ORGANIC CHEMISTRY.1. INTRODUCTION.THE year under review, like its predecessors, has witnessed rapidadvances in numerous and diverse fields of organic chemistry. Onthe more general and theoretical side the liaison with recent develop-ments in physics and mathematics has been maintained, andphysical, stereochemical and kinetic studies have all played theirpart in the elucidation of the mechanisms of the reactions of organiccompounds. Simultaneously the organic chemist has continued histask of determining the structures of many complex substances ofgreat biological significance.In last year’s Report, considerable attention was given to themechanism of aliphatic substitution, prototropy and anionotropy,and the progress of our knowledge over a period of some five yearswas reviewed; a survey of work upon molecular rearrangements,covering approximately the same period, is now included.Studiesof the Hofmann, Lossen and Curtius degradations have establishedthe intramolecular character of these changes ; their detailedmechanisms probably resemble those of the Beckmann, imino-etherand amidine transformations. The rearrangements of hydroxy-sulphones and allied compounds are similar. There are also indica-tions that changes of the pinacol-pinacolone type are intramolecular ;a mesomeric organic cation appears to be involved. Migrationsfrom side-chain nitrogen to nuclear carbon have received muchattention in the past, and the chloroamine transformation inaqueous solvents was long ago shown to be intermolecular; thishas now been confirmed by experiments using radioactive hydrogenchloride, but the position is less simple than was formerly supposed,and in anhydrous solvents a different mechanism is indicated.The intermolecular character of the diazoamino-aminoazo-con-version has been established more definitely, and the Hofmann-Martius rearrangement of alkylanilines appears to involve theseparation of the alkyl group as cation.The conversion of arylalkyl ethers into nuclear-alkylated phenols has been the subjectof a number of recent studies; there are indications that the alkylgroup becomes free as a neutral radical except in the o-migrationsof ally1 groups, which are intramolecular.In recent discussions of condensations of the aldol, Perkin,Knoevenagel and Claisen types, the close relationship existingbetween these processes has been emphasised.An aldol additioM!CRODUC!I!ION. 187product is envisaged in every case, and the function of the catalysthas been considered. Recent studies of the Perkin synthesis supportthe view that the aldehyde condenses with the anhydride, the saltacting catalytically.A comprehensive review of the effects of chemical constitutionupon the dissociation constants of organic acids has appeared.* Itis now shown that the strengths of the o-halogeno- and o-nitro-benzoic acids and -anilines do not indicate the operation of anyeffect which is not also found in the isomeric p-compounds; the'' ortho-effects " observed in many of the reactions of o-substitutedacids, acid derivatives and amines may thus often be absent inionic equilibria.The discovery of the Kharasch peroxide effect in 1933 has led toa revision of the whole subject of additions at olefinic linkages.Ithas been shown that the orientation of addition of hydrogen bromide(but not of the other halogen hydrides) is determined by the oxygenor peroxide content of the system. Further examples of theinfluence of oxygen and peroxides are rapidly emerging. Thusaddition of bromine to olefins and bromination of the side chains ofbenzene homologues are greatly affected by oxygen, and the Can-nizzaro reaction of aldehydes and the addition of bisulphite toolefins do not proceed in the absence of oxygen. Sulphuryl chloridein presence of benzoyl peroxide is a specific reagent for the chlorin-ation of benzene side chains.In the postulated mechanisms ofthese reactions, halogen atoms or organic free radicals play animportant part.Recent advances in the study of stereochemistry have beenassisted by the introduction of the valuable Stuart models withtheir accurate interatomic distances and valency angles. Opticalactivity due to restricted rotation has been observed in the benzeneseries, and the effects of numerous groups upon the rate of racemis-ation of diphenyl derivatives have been further studied. Themechanism of the asymmetric catalytic dehydration of an ally1alcohol to give an optically active allene hydrocarbon remainsobscure. Further investigation of compounds containing sym-metrically placed hydrogen and deuterium has shown that thedifference between these isotopes is insufficient to give rise todetectable optical activity.In recent work upon asymmetrictransformation or optical activation, the problem has for the firsttime been studied kinetically.The resolution of a tervalent nitrogen compound remains one ofthe outstanding problems of stereochemistry. The evidence for thenon-planar codguration of the tervalent nitrogen valencies is* J. F. J. Dippy, Chem. Review4 1939, 25, 161188 ORGANIC CHEMISTRY.reviewed in this Report, and some promising investigations onsubstituted ethyleneimines are described. The isolation of thecis-form of azobenzene by irradiation of the normal form has stimul-ated the investigation of cis-trans-isomerism in general, and it hasbeen found that the two forms of many azo-compounds can beseparated by chromatographic analysis with alumina or charcoalas adsorbent. The equilibrium between syn- and anti-diazocyanideshas also been studied, and measurements of the dipole moments haveconfirmed Hantzsch’s conclusions regarding their configurations.The elegant method employed for demonstrating the planar con-figuration of the 4-covalent plktinous atom has now been applied topalladium, and it has been shown that this metal too, in the palladouscondition, must possess planar valencies.The Report on carbohydrates deals with two aspects of thesubject only; first, the development in conceptions of opticalinversion in the sugar group makes opportune a review of work onthe anhydro-sugars and on glucosamine, and secondly the constitu-tions of starch and cellulose are discussed.It is established that thealkaline hydrolysis of a sugar p-toluenesulphonate takes placereadily and is accompanied by Walden inversion on the carbonatom to which the sulphonic acid group is attached, provided thatthere exists a free hydroxyl in the trans-position on an adjacentcarbon atom ; an anhydro-ring of the ethylene oxide type is formed.When an adjacent trans-hydroxyl group is not available, then eitherthe sugar toluenesulphonate is not hydrolysed, or hydrolysis occurswith extreme difficulty and without optical inversion. Scission ofthe anhydro-ring is effected by the further action of alkaline reagents ;the rupture is accompanied by inversion and takes place on eitherside of the oxygen atom, with the result that a mixture of twosugars is obtained from each anhydro-compound.A derivativeof dimethyl 2-amino-~-methylglucopyranoside (which is obtainableas a product of the action of ammonia on dimethyl 2 : 3-anhydro-p-methylmannoside) proves to be identical with the correspondingderivative prepared from natural glucosamine, which must thereforehave the configuration of d-glucose and not that of d-mannose.Two other types of anhydrohexoses are discussed. In the glucosanclass the reducing group is involved in the anhydro-bridge formation,and in the pentaphan ring type the anhydro-ring is six-memberedand does not involve the reducing group; the outstanding featureof the latter is the great sta.bility of the anhydro-ring, which is notruptured by the strongest acid or alkaline hydrolytic agents.Variations in the conditions of methylation of starch or the useof starches from different biological sources all yield productshaving the same proportion of end-group (tetramethyl glucose) INTRODUCTION.189corresponding to a chain length of 24-30 glucose units. Starchthus appears to be best represented by the laminated formulabelow, in which each repeating unit of 24-30 glucose residues isrepresented by a straight line terminated by an arrowhead indicatingthe reducing glucose unit.J. II4 IJ. IProlonged methylation or treatment with weak acid brings about ascission of the ( ( polymeric links ” between the repeating units and aconsequent diminution of molecular size as determined by physicalmethods. A study of the kinetics of this disaggregation processindicates that the (‘ polymeric bonds ” are of the glycosidic type.Evidence for the existence of basal repeating units in cellulose isnot so definite.The amount of end-group isolated from a methyl-ated cellulose is more or less proportional to the number of methyl-ations to which it has been submitted; if air is excluded, themolecular size diminishes as methylation proceeds, but the amountof end-group isolated from the product bears no simple relation tothe molecular weight determine’d osmotically, and indeed end-group is entirely absent when the methylation in nitrogen is notcarried beyond five or six repetitions.A simple explanation repre-sents a methylated cellulose of high molecular weight by a largeloop (a), further methylation leading to scission simultaneously attwo points as shown by the dotted line. In the absence of oxygen,the ‘( healing ” process represented by the union of the “ head ” and(( tail ” of each fragment in ( b ) produces the smaller loops ( c ) inwhich no end-group will be apparent. Oxygen appears to inhibitthis healing process, for when the methylation is conducted in airthe amount of end-group in the product indicates that most of theunits exist as terminated chains ( b ) . The retention of the loopedstructure in the fragments ( b ) is ascribed to cross-linkages as shownin ( d ) ; the nature of these cannot at present be specified.IReference is also made to investigations of the relationships ofhydrocellulose and oxycellulose with cellulose, and of the dis-ruption of the glucopyranose units in cellulose or starch when thepolysaccharides are oxidised with periodic acid190 ORGANIC OHEMISTRY.Recent researches have shown that the a-naphthaquinone nucleusis a frequent constituent of the molecular structure of naturalproducts, and 8 short account of the pigments based upon thisskeleton introduces a topic which has not previously been dealtwith in these Reports.Structural relationships have been estab-lished between juglone, plumbagin, lawsone, and phthoicol, all ofwhich are hydroxy- or methyl hydroxy- a-naphthaquinones, andechinochrome A is probably an ethylpentahydroxy-a-naphtha-quinone. Extensive studies have shown that lapachol and lomatiolare derivatives of 2-hydroxy-3-isopropenyl-1 : 4-naphthaquinone.It is probable that dunnione is a related 1 : 2-naphthaquinone, andalkannin, alkannan and shikonin are isopropenyldihydroxy- 1 : 4-naphthaquinones.The antihzemorrhagic vitamin Kl is S-phytyl-2-methyl-1 : 4-naphthaquinone ; the constitution has been establishedby degradation and by three independent syntheses of the vitamin.The marked physiological activity of 2-methyl-1 : 4-naphthaquinoneis noteworthy. Intensive attacks on the structure of gossypol,the yellow toxic principle of cotton seed, have been made; thepigment is not a quinone, but the attractive hexahydroxydinaphthylstructure composed of six isoprene units which has been suggestedis supported by degradation and synthetic evidence of considerablemagnitude and importance.Great advances have been made in the synthesis of steroidsduring the period 1937-39.The complete synthesis, by establishedmethods, of the sex hormone equilenin has been announced. z-Nor-equilenin and 2-noroestrone have been prepared by an ingeniousdouble cyclisation process. The diene synthesis has been success-fully applied to the formation of a stereoisomer (or perhaps a,structural isomer) of oestrone. In view of the difliculty of achievingthe complete synthesis of any sex hormone, special interest attachesto the development of the potent synthetic oestrogens, the C-alkylderivatives of 4 : 4’-dihydroxystilbene (“ stilbastrol ”).Con-siderable progress has been made in the synthesis of the tricyclichexatriene skeleton of the antirachitic vitamins.The review of heterocyclic compounds is again very largelydevoted to the chemistry of natural products. The section onoxygen ring compounds is a development of the topics discussed inthe Reports for 1937 and 1938; benzofuran structures have beenadvanced for euparin and egonol, extensions have been made in thecoumarin, furocoumarin, furochromone, and furoflavone groups,and an investigation on fustin has established the natural occurrenceof the flavanonol structure. Among heterocyclic nitrogen ringderivatives, adermin, the rat-dermatitis-preventing factor of thevitamin B complex, proves to be a comparatively simple thougWATSON : REACTION MECHANISMS.191highly substituted pyridine derivative. Interesting advances havebeen made in the chemistry of the bicycb-aza-alkanes. Newmethods have been devised for the preparation of bicyclic basesof types I (n = 1 or 2), I1 (n = 2 or 3), I11 (1% = 3 or 4) and IV.An earlier unconventional stereochemical explanation of the exist-ence of two forms of norlupinan (I11 ; n = 4) must now be discarded,as norlupinan B has been proved to be identical with the syntheticbase (IV). Attempts to utilise these bicyclic bases in the synthesisof quinine derivatives have been made, and developments in thisdirection will be awaited with interest.The alkaloidal field has not produced any spectacular advancesduring the year, and this section of the Report is devoted mainlyto the speculative structures with the aid of which the chemistryof calycanthine, matrine, artabotrine and the ergot bases is inter-preted. An extremely ready rupture of the terminal pyrrolidinering of the eserethole molecule has been discovered.R.D. HAWORTH.P. MAITLAND.S. PEAT.J. C. SMITH.H. D. SPRINGALL.H. B. WATSON.2 . REACTION MECHANISMS.(Continued from Ann. Reports, 1938, 35, 208.)(a) Rearrangements.Attention continues to be given to those changes in which agroup migrates from one position in a compound to another, eitherwith or without the elimination of water, halogen acid or someother simple molecule.Work in this field up to 1933 has beensummarised in earlier Reports,l and a comprehensive account of1 Ann. Reports, 1923,20,116; 1924,21,96; 1925, 22, 113; 1927, 24, 154;1928,25, 133; 1929, 26,122; 1930, 27,114; 1933,30,176192 ORGANIC CHEMISTRY.the mechanisms which have from time to time been suggested torepresent certain rearrangements has recently appeared.2 In anumber of instances there is definite evidence for the intermolecularcharacter of the change, the migrating group becoming detachedfrom the molecule at some intermediate stage, while certain otherrearrangements (e.g., the benzidine change) have been. shown tobe truly intramolecular. The evidence is often far from con-clusive, however ; moreover, even if the transformation is knownto be intermolecular, the nature of the intermediate may not beclear, while the exact mechanism of an intramolecular rearrange-ment is not always easy to determine.It is therefore not sur-prising that the literature should contain a large number of postu-lated schemes in which compounds such as olefms or alkyl halides,cyclic compounds, ions, radicals or molecules with free valenciesappear as intermediates, in addition to mechanisms of an intra-molecular type involving partial valencies or their electronicequivalent,The Hofmann, Lossen, and Curtius Rearrangements.-Accordingt o the general scheme put forward by F. C. Whitmore,* thesechanges (in all of which an isocyanate precedes the ultimate product)are to be represented as follows :The removal of A as a negative ion (or nitrogen molecule in theCurtius degradation) leaves the adjacent nitrogen atom with anincomplete electronic group (sextet), and R is then transferred,with the R-C electron pair, from carbon to nitrogen.This schemedoes not differ in principle from C. K. Ingold’s representation ofthe pinacol and Wagner transformations.Kinetic studies of the rearrangements of the salts of a numberof nuclear- subst itut ed benz obromoamides , X C,H,*CO*NHBr,and dibenzhydroxamic acids, X*C,H4*CO*NH*O*CO*C,H5 and~ , H , * ~ O * ~ ~ O ~ C O * ~ , H , X ‘ , in aqueous ammonia have been con-ducted by C. R. Hauser and his collaborators,6 who conclude thatthe rate-determining step is the release of Br, O*CO*C,H, orE.S. Wallis in Gilman’s “ Organic Chemistry,” New York, 1938, Vol. 1,p. 720.See Ann. Reports, 1933, 30, 178.Ibid., 1928, 25, 133.4 3. Amer. Chem. SOC., 1932, 54, 3274; Ann. Reports, 1933, 30, 176.6 C. R. Hauser and W. B. Renfrow, J . Amer. Chem. SOC., 1937, 59, 121,2308; R. D. Bright and C. R. Hauser, ibid., 1939, 61, 618w m s o N : REACTION MECHANISMS. 193O*CO*~,H,X’ as anion. The rearrangements are facilitated byelectron-repulsive characters in X and electron-attractive charactersin X’. Further, the logarithms of the velocity coefficients for thetransformations of C,H5’CO*NH’OoCO*C6H4X‘ give a straight linewhen plotted against the values of log K for the acids X’*C,H,*CO,Heven when X’ is in the o-position, but the linear relationship is notfound for the series X*C,H,*CO*NH*O*CO*C,H, ; i.e., the quanti-tative correlation of log k with log K holds only when the variablegroup is in the anionic portion of the compound (as it is, of course,in the acid).The further step, R*CO*N-+ CO:NR, consists in the trans-ference of R from carbon to nitrogen, and the absence of racemis-ation during the Curtius, Hofmann, and Lossen degradations ofoptically active derivatives of benzylmethylacetic acid, and theidentity of the rotations of the amines or ure2s obtained from theazide, the bromoamide and the hydroxamate, observed by E.S.Wallis and his co-~orkers,~ point, very strongly to the conclusionthat the process is truly intramolecular. This view is confirmedby the production of only one isocyanate when benzylmethylacet-azide is degraded in presence of triphenylmethyl radicals,S by thequantitative yield of tert.