there was no replacement of methyl by phytylat Cl0.Other esters of phzophorbide were prepared by this processand the interesting observation was made that the enzymic actionof chlorophyllase is relatively but not absolutely specific; e.g., itcatalysed the hydrolysis of geranyl and cetyl phEeophorbides, butnot that of the methyl and the bornyl ester.5862 Annalen, 1913, 396, 180; A., 1913, i, 499.63 Naturwiss., 1932, 20, 630.64 H. Fischer and M. Diirr, Annalen, 1933, 501, 107; A , , 1933, 515; H.5 5 H. Fischer and G. Spielberger, ibid., 1934, 510, 156; d., 1934, 785.G 6 Ibid., 1911, 380, 148; A., 1911, i, 391.17 Ibid., 1935, 519, 244; A., 1382.68 Ibid., 1898, 301, 95; A., 1898, i, 577,Fischer and J.Riedmair, ibid., 1933, 506, 118; A., 1933, 1308LMSTEAD : THE POXPHYRIN GROUP. 395Phaeoporphyrin a5 has not yet been synthesised 59 and it appearsthat the porphin --+ isoporphin conversion will be difficult.Synthesis from 2-de-ethylporphyrins, followed by introduction ofthe vinyl group, may be necessary.Chlorophyll b.-The study of this compound is not so welladvanced as that of chlorophyll a, and only a brief summary ofprogress will be given here. As has already been stated, chloro-phylls a and b are very similar in general properties, and phyllidesand phorbides of the b series can be prepared by the usual methods.The formula proposed by Willstatter for chlorophyll b(C,,H,,O,N,Mg) is generally accepted. This contains one moreatom of oxygen and two less of hydrogen than that of chlorophyll a.Drastic alkaline degradation of chlorophyll b gives pyrro-, rhodo-,and phyllo-porphyrins, but in worse yield than in the a series; 6ohence the essential arrangement of the substituents is the same.Phzophorbide b on complete reduction with hydrogen iodideyielded hzemopyrrole and hzmopyrrolecarboxylic acid.,1 Will-statter showed that phaeophorbide b is converted by rapid hydrolysiswith hot alkali into rhodin g, which is a tribasic acid containing7 atoms of oxygen and corresponds to chlorin e of the a series.Theextra oxygen atom of the b series was first proved by Conant 62t o be located in a carbonyl group by the preparation of a semi-carbazone from phBophorbide b.This was shown independentlyby 0. Warburg, who converted phzophorbide b into phzoporphyrinb6 (corresponding t o phzoporphin as) by hydrogen iodide, andprepared an oxime from this.63 Fischer and his collaboratorsalso have obtained a monoxime from rhodin g 64 and one fromphsophorbide b.65 The porphyrins obtained by hydrogen iodidedegradation of phaeophorbide b and rhodin g showed reactionsparallel to those of the compounds of the a series; which indicatedthat a carbocyclic ring containing a $-acidic keto-group was alsopresent in phaeophorbide b.", 66 I n 1934, A. Stoll and E. Wiede-59 Cf. H. Fischer and T. Scherer, Annden, 1935, 519, 236; A., 1382.60 Willstiitter, " Chlorophyll," p. 334; cf. A. Treibs and E. Wiedemann,Annalen, 1929, 471, 146; A., 1929, 941.61 H.Fischer, A. Merka, and E. Plotz, ibid., 1930, 478, 299; A., 1030, 620.6% J. B. Conant, E. M. Dietz, and T. H. Werner, J . Arner. Chern. Soc.,1931, 53, 4436; A., 1932, 174.63 Biochern. Z., 1931, 235, 1 ; 1932, 244, 9; A., 1931, 661.64 H. Fischer, F. Broich, S. Breitner, and L. Nussler, AnnaZen, 1932, 498,228; A., 1932, 1263.66 H. Fischer, S. Breitner, A. Henhchel, and L. Niissler, ibd., 1933, 503,1 ; A., 1933, 839; cf. A. Stoll and E. Wiedemann, Helv. Chirn. Acta, 1932,15, 1132; A., 1932, 1265.6 6 H. Fischer, A. Hendschel, and L. Niissler, Annalen, 1933, 506, 83; A.,1933, 1173; H. Fischer and S. Breitner, ibid., 1934, 510,183; 611, 183; A.,1934, 785, 907396 ORQANIC CHEMISTRY.mann prepared a dioxime of methyl phzeophorbide b and thusconclusively proved the presence of two