ORGANIC CHEMISTRY,1. INTRODUCTION.THE subjects selected for inclusion in this Report deal with some recentlyintroduced general methods, developments in the reactions of fiee radicalsand atoms, mono- and di-terpenes, colchicine and related compounds, somereactions of organic sulphur compounds, and certain groups of heterocycliccompounds.I n a survey of new general methods reference is made to the introductionof lithium aluminium hydride, which has proved to be a most useful reagentfor effecting a wide variety of reductions, and in particular the ready con-version of -CO,H into -CH,*OH. The process of cyanoethylation is findingnumerous applications in synthetic organic chemistry, and a considerablevolume of new work has centred around the preparation and properties ofcyanides.Particular refhence should be made to the conversion of cyanidesinto amidines by a number of new reactions. The introduction of the-CH,=CO,H group directly into an aromatic nucleus has been effected bya dehydrogenating condensation with acetic anhydride and potassiumpermanganate. A general method for preparing amides has been reportedwhich consists of heating the acid with urea. Several new general methodsfor the preparation of certain amino-acids have been published, and mentionis made of the formation of a p-lactam from diazomethane and an isocyanate.Great advances have been made in the utilisation of simple compoundssuch as acetylene, ethylene, and carbon monoxide, and the processes ofhydroxylation, chloromethylation, aminomethylation, sulphomethylation,nitrosation, oxynitration, and nitration (with sulphuric acid) have receivedconsiderable attention.Certain general reactions among organic com-pounds of phosphorus have now become h l y established. Dibenzylchlorophosphonate has been employed in the synthesis of complex sub-stances such as adenosine triphosphate, and esters of fluorophosphoric acid(having powerful anti-cholinesterase activity) are readily prepared by simplemethods.A survey of developments in the reactions of free radicals and atomsduring the past four years shows that reactions of this type are more wide-spread than seemed likely a comparatively short time ago. Many newexamples have been brought to light, and the added knowledge thus gainedconfirms and extends the main theoretical views which had already beendeveloped. There is now clear evidence of a duality of function in manycompounds which, under appropriate experimental conditions, can takepart in either a homolytic or a heterolytic process.M. S. Kharasch andhis collaborators have made an extensive study of the uses of acetyl per-oxide, which acts as a convenient source of methyl radicals, and theseinvestigations have led to the development of many novel synthetic methodsHEY AND JONES : INTRODUCTION. 119The use of acyl peroxides for the alkylation of quinones has also been furtherextended. Considerable attention has been devoted to the kinetics of thedecomposition of benzoyl peroxide in solvents; this is now considered toinvolve a secondary chain reaction in addition to the true unimolecularprocess.The remarkable properties of tert.-butyl peroxide have attractedwide interest ; its reactions have been studied both in the vapour phase andin a variety of solvents, and a simple rate-determining dissociation process,involving the scission of the 0-0 bond, is common to both types of reaction.New contributions have been made to the theory of free-radical additionreactions, and a considerable extension of our knowledge of these reactionshas resulted from further work on the addition to olefins under peroxidicconditions of various halogen derivatives of methane, derivatives of halo-genated acids, phosphorus trichloride, trichlorosilane, and various sulphurcompounds.Me* + XY -+ MeX + Y*These reactions can be represented by the general scheme :R*CH:CH, + Ye + R*(?H*CH,YR*kH*CH,Y + XY + R*CHX*CH,Y + Y*The substitution reactions of atomic halogens have also been furtherexplored.For bromination great use has been made of N-bromosuccin-imide, and for chlorination further reactions have been described withsulphuryl chloride in the presence of a peroxide. New reactions whichreveal a free-radical mechanism include those of diphenyliodonium hydroxideand phenyl iodosoacetate, and oxidation processes using selenium dioxideand chromic anhydride have also shown free-radical characteristics. Theuse of cobaltous chloride to induce free-radical reactions with Grignardreagents has been further extended.The review of recent advances in the terpenes is limited to some aspectsof the chemistry of the mono- and di-terpenes. The monoterpene sectionis devoted to the important series of esters derived from the chrysanthemummono- and di-carboxylic acids (I ; R = Me, and I ; R = C02Me, respectively),monoterpenic acids containing the unusual feature of an unfused cycEo-propane ring, and cyclic keto-alcohols of the type (11; R = alkyl oralkenyl).The esters derived from (I1 ; R = CH,*CH:CHMe orCH,*CH:CH*CH:CH,) are the constituents of pyrethrum flowers (Chrys-anthemum cinerariifolium) responsible for their remarkable insecticidalproperties. As these substances have not been reported on before, a briefsurvey has been made of the development of our knowledge of their struc-tures from the time of their discovery by H.Staudinger and L. Ruzickamore than twenty-five years ago. Although their total synthesis has notyet been accomplished, the remaining difficulties should not prove insuramouhtable. By the study of synthetic analogues much light is bein120 ORGANIC CHEMISTRY.thrown on the relationship of toxicity to structure. The section on thediterpenes is devoted to the resin acids containing the hydrophenanthreneskeleton. Although the structure of abietic acid was settled by 1941, thestructures of sapietic, pimaric, and other resin acids have remainedin doubt until recently-. Conspicuous advances have been made in theisolation of primary resin acids, which have disclosed that abietic, di-hydroabietic, and dehydroabietic acids, formerly regarded as artefacts,do in fact occur as such in pine oleoresin.The structure of pimaric acidhas finally been settled as one containing a gem.