-butylmethylamine (neopentylamine) fromtert.-butylacetamide (whereas reactions in which dissociation maybe supposed t o occur lead to tert.-amyl derivatives by isomerisationof the cation CMe,*CH,+), and by the complete preservation ofoptical activity (due in this case t o restricted rotation) during theHofmann degradation of d-3 : 5-dinitro-6-a-naphthylbenzamide7where dissociation a t any stage of the reaction would have per-mitted free rotation and resulted in a racemic product.lO More-over, C. L. Arcus and J. Kenyon11 have recently observed analmost quantitative retention of asymmetry when optically activehydratropamide , CHPhMe- CO *NH,, undergoes the Hofmanndegradation ; this offers a striking contrast to the very considerableor almost complete racemisation which occurs when optically activegroups migrate in certain other systems, e.g., from oxygen to sulphurin the change CHPhMe*O*SOC,H, --+ CHPhMe-SO,*C,H,.l2 It7 L. W. Jones and E. S . Wallis, J . Anzer. Chem. Soc., 1926, 48, 169;5:. S . Wallis and S. C. Nagel, ibid., 1931, 53, 2787; E. S. Wallis and R. D.Dripps, ibid., 1933, 55, 1701.8 E. S. Wallis, ibid., 1929, 51, 2982.9 P. C. Whitmore and A. H. Homeyer, ibicl., 1932, 54, 3435.lo E. S . Wallis and W. W. Moyer, ibid., 1933, 55, 2598; Ann. Reports,1 1 J . , 1939, 916.l!):3:3, 30, 178. Compare F. Bell, Cherrh. ant1 Id., 1932, 52, 584.<J. Ihrlyorl ant1 H. l’hillip, J., 1930, 1 ( i i c i ; CJ.I;. Arms, &I. P. Balfe,mid J. Kenyon, J . , 1938, 483.REP.-VOL. XXXVI. 194 ORGANIC CHEMISTRY.must be concluded, on the basis of the above evidence, that intransformations of the Hofmann, Lossen, and Curtius types themigrating group never achieves freedom as radical or as ion. Al-though it appears that a carbanion ( L e e , a negatively charged alkylgroup which has separated from a molecule in such a way thatthe carbon has its complete octet) may retain its asymmetry to alarge degree,13 the observations referred to above are not easilyexplicable in any other manner.F. C. Whitmore considers that the rearrangement is intramole-cular,14 but it is nevertheless difficult to visualise the transferenceof the group R with its electron pair without postulating its freedomfor an instant of time.This difficulty is overcome in the followingslightly modified scheme,15 where the arrows indicate movementsof electrons :The removal of A is followed (or accompanied) by the migrationof the R-C electron pair to the C-N bond and the attachment ofR by the unshared electrons of the nitrogen (Le., R is severed fromcarbon and linked to nitrogen at one and the same time). In thecase of a dibenzhydroxamate the mechanism becomes‘A! R’ d 7R Nk-O-C/ R N R-N\ / i \O \A!/ //- - - f G - - - + f l0 0 G 0Electron-attractive substituents in R‘ will promote the removal ofOCOR’ as anion, and the observed relationship with the dis-sociation constants of substituted benzoic acids is to be anticipated ;on the other hand, electron-repulsive groups in R will facilitate themigration of electrons from the R-C bond, and the analogy withtheir effects upon the ionisation of a benzoic acid disappears.Theinfluence of the nature of R upon the R-C bond also explains theeffect of phenyl in the series CPh,*CO*NH*OH > CHPh,*CO*NH*OH> CH,Ph*CO*NH*OH, and NPh,*CO*NH*OH > NHPh*CO*NH*OH.16The Beckmann rearrangement may also be represented by al3 E. S. Wallis and F. H. Adams, J . Anter. Chem. SOL, 1933, 55, 3838.1 4 Sce J . Amer. Claeiir. SOC., 1934, 56, 1427; J., 1934, 1269.15 H. B. Watson, “ Modern Theories of Organic Chemistry,” Oxford16 L. W. Jones and C. I>. Hurd, J . Amer. Cfien,.b‘oc., 1921, 43, 2422; C. D.University Press, 1937, Chap. X.Hurd, {bid., 1923, 45, 1472WATSON : REACTION MECHANISMS. 195mechanism of a similar type (the group A does not here leavethe molecule), and the transformations of imino-ethers and amid-ines l7 and the rearrangements of quaternary ammonium saltsobserved by Stevens and co-workers l7 appear to be comparable.The following mechanism for the last-named change is the same inits essentials as that suggested by the workers themselves, and isin harmony with the observed effects of variations in the natureof R and R’ :tR-CH2-NMe, R-CH-he, c ( -+ R*CHR’*NMe2L WI +R’All these rearrangements have one feature in common; thegroup migrates to an atom possessing unshared electrons, and it iseasy to visualise its attachment by these electrons simultaneouslywith the release of the pair by which it was linked originally.It isless easy to visualise an intramolecular process which occurs underconditions other than these.The Benxilic, Pinacol, Wagner, and Allied Transformations.-Akinetic study of the transformation of benzil into benzilic acid in32% alcohol (where interference by production of benzoate andbenzaldehyde is but slight) has shown that it is of the first orderwith respect to both diketone and hydroxyl ion.l8 This confirmsthe view, already put forward by R. Robinson and others, that anintermediate negative ion Ph*CO*C(OH)Ph*6 is formed.19 Thesame conclusion is drawn from the observation that benzil exchangeswith heavy oxygen water far more rapidly in alkaline than inneutral solution.2° Further, if the o-positions of both nuclei areoccupied by methyl groups (so that addition to carbonyl is “ steric-ally ” hindered), the rearrangement does not occur.21P.D. Bartlett and I. Pockel’s observations of the pinacolicrearrangements of cis- and trans-1 : 2-dimethylcyclohexane-1 : 2-diols indicate that the elimination of hydroxyl and the attach-ment of the migrating group occur on opposite sides of the ringcarbon atom.22 Confirmation is found in the further observation 23l7 See Ann. Reports, 1930, 27, 121.l 8 F. H. Westheimer, J . Amer. Chewe. SOC., 1936, 58, 2809.Ann. Reports, 1923, 20, 118. The author here summarises a number ofobservations which are. in harmony with this conclusion, particularly theisolation of an addition compound of benzil with potassium hydroxide, whichwnrrangcs to give potassium benzilate (G.Scheuing, Ber., 1923, 56, 252).2o I. Roberts and H. C. Urey, J . Amer. Chem. SOC., 1938, 60, 880.2 1 13. P. Kohler a i d R. Baltzly, ibid., 1932, 54, 4019.22 Ibid., 1937, 59, 820.23 P. D. Bartlett and A. Bavley, 2’6id., 1938, 60, 2416196 ORGANIC CHEMISTRY.that when the geometrical isomerides of 1 : 2-diniethylcyclopentane-1 : 2-diol are refiuxed with 30% sulphuric acid, the cis-form gives2 : 2-dimethylcycbpentanone in 137% yield whereas the trans-isomeride produces only brown tars ; the migration occurs Onlywhen the methyl group can displace the hydroxyl by attacking theopposite side of the carbon atom holding it.A similar " rearwardattack " is indicated by the fact tha,t the semipinacolinic deamin-ation of (I) to (11) involves a Walden inversion : 2pPh PhMe II I)C-c-Ph M e \ L - P h (11.)H' II (1.)0 H2N OHTetramethylethylene bromohydrin loses hydrogen bromide at 100"or on treatment with silver oxide or nitrate or with sodium thio-sulphate, to yield pinacolone. The corresponding iodohydrin isstable only in solution ; the solid decomposes spontaneously togive pinacolone and otherCertain new facts have emerged regarding the Wagner-Meerweinrearrangement, which is closely related to the pinacol transform-ation. P. D. Bartlett and I. Pockel find26 that the change ofcamphene hydrochloride t o isobornyl chloride does not proceedspontaneously, but is induced by hydrogen chloride or o-cresol ;the chloride ion is not a catalyst.T. P. Nevell, E. de Salas, andC. L. Wilson have made the further observation 27 that in presenceof hydrochloric acid containing radioactive chlorine the rate ofchlorine exchange in chloroform medium a t 0" is about fifteentimes greater than that of rearrangement, and they conclude thatthe rapid establishment of an ionic equilibrium (as in Meerwein'sscheme) is followed by a slow bimolecular reaction of the organiccation with a molecule of hydrogen chloride.The benzilic, pinacol, and Wagner-Meerwein transformationsdiffer from those dealt with in the preceding section (Hofmann,Lossen, etc.) in that they involve a migration not t o an atomwhich has unshared electrons but to one which is actually deficientin electrons.Nevertheless, the opposite behaviour of cis- andtrans-1 : 2-dimethylcyclopentane-1 : 2-diols (see above) and theabsence of by-products which might be expected if an alkyl groupwere a t any time free as an ion point to an intramolecular migration.Perhaps the organic cation or anion which is presumably firstO4 H. I. Bernstein and F. C. Whitmore, J . Awer. G'hem. Soc., 1939. 61, 1394.25 (2. W. Ayers, ibid., 1938, 60, 2957.O G Ibid.. p. 1585.2 7 J., 1939, 1188WATSON : REACTION MECHANISMS. 197formed in each case is actually a mesonieric structure ; * for pinacoland camphene hydrochloride this ma07 be represented as follows :This view appears to be in harmony with the relative “ migratoryaptitudes ” of different groups in the pinacol change (e.y., aryl >alphyl ; anisyl > phenyl) .z8A number of rearrangements of alkyl groups have been studiedduring recent years by P.C. Whitmore and his collaborators.P. C. Whitmore, E. L. Wittle, and A. H. Popkinzs find that,whereas the reaction of neopentyl iodide with concentrated alcoholicpotassium hydroxide leads to no rearranged products, the far morefacile attack by silver nitrate or mercuric nitrate gives almostcomplete rearrangement to tert.-amyl alcohol. The authors suggestthat in the first example the addition of hydroxyl or ethoxyl ionand the removal of iodide ion proceed simultaneously; the carbonatom is at no time deficient in electrons, and rearrangement doesnot occur.The preparation of neopentyl iodide and bromide byconversion of neopentylmagnesium chloride into the correspondingmercury compound, followed by treatment with the appropriatehalogen, is probably made possible by the same factor; the neo-pentyl group retains its full quota of electrons throughout.30 Thesilver or mercuric ion, on the other hand, first removes iodide ionand the positively charged neopentyl system rearranges to tert.-amyl. Mechanisms of both of these types appear to be involvedin the reaction with potassium acetate, which leads t o both normaland rearrangement products, their proportAons depending upon themedium .z9Rearrangements of Hydroxy-sulphones and Related Compounds,-The investigations of S.Smiles and his collaborators, t o which refer-28 Ann. Reports, 1928, 25, 134; 1930, 2’7, 118.29 J . Arner. Chem. SOC., 1939, 61, 1586.30 F. C. Whitmore, E. L. Wittle, and B. R. Harriman, ibid., p. 1585.* Suggestion by Professor C. K. Ingold (private communication) (comparealso ref. 27). It may be regarded as the electronic equivalent of the earlier‘‘ partial valencies ” view (R. Robinson, Mem. Munchester Lit. Phil. SOC., 1920,64, No. 4).c1 c1 c1Phosphorus pentachloride may prosent an analogous case, ‘uiz.,\hcI . cf ’Y-1L-----198 ORGAMC CHEMISTRY.ence was made in the Report for 1933,31 have been continuedand extended to include the rearrangements of further hydroxy- ,amino-, and acetamido-siilphones, -sulphoxides, and -sulphides.These changes appear to be intramolecular.Thus, in the analogousconversion of a thiol oxide into a hydroxysulphide (X = 0,YH = SHY in I11 and IV), no evidence of fission of the moleculecould be found,32 and, further, the o-hydroxysulphone (I) rearrangeseasily whereas the p-hydroxy-derivative (11) does not rearrange.33It is probable that the transformation occurs in the ion, since it ispromoted by alkali and becomes more facile when the strength orconcentration of the alkali is increased ; for example, when differentbases are used,34 the velocity increases in the order NaOH <NaOMe < NaOEt < NaOPrb, this being the order of effects of thesame bases on three-carbon taut0merism.3~The rearrangements may thus be represented by the generalscheme (111) + (IV), where X = SO,, SO, or S and YH = NHAc,NH,, or OH.The ionisation of the proton from Y may, of course,occur simultaneously with the processes indicated by the curvedarrows. n(111.)c/BCZy/ArThis mechanism is similar in principle t o that suggested above(p. 194) for the Hofmann and allied rearrangements, and is in har-mony with the observed effects of varying X or Y or of introducingsubstituent groups. The constitutional factors which determinethe ease of the rearrangement are (a) the positive character of thecarbon atom of Ar which is linked directly t o X, (b) the positivecharacter of X, ( c ) the tendency of YH t o yield its proton to theacceptor present in the medium, and (a) the capacity of Y t o actas electron donor.The effect of (a) is illustrated by examples in which Ar containsvarious electron-attractive groups. Thus, qualitative experiments31 Ann.Reports, 1933, 30, 188.32 L. A. Warren and S. Smiles, J . , 1931, 914.33 A. A. Levi and S. Smiles, J . , 1932, 1488.34 B. A. Kent and S. Smiles, J., 1934, 422.G . A. R. Ron and R. P. Linstead, J., 1929, 1269WATSON : REACTLON MECHANISMS. 199indicated a decreasing facility of rearrangement in the series%NO,*C,H, > 4-N02*C,H, > 4-MeS0,*C,H,.33. 36 Quantitative de-terminations (by a colorimetric method) of the times necessaryfor the completion of rearrangement of the sulphone (V) to thesulphinic acid (VI) under fixed conditions gave the order 2 : 4-di-NO, > 2-N02 : 4-C1 > 2-N02 > %NO2 for substituents in the Bn~cleus.3~ This is the order of decreasing electron-attractivecharacter as indicated by the dissociation constants of the corre-sponding benzoic acids, and therefore the order of decreasing posi-tive nature of the carbon marked with an asterisk.Other vari-ations of the B nucleus have also been studied.37A similar acceleration by electron-attractive substituent,s in themigrating group has been found in the rearrangements of imino-ethers, amidines,l7 and quaternary ammonium salts,17 and it wouldappear that in all these transformations the important factor isthe facility with which the group can accept electrons from thedonating atom to which it migrates. On the other hand, in theBeckmann, Hofmann, and Lossen degradation^,^^ electron-repulsivesubstituenhs in the migrating group have a favourable effect; thenitrogen which here acts as donor is actually deficient in electrons(owing to the departure or tendency towards departure of ananionic group), and the important factor is the supply of electronsto the C-N bond (a process which has no analogue in the othersystems referred to).Substituents in the A nucleus (B = 2-N02*C,H4) facilitate therearrangement in accordance with their electron-repulsive char-acters, which will, of course, increase the electron-donating capacityof the phenolic oxygen (factor d above).This is in harmony withthe effects of substituents in the non-migrating groups of imino-ethers, amidines and quaternary ammonium salts.It is indicated 34by the sequences 5 : 6-C4H4 > 3 : 5-diMe > $Me, and 5-6 >5-OMe > 4-0. The approximate equality of the effects of 5-Meand 5-OMe is probably due to the intervention of factor ( c ) whenthe group is powerfully electron-releasing. A 6-Me group is extra-s~ L. A. Warren and 5. Smiles, J., 1932, 1040.37 F. Galbraith and S. Smiles, J., 1935, 1234.38 See above, p. 192200 ORGANIC CHEMISTRY.ordinarily effective in facilitating the change ; 39 this appearspeculiar, but may perhaps be due to hydrogen bonding with thesulphone oxygen, the positive character of the sulphur atom thusbeing increased (factor b ) ; the effect is probably analogous to therelatively high strength of o-toluic acid.*The co-operation of factors ( b ) , (c), and (d) is well illustrated bythe effects observed when X and Y (in 111) are varied.34.409 41When Ar = 2-N0,*C,H4 and YH = NHAc, the rearrangementoccurs (under the experimental conditions employed) in the sul-phone, the sulphoxide and the sulphide (X = SO,, SO or S),but when YH = OH (aryl; i.e., C-C in I11 is a portion of thearomatic nucleus) only the sulphone rearranges, and if YH = OH(alphyl) or NH, (aryl) the sulphide does not change.The sulphuratom becomes progressively less positive in the series SO, > SO > S,and hence, owing to factor ( b ) , the rearrangement is most facile inthe sulphone and least in the sulphide. For different YH groups,the ease of rearrangement decreases in the series NHAc > NH, > OH,and OH (alphyl) > OH (aryl).The superiority of alcoholic tophenolic hydroxyl is probably due to the smaller electron-donatingpower of the oxygen in the latter owing to mesomerism. Theorder of basic (electron-donating) characters of the amino- andacetamido-groups is NH, > NHAc, and the above reversal of thisorder must be attributed to the intervention of factor (c). Thecombination of factors ( c ) and (d) is shown further in a moreextended series of substituted amino-gr~ups.~~ In the series "€€Me,NH,, NHAc, NH*SO,Ph, the donor capacity of the nitrogen decreasesand the tendency to release the proton increases. These opposingeffects lead, in the sulphone (VII), to a maximum facility ofrearrangement a t the middle members, where the optimum balanceof the two appears t o be achieved.Moreover, in the correspond-ing sulphides (VIII); where YH is NMe,, NHMe, NH,, NHAc,SO 42 H *NO,(o) /\/S*C,H4*NO,(o)(VIII.) /\/I t (VII.) I IV \ Y H V \ Y HNH*COC,H,*NO,(o), NH*C,H,(NO,),, NH*SO,Ph (in decreasingorder of basicity) only the acetamido- and 2-nitrobenzoyl derivativesrearranged in hot ~/3-sodium hydroxide.The rearrangements of o-aminodiphenyl ethers, studied by K. C.39 C. S . MeClement and S . Smiles, J., 1937, 1016.4O A. Levi, L. A. Warren, and S. Smiles, J., 1933, 1490.41 W. J. Evans and S. Smiles, .7., 1935, 181.* Seep. 216WATSON : REACTION MECHANISMS. 20 1Roberts and his collab0rators,~2 (IX -+ X : R = H or acyl), areof the same type as those considered above. They occur in solvents(X.)