ORGANIC CHEMISTRY.1. INTRODUCTION.THE section on General Methods deals with ion exchange resins, chromato-graphy, reduction, and phosphorylation.Within the last twenty years ultra-violet absorption spectrometry hasbecome a valuable tool for purposes of both qualitative and quantitativeanalysis, as well as in physicochemical studies, such as the determination ofequilibria and reaction rates. Purely structural applications have, however,remained perhaps the most important and have played an invaluable partin many recent investigations on natural and synthetic products, includingvitamins, hormones, and plant pigments. Some of the more theoretical aspectsof the subject have been dealt with in the physicochemical sections of pre-vious Reports, but no comprehensive summary has been hitherto availableof the well-defined empirical relationships between the ultra-violet lightabsorption and the constitution of organic compounds.It is hoped, there-fore, that the present survey will supply a real need. Experimental methodsof determining extinction curves, and in the interpretation of the results,are first brieff y discussed. Selective absorption in the region easily accessiblewith quartz instruments (> 2000 A.) depends on the presence of atoms orgroups containing " mobile " electrons, i.e., electrons of comparativelylow excitation energies. The most important type of mobile electrons metwith in organic compounds is the unsaturation electrons of multiple bonds, andthe vague classical concepts of chromophore and auxochrome are convenientlyre-defined to refer respectively to covalently unsaturated and covalentlysaturated groups.Single chromophores and auxochromes usually give riseto bands in the Schumann region or on the edge of the quartz ultra-violet,but the interaction between two or more of such groups generally results ina displacement of absorption towards higher wave-lengths. Conjugationis strongest when the groups are in vicinal positions, and ultra-violet lightabsorption is still most useful when dealing with compounds containingclassical conjugated systems. The wave-length positions and intensitiesof the maxima depend on the nature of the chromophores and auxochromespresent and on their number, increasing with the latter and eventuallyresulting in visible colour.The more important empirical relationshipsare thus outlined, and their interpretation in terms of electronic mobilityand their correlation with chemical reactivity are indicated. Sufficientdata are now available to make ultra-violet light absorption a convenientcriterion of identity for mahy classes of organic compounds, to allow thereliable prediction of spectral properties, and to judge the compatibility ofthose observed with assigned structures.Recent developments in knowledge of the characteristic reactionsof free neutral radicals have led to a great clarification of outlook in regardto mechanisms of oxidation, and it is now thought that most, though prob-ably not all, organic oxidation processes are chain reactions in which transienINTRODUCTION.93uncharged radicals pa,rticipate. The autoxidation of hydrocarbons has beenstudied intensively by many workers during the past decade in view of itstechnical importance in connection with the chemistry of fats, and with therubber, paint, and varnish industries. Unconjugated olefins, and the sidechains of aromatic hydrocarbons, are, at moderate temperatures, undoubtedlyattacked first a t active methylene groups to give hydroperosides,R”CH*O*OH, several of which have been isolated by R. Criegee, H. Hock,E. H. Farmer, and others. In the case of olefins the initial point of attackis the methylene group adjacent to the double bond. These hydroperoxidessubsequently break down to other products, and in so doing catalyse theattack of oxygen on the original hydrocarbon, thus rendering the wholeautoxidation process autocatalytic.Strong evidence has been forthcomingto indicate that the initiating step in the autoxidation of hydrocarbons is ahomolytic dehydrogenation by a neutral radical, or equivalent catalyst,acting thus :Re + R”CH, + R-H + R”CH*giving a hydrocarbon radical, R”CH-, which then combines with oxygenby the repetitive chain process :R’’CH0 + 0, --+ R”CHaO*O* ‘IR”CH.00 + R”CH, -+ R”CH*O*OH + R”CH*It has been suggested that secondary radicals, such as *OH and R”CH*O*,from hydroperoxide decomposition may function as initiating catalysts(Re). Inhibitors of autoxidation are now thought to be substances whichcombine rapidly with free radicals, for it has been shown that they are notnecessarily substances which rapidly destroy hydroperoxide molecules.Conjugated diolefins, in contrast, seem to add on oxygen in the 1 : 4 positions,giving cyclic peroxides, which, however, can function as catalysts for otherautoxidations.T. P. Hilditch and his colleagues have shown, however,that at elevated temperatures the primary attack of oxygen on an olefinmay be upon the double bond, and this is possibly a hydroxylation ratherthan a peroxidation process. The oxidising reactions of lead tetra-acetatecan be explained consistently as reactions of the two radicals, acetate,Me*CO*O*, and methyl, *Me, which are formed successively by the thermaldecompositions :Pb(O*CO*Me)4 + Pb(O*CO*Me), + 2*O*CO*Me*O*CO*Me -+ CO, + *MeBoth these radicals have dehydrogenating properties.The more activeradical, methyl, is believed to be concerned in oxidations employing leadtetra-acetate in hot solutions, though only the acetate radical may be con-cerned in the selective glycol fission process discovered by R. Criegee.Hydrogen peroxide has been shown to yield free neutral hydroxyl radicals,both by photochemical decomposition and by “ reduction activation ”by a single electron donor, such as a ferrous salt : Fe++ + HOoOH --+Fe+++ + HO. + (:OH)-. The consequent oxidation reactions of hydrogenperoxide are thus hydroxylationa due to the hydroxyl radical, which, a94 ORGANIC CHEMISTRY.shown by J. H. Baxendale, M. G. Evans, and G. S. Park, will add on to olefins,and SO may bring about their chain polymerisation.Fenton’s reaction-the oxidation of a-hydroxy-acids with hydrogen peroxide and a ferroussalt-as well as many typical reactions of the inorganic and organic per-acids can be explained easily as reactions of free hydroxyl radicals. Manyanalogues of these processes can be traced in the organic chemistry of sul-phur. Several quinones have been shown to undergo reversible reductionin two stages, and may give, by gain of one electron, reactive “ semi-quin-ones ” which can play important r6les as “ potential mediators ” in manyother oxidation processes. On these grounds both P. A. Schaffer andL. Michaelis have suggested that the oxidations of both organic and inorganiccompounds can proceed a t measurable speeds only in steps of one electroneach.This generalisation accords with the finding of W. A. Waters thatmany inorganic oxidising agents, such as chromic acid, potassiumpermanganate, and periodic acid, can act upon organic compounds by thehomolytic dehydrogenating mechanismOx* + H*R + 0x.H + *Rbut should as yet be viewed with caution, since some oxidations, e.g., thoseinvolving nitrous acid or selenium dioxide, may proceed by heterolytic stages.The increasing importance of furan and its derivatives is reflected in theallocation of one section of this Report to this topic. Studies have beenmade on the preparation of furans from carbohydrates, unsaturated diketones,ethylenic ethynylcarbinols, acetylenic acetals, acetylenic and ethylenicdiols, long chain cco-diols, 78-ethylenic alcohols, and y-bromo-alcohols ;where the formation of both a tetrahydrofuran and a tetrahydropyranring is possible, the former appears to be obtained preferentially.Several3 : 4-diaminotetrahydrofurans have been prepared from 3 : 4-dicarbethoxy-furans, and in the course of this work the 0-analogue of p-biotin has beensynthesised. Furan mercurials have proved to be of particular value forthe introduction of certain substituents, such as iodine and arsenic, into thefuran ring, and for the elimination of carboxyl groups in the 2- and 5-positions. A considerable amount of information has accumulated on thehydrogenation of furans, and by a suitable choice of catalyst and control ofconditions it is now possible, for example, to obtain furfuryl alcohol, tetra-hydrofurfuryl alcohol, 2-methylfuran, or 2-methyltetrahydrofuran fromfurfuraldehyde, although a satisfactory method has yet to be found for thepreparation of tetrahydrofurfuraldehyde, even by an indirect route.Elim-ination of side chains in furans by passage of the vapour over heated metalsor other catalysts is a process which is of particular interest from the indus-trial viewpoint, since it has recently been shown that in this way it is possibleto obtain good yields of furan from furfuraldehyde, and tetrahydrofuranfrom tetrahydrofurfuryl alcohol. Similar treatment of 2-cyanotetrahydro-furan and of methyl tetrahydrofuroate gives 2 : 3-dihydrofuran. Theseobservations are of some importance in view of the extensive studies whichhave been made on the conversion of furan derivatives by treatment witIHTRODUCTION.95ammonia or hydrogen sulphide into pyrroles and thiophens, processes whichin the absence of ample and cheap supplies of furan and tetrahydrofuranwould be only of academic interest. Improved results have also been claimedin the production of piperidine from tetrahydrofurfuryl alcohol. Ring fissionof furans, particularly of the tetrahydro-derivatives, may lead to the form-ation of valuable aliphatic compounds, and the optimum conditions for manyof these reactions have’ been carefully studied. Reagents which will bringabout this fission include hydrogen, carbon monoxide, hydrogen halides,acyl halides, and acetic anhydride.With unsymmetrically substitutedcompounds, the reaction may theoretically occur in two ways : A+ RCH( OX)*CH2*CH2*CHY*R’------+ B R CHY CH ,*CH ,*CH ( OX) OR’R(,?R’ +xyB\O/AIn some instances, however, the scission takes place almost entirely in onedirection.In continuation of last year’s Report on nucleosides and nucleotides anaccount is now presented of the chemistry of the adenine nucleotides func-tioning as coenzymes in biological systems. They fall into two classes :(a) derivatives of adenosine-5‘-phosphoric acid, which contain labile PJTO-phosphoryl residues by virtue of which they are active in phosphate transfer,and (b) dinucleotides in which adenosine-5’-phosphoric acid is united througha pyrophosphoryl linkage with a moiety containing either quaternarilybound nicotinamide or an alloxazine residue.These residues confer redoxproperties upon the molecules, and condition their activity as coenzymesof hydrogen transfer. The investigation of these coenzymes provides astriking example of the mutually beneficial way in which organic and bio-logical chemistry may interact; knowledge of the chemistry of thesecompounds has greatly furthered the insight gained during the last decadeinto vital processes, whilst in the constitutional investigations extensiveuse was made of biological test methods to determine the nature of fissionproducts produced in degradation reactions.This Report also includes a small section on the chemistry of pyrazineand its derivatives; it was originally planned to include similar sectionsdealing with other “ minor ” heterocyclic systems which have not hithertofound a place in these Reports, but these have had to be deferred.In report-ing upon the chemistry of such heterocyclic systems it has been consideredexpedient to present a representative account of the subject as a wholerather than limit it to a statement of recent developments.R. A. BASTER.E. A. BRAUDE.R. LYTHGOE.G. T. KEWBOLD.F. S. SPRENG.W. A. WATERS.1,. N. OWEN96 ORaANIC CHEMISTRY.2. GENERAL METHODS.1. Ion Exchange Resins.