carbonyl groups.The compounds of the b series therefore contain one carbonylgroup in place of a methylene group of the a series.Conant firstplaced the additional carbonyl group as a ketonic bridge (at Cp)and Fischer put it in the propionic acid side chain (at C,). Sub-sequently Pischer showed that it was present as a nuclear aldehydo-group on C,.67 Rhodin g trimethyl ester was degraded t o 3-de-methyldeoxophylloerythrin, the structure of which was proved byThese results are embodied in the formula (XLVII) ofH. Fischer and A. Stern,69 which accommodates the optical activityof chlorophyll b and phzophorbide b discovered by A. Stoll andE. Wiedema~m.~O This is the most recent formula proposed by(XLVII.)I C0,PhytylFischer, but it would presumably now be modified with respect tothe position of the two '' extra " hydrogen atoms so as to come intoline with formula (XLVI) for chlorophyll a.The presence of thevinyl group a t C3 is inferred by analogy, but has not been proved.There now appears to be no justification for the supposition thatchlorophyll a and b are readily interconvertible either in the plantor in the laboratory.71Other Natui4al Porphyrin Derivatives.BacteriochZoro~h.yZZ.-- Some bacteria have the power of assimilat-ing carbon dioxide and recent investigations have shown that theycontain a pigment closely resembling chlorophyll. K. Noack andE. Schneider 72 first isolated the pigment of certain red sulphur orpurpurbacteria of this type.They found that this compound,bacteriochlorophyll, contained magnesium removable by acids to6 7 H. Fischer, A. Hendschel, and L. Nussler, Annalen, 1935,516, 61 ; A., 530.6 8 H. Fischer and W. Rose, ibid., 1935, 519, 1 ; A., 1134.70 Helv. Chim. Acta, 1933, 16, 307; A., 1933, 515.71 A. Stoll and E. Wiedemann, Naturwiss., 1932, 20, 889; A., 1933, 167.72 Ibid., 1933, 21, 836; A . , 1934, 112.Ibid., p. 5 8 ; A., 1134LINSTEAD : THE PORPHYRIN GROUP. 397give a substance resembling phaeophytin. Two carboxyl groupswere found to be present, one esterified by methyl, the other byphytyl or a similar alcoho1,73 and v:Lrious derivatives were prepared.Noack and Schneider considered that two pigments were present, ofwhich the one in larger amount resembled chlorophyll b and contained6 atoms of oxygen in the molecule.H. Pischer and J. Hasenkamp 74investigated the same (or a similar) substance, isolated by H.Gaffron from pure cultures of Thiocystis violacea. They foundthat the predominating pigment belonged to the a series and hadan absorption spectrum of the “0x0 ”-type (see p. 381). Thecrystalline bacterio-methyl phzophorbide a prepared from it gave,on treatment with hydrogen iodide, two porphyrins, one of whichwas identical with oxophzoporphyrin a5 ester. From this resultthe sixth oxygen atom was placed in an acetyl group on C,.Fischer’s formula 75 for bacterio-methyl phaeophorbide a is(XLVIII), but the fine structure would presumably now be modifiedVH, C02MeC02Meto agree with that of the last formula(XLVI). On Fischer’s formulation, thischlorophylls are all structurally derivedparatively simple processes of oxidationgroups and central metal being neglected) :for chlorophyll a itselfsubstance and the twofrom haemin by com-or reduction (the esterSubstituent at7 -----A -___.-- ~Substance. c,. c,. cp C6. c, -H s m i n ............ CHXH, CH, CH:CH, CH,:CH2*CO2H CH,.CH,CO,HBacterio-chloro-Chlorophyll a.. . . . . . . C2H, CO*CH.CO,E *Chlorophyll b.. . . . . . . CHO .. 1, 9 ,9 ,phyll ............ CO-CH, CH3 .. 9 9 Y ,* Cyclised on t o Cy.73 E. Schneider, 2. physiol. Chem., 1934, 226, 221; A., 1934, 1265.74 Annalen, 1935, 515, 148; A., 362.7 5 H. Fischer and J. Hasenkamp, &id., 1935, 519, 42; A., 1134398 ORGANIC CHEMISTRY.Porphyrins in Minerals.