./\/OR O*C6H3(NO‘3),(IX.) fi/ _3 I\ANHR ‘/\NH*c6H3(No2) ,such as alcohol and pyridine, and particularly in mixtures of thesewith water, but are retarded by the presence of ions.They arecontrolled by the same constitutional factors as are the rearrange-ments of sulphones and other compounds studied by Smiles, andwhen different substituents, ranging from strongly electron-attrac-five to strongly electron-repelling groups, are introduced into thenucleus, the opposite effects upon factors ( c ) and (d) (above) leadto the observation of a decreased rate of transformation by groupsof both types. The peculiarity of the rearrangement lies in thefact that, when R is an acyl group, this group is found linked tooxygen in the product; the authors suggest that, like hydrogen.the acyl group migrates as a positively charged radical.An extension of the work of Smiles and his collaborators hasembraced systems of the type (XI) ---+ (XII)Ar (XII.)/\/XH - v\,.y/ (XI.)where Ar = C,H,*NO,, Z = CO or SO,, X =SO,, S or 0, andYH = NH, , NHPh, or NHMe.43 Analogous cases are (XIII) ->(XIV) and (XV) --+ (XVI).(XTV.) ~H,*SO,*C,H,*NO, CH,*SO,HCO*NHPh CO*NPh*C,H,*NO,CH,*O*C6H40N0, - CH,*OH ,(xv*) CO-NHAc CO*NAc*C,H,=NO,(XTII.)(XVT.)The reverse transformation ( X I 1 4 XI) occurs on treatment ofcertnin aryl salicylates and phenol-o-sulphonates (Z = CO, SO, ;/\/O9C6H4*~02(XVII.) (\(OH - 1 1 (XVIII.)\/\Co *o *C6H,*NO, \/\CO*OH42 I<. C. Roberts and C. G. M. de Worms, J ., 1934, 727; 1935, 1309;43 W. J. Evans anti S . Smiles, J . , 19.36, 329; R. T. Tozer and S. Smiles,K. C. Roberts, C. G. M. de Worms, and (Miss) H. B. Clark, J., 1935, 196.J., 1938, 2052202 ORGANIC CHEMISTRY.X = 0; YH = OH) with N-sodium hydroxide at lOOO.44 Ex-amples are (XVII) --+ (XVIII) and (XIX) ---+ (XX).0 C H4*N0,SO,*OH(XX.) (XIX.)Migrations from Side-chain to Nucleus.-The well-known work ofK. J. P. Orton and his collaborators upon the transformation ofN-chloroacetanilide into o- and p-chloroacetanilides in aqueousmedia has led to the acceptance of this “rearrangement ” as atypical intermolecular change, involving the intermediate form-ation of acetanilide and chlorine. A recent investigation of thetransformation in 20% alcohol under the influence of radioactivehydrogen chloride 45 has provided further confirmation; all theN-ch1oroacetanili.de which rearranges comes into radioactiveequilibrium with chloride ions and therefore passes through thechlorine stage.Although, however, the production and disappear-ance of chlorine are well established, the position appears to beless simple than was formerly supposed. The existence of a sidereaction producing chloride ions was observed by J. J. Blanksma.46This side reaction was catalysed by acids other than hydrochloric,and Orton and W. H. Gray4’ concluded that hypochlorous acid,formed by hydrolysis of the chloroamine, was reduced by theaniline resulting from further hydrolysis of the anilide. In therecent experiments with radioactive hydrochloric acid 45 a con-siderable correction for the chloride ions thus formed was foundt o be necessary, and A.R. Olson and J. C. Hornel have made afurther study of the side reaction.48 They conclude from theresults of kinetic experiments that three mols. of N-chloroacetanilidegive a compound X and two chloride ions, the reaction beingcatalysed by hydrogen ion. The compound X, which liberatesiodine instantaneously from hydrogen iodide, decomposes slowly,giving another chloride ion and a compound Y, which does notoxidise hydrogen iodide. The proportion of the N-chloroacet-44 Idem, J . , 1938, p. 1897.45 A. R. Olson, C. W. Porter, F. A. Long, and R. S. Halford, J . Amer.Chem. SOC., 1936, 58, 2467; A. R. OIson, R.S . Halford, and J. C. Hornel,ibid., 1937, 59, 1613.Compare F. D. Chattaway and K. J. P.Orton, Proc. Chem. SOC., 1902, 18, 200; K. J. P. Orton and W. J. Jones,Rep. Brit. Assoc., 1910, 85.46 Rec. Trav. chim., 1903, 22, 290.47 Rep. Brit. ASSOC., 1913, 136.4 8 J . Org. Chem., 1938, 3, 7 6 WATSON : REACTION MECHANISMS. 203anilide which gives C-chloroacetanilides in 20% alcohol a t 40” fallsfrom 70% to about 25% when the concentration of hydrochloricacid is decreased from 0 . 0 4 ~ to 0.005~. They conclude that‘‘ the second reaction apparently is prominent enough t o warranta very full investigation of it before any general conclusions basedupon the quantitative study of this rearrangement can be accepted.”The simultaneous occurrence of more than one reaction may per-haps be not unconnected with the apparent dependence of theenergy of activation upon the temperature, as observed by J.0.Percival and V. K. LaMer.4gThe transformations of N - halogenoacylanilides in anhydroussolvents have been examined by R. P. Bell and his co-~orkers.~OI n the case of N-chloroacetanilide, of N-bromoacetanilide and alsoof N-iodoformanilide, general acid catalysis is observed. More-over, no free halogen could be detected in the bromoacetanilidesystem and a small amount of iodine which appeared in the solutionof N-iodoformanilide was quite insufficient to account for the rateof production of p-iodoformanilide. It is concluded that the trans-formations are intramolecular in these solvents. The behaviourof N-bromobenzanilide also resembles that of N-bromoacetanilide.51There has long been evidence (e.g., the wandering of the ArN,group to a foreign nucleus) that the diazoamino-aminoazo-conversionis an intermolecular process. H. V. Kidd 52 has now reported theisolation of benzeneazo-p-naphthol and aniline in 90% of thetheoretical quantities from an acidic solution of diazoamino-and he finds further that, under the correct conditionsof acidity, benzenediazonium chloride and aniline condense to givep-aminoazobenzene without the intermediate production of diazo-amin~benzene.~~ Similarly, E. Rosanhauer and H. Unger 55 findthat the coupling of benzenediazonium chloride with aniline inneutral, feebly basic, or weakly acidic solutions proceeds by twoindependent routes, one leading to diazoaminobenzene and theother to p-aminoazobenzene ; increasing acidity favours the latter,which becomes the exclusive product a t a certain acid concen-tration. All these results lead inevitably t o the conclusion thatthe conversion of diazoaminobenzene into p-aminoazobenzene pro-4 9 J .Amer. Chem. Xoc., 1936, 58, 2413.50 Proc. Roy. SOC., 1934, A , 143, 377; 1935, A , 151, 211; J., 1936, 1154,51 R. P. Bell and 0. M. Lidwell, J . , 1939, 1096.52 J . Org. Chem., 1937, 2, 198.53 Compare J. C. Earl, Chem. and Ind., 1936, 55, 192.54 Compare K. H. Meyer, Ber., 1921, 54, 2265.5 5 Ibid., 1928, 61, 39?.1520; 1939, 1774.204 ORGANIC CHEMISTRY.ceeds through a fission of the molecule giving benzenediazoniumchloride and aniline, followed by p-~oupling,5~ and Kidd has putforward the following scheme :PhN,*NHPh + HCl PhN,Cl + PH < 7, fastpH > 7, very fastpH < 7, very slowPhNH, <-- -+ PhN,*C,H,*NH, 4- HClHe suggests that benzenediazoaminoazobenzene, which has beenpostulated as an intermediate product,57 may be formed by a sidereaction between the aminoazobenzene and benzenediazoniumchloride.Certain of Goldschmidt's earlier results are also dis-cussed in their relation to the newly-proposed mechanism.W. J. Hickinbottom has suggested 58 that the production ofnuclear-alkylated anilines by the Hofmann-Martius method or byheating N-alkylanilines with metallic salts may involve the separ-ation of the alkyl group as a positive ion ; the production of neutralradicals is unlikely, since the known products of decomposition ofthese radicals have not been detected.On this view, olefins andalkyl halides, which are frequently produced, and which havebeen suggested as intermediates in the rearrangements, are theresult of side reactions. Thus, an olefin would arise from thestabilisation of the positive alkyl group by loss of a proton assuggested by Whitrn~re.~ I n the case of certain branched groupssuch as isobutyl, tert.-butyl and isoamyl, however, the formationof olefin appears to become of greater importance, and it is sug-gested that the rearrangement may here proceed largely throughthe action of the olefin upon the nucleus,59 since aminoalkyl-benzenes are shown to be formed in this way.60 The Hofmann-Martius rearrangement now becomesprr > I .I(% \low+ R.c,T-&H~+ l t ' Y I C,H,-NH,R -+ C,H,*NH,C,H,*NH;+ + olefinThe necessity for postulating two routes is shown by the factthat, whereas toluidines are formed in considerable quantities by56 Compare C. K. Ingold, E. W. Smit.h, and C. C. N. Vass, J., 1927, 1245.5 7 J. C. Earl, Ber., 1930, 63, 1666.5 8 J., 1934, 1700.59 W. J. Hickinbottom and S. E. A. Ryder, .7., 19.31, 1281 ; IT. J. Hickin-bottom, J., 1932, 2396 ; 1937, 404.60 Idem, J., 1935, 1279WATSON : REACTION MZCHANISMS. 205heating methylaniline hydrobromide or hydriodide a t 300" in openvessels,5* very little if any p-amino-tert.-butylbenzene is obtainedby heating tert.-butylaniline hydrochloride a t 212" or by heatingthe base with cobalt chloride or bromide under such conditionsthai; volatile products can escape; high yields of butylene areformed, h0wever.5~The positive-alliyl-group mechanism provides a simple explan-ation of the presence of nuclear-alkylated alkylanilines in theproducts of some rearrangements (migration to a foreign nucleus)and of the formation of 2 : 4-dibenzylaniline when benzylaniline isheated with cobalt chloride.61 Branched alkyl groups frequentlyundergo isomerisation, and this change probably occurs in theion.Sometimes isomerisation of the alkyl group occurs when thehydrobromide is heated, but not when the amine is heated with ametallic salt. Hickinbottom suggests that in the latter case thesalt may stabilise the ion or accelerate its reaction with thenucleus.In the rearrangements of allyl (and subst,ituted allyl) aryl ethersto give nuclear-slkylated phenols, shown by L.Claiseii 62 to occuron heating a t about 200°, the allyl group takes up the o-positionpreferentially ; thus, no detectable quantity of p-allylphenol isformed when phenyl allyl ether is heated.63 The group becomeslinked to the nucleus by the terminal (7) carbon atom,64 and thechange is kinetically uniinolecular (this has recently been demon-strated for p-tolyl allyl ether in diphenyl ether solution).65 Vinylallyl ether, containing the essential part of the allyl aryl etherskeleton, undergoes a similar rearrangement to yield allylacet-aldehyde.G6 Definite proof of the intramolecular character of themigration of a substituted allyl group from oxygen t o the o-carbonatom has been provided by C.D. Hurd and L. Schmerling.67When a mixture oE phenyl cinnaniyl ether and p-naphthyl allyl etherwas heated, each compound rearranged independently ; the pro-duct contained no o-allylphenol, for example. Similar results wereobtained when a mixture of P-naphthyl allyl ether and 2-hexenylphenyl ether was heated. There can be little doubt, therefore,61 Iclena, J., 1937, I 1 19.63 Ber., 1912, 45, 3157.63 W. M. Lauer and R. M. Leekley, J . Amer. Chefre. SOC., 1939, 61,30-13.6 i L. Claisen and E. Tietze, Ber., 1925, 58, 275; C. D. Hurd and F. L.Cohen, J. Amcr. Chem. SOC., 1931, 53, 1917.G 5 J. F. Kincaici and D.S. Tarbell, ibid., 1939, 61, 3085.6 6 C. D. Hurd and M. A. Pollack, ibid., 193X, 60, IOO.?; J . Or!/. Chem.,67 J . Anier. Clietts. Soc., 1937, 59, 1 O i .1938, 3, 550206 ORGANIC CHEMISTRY.that the thermal rearrangements of aryl allyl ethers to o-ally1phenols are to be represented by a scheme such as the following :Such a mechanism would be possible only in a system which permitsthe electromeric changes indicated by the curved arrows, and it isin systems of this type that ortho-migration has been observed mostfrequently on the application of heat alone. In the rearrangementof phenyl y-ethylallyl ether (I), o- ay-dimethylallylphenol (111) isproduced in addition to o-y-ethylallylphenol (I1 ; the anticipatedproduct, with the y-carbon atom of the allyl group linked to thenucleus), and a similar " abnormal '' product has been found inother rearrangements. 68(1.) (11.) (111.)If the o-positions are already occupied, the migrating allyl grouptakes up the p - p o ~ i t i o n , ~ ~ and there is evidence that the carbonatom which now becomes linked t o the nucleus may be that whichwas originally attached t o oxygenY70 e.g., (IV) to (V).The mechanism of this p-migration is clearly different from thatof the ortho-rearrangement. If the p-position and both o-positionsare occupied (as in 2 : 4 : 6-trialkylphenyl allyl ethers), olefin iseliminated on heating, and, indeed, products which could arise6 8 W.M. Lauer et al., J. Amer. Chem. Soc., 1936,58, 1388; 1939,61,3039,3043, 3047; C.D. Hurd and M. A. Pollack, J . Org. Chem., 1938, 3, 550.69 L. Claisen and C. Eisleb, Annalen, 1913, 401, 21.70 0. M u m and F. Mailer, Ber., 1937, 70, 2214. Compare 0. Muminet al., ibid., 1939, 72, 100WATSON : REACTION MECHANISMS. 207only by an initial scission of the molecule have been found in somecases where the normal reawangement occurs; 71 e.g., during therearrangement of (VI) to (VII) some (VIII) is formed.It is quite reasonable, therefore, to suppose that, when thep-alkylated phenol is formed by rearrangement of the ether, thechange is intermolecular, the mobile group being removed com-pletely from the molecule and replaced at the p-position. Theortho-rearrangement of allyl aryl ethers must be regarded, however,as essentially intramolecular, with the possibility of an inter-molecular process as a subsidiary change.In all the rearrangements of ethers to which reference has thusfar been made, the migrating group is allyl or substituted allyl,and the electromeric changes depicted in the suggested scheme(above) can occur.Several cases are on record, however, in whicha tertiary alkyl or a benzyl group migrates from oxygen to thenucleus under the action of heat. The new linkage is now formedby the carbon atom which was originally bound to oxygen, and itmay take up either the o- or the p-position; e.g., phenyl tert.-butylether gives p-tert.-b~tylphenol,~~* 73 benzyl phenyl, benzyl o-tolyl,and benzyl guaiacyl ethers on heating at about 250" yield thecorresponding p-benzyl and dibenzyl derivatives in addition tophenol, o-cresol and guaiacol respectively, and benzyl a- and p-naphthyl ethers give respectively 4-benzyl-a-naphthol and l-benzyl-p-naphthol together with the naphthols fhem~elves.7~ The migra-tion of the benzyl group to a foreign nucleus has been observed incertain cases ; e.g., a mixture of benzyl phenyl ether and p-naphtholyields l-ben~yl-~-naphthol,~~ and when a quinoline solution ofbenzyl phenyl ether is heated at 250°, the products are o- andp-hydroxydiphenylmethanes, benzylquinolines, phenoxyquinolines,phenol and toluene ; by controlling the experimental conditions,the yields of benzylquinolines and phenoxyquinolines may beincreased considerably at the expense of the normal products(hydroxydiphenylmethanes) .7571 C. D. Hurd and W. A. Yarnall, J. A ~ n e r . C h e m Soc., 1937, 59, 1686.74 R. A. Smit,h, ibid., 1933, 55, 3718.7 3 S. Natelson, ibid., 1934, 56, 1583.$4 0. Behagel and H. Freiensehnor, Ber., 1934, 61, 1368.7 5 W. J. Hickinbottom, Nature, 1938, 142, 830; 1939, 143, 520208 ORGANIC CHEMISTRY.There are also numerous instances in which phenyl alkyl ethersrearrange to alkylated phenols, not under the influence of heatalone, but in presence of a catalyst. The groups which migrateunder these conditions include isopropyl, isobutyl and see.-butyl,but not n-alkyl groups; isomerisation of the group may occurduring the rearrangement. The migrations of tert.-butyl and benzylhave also been studied in presence of a catalystl.The catalystswhich promote these changes are hydrogen chloride,76, 77 sulphuricacid in glacial acetic zinc chloride,77 zinc chloride andhydrogen chloride,79 zinc chloride and acetic acid,80 boron fluoride 81and aluminium chloride.s2 These may all be classed as electron-acceptors, and it is quite probable that the ether is activated byoxonium compound formation; 77p 78 an addition compound ofboron fluoride with a phenolic ether is actually known. As inthe migrations of benzyl and tert.-alkyl groups under the action ofheat alone, the atom severed from oxygen becomes attached tothe nucleus, and the group may enter either the ortho- or the para-position. Moreover, in a number of these cstalysed rearrange-ments the migration of the group to a foreign nucleus has beenrealised.Thus, when a mixture of 2 : 4 : 6-triisopropylphenylisopropyl ether with phenol was treated with boron fluoride, allthe possible o- and p-alkylated phenols and ethers were isolated,and on passing boron fluoride into phenyl isopropyl ether con-siderable quantities of the mono-, di-, and tri-alkylated phenolsand ethers, in addition to phenol, were found in the product ; 81 thetolyl isopropyl ethers gave similar results. W. F. Short andM. L. Stewart 79 have observed that benzyl phenyl ether in presenceof zinc chloride and hydrochloric acid gives up more than half itsbenzyl to an anisole nucleus; and R. A. Smiths3 has shown thatphenyl and p-tolyl isopropyl ethers on treatment with aluminiumchloride in presence of diphenyl ether give alkylated diphenylethers, while p-tolyl isopropyl and phenyl isobutyl ethers in presenceof benzene and aluminium chloride give isopropyl- and tert.-butyl-benzenes respectively as main product (the alkyl group isomerisesin the last case).7 6 N. I. Kursanov, J . Rms. Phys. Chem. Soc., 1914, 46, 815.7 7 J. van Alphen, Rec. Trav. chim., 1927, 46, 799.J. B. Niederl and S. Natelson, J . Amer. Chena. Xoc., 1931, 53, 1928;J. B. Niederl and E. A. Storch, ibid., 1933, 55, 284.79 W. F. Short, J., 1928, 528; W. F. Short and M. L. Stewart, J., 1929,553.8o J . Ainer. Chewi. SOC., 1934, 56, 1715.81 F. J. Sowa, H. D. HiIlton, and J. ,4. Nieuu71aiic1, ibid., 1932, 54, 201'3;82 R. A. Smith, ibid., 1933, 55, 849, 3718. '1933, 55, 3402.83 Ibid., 1934, 56, 7 WATSON : REACTION MECHANISMS.20'3M. M. Sprung and E. S. Wallisso find that some optical activityis retained in the products of rearrangement of certain aryl sec.