-Ion exchange processes have long been used forsoftening water; the use of zeolites and related substances was limited toreplacing calcium and magnesium by sodium cations. Hydrogen exchangewas impracticable since the zeolites are decomposed by acids; it becamepossible with the introduction of exchangers of the .sulphonated coal type.The development in this country of synthetic resin ion exchangers hasgreatly increased the scope of ion exchange reactions 2 and these reactionsare being increasingly applied in various commercial processes involvingorganic materials.In the belief that the technique will become of valuein certain other branches of organic chemistry, a short account of the re-latively few published reports of the use of ion exchangers in this field isappended.The firstis a cation exchanger and usually consists of a sulphonated phenol-form-aldehyde resin which can be re-activated by washing with a dilute solutionof acid.The second type usually consists of a polyamine-formaldehydetype resin and it allows the removal of an acid from solution. The resincan be reactivated and the acid recovered by washing with an alkali. Thistype of resin is not a true “ anion exchanger ” but rather an “ acid adsorbent.”I n addition, carbonaceous cation exchangers of the sulphonated coal typeare available which also allow the use of the hydrogen cycle exchange.The use of ion exchange resins for the removal of impurities in sugarjuice before crystallisation has constituted a major advance in sugartechn~logy.~ A very ingenious use of ion exchangers is described by D. T.Englis and H. A. Fiess 5 in the preparation of a high quality fructosesyrup from an aqueous solution of the polysaccharides of Jerusalem arti-chokes.The solution, which contains a relatively high percentage ofpotassium salts, was treated with a hydrogen exchange resin which increasedthe acidity of the solution to such an extent that hydrolysis was subse-quently effected without the addition of mineral acid. After hydrolysis,acid was removed by treatment with an acid adsorbent, followed by con-centration to a syrup.J. R. Matchett, R. R. Legault, C. C. Nimmo, and G. K. Notter describeexperiments leading to the recovery of tartaric acid from waste naturalsources such as the pomace from grape juice manufacture, and the slopfrom brandy making. The process consists in passing the slop through abed of acid-regenerated cation exchanger which frees the tartaric acid,Two types of synthetic resin ion exchangers are available.B.A. Adams and G. L. Holmes, J. SOC. Chem. Ind., 1935,54, 1.F. J. Myers, Ind. Eng. Chem., 1943, 35, 858.* F. J. Myers quoted by C. S. Cleaver, R. A. Hardy, and H. G. Cassidy, J . Amer.Chem. SOC., 1945, 87, 1344.F. N. Rawlings and R. W. Shafor, Sugar, 1942, 37, I, 26; F. W. Weitz, ibid.,1943, 38, I, 26; W. Meyer, Centr. Zuckerind., 1943, 51, 37; 0. Spengler and F. Todt,2. Wirts. Zuckerind., 1942, 152.Ind. Eng. Chem., 1942, 34, 864. Ibid., p. 486BAXTER AND SPRING : GENERAL METHODS. 97followed by treatment of the solution with an acid-adsorbent which pre-ferentially adsorbs the tartaric acid. The latter is recovered by washingthe resin with sodium carbonate eolution, and is finally converted into theinsoluble calcium salt.A new method for the separation of basic amino-acids from proteinhydrolysates has been described by R.J. Block.' The protein (bloodfibrin) was hydrolysed with hydrochloric acid, and the protein hydrolysate(amino-acid hydrochlorides) freed from 'excess of mineral acid and thenstirred with an acid-binding resin. After removal of this resin the solutionof amino-acids was treated with a cation exchange resin, from which thebasic amino-acids were regenerated by subsequent washing with dilutehydrochloric acid. From the solution of hydrochlorides so obtained, lysine,arginine, and histidinc were isolated by standard procedures. A method forthe separation of acidic amino-acids (glutamic and aspartic acids) fromprotein hydrolysates has also been developed by R.K. Cannan.8 Thehydrolysate is treated with a basic resin which binds the acidic amino-acidsbut not the neutral or basic amino-acids. The resin is then washed withhydrochloric acid which elutes the acidic amino-acids, and from the acidsolution, glutamic acid, as its hydrochloride, and aspartic acid, as its coppersalt, are readily isolated. Separation of amino-acid mixtures by meansof ion exchange resins has also been claimed by K. Freudenberg, H. Walch,and H. Molter? A study of the behaviour of various amino-acids on ionexchange resins has been made by D. T. Englis and H. A. Fiess,l0 and byC. S. Cleaver, R. A. Hardy, and H. G. Cassidy.ll The latter authorsinvestigated the infiuence of various factors upon the exchange processincluding type of resin, resin particle size, length of adsorption column,hydrogen -ion concentration of the solution, and concentration of the amino-acid solution.p-Alanine can best be prepared from its hydrochloride by passing asolution of the latter through a bed of an acid-binding resin followed byconcentration of the effluent.12A simplification of the preparation and purification of glucose- 1 -phosphate(Cori ester) is claimed by R.M. McCready and W. Z. Hassid.13 The essentialstep in the new procedure is that the reaction mixture obtained by thephosphorylation of starch followed by removal of inorganic phosphate istreated with a cation exchange resin. The effluent is then run through anacid-adsorbing resin which holds the glucose- 1 -phosphate but allows thepassage of dextrins, proteins, and weak organic acids.The acid-adsorbentresin was washed with aqueous ammonia, the Cori ester being isolated fromthe effluent as its dipotassium salt.Proc. SOC. Exp. Bwl. Med., 1942, 51, 252.Naturwiss., 1942, 30, 87; see also F. Turba, M. Richter, and F. Kuchar, ibid.,J. Biol. Chern., 1944, 152, 401.1943, 31, 508.10 I n d . Eng. Chem., 1944, 36, 604.l2 S. R. Buc, J. H. Ford, and E. C. Wise, ibid., p. 94.l3 Ibid., 1944, 66, 560.11 J . Amer. Chem. SOC., 1945, 67, 1343.REP.-VOL. XLII. 98 ORGANIC CHEMISTRY.An attractive procedure for the isolation of alkaloids from crude totaquine(the product obtained by alkaline precipitation of an acid extract of cinchonabark) by means of a cation exchange has been described by N.Applezweig l4who stresses the potential value of the new technique in alkaloid chemistryand reports its successful application in the isolation of atropine, scopol-amine, and morphine. When a solution of quinine in 1% sulphuric acid istreated with a cation exchange resin, the alkaloid is removed from solution;it can be liberated from the exchanger by washing with ammoniacal alcohol,a treatment which also reactivates the resin.A new method for the deacetylation of sugar acetates is described byW. W. Binkley, M. G. Blair, and M. L. Wolfrom; l5 in the case of inositolhexa-acetate saponification was effected with sodium hydroxide and thesolution passed over a column of cation exchanger resin which removedsodium ions; the effluent was then passed over a column of acid-bindingresin which removes acetic acid.Evaporation of the final effluent gaveinositol.2. Chromatography.-Reference has been made in previous Reports l6to the employment of chromatographic methods in the separation of mix-tures of amino-acids. T. Wieland and H. Fremerey l7 have applied thepartition-chromatography method of A. J. P. Martin and R. L. M. Syngeto effect a separation of copper complexes of amino-acids by partitionbetween two phenol-water phases on a silica gel column. In this way aseparation of alanine from valine and of leucine from valine and proline wasachieved, V. Prelog and P. Wieland l9 have claimed the resolution of anunsymmetrical tervalent nitrogen compound, Troger’s base,2o using thechromatographic technique of G.M. Henderson and H. G. Rule; 21 theadsorbent consisted of d-lactose activated by drying and grinding. G. T.Newbold and F. S. Spring,22 using a chromatographic method, have effected aready separation of two alcohols, a-euphorbol and euphol, from euphorbium.Several new applications of chromatography to specific problems incarbohydrate chemistry have been reported. Using the previously men-tioned method of A. J. P. Martin and R. L. M. Synge,18 D. J. Bel123 hasquantitatively separated 2 : 3 : 4 : 6-tetramethyl glucose from 2 : 3 : 6-trimethyl glucose by partition between organic solvents and water held ina column of silica gel.Separation of dimethyl glucoses from tri- and tetra-methyl glucose can also be effected. Using activated alumina as adsorbent,J, K. N. Jones2* has achieved a quantitative separation of tetramethylmethylglucosides from trimethyl methylglucosides. The same methodallows a partial separation of a constant-boiling mixture of trimethyl methyl-Z-arabofuranoside and trimethyl methyl-d-xylopyranoside ; the separationl4 J . Amer. Chem. SOC., 1945, 67, 1990. l6 Ibid., 1945, 67, 1791.l7 Ber., 1944, 77, 234. Ann. Reports, 1944, 41, 127; 1942, 39, 237.Biochem. J., 1941, 35, 1358; 1943, 37, 79, 86, 92.Helv. Chim. Acta, 1944, 27, 1127.*O J . Trdger, J . p r . Chem., 1887, 36, 227.22 Ibid., 1944, 249.21 J., 1939, 1568.24 Ibid., p. 333. 23 Ibid., p. 473BAXTER AND SPRING : GENERAL METHODS.99was complicated by the fact that separation of the a- and p-forms of thetwo glycosides occurred. The chromatographic separation of tetramethylglucose has been used in the assay of end-groups in polysa~charides.~~E. A. Tulley, D. D. Reynolds, and W. L. Evans 26 have used a chromato-graphy technique in the purification of a sugar acetate. Using magnewl,a synthetic hydrated magnesium acid silicate, as adsorbent and employingthe brush technique of L. Zechmeister 27 with aqueous alkaline permanganateas the brush reagent, W. H. McNeely, W. W. Binkley, and M. L. Wolfrom 28have achieved many clear-cut separations of mixtures of sugar acetates ;of these may be mentioned the separation of p-maltose octa-acetate fromsucrose octa-acetate.W. W. Binkley, M. G. Blair, and M. L. Wolfrom29in an analytical study of various molasses have isolated inositol (as itshexa-acetate) and d-mannitol (as its hexa-acetate) by chromatographicmethods. The method developed by W. S. ReichY3O in which separation ofmixtures of monosaccharides is achieved by conversion into a mixture ofthe (coloured) p-phenylazobenzoyl esters followed by chromatographicseparation of the latter, has been extended.31 J. K. Mertzweiler, D. M.Carney, and F. F. Farley 32 have also used the method to separate mixturesof the p-phenylazobenzoyl esters of 2 : 3-dimethy1, 2 : 3 : 6-trimethyl, and2 : 3 : 4 : 6-tetramethyl glucose. B. W. Lew, M. L. Wolfrom, and R. M.Goepp33 report a new method for the chromatography of carbohydratesand related polyhydroxy-compounds in which the adsorbents are clays suchas Florida clay and the developers comprise such solvents as alcohols,ethers, and pyridine. Water was the eluting agent and the brush techniquewas employed to detect zones.Separations such as sorbitol from d-glucose,d-mannitol, and dulcitol were achieved.3. Reduction.-L. W. Covert and H. Adkins34 first reported that Raneynickel, prepared by treatment of a nickel-aluminium alloy with aqueoussodium hydroxide followed by washing, is extremely reactive ; for examplethe authors observed that, when the catalyst is mixed a t room temperaturewith nitrobenzene in an open beaker, the nitrobenzene is reduced to a mix-ture of azo- and azoxy-benzene. J.Bougault, E. Cattelain, and P. Chabrier 35demonstrated that Raney nickel catalyst, prepared in the usual manner,retains hydrogen. Although the nature of the retention was not established,it was shown that the catalyst is capable of effecting a variety of reactionssuch as direct reduction of ethylenic bonds and the decomposition of an2 5 F. Brown, S. Dunstan, T. G. HalsaII, E. L. Hirst, and J. K. N. Jones, Nature,26 J . Amer. Chem. SOC., 1943, 65, 575.27 Bull. SOC. Chim. biol., 1936, 15, 1885; 1940, 22, 458; J , Amer. Chem. 80% 1946,28 Ibid., 1945, 67, 527.30 Compt. rend., 1939, 208, 589, 748; Biochem. J . , 1939, 33, 1000.31 G. H. Coleman, A. G. Farnham, and A. Miller, J . Amer. Chem. SOC., 1942, 84,32 Ibid., p. 2367.34 Ibid., 1932, 54, 4116.1945, 156, 786.67, 1919, 1922.29 lbid., p.1789.1501 ; G. H. Coleman and C. M. McCIoskey, ibid., 1943, 65, 1588.33 Ibid., 1945, 67, 1865.35 Bull.Soc. chim., 1938, 5, 1699100 ORGANIC CHEMISTRY.aqueous solution of potassium permanganate. Later,36 the same authorsshowed that Raney nickel, without addition of gaseous hydrogen, quantit-atively converts thioglycollanilide, HS*CH,*CO*NHPh, using aqueous oralcoholic solutions, into acetanilide and that similar treatment of thioglycollicacid, HS*CH,-CO,H, and dithioglycollic acid, ( SCH2C0,H),, gives acetic acid.A simple method of rendering benzene and toluene free from thiophen andmethylthiophen by treatment with Raney nickel is also reported. Theseearly observations have been considerably extended by R.Mozingo and hiscollaborators 37 who show that Raney nickel catalyst without gaseoushydrogen, at a moderate temperature and in the presence of a solventsuch as alcohol, removes sulphur from a variety of organic compoundsaccording to the scheme :Ni(H) R0S.R’ ~-> RH + R’H.As an example diphenyl sulphide in aqueous alcohol was refluxed withRaney nickel for 4+ hours (the authors observe that the time and tem-perature of reaction are probably in excess of those necessary) to give a68% yield of benzene. A similar ease of hydrogenolysis was observed inthe case of sulphones and sulphoxides, diphenyl sulphone, Ph*SO,*Ph , anddiphenyl sulphoxide, Ph*SO*Ph, giving 65% and 76% yields of benzenerespectively. H. R. Snyder and G. W. Cannon 38 report that with certainethers of ethylenedithiol two types of cleavage may occur when they aretreated with Raney nickel.In addition to the normal reaction leading to theformation of ethane, carbon-carbon cleavage occurs to a certain extentwith formation of methane :The method was used by V. du Vigneaud, R. Mozingo, and their collaboratorsto convert biotin methyl ester into dethiobiotin methyl ester.39 An interest-ing case of hydrogenolysis effected by Raney nickel has been described byM. L. Wolfrom and J. V. Karabinos40 who have developed a method forthe reduction of carbonyl compounds to the corresponding methylenecompounds. The carbonyl compound is first converted into the thioacetal(thioketal) which is then subjected to hydrogenolysis in dilute alcoholaolution with Raney nickel :Using this method acetophenone, benzophenone, benzaldehyde, and heptan-%one are converted into ethylbenzene, diphenylmethane, toluene, andR.Mozingo, D. E. Wolf, S . A. Harris, and K. Folkers, J . Amer. Chem. Soc.,s6 Bull. SOC. chim., 1940, 7 , 781.1943, 65, 1013.s8 IbicE., 1944, 66, 155.4 0 J . Amer. Chem. SOC., 1944, 66, 909.89 Ann. Reports, 1943, 40, 175BAXTER AND SPRING: GENERAL METHODS. 