-vinylmethyl group at C,in preference to a structure with an angular vinyl group at C14. Workon agathic acid (a dicyclic diterpene) and on podocarpic acid, which is nota terpene although containing the hydrophenanthrene skeleton, is described,since this has an important bearing on the elucidation of the stereochemistryof the resin acids. The configurations of C,, C,,, and CI2 have now beenelucidated with reasonable certainty by a study of their interrelationships,by degradation to a common tricarboxylic acid in which ring A remainsintact, and by a study of lactone formation in the dihydro-resin acids.The application of a theoretical analysis of the dissociation constants ofthe tricarboxylic acid to this configurational problem together with thesuggestion that lactone formation is preceded by structural rearrangement(the angular methyl group migrating from C,, to CIS) constitutes a usefuadditional approach to these problems.The concept of cydoheptatrienolone (tropolone) as a new type of reson-ating aromatic system has proved extremely useful for interpreting thestructures of a numbe’r of natural products.These are shown to range incomplexity from the relatively simple mould-product stipitatic acid and agroup of products isolated from the heartwoods of red cedar, throughpurpurogallin and its analogues which are now regarded as benztropolonederivatives, to the tricyclic system of the acetylated alkaloidal amine,colchicine.I n these compounds, generally, ethylenic and carbonyl func-tions are revealed by the results of hydrogenation but are masked towardsthe usual reagents and become incorporated in a carboxylated benzenoidring through rearrangement in strongly alkaline media ; there is supple-mentary evidence for the tropolone structure in: most of them, and it nowrewins for the rational synthesis of a typical tropolone to set the seal onthe degradative and interpretative work. The present state of the structuralchemistry of colchicine is reviewed. Recent work indicates that the com-pound mag contain a second 7-membered ring in addition to the tropolonesystem.After the tropolone ring has been converted into a benzenoidform, further degradation yields products which are identified as dibenz-cycloheptatrienes. Attention is thereby directed to the synthesis andtransformations of compounds of this type.I n recent years, interest in the chemistry of organic sulphur compoundshas been stimulated by the discovery of several physiologically activenatural products containing sulphur-for example, aneurin, biotin, andpenicillin. A t the same time, an appreciation of the use of organic sulphuHEY AND JONES : INTRODUCTION. 121compounds, which are often more reactive than their oxygen analogues, hasled to new synthetic methods and improvements on previously known ones.In 1939 the remarkable discovery was made that sulphur can be removedfrom, or replaced by, hydrogen in many different ty-pes of inorganic andorganic substance by means of Raney nickel.This reaction has alreadyfound numerous applications in synthetic work, and has been particularlyuseful for the elucidation of the structure of organic sulphur compounds :thus it was of great importance in connection with biotin and penicillin.Raney-nickel hydrogenolysis does not cause racemisation of opticallyactive molecules ; for example, natural( -)rnethionine yields L( +)-%-amino-butyric acid, thereby confirming that the former has the same absoluteconfiguration as the other natural (L) a-amino-acids. Similarly, the stereo-chemistry of penicillin and of the natural and synthetic stereoisomers ofbiotin has been studied.The ease of cyclisation of many thioacylamido-compounds has led to new methods for the synthesis of derivatives ofglyoxaline, oxazole, and thiazole. The required intermediates can some-times be made by the action of phosphorus pentasulphide on the corre-sponding amides, but this reaction is not always satisfactory. A muchbetter method is the direct thioacylation of amines by means of dithio-acids, their salts or esters, or by thion-esters. The Willgerodt and Kindlerreactions have been studied extensively in the last few years and so modifiedthat excellent yields of substituted phenylacetic acids are now obtainable.The preparation of aromatic aldehydes by the McFadyen-Stevens and theWuyts reaction, the use of thiourea for the synthesis of pyrimidines andpurines, of 5-methylthiouronium salts for guanidines and rhodanine, andof 2-thiothiazoline for a-amino-acids have also been outlined.In the heterocyclic series the chemistry of aziridines, P-biotin, and thepterins is reviewed.Under the first heading attention is directed to themore recent interest shown, since 1941, in the preparation and propertiesof ethyleneimines. Under the second heading recent syntheses of three ofthe stereoisomeric racemates of the biotin structure including the hithertounknown ( f )-epibiotin are outlined. By chemical control of