-butyl ethers, although extensive racemisation occurs ; this con-trasts with the almost complete retention of optical activity in theCurtius, Hofmann, and Lossen degradations,7? l1 and can hardlybe regarded as evidence of an intramolecular mechanism. Thebalance of evidence, indeed, leaves no doubt as to the intermolecularcharacter of the changes considered above. It is less easy todraw conclusions regarding the form in which the group migrates.The presence of considerable quantities of olefin as by-product inmany cases might perhaps be considered to lend probability to theformation of an alkyl cation which could stabilise itself by loss ofa proton ; the absence of dibenzyl in the products of rearrangementof benzyl phenyl ether has been used as an argument againstneutral radical formation, 79 but union of phenyl radicals in solutionwas not observed by D.H. Hey and W. A. Waters.s4 W. J. Hick-inbottom finds, ix~oreover,~~ that on heating benzyl phenyl ether inquinoline solution the benzyl and phenoxy-groups attack the sameposition of the quinoline nucleus ; this renders it unlikely that theyreact as oppositely-charged ions, but is consistent with the viewthat the ether dissociates into neutral radicals.The catalysed rearrangements of phenolic ethers appear to becomparable with the formation of liydroxy-aromatic ketones fromphenolic esters in presence of aluminium chloride, zinc chloride orferric chloride (Fries reaction).Migration of the acyl group to aforeign nucleus has been demonstrated here also by S. Skraup andK. Poller s5 and by K. W. Rosenmund and W. Schnurr; 86 E. H.Cox finds, too, that, if the reaction is carried out in diphenyl etheras solvent, the acyl group migrates to the ether, while the presenceof alcohol leads t o the product'ion of ethyl ester.87The evidence appears to lead t o the following general conclusions :(1) The rearrangements of aryl ally1 ethers under the influence ofheat to give o-allylphenols are iiitramolecular ; the necessaryelectromeric changes are here possible. (2) p-Rearrangements ofthe same ethers are intermolecular. (3) Benzyl and tert.-alkyl arylethers rearrange under the action of heat by an intermolecularprocess, the group taking up either the o- or the p-position. Thereis evidence that the group migrates as a neutral radical.(4) Therearrangements of sec.- and tert.-alkyl and of benzyl aryl ethers inpresence of catalysts (the group migrating to the o- or the p-position)R4 Chem. Reviews, 1937, 21, 19.1.8 B B e y . , 1994, 57, 2033.R i J . ArrLer. Chenz. SOC., 1930,52, 352. This paper contains a useful summaryof suggested mechanisms.X G Annalen, 1928, 460, 56210 ORGANIC CHEMISTRY.are intermolecular. It appears that the group must be sufficientlyelectron-repulsive, since n-alkyl groups do not migrate ; this mayalso be a factor in (3).The catalyst serves to weaken the bondlinking the mobile group to oxygen; the withdrawal of electronsfrom this group by the linking of an acceptor to the molecule is afeature of many intermolecular rearrangements.56The transformations of alkyl salicylates to alkylated salicylicacids by boron fluoride B8 are also intermolecular, since salicylicacid and dialkylated acids are found in the products, and migrationto a foreign nucleus has been observed. The alkyl group frequentlyisomerises (e.g., n-propyl to isopropyl, n-butyl to sec.-butyl, isobutylto tert.-butyl).A summary of our knowledge of the migrations of alkyl groupsor halogens from one position in the nucleus to another in presenceof aluminium chloride or concentrated sulphuric acid (Jacobsenreaction) has a~peared.8~(b) Condensations of Carbonyl Compounds.A recent investigation by D.S. Breslow and C. R. Hauser 1 hasled them to the conclusion that the Perkin synthesis of cinnamicacids involves the condensation of the aldehyde with the anhydride,the salt acting catalytically. This is in harmony with Perkin'soriginal view, which has been supported by Michael and others,but contrary to the interpretation of the reaction first put forwardby Fittig and assumed in most recent writings on the subject.Breslow and Hauser obtained the same relative quantities ofcinnamic and a-ethylcinnamic acids when benzaldehyde was treatedwith a mixture of either acetic anhydride and sodium butyrateor butyric anhydride and sodium acetate which had in each casebeen heated for several hours in order to establish equilibrium.At100" the equilibrium mixture contains far more butyric anhydridethan acetic anhydride, and the yield of ethylcinnamic acid is fourtimes greater than that of cinnamic acid; at 180" there is a largerproportion of acetic anhydride in the mixture, and a correspondingrelatively higher yield of cinnamic acid results. The effect oftemperature upon the proportions of the two acids obtained frombenzaldehyde, acetic anhydride, and sodium butyrate was, of course,observed in the original work of Fittig, who, however, interpretedit wrongly. Breslow and Hauser have shown further that theW. J. Croxall, F. J. Sowa, and J. A. Nieuwland, J . Org. Chem., 1937, 2,253.See also L.I.Smith and M. A. Kiess, J . Amer. Chem. SOC., 1939, 61, 989.89 C. L. Moyle and L. I. Smith, J . Org. Chem., 1937, 2, 112.* J . Amer. Chem. SOC., 1939, 61, 786WATSON REACTION MECHANISMS. 21 1condensation of benzaldehyde with sodium malonate (which cannotgive an anhydride by reaction with acetic anhydride), reported byStuart in 1883, does not occur. These observations provide sub-stantial evidence that it is the anhydride and not the salt whichundergoes condensation with the aldehyde. Certain results ofother workers lead to the same conclusion. P. Kalnin has shownthat benzaldehyde condenses readily with acetic anhydride inpresence of inorganic and organic bases (e.g., potassium carbonate,triethylamine), whereas it does not condense with sodium acetatein presence of the same substances; the significance of this observ-ation may have been overlooked on account of the very improb-able mechanism which Kalnin put forward, and which has beendi~proved.~ Moreover, benxaldehytie does not condense withsodium acetate in presence of inorganic dehydrating agents.4 Afurther related observation is that of R.Kuhn and S. I~hikawa,~who find that a-vinylcinnamic acid is formed from benzaldehydeand crotonic anhydride in presence of tertiary bases but not inpresence of potassium crotonate.In recent discussions of the mechanisms of the Perkin, Knoeve-nagel and aldol condensations,6, the close relationship of theprocesses with one another and with the Claisen acetoacetic estercondensation has been recognised.Reactions of the Perkin andKnoevenagel types may be represented by appropriate modificationsof the general schemeR*CHO + CH,R’It” + RCIXCR’R” + H,O(aldehyde (methylenecomponent) component)where at least one of the groups R’, R” is of such a character thatthe methylene group is activated (Le., the protons are incipientlyionised). As early as 1904, A. C. 0. Hann and A. Lapworth 8suggested that, in reactions of the Knoevenagel type, the methylenecomponent becomes an anion, which then forms an aldol additionproduct with the aldehyde. This was an alternative to Knoeve-nagel’s original view that the aldehyde and the base react initially;it accounts for the catalytic influence of tertiary bases, and forthe favourable effect of an increase in the strength of the base.Helv.Chim. Acta, 1938, 11, 977.E. Miiller, Annalen, 1931, 491, 251 ; C. D. Hurd and C. L. Thomas, J .Amer. Chenz. Xoc., 1933, 55, 278.* (Signa) M. Balmnin and D. Peccerillo, CTazxctta, 1935, 65, 1145.6 F. Arndt and R. Eistert, Ber., 1936, 69, 2381.Ber., 1931, 64, 2347.C. R. Hauser slid D. S. Breslow, J . At,cer. Chew. Xoc., 1939, 61, 793.t J . , 1905, 85, 46212 ORGANIC CHEMISTRY.The aldol addition product a’ppears as an intermediate in practicallyall subsequent 10 and in some cases has been i~olated.~A similar view has been held with regard to the Perkin reaction.It was based originally upon Fittig’s reported preparation of an“ aldol ” by the interaction of benzaldehyde, isobutyric anhydride,and sodium isobutyrate (where subsequent elimination of water isnot possible), but E.Muller and his co-workers failed to repeatFittig’s result.lf C. R. Hauser and D. S. Breslow have nowprepared this compound, using sodium triphenylmethyl as con-densing agent; they have also obtained “ aldols ” from benz-aldehyde and ethyl isobutyrate [CHPh( OH)-CMe,*CO,Et] andfrom benzaldehyde and ethyl acetate [CHPh( OH)*CH,*CO,Et],using the same agent. The “aldol ” view of the Perkin andKnoevenagel reactions may thus be regarded as established ex-perimentally.The mechanisms recently put forward by Arndt and Eistertand by Hauser and Bre~low,~ which incorporate the essentials ofthat suggested by Hann and Lapworth, may be written as follows(B = basic catalyst) :R + CH,R’R” 2 BJ? + 8HR’R’’/H H(1.) RW’ + EHR’R” (11.1 2 RC-CHR’R” (111.1\O \O” *H@ /H - H,O --+ ROC-CHR’R” (1v.1 -+ R*CH:CR’R‘’ (17.1 <- ‘OHArndt and Eistert, however, omit the addition of hydrion to formthe true aldol (IV); they also suggest that, as an alternative t oforming the unsaturated compound R*CH:CR’R‘’, the ion (111)may react with another molecule of the methylene component asfollows :CHR’R”R-C-CHR’R” /H + CH,R’R” -+ RCH/ + OH’\O” \CHR/RIIIt may be pointed out, however, that, if the unsaturated compoundis first formed, the addition of a second molecule of the methylenecomponent is an ordinary Michael reaction.See, e.g., E. P.Kohler and B. B. Corson, J . Amer. Chem. Soc., 1923, 45,1975.lo See, e.g., A.C. Cope, ibid., 1937, 59, 2327.l1 Annalen, 1935, 515, 97WATSON : REACTION MECHANISMS. 213For condensations occurring under the catalytic influence ofacids, Arndt and Eistert write a mechanism of the following type :R.C/H -???!+ k*C<:J C1 A+ R;C*CHR’R” + HC1 CEI R’R” /H\O ‘OHIn the Claisen acetoacetic ester synthesis,ROC/,,OEt + CHR’R”*COX _I, R*COCR’R’‘*COX + HOEt(second\‘O(estercomponent) component)the ester which replaces the aldehyde of the Perkin and IZnoevenagelcondensations will take part in the addition process less readily,owing to the resonance of the carbethoxyl group, but it possessesan anionic group which is split off more easily than is the carbonyloxygen itself.6 It follows that the aldol phase is less likely to beisolated here than in the Perkin and Ihoevenagel syntheses.Arndtand Eistert suggest that the necessary driving factor for the con-densation is the tendency for the formation of a conjugated systemwith the additional stability gained from mesomerism. Thepresence of two cc-hydrogen atoms in the second component is notnecessary here. Thus, although ethyl isobutyrate does not con-dense with a second molecule of itself under the influence of sodiumethoxide, condensations where the “ second component ” has onlyone cc-hydrogen have been effected by this agent in certain caseswhere both reacting groups are in the same molecule.12 Moreover,it has recently been shown l3 that sodium triphenylmethyl, theaction of which is more powerful than that of sodium ethoxide(since it is a stronger base) but does not appear to differ from it inprinciple, brings about the condensation Jf ethyl isobutyrate withitself (to give CHMe,*CO*CMe,*CO,Et) or with ethyl benzoate(giving Ph*CO*CMe,*CO,Et). Ethyl isobutyrate can also be con-densed by means of mesitylmagnesium bromide, which also inducesreaction in certain cases (e.g., CMe,*CH,CO,Et) where two a-hydro-gen atoms are present but where sodium ethoxide is neverthelessineffective. l4Some of the earlier mechanisms for the acetoacetic ester synthesis(Claisen, Dieckmann) postulate reaction of the second component3934, 58, 1173.Compare C.R. Hauser, ibid., 1938, 60, 1967.13 R. I?. B. Cox, E. H. Kroeker, and S. M. McElvain, J .Anaer. Chem. SOC.,13 C. R. Hauser and W. B. Renfrow, ibid., 1937, 59, 1923; 1938. 60, 483.14 M. A. Spiclman and M. T. Schmidt, ibid., 1937, 59, 2009214 ORGANIC CHEMISTRY.with an addition complex of ester and ethoxide, to form thesodium derivative of the p-diketone or ketonic ester. A schemein which the ion of Ohe second component adds to the ester wasput forward by A. La~w0rth.l~ H. Scheibler has more recentlysuggested a mechanism l6 involving the formation of certain inter-mediates which, as Arndt and Eistert point out, have no experi-mental foundation.6 C. R. Hauser and W. B. Renfrow 13 havenow adopted Lapworth’s “ ionic ” conception in a scheme whichresembles their representation of the Perkin and Knoevenagelcondensations (see above) :OEt /OEtRO’ + ~R~RWOX-> +- R*C-CR’R”*COX --+N O \oe R*CO*CR’R”*COX + OE?The product obtained when R” = H, vix., the sodium derivative ofthe diketone or ketonic ester, is regarded as being formed by afurther reaction of the ketonic form with the base.In anothermechanism, due to Arndt and Eistert,6 the enol ion is formed byelimination of alcohol from the addition complex ; this, however,makes necessary the presence of two or-hydrogen atoms in thesecond component (R” = H), which is not the case. If the reactingentity is the anion CR!R”*COX, the postulate of enolisation of thesecond component is redundant, since the ion is a mesomeric-structure C-C-O:There is no proof that the only function of the base (or even itsmain function) is the conversion of the “ second component ” intoanion. Undoubtedly such ionisation can and does occur; e.g.,J.Kenyon and D. P. Young have recently observed the almostcomplete (though slow) racemisation of certain optically activeesters of formula CHR’R’’*CO2Et in contact with sodium ethoxide,l’which would indicate an equilibrium between ester, base, andmesomeric anion. Nevertheless, the most important pa,rt playedby the base in the Knoevenagel, Perkin, and Claisen condensationsmay be the activation of the carbonyl group of the aldehyde orester by the formation of an addition product which reacts rapidly(perhaps instantaneously) with the “ second component ” (or itsion). Such a mechanism might account for the unimolecularkinetics of the aldol condensation of acetaldehyde observed byR.P. Bell,ls who interprets this feature by supposing’that thel5 J., 1901, 79, 1269. Compare J. U. Nef, Annden, 1897, 298, 218;A. Michael, Ber., 1900, 33, 3731.l6 See Ann. Reports, 1934, 31, 200.l 7 J., 1940, 216. J., 1937, 1637WATSON : REACTION MECHANISMS. 215dehydration of a hydrated aldehyde molecule a t measurable speed(catalysed by the base) is followed by a rapid condensation (alsobase-catalysed) with a second hydrated molecule. The rapidreaction might, however, be that between the complex and thesecond aldehyde molecule. If the base adds in this way to thealdehyde or ester, the Knoevenagel and Claisen condensationswould be represented in the following way :Knoevenagel.CH,R’R” +/H _____, /HN O \0- 1 0 -HROC/ + B 3 ROC-B -+- ROC-CHR’R’’ + BH+fr R-C-CHR’R’’ /H + B ---f R*CH:CR’R” + H,O + 13\OHClaisen.ROC/ + OEt- 3 R d O E tOEt OEt CHR’R”*COX /OEtR*C-CR’R”*COX + HOEtN O ‘0- \0-4 R*CO*CR‘R”COX + OEt- + HOEt/OEt CR’COX(or, if R” = H, RC-CHRWOX + ROC// + HOEt)\O- \O-(c) The ortho-Eflect.(Continued from Ann.Reports, 1938, 35, 243.)H. 0. Jenkins l9 has calculated values of the field intensity, atthe nuclear carbon atom to which the carboxyl group is linked,due to the C-Hal dipole in the halogenobenzoic acids. The relevantexpression is P = ~ ( 1 + 3 cos2 0)*/r3, where p is the dipole momentof the appropriate halogenobenzene, and r is the distance from thecentre of the dipole, the axis of which is inclined at an angle 8 to r.He finds that, for each series of halogenated acids, the plot of fieldintensities against logarithms of dissociation constants is linear ;the same applies when the values for the nitrobenzoic acids aretreated similarly.Actually the p-nitro- and p-fluoro-acids showsmall divergences from the linear relationship, owing presumablyto mesomeric effects which may not be quantitatively identical inC,H,X and C0,H*C6H4X.20 The conclusion is drawn that theortho-substituted acids show no abnormality, the relatively highvalues of their dissociation constants being due merely to theshort distances from which the inductive effects operate. Suchl9 J., 1039, 640. ao Compare Ann. Reports, 1938, 35, 341216 ORGANIC CHEMISTRY.an interpretation is probably applicable also to o-methoxybenzoicacid, but not to salicylic and o-toluic acids.All three are strongerthan benzoic acid, whereas their p-isomerides are weaker. In thefirst case, however, the inductive effect of OMe (- I ) operatinga t close range might overcome the mesomeric effect which isresponsible for the low value of the dissociation constant of p-anisicacid, but the much higher strength of salicylic acid and the increaseof strength conferred by o-Me (a + I group) make necessary thechelation hypothesis, to which reference was made in last year’sReport.When, instead of the field intensity, the electrostatic potential# = 1.1 cos e p , at the same carbon atom, is plotted against log K ,a linear relationship is found for the o-, m-, and p-substituted acidsof the above series, but the lines do not pass through the pointfor the unsubstituted acid.Again no ortho-effects are found, andthe slopes of the lines agree quite well with the values calculatedfrom the expression NEIRT. A similar treatment of the halogeno-phenols and -anilines 21 gives a linear relationship between the pointsfor the unsubstituted, o- and p-compounds of each series, and againthe slopes approximate to NE/RT, but the m-halogeno-compoundsnow diverge. This divergence is ascribed to the non-operation ofthe mesomeric effects from the m-position, and the geometry ofthe figure indicates a proportionality between the inductive andmesomeric effects (which would, of course, be expected if they hada common origin).In the case of the nitroanilines, the plot of $against log K is linear for unsubstituted, o- and m-compounds, butthe p-derivative diverges as in the acids.The significant result of this work is that the o-halogeno- ando-nitro-benzoic acids, -anilines and -phenols appear to show noeffects which are not found also in the p-isomerides, except thato-nitrophenol has an abnormally low dissociation constant , due nodoubt t o chelation of the phenolic hydrogen with the elcctron-donating oxygen of N02.22A kinetic investigation of the addition of methyl iodide to anumber of o-substituted dimethylanilines in methyl-alcoholicsolution has been reported recently.23 The energies of activationhave throughout higher values than those found for the corre-sponding p-substituted compounds ; the differences between theE values for the u- and p-isomerides (Eo - E,) cover a wide rangeand sometimes exceed 6000 calories.For o-fluorodimethylanilinethe magnitude of the PZ term of the kinetic equation does not21 J., 1939, 1137.22 Compare N. V. Sidgwick and R. K. Callow, J . , 1924, 125, 527.2n D. P. Evans, 13. B. Watson, and R. Williams, J . , 1939, 1348WATSON : REACTION MECHANISMS. 