101heptane respectively and in the same way the diethyl thioacetal of aldehydo-d-glucose -penta-acetate is converted into l-deoxy-d-glucitol penta-acetate.Further examples of the application of Raney nickel catalyst withoutthe addition of gaseous hydrogen are given by R. Mozingo, C. Spencer, andK.Folkers 41 who show that, using the general reaction conditions outlinedabove for sulphides, ethylenic bonds are saturated, e.g., conversion of eugenolinto dihydroeugenol (75%), and carbonyl compounds are reduced to thecorresponding alcohols, e.g., conversion of acetone, ethyl acetoacetate, cycb-pentanone, and benzylideneacetone into isopropyl alcohol (78%), ethylP-hydroxybutyrate (96 yo), cycbpentanol (61 %), and 4-phenyl-2-butanol(79%) respectively. Benzaldehyde gives toluene (78%) and not benzylalcohol, indicating that the latter undergoes hydrogenolysis under theseexperimental conditions. Benzene rings, aliphatic acids, and esters are notreduced under these conditions.Raney (nickel-aluminium) alloy added to an alkaline solution of acarbonyl compound has been used by B.Whitman, 0. Wintersteiner, andE. Schwenk42 to reduce oestrone to a mixture of ct- and p-oestradiols.D. Papa, E. Schwenk, and B. Whitman 43 have found that this method whenapplied to phenyl ketones, Ph*CO*R, results in reduction to the correspondinghydrocarbon, Ph*CH,*R. On the other hand, ketones of the typePh*[CH,];CO*R, where x is 1 or greater, give the corresponding carbinol.In the case of ketones which are not soluble in aqueous alkali, the reactioncan be carried out in the presence of toluene or alcohol. The authorssuggest that the reduction is due to the liberation of hydrogen which is thenactivated by the presence of the nickel catalyst since reduction of thecarbonyl group also occurs using aluminium in conjunction with a previouslyprepared Raney nickel catalyst.If the nickel catalyst is omitted and thealkaline solution treated with aluminium, no reduction occurs or amorphousproducts are produced.44Reduction by means of nickel-aluminium alloy in the presence of alkalihas been further studied by E. Schwenk, D. Papa, B. Whitman, andH. Ginsberg 45 who show that using this technique various groups attachedto the aromatic nucleus are displaced by hydrogen. For example, halogenis displaced, bromobenzene giving benzene (100 yo) and m-chlorobenzoicacid giving benzoic acid (100%). Simultaneous replacement of halogenby hydrogen and reduction of carbonyl to methylene was observed in variouscompounds, p-chlorobenzaldehyde yielding toluene (60%) and p-bromo-acetophenone yielding ethylbenzene (67 %) .Reductive displacement ofthe sulphonic acid group occurs for example in o- and m-sulphobenzoic acidswhich gave 40% and 50% yields of benzoic acid respectively. Alkoxylgroups are displaced from disubstituted benzene derivatives when they aresituated in the o- and p-positions with respect to a m-directive group. Forexample, quantitative displacement of the methoxyl group occurs with o-4l J. Amer. C k m . SOC., 1944, 66, 1859.42 J . Biol. Chem., 1937, 118, 792.4 4 E. Schwenk and D. Papa, ibid., 1945,10, 232.4 5 J . Org. Chem., 1942, 7, 507.45 Ibid., 1944, 9, 1102 ORGANIC CHEMISTRY.and p-methoxybenzoic acids, but the m-isomer is recovered unchanged. Aninteresting case is provided by p-nitroanisole which gives a mixture of aniline(20%) and p-anisidine (70%) ; elimination of the methoxyl group can onlyoccur prior to the reduction of the (m-directive) nitro-group to the (o-p-directive) amino-group.Other alkoxy groups are similarly displacedprovided that a m-directive group is located in the.p- (or o - ) position. Aquantitative method for the estimation of halogen in organic compoundsbased upon the replacement reaction has been describedSP6 Various re-ductions of ethylenic compounds by means of nickel-aluminium alloy inthe presence of alkali have been reported,d7 oleic acid giving stearic acid(100%) and maleic acid giving succinic acid (90%). An acetylenic com-pound, phenylpropiolic acid, undergoes complete reduction to p-phenyl-propionic acid.In a study of reduction by means of sodium in liquid ammonia, A.J.Birch48 shows that in a number of benzene and naphthalene derivativesthe course of reduction is very considerably modified by the addition of analcohol. Thus with sodium a- and p-napththoxides, sodium in liquid am-monia effected little reduction, but addition of lert-amyl alcohol as a protonsource gave good yields of dihydro-derivatives (5 : 8-dihydro-a-naphtholand p-tetralone respectively). Reduction and demethylation of methoxy-alkylbenzene derivatives occur when they are treated with sodium in liquidammonia in the presence of an alcohol, whereas treatment with metal andliquid ammonia alone merely leads to demethylati~n.~~ The methoxyalkyl-benzenes when treated with sodium in liquid ammonia in the presence ofalcohol give a small quantity of the corresponding phenol (simple demethyl-ation) together with a difficultly separable mixture of starting material andreduction products.The reduction products were characterised as dihydro-derivatives since treatment of the reaction product with mineral acid gavean ap-unsaturated ketone produced thus :Hydrogenolysis of a number of vinylcarbinols by means of sodium and analcohol in liquid ammonia has also been studied by A. J.Reductions by alkali metals and liquid ammonia have also been studiedby C. M. Knowles and G. W. Watt.51 Quinoline is reduced to a dihydro-derivative which, it is suggested, is the 1 : 4-compound. Reduction of4 6 E. Schwenk, D. Papa, and H.Ginsberg, Ind. Eng. Chem. Anal., 1943, 15, 576.4 7 E. Schwenk, D. Papa, B. Whitman, and H. Ginsberg, J. Org. Chem., 1944,4 8 J., 1944, 430. '' I(. Freudenberg, W. Lautsch, and G. Piazolo, Ber., 1941, 74, 1886.I0 J . , 1945, 809.I1 J . Amer. Chem. SOC., 1943, 85, 410.9, 175BAXTER AND SPRING GENERAL METHODS. 103nitroparaffins by sodium in liquid ammonia is also described ; 52 the reactionis slow and incomplete and yields the corresponding alkylhydroxylamine~.~~A very interesting modification in the technique of the Wolff-Kishnerreduction of carbonyl to methylene compounds is reported by M. D. Soffer,M. B. Soffer, and K. W. Sherk.54 The essential part of the modification isthe use of a high-boiling solvent such as a glycol to obviate the use of itbomb-tube or high-pressure apparatus.As an example may be quoted theprocedure for the reduction of propiophenone to n-propylbenzene (79%).Sodium is dissolved in an excess of diethylene glycol, hydrazine hydrateand the ketone are added, and the mixture is refluxed for 49 hours.H. Houber 55 claims high yields of primary amines by the catalyticreduction of basic nitriles in the presence of ammonia and Raney nickel.Using this method secondary amine formation is negligible and reduction israpid. Perchloric acid is found to be an effective activator for certain reduc-tions using a palladium-barium sulphate catalyst. Using this catalyst andacetic acid as solvent, ethyl benzoylacetate is reduced to ethyl p-phenylpro-pionate in the presence of perchloric acid.56 Using similar conditionsreduction of a p-arylallcanolamine, e.g., Ph*CH( OH)*CHMe*MH,, gives thecorresponding p-arylalkylamine, Ph*CH,*CHMe*NH,.It has been shown by R.Mozingo and associates 57 that hydrogenationof carbon-carbon double bonds can be effected in some sulphur-containingcompounds, using a supported palladium catalyst. Thus thiophen and2-bromothiophen are reduced to tetrahydrothiophen, and the successfulreduction of certain biotin intermediates and other sulphides is described.4. PhosphoryZation.-The known phosphorylation procedures have beenreviewed by F. R. Atherton, H. T. Openshaw, and A. R. Todd,58 who dis-cuss the limitations of existing methods for the phosphorylation of alcohols(particularly carbohydrates) and amines, and state the chief requirementsto be met by a convenient method.In seeking such a method, the authorshave prepared and examined the reactions of dibenzyl chlorophosphonate(11). L. Zervas 59 had previously prepared this acid chloride but reportedthat it was so unstable as to be of little use. The method of preparationnow used is treatment of phosphorus trichloride with benzyl alcohol in thepresence of dimethylaniline to give dibenzyl hydrogen phosphite (I), whichis treated with chlorine in carbon tetrachloride solution, a method developedby H. McCombie, B. C. Saunders, and G. J. Stacey.60 The solution ofdibenzyl chlorophosphonate in carbon tetrachloride thus obt'ained is normallyused directly without isolat,ion of the ester since the latter decomposed ons2 G.W. Watt and C. M. Knowles, J. Org. Chem., 1943, 8, 540.63 For a general review of the reduction of nitroparaffins to the corresponding64 J . Amer. Chenz. SOC., 1945, 67, 1435.65 Ibid., 1944, 66, 876.66 K. Rosenmund, E. Karg, and F. K. Marcus, Be?., 1942, 75, 1850.5 7 J. Amer. Chem. SOC., 1945, 67, 2092.li9 Naturwiss., 1939, 27, 31T.hydroxylamines see H. B. Hass and E. F. Riley, Chew&. Reviews, 1943, 32, 390.J . , 1945, 382.6o J., 1945, 380104 ORQANIC CHEMISTRY.attempted distillation. The solution in an inert solvent reacts readilywith amines, or with alcohols in the presence of pyridine, and the productscan be debenzylated by the standard hydrogenolysis procedure. Thus withalcohol in the presence of pyridine, dibenzyl ethyl phosphate was obtained,hydrogenolysis of which gave ethyl dihydrogen phosphate.The chloro-phosphonate (11) does not react readily with phenols but reacts readily(Ph*CH,*O),P*OH --+ (Ph*CH,*O),POCl(1.1 P a /with sodium phenoxides. More recently D. Deutsch and 0. Ferne 61 haveindependently developed a phosphorylation technique also employingdibenzyl chlorophosphonate. I?. R. Atherton, H. T. Openshaw, and A. R.Todd 62 have described a new method for the phosphorylation of amines inwhich dibenzyl phosphite reacts with carbon tetrachloride in the presenceof a strong primary or secondary amine. The reaction appears to occur intwo stages in the first of which the carbon tetrachloride reacts with thephosphite to give a trichloromethyl phosphonate, the base acting as hydrogenchloride acceptor :(Ph*CH,*O),PH + CCl, + B -+ (Ph*CH,*O),P*CCl3 + B,HC1146 ORGANIC CHEMISTRY.reaction mechanisms which may be involved, but one can discriminatebroadly between the rapid oxidations which can be effected a t room temper-ature, usually in the presence of light or catalysts, and many slower oxidationswhich can be effected only a t higher temperatures or by using the reagentin large excess. Into the first category come reactions with polyphenols,x-hydroxy-acids, and olefins, and into the second, oxidations of saturatedfatty thio-ethers, and aromatic azo-compounds.Evidence is now accumulating to indicate that the rapid reactions ofhydrogen peroxide involve the presence, in aqueous solution, of the transientfree hydroxyl radical, *OH.As mentioned in these Reports for 194397it' is considered by N. A. Milas, P. F. Kurz, and W. P. Anslow that the photo-chemical hydroxylations of ally1 alcohol, crotonic acid, and maleic acidby hydrogen peroxide are reactions of free hydroxyl radicals :H202 --%+ 2 *OHCH,:CH*CH2*OH + 2 *OH --+ CH2( OH)*CH( OH)*CH,-OHsince the radiant energy supplied is amply sufficient to split the weak0-0 bond. Since a similar cis addition of two hydroxyl groups to olefinscan be effected much more easily by catalysing the peroxide reaction with alittle osmium tetroxide, vanadium pentoxide, or, less effectively, chromiumt r i o ~ i d e , ~ ~ * loo* lol it is possible that in these cases too the free hydroxylradical is concerned, its immediate precursor being a relatively unstableper-acid.More stable per-acids, such as perbenzoic acid, do not easilyoxidise olefinic substances such as ethyl fumarate or ethyl maleate whichare attacked by the Milas reagents,lm, lo2 which will even oxidise benzeneto phenol, toluene to cresols, and naphthalene to naphthols.99 Since thiscatalysed hydroxylation can also be effected by means of anhydrous tert.