configurationsthe methods make feasible the preparation of all four racemates from asingle intermediate, and the need to separate mixtures of racemates isthereby obviated.Finally, a review is included of the advances made inthe field of pterin chemistry since the previous Report of two years ago,New syntheses of vitamin B, and pteroic acid are reported together withexperiments concerning the degradation and synthesis of the fermentationL. casei factor and a new pterin growth-factor, rhizopterin. Attention isdirected to investigations on the reduction of some pteridines, includingvitamin B,, the synthesis of compounds antagonistic to vitamin Bc, andpreparations of numerous simpler pteridines. Experiments on the utilis-ation of sugars, and recent views concerning the authenticity of preparationsof hgdroxy- or halogeno-methylpteridines, are ttlso reported.13. H. H.B. J122 ORGANIC CHEMISTRY.8. (3ENERAL mTHODS.Reduction.-Since the last Report on General Methods,l outstandingadvances have centred round the use of lithium aluminiiiin hydride, LiAIH,.Th4 organic chemist has always been handicapped by the lack of a suitablegeneral method for converting the -CO,H group into -CH,*OH.This newreagent now promises to give the answer to this age-long problem, and tobe of considerable general use in effecting a wide variety of reductions.Lithium aluminium hydride was discovered in 1947 by A. E. Finholt,A. C . Bond, and H. I. Schlesinger who treated finely divided lithiumhydride with an ethereal solution of aluminium chloride. Lithium aluminiumhydride was thus obtained as an ether-soluble solid according to theequation :4LiH + AICl, --+ LiAlH, + 3LiC1.Larger quantities of aluminium chloride gave aluminium hydride :3LiAlH, + AlC1, -+ 4A1H3 + 3LiC1.By the use of lithium aluminium hydride, new methods, simpler than thosealready available, were a t once developed €or the preparation of hydridessuch as silane and stannane and of their alkylated derivatives. (Thehitherto unknown hydridcs of zinc and beryllium have also been prepare’d.) :LiAIH, + SiCl, -+ LiCl + AlC1, + SiH,LiAlH, + 2Me,SnCl, --+ LiCl + AlCl, + ZMe2SnH,LiAlH, + ZnMea -+ LiAIMe,H, + ZnH,These reactions usually proceed smoothly at room temperature, and givehigh yields of pure products. The authors point out that in certain reduc-tions of this type lithium hydride and aluminium hydride can be used inplace of lithium aluminium hydride. With lithium hydride, however, thereacticns are slower and the yields poorer.R.F. Nystrom and W. G. Brown3 used lithium aluminium hydride forrcducing aldehydes, ketones, esters, and acid chlorides to the correspondingalcohols. The reaction proceeds in ether, at room temperature, and yieldsare of the order 70-98%. Points in favour of this reagent are : (a) nospecial apparatus is required, ( b ) it is claimed to be indefinitely stable atroom temperature. The most notable advance, however, by these authors,concerns the smooth reduction of carboxylic acids to the correspondingprimary alcohol by lithium aluminium hydride. An ethereal solution ofthe acid is added dropwise to an ethereal solution of the reagent at such arate as to maintain a gentle reflux.After 15 minutes, the cooled productis diluted with water and treated with 10% sulphuric acid (or loo/, sodiumhydroxide solution), leaving the alcohol in the ethereal layer. Free hydroxyland amino-groups do not interfere. The double bond in cinnamic acid,but not those in sorbic or furoic acid, is simultaneously hydrogenated.1 R. A. Baxter and F. S. Spring, Ann. Reports, 1945,42, 96.* J . Amer. Chem. SOC., 1947, 69, 1199.208 ORGANIC CHEMISTRY.Dithio-acids are unstable and decompose on keeping. Dithioformic 58 anddithiophenylacetic 69 acids are best stored as the stable potassium salts.Dithio-esters are stable, and the methyl estcrs of dithiophenylacetic 68, 7Oand dithio-n-hexoic 7 l acids have been used for thioacylation.Dithio-benzoic acid is less reactive than the aliphatic dithio-acids and does notR*CS,Me + NH,*CHR’*CO,H .“.‘k“‘t R*CS*NH*CHR’*C02Hreact with glycine or leucine even on warming.7, Amino-acids can, how-ever, be thiobenzoylated by means of carboxymethyl dithi~benzoate.?~Ph*CS,*CH,*CO,H + NH,*CHR*C02H ,-f%+ Ph*CS*NH*CHR*CO,HUnfortunately aliphatic dithio-acids (except dithioformic 74 and dithio-phenylacetic 75) cannot be obtained in good yield. However, aliphatic 69and aromatic 677 76 thion-esters (R*CS*OR’) can be used in place of dithio-esters, and they are easily prepared by the action of hydrogen sulphide oniminoethers.7’ Thiobenzoyl chloride 78 (Ph-CSCI), prepared by the actionof oxalyl chloride or thionyl chloride on dithiobenzoic acid, is reported toreact vigorously with aniline to give thi~benzanilide,~~ but does not reactin the expected way with amino-a~ids.