-A. Treibs has made an interestingstudy of the occurrence of porphyrins in a number of minerals.In 1934, from the bitumen of an oil shale from the KarwendelMountains, he isolated two porphyrins,76 deoxophylloerythrin(XLIX, R = CO,H), and the decarboxylated compound, deoxo-phylloerythro-ztioporphyrin (XLIX, R = H).CH,R CH2RThe structure of the first was known, that of the latter was provedshortly afterwards by synthesis.77 Porphyrins were next foundin other shales, in petroleums 78% 79 (from Galicia, Trinidad, etc.), inasphalts 78 (Trinidad, Dead Sea, etc.), in phosphorite~,7~ and inmany Mineral waxes appeared to contain very littlep o r ~ h y r i n . ~ ~ The richest sources were Trinidad oil (which con-tained about 0.04% of total porphyrin) and particularly a bituminousmarl from Switzerland (0.4%).79 These quantities are astonish-ingly high when compared with the chlorophyll content of driedleaves (0.8%).The mineral was extracted successively with acetic acid andchloroform, and the extract treated at 50" with hydrobromic andacetic acids.76 The porphyrins were finally purified by acidfractionation. Besides the two compounds mentioned above,which are derivatives of chlorophyll, Treibs has identified meso-ztioporphyrin (L, R = K) and mesoporphyrin (L, R = C0,H) of thehamin series.78 Coproporphyrin was detected in Thechlorophyll porphyrins are always present in greater amount thanthose from hzmin, and the decarboxylated compounds predominateover the free acids. The deoxophylloerythrin and the derivedztioporphyrin from the Swiss marl occur naturally combinedwith vanadium in the centre of the large ring, as compounds of76 Annalen, 1934, 509, 103; A., 1934, 387.7 7 H. Fischer and H. J. Hofmann, ibid., 1935, 517, 274; A., 871.78 A. Treibs, ibid., 1934, 510, 42; A., 1934, 629.'9 Idem, ibid., 1935, 517, 172; A., 727.Idem, ibid., 1935,520,144; A., 1347LTNSTEAD : THE PORPHYRIN GROUP. 399the type >VO, >VO, or >V(OH),.7g It appears probable that thevanadium enters the complex as a result of a secondary reaction.A stable ferrous complex also has been detected.79 There can belittle doubt that chlorophyll and h%min are the parent substancesfrom which these porphyrins are formed by geological processes.The purity of the products makes it unlikely that there has beenring opening and resynthesis. The production of vanadyl deoxo-p hy lloerythr o - ze t iop orp h yrin from c hloroph y 11 a involves the st ages :(i) replacement of magnesium by vanadium, (ii) elimination of thelabile carbomethoxy-group at Clo ; (iii) conversion into the porphyrinsystem, and (iv) decarboxylation a t C,. The decomposition ofphaophytin to deoxophylloerythro-:etioporphyrin in petroleum at360" has been reproduced artificially.A number of interesting deductions have been drawn by Treibs.The results prove that plants play a major part in the formationof petroleum. The r81e of animals appears less important. Thedecompositions which they undergo must be comparatively mild ;this agrees with the discovery of optical activity, and recently ofcestrogenic properties, in petroleums. Moreover the two carb-oxylated porphyrins, mentioned above, act as recorders of themaximum temperature to which the oil has been subjected, fordirect experiment shows that mesoporphyrin is decarboxylated bya week's heating at 240". Treibs estimates that oils or mineralscontaining acid porphyrins cannot have been subjected t o a tem-perature of more than 200" a t any stage during their formation.It has been necessary to omit a great deal of interesting materialin compiling this Report on the porphyrin field. As usual, themore complete topics have been discussed.R. P. L.E. H. YARMER.E. L. HIRST.R. P. LIXSTEAD.S. PEAT.E. E. TURNER.I?. 8. SPRING