-butyl hydroperoxide lo3 it is evident that organic hydroperoxides Alk-O-OH,and perhaps also their inorgamic per-esters, must also be looked upon assources of active hydroxyl radicals. The fact that chromium trioxideaccelerates the decomposition of aqueous hydrogen peroxide 104 tends tosupport this view of the action of the catalyst. The possibility that thehydroxylation catalysed by osmium tetroxide proceeds via addition of thetetroxide to the double bond cannot be ruled out in all cases, though, asF.S. Spring has pointed O U ~ , ~ ' it is discounted by the fact that hydrogenperoxide plus a little osmium tetroxide will hydroxylate ap-unsaturatedketones, whilst osmium tetroxide in dry ether is inert to these substances.96 H. D. Dakin, J. Biol. Chem., 1908, 4, 63, 227; 1909, 5, 409; cf. H. D. Dakin,o 7 Ann. Reports, 1943, 40, 107.9a J . Amer. Chenz. SOC., 1937, 59, ,543.100 N. A. Yilas, S. Sussmann, and H. S. Mason, ibid., 1939, 61, 1845.101 N. A. Milas and L. S. Maloney, ibid., 1940, 62, 1841.l o 2 Cf. J. Boeseken, Rec. Trav. ciiirn., 1926, 45, 838.lo3 N.A. Milas and S. Sussmann, J. Arner. Chem. SOC., 1936,58, 1302.lo4 M. Bobtelsky, A. Glasner, and I,. Bobtelsky-Chaikin, ibid., 1945, 67, 916.Oxidations and Reductions in the Animal Body," London, 1922.99 N. A. Milas, ibid., p. 2342WATERS : MECHANISMS OF OXIDATION. 147W. Treibs 1°5 has shown that hydrogen peroxide with a vanadate cata-lyst will rapidly oxidise cyclic ketones to aldehydic acids, and has suggestedthat this action involves the addition of two hydroxyl groups to the enolicforms of the ketones :OHoxidat.ive 3 pgg GF fission of a-glycol v \ HThis view is probably an over-simplification of the mechanism of this reaction,since crystalline addition products of anhydrous hydrogen peroxide andvarious cyclic ketones have been isolated 106% lo7 (e.g., XIV).These may bedehydrated by cold sulphuric acid to unstable cyclic peroxides (XV) orpossibly (XVI) lo7 whilst warm sulphuric acid converts the ketone-hydrogenperoxide complex into a lactone.HO,,O*OH 0(XIV.) (XVI.)M. Stoll and W. Scherrer lo6 have suggested that this change, which is ofcourse the normal action of Caro’s acid on a ketone (compare p. 140), proceedsvia an epoxide as follows,the final ring fission being brought about by a prototropic change :Definite evidence for the participation of the neutral hydroxyl radicalin reactions of hydrogen peroxide has been obtained from studies of decom-positions and oxidations catalysed by mild reducing agents, such as ferroussalts. In 1931 F. Haber and R.Willstatter,lo8 in an attempt to explain themechanism of enzyme oxidation, suggested that ferric compounds couldcatalyse many dehydrogenations by abstracting one electron from a H-Rbond :Fe+++ + H-R + Fe++ + H+ + Re105 Ber., 1939, 72, 7, 1194.107 N. A. Milas, S. A. Harris, and P. C. Pangiotakos, J . Amer. Chern. SOC., 1939,108 Ber., 1931, 64, 2644.Helv. Chirn. Acta, 1930, 13, 142.01, 2430148 ORGANIC CHEMISTRY.The action of the enzyme cablase in decomposing hydrogen peroxide wasrepresented as followsfollowed by the chain processFe++' + HO-OH _c, Fe++ + H+ + 00-OH . . (i)*O-OH (-0-0:)- + H+ . . . . . . . . . (ii)(-0-0:)- + HO-OH + (*O-O*) + H-0. + ( : O H ) - . (iii) i H-O*+HO-OH+HO~H+~O-OH . . . . . (iv)This scheme was amended in 1934 by F.Haber and J. Weiss.lo9 Theypointed out that ferrous salts were much more effective catalysts for thedecomposition of hydrogen peroxide than ferric salts, and consequentlyformulated the primary reaction asFe++ + HO-OH --+ Fe+++ + H-0. + (:O-H)- . (v)Fe++ + -0-H -+ Fe+++ + (:O-H)- . . . . (vi)and introduced, a.s the chief chain-breaking reaction,Flow experiments showed that the kinetics of the iron-sa,lt-catalysed decom-position of hydrogen peroxide accorded with the Haber-Weiss scheme overa wide range of pH. However, ferric salts do catalyse the decompositionof hydrogen peroxide, though slowly, so that the primary reaction of theHaber-Willstatter scheme does occur. R. Kuhn and A. Wassermann 110have shown that the reduction of ferric ions to ferrous ions by decomposinghydrogen peroxide can be demonstrated by complex formation with act'-dipyridyl or phenanthroline. The resulting complexes have low but stillobservable catalytic activity.J. H.Baxendale, M. G. Evans, and G. S. Park ll1 have, in a paper ofimportant technical as well as theoretical value, recently given a conclusiveexperimental proof of the Haber-Weiss theory. On the addition of ferrousions, oxygen-free aqueous hydrogen peroxide immediately brings about thechain polymerisations of methyl acrylate, methacrylic acid, methyl meth-acrylate, vinyl cyanide, and styrene, both in solution and in the form ofemulsified droplets. I n the presence of the monomeric olefin no oxygen iaevolved from the hydrogen peroxide, and the stoicheiometry of the reaction,with ferrous salt in initial excess, changes from the oxidation of 2 equivalentsof ferrous ion per mol. of hydrogen peroxide [i.e., an overall reaction2Fe++ + H202 + 2Fe+++ + 2(:OH)- due to the occurrence of reactions(v) and (vi) in quick succession] to oxidation of 1 equivalent of ferrous ionper mol.of peroxide [i.e., reaction (v) alone], for the free hydroxyl radicalsare removed very rapidly by the addition reaction (I%)HO* + CH2=CHR 4 HO--CH2-6HR . . . (vii)and thus start a polymerisation chain, which continues :HO*CH,*eHR + CH2:CHR -+ HO*CH2*CHR*CH,*6HR etc.lo9 Proc. Roy. SOC., 1934, A , 147, 333.ll1 Tran8. Faraday SOC., 1946, 42, 155.110 Annalen, 1933, 503, 203WATERS : MECHANISMS O F OXIDATION. 149until terminated by the union of two radicals, for example ByHO*[CH,*CHR],L*CH2*bHR + *OH --+ HO*[CH,*CHR],, + ,*OH .(viii)J. H. Baxendale, M. G. Evans, and G. S. Park estimate that reaction (vii)between methyl acrylate and hydroxyl radicals is about 5 times as fast asthe reaction between hydroxyl radicals and ferrous ions [reaction (vi)], whilstvinyl cyanide is attacked by hydroxyl still more rapidly. Even ethyleneitself is attacked by hydrogen peroxide-ferrous sulphate mixture.It is generally known that the chain polymerisation of olefins, CH,:CHR,cannot be performed reproducibly in the presence of air, since oxygen actsas an (' inhibitor '' of the free-radical reaction. When studying the reactionsdescribed above, it was found that oxygen could be absorbed by the reactingsystem, though only when ferrous ions, hydrogen peroxide, and monomericolefin were all present.This suggests that oxygen acts as a chain-terminator :HO*[CH,*CHR],* + 0, -+ HO*[CH,*CHR],*O*O* . (ix)HO*[CH,-CHR], - ,*CH,.CHR*O*O* + Fe++ --+HO*[CH,-CHR],, - ,*CH,*CR:O + (:OH)- + Fe+++This action is undoubtedly significant also in the autoxidation of olefinsin the presence of hydroperoxides which can, by thermal decomposition,generate free hydroxyl radicals. Whilst reaction (ix) is an addition of mole-cular oxygen to an ethylenic bond, the immediate reaction product is ahydroperoxide radical, which may (as indicated on pp. 132-140) thenabstract hydrogen from a reactive methylene group. In the case of styrene,with which the spontaneous autoxidation is accompanied by polymerisation,S.Medvedev and P. Zeitlin 112 have shown that the autoxidation andpolymerisation chains must involve the same radicals, for the ratio (amountoxidised) /(amount polymerised) is a constant, independent of the reactiontime, the temperature, and the presence of inhibitors.Ferrous salts are by no means the only reducing agents which can liber-ate hydroxyl radicals from hydrogen peroxide. J. H. Baxendale, M. G .Evans, and G. S. Park ll1 have found Cr", Cu', Ti+++ and Mn++ cations,and also metallic mercury, to be effective polymerisation catalysts, whilstR. G. R. Bacon113 and L. B. Morgan 11* have described many furtherapplications of this (' reduction activation " of peroxidic compounds, suchas potassium persulphate, in emulsion polymerisation.Sulphites, thio-sulphates, sulphides, organic thiols, hydroxylamine, quinol, and pyrogallol,and also clean metals such as copper, iron, and silver must all be regarded assubstances capable of giving electrons singly to hydrogen peroxide, accordingt o the generalised equationJ. Weiss 115 has discussed the application of the hydroxyl radical theory tothe metal-catalysed decomposition of hydrogen peroxide, and also to thellS Trans. Faraday SOC., 1946,43,140.115 Ibid., 1935,31, 1647.Red + HO-OH --+ Ox + KO* + (:OH)-; where Red += Ox + e112 Acta Physicochim. U.R.S.S., 1945,20,3.114 Ibid., p. 169150 ORGANIC CHEMISTRY.action of peroxidase enzymes.l16 As would be expected of reactions involvingthe neutral *OH radical, H.Wieland and W. Francke 117 have shown thatmany of the low-temperature oxidations involving hydrogen peroxide areaccelerated enormously upon the addition of a little ferrous salt, and there-after proceed at a steady rate, though the reacting solution contains through-out both ferrous and ferric ions. Oxidations of arsenites and phosphites,as well as of formic acid, a-hydroxy-acids, and a-amino-acids all show thisbehaviour, though the initial accelerated oxidation is not evident with aro-matic substances such as quinol, pyrogallol, and p-phenylenediamine, theoxidations of which can be catalysed equally well by ferrous and by ferricsalts. This has led J. Weiss 116 to discriminate between two types ofcatalysed oxidations.I n Type A reactions, which are exemplified by the hydrogen peroxide-formic acid system, the oxidisable substance is attacked only by the com-plete system (H,O, + Fe"), and oxidation stops when all the-iron is in theferric state, though excess of hydrogen peroxide may still be present.Theactive oxidising agent is the neutral hydroxyl radical, and only short reactionchains may be involved, e.g. :*OH + H*CO*OH + H0.H + *CO*OH*CO*OH + HO*OH + HO*CO*OH + *OHHO*CO*OH =+ H,O + CO,} chainwith *CO*OH + *OH --+ H,O + CO, chain breakingI n Type B reactions, which are exemplified by iodide anion and by thephotographic developers, the oxidisable substance is attacked by the Fe+++cation, and the function of the hydrogen peroxide may consist in the repeatedrapid re-oxidation of the ferrous cation to the ferric state.Fenton's reaction l18-the oxidation of a-hydroxy-acids at 0" with hydro-gen peroxide in the presence of a trace of a ferrous salt-is obviously a processof Type A , and can be written either asHOH OHR--$-CO,H + *OH ---+ R-vLCO,H + H,O?HR-(YCO,H + HO-OH ras/ R-Y-CO,H + *OH __+ R-CO-CO,HOH OHor as 1 H HOH 0.R-(Y-CO,H + *OH + R-V-CO,H + H,OH + HO-OH11% J .Physical Chern., 1937, 41, 1107.118 H. J. H. Fenton, J . , 1894, 65, 899; 1899, 75, 1 ; 1900, 77,69.11' Annalen, 1927,457, 1 ; 1929,478,1, 19WATERS : MECHANISMS OF OXIDATION. 151according to whether or not the initial dehydrogenation can be believed toattack R C-H or an O-H link.