~~ Attempts to prepare thiophenyl-acetyl chloride (Ph*CH,*CSCI) were unsuccessful ; oxalyl chloride andpotassium dithiophenylacetate yielded 4 : 5-diketo-2-benzylidene- 1 : 3-di-thiolan (XXXIV) .Attempts to prepare the potentially-useful thiophenyl-Ph*CH:C<S,CO S-YO Ph*CH,*wN-N>C*CH,PhN=N(XXXIV.) (XXXV.)acetyl azide (Ph*CH,*CS*N,) were also unsuccessful, since the dithio-aciddid not react with sodium azide, while methyl dithiophenylacetate andhydrazine gave 3 : 6-dibenzyl-1 : 2 : 4 : 5-tetrazine (XXXV) and otherproducts instead of the desired hydrazide.80 Thiobenzhydrazide has been69 W.Baker and J. F. W. McOmie, unpublished results; J. F. W. McOmie, D.Phil.Thesis, Oxford, 1946.70 A. R. Todd and A. Topham, CPS 93; J. Wardleworth, A. R. Todd, P. Sykes,5. Baddiley, and H. T. Openshaw, CPS 351; R. Bentley, J. R. Catch, A. H. Cook,(Sir) I. M. Heilbron, and G. Shaw, CPS 267, 328; E. P. Abraham, W. Baker, E. Chain,and (Sir) R. Robinson, CPS 342; Lilly Research Laboratories, CPS 286, 364; W. E.Bachmann, CPS 335, 358 ; Squibb Institute for Medical Research, CPS 452.71 A. H. Cook, J. A. Elvidge, and (Sir) I. M. Heilbron, CPS 273.72 Squibb Institute for Medical Research, CPS 278.73 B. Holmberg, “ The Svedberg Memorial Volume,” p. 299, Stockholm, 1944;74 T. G. Levi, Cazzetta, 1924, 54, 395.7 5 J. Houben, Ber., 1906, 39, 3227.78 (a) E.P. Abraham, E. Chain, W. Baker, and (Sir) R. Robinson, B.P. 588,101,1947; (b) A. A. Goldberg and W. Kelly, J., 1948, 1919.7 7 Y. Sakurada, Mem. Coll. Sci. Kyoto, 1926, 9, 237.78 H. Staudinger and J. Siegwart, Helv. Chim. Acta, 1920, 3, 824.7s Squibb Institute for Medical Research, CPS 301.B0 A. H. Cook, J. A, Elvidge, and (Sir) I. M. Heilbron, CPS 328.Arkic Kemi Min. Cr’eol., 1944, 17 A, 1 ; D. F. Elliot, Nature, 1948, 162, 658MCOMIE : ORGANIC SULPHUR COMPOUNDS. 209prepared by B. Holmberg,73 but its reaction with nitrous acid has not beentried.218 ORGANIC OHEMISTRY.presence of water,16 aluminiun chloride,17 or aqueous mineral acid,ls butnone of the methods is suitable for preparing the anhydrous diamines.L. B. Clapp finds that the reaction proceeds well under anhydrous con-ditions at 100” with ammonium chloride as catalyst.In this manner2-ethylethyleneimine with liquid ammonia (40-fold excess) gives 1 : 2-butylenediamine, in 55 yo yield, and very little polymer, while with variousalkyl- and cycloalkyl-amines (in %fold excess) 2-amino-l-alkylamino-n-butanes (R = Et, R’ = H) are formed. 2-Amino-l-alkylaminoisobutanes(R = R’ = Me) arise from 2 : 2-dimethylethyleneimine :RR’C-CH, I I RR’C-CH, RtiR’tjNgNH, NR”R”’\ / - NHRing opening therefore occurs preferentially at the primary carbon. How-ever, in the case of reactions with aniline appreciable amounts (9--22%)of the alternative ring-scission products are also formed.Only a few examples of the above type of reaction (substituted-aziridinering-scission) have previously been recorded.Catalytic hydrogena’tion 2oand reaction with hydrogen bromide21 proceed with ring opening a t theprimary carbon, whereas hydrolysis lo occurs with scission at the tertiarycarbon rather than at the primary (or secondary,13 see above) :RR’$F--CH,*NH,OHRR‘C--CH,\ /NH % RR’$F--CH,X(X = H or Br)NH2When a mixture of ethyleneimine and acetic acid a t -78” warms toroom temperature 2-acetoxyethylammonium acetate (V) is formed,22 whichHOAc VHz>NH ---+ AcO*[CH2],*NH,,HOAc (V.)CH2l6 U.S.P. 2,318,729.l7 A. L. Coleman and J. E. Callen, J . Amer. Chem. Soc., 1946, 68, 2006.l8 G. I. Braz amd V. A. Skorodumov, Compt. rend. Acad. Sci. U.R.S.S., 1947, 55,l9 J . Arner.Chem. SOC., 1948, ‘SO, 184.2o J. V. Karabinos and K. T. Serijan, ibid., 1945, 67, 1856; K. N. Campbell, A. H.Sommers, and B. K. Campbell, ibid., 1946, 68, 140.21 S. Gabriel and H. Ohle, Ber., 1917, 50, 804.22 G. D. Jones, J. Zomlefer, and K. Hawkins, J . Org. Chem., 1944, 9, 500.315ELVIDGE : HETEROCYCLIC COMPOUNDS. 219on heating in an open vessel cf- 23 loses acetic acid and water and cyclises to2-methyl-A2-oxazoline (VI) ; in a closed system re-arrangement to 2-acet-amido ethanol (VII) occurs.An example of possible further applications of ethyleneimines in syntheticwork is the preparation by H. Gilman et aZ.24 of ethyleneiminyl-lithium(VIII) which was employed for the synthesis of a quinoline derivative (IX)as follows :(VIII.)A new class of aziridine derivatives has been discovered by N.H. Crom-well and his co-workers 25 who observed that interaction of phenyl a-bromo-styryl ketone (X) and phenyl 1 : 2-dibromo-2-phenylethyl ketone (XI)with benzylamine and cyclohexylamine afforded colourless products. Fromultra-violet absorption data 26 and chemical evidence these products wereconcluded27 to be ethyleneimine ketones and not compounds of the typePh*CO-CH:CPh*NHR as had earlier been suggested : 28, cf*Ph-CH:CBr COPh* Ph*CH-CH-COPhNR(X.) 1\ /Ph*[CHBr],*COPh(XI-)Ph*vH--vH*COPh Ph*vH*CHCl*COPhPh*CH,-NH,+ SO,- R*NH,HCl(XIV.) (XIII. )Treatment 25 of the products (XII) with hydrogen chloride resulted in theuptake of two molecules of the acid and formation of chloro-amine salts23 H.Wenker, J. Arner. Chem. SOC., 1935, 57, 1079.24 H. Gilman, N. N. Crounse, S. P. Mwsie, junr., R. A. Benkeser, and S. M. Spatz,25 N. H. Cromwell, R. D. Babson, and C. E. Harris, ibid., 1943, 65, 312.26 N. H. Cromwell and R. S. Johnson, ibid., p. 316.2' N. H. Cromwell and J. A. Caughlan, ibid., 1945, 67, 2235.2 8 S. Ruhemann and E. R. Watson, J., 1904, 1181.29 J. Agar, A. Hickey, and P. G. Sherry, Proc. Roy. I&h Acad., 1943, 49, B, 109.ibid., 1945, 67, 2106220 ORGANIC CHEMISTRY.(XIII). In dry ether, the hydrochloride of (XII; R = CH,Ph) could alsobe obtained. With sulphuric acid, (XII; R = CH,Ph) yielded a high-melting, sparingly soluble substance which, in view of its ready recon-version into (XII; R = CH2Ph) by treatment with alkali, was formulatedas the internal salt (XIV).Methylamine was also found 27 to condense with (XI) to form an imine(XII; R = Me), the aziridine structure of which was well substantiated byits reaction with hydrogen chloride to yield (XV), ring opening havingoccurred in opposite sense to that previously observed :Additional evidence in support of the aziridine structures of theseimino-ketones has recently been obtained.30 Thus the products (XVI ;R = Ph, R‘ = p-Me*C,H,) and (XVI; R = p-Me*C,H,, R’ = Ph), derivedfrom benzylamine and the appropriate ap-dibromo-ketones in ethanol at>40”, were found to react merely as normal ketones with Grignard reagents,affording the corresponding carbinols :(XVI.)Possible open-chain structures are consequently ruled out, and incidentallythe behaviour indicates that the ethyleneimine ring is stabilised by theintroduction of an alkyl substituent in the l-position.p-Biotin.Stereochemicd Studies.