21 7differ perceptibly from that relating t o the same reaction of unsub-stituted dimethylaniline and its m- and psubstituted derivatives(this term is not changed by rn- or p-substitution).24 As in thealkaline hydrolysis of ethyl o‘-flu~robenzoate,~~ therefore, fluorineexhibits no unusual effect when present in the o-position to thereacting group.The remaining six groups examined, however,cause an increase in the value of PZ by a factor which varies from3 to 400. In this reaction, therefore, as in the esterification ofcarboxylic acids and the acid hydrolysis of esters,26 an “ortho-effect ” is indicated by a simultaneous increase in E and in PZ.The magnitude of the effect increases in the order F < OPh <OMe (NO,) < C1 < Me < Ph, a sequence which is quite uncon-nected with the weights, volumes, or chemical characters of thegroups. The effect is ascribed to an interaction, in the transitioncomplex, between the group and the unshared electrons of thenitrogen ; this is actually an electronic expression of J. von Braun’ssuggestion that the free affinity of the nitrogen is to some extentsaturated by the group attached t o the o-carbon atom.27 Thesame conception was discussed in last year’s Report 25 in connectionwith esterification and ester hydrolysis, where the electron-donatingatom is carbonyl oxygen.I n the present case tervalent nitrogenis the donor atom, and it is possible to envisage a series of groupswhich, when suitably placed, would give a steadily increasing‘’ degree of interaction.” The extreme members of this serieswould be fluorine, where interaction is impossible, and carboxyl,which transfers a proton completely to nitrogen, forming an electro-valent bond. Intermediate between these would stand groupscapable of forming a hydrogen bond with the nitrogen, and it issuggested that o-Me and o-Ph come into this category.The pro-cess would render the unshared electrons of the nitrogen less avail-able for reaction with the alkyl halide, thus raising E ; it wouldalso confer a charge upon the reactive portion of the complex, thelife of which would be increased with a resulting rise in the valueof PZ.25 This is in harmony with observation. The remaininggroups studied (Cl, NO,, OMe, OPh) have much smaller effects,and it is postulated that there is here some form of interactionwithout the formation of an actual “ bond.’’ It was suggestedin last year’s Report that some interaction of this nature might24 I(. J. Laidler, J . , 1938, 1786; D. P. Evans, H. B. VC7atson, and R.Williams, J ., 1939, 1346.z 5 See ibid., p. 246.26 C. N. Hinshelwood and A. R. Legard, J., 1935, 587; E. W. Timm a.nd2 7 Rer., 1916, 49, 1101; 3918, 51, 282.Compare Ann. Reports, 1938, 35, 237.C. N. Hinshelwood, J., 1938, 862218 ORGANIC CHEMISTRY.explain the rather low strengths of some o-substituted anilines. Inview of the recent findings of H. 0. Jenkins,21 however, the con-ception of an " ortho-effect '' in o-chloro- and o-nitro-anilines mustbe abandoned; the postulated interaction is manifested only inreactions and not in equilibria. A chelation hypothesis still appearsto be necessary to interpret the basic strength of o-toluidine, how-ever (compare o-toluic acid, above).The order o-Me > o-C1 > o-OMe for the energies of activationin the addition of methyl iodide to dimethylanilines is in harmonywith the relative yields of quaternary iodide obtained by vonBraun2' from the same compounds under fixed conditions.Healso found that the p-positions were rendered less active, the sameorder of effects being maintained. It is now suggested23 that theinteraction of the unshared electrons of nitrogen with the o-sub-stituent reduces the electromeric effect of the dimethylamino-group.A reduction of mesomerism by o-methyl groups has been demon-strated by C. E. Ingham and G. C. Hampson28 by determinationsof the dipole moments of a number of derivatives of durene andmesitylene. The moments of aminodurene and mesidine areidentical within experimental error, but slightly lower (by 0.13 D)than that of aniline; a larger difference (0.55 D) is found betweenthe moments of dimethylmesidine and dimethylaniline.A relatedfact is the nitration of the acetyl derivative of m-2-xylidine atCompare R. H. Birtles and G. C. Hampson, J . , 1937, 10. 28 J., 1939, 981SMITH : THE PEROXIDE EFFECT. 219positions 4 and 6 ; the mesomeric effect (and by implication theelectromeric effect also) of the acetamido-group is so reduced thatthe directive influence of the methyl groups takes control. Nitro-aminodurene also has a lower moment (by 1-12 D) than p-nitro-aniline, and almost the same lowering is observed in 2-nitro-m-5-xylidine, which has methyl groups o- t o the nitro- but not t o thoamino-group. The greatest depression observed is in nitrodi-methylaminodurene, which has a moment lower by 2.76 unitsthan that of p-nitrodimethylaniline. The results are summarisedabove, where the figures in the centre of the formulz denote thedipole moments .The authors ascribe this lowering of the dipole moment to asteric effect of the o-methyl groups which prevents the system fromtaking up a planar configuration and thus reduces the mesomerism.The formation of hydrogen bonds between the methyl groups andthe adjacent electron-doncrs (oxygen of NO,, nitrogen of NH,or NR,) appears to furnish an alternative explanation, however.It is also in harmony with the absence of the effect in durenol(compare absence of abnormal properties in o-cresol, for example) .22The greater effect in the dimethylamino- than in the amino-com-pounds is not interpreted so easily on this basis; it is strange thatthe reverse is found in the basic strengths of o-toluidine and di-methyl-o-toluidine.29 A complete interpretation of the observedphenomena would perhaps include both factors.H. B. W.3. THE PEROXIDE EFFECT.According to a patent applied for by W. Bauer in 1922 acetyleneadds hydrogen bromide at the ordinary temperature in presence ofgaseous oxidising agents (air, oxygen, nitrogen peroxide) to give90% yields of ethylene dibromide ; the second molecule of hydrogenbromide is oriented contrary to the Markownikoff rule.2CH,:CHBr + HBr + CH,Br*CH,Br(Bauer’s claim for a similar effect with other halogen acids has notbeen substantiated.) This publication seems to have been overlookedand additions of hydrogen bromide to olefins continued to giveconfusing and inexplicable results until 1933.Y. Urushibara and29 Compare Ann. Reports, 1938, 35, 248.U.S. Patent, 1,540,748 (1925); British Chem. Abstr., 1925, B, 692.W. B. Markownikoff, Annalen, 1870, 153, 256. (i) “If an unsym-metrical hydrocarbon combines with halogen acid the halogen adds to thecarbon atom with fewer hydrogen atoms, i.e., t o the carbon which is moreunder the influence of the other carbon atoms; (ii) by addition of halogenacid to vinyl chloride or chlorinated propylene, etc., the halogen will alwaysadd to the carbon which is already combined with halogen.220 ORGANIC CHEMISTRY.R. Robinson reported that the orientation of addition to Ale-undecenoic acid depended on whether the reaction vessel was opento the air or not (they provisionallyattributed the effect to moisture)."M.S. Kharasch and F. R. may^,^ carrying out reactions betweenhydrogen bromide and ally1 bromide in sealed tubes, noticed thatwhen air was left in the tubes yields of 85-90% of 1 : 3-dibromo-propane were obtained; if the tubes were cooled and evacuatedbefore sealing, reaction was much slower and the product wasmainly 1 : 2-dibromopropane. The American authors extended thework to vinyl bromide and to propylene,6 finding in each casethat the presence of oxygen and/or peroxides €avowed an orient-ation of addition opposite to that predicted by the Markownikoffrules. The results obtained by Kharasch and Mayo have beenconfirmed in several laboratories and the hypothesis that the" abnormal " or " peroxide catalysed " reaction is due to the presenceof bromine atoms (formed from hydrogen bromide and the " oxidant")is now widely accepted.The subject has already been dealt within two reviews; 7 9 7a in this Report the emphasis will be on thepublications of 1938 and 1939.oxide in benzene to yield the saturated diketone (LXXXII), ofwhich the ll-keto-group could be reduced catalytically to CH,.Demethylation of the product gave x-norequilenin (LXXXIII).(LXXXII.) (LXXXIII.)Recently 37 x-noroestrone has been prepared from (LXXXI,R = OMe) (see p. 293). H. A. Weidlich and G. H. Daniels 59 haveused the ring closure of y-diketones in the preparation of the interest-ing 3-naphthyl-2-methylcyclopentttnone (LXXXVI) (the methodis a development of early work by W.Borsche60). They con-densed ethyl sodio- P-ketovalerate with w-bromo-p-acetylnaphth-alene, obtaining (LXXXIV). This, on decarboxylation and ringclosure, gave the naphthylmethylcyclopentenone (LXXXV) , whichwas hydrogenated to (LXXXVI).(LXXXIV.) (LXXXV. ) (LXXXVI. )C. K. Chuang and his collaborators58 J., 1938, 1994.60 W. Borsche and A. FeIs, Ber., 1906, 39, 1813; W. Borsche and W.Mentz, ibid., 1908, 41, 194.C. K. Chuang,Y. L. Tien, and Y. T. Huang, Ber., 1937, 70, 858; C. K.Chuang, Y . T. Huang, and C. M. Ma, ibid., 1939, 72, 713 ; C. K. Chuang,C. M. Ma, Y . L. Tien, and Y. T. Huang, ibid., p. 949.have described severalSD Ber., 1939, 72, 1590SPRINGALL : SYNTHESIS OF STEROIDS.299syntheses by the Robinson-Schlittler-Walker method.62 Themethyl 5-keto-8-m-methoxyphenyloctoate of (Sir) R. Robinsonand E. Schlittler (LXXXVII) was cyclised with sulphuric acid togive methyl y-(6-methoxy-3 : 4-dihydro-l-naphthy1)butyrate(LXXXVIII) . Hydrolysis, dehydrogenation with sulphur, andesterscation yielded the corresponding naphthalene ester (LXXXIX) .Repetition of the Robinson-Schlittler condensation with ethylsodio-a-acetylglutarate led to 5-kefo-8-(6’-methoxy-l’-naphthyl)-octoic acid (XC). The corresponding keto-ester was cyclised withsodium ethoxide t o the 1 : 3-diketocyclohexane derivative (XCI),which in turn was cyclised in the presence of either phosphoricoxide in benzene or cold sulphuric acid to the methoxyketohexa-hydrochrysene (XCII).(LXXXVII.) (LXXXVIII.) (LXXXIX.)The cyclopentenophenanthrene (XCIII) corresponding to (XCII)was prepared from the acid of (LXXXIX) by the same process,ethyl sodioacetylsuccinate being used. An attempt to introducethe angle methyl group on C,, (of the cholane system) failed. Thea-methyl-y-( 6-methoxy-l-naphthy1)butyryl chloride appeared tocondense with the ethyl sodioacetylsuccinate, but hydrolysis gaveonly the original acids and not the desired (XCIV). (Severale2 (Sir) R. Robinson and E. Schlittler, J., 1935, 1288; (Sir) R. Robinsonand J. Walker, J., 1936, 192; Ann. -Reports, 1936, 33, 334300 ORGANIC CHEMISTRY.similar failures of the Robinson keto-acid ~ynthesis,~~ in caseswhere a methyl group should appear a to the keto-group, have beenr e ~ o r d e d .~ ~ ~ ) This development of the Robinson-Schlittler-Walkermethod had been independently investigated by (Sir) R. Robinsonand J. M. C. T h ~ m p s o n , ~ ~ ~ who also prepared the ketomethoxy-hexahydrochrysene (XCII). G. Haberland 64b built up the 4-keto-5-methyllieptoic acid system from a-methyl-y-( 6-methoxy- 1 -naphthy1)butyryl chloride by the successive action of diazomethane,hydrobromic acid and ethyl sodiomalonate. The resulting ester(XCV) was cyclised with sulphuric acid to give methyl 7-methoxy-.%methyl3 : 4-dihydrophenanthrene-l-propionate (XCVI).The possibility of preparing keto-esters of the type (XCVII) bythe Friedel-Crafts reaction with ethyl y-m-methoxyphenylbutyrate and y-carbo-p10red.~~ By double ring closure suchMe01 ,)!,) compounds should give valuable oestrone\\ (xcvll') intermediates.Difficulties arose, how-ever, due to substitution o to the methoxy-group and to prematurering closures.The ring closure of diketones and keto-acids has been used inthe preparation of hydrophenanthrenes related to morphine 66and to the diterpene~.~'co /\/co2Me C0,E.t methoxybutyryl chloride has been ex-/\/ /(vi) Formation of Cyclic Ketones from Monocarboxylic Acids andtheir Derivatives.The early work on the preparation of cyclic ketones by the lossof water from acids (in the presence of sulphuric acid) and of hydro-gen chloride from acid chlorides (in the presence of aluminium63 (Lady) G.M. Robinson, J., 1930, 745.64 (a) D. A. Peake and (Sir) R. Robinson, J., 1937, 1581; (b) G. Heberland,Ber., 1939, 72, 1215; (c) (Sir) R. Robinson and J. M. C. Thompson, J., 1939,1739.65 (Sir) R. Robinson and J. Walker, J., 1937, 60; K. H. Lin, J. Resuggan,(Sir) R. Robinson, and J. Walker, ibid., p. 68; (Sir) R. Robinson and J.Walker, J., 1938, 183; K. H. Lin and (Sir) R. Robinson, ibid., p. 2005;(Sir) R. Robinson and J. M. C. Thompson, ibid., p. 2009.G 6 L. F. Fieser and H. L. Holmes, J. Arner. Chem. SOC., 1938, 60, 2548.6 7 H. Plimmer, W. F. Short,, and P. Hill, J., 1938, 694SPRINGALL : SYNTHESIS OF STEROIDS. 301chloride or stannic chloride) has been reviewed previously.68 Thering closures may be performed on aromatic rings (following F.S.Kipping or on unsaturated alicyclic rings (following G. Darzens 70).Both methods have been employed in recent steroid work.(a) Ring Closures involving Aromatic Rings.-G. Haberland andE. Blade 71 condensed 6-methoxy-l-~-bromoethyltetralin withethyl sodiomethylmalonate and decarboxylated and dehydro-genated the product, obtaining the acid (XCVIII). Cyclisation ofthis acid with sulphuric acid, followed by demethylation, yieldedthe valuable l-ket,o-7-methoxy-2-methyl-l : 2 : 3 : 4-tetrahydro-phenanthrene (XCIX). This was subjected to the Reformatskyreaction with ethyl p-bromopropionate, giving the lactone (C) ,which was converted into the chloride of the dehydrated acid (CI)and cyclised with stannic chloride, yielding either (CII) or (CIII) 72(dehydrogenation occurring during the cyclisation).0The methyl 7-methoxy-2-methyl-3 : 4-dihydrophenanthrene-l-pro-pionate (XCVI) (p. 300) was also converted into the acid chloride(CI) and gave (CII) or (CIII) on treatment with stannic chloride.64bWhen the cyclisation was performed a t - Z O O , dehydrogenationwas avoided and the 3 : 4-dihydrophenanthrene compound corre-sponding to (CII) or (CIII) was obtained.KetocycZopentenophenanthrenes with the five-membered ringfused in various positions have been prepared from p-1-, p-2-,Ann.Reports, 1936, 33, 336.C m p t . rend., 1910, 150, 707.69 J., 1894, 65, 480.'il Ber., 1937, 70, 169.'i2 G. Haberland and E. Heinrich, Ber., 1939, 72, 1222302 ORGANfC CHEMfSTRY.p-3- and p- 10-phenanthrylpropionyl chlorides.73 The first twooyclisations bear on the steroid problem. The 1 -substitutedphenanthrene gave only 4% of the 1 : 2-cyclisation product (CIV),and 25% of the 1 : 10-product (CV). The 2-substituted phen-anthrene gave exclusively the 1 : 2-product (CVI).The 9 : 10-dihydro-derivative of P-2-phenanthrylpropionic acid,however, gave both 1 : 2- and 2 : 3-cyclisation products.74 Thering closure of the 1 : 2 : 3 : 4-tetrahydro-derivative of p-l-phen-anthrylpropionyl chloride has been studied by J. Hoch : 75 thecompound ring closes to the 10-position of the phenanthrenenucleus. From y-l-phenanthrylbutyric acid, the ketotetrahydro-chrysene was obtained. A synthesis along these lines of a nor-equilenin methyl ether has been described in a German patent.767-Methoxy- 1 : 2 : 3 : 4-tetrahydrophenanthrene-1 -propionic acid(CVII) was cyclised, and the product described as a norequileninmethyl ether (CVIII).It would appear more probable, however,that ring closure would occur to the lo-, rather than the 2-position,giving (CIX), especially as in this case the C, atom is part of asaturated alicyclic system (such saturated systems are known not tocondense readily, if at all, with carboxylic acid derivatives77).0(CVII.)The method was also employed in the preparation of (CXI),'878 W. E. Brtchmann and M. C. Kloetzel, J. dmer. Chern. SOC., 1937, 59,74 A. Burger and E. Mosettig, ibid., p. 1303.713 Bull. Soc. dim., 1938, [v], 5, 264; Compt. rend., 1938, 207, 921.76 Chem.Abstracts, 1938, 32, 4176.7 7 A. E. Bradfield, E. R. Jones, and J. L. Sirnonsen, J., 1934, 1810; J. W.Barrett, A. H. Cook, and R. P. Linstead, J., 1936, 1067; J. W. Cook mid C. A.Lawrence, ibid., p. 1637.2207.7 8 (Sir) R. Robinson and J. M. C. Thompson, J., 1938, 2009SPRINGALL : SYNTHESIS OF STEROIDS. 303a model for (CXII) desired as an oestrone intermediate. TheMichael addition of ethyl cyanoacetate to ethyl Ab-dihydromuconate,C02Et/ CO,Etand condensation of the product with p-phenylethyl bromide, gave(CX), which, as the free acid, cyclised with sulphuric acid to (CXI).In the course of syntheses in the diterpene series the formation ofa, seven-membered ring by this cyclisation method has been ob-served.7g The acid (CXIII) gave (CXIV) as well as the expectedproduct (CXV).,rl CH2TH2 70 p 2\/\//\/\ A/\/\() fl /v\ co2I-p. I II I I II I I II I\A/ \/\/OMe OMe(CXIII.) (CXIV.) (CXV.)The preparation of keto- and hydroxy-chrysenes starting frombenzalacetophenone, two such ring closures being used, has beendescribed.sO The introduction of two new reagents for these cyclicketone syntheses, namely, acetic anhydride, acetic acid and zincchloride,s1 and liquid hydrogen fluoride,82 has been announced.