*In accordance with this theory, J.H. Baxendale, M. G. Evans, and G. 8.Park ll1 have shown that the addition of a monomeric olefin, such as iso-propenyl cyanide or methyl acrylate, greatly reduces the rate of oxidationof glycollic acid by Fenton’s reagent, thus proving that the entity (hydroxylradical) responsible for starting the chain oxidation of glycollic acid is thesame as that responsible for initiating polymerisation of the olefin.I n accordance with this recent evidence, the hydroxylation of olefinsby the inorganic per-acids may be represented as a chain processRO-OH -+ RO* + =OH- H-GH--CH=CH- + *OH + 8Hin which the formation of an epoxide is an obvious alternative as a secondstage :-CH-GH- -CH--GH-f l +- Ph-CO-OH + *OH - I -O-i-H + Ph-CO-O-/-OHThe slower reactions of hydrogen peroxide, for which a large excess of thereagent is usually employed, may also be reactions of hydroxyl radicals,but not chain processes in which a second active radical, capable of attackinghydrogen peroxide molecules, is concerned.It is significant that all thereactions of hydrogen peroxide indicate that the free hydrosyl radical isincapable of attacking the C-H bond of a paraffin chain a t room temperatures.Oxidations of thio-ethers, tervalent arsenic or antimony compounds,etc., may also be represented as radical addition processes, for the oxidisableatom can, at least temporarily, increase its electrovalency shell :Hydroxylations of olefins, apparently anaslogous to those effected by theFenton and the Milas reagents, can also be carried out by using a cold aqueoussolution of a chlorate and a trace of osmium tetroxide or vanadium pentoxideas a catalyst.llg$ l 2 O $ 121 G.Braun 122 considers that this chlorate oxidation1111 K. A. Hofma,nn, 0. Erhardt, and 0. Schneider, Ber., 1913,46,667.120 N. A. Milas and E. M. Terry, J. Amer. Chem. SOC., 1925,47, 1412.lzl J. W. E. Glatfield and S. Woodruff, ibid., 1927, 49, 2309.1z2 Ibid., 1929, 51, 228.* This is at present a moot point, which is of significance also in connection with themechanism of oxidation of alcohols (compare W. A. Waters, Trans. Faraduy SOC.,1946, 42, 194). The alternative radicals, >CH-0- and >&OH, would certainly betautomeric, and the removal of the second hydrogen from either form to give thestable structure >CO would undoubtedly be so facile that it might be quite impossibleto characterise the primary radical152 ORGANIU CHEMISTRY.of olefins invariably gives cis-glycols, in contrast to perbenzoic acid oxidation,which often yields trans-glycols on account of the trans addition of water tothe intermediate epoxide.There is some evidence to suggest that chloratehydroxylation requires solutions of low pH, and that it may involve freehypochlorous acid as a possible source of hydroxyl radicals :HO-Cl + e -+ C1- + *OHChlorate reagents are also convenient for effecting the smooth oxidation ofquinols to quinones.l23Isolated experiments also indicate that active radicals are possibly pro-duced by the ‘‘ reductive activation ” of both bromates 12* and iodates,12jbut it may be unwise to speculate too much upon old evidence.Oxidations Involving Sulphur Compounds.The autoxidation of aqueous solutions of sulphites was one of the firstreactions to be explained successfully by the Haber-Willstatter theory ofone-electron transfer.lo** 126 In this, cupric salts are particularly activecatalysts, and have a noticeable effect in concentrations as low as 10-13molar.CU++ + SO,= --+ Cu’ 4- (*SO,)-(*SO,)- + 0, --+ (*OOSO,)-(*O*O*SO,)- + (HSO,)- --+ (HO-O-SO,)- + (*SO,)-giving, in (HS0,)-, the anion of Caro’s acid, which is it sufficiently powerfuloxidiser to convert sulphites directly into sulphates by hydroxylation.(HO*OeSO,)- + SO,= --+ (O*SO,)= + (HO*S03)-Evidence for the participation of the (*SO,)- radical-ion in this processis afforded by the fact that, in the absence of oxygen, copper sulphate andsodium sulphite react to give cuprous oxide and sodium dithionate.l27Persulphate radicals, (.O-O*SO,)-, are evidently formed since the reactingsystem can induce autoxidations of arsenites, nitrites, and alcohols, aldehydes,and other organic substances, many of which can be used a6 “ stabilisers ”of sulphite solutions.The action of photographic developers is of interest in this connection.They consist essentially of buffered solutions of easily oxidisable polyphenolsor amino-phenols, together with a large excess of sodium sulphite.Theinitial reduction of the sensitised silver salt is undoubtedly due to the organiccomponent, which, giving a semi-quinone radical, is then reduced again byThe chain process can be written asE.M. Terry and N. A. Milas, J . Amer. Chem. SOC., 1926,48, 2647; W. Baker and(Miss) I. Munk, J., 1940, 1092; cf. K. A. Hofmann, Ber., 1912,45, 3329.12‘ F. Wachholtz, 2. Elektrochem., 1927, 33, 545.126 H. Wieland and F. G. Fischer, Ber., 1926, 59, 1171.12E H. J. L. Backstrom, 2. physikal. Chern., 1934, B, 34, 122.12’ H. Baubigny, Compt. rend., 1912, 154, 701 ; Ann. Chim. Phys., 1910, 20, 12;1914,1, 201WATERS : MECHANISMS OF OXIDATION. 153the sulphite, and so acts as a '' potential mediator " to the whole system,though it is the sulphite which is the main reducing agent in the end.12*Ag+ + HO*C,H,*OH --+ Ag+ + H+ + *O*C,H,*OH-O*C,H,-OH + (SO,)= -+ (:O*C,H,*OH)- + (*SO,)-(*SO3)- + (:OH)- + *O*C,H,*OH --+ (HS0,)- + (:O*C,H,*OH)-The final reaction involves the mutual destruction of two radicals, and henceeach component of the developer can be regarded as acting as an anti-oxidantfor the other.Many other reactions undoubtedly involve the oxidised sulphite radical-ion.Thus nuclear sulphonation of phenols can be eEected by blowing airthrough their solutions in ammonium sulphite in the presence of a trace ofa copper salt ,129 and similar reactions have been reported with heterocycliccornpo~nds.~30 This action of the (*SO,)- radical is analogous to the alkyl-ation of quinones, etc., with lead tetra-acetate (p. 144). Again, quinonescan be substituted directly by thiols, with simultaneous reduction toquin01s.l~~ The " peroxide-catalysed " additions of sulphites to olefins 132also show the chemical importance of this radical-ion.Thiol radicals, RS*, formed by peroxide catalysts, are undoubtedlyresponsible for the similar " abnormal " addition of thiols to olefins.Theseradicals probably take part in many oxidation reactions involving sulphurcompounds, notably in biochemical processes in which the equilibriumRS: += RS* + e may often be involved as an essential potential mediatoreven when sulphur compounds do not appear as final reaction products.92In this connection copper porphyrins and other " trace-metal " compoundsmay play an essential part in initiating radical formation.K.Ziegler and K. Giinicke 133 have shown that thiophenol can markedlyaccelerate the autoxidation of olefins once radical formation has been initi-ated by the addition of a trace of triphenylmethyl, though phenyl disulphide,Ph*S*S*Ph, has no catalytic action. The essential reaction involving thetkiol is therefore hydrogen abstraction by the thiol radicalPh*S* + HR + Ph*SH + *Rrather than the activation of oxygen by it. At higher temperatures, however,organic disulphides can dissociate thermally to thiol radicals, and can be usedas selective dehydrogenators. J. J. Ritter and (Miss) E. D. Sharpe 13*for instance have shown that tetralin can be smoothly oxidised to naphthal-ene by distilling it a t 250" with isoamyl disulphide through a fractionatingcolumn.isoAmylthio1 gradually distils over, and can easily be collected inCf. A. Weissbcrger, D. S. Thomas, and J. E. Lu Valle, J. Amer. Chenz. Xoc.,1943,65, 1489.lze (Mlle.) Y . Garreau, Bull. ,Sot. chim., 1934, 1, 1563; Compt. rend., 1936, 203,1073.130 H'. McIlwain, J., 1937, 1704.131 J. M. Snell and A. Weissberger, J . Amer. Chem. SOC., 1939, 61, 450.13' M. S. Kharasch, E. M. May, and F. R. Mayo, J . Org. Chem., 1938,3, 175.133 Annalen, 1942, 551, 213. lS4 J . Amer. Chem. SOC., 1937, 59, 2361154 OHQANIC CEiEM1STK.Y.dilute hydrogen peroxide and so immediately re-oxidised for further use.This reaction has been shown to involve the production of transient hydro-carbon radi~a1s.l~~ It is probable that dehydrogenations by sulphur orselenium have a similar mechanism, though in a heterogeneous system.136The fact that thiols can dehydrogenate reactive methylene groups may havean important bearing upon the mechanism of vulcanisation of rub-ber.136, 13’% l38 Thiols, such as thiobenzthiazole, C,H,<S >C*SH, and di-sulphides, such as tetramethylthiuram disulphide, Me,N*CS*S*S*CS*NMe,,are technically valuable “ accelerators ” of the vulcanisation process.Theymay act as Bources of thiol radicals which dehydrogenate active methylenegroups in the rubber molecule, and thus promote the cross-linking of hydro-carbon chains, either by dimerisations :RS- + >CH,+ R*SH + >CH*w2>CH* + >CH*CH<or by union with sulphur, giving thio-ethers or polysulphides :2>CH* + * S o + >CH*S*CH<whilst from the free sulphur there may again be formed fresh active thiolradicals which could continue the chain process :>CH* + *S* ---+ >CH*S*The close connection between the autoxidisability of rubber and its degreeof vulcanisation 139 is an indication that radicals of similar types are involvedin both.Attention may be directed to the fact that thiol radicals (R-S) can evid-ently dehydrogenate some C-H links, whereas hydroxyl or alkoxy (R-0)radicals apparently cannot.This may be associated with the fact thatlthe bond energy of disulphide links is apparently higher than that of peroxidelinks.*Reactiom of Quinones.It is now well-established that the reversible oxidation-reduction re-actions of many quinones and quinonoid dyes take place in two successivestages, which may be representedQH, + QH- + H+ ; QH- + QH + eQH eQ- + H + ; Q- +=Q + esince the independent existence of semi-quinonoid radicals, QH (e.g., XVII) ,135 W; A.Waters, Trans. Faraday SOC., 1946,42, 184.136 E. H. Farmer, ibid., 1942, 38, 345.13’ Idem, ibid., p. 360.138 E. H. Farmer and S. E. Michael, J., 1942, 513.139 S. Horrabin, R. G. A. New, and D. Taylor, Trans. Faraday SOC., 1946,42,262.* Pauling (“ The Nature of the Chemical Bond,” p. 63) gives S-S 64 kg.-cal. and0-0 35 kg.-cal. though both these values may be much too low (compare J. L. Bollaxidand G. Gee, Trans. Paraday SOC., 1946,42,244)WATERS : MECHANISMS OF OXIDATION. 155which are often deeply coloured, has been conclusively established bothmagnetically and electrochemically. 140Compounds of this group include most of the important oxidation-reduction indicators, and also such biochemically important substances as.. - pyocyanine and riboflavin.As well as being important onaccount of their colour changes, several of these semi-quin- :P: onoid systems are important as " potential mediators "Me\//\/Me which, by two successive, reversible, stages of one-electronI I] transfer can bring about oxidations or reductions which,M e / \ \ f \ ; M e though thermodynamically possible, are otherwise exceed-: 0 : ingly slow ; 141, 142 e.g., 2Ti+++ + I, --+ 2Tiff++ + 21-.P. A. Schaffer in 1936 142 pointed out that oxidations and (XVII.) reductions involving simultaneous two-electron changes, suchas Tl' --+ Tl+++, or I, --+ 21-, were generally slow, whereas all oxidationsor reductions involving one-electron changes seemed to be rapid, and ascribedthe catalytic powers of quinonoid dyes to their abilities to accept electronssingly.He pointed out in particular the significance of this view in con-nection with autoxidation and with biological respiration involving the reduc-tion of free oxygen to hydrogen peroxide. L. Michaelis 143 has gone furtherto propound a " principle of compulsory univalent oxidation " accordingto which oxidations of organic compounds can proceed a t a measurablespeed only in steps of one electron each. He postulates the productionof transient radicals in all oxidations, and considers that an oxidation orreduction process is slow when the formation of this intermediate radicalinvolves a high energy increment.Rapid oxidation can occur if, owing toresonance, the formation of the radical involves comparatively little energy.Consequently it is to the resonance-stabilisation of semi-quinone radicals,and to the formation of dimeric quinhydrone complexes, that one can ascribethe potential mediating powers of quinonoid dyestuffs.The preceeding pages will have shown that, in broad outline, the Michaelisprinciple of compulsory univalent oxidation is applicable over a very widefield of organic chemistry, but it may still be premature to accept it, withoutexperimental confirmation, for all oxidation reactions.The Schaffer-Michaelis theory of one-electron transfer was developedfor reactions of quinones in aqueous solution, in which acid-base ionisation,QH Q- + H', was immediate.There is, however, a considerableamount of evidence to suggest that in non-electrolytes quinones can oxidiseby hydrogen atom transfer :Q+H-R---+*QH+*R140 L. Michaelis, Chem. Reviews, 1935, 16, 243; L. Michaelis and M. P. Schubert,ibid., 1938, 22, 437 ; cf. A. E. Remick, " Electronic Interpretations of Organic Chem-istry,', Chap. VIII. (Wiley, New York, 1943.)141 P. A. Schaffer, J. Amer. Ghem. SOC., 1933,55,2169.142 J. Physical Chem., 1936,40, 1021.Trans. Electrochem. SOC., 1937, 71, 107; Ann. Rev. Biochem., 1938, 7, 1 ; J .Amer. Chem. Xoc., 