-In the previous Report on biotin 31 evidenceconcerning the configurations of the four racemates of (XVII), ( A)-,(&)-epi-, (f)-epiaZlo-, and ( f)-allo-biotin, was reviewed.This evidencesuggested that the configurations about the C3-C4 linkage were trans in( f)-allo- and (&)-epiaZZobiotin, whereas in (&)-biotinthe C,-C, configuration was cis. More recently cothese conclusions have been rigidly verified by NH NHB. R. Baker and his associates. These workers have 1-1 ;CH,14.co2H elaborated methods of obtaining singly any of the\5/2 four racemates of (XVII), starting from only one1 (xvll.) intermediate (XXI; R = [CH,],*CO,H) : by chemicalcontrol of isomers the need to effect separations by fractional crystallisationwas obviated.Besides providing new syntheses of (&)-biotin and (&)-/ \so N. H. Cromwell, J . Amer. Chem. S O ~ . , 1947, 69, 258.31 Ann. Reports, 1946, 43, 239ELVIDOE : HETEROCYCLIC COMPOUNDS. 221epiaZZobiotin the procedures have enabled the hitherto unknown ( &)-epibiotinto be obtained. Though perfectly feasible, the synthesis of (&)-aZZobiotinfrom (XXI; R = [CH,],*CO,H) was not attempted. The work may con-veniently be considered under two headings : ( A ) preparation of inter-mediates ; (B) synthesis of biotin isomers.( A ) Dieckmanii cyclisation of esters (XVIII; R’ = C0,Et or C0,Me)prepared from ethyl thioglycollate and a@-unsaturated esters gave mainly4-ketothiophan-3-carbovylates (XIX ; R’ = C0,Et or C0,Me) rather thanthe 5-isomers (XX).321 33 This was shown by the fact that the derivedacids (XIX ; R’ = C0,H) were identical with corresponding acids obtainedvia the unambiguous cyclisation of the appropriate cyanides (XVIII ;R’ = CN) to (XIX ; R’ = CN). Preparation of the 2-substituted thiophan-3 : 4-dicarboxylic acids (XXI) was completed along conventional lines :The double bond in (XXII) was shown to be in the 3.: 4-position by theobservation that the same product could be derived from the structurallyestablished intermediate (XXIII). The synthesis of (XXI ; R =[CH,I4*CO,H) by the above route proved highly unsatisfactory, however,since in this case (XX; R = [CH&*CO,Me) (actually the 5-methyl ester)was the main product from the Dieckmann cy~lisation:~~ Attempts toconvert the compound (XXI; R = [CH,],*OPh) via the bromide (R =32 B.R. Baker, M. V. Querry, S. R. Safi, and S. Bernstein, J. Org. Chem., 1947, 12,33 G. B. Brown, B. R. Baker, S. Bernstein, and S. R. Safir, ibid., p. 155.34 G. B. Brown, M. D. Armstrong, A. W. Moyer, W. P. Anslow, junr.,.B. R. Baker,138.M. V. Querry, S. Bernstein, and S. R. Safir, ibid., p. 160222 ORQANIC CHEMISTRY.[CH,],*Br) into (XXI; R = [CH,],*CO,H) also failed, as did attemptsfrom the corresponding urethane, for the reasons indicated : 32(XXI; R = [CH,],*OPh) -> HBr ';;l-F..Curtius Br- \CH,* H,2Br- .1Et0,Q Q02EtNH NH N - , N+K, $H3 dH2-vH2-----+ + t--= (7%\ s r C H 2HBr H k- *[CH213*oph \ s r C H 23Br-+ \CH,*bH,However, synthesis of (XXI; R = [CH,],*CO,H) was accomplished insatisfactory overall yield starting from pimelic acid, the derived triester(XXIV) on Dieckmann cyclisation giving the required intermediate (XIX ;R = [CH,],*CO,H, R' = CO,Me), and not the cyclohexanone (XXV).35The acids (XXI) as initially obtained were mixtures of isomers, but ineach of the cases (R = Pry [CH,],*OPh, [CH,],*CO,H) a single racematehaving a tra?zs-C,-C, configuration readily crystallised from the mix-t ~ r e , ~ , , ,,, 35 and the residue could be made to yield more of the sameracemate by esterification and treatment with methanolic sodium methoxide(to invert the corresponding cis-racemate) .329 35 The trans-configurationof each acid was shown by stability to heat, formation of di-acid chlorides,etc.Conversion into the cis-isomer was achieved by heating under refluxwith acetic or propionic anhydride, followed by 339 36 Thetrans-di-acids (XXI) were converted via the dimethyl esters and hydrazides,and using the Curtius reaction, into the trans-diamines (XXVI). Thecis-esters, however, also afforded the trans-di-acid hydrazides so that thecis-diamines could not be obtained, by this route, or by other routes involv-ing simultaneous degradation of both carb0xyls.~~9 33 Inversion on form-ation of acid hydrazides or azides had not previously been experiencedwith carbocyclic compounds, but the unusual behaviour was also shownby penthian derivative^.^'(B) It became apparent that, in order to preserve the configurations atC3 and C,, methods for the stepwise degradation of the carboxyl groupswould have to be used.The procedures were developed using first theunsubstituted di-&id (XXI; R = €€).36Application of these procedures to the intermediate (XXI; R =35 B. R. Baker, M. V. Querry, S. Bernstein, 8. R. Safir, and Y . SubbaRow, J . Org.36 B. R. Baker, M. V. Querry, S. F. Safir, W. L. McEwen, and S. Bernstein, ibid.,37 B. R. Baker and F. Ablondi, ibid., p. 328.Chem., 1947,12, 167.p. 174ELVIDGE : HETEROCYCLIC COMPOUNDS. 223[CH,],*CO,H) led to (&)-biotin and (-~-)-epiaZlobiotin.~~ The chemical andstereochemical configurations of the intermediates were checked a t criticalc + t- (XXI; - R = H)(XXVII.)vO*NHPh J3°2Me (XXVIII.)/*\co co CO*NHPh C/O*NHPhNH NH*CO,Et NH CO*NH*NH2R = H) + EbOH I MeOHJ.