(b) Ring Closures involving Unsaturated Alicyclic Ring8.-Developing earlier M T O ~ ~ , ~ ~ J. W. Cook and C. A. Lawrcnce 84 haveOMe79 G. A. R. Kon and F. C. J. Ruzicka, J . , 1936, 187; P. Hill, W. F. Short,and H. Stromberg, J., 1937, 1619; J. Lockett and W. F.Short, J., 1939,787; G. A. R. Kon and H. R. Soper, ibid., p. 790; see also L. F. Fieser andM. A. Peters, J . Amer. Chem. Soc., 1932, 54, 4347; L. F. Fieser and M. Fieser,ibid., 1933, 55, 3342.8o M. S. Newman, J . Amer. Chem. SOC., 1938, 60, 2947.L. F. Fieser and E. B. Hershberg, ibid., 1937, 59, 1028.82 L. F. Fieser and E. B. Hershberg, ibid., 1939, 61, 1272; W. S. Calcott,83 J. W. Cook and C. A. Lawrence, J., 1935, 1637.J- M. Tinker, and V. Weinmap, ibid., p. 949.84 J., 1937, 817304 ORGANIC CHEMISTRY.prepared the methyloctalone (CXVII) [and, by reduction, themethyldecalone (CXVIII)] from 1 -methyl-A1-cycZohexene-2-y-butyrylchloride (CXVI). Similar preparation of (CXVIII) had beenreported previously 85 and was again described subsequently.Inthe hope of obtaining the methoxymethyldecalone (CXX), 4-methoxy-l-methyl-A1-cycZohexene-2-y-butyric acid (CXIX) was prepared.86Cyclisation was accompanied by loss of methyl alcohol, however,and only the ketohexahydronaphfhalene (CXXI) was isolated.CO,H 0 0 I \ II ...c/v Me0 p,J) PI", \/\/(CXXI. )\/\/(CXIX.) (CXX.)(Sir) R. Robinson and J. WalkerJ8' extending their original workon this cyclisationJs8 prepared y-6-methoxy-3 : 4-dihydronaphth-alene-l-butyryl chloride (CXXII) from methyl 5-keto-8-m-methoxy-phenyloctoate. It was hoped that the action of aluminium chlorideon this might cause simultaneous ring closure and reduction togive the already known, but inaccessible 62 (CXXIV). The reduc-tion did not occur, however, and the known unsaturated ketone(CXXIII) 62 was produced.The cyclisation of monocarboxylic acids has been used in thediterpene g9 and morphine fields.85 C.K. Chuang, Y. L. Tien, and C. M. Ma, Ber., 1936, 69, 1494.a7 J . , 1937, 60. J., 1936, 192.8Q D. E. Adelson and M. T. Bogert, J . Amer. Chem. SOC., 1937, 59, 399;P. Hill, W. F. Short, and H. Stromberg, J., 1937, 937; H. Plimmer, F. W.Short, and P. Hill, J., 1938, 694; G. A. R. Kon, E. S. Narracott, and C.Reid, ibid., p. 778.90 G. Haberland and G. Kleinert, Ber., 1938, 71, 470; G. Haberland andH. J. Siegert, ibid., p. 2619; G. Haberland, G. Kleinert, and H. J. Siegert,ibid., p. 2623.J. W. Cook and C. A. Lawrence, J., 1938, 58SPRINGALL : SYNTHESIS OF STEROIDS. 305(vii) Formation of Cyclic Ketones from Dicarboxylic Acids.Much work has been done on hydrophenanthrene derivativeshaving carboxylic acid residues on C, and C, suitable for pyrolyticor Dieckmann cyclisation to ketoc yclopentenohydrophenanthrenes.The preparation of the ketomethoxyoctahydrophenanthrene(CXXIV) from the hexahydro-ketone (CXXIII) 62 via the saturatedalcohol (CXXV) has been improved.The oxidation (CXXV) +(CXXIV) was originally performed with chromic oxide. (Sir)R. Robinson and J. Walker found cupric oxide to be betterJgl andthe Oppenauer method best of all.92,\/CO*C02EtP V H I 1 (CXXVI.)/ \ / \ A 0(CXXV. )/ \ / \ / \ O ~\/\/ Me01 11 ,) \/\ Me01 11 IIn the latter paper, improved conditions are given for the con-version of (CXXIV) through the glyoxylic acid derivative (CXXVI)into the 2-carbethoxy-2-methyl derivative (CXXVII) first preparedin 1936.93 The synthesis of oestrone from (CXXVII) has beeninvestigated.In the course of model experiments on 2-carbo-methoxy-2-methylcyclohexanone (CXXVIII) 94 [from which 1-carbomethoxy- 1 -methylcycEohexane-2-acetic acid (CXXIX) had(CXXVII.)I /\//co2Et/\/\/I\()Me01 II I \/v (CXXVIII.) (CXXIX. )already been obtained '1 (CXXX) was finally prepared by theaction of y-methoxypropylmagnesium chloride, and, thence, theacid (CXXXI). Pyrolysis of the barium salt of this acid gave8-methylhydrindan-l-one (CXXXII).(CXXX.) (CXXXI. ) (CXXXII.)G. A. R. Kon, R. P. Linstead, and C. Simons 95 had independentlyThe applic- prepared (CXXXII) by t'he same series of reactions.B1 J ., 1937, 60.93 (Sir) R. Robinson and 6. Walker, J . , 1936, 747.m4 Idem, J., 1937, 1160.B2 J., 1938, 183.95 Ibid., p. 814306 ORGANIC CHEMISTRY.atian of such reactions to oestrone (CXXXIII) was studied by F.Litvan and (Sir) R. Robinson,96 who oxidised the methyl ether ofnatural oestrone via the isonitroso-compound (CXXXIV) to themethyl ether of the oestric acid (CXXXV) first obtained by G. F.Marrian and G. A. D. Haslew~od.~~ The methyl half-ester wassubjected to Amdt-Eistert 98 chain-lengthening to give the half-ester of (CXXXVI). The lead salt of (CXXXVI) on pyrolysis anddemethylation gave again oestrone.0 0(CXXXIII.) /\ /\ I1 / \ , A 7/\A+ (CXXXIV.)I l lA/\/--Me01 11 I \/v(Sir) R. Robinson and H. N. R y d ~ n , ~ ? having obtained thecompound (CXXXVII) by the new Robinson synthesis (p.293) andfound direct hydrogenation unprofitable, opened the five-memberedring by a, modification of the Litvan-Robinson procedure via theisoformyl derivative (CXXXVIII) and the nitrile (CXXXIX),obtaining the acid (CXL). Hydrogenation of the methyl ester of(CXL) with the Adams catalyst gave some of the compound (CXLI).Pyrolysis of the lead salt of the corresponding acid and demethyl-ation yielded x-noroestrone (CXLII), which probably has the cis-cis- configuration.0~XXXVII.) (CXXXVIII. ) (CXXXIX. )O6 J., 1938, 1997. J. SOC. Chem. Ind., 1932, 51, 277.F. Arndt snd B. Eistert, Ber., 1935, 68, 200SPRINGALL : SYNTHESIS OF S!i!EROIDS, 307W. E. Bachmann, W.Cole, and A. L. Wilds have announced ina, letter gQ the synthesis of equilenin by the use of methods similarto those outlined above. The l-keto-7-methoxy-1 : 2 : 3 : 4-tetra-hydrophenanthrene of A. Butenandt and G. Schramml was pre-pared from l-aminonaphthalene-6-sulphonic acid and was condensedwith methyl oxalate. The glyoxylic acid derivative (CXLIII,R = OMe) lost carbon monoxide on heating, giving the 2-carbo-methoxy-compound (CXLIV). [R. D. Haworth 2 had preparedthe corresponding ethyl glyoxylate (CXLIII, R = H) but had beenunable to degrade it.] This was methylated in the 2-position andsubjected successively to the Reformatsky reaction with methylbromoacetate, dehydration, and reduction, yielding a mixture ofthe cis- and truns-esters (CXLV).The free acids were separatedand both were subjected to the Arndt-Eistert chain-lengthening,giving the propionic acids in cis- and truns-forms. The methylester of the trans-acid (CXLVI) was submitted to the Dieckmannreaction and the resulting P-keto-ester (CXLVII) was hydrolysed,decarboxylat ed and deme thylate d, yielding dl- equilenin. Resolu-tion through the Z-menthoxyacetic ester gave the form identicalwith the natural hormone. In view of the reduction of equileninto oestrone reported by R. E. Marker the' above synthesis mayconstitute also an oestrone synthesis./\//COzMe/w"02M;*2Me /\ /Pi9 \//\/\/ \CH,*CO,Me /\ / Y O z M e \\ No Me01 I1 I\/\/(CXLV.)Me01 It I \/\/(CXLIV.)011 C0,MeA/ \CH2,CHZ Me01 11 I \/\/(CXLVI.) (CXLVII.)IOa J. Amer. Ckm. Soc., 1939,61,974. 1 Ber., 1935, 68, 2083.3 J. Amer. Chem. Soc., 1938,60,1897. J., 1932, 1126308 ORGANIC CHEMISTRY.The pyrolysis of barium salts of carboxylic acids has been usedby R. P. Linstesd and his collaborators for the formation of hydrind-anones in the degradation of hydronaphthalene derivatives. 6pA. Cohen and F. L. Warren used a Dieckmann cycli~ation.~~They treated 1 : 2-dicarbomethoxy-1 : 2 : 3 : 4-tetrahydrophen-anthrene (CXLIX) with ethyl acetate, but the product on hydrolysisand decarboxylation gave not the expected diketone but its de-hydrogenation product (CL).Synthetic Oestrogens.The great difficulties involved in the complete synthesis of thesteroid sex hormones have led to the search for synthetic oe~trogens.~The discovery of the slight but definite oestrogenic activity ofl-keto-1 : 2 : 3 : 4-tetrahydr0phenanthrene,~ 4 : 4‘-dihydroxy-diphenyl and 4-hydro~y-n-propylbenzene,~ 4 : 4’-dihydroxydi-phenylethane and 4 : 4’-dihydroxystilbene 7 indicated that moleculesof weight, shape, and hydroxyl situation similar to those of oestronemight indeed prove valuable oestrogens despite the absence of thecharacteristic steroid tetracyclic system ; E.C. Dodds, L. Golberg,W. Lawson, and (Sir) R. Robinson 13 have therefore prepared C-alkylated derivatives of 4 : 4’-dihydroxystilbene (“ stilboestrol ” )and have found in trans-diethylstilboestrol (CLI) the most potentFor a review of the early work, see G. F. Marrian, Ergebn. Vitamin- u.ti J.W. Cook, E. C. Dodds, C. L. Hewstt, and W. Lawson, Proc. Roy. SOC.,Horrnonforsch., 1938, 1, 443.1934, By 114, 272.E. C. Dodds and W. Lawson, ibid., 1938, €3,125, 222.E. C. Dodds and W. Lawson, Nature, 1937, 139, 627, 1068; E. C. Dodds,M. E. H. Fitzgerald and W. Lawson, ibid., 1937, 140, 772.Ibid., 1938, 141, 247 ; 1938, 142, 34, 211 ; Proc. Roy. SOC., 1939, B, 127,140SPRINGALL : SYNTHESIS OF STEROIDS. 300oestrogen known. These C-alkylstilboestrols were prepared by thefollowing general procedure. Deoxyanisoin (CLII) was treatedwith an alkyl halide, RI, and sodium ethoxide to give (CLIII),which with the Grignard reagent, R’MgI, gave the carbinol (CLIV).This was dehydrated, by phosphorus tribromide in chloroform orby potassium hydrogen sulphate, and the product (CLV) de-methylated by alcoholic pot’assium hydroxide, yielding the dialkyl-stilboestrol.Me0 C,H,*CO*CK,*C,H,*OMe + MeO*C,H,*CO -CHR*C6B4*OMe -->MeO*C,H,*CR’( OH)*CHR*C,H,*OMe --+(CLII.) (CLIII.)(CLIV.)MeO*C6H4.CR’:CR*C6H4*OMe: --+ HO*C,H,.C~R‘:CR*C,H,.OH(CLV.)A nt i rachi t i c V it am i ns .The establishment of the tricyclic hexatriene structures ofvitamin D, (calciferol) (CLVI) and tachysterol (CLVII) or?9*17/\/\/\I \/--\/\/(CLVIII.)C%l 1 1\/\/ I II HO(CLVI.) (CLVII.)HO!(CLVIII) has led to attempts at their synthesis.The preparation,from 1 -formyl-2-~-cyclohexylethylcycZohexane, of (CLIX) con -taining the correct ring system, was early anno~nced.~Two independent approaches depending on the condensation ofcyclohexylideneacetaldehyde (CLX) with cyclohexanones haveappeared.(CLIX.) (CLX.) (CLXI.),J. B.Aldersley and N. Burkhardt prepared (CLXI, R = OAc) ;K. Dimroth l1 prepared (CLXI, R = H) and investigated replace-S. Natelson and s. P. Gott’fried, J . Amer. Chem. SOC., 1936, 58, 1432.l o J., 1038, 545. l1 Ber., 1938, 71, 1333, 1346310 ORGAMO CHEMISTRY.ment of the oxygen atom by CH,, via the Grignard reaction andthe compound (CLXII, R = CH3), and via the Reformatskyreaction and the compound (CLXII, R = CH,*CO,Et) (following0. Wallach 12).Such compounds on dehydration should yield either (CLXIII)(calciferol type) or (CLXIV) (alternative tachysterol type). Theabsorption spectra of the products indicate that they are of thelatter type.Extending this work, K. Dimroth and H. Jonsson l3have prepared (CLXI, R = OMe) and (CLXV).N. Burkhardt and N. C. Hindley 14 have obtained from l-ethinyl-cyclohexanol the dicyclohexenylethylene (CLXVI) containing thetriene system of the other alternative tachysterol type.As a model for calciferol, 3-( 2’-methylenecyclohexylidene-1‘-)-propene (CLXVIII) has been prepared from the methiodide ofthe Mannich base (CLXVII).15CH,(CLXVII.) 0’cH2*NMe311 VN0 C>(,iH2 (CLXVIII.)H. D. S.8. .HETEROCYCLIC COMPOUNDS.Oxygen Ring Cmpounds.Raney nickel is an effective catalyst for the reduction of furansto the corresponding tetrahydro-derivatives,l and an examinationof the action of acetic anhydride-zinc chloride on tetrahydrofuransindicates that the ring is ruptured more readily than the hydropyranring.The products are either diacetates (I) or unsaturated mono-la Annalen, 1909, 365, 255. 14 J., 1938, 987.1E N. A. Milss and W. L. Alderson, J . Arner. Chem. SOC., 1939, 61, 2534.R. Paul, Bull. SOC. chim., 1937,4, 846; 1939,6, 1162; Compt. rend., 1938,206,1028; R. Paul and G. Hilly, ibid., 1939,208,359; N. I. Shinkin and V. I.Bunina, J . Gem. Chem. Russia, 1938, 8, 669.l8 Ber., 1938, 71, 2658HAWORTH : HETEROCYCLIC COMPOUNDS. 31 1acetates (11) and the presence of a-side chains containing carb-ethoxy-groups facilitates the formation of (I) ., The greater stabilityof the six-membered ring is also indicated by the conversion of (111)into (IV) by the action of alumina at 400°.3/"\HZAcO*[CH,],,,,*OAc H 2 7 7 H 2 H& GHAcO*[CH,],,~*CH:CH~ 0(1.) E&C CH-QH-OH H2C CMe\/ 0Me(11.) (111.) (IV.)The formation of hydroxytetrahydrofurans of types (V) and (VI)during the reduction of @-unsaturated aldehydes with magnesiumand acetic acid is of interestY4 and 2-hydroxyfuran has been preparedby the action of sodium hydroxide and a trace of potassium chlorateupon 5-sulphofuroic acid at 3-Hydroxyfuran has beenprepared by debromination of 2-bromo-3-hydroxyfuran obtainedCH*C02HR*CH:CH*CH CH*OH R*CKCH*CH CHR 02c*HQ 11 11 R-QH-VH, HO*VH-QH, H LH 2 c v vCH, 0(v=)\d \4(V.1by the action of bromine and water on furoic acid, and the nitros-ation and nitration of the hydroxyfurans have been investigated.6The difficulty of preparation of simple 2- and 3-aminofurans isillustrated by the work crf B.H. Stevenson and J. R. Johnson andH. M. Singleton and W. R. Edwards.8 The former prepared3-amino-2-methyl- and 3-amino-2 : 5-dimethyl-furan ; the amino-group was introduced by conversion of a 3-carbethoxy-group intothe azide, which with formic acid gave the 3-formamido-derivative,hydrolysed by steam to the unstable amine. This yielded a diazo-solution which coupled with @-naphthol but showed no other diazo-reactions. The second authors obtained furyl-2-carbimide by heat -ing the azide of furoic acid, but although the carbimide was convertedinto the 2-alkylamido-derivative by treatment with Grignardreagents, attempts to obtain 2-aminofuran were unsuccessful.The diene synthesis proceeds normally with furan and its homo-logues: and the products have the endo-configuration because theyR.Paul, Compt. rend., 1939, 208, 587.Idem, Bull. Xoc. chim., 1938, 5, 919.Z. C. Glace1 and J. Wiemann, Compt. rend., 1939, 208, 1233, 1323.H. H. Hodgson and R. R. Davies, J., 1939, 806.Ibid., p. 1013.J . Amer. Chem. Soc., 1937, 59, 2525.K. Alder emd K. H. Backendorf, Annalen, 1938,535, 101.Ibid., 1938, 60, 540312 ORGANIC CHEMISTRY.are converted into bromolactonic acids inaqueous + Br,accordance with theR CO,H3-Hydroxyfuran also condenses normally with maleic anhydride,but the extranuclear double bond of furylethylene enters into thereaction ; the product, containing two double bonds, is regarded as(VII) because no formaldehyde was detected on ozonisation.1°Natural Products containing Oxygen Rings.-Euparin, isolatedfrom gravel root, is a phenolic ketone containing two double bonds.llOxidation of the O-methyl ether with permanganate yielded an acidof known structure (I; R = C0,H) and, as ozonisation gave the0H O / W \ - :CH, (11.1 I lie (1.1 M~O/\OHCH,*CO//R CH,*CO(,--corresponding aldehyde (I ; R = CHO) and formaldehyde, structure(11) was advanced for euparin.This has been confirmed by thesynthesis of tetrahydroeupa,rin (111) as follows :0redn. of 0HOf'\/\FH*CHMe,oxime J7-CH.bH20 0HO()A?H*CHMe, & HO\/\$*CHMe,J/--CH0HOm\FH*CHMe, (111.)CH3*CO"-CH210 R. Paul, Compt. rend., 1939, 208, 1028.l1 B. Kamthong and A.Robertson, J., 1939, 933HAWORTH : HETEROCYCLIC COMPOUNDS. 313The benzofuran structure (I) has been suggested for egonolobtained from ethereal extracts of the fruits of Styrux Juponica.12Egonol contains one hydroxyl and one methoxyl group and it yieldspiperonylic acid on oxidation with permanganate. Ozonisation ofacetylegonol gave the acetyl derivative of styraxin aldehyde (11),which was hydrolysed to piperonylic acid and styraxinolic aldehyde(111; R = CHO). Styraxinolic acid (111; R = C0,H) on methyl-ation and subsequent oxidation gave (IV), and on distillation itHO*[CH,],*/\R/)OH (111.)yielded dihydroconiferyl alcohol (I11 ; R = EL). The structures ofthe products (111) are well established and consequently the positionof the substituents in egonol is proved.Structure (I) for egonolhas been confirmed synthetically. Dihydroconiferyl alcohol (I11 ;= H) was converted by the Reimer-Tiemann rea,ction into thealdehyde (I11 ; R = CHO), which yielded (V) on condensation withethyl a-chloro-3 : 4-methylenedioxyphenylacetate ; hydrolysis anddecarboxylation of the ester (V) gave egonol (I). When acetyl-egonol was oxidised with hydrogen peroxide, it was converted intothe highly coloured noregonolidene acetate ; this lacks methoxylgroups but exhibits properties consistent with those of thequinone (VI) .A decision has now been made between the alternative structures(I) and (11) discussed in these Reports for 1938 (p. 311) for equol.As equol dimethyl ether did not give acetic acid by the Kuhn-Rothl2 S.Kawai and T. Miyahi, Ber., 1938, 71, 1457; S. Kawai and M. Snga,ibid., p. 2071 ; S. Kawai and F. Yoshimura, ibid., p. 2415; S. Kawai and N.Sugiyama, ibid., p. 2421 ; Proc. Imp. Acad. Tokyo, 1938, 14, 352 ; 1939, 15,46; Ber., 1939, 72, 367; S. Kawai, K. Sugimoto, and N. Sugiyama, Ber.,1932,72,963 ; S. Kawai, T. Xakamura, and N. Sugiyama, ibid., p. 1146314 ORGANIC CHEMISTRY.test, structure (I) becomes unlikely. Reduction of diadzin gave animpure dl-form of (11), but the absorption spectra of the purifieddiacetate and dimethyl ether were identical with those of diacetyland dimethyl equol re~pective1y.l~ More satisfactory evidence infavour of (11) has been obtained by oxidising equol dimethyl etherwith chromic acid.14 The resultant lavorotatory ketone (111) wasracemised by acids or alkalis, and the dl-form has been synthesisedfrom p - met hox yp hen ylace t onit rile and re sor cinol monome t h y 1 ether.The intermediate ketone (IV) was converted into the isoflavone (V)by condensation with ethyl formate, and catalytic reduction of (V)co(IV.) (V.)yielded the dl-ketone (111), reduced by Clemmensen’s method todl-equol dimethyl ether.Progress continues to be made in the natural coumarin field (seeAnn. Reports, 1937, 34, 343). Aurapten, a fish poison obtainedfrom orange-peel oil, has been assigned structure 15 (I; R =CH-CH-CMeJ. Oxidation gave acetone and 7-methoxy-coumarin-8-acetic acid, identical with the acid obtained fromosthol.16 The oxide ring is indicated by hydration with dilute oxalicacid and oxidation of the hydrate with lead tetra-acetate to 7-meth-oxycoumarin-8-aldehyde. Osthol (I ; R = CH,*CH:CMe,) has beenoxidised with perphthalic acid to dl-aurapten, which was isomerised13 F.Wesely and F. Prillinger, Ber., 1939, 72, 629.1* Miss E. L. Anderson and G. F. Merrian, J. Bhl. Chem., 1939,127, 649.15 H. Bohme and G. Pietsch, Arch. Pha~m., 1938, 276, 482 ; Bw., 1939, 72.773; H. Bohme and E. Schneider, ibid., p. 780.16 E. Spiith and 0. Pesta, Ber., 1933,66, 764./oHAWORTH : HETEROCYCLIC COMPOUNDS. 