1937, 59, 1246156 ORGANIC CHEMISTRY.E.Clar and F. John 144 introduced phenanthraquinone in boiling nitrobenz-ene, and chloranil in boiling xylene, as convenient reagents for the smoothdehydrogenation of hydroaromatic hydrocarbons, and the same methodhas since had extended application by R. Criegee 145 and by R. T. Arnoldand C. J. The latter workers point out that as a cheap, clean,dehydrogenating agent, chloranil, which can easily be made by oxidisingquinone in concentrated hydrochloric acid with perhydrol, is often superiorto selenium, which reacts only at much higher temperatures.R. Criegee 145 showed that quinones reacted with tetralin to give quinoltetralyl ethers, RO*C,H,*OH, which broke down at higher temperatures todihydronaphthalene and quinols. W. A. Waters 135 has shown that quinonesare partial inhibitors of the autoxidation of tetralin, and considers that thisis due to the facile combination of a-tetralyl and semi-quinone radicals,which leads to enhanced chain-breaking in the oxidation cycle, but pointsout that initially the quinone must be considered as abstracting atomichydrogen from the tetralin.Other Oxidising Agents.It has been impossible to consider in this report the mechanisms of actionof many oxidising agents commonly used in organic chemistry.There isoften no experimental evidence to substantiate theories that can be advancedon paper.Potassium perrnungunate, for instance, is usually regarded as the simplesthydroxylating agent for olefins, but attack on C-H bonds, e.g., of aromaticside chains, is one of its general uses.Many oxidations by permanganate,as for instance the volumetric permanganate-oxalate reaction, are essentiallychain processes in which radicals must presumably be formed, and freehydrocarbon radicals are evidently produced when permanganate is made tooxidise tetralin. It would be premature, however, to classify potassiumpermanganate and lead tetra-acetate or benzoyl peroxide as reagents of anentirely similar type.Selenium dioxide may perhaps be an exception to the Michaelis rule,though from all its reactions with organic compounds 14' scarcely any evidencecan be adduced for its mode of action, H. L. Riley 14* has suggested thatunstable intermediate compounds of selenium are often formed, but haspointed out that subsequently there must occur a very complicated stagewhich brings about the combination of carbon and oxygen as a C-0 group.N.N. Melnikov and M. S. Rokitskaya 149 have shown that selenium dioxidereacts with alcohols to form alkyl selenites, which at 300" decompose told4 Ber., 1930, 63, 2967.146 J . Amer. Chem. SOC., 1939, 61, 1407; cf. ibid., 1940, 62, 983.14' For reviews, see G. R. Waitkins and C. W. Clark, Chem. Reviews, 1945, 36,l48 Sci. J . Roy. Coll. Science, 1935, 5, 7 ; S. Austin, L. de V. Moulds, and H. L.14e J . Qen. Chem. RUGS., 1937, 7, 1532.145 Ibid., 1936, 69, 2758.235.RiIey, J . , 1935, 901OWEN : FURANS.aldehydes, selenium, and water.perhaps be similar :The attackI I ,OH I L/?! _------------ t, I /157on methylene groups may>CH2 + SeO, + >C/ /,S;? -+ >CO +Se+H,O\O/l--------------The fact that the oxidations of methyl and carbinol groups, as in acetoneand alcohol, stop a t the aldehyde stage rather discounts dehydrogenationhypotheses, since aldehydes are usually attacked by free radical reagentsfar more easily than is methyl.In considering the mechanisms of these and other oxidations, the theorieewhich have been described on the preceding pages may serve to indicatenew lines of approach for future crucial experiments.The rapidity of de-velopment of this whole field has been remarkable, for the very conceptof the intervention of transient free radicals in simple organic reactionsin solution was startlingly novel but ten years ago.W. A.W.5 . FURANS.The chemistry of furan has undergone a remarkable expansion withinthe last 25 years, owing in the main to an increasing realisation of the potenti-alities awaiting investigation within this field. It is not intended in thisReport to give any account of the multitudinous uses to which furan deriv-atives have been applied in the commercial production of solvents, pre-servatives, fungicides, dyestuffs, etc., particularly since brief reviews of suchindustrial applications are already available.lY Rather has an attempt beenmade to survey some of the more important reactions which have been studied,both from the point of view of the formation of furan derivatives, and alsowith regard to their synthetic uses, such as the preparation of aliphaticcompounds by ring scission.Formation.-Although considerable interest has recently been shown inthe preparation of furans from acetylene (see p.lSO), the principal source isstill to be found in the naturally occurring pentosans,2 such as those presentin oat-hulls, which when subjected to acid hydrolysis yield pentoses, andfinally furfuraldehyde, from which most of the other derivatives can beprepared. Other materials of carbohydrate nature may also serve assources of furans, but it is only recently that a thorough investigation hasbeen made into the optimum conditions for the preparation of 5-hydroxy-methylfurfuraldehyde from sucrose.have obtained a 54% yield by the use of 0.25% aqueous oxalic acid a t 130"under pressure, the yield being based on the fructose portion of the molecule,since the glucose portion takes no part in the reaction.4 This is probablyW.N. Haworth and W. G. M. JonesP. N. Peters, I n d . Eng. Chem., 1936,28, 755; 1939,31, 178.A. Wacek, Angew. Cheve., 1941, 54, 453.Compare A. D. Braun, Biokhimiya, 1939,4, 276; B. L. Scallet with J. H. Gardner,J., 1944, 667.J. Amer. Ghem. SOC., 1945, 67. 1934158 ORGANIC CHEMISTRY.due to the fact that fructose has a much greater tendency than glucose toreact in the furanose form (I), which by successive dehydrations can readilypass into 5-hydroxymethylfurfuraldehyde (11). They have also shown thatglucose (111) will undergo the reaction, provided it is pretreated with dilutealkali to facilitate the formartion of the 1 : 2-enediol (IV), which by loss ofwater can give (V), a postulated intermediate in the above scheme.HO- OHHO~CH,! )-OH 4 HOTOH --+0 \CH,*OH HO-CH,! ,LCH.OH HOGH,QO!/CHO 0(11.)(I?.) \ (1.)HO*VH-YH*OH HO vH-vH*OHOH OH OH OH(111.) (IV.)HO*CH,*VH YH*CHO + HO*CH,*yH $XCH*OHOptimum conditions have also been worked out for the preparation of5-chloromethylfurfuraldehyde by the action of hydrogen chloride onsucr~se.~ M.L. Wolfrom, E. G. Wallace, and E. A. Metcalf have shownthat acid treatment of 2 : 3 : 4 : 6-tetramethyl d-1 : 2-glucoseen gives5-methoxymethylfurfuraldehyde. The electronic interpretation of suchreactions has been discussed by H. S. Isbell.6Carbohydrates, however, can be utilised in another way for the pre-paration of furan derivatives.The condensation of glucose or mannose with1 : 3-diketones, HA*CO*CH,*CO*R’, in the presence of zinc chloride gives avariety of substituted 5-tetrahydroxybutylfurans (VI), the constitutionsof which have been proved by oxidative degradation with lead tetra-acetate 79 s or periodic acid to give 5-formylfurans (VII), which on furtheroxidation with silver oxide yield substituted furoic acids (VIII). Other- C0.R’ -- CO*R’ --CO*R’+ OHC~!, )IR + HO,C/!~)/R HO*CH2*[CH*OH]3*[o)1R 0(VI.) \ (VII.) (VIII.)-CO,H/\CO.Rt -1 \HO(),o.h3 OH H0,C[O)1C02H(X.1sugars do not readily take part in this condensation.stated to undergo no reaction with acetylacetone,s but JonesGlucose was originallyhas shownJ . Amer. Chem. SOC., 1942, 64, 265.J. Res.Nut. Bur. Stand., 1944, 32, 45.S. Muller and I. Varga, Ber., 1939, 72, 1993; I. Varga, Chem. Abs., 1941,35,1034.8 T. Szeki and E. Laszlo, Ber., 1940, 73, 924. J. K. N. Jones, J., 1945, 116OWEN : FURANS. 169that it readily gives (VI; R = R’ = Me), which rearranges with lossof water in boiling dilute acid solution to form the pyran derivative (IX;R = R’ = Me). The product (VI; R = Me, R’ = OH) from ethyl aceto-acetate behaves similarly. Furan-2 : 3 : 5-tricarboxylic acid (X) can beobtained by permanganate oxidation of the condensation product (VI ;R = CH,*CO,Et, R’ = OEt) from glucose and ethyl acetonedicarboxylate.8This direct oxidation of the methylene group in the side chain a t C, is unusual,since the oxidation of alkylfurans usually results in extensive degradation.It is possible that the necessary stability is secured by the prior formationof carboxyl groups a t C, and C,, since E.V. Brownlo has shown that5-methylfuroic acid, unlike 2 : 5-dimethylfuran, can be oxidised to furan-2 : 5-dicarboxylic acid by ferricyanide. If this is so, oxidation of (VIII)to (X) should be possible.S. Archer and M. G . Pratt l1 have investigated the condensation of ethylbromopyruvate with ethyl P-ketosuberate (XI ; R = [CH,],*CO,Et).If C-alkylation occurs, as originally postulated by H. Sutter l2 for a similarcondensation between ethyl bromopyruvate and ethyl oxalacetate, theproduct would be ethyl 3 : 5-dicarbethoxyfuran-2-valerate, but i t is clearthat the reaction proceeds by O-alkylation, since it gives the isomeric 3 : 4-di-carboxylate (XII; R = [CH,],*CO,Et).The acid obtained on saponific-ation is identical with that prepared by K. Hofmann l3 by an applicationof a reaction studied by K. Alder and H. F. Rickert.14 The latter authorsshowed that whilst the adduct (XIII) from furan and maleic anhydride isreconverted into its components on heating, the ester (XIV; R = H),obtained by semihydrogenation of tlhe adduct from furan and ethyl acetyl-enedicarboxylate, loses ethylene and gives ethyl furan-3 : 4-dicarboxylate(XII; R = H).”R*CO*CH,*CO,Et (XI.) R*F:CH*CO,Et Rlo Iowa State Coll. J . Xci., 1937, 12, 227.l1 J . Amer. Chem. SOC., 1944, 66, 1656.l2 Annulen, 1932, 499, 54; compare T. Reichstein, A. Griissner, K.Schindler, andE. Hardmeier, Helv. Chim. Acta, 1933, 16, 276.lS J . Amer. Chem. SOC., 1944, 68, 51. * W. Nudenberg and L. W. Butz ( J . Amer. Chem. SOC., 1944,86, 307) have obtained3 : 6-epoxycyclohexene, the parent compound of (XIV), by condensation of furan withethylene under high pressure, thus demonstrating the reversibility of the second typeof reaction.l4 Ber., 1937, 70, 1354160 ORGANIC CHEMISTRY.By condensation of furan-2-valeric acid with ethyl acetylenedicarboxyl-ate, followed by semihydrogenation, the ester (XIV; R = [CH2],*C02H)is obtained, which on pyrolysis gives (XI; R = [CH2],*C0211). Similarly,furan-2-pentanol gives (XI1 ; R = [CH,],*CH,*OH), an important inter-mediate in the synthesis of O-heterobiotin (see p. 165).The investigations of R.E. Lutz on the preparation of heavily substitutedfurans have been ~0ntinued.l~ By reductive cyclisation of an unsaturateddiketone (XV), the furan (XVI) can be prepared, the saturated diketone(XVII) probably being an intermediate product. l6 The reverse reaction,oxidative fission, may be accomplished by treatment of some furans with anitric acid-acetic acid mixture. This usually gives the cis-form of theunsaturated diketone, and by the application of these reactions it has beenpossible to prepare certain &-compounds which are difficult to obtain bydirect isomerisation of the trans-form. For example, trans-dibenzoyl-methylethylene (XV; R = Ph, R’ = Me) is converted in high yield into2 : 5-diphenyl-3-methylfuran (XVI; R = Ph, R’ = Me), which on oxidativescission gives the pure ~is-is0rner.l~ The effect of mesityl groups on thestability of the furan ring is exemplified by the resistance of 2 : 5-dimesityl-furan (XVI; R = C9Hll, R’ = H) towards fission with nitric acid.l*Although it has not been found possible to isolate free 3-hydroxyfurans,probably owing to their tautomerism with 3-ketodihydrof~ran.s,~~ the3-acetoxy-derivatives (XVIII), prepared from the unsaturated diketonea(XV) by the action of acetyl chloride, react with a Grignard reagent togive (XIX), from which, by treatment with the appropriate halides, several3-alkoxy- and 3-acyloxy-compounds (XX) have been obtained.20It has recently been shown21 that substituted furans can readily beprepared from ethylenic ethynylcarbinols (XXI), obtained by the condens-ation of ap-unsaturated aldehydes with acetylene.Compare T.S. Stevens, Ann. Reports, 1941, 38, 315.R. E. Lutz and C. J. Kibler, J . Amer. Chem. SOC., 1939, 61, 3007; R. E. Lutzand W. G. Reveley, ibid., 1941, 63, 3180; R. E. Lutz and W. P. Boyer, ibid., p. 3189.l7 R. E. Lutz and C. E. McGinn, ibicl., 1942, 64, 2585.R. E. Lutz and C. J. Kibler, ibid., 1940, 62, 1520.lQ E. P. Kohler, F. H. Westheimer, and 31. Tishler, ibid., 1936, 58, 264; E. P.2o R. E. Lutz, C. E. McGinn, and P. S. Bailey, ibid., 1943, 65, 