(XXIX.)c- (XXVII) Ph*NH*CO CO\/ \c- (XXVI; '- (xxvlll) '- (xxlx) -2 ,."-r* R = H) + EtOH[ t = trans, c = cis, (i) = +COCl_3 CON, +NCO, (ii) = +CON,+ NCO.]stages, e.g., a t (XXX), by chemical methods. It will be observed thatinversion a t C, was effected during stage (XXXI) -+ (XXXII) and againduring stage (XXXII) + (XXXIII) so that the C,-acid, correspondingto the ester (XXXIII), was identical with (XXXI).Of the known biotin racemates, ( f)-allo- and (A)-epiallo-biotin bothgive microbiologically inactive dethiobiotin (XXXIV) and must con-sequently have the same configurations about C3-C,.(&-)-Biotin givesrise to an active dethiobiotin and ( A)-epibiotin would also afford the sameisomer. The melting point of the trans-product t-(XVII) corresponded tothat of the (&)-epiallobiotin of S. A. Harris et al.,39 and was identified withcertainty by Raney-nickel desulphurisation to a biologically inactive pro-duct : thus t-(XVII) could not have been the unknown (&)-epibiotin. Atrans-configuration for the C3-C, link in (&)-epiaZZo- and ( f)-allobiotinwas thus rigidly establi~hed.~~ The cis-product c-(XVII) had 60% of theactivity of natural biotin in assay against L. arabinosus, and resolution withL-arginine yielded (+)-biotin, identical with the natural vitamin.Itfollowed that the remaining, unknown racemate, (&)-epibiotin, was also acis-isomer, necessarily epimeric with ( &)-biotin a t C,.and Y . SubbaRow, J . Org. Chem., 1947, 12, 186.J . Amw. Chem. SOC., 1944, 66, 1800.38 B. R. Baker, M. V. Querry, W. L. McEwen, S. Bernstein, S. R. Safir, L. Dorfman,39 S. A. Harris, R. Mozingo, D. E. Wolf, A. N. Wilson, G. E. Arth, and K. Folkers224 ORGANIC CHEMISTRY.Bemuse the synthetic methods enabled the configurations a t C, and C,to be inverted selectively, a synthesis of ( f)-epibiotin became possible.4o Itt- (XXI)YO*NHPh vO*NHPhNH C0,Me NH C0,H.1Me0,C C0,Me H0,C C0,Met- t - R l NaOH t- dR,l *t- dR,f 3 t- d.(XXX.) (XXXI.)k % OYOONHPh YO-NHPh VO-NHPh Y O-NPhNH NH*C02Me NH CO*NH*NH, NH C02Me NH bo(ii) MeOHt- ‘ d R t- U R f<NH,x t- F I R + = c- b { RI (XXXII.) (XXXIII.)\S/+t- (XXVI.) cooll,coNH NH/ \Me l - i CH,Rc- (XVII.) YO-NPh(*)-biotin NH 60 + Ic- (XXVI.)t- (XVII.)(&)-epiaZbbiotin(XXXVII .)YO-NHPhPh*NH*CO COC-was apparent that the trans-racemate (XXI; R = [CH,],*CO,H) could haveeither of the total configurations (XXIa) or (XXIb). By inversion at C,and C, as indicated, both of the possible cis-C,-C, racemates of biotin,(XVIIa) and (XVIIb), were obtainable irrespective of whether the actualconfiguration of the starting material was (XXIa) or (XXIb). Now in theprevious reaction sequence a cis-configuration had been obtained by inver-sion a t C, during the stage (XXXI) + (XXXII).If instead, a cis-40 B. R. Baker, W. L. McEwen, and W. N. Kinley, J . Org. Chem., 1947,12, 322ELVIDGE : HETEROUYCLIC COMPOUNDS. 226configuration were obtained by inversion at C,, a t a suitable stage, the endproduct would, as already indicated, then be epimeric with the product(XXIU.)c,+ - invert c, f- - iriver t c, + - c,+- -- c, (3,-+ -y--+ c3-+c4+ - c4+ - c4 - +(XVIICC.) (XVIIb . )'2 - + invert '2 - + invert c2 - + c3-+ - c3-3- -- c,+-c, - + c4+ - c4+ - C' c*(XXIb.)previously given, i.e., it would be (f)-epibiotin. The correctness of thisreasoning was demonstrated by the synthesis of ( 5)-epibiotion c'- (XVII)as follows, the structures of intermediates being proved where necessary :t- (XXX) THPh YHPh(XL.) (XXXVIII.)c'- (XXVI) c'-(XVII)( f )-epibiotin(R, R', R" as previously.)It is to be noted that the reagents used for effecting the stage (XXXV) -->(XXXVI), vix.acetic anhydride-sodium acetate, caused inversion a t C4but not a t C,. Already established was the fact that in the Curtius re-arrangement of an azide to an isocyanate no inversion occurred except whentwo adjacent carboxyls, attached to the thiophan (or penthian) nucleus,were degraded simultaneously. The need to convert the C, side chaincarboxyl into anilide as a t (XXXVII) 38 and (XXXVIII) 40 was occasionedby the fact that otherwise partial inversion took place a t C3 and C4, re-spectively, during the subsequent conversions into the acid hydrazides(XXXIX) and (XL). The primary effect of introducing the anilide groupwas to reduce very markedly the solubility of these compounds in thereaction medium.REP.-VOL. XLV, 226 ORGANIC UEPMISTRY.To reiterate briefly, it is apparent that a configurational change can beeffected selectively a t nuclear substituents as follows : (a) a t a carboxylby boiling acetic or propionic anhydrides; ( b ) a t an ester group by boilingalcohol containing a trace of sodium alkoxide; ( c ) a t an anilide group byboiling acetic anhydride containing sodium acetate.Peculiar to thethiophan (and penthian) systems, and directionally uncontrollable, are (d)the inversion a t one of two adjacent nuclear carboxyls during their simul-taneous Curtius degradation via acid chlorides and sodium azide, and (e)the inversion a t one of two adjacent ester groups during their conversionwith hydrazine into acid hydrazides.Pterins.Since work in this field was previously reviewed*l progress has becnmade inseveral directions : new syntheses of vitamin B, (pteroylglutamic acid) and ofpteroic acid have been devised ; the structure of the fermentation Lactobacillusm e i factor has virtually been established; a new microbial growth factor,rhizopterin, has been isolated, identified, and synthesised ; some experi-ments concerning reduction of the pteridine nucleus are reported; and acontinued interest has been shown in the preparation of pteridine deriv-atives generally. .Results previously announced in a preliminary manner,e.g., the deduction of the structure and the synthesis of the liver L.caseifactor (vitamin B,) have now been established by the publication of fullchemical details.42, 43, 449 45, 46 Details of the method of isolation andpurification of this factor have also been disclosed.47 The precise natureof the anti-anEmia factor, " folic acid ", originally obtained from vegetablesources by H. K. Mitchell and his co-workers 48 remaiiis undeterminedthough evidence increasingly points to its being a mixture of closely relatedcompounds with differing physiological activity.49 On the other hand, a" norite eluate factor " isolated from liver and yeast in 1940,50 is, according41 Ann. Reports, 1946, 43, 250.