315by sulphuric acid into the ketone (I ; R = CH,*CO-CHNe,) ; thisketone was also obtained in a similar ma.nner from natural aurapten.OR OMeToddalolactone l7 is regarded as (11) because hydrolysis ethylationyielded an o-ethoxycinnamic acid giving 2 : 4-dimethoxy-6-ethoxy-benzene-1 : 3-dicarboxylic acid on oxidation ; the constitution of thedibasic acid was established synthetically.Advances in the furocoumarin group include a synthesis ofisoimperatorin (I11 ; R = CH,*CH:CMe,) from bergaptol (I11 ;R = H) and y-methyl-Ab-butenyl bromide.18 Byak-angelicin [IV ;R = CH,-CH(OH)-CMe,*OH] and byak-angelic01 (IV; R = .CH2CHGMe,) have been isolated from the roots of Angelica gl~ba.1~The former, containing two hydroxyl groups, gave furan-2 : 3-di-carboxylic acid, ct-hydroxyisobutyric acid, and bergapten quinoneon oxidation with peroxide, permanganate, and chromic acid respect-ively, and on treatment with sulphuric-acetic acid it yielded 8-hydr-oxy-5-methoxypsoralen (IV; R = H), which was synthesised fromaminobergapten.Byak-angelic01 was also converted into (IV ;R = H) by the action of sulphuric-acetic acid and into byak-angelicin by hydration.The furochromone structure (I) has been established for kellin,isolated from Ammi visnagu in 1897.20 The presence of the furanring is proved by degradation to furan-2 : 3-dicarboxylic acid, andAOMe CO OMe OMe(1.) (11.1 (111.)B. B. Dey and P. P. Pillay, Arch. Pharm., 1935, 273, 223; E. SpBth,E. Spath and E. Dobrovolny, Ber., 1939, 72, 52.T. Noguchi and M. Kawanami, Ber., 1938,71,344,1428; 1939,72, 483.B. B. Dey, and E. Tyraz, Ber., 1938,71, 1825; 1939,72,53.2o P. Font1 and S. I. Salem, Biochem. Z., 1930, 226, 166; E.Spiith and W.Gruber, Ber., 1938, 72, 10631% ORGANIC CHEMISTRY.alkaline hydrolysis yielded acetic acid and kellinone (11). Ethyl-ation and subsequent ozonisation gave (111; R1 = CHO, R2 = H),which was ethylated, oxidised, and decarboxylated to (I11 ; R1= H,R2 = Et), the structure of which was established synthetically.Kellinone (11) may be reconverted into kellin (I) by the action ofsodium acetate and acetic anhydride.Karangin, obta,ined from Pongamia glabra, is regarded as the furo-flavone 21 (I ; R = H). Alkaline hydrolysis gave benzoic acid, aphenolic ketone, probably (I1 ; R = CH,*OMe), and a phenolic acid(11; R = OH). The structure of the acid (11; R = OH) wasconfirmed by oxidation to furan-2 : 3-dicarboxylic acid, decarboxyl-ation to 4-hydroxybenzofuran, and ozonisation to (111).The ketoneR W \ , 0- <? f h O H(11.1',,,,,J\/C*OMe .-' \ P O RH*CO 0 (1.1 f70HO/\/'\C /-\ CHO L,, )\,tLoMe'-/'(111.) HO/\OH \/ b02H 0 Fl w.1(I1 ; R = CH,*OMe) was reconverted into karangin by condensationwith benzoic anhydride and sodium benzoate, but the completion ofthe synthesis by the preparation of (11; R = CH,-OMe) has notbeen effected. An acid (I; R = CO,H), prepared from ethylbromoacetate and the flavone aldehyde (IV), resisted decarboxyl-ation to karangin (I ; R = H).22The chromene structure (I) 23 has been established for seselin,obtained from Slcimmia japonica 2* and Seseli indi~in.~~ Seselinwas converted into umbelliferone (11) by the action of sulphuric-acetic acid, and it gave acetone and resorcinol-2 : 4-dialdehyde onozonisation and cc-hydroxyisobutyric acid on permanganate oxid-ation.Hydrolytic methylation yielded an o-methoxycinnamic acid,which was oxidised to the acid (111), and the structure of (111) wasproved by synthesis. Seselin (I) has been synthesised in small21 B. L. Manjkath, A. Seetharamiah, and S. Siddappa, Ber., 1939, 72, 93.22 S. Rangaswami and T. R. Seshadri, Proc. Indian Acad. Sci., 1939, 9, 259.23 E. Spath, P. K. Bose, J. Matzke, and N. C. Guha, Ber., 1939, 72, 821;24 Y. Asahina and M. Inubuse, Ber., 1930, 63, 2052; E. SpBth and 0.z 5 P. K. Bose and N. C. Guha, Science and Culture, 1936, 2, 326.E. Spath and R. Hillel, ibid., p. 963.Neufeld, ibid., 1938, 71, 353HAWORTH : HETEROCYCLIC COMPOUNDS. 317yield from umbelliferone (11) and p-methyl- Ar-butin- p-01, and theformation of the angular structure of seselin is noteworthy in viewof the previous synthesis of dihydroxanthyletin (IV) from umbellifer-one and isoprene.26Structure (I) has now been established for the substance firstisolated by T.A. Buckley27 from derris. The Izvorotatory pre-cursor, known as Z-elliptone, has been isolated from Derris eZZiptica,and racemised by sodium acetate to Buckley's compound (dl-ellipt-one).28 The identity of rings A, B, and C with those of rotenone isproved by the conversion of Z-elliptone into derric acid, and thestructure of rings D and E was established by hydrolysis of Z-elliptoneinto (11), which was synthesised from 4-hydroxycoumarone byKolbe's method.Dehydrotetrahydrosumatrol has been syn-thesised ; 29 the nitrile (111) was converted by Hoesch's method intoMe0(IV), which on condensation with acetic anhydride and sodiumacetate yielded a diacetate, hydrolysed to dehydrotetrahydro-sumatrol (V.)26 E. Spath and W. Mocnik, Bey., 1937, 70, 2276.27 J. SOC. Chem. In&., 1936, 55, 2 8 5 ~ .28 D. R. Koolhaas and T. M. Meyer, Rec. T ~ a v . chim., 1939, 58, 207, 875;28 T. S . Kenny, A. Robertson, and S . W. George, J., 1939, 1601.S . H. Harper, J . , 1939, 1099, 1424318 ORGANIC CHEMISTRY.Formulm (I) and (11) suggested previously for rottlerone areuntenable, as tetrahydrorottlerone differs from the synthetic tetra-hydro-derivatives of (I) and (II).3O Rottlerone is now regarded as9H:CHPh7H:CH"Ph $lH:CHPhP e 2 C / y j O H 0 (P q;;"I..e2 FO// -CH,OH (111.) OH(111) because hydrolysis of octahydrorottlerone (previously calledtetrahydrorottlerone) yielded a substance identical with the synthetictetrahydro-derivative of (11) ; in addition the tetrahydro-derivativeof (11) and formaldehyde reacted to yield octahydrorottlerone.These modifications do not invalidate previous views on the structureof rottlerin, but of the two structures (111) and (IV) discussed in theAnnual Reports for 1938 (p.314) the latter formula is now preferred.A modification in the structural formulz of isorottlerin has beensuggested .31The difficulties arising from demethylation of flavone derivativeshave been mentioned in earlier Reports.32 It has now been shownthat 2 : 4-dihydroxy-3 : 6-dimethylacetophenone and benzoic an-hydride react to give the flavone (I), which with aluminium chloridemay be completely or partially demethylated to 5 : 7 : 8-trihydroxy-flavone or to wogonin (a dihydroxymethoxyflavone).Structuralchange does not occur, because methylation of (I) and of wogonin\\gives the same trimethoxyflavone. WithOMe 0 0OH CO(1.) (11.1hydriodic acid, however,0(111.)30 T. Backhaus and A. Robertson, J., 1939,1257; A. McGookin, A. Robert-son, and E. Tittensor, ibid., p. 1579.31 Formula (VI) given in an earlier report (Ann. Reports, 1938, 35, 314) iserroneous ; the 2-keto-4-phenylchromone should be replaaed by a 4-keto-2-phen y lchromone structure .32 Ann.Reports, 1931, 28, 148HAWORTH : HETEROCYCLIO COMPOUNDS. 31 9structural change accompanies the demethyhtion, and baicalein(11) is obtained.33 Primetin, isolated from the leaves of Primulumodesta, has previously been assumed to be 5 : 6-dihydroxyflavone(III).34 This substance (111) was synthesised in an impure formby S. Sugasawa,36 who regarded the product as 5 : 8-dihydroxy-flavone. An unambiguous synthesis of 5 : 6-dihydroxyflavone hasnow been effected ; 36 the compound differs markedly from primetinand it is considered that the latter is probably 5 : 8-dihydroxy-flavone. Nobiletin, isolated from Citrus nobilis, is 5 : 6 : 7 : 8 : 3' : 4'-hexamethoxyflavone ; it contains six methoxyl groups, yields veratricacid and acetoveratrone on hydrolysis, and a hexahydroxyflavone,which is only pentamethylated with dia~omethane,~' on demethyl-ation.An investigation on fustin,38 a colourless crystalline substanceisolated from various Rhw species, disproves the earlier ideas of theglycosidic nature of the compound and reveals for the first time thenatural occurrence of the flavanonol type (I).Methylation of fustinyielded a trimethyl ether (I ; R = Me), which on alkaline hydrolysisgave trimethylfisetin (11), presumably by oxidation, and a phenolicacid known as trirnethylhazeinic acid (111). In accordance withstructure (111) the acid may be converted into a y-lactone, anMeO/\OH Me0a@-unsaturated acid, or a ketone (IV). The constitution of the last(IV) was deduced from an examination of the Beckmann change onthe oxime, which yielded an amide identical with that synthesisedfrom homoveratric acid and 4-aminoresorcinol dimethyl ether.Theformation of (111) from (I; R = Me) is accounted for by a benzilicacid transformation of the intermediate diketone (V). The con-33 R. C. Shah, C. R. Mehta, and T . s. Wheeler, J., 1938,1555.34 S. Rattori and W. Nagai, J . Chem. SQC. Japan, 1930,51, 162.35 J., 1933,1621 ; J . Pharm. Soc. Japan, 1936,56, 105.37 K. F. Tseng, J., 1938,1003; R. Robinson and K. F. Tseng, ibid., p. 1004.38 T. Oyamah, Annalen, 1939,538, 44.W. Baker, J., 1939, 956320 ORGANIC CHEMISTRY.stitution of fustin (I ; R = H) was confirmed by the synthesis of thetrimethyl ether (I ; R = Me) ; the dibromo-derivative (VI ; R = MeoooAc -CO*QH*QH- OoMe OMeR R(VI.)OMeMe,@ vH2-C>OMePo co (V.)Br), obtained from the corresponding chalkone, was converted intothe diacetate (VI ; R = OAc), which with hydrochloric acid yieldeda compound identical with trimethylfustin (I; R = Me).Nitrogen Ring Compounds.Pyridines, Quinoline and isoQuino1ine.-The reactivity of themethyl group of a- and y-picolines has been utilised in the prepar-ation of homologues of pyridine.Thus a-picoline, heated with analkyl halide in presence of finely powdered sodamide, gives alkyl-picolines in good yield and dialkylpicolines are frequently obtainedas higher-boiling fractions. The reaction is of wide application ;simple, complex and unsaturated alkyl groups and other groupssuch as p-ethoxyethyl and P-dimethylaminoethyl may be introducedinto the molecule of reactive Similar reactivity is displayedwhen a-picoline and benzophenone are heated in presence ofsodamide; the carbinol (I) is obtained, but when() esters, aldehydes or nitriles are employed, smallyields only of condensation products are obtained.4ON (1.1 When 2 : 4-dimethylpyridine is condensed withbenzaldehyde, a mixture of 2-styryl-4-methylpyridine and 2 : 4-distyrylpyridine is produced ; oxidation of the former to 4-methyl-pyridine-2-carboxylic acid and subsequent decarboxylation providesa convenient method for the preparation of y-~icoline.~lF.W. Bergstrom 42 had previously shown that the hydrogenatoms of the methyl groups of a- and y-picoline were replaceable bysodium and used the metallic compounds in the preparation of alkylderivatives.This author has since shown that 2-aminoquinolineis obtained in good yield from quinoline and (a) barium amide inliquid ammonia at room temperature 43 or (b) potassamide in presenceof oxidising agents such as potassium nitrate.44 Potassamide hasYH.CH2*CPh239 A. E. Tschitschibabin, Bull. SOC. chim., 1936, 3, 1607; 1937, 5, 429, 436;40 A. E. Tschitschibabjn, Rec. Trav. chim., 1938, 6, 582.4 1 G. R. Clerno and W. M. Gourlay, J., 1938, 478.42 J . Amer. Chern. Soc., 1931, 53, 1846, 3027, 4065.43 Ibid., 1934, 56, 1748.G. A. Knight and B. D. Shaw, J . , 1938, 682.44 J . Org. Chem., 1938, 2, 411HAWORTH : HETEROCYCLIC COMPOUNDS. 321also been used for the conversion of iso- and 2-phenyl-quinoline into2 - aminoiso quinoline and 4 -amino - 2 - p heny lquinoline respectively .45It has been shown 46 that a-picolinic, quinaldinic and isoquin-aldinic acids are decarboxylated by heating with aldehydes orketones, and carbinols containing heterocyclic radicals are producedin accordance with the equation :R’*C02H + R”R”‘C0 -+ R’R”R”’C*OH + CO,The reaction is specific for a-imino-acids and it is suggested that itdepends upon the intermediate formation of an anion radical con-taining the modified cyanide ion group [-5--C-]-.The additionof this ion to the carbonyl group would be analogous to cyanohydrinformation.bicycloAxa-aZEanes.-Considerable developments have occurredin the synthesis of these compounds since the subject was reviewedin 1937.*’ It has been shown that bicycZo[l : 2 : 21-aza-l-heptane(I; n = l), prepared from piperidine-4-carboxylic acid, is not anoil as previously reported but a crystalline solid identical with thatobtained from tetrahydropyran-4-carboxylic acid (I1 ; R = C02H).48Tetrahydropyran-4-carboxylic acid has been employed in severalnew directions for the preparation of bicyclo-compounds.By theaction of organo-metallic compounds on the acid chloride, alkyl oraryl tetrahydropyranyl ketones are obtained which yield the corre-sponding carbinols on reduction with sodium in presence of aqueoussodium carbonate and ether. The carbinols are converted intotribromides (111) by treatment with hydrogen bromide and thesubsequent action of ammonia leads to the formation of alkylbicycZo[l : 2 : 21-aza-l-heptanes (IV).49 Ethyl tetrahydropyran-4-carboxylate (11; R = C0,Et) is converted by means of hydrogenbromide into the dibromide (V ; R = C02Et), which with potassium45 J .Org. Chem., 1938, 3, 233, 424; Annalen, 1934, 515, 34.4 6 P. DysonandD. L. Hammick, J., 1937, 1724; M. R. F. Ashworth, R. P.4 7 Ann. Reports, 1937, 34, 380.4 8 G. R. Clemo and V. Prelog, J., 1938, 400.49 V. Prelog, E. Cerkovnikov, and (Miss) S . Heimbach, CoZZ. Czech. C’hem.REP.-VOL. XXXVI. LDaffern, and D. L. Hammick, J., 1939: 809.Coinm., 1938, 10, 399322 ORGANIO CHEMISTBY.sulphide yields the thiopyran (VI; R = C0,Et). This ester maythen be reduced to the corresponding primary alcohol (VI; R =CH2*OH) , which is transformed by concentrated hydrochloric acidCHR CHRH 2 d ' p 2BrH,C CH2Br H&, ,CH,/ \H2Q p 2(V-) s (VI.)into bicycZo[l : 2 : 21-thianium-l-heptane salts (VII; n = l).m Ina similar manner ethyl tetrahydropyran-4-acetate (I1 ; R =CH,*CO,Et) has been converted into (VII ; n = 2).514-Amino-, 4-aminome t hyl- , and 4- p-aminoet hyl- tetrahydropyran(I1 ; R = NH,, CH,*NH,, and CH2*CH2*NH, respectively) havebeen obtained by the action of sodium azide and sulphuric acid onthe appropriate carboxylic acids. A bicyclic base could not beobtained from (11; R = NH,), but in the other cases treatment ofthe amine with hydrogen bromide gave the dibromides (V; R =CH,*NH,) and (V; R = CH,*CH,*NH,), which with sodium hydr-oxide gave the bicyclic bases (I ; n = 1) and (I ; n = 2) respectivelyin good yield.52The meta-bridging of a piperidine derivative has been realised.533-p-Hydroxyethylpiperidine (VIII ; n = 2), synthesised from nico-tinic acid,= was converted into the corresponding bromide andthence into bicycZo[l : 2 : 31-aza-l-octane (IX; n = 2).bicycb-[l : 3 : 31-Aza-l-nonane (IX ; n = 3) has also been prepared ; ethylnicotinylacetoacetate was catalytically reduced to the P-piperidyl-3-propionic acid, which on Bouveault-Blanc reduction gave thealcohol (VIII ; n = 3), and the corresponding bromide was readilyconverted into the bicyclic base (IX; n = 3).5560 V. Prelog and E. Cerkovnikov, Annulen, 1939, 537, 214.5 1 V. Prelog and D. Kohlbach, Ber., 1939,72, 672.52 V.PreIog, E. Cerkovnikov, and G. Ustrichev, Annulen, 1938, 535, 37.53 V. Prelog, (Miss) S. Heimbach, end E. Cerkovnikov, J., 1939, 677.64 R. Marchant and C. S. Marvel, J . Amer. Chem. SOC., 1928, 50, 1197.56 V. Prslog, (Miss) S. Heimbach, and R. Seiwerth, Ber., 1939, '72, 1319HAWORTH : HETEROCYCLIC COMPOUNDS. 323Another method for the synthesis of bicyclic bases is indicatedby the following scheme :In this way bicycZo[O : 3 : 31-aza-l-octane (X; x = y = 3) 56 andbicycZo[O : 4 : 41-aza-l-decane (X ; x = y = 4) 57 have been prepared.The latter base was identical with norlupinan A, which was obtainedby G. R. Clemo and G. R. Ramage 58 during experiments on thelupin alkaloids and later by G. R. Clemo, T. P. Metcalfe, and R.Raper 59 by Wolff-Kishner reduction of l-ketonorlupinan (XI ;R = H).Clemmensen reduction of (XI ; R = H), however, yieldedan isomeric base, norlupinan B, which was considered to be astereoisomer of the A base, and the isomerism was regarded asresulting from the non-planar arrangement of the nitrogen valencybonds. This explanation is no longer necessary, because norlupinanB has now been shown to be identical with bicycb[O : 3 : 5l-aza-l-decane (X; x = 5 ; y = 3), which has been synthesised by themethod outlined above.60 It is suggested, therefore, that structuraland not stereochemical differences probably account for the isomericCHR CO CH2 CH, CO/ \ / \ / \ / \ / \H27 7H ( 7 3 2 H2C CH-?O H27 p 9H2\ / \ / \ / \ / \ / \ /H,C N CH, H,C N CH, H,C CH CH,CH, NH CH, CH,(XI.) (XII.) (XIII.)CH2 CH2CH, CH,/ \ / \H2V p CH2\ / \ / \(XIV.) \ / \ /CH2 CH2CH2H,C CH CH "' \7H2(XV-1 I ( CH\/ / \ '*CH2 y pH,C CH, r 70 H,C CH,5 6 V.Prelog and (Miss) S. Heimbach, Ber., 1939, 72, 1101.6 7 V. Prelog and K. Bozicevic, ibid., p. 1103.5 8 J., 1931, 437.60 V. Prelog and R. Seiwerth, Ber., 1939, 72, 1638.59 J., 1936, 1430324 ORGANIC CHEMISTRY.bases encountered during reduction of (XI ; K, = Me), (XII), (XII1) ,(XIV) and (XV).slInteresting results have been obtained during attempts to extendthese experiments to the preparation of quinine derivatives.62 Ethyltetrahydropyran-4-p-propionate (I1 ; R = CH2*CH2*C02Et) con-denses with ethyl cinchonate and quinate to yield the ketones(XVI; R = H) and (XVI; R = OMe) respectively. Successivetreatment with hydrobromic acid and bromine converts (XVI) intothe tribromide (XVII), but attempts to transform this into therubatoxanone-9 (XVIII) haveco-(XVI.)P \ C H 2FH212 I I ,,CH2- CH(XVIII.)been unsuccessful.The piperidine@r Br Br(XVII.)