843; R. E. Lutz and*l I. M. Heilbron, E. R. H. Jones, Peter Smith, and B. C. L. Weedon, J . , 1946, 54.Kohler and D. W. Woodward, ibid., p. 1933.C. E. McGinn, ibid., p. 849OWEN : FURANS.161On treatment with acids, these carbinols rearrange22 to give products(XXII) which when steam distilled in the presence of mercuric chloride givefurans (XXIII). Propenylethynylcarbinol (XXI ; R = Me, R’ = H)gives a 55% yield of 2 : 5-dimethylfuran (XXIII; R = Me, R’ = H),*and from 4-ethyloct-4-en-1-yn-3-01 (XXI; R = Pr, R’ = Et) a 64% yieldof 5-methyl-3-ethyl-2-propylfuran (XXIII ; R = Pr, R’ = Et) can beobtained. This process can be regarded either as a direct intramolecularhydration to (XXIV), followed by isomerisation, or as a normal hydrationof the acetylenic linkage to give (XXV), followed by cyclodehydration.1 : 4-Diketones (XXVI) are formed simultaneously.R’*C-CH*OH R’*((==FH ,/RCH CiCH‘%_3R*CH CiCH II 1(XXI.) I ‘\OH(XXII.)R*CO*CHR’*CH,*CO*Me (XXVI.)R’RInlMe‘0’(XXIIT. )The tautomerism exhibited by hydroxy-aldehydes and hydroxy- ketones 23is well exemplified by y- hydroxycarbonyl compounds (XXVII), whichexist almost entirely in the form of 2-hydroxytetrahydrofurans (XXVIII).On treatment with meth anolic hydrogen chloride, the corresponding‘‘ furanosides ” (XXIX) are formed, whilst dehydration gives 2 : 3-dihydro-furans (XXX).The latter react readily with water to regenerate thehydroxyfurans. The simplest member of this series (XXVIII ; R = R’ = H)has recently been prepared by two new routes, vix. the oxidation of pent-ane-1 : 2 : 5-trio1 with lead tetra-acetate or periodic acid,M and the hydrogen-ation of butyne-1 : 4-diol (see below). J. R.Stevens and G. A. Stein25R(0)<Ek --+ R r l < R ’(XXVIII.) (XXIX.)R*CH( OH)*CH,*CH,*CO*R‘ 6-(XXVII.) \o/ OMeRQR’ (XXX.)22 E. R. H. Jones, Ann. Reports, 1944, 41, 175.z3 L. N. Owen, ibid., p. 140.24 R. Paul, Gompt. rend., 1941, 212, 492; 1942, 215, 303; Bull. SOC. chim., 1941, 8,911.z6 J . Amer. ChemSoc., 1940,62,1045; U.S.P. 2,123,653; see also D.R.-P. 705,034;723,052.* 2 : 5-Dimethylfuran is obtained industrially as a by-product in the preparationof acetaldehyde from acetylene (Reports Appl. Chem., 1938, 23, 142) or by the pyrolysisof acetone at 700” (U.S.P. 2,098,592).REP.-VOL. XLII. 162 ORGANIC CHEMISTRY.have obtained 3-chloro-2-ethoxy-2-methyltetrahydrofuran (XXXI) by theaction of boiling acid-ethanol on chloroacetobutyrolactone (XXXII) :CH,*CH,*CCl*CO*Me c1 I + [CH,(OH)*CH,*CHCl*CO*Me] --+ Me(XXXII.) (XXXI.)\O/<OEt co 0---A closely related instance has been encountered by I.M. Heilbron, E. R. H.Jones, and H. P. Koch 26 who have shown that semihydrogenation of theacetylenic acetal (XXXIII) gives, instead of the expected ethylenic compound(XXXIV), 2-ethoxy-5-methyl-5-ethyl-2 : 5-dihydrofuran (XXXV).I'YC( OH)*CiC*CH( OEt), + Me (XXXIII.)Reference was made in a recent Report 22 to the formation of di- andtetra-h ydrofurans from acetylenic glycols under conditions of hydration,halogenation, hydrogenation, etc. The commercial production of but-2-yne-1 : 4-diol (XXXVI), by the condensation of acetylene with formaldehyde inthe presence of a copper or silver acetylide catalyst,27, 28 has led to a renewedinterest in its hydrogenation products, but-2-ene-1 : 4-diol (XXXVII)and butane-1 : 4-diol (XXXVIII), since dehydration of these glycols overacid catalysts 29 gives dihydro- and tetrahydro-furan, respectively.It hasalso been shown that hydrogenation of butynediol a t 100" over a palladiumcatalyst, on an acid carrier, gives 2-hydroxytetrahydrof~ran.~~?--- (7 H,CH2*OH CH2*OH --+(XXXVI.) (XXXVII.) 4- Ha0I=={\O'?H=--Xp HZ 7H2-?H2CH2*OH CH2*OH -+ CH2*OH CH2*OH(XXXVIII.)I 4- HaoA striking example of t,,e different behaviour of cis- ank tram-unsaturateddiols has been provided by J. R. Johnson and 0. H. whohave shown that whereas trans-2 : 5-dimethylhex-3-ene-2 : 5-diol with 15%sulphuric acid gives only 2 : 5-dimethylhexa-1 : 3 : 5-triene, the cis-isomergives an excellent yield of 2 : 2 : 5 : 5-tetramethyldihydrofuran.E.Beati and G. Mattei 32 have studied the conditions for the dehydrationof pentane-1 : 4-diol to 2-methyltetrahydrofuran, whilst C. S. Marvel andE. H. Dunlap 33 have shown that reduction of ethyl aa'-dipropylsuccinate26 J., 1942, 735.28 Ind. Eng. Chem. (News), 1945, 23, 1516, 1840, 1846.28 B.P. 508,548, 510,949; U.S.P. 2,251,835, 2,251,895.31 J . Amer. Chem. SOC., 1940, 82, 2615.s3 J. Amer. Chem. SOC., 1939, 61, 2714.2' l3.R.-P. 721,004.30 U.S.P. 2,333,216.32 Ann. Chim. appl., 1940, 30, 21OWEN : FURANS. 163over copper chromite at 260" gives 3 : 4-dipropyltetrahydrofuran, the expected1 : 4-diol undergoing dehydration under these conditions.The latter ob-servation is of interest, inasmuch as copper chromite a t high temperaturesfrequently favours the reductive scission of the tetrahydrofuran ring (seep. 167).Long chain ao-diols were originally considered 34 to give 2-alkyltetra-hydropyrans on dehydration with sulphuric acid, but more recently 3 5 9 3 6 ithas been shown that the main product of this reaction is the 2-alkyltetra-hydrofuran (XXXIX). It would therefore appear that the alleged formationHO*[CH,l?I*OH + (01.[CH21n-s*Me (XXXIX.)of tetrahydropyran from pentane-1 : 5-diol under these conditions 37 isprobably erroneous; the physical constants of the product suggest, in fact,that it is 2-methyltetrahydrofuran7 and a further investigation of thisdehydration would be of interest.Preferential formation of the five-membered ring system is also to be found in the dehydration of pentane-1 : 2 : 5-tri01,~~ which gives tetrahydrofurfuryl alcohol,* and in the isomeris-ation of y8-unsaturated alcohols by treatment with concentrated sulphuricacid ; pent-4-en-1-01, for example, gives an 88% yield of 2-methyltetra-hydr~furan.~~ Closely related is the dehydrobromiiiation of 4 : Ei-dibromo-pentanol, which forms only tetrahydrofurfuryl bromide.38 Ring closure byremoval of hydrogen halide has also been studied by C. L. Wilson,40 whohas shown that 1 : 5-dibromopentan-2-01 (XL) gives approximately equalamounts of tetrahydrofurfuryl bromide and y-bromopropylethylene oxide,whereas 1 -chloro-5- bromopentan-2-01 (conveniently prepared from the oxideby treatment with hydrochloric acid) gives mainly tetrahydrofurfurylchloride, which indicates that removal of hydrogen bromide is more facilethan that of hydrogen chloride :$L)CH2Cl Br*[CH,],*CH( OH)*CH,Cl34 A.Franlre and A. Kroupa, fionatsh., 1930, 56, 340.s b Idem, ibid., 1936, 69, 167.36 J. K. Juriev, V. I. Gusev, V. A. Tronova, and P. P. Yurilin, J. Gen. Chein. Russia,37 N. Demjanov, J. Russ. Phys. Chem. SOG., 1890, 22, 388.38 R. Paul, Ann. Chim., 1832, 18, 303; see also ref. 97.39 R. Paul and H. Normant, Bull. SOC. chiin., 1944, 11, 365.* Further dehydration of the latter, however, over alumina at 300-350" gives a70% yield of dihydropyran (U.S.P. 2,365,623 ; Org. Synth., 1943, 23, 25 ; C.H. Klhoand J. Turkevich, J . Amer. Chem. SOG., 1945,67, 498) which is known t o be more stableat high temperatures than thc expected 2-methylenetetrahydrofuran (R. Paul, Bull.SOC. chim., 1935, 2, 745).1941, 11, 344.40 J., 1945, 48164 ORGANIC CHEMISTRY.Cycliaation of py-dibromoalcohols, CH,Br*CHBr*CH,*CH(OH)*R, yields4- bromo-2-alkyltetrahydrofurans,411 42 which on further dehydrobrominationwith alcoholic alkali give 2-alkyl-2 : 5-dihydrofurans (XLI).Amino-derivatives.-Comment was made in an earlier Report 43 on thedifficulty which has been experienced in the preparation of simple amino-furans in which the amino-group is directly attached to the nucleus. The fewcompounds of this type which have been described are very unstable, andit has been suggested that this is due to tautomeric change into the imino-form.It would therefore be anticipated that aminotetrahydrofurans wouldnot show such instability, particularly if the amino-group were in the 3-or 4-position; this is true of the three such compounds known, all of whichwere prepared by cyclisation of 2-aminopolyalkyl-1 : 4-diols. No diamineshad been obtained before the recent work of K. Hofmann and A. Bridg-water. These authors 45 have converted 2-methylfuran-3 : 4-dicarboxylicacid (I) via the acid chloride and azide into 3 : 4-di(carbethoxyamino)-2-methylfuran (11) 46 which is hydrogenated in acetic acid solution over apalladium-barium sulphate catalyst to give the tetrahydro-derivative (111).When this is boiled with 10% aqueous barium hydroxide it yields the bicycliccompound (IV) instead of the expected diamine (V).The latter is formedonly under more drastic conditions, by heating with barium hydroxide underpressure at 140°, and it can be reconverted into (IV) by treatment withcarbonyl chloride in sodium bicarbonate solution. This cyclisation wouldbe expected to occur only with a cis-disposition of amino-groups a t C, andC,, and it is significant that the conditions of hydrogenation were such as tofavour the formation of a cis-compound./\NH NH/\NH NH(VI.) (V. 1 W.)This simple method of preparation is of particular interest in view of theclose relationship of the new ring system to that present in p-biotin, and by a41 0. Kiun-Houo, Ann.Chim., 1940,13, 175; D.R.-P. 696,725.4a E. D. Amstutz, J. Org. Chenz., 1944, 9, 310.43 R. D. Haworth, Ann. Reporb, 1939,36, 310.44 M. Kohn and A. Ostersetzer, Monatsh., 1916, 37, 47; S. Kanao and S. Inagawa,J. Pharm. SOC. Japan, 1928,48, 238. 3-Arylarninotetrahydrof~rans~ prepared from the3-chloro-compound, are described in U.S.P. 2,278,202. 2-Amino- and 3-amino-tetra-hydrofuran are mentioned as typical amines in U.S.P. 2,150,422, but there is no evidencethat they have actually been prepared.45 J. Amer. Chem. SOC., 1945, 67, 738, 1165. 4~ Compare G. Stork, ibid., p. 884OWEN: FDRANS. 166;further application of these reactions47 it has been possible to prepare theO-analogue of p-biotin. For this purpose, 3 : 4-dicarbethoxyfuran-2-pentanolwas synthesised by the method already discussed on p.160, and convertedby the above process into (VI; R = CH,*OH); this on oxidation withchromic acid gave O-heterobiotin (VI; R = CO,H), which differs from@-biotin only in having a tetrahydrofuran in place of a tetrahydrothiophenring. This result has beenconfirmed by R. Duschinsky, L. A. Dolan, D. Flower, and S. H. Rubin.,*The Hofmann method of synthesis, described above, should obviouslybe applicable also to the preparation of monoaminotetrahydrofurans.MercuriaZs.-The ease with which furan and its derivatives can bemercurated, coupled with the great reactivity of the mercuri-group so intro-duced, renders these compounds of particular value for synthetic purposes.Mercuration occurs most readily in the 2- and 5-positions.By controlledtreatment of furan with mercuric chloride and sodium acetate, 2-chloro-mercurifuran (I) is obtained; under more vigorous conditions the 2 : 5-di-and 2 : 3 : 4 : 5-tetra-chloromercurifurans are formed.49 It has been shownthat an alkyl group a t C, or C , orientates the chloromercuri-group into acontiguous 2- or 5-position, if available ; thus, 3-isopropylfuran gives 2-chloro-merc~ri-3-isopropylfuran.~~ Furfuryl alcohol gives 5-chloromercurifurfurylalcohol,s1 and methyl 5-bromofuroate undergoes mercuration in the 4-positionto give (II).62 A chloromercuri-group can also be introduced by replace-ment of a carboxyl residue, provided that this is in the 2- or 5-position,by treatment of the sodium salt of the acid with mercuric chloride.49, 53 Themercuri-chlorides are readily decomposed by acid, the group being replacedby hydrogen, and it is only necessary to treat a furan-2- or-5-carboxylic acidwith mercuric chloride in order to bring about a smooth decarboxylation,since the free hydrochloric acid produced during mercuration is sufficient tobring about the subsequent decomposition.49* 60 Some further typicalreactions involving the replacement of a chloromercuri-group are shownbelow.For simplicity, 2-chloromercurifuran (I) has been chosen as theexample, but alkyl-, hydroxymethyl-, halogeno-, and carbomethoxy-derivatives behave in the same way. Of particular interest is the reactionwith arsenic t r i ~ h l o r i d e , ~ ~ ~ 54 which can be controlled to give furyldichloro-,difurylchloro-, and trifuryl-arsines, compounds which have not been success-The product showed high yeast growth activity.4 7 K.Hofmann, J. Amer. Chem. SOC., 1944, 66, 157; 1945, 67, 421, 1459.4 8 Arch. Biochem., 1945, 6 , 480.4 9 H. Gilman and G. F. Wright, J. Amer. Chem. SOC., 1933, 55, 3302; compare50 H. Gilman, N. 0. Callowny, and R. R. Burtner, J. Amer. Chem. SOC., 1935, 57,6 1 W. J. Chute, W. M. Orchard, and G. F. Wright, J . Org. Chem., 1941, 6, 157.5 2 W. W. Beck and C. S. Hamilton, J . Amer. Chem. SOG., 1938,80, 620.s3 H. Gilman and R. J. Vanderwal, Rec. Trav. chim., 1933,52,267; H. Gilman andR. R. Burtner, J. Amer. Chem. Soc., 1933, 55, 2903; W. G. Lowe and C. S. Hamilton,.ibid., 1935, 57, 2314.H. Gilman and N.0. Calloway, ibid., p. 4197; R. Paul, Compt. Tend., 1935, 200, 1481.906.S4 Idem, ;bid., p. 1081166 ORGANIC CHEMISTRY.fully prepared by any other route. Conversion of (I) into the bisfurylmercury(I11 ; R = H) can be effected by hydrazine, diazomethane, sodium iodide, orIII'0'ClHg-- -BAl lICO,MeFurfurylchloride\O/(11.)thiosulphate.49~ 51 The product (I11 ; R = CH,*OH), from 5-chloromercuri-furfuryl alcohol, on oxidation with potassium permanganate gives bis-(5-formylfuryl-2)-mercury (I11 ; R = CHO), which when treated with mercuricchloride in boiling ethanol gives 5-chloromercurifurfuraldehyde ( IV).51The chloromercuri-derivatives are useful for the characterisation of furans,and can be quantitatively estimated by titration with iodine.Furan-3-mercurials generally undergo the same reactions as their 2-isomerides,but not so readily.Reduction.