* A suggestion (M. Gordon, J. M. Ravel, R. E. Eakin, and W. Shive, J . Amer. Chern.Soc., 1948, 70, 878) that pteroyl derivatives function as carriers of formate (cf.thestructure of rhizopterin) in the biosynthesis of purines, possibly being involved in theinsertion of a single carbon unit into the pyrimidine ring, must be regarded with extremecaution since the evidence adduced in its support consists of microbiological experimentswith this “methylfolic acid.” Later work (W. Shive, J. M. Ravel, and R. E. Eakin,ibid., p. 2614; W. Shive, J. M. Ravel, and W. M. Harding, J . Biol. Chem., 1948, 176,991) seems to afford no substantiation234 ORGANIC CHEMISTRY.NH,, R, = OH), and 2 : 6-diamino- (LVII; R, = R, = NH,) derivativeswith glyoxal and diacetyl, and in addition, by reaction of the latter pyrim-idine with benzil, phenanthraquinone, and acenaphthaquinone, haveprepared a series of symmetrically 8 : 9-disubstituted pteridines :W.Steinbuch 77 condensed (LVII; R, = R, = NH,) with mesoxalic esterand saponified the product to obtain 6-aminoisoxanthopterincarboxylicacid (LVI; R, = R, = NH,, R, = CO,H, R, = OH). A similar reactionwith (LVII; R, = R, = OH) provided the hitherto inadequately describeddeaminoisoxanthopterincarboxylic acid (LVI ; R, = R, = R, = OH ; R, =C0,H). It is to be noted incidentally that the nomenclature of some ofthese products could be improved. Diacetyl, phenanthraquinone, andacenaphthaquinone have also been condensed with 4 : 5-diamino-6-hydroxy-2-ethylthiopyrimidine (LVII; R, = SEt, R, = OH) to give three new2-ethylthiopteridines (LVI; R, = SEt, R, = OH). 6 : 8 : 9-Trihydroxy-2-mercaptopteridine was prepared by G.B. Elion et from the mercapto-pyrimidine (LVII; R, = SHY R, = OH) and oxalic acid. These workersfound also that, contrary to the results of 0. I ~ l a y , ~ ~ reduction of 2-chloro-5-nitro-4-aminopyrimidine can be effected by an excess of alcoholic potassiumhydrogen sulphide to yield 4 : 5-diamino-2-mercaptopyrimidine (LVII ;R, = SH, R, = H). Condensation of the latter with glyoxal then afforded2-mercaptopteridine (LVI; R, = SHY R, = R, = R, = H) itself.A number of aminohydroxypteridine mono- and di-carboxylic acids andmethyl esters (LVI; R, = OH or NH,, R, = OH or NH,, R, = H, CO,Me,or CO,H, R, = COzMe or C0,H) have been prepared by C. K. Cain andhis collaborators 8O for testing as microbial growth factors and for studieson growth and hsmoglobin formation in chicks. The acids were obtainedfrom the corresponding (known) 9-methyl- (R, = H, R, = Me) and 8 : 9-di-methyl- (R, = R, = Me) pteridines by oxidation in alkaline solution withpotassium permanganate, and were converted into the methyl esters withmethanolic hydrogen chloride.It was observed that a carboxyl group inthe 8-position is less stable than in the 9-position of the pteridine nucleus,for on heating the 8 : 9-dicarboxylic acid (LVI; R, = R, = OH, R, = R, =C0,H) in quinoline monodecarboxylation to the 9-carboxylic acid (LVI ;R, = R, = OH, R, = H, R, = C0,H) took place.The potentialities of sugars and related compounds for the synthesis ofpteridines from 4 : 5-diaminopyrimidines have been investigated by P.Karrer7 7 Helv. Chim. Acta, 1948, 31, 2051.78 G. B. Elion and G. H. Hitchings, J . Amer. Chem. SOC., 1947, 69, 2553.79 Ber., 1906, 39, 250.80 C. K. Cain, M. F. Mallette, and E. C. Taylor, junr., J. Amer. Chem. SOC., 1948,'SO, 3026ELVIDGE : HETEROCYULIC COMPOUNDS. 235and his co-workers.81 By performing the reaction under carbon dioxide inboiling water containing a little acetic acid, pteridines were prepared from2 : 4 : 6-triamino-6-hydroxypyrimidine (XLI) and the aldoses, arabinose,xylose, glucose, galactose, and glyceraldehyde. The configuration of theproducts was not a t first rigidly established, but it was suggested that theywere probably 9-hydroxyalkylpteridines (LVIII) (e.g., R = CH2*OH in thecase of reaction with glyceraldehyde) formed in the following way :The ketose, fructose, gave a different product from glucose, so that, onthe preceding ideas, it was formulated as the isomeric 2-amino-6-hydroxy-8-~-arabotetrahydroxybutylpteridine (XLV ; R = [CH*OH],*CH,*OH).Thespectra of the pteridines from glucose and glyceraldehyde were closelysimilar to one another, but differed from the spectra of the products fromfructose and dihydroxyacetone.H. G. Petering and D. I. Weisblat 82 in a preliminary report stated, inagreement, that under Karrer's conditions D-glucose reacts with (XLI) to formthe 9-D-arabotetrahydroxybutylpteridine (LVIII ; R = [CH*OH],*CH,*OH),whereas D-glucosone a t pH 5-0 yields mainly the %isomer. These authorsfound in addition that in strongly acid solution both these two reactionsproceeded in the opposite senses since mixtures Were obtained richer in theisomer of the product previously isolated (h,, glucose gave the %isomer ;glucosone the 9-isomer).Following up their initial experiments, Karrer and Schwyzer 56 sub-stantiated their ideas that aldoses condensed with the pyrimidine (XLI)to yield 9-substituted pteridines, whereas ketoses gave the 8-isomers.Thusit was shown thet the product from (XLI) and dihydroxyacetone [evidently(XLV ; R = CH,*OH)] when condensed with p-aminobenzoyl-L-glutamicacid afforded a product containing 15% of vitamin B, (an 8-substitutedpteridine). A similar reaction with the 9-hydroxymethylpteridine derivedfrom (XLI) and glyceraldehyde, on the other hand, gave rise to no micro-biological activity, though a modification in which (XLI) , p-aminobenzoyl-L-glutamic acid, and glyceraldehyde ditoluene-p-sulphonate were condensedin the presence of potassium iodide led to a product containing 6% of thevitamin.81 F.Karrer, R. Schwyzer, B. Erden, and A. Siegwgrt, Helu. Chim. Acta, 1047,@a J. Amer. Chem. Soo., 1947, 09, 3566.30, 1031236 ORQANIC CHEMISTRY.Independently of Karrer et al., H. S. Forrest and J. Walker 83 hadtreated D-glucose and D-fructose with 2 : 4 : 5-triamino-6-hydroxypyrimidineand obtained from each condensation the same product, which they sug-gested was probably 2-amino-6- hydroxy- 8-D -arabotetrahydroxybutylpter-idine (XLV ; R = [CH*OH],*CH,*OH).However, P. Karrer and R.Schwyzer 84 pointed out that the reaction conditions involved the presenceof phenylhydrazine. Since glucose and fructose both give the same phenyl-osazone, it was to be expected that under such conditions the same pteridinewould arise from each sugar. There was no doubt that in the absence ofphenylhydrazine different products were formed.Since in the structural determination of the liver L. casei factor the 8- and9-methyl- and -carboxy-pteridines, (XLV ; R = Me and C0,H) and (LVIII ;R = Me and CO,H), were prepared and characterised unambiguously,44 itwill in future be a simple matter to orientate pteridine products whicharise from ambiguous condensations. Thus Forrest and Walker 57 haveshown that the product from reaction of 2 : 4 : 5-triamino-6-hydroxy-pyrimidine (XLI) with reductone and methyl p-aminobenzoate is mainlypteroic ester (an 8-substituted pteridine) since aerobic alkaline hydrolysisyielded 2-amino-6-hydroxypteridine-8-carboxylic acid (XLV ; R = C0,H).Independently, H. J.Backer and A. C. Houtmann 85 had suggested thatthe reaction between reductone and the same pyrimidine (XLI) producedthe 9-hydroxymethylpteridine (LVIII; R = CH,*OH) by analogy with thebehaviour of methylglyoxal. Clearly though, analogies are unreliable inthis field.The condensation of methylglyoxal with 4 : 5-diamino-2 : 6-dihydroxy-pyrimidine (LIX) had earlier been shown by J. Weijlard et aZ.86 to yieldonly 2 : 6-dihydroxy-9-methylpteridine (LX), since degradation gave noneof the known 2-amino-5-methylpyrazine.Subsequently, Cain et aLso haveconfirmed the structure (LX) in a more positive manner by degradingthe substance to the known 2-amino-3-carboxy-6-methylpyrazine (LXI) :MeGO.CHOI____, 4 H02C/NN NH2(flh!le(LXI.)In this connection it is of interest that pyruvic acid condenses with 2 : 4 : 5-triamino-6-hydroxypyrimidine to yield both of the theoretically possibleproducts : 78OH OH83 Nature, 1948; 161, 308.86 Rec. Truv. chim., 1948, 67, 260.86 J. Weijlard, M. Tishler, and A. E. Erickson, J . Arne?. Chem. Soc., 1945, 67, 802.a4 Helv. Chim. Acta, 1948, 31, 782ELVIDQE : HETEROCYCLIC COMPOUNDS. 237In boiling B~-sulphuric acid 9-methylxanthopterin (LXII) is formed, whilstin dilute acetic acid a mixture results, containing 8-methylisoxanthopterin(LXIII). This behaviour is comparable with that of glucosone (see earlier).Recently, the variously-reported preparations of hydroxymethylpteridineshave been questioned. According to R. B. Angier and his co-workers *'the condensations in 6~-hydrochloric acid of glyceraldehyde, s-dichloro-acetone, 2 : 3-dichloropropaldehyde (cf. reaction of the dibromo-compounda t pH 4), and a-bromotetronic acid (presumed to hydrolyse to 1-bromo-3-hydroxypropan-2-one) with 2 : 4 : 5-triamino-6-hydroxypyrimidine(XLI) all lead to 2-amino-6-hydroxy-9-methylpteridine (LXIV), and notto the expected 9-hydroxymethyl or -halogenomethyl compounds (LXV).Similarly ethyl ccy-dibromoacetoacetate reacts with (XLI) to give, aftertreatment with alkali, 2-am'ino-6-hydroxypteridine-8-acetic acid (XLV ;R = CH,*CO,H) instead of a bromo- or hydroxy-acetic acid derivative.It is suggested that intermediately-formed dihydropteridines such as (LXVI)aromatise by loss of the elements of water or hydrogen halide (as the casemay be) rather than by dehydrogenation :CHO(where X = C1 or OH)N/\/" OH /No;-Hx // /N\(Lxv.) NHJ\ ,,!I bH,X N I I , ! N / \ N p II I (LXIV.)N \N/There is, however, the possibility that under certain conditions oxidation ofthe dihydro-intermediate (LXVI) might take place preferentially, and asubstituted-methylpteridine (LXV) would then result. Further work isobviously required : especially is it desirable that better analyses beobtained for the alleged hydroxymethyl compounds. In view of the furtherreactions which have been achieved with these products there can be littledoubt that they are not pure methylpteridines, but their precise nature,e.g., whether they are mixtures containing (LXV) and/or (LXVI), is inneed of clarification.J. A. E.J. A. ELVIDGE. J. D. LOUDON.S. H. HARPER. J. F. W. MCOMIE.D. H. HEY. B. C. SAUNDERS.B. JONES.*' R. B. Angier, C. W. Waller, J. H. Boothe, J. H. Mowat, J. Semb, B. L. Hutchings,E. L. R. Stokstd, and Y. SubbaRow, J. Amr. Chem. SOC., 1948, 70, 3029