(XIX.)(XIX; X = NH) and thiopyran (XIX; X = S) derivatives havebeen prepared from (XVI) by Wolff-Kishner reduction of thecarbonyl group, followed by rupture of the pyran ring with hydro-bromic acid and subsequent ring closure to (XIX) by ammonia orpot a ssium sulp hide.Another interesting approach to the quinine structure dependsupon the condensation of 3-ketoquinuclidine (XX) with quinoline-4-CHoc’ j \CH,I ECH212ITH=C 1 ,CH,/’‘\,,/\ \N/i l l\/‘W (XXI.)aldehyde to give the unsaturated ketone (XXI), which was reducedto 5-ketoruban (XXII).63 The latter ketone was converted byG.R. Clemo and G. R. Ramage, J . , 1932, 2970; G. R. Clemo, J. G. Cook,andR. Raper, J., 1938, 1184, 1318.V. Prelog, R. Seiwerth, V. Hahn, and E. Cerkovnikov, Ber., 1939,72,1325.63 G. R. Clemo and E. Hoggarth, J., 1939, 1241HAWORTH : HETEROCYCLIC COMPOUNDS. 325aluminium isopropoxide and ethylmagnesium iodide into ruban-5-01(XXIII; R = H) and 5-ethylruban-5-01 (XXIII; R = Et) respect-ively, but the constitution of the product obtained by the action ofethylmagnesium iodide on the unsaturated ketone (XXI) has notbeen established. /?*\ OC I CH,-CHI ,CH,I PH212I\N* '(XXII.)Adermin.-This name has been suggested for vitamin B,, therat-dermatitis-preventing factor of the vitamin B complex.Adermin has been obtained from rice bran as a crystalline hydro-chloride of an optically inactive, weak tertiary base, C,Hi,O,N,containing one C-methyl, one phenolic, and two primary alcoholicgroups.64 The absorption spectrum of adermin resembles that of3-hydroxypyridine and differs considerably from those of 2- and4-hydro~ypyridine,6~ and tests with the Folin-Denis reagent confirmthe 3-hydroxypyridine structure.The methyl ether of adermin isunattacked by lead tetra-acetate, but oxidation with alkalinepermanganate gives a methoxypyridine tricarboxylic acid whichloses carbon dioxide and yields a methoxypyridinedicarboxylic acidon heating.The carbon dioxide is probably eliminated from thea-position, because the dicarboxylic acid, unlike the tribasic acid,does not give the ferrous sulphate test. Structure (I), assigned tothe dibasic acid on the basis of the degradation work, has been con-firmed synthetically,66 and the tribasic acid must therefore beeither (11; R = CO,H) or (111; R = C0,H). Oxidation ofadermin methyl ether with barium permanganate 67 gives a methoxy-methylpyridinedicarboxylic acid, which gives a negative ferroustest and yields a 3-hydroxy-a-picoline on decarboxylation ; adecision between (I1 ; R = Me) and (I11 ; R = Me) has been madein favour of the former by synthesis from 4-methoxy-3-methyliso-64 P. Gyorgy, J . Am.er. Chern. SOC., 1938,60, 983; R.Kuhn and G. Wendt,Ber., 1938, 71, 780, 1118,1534; J. C. Keresztesy and J. R. Stevens, Proc. Exp.Biol. Med., 1938, 38, 64; J . Amer. Chem. SOL, 1938, 60, 1267.65 R. Kuhn and G. Wendt, Ber., 1939, 72, 305; E. T. Stiller, J. C.Koresztesy, and J. R. Stevens, J. Amer. Chem. SOC., 1939, 61, 1237.6 6 R. Kuhn, H. Andersag, K. Westphal, and G. Wendt, Ber., 1939,72, 309.67 R. Kuhn, G. Wendt, and K. Westphal, ibid., p. 310; E. T. Stiller, J. C.Keresztesy, and J. R. Stevens, loc. cit326 ORGANIC CHEMISTRY.quinoline by oxidation.ss Additional evidence is provided by theoxidation of adermin methyl ether with neutral permanganate ; 69HO,C/\,OMe H02C/)gMe H0,CPOMeC0,H C0,H C0,H\N/ R\J(111.)\N/(1.) (11.)O-FH, CH,*OHHO*CH,/)g: (V.)OqJ:? (N/(IV.)a lactone, C9H,0,N, is obtained for which structure (IV) has beenestablished synthetically. It follows that adermin must havestructure (V), which has been confirmed by the two independentsyntheses (A) and (B) outlined below :CH,eOEt CH,*OEt CH,*CO*CH,*CO*CH,*OEt (4 'O + N C A HE0; NC/\NO,NC*CH,*CO*NH, --+ HO HO. &,!MeA lkaloids.Indole Group-Final confirmation of the structure of eserine( I ; R = NH*CO,Et) was obtained by the synthesis of 1- and68 R. Kuhn, K. Westphal, G. Wendt, and 0. Westphal, Naturwiss., 1939,27, 469.69 E. T. Stiller, J. C. Keresztesy, and J. R. Stevens, J. Amer. Chem. SOC.,1939, 61, 1237; A. Ichiba and K. Micbi, Sci. Papers Inst. Phys. Chem. Res.Tokyo, 1938, 35, 73.70 E. A. Harris, E.T. Stiller, and K. Folkers, J . Amer. Chem. SOC., 1939,61, 1242, 1245HAWORTH : HETEROCYCLIC COMPOUNDS. 327&I-eserethole (I; R = Et) by P. L. Julian and J. Pikl in 1935.71An isomeric synthetic base was regarded a8 a cis-trans-isomer ofdl-eserethole by F. E. King, M. Liguori, and R. Robinson, but otherworkers preferred structure (11) . This improbable structure (11)has now been replaced by structure (I11 ; R = NMe2).V2 The base,which contains a dimethylamino-group but no active hydrogen atom,yielded on thermal decomposition 5-ethoxy-3-methylindole, andstructure (I11 ; R = m e , ) has been established synthetically.Me CH,(1.)NMe m e 2Magnesium 5-ethoxy-3-methylindole was condensed with ethylenedibromide, and the product (I11 ; R = Br) yielded (I11 ; R = NMe,)with dimethylamine.The formation of this indole derivative (111 ;R = NMe,) takes place during methylation of (IV; R1 = R2 = H),(IV; R1 = H ; R2 = Me), and, under certain conditions, from(IV ; R1 = Me ; R2 = H) and the fission of the terminal pyrrolidinering is noteworthy.Calycanthine, isolated from Calycanthus species , mas given theformula C,,H2,N,,H20 by E. Spath and W. Stroh in 1925.73 Thealkaloid contains an NMe group, the presence of a secondary amino-group was indicated by the formation of a nitrosoamine, andZerewitinoff determinations gave results varying with the temper-ature of the experiment. Oxidation of the amorphous benzoylderivative with permanganate yielded benzoyl-N-methyltryptamine(I) 74 and selenium dehydrogenation gave norharman (11) and abase, C,6H182.75 G.Barger, J. Madinaveitia, and P. Streuli i 6 haverecently suggested C22H26N4 as the molecular formula for caly-canthine. They obtained N-methyltryptamine by heating caly-canthine with soda lime, and a base, possibly a methyl-4-carboline,by distilling the alkaloid with lime. The two products are regardedas arising from different parts of the molecule of the alkaloid. WhenAnn. Reports, 1935, 32, 343.T. Yobayashi, Annalen, 1938,536, 143; 1939, 539, 213.73 Ber., 1925, 58, 2131.74 R. H. F. Manske, Canadian J. Res., 1931, 4, 275.7 5 L. Marion and R. H. F. Mmske, {bid., 1938,16,432.76 J., 1939, 610328 ORGANIC CHEMISTRY.calycanthine is distilled with numerous reagents or oxidised withchromic acid, it yields the extremely stable, weak base, C,,HIoN2,CH,NH CHMenamed calycanine, for which structure (111) is tentatively advanced.The base contains both secondary and tertiary nitrogen atoms andthe structure is supported by the formation of quinoline by theaction of hydriodic acid on the alkaloid or of soda lime on benzoyl-calycanthine.Structure (IV) suggested for the alkaloid must beregarded as highly speculative, but a reversible colour reaction withp-dimethylaminobenzaldehyde, which appears t o be characteristicof tetrahydroharman derivatives) leads to the assumption of areduced pyridine nucleus.A new base, calycanthidine, which accompanies calycanthinehas the formula C,,H,,N2. It contains an NMe group, a secondaryamino-group, and possibly a C-methyl group and it may be Z-N-methyltetrahydroharman (V).The dl-form of (V) has been syn-thesised, but attempts to resolve this base or t o racemise caly-canthidine were unsuccessful and, as some differences in reactivitywere observed between the synthetic and the natural product, thestructure of the alkaloid remains uncertain.Lupinan Goup-Structure (I), assigned to cytisine by H. R. Ingin 1931, was modified to (11) by E. Spiith and F. Galinovsky in 1932.i \ / Y H - P (\( VH2 VHCH2QH-(F2 / \N CH-CH, CH, CH,CH,*[CH,],*yH FHMe N yH2 SJH\/ \CMe-CH, \do \CH2 \ /NAcco(1.1 (11.) (111.)'' G. Bargor, (Miss) A. Jacob, and J. Madinaveitia, Rec. Trav. chim., 1938,57, 548HAWORTH : HETEROCYCLIC COMPOUNDS.329By a combination of reduction and exhaustive methylation methodsthe alkaloid was converted into (111), the constitution of which wasindicated by dehydrogenation and oxidation to S-methylpyridine-5-carboxylic acid,78 thus establishing the piperidine structure (11).Further evidence supporting (11) was obtained by degrading cytisineto (IV),79 the structure of which has recently received the followingsvnthetical confirmation : 8oCHMe CH,/ / \ Me/\(A\ / \ / \/ b< H 2 7 MeCH b N YH2 CH2 --+ p,, hIeI N I (IV.1CH, COdl-Lupinine (I), the synthesis of which was reported in 1937,81has been resolvedB2 and the Z-base is identical with the naturalalkaloid. The alkaloids aphyllidine and aphylline, isolated fromAnabasis aphylla, are members of the lupinan group ; aphylline is astereoisomer of oxysparteine (11) , and aphyllidine, which is con-verted into aphylline by catalytic reduction and into d-sparteineby electrolytic reduction, is regarded as A5: 6-dehydro-oxysparteine.83Octalupine, isolated from Lupinus sericeus, yields sparteine andd-lupanine on reductionIand it is regarded as 2 : 16-diketo~parteine.~~CH, CH?H,*OH /'I \CH, CH H274 6VH 177H287H2H,C3 'N CH15N9 (TI.)\"/ v A 1 0w/CH2/\A=/ CH, CO >H14 7H2 (1.) Hz? p 7H2H,C N CH,H,C13 CH,"CH, CH,The alkaloid matrine, isolated from Sophora species, is isomericIt is a tertiary with and shows certain resemblances to lupanine.E. Spath and F. Galinovsky, Ber., 1933,66, 1338.Idem, ibid., 1936, 69, 761.81 Ann.Reports, 1937, 34, 359.B3 G. R. Clemo, W. McG. Morgan, and R. Raper, J . , 1938, 1574.83 A. Orhkhov and G. Menachikov, Ber., 1932,65, 234; A. Orekhov, J . Gen.84 J. F. Couch, J . Amer. Chern. SOC., 1939, 61, 1523.80 Idem, ibid., 1938, 71, 721.Chem. Russia, 1937,7, 2048330 ORGANIC UHF&tISTRY.base and contains a Iactam group which is readily hydrolysed to givematrinic acid. When matrinic acid is distilled with zinc dust, amixture of matridine, C16H26N2, and p-lupinan (I; R = Me) isobtained, and distillation with soda lime yields a- and p-matridines,C , , H , ~ , , together with a fraction giving 2-butylpiperidine andnorlupinan (I; R = H) 85 on reduction.CH, CHRH,C N CH2\A/ CH, CH,NHAc-CH, COGH, 03 (111.)N0 5 H 2 H 2 6 Z 2 CH CHStructure (11), which is suggested for a-matridine, is consistentwith its reduction to a dihydro-derivative, the formation of theketone (111) with acetic anhydride, and with the dehydrogenationto (IV; R =Me), which gives a benzylidene derivative.86 Thedehydro-derivative (IV ; R = Me) gives a lithium derivative whichwith ethyl bromide yields (IV; R = CH,*CH,Me), identical with abase, C14H20N2, obtained by dehydrogenating matrine with palladiuma t 300".On the basis of (11), structure (V) 87 has been suggestedfor matrine and the piperidine nature of ring A is more probablethan a methylpyrrolidine structure because of the isolation ofglutaric and succinic (and not methylsuccinic) acids as productsof oxidation of methyl matrinate. The structure of the alkaloid is,however, still unsettled and alternative formulae more closely relatedthan (V) to other lupinan alkaloids cannot be excluded. Oxy-matrine, isolated from Xophra JEaveScens,88 is probably matrineN-oxide and it may be prepared by the action of hydrogen peroxideon matrine.89isoQuinoZine &oup.-Structure (I; R = H) has been established85 H.Kondo, E. Ochiai, K. Tsuda, and S. Yoshida, Ber., 1935, 68, 570.8 6 H. Kondo, E. Ochiai, andK. Tsuda, ibid., 1936, 68, 1899.8 7 K. Tsuda, ibid., 1936, 69, 429; J. Pharm. SOC. Japan, 1937, 57, 68.** H. Kondo, E. Ochiai, and K. Tsude, Arch. Pharm., 1937, 275, 493.** E. Ochiai and Y . Ito, Ber., 1938,71, 938HAWORTH : HETEROCYCLIC COMPOUNDS. 33 1for Z-tudarinine;90 the diethyl derivative (I; R = Et) has beensynthesised and the Hoffmann degradation products are identicalwith those obtained from the naturally occurring base.Arta-botrine, isolated from the bark of Artabotrys swveolens Bl., is anaporphine base containing an alcoholic hydroxyl group for whichE Z g Y R E g f O N M e MeOpH:CH, M e 0 8structure (11; R = Me) has been sugge~ted.~~ Hofmann degrad-ation yielded first an optically active methine and finally aphenanthrol, probably (111), which gave a, stable dimethoxy-lactonic acid on oxidation with permanganate. The properties ofthe lactonic acid are not consistent with a structure such as (IV),and structure (V) is regarded as probable. On this assumption itis concluded that ring A of the alkaloid contains two methoxylgroups and the 5 : 6-arrangement is based on analogy. Artabotrine(11; R =Me) is the O-methyl ether of an accompanying base,suaveline (I1 ; R = H), which gives the Pellagri reaction, indicatingRO DycH*oH Me0(111.)RO( 2/(1.) ( 11.1coMeO( Meo2 CO,H(V. 1 (VI.)that the p-position of the phenolic group is unsubstituted. Thisexcludes positions 2 and 3 for the hydroxyl and the methoxyl groupin ring D of suaveline and artabotrine respectively and position 4is selected because the remaining position 1 is unoccupied in allknown alkaloids of the group. A third alkaloid, artabotrinine,contains a methoxyl, a methylenedioxy-, and a secondary amino-group, and it probably represents the O-methyl ether of anolobineErgot AZkuZ~ids.~~-The difficult constitutional problem of these(VI) .92K. Goto and H. Shishodo, Annalen, 1939, 539, 262.O1 G. Barger and L. J. Sargent, J., 1939, 991.s2 Ann. Reports, 1938, 35, 326. 93 Ibid., 1935, 32, 345; 1936,33,374332 ORGANIC CHEMISTRY,alkaloids is still unsolved. Since the last review a fifth pair ofalkaloids, zlix., ergocristine and ergocristinine, has been isolated,g4the identity of erg~metrine?~ erg~basine,~~ and ergotocine 97 hasbeen confirmed,98 and ergosine and ergosinine have been shown toyield lysergic acid, ammonia, d-proline, l-leucine, and pyruvic acid onhydrolysis .99An important advance concerning the mode of attachment oflysergic acid to the amino-acids of the molecule has been made.The ergotoxine-ergotinine and ergotamine-ergotaminine pairs havebeen reduced to their corresponding dihydro-derivatives, which onhydrolysis yield isobutyrylformic and pyruvic acids respectively.As these keto-acids are reduced to the corresponding hydroxy-acidsunder similar conditions, it follows that the keto-acids do not occuras such in the alkaloids and it is suggested that they occur ascc- hydroxyvaline and cc- hydroxyalanine residues respectively asshown in (I), where L = lysergyl and R = Me for ergotamine andCHMe, for erg0toxine.l The outstanding problem in connectionY (11.1 (111.)with the ergot bases is the structure of the lysergic acids which areobtained by hydrolysis of the alkaloids. S. Smith and G. M.Timmis 2 established the following relationships of the isomericlysergic acids :boil H,Oalkalidl-lysergic acid (Ba(oH)a d-lysergic acid -f d-isolysergic acidA. Stoll and A. Hoffmann treated ergot bases with hydrazineand obtained dl-isolyserghydrazide, which gave dl-isolysergic acid94 A. Stoll and E. Burckhardt, 2. ph,ysioZ. Chem., 1937, 250, 1 ; 1938, 251,287.96 H. W. Dudley, Phurm. J . , 1935,134, 709.913 A. Stoll and E. Burckhardt, Compt. rend., 1935, 200, 1680.97 M. S. Kharasch and R. R. Legault, J . Amer. Chem. SOC., 1935, 5'9, 956.98 A. Stoll and E. Burckhardt, Schweiz. med. Woch., 1936, 66, 353.99 S. Smith and G. M. Timmis, J., 1937, 396.1 W. A. Jacobs and L. C. Craig, J . Biol. Chem., 1938,122,419.2 J., 1936, 1440.9 2. physiol. Chem., 1937, 250, 7 ; 1938, 251, 155HAWORTH : HETEROCYCLIC COMPOUNDS. 333by hydrolysis of the corresponding azide with sodium bicarbonate,but hydrolysis of the hydrazide with strong alkali gave d2-lysergicacid. d2-isolysergic acid was resolvea with nor-Z-ephedrine andhydrolysis of the salts with strong alkali was accompanied byisomeric change, and the formation of d- and Z-lysergic acids. Bycombination of dl-isolysergazide with d-p-aminoisopropyl alcoholand separation with alumina, d-isolysergo-d-p-hydroxyisopropyl-amide, identical with d-ergometrinine, was obtained ; it was con-verted by phosphoric acid into d-ergometrine. The 2-forms of thealkaloidal pair were obtained simildy. As a result of these experi-ments it is concluded that the &(physiologically weak) and 2-(physio-logically active) ergot bases are related to the isolysergic andlysergic acid series respectively. Structure (11), introduced in 1936for lysergic acid, accounts for many properties of the acid, and theconversion into isolysergic acid is interpreted by a migration of the5 : 10 double bond. A base, ergoline (111), giving several colourreactions of lysergic acid, has been synthesised as follows : 4%Amino- 1 -naph-thoic acidSkraupreaction +-INa + butylalcohol(WRecently a modification has been introduced and lysergic acid isregarded as the p-amino-acid (V).5 Since the basicity of lysergicacid is smaller than that of isolysergic and dihydrolysergic acids, thedouble bond is placed in the 5 : 10- and 9 : 10-positions in lysergicand the iso-acid respectively. Two-stage reduction of the meth-iodide of base (IV) yielded 6-methylergoline (I11 ; with NMeinstead of NH in position 6),6 and the difference in basic dissociationconstant between this base and lysergic acid corresponds to thesubstitution of a carboxyl group in the P-position with respect tothe NMe group. Structure (V) for lysergic acid is consistent withthe formation of 3 : 4-dimethylindole by fusion of dihydrolysergic4 W. A. Jacobs and R. G. Gould, J . Biol. Chem., 1937,120,141.6 W. A. Jacobs, L. C. Craig, G. Shedlovsky, and R. G. Gould, ibid., 1938,6 W. A. Jacobs and R. G. Gould, ibid., 1938,126, 67.W. A. Jacobs and L. C. Craig, ibid., 1939, 128, 715.125, 289334 ORGANIC CHEMISTRY.acid with alkali, and a new compound, C,,H1,ON,, obtained bydistillation of dihydrolysergic acid at 25 mm., is represented bystructure (VI).8ergoline, the dl-form of which has been synthesised.9Catalytic reduction of (VI) gave 6 : 8-dimethyl-R. D. H.W. A. Jacobs and L. C . Craig, J . Amer. Chem. SOC., 1938,60,1701.* W. A. Jacobs and R. G. Gould, J . Biol. Chem., 1939,130,399