-The rules which apply to the hydrogenation of benzenoidcompounds are generally applicable to furan derivatives, with the reservationthat the ease of fission of the latter may lead under vigorous conditions to theformation of open-chain 559 56 Hydrogenation of the furannucleus usually proceeds very readily under pressure in the presence of Raneynickel a t 150", but a noteworthy exception has been encountered by W. N.Haworth, W. G. M. Jones, and L. F. wig gin^,^' who have failed to reducemethyl furan-2 : 5-dicarboxylate. The corresponding acid requires atemperature of 235" for hydrogenation to occur, and gives a polymerisedproduct ; by boiling with methanolic hydrogen chloride this is convertedinto the monomeric ester, from which a poor overall yield of tetrahydrofuran-2 : 5-dicarboxylic acid is obtained on saponification.Polymer formation alsooccurs in the hydrogenation of 5-hydroxymethylfuroic acid, but the sametechnique of depolymerisation through the ester gives a good yield of themonomeric tetrahydro-acid. In this case, however, polymerisation can beentirely avoided by the use of methyl 5-hydroxymethylfuroate, whichhydrogenates smoothly over Raney nickel a t 140".If the molecule contamins an easily reducible side chain, this also is hydro-65 Compare H. E. Burdick and H. Adkins, J . Amer. Chem. Xoc., 1934, 56, 438;5 6 G. Natt.n, R. Rigamonti, and E.Beati, Chim. e Z'Ind., 1941, 23 117.6 7 J., 1945, 1." Organic Chemistry " (ed. Gilman), 2nd. edition, p. 779OWEN : FURANS. 167genated under the conditions necessary to reduce the nucleus. Furfuryl-ideneacetone, for example, gives 2- (3'-hydroxybutyl)tetrahydrofuran j8, 59whilst 5-hydroxymethylfurfural gives 2 : 5-bishydroxymethyltetrahydro-furan.57Reduction of the nucleus can be avoided by the use of mild conditionsor less active catalysts. W. Huber 60 has shown that fury1 cyanide, pre-ferably in the presence of ammonia, gives furfurylamine when hydrogenatedover Raney nickel a t room temperature. The same amine can be preparedby the controlled hydrogenation of furfuraldehyde in methanolic ammonia,61and also from furfuraldoxime ; 62 under more drastic conditions, tetrahydro-furfurylamine is formed.63 It has been shown 58 that the hydrogenation offurfurylideneacetaldehyde (I) over Raney nickel to 2-(3'-hydroxypropy1)-tetrahydrofuran (11) proceeds in two stages, the side chain being reduceda t 80", and the nucleus a t 175"; heptane-1 : 3 : 7-trio1 is said Lo be obtainedas a by-product, but there is no evidence to support this formulation ratherthan that of the 1 : 4 : 7-isomer, the formation of which would not be sodifficult to explain.The reduction of furfuraldehyde to furfuryl alcohol and to tetrahydro-furfuryl alcohol has been extensively 55, 569 64, 65 The me of anickel-cobalt catalyst has been recommended for the first stage.Thiscatalyst possesses enhanced activity for side-chain reductions, and undermore vigorous conditions can be used for the preparation from furfuraldehydeof 2-methylfuran and 2-methyltetrahydrofuran.56 I n the benzenoid series,copper chromite is recognised to be particularly effective for the preferentialreduction of side chains, but when applied to furan compounds the temper-ature must be carefully controlled to avoid scission of the heterocyclic ring.For example, furfuraldehyde is rapidly and almost quantitatively reducedto furfuryl alcohol over copper chromite a t 150" under high pressure,j5 butif the temperature is raised it gives a mixture of pentanedi~ls.~~ C.L.Wilson,66,67 however, has shown that 2-methylfuran can be obtained in80% yield by the hydrogenation of furfuraldehyde in the vapour phaseover copper chromite a t 280".Mention must also be made of the applica-tion by A. M. Berkenheim and T. F. Dankova 68 of the general method of5 8 A. Hhz, G. Meyer, and G. Schucking, Ber., 1943, 76, 676.13* W. Huber, ibid., 1944, 66, 876. '' E. J. Schwoegler and H. Adkins, ibid., 1939, 61, 3499; U.S.P. 2,109,159; Org.6 2 R. Paul, Bull. Xoc. chim., 1937,4, 1121.63 U.S.P. 2,338,655. U.S.P. 2,201,347.65 I. B. Rapoport and B. Rapoport, J. Appl. Chenr. Russia, 1938, 11, 723.'I3 J . , 1945, 61.68 J. Gen. Chem. Russia, 1939, 9, 924.R. D. Kleene, J. Amer. Chem. SOC., 1941, 63, 3539.Synth., 1943, 23, 68.6 7 Compare ref. 85168 ORGANIC CHEMISTRY,reducing aldehydes by treatment with formaldehyde and alkali. In thisway they have prepared furfuryl alcohol in 90% yield from furfuraldehyde.It is usually not possible to effect reduction of the nucleus without affect-ing unsaturated centres in the side chain, unless these are suitably protected.The preparation of tetrahydrofurfural has proved to be particularly difficult)and a satisfactory method has yet to be discovered.The hydrogenation offurfuraldehyde diethylacetal or diacetate, followed by removal of theprotecting groups, gives only a very poor yield, and in the experience ofrecent workers is ~nreliable.~B$ 69 Small amounts have been obtained bycatalytic dehydrogenation of tetrahydrofurfuryl alcohol, and by oxidationof the octahydropinacol(II1) with lead tetra-acetate : 69~-~*cH(oH)cH(oH)~J \O ,,J + QHO(111.)'0'No aldehyde was obtained by reduction of 2-cyanotetrahydrofuran withstannous chloride, reduction of barium tetrahydrofuroate with bariumf ~ r m a t e , ~ ~ oxidation of tetrahydro€urfuryl alcohol, or saponification of2 - dic hlor o me t h yl t e t r ah y dr o f uran .70Elimhalion of Bide Chains.-C.L. Wilson 66, 71 has shown that a ttemperatures above 200°, preferably a t 280", furfuraldehyde vapour is de-composed into furan and carbon monoxide on contact with a catalyst con-taining nickel or cobalt. An interesting observation is that the yield offuran may be raised to 65% by the introduction of a limited quantity ofhydrogen (about 2/3 mol. per mol. of aldehyde); as would be anticipated,a small amount of 2-methylfuran is also formed under these conditions, but,apart from that required for this minor reaction, all the hydrogen issuesunchanged.The added hydrogen also has the effect of greatly prolongingthe effective life of the catalyst, which otherwise deteriorates rapidly.Nitrogen or carbon dioxide is ineffective. Nickel gauze is the most suitablecatalyst for continuous operation, since it has a long active life and gives a50% yield of furan, together with 8% of 2-methylfuran. Monel metal gives65% of furan with only 2% of 2-methylfurnn, but unfortunately deterioratesmuch more rapidly. The conversion of furfuraldehyde into furan is alsoreported to occur over lime a t high temperature^.^^in whichfurfuryl alcohol, passed over Raney nickel a t 150") is shown to give a mixtureof furfuraldehyde, furan, and 2-methylfuran.The reactions involved areshown below, the hydrogen necessary for the production of the 2-methylfuranbeing derived from the dehydrogenation of furfuryl alcohol :Closely related to the above are the experiments of R.69 C. L. Wilson, J., 1945, 52.70 H. Paillard and R. Szasz, Helv. Chim. Acta, 1943, 26, 1856. 'l B.P. 553,175.7 2 U.S.P. 2,337,027. 7' Bull. SOC. chim., 1938, 5, 1692; 1941, 8, 607OWEN : FURANS. 169The side chain in tetrahydrofurfuryl a,lcohol, also, is eliminated whenthe vapour is passed over various nickel catalysts, preferably nickel gauze,a t about 260°.69, 74 The principal product is tetra,hydrofuran, accompaniedby carbon monoxide and hydrogen in approximately equimolecular amounts.I n this reaction the addition of hydrogen is unnecessary; presumably thehydrogen formed during the pyrolysis is capable of preventing deactivationof the catalyst.By-products include tetrahydrofuryl tetrahydrofurfurylether (I) and 2 : 3-dihydrofuran (11). The formation of (11) is not readilyexplicable, since it has been proved by direct experiment that tetrahydrofuranis not dehydrogenated to dihydrofuran over nickel. The ether (I) canobviously arise by condensation of dihydrofuran with tetrahydrofurfurylalcohol. The dihydrofuran content of the crude tetrahydrofuran obtainedis usually about 4%, but may be increased to 38% by the use of a cupro-nickel catalyst. This substance is also formed 75 when the vapour of either2-cyanotetrahydrofuran (111) or methyl tetrahydrofuroate (IV) is passedover a dehydrating catalyst a t 300--400", the reaction proceeding by lossof hydrogen cyanide, or of carbon monoxide and methanol, respectively.A t higher temperatures, cyclopropanealdehyde is produced at the expenseof dihydrofuran, from which it is probably formed by rearrangement :i-71'0'I ) C H O +-(11.)Nuclear Transformations.-Brief mention was made in an earlier Report l5of the conversion of furan derivatives into pyrroles and thiophens.Theseinvestigations have been continued by several workers, and sufficientinformation has now been acquired to render possible the evaluation ofthe reactions for preparative purposes, and to give some insight into theprobable mechanisms involved.Treatment of furan, in the vapour phase, with ammonia or with alkyl-or aryl-amines, a t 400-450" over alumina or alumina-chromia, gives 25-30% yields of pyrrole or N-substituted pyrr01es.~~ Although at first sightthese results appear to be unsatisfactory, account must be taken of the recentwork, discussed in the previous section, which has led to tlhe ready avail-ability of furan.C. L. Wilson 77 has investigated the nature of the high-boiling by-products which are formed during the preparation of pyrrole by74 B.P. 550,105.76 D.R,.-P. 706,095.75 C. L. Wilson, J . , 1945, 58.7 7 J., 1945, 63.F 170 ORGANIC CHEMISTRY.this method, and has shown that indole, carbazole, and pyrrocoline (I) arepresent. He considers that these are probably formed by/\--- interaction of pyrrole with itself, since the yield is independentI 1-7 of the concentration of unchanged furan present in the reaction N/N\/ mixture.As yet, however, there is no direct evidence for such con-('') Tetrahydrofurancan be converted into pyrrolidine or A7-substituted pyrrolidines in yields ofup to 56%. Less favourable results a.re observed when the nucleus carriesa C-alkyl substituent, 78 and such homologues of furan and tetrahydrofurangive yields of only 10-12% and 27-34% respectively. The presence ofan efficient dehydration catalyst is essential, and, in view of the readinesswith which the furan ring undergoes scission, it is probable that themechanism involved is one of amination to give a y-amino-alcohol, followedby dehydrati~n.~~ Thus the conversion of furan into pyrrole may be re-presented as proceeding through the intermediate formation of 4-amino-butadien- 1-01 :densations under the conditions employed.I n [ TE+ ifI---CH --++ H*O\\O/ I I CH-OH CH*NH,whereas tetrahydrofuran would give 4-aminobutanol, andiS-lI \ /NHthence pyrrolidine.are obtainable by Only very s m h yields of thiiphen or its homologuessimilar reactions in which hydrogen sulphide is used in place of ammonia, butthe method is excellent for the preparation of tetrahydro-derivatives (thio-phans).2 : 5-Dimethyltetrahydrofuran, for example, gives a 68% yield of2 : 5-dimethylthi0phan.'~ In these cases, also, it is probable that themechanism is similar, since it has been shown that 4-hydroxybutanethiolreadily gives thiophan over alumina a t 400".C.H. Kline and J. Turkevich 8o have studied the vapour phase reactionbetween tetrahydrofurfuryl alcohol and ammonia over alumina a t 400".The products contained only 3% of pyridine and 9% of piperidine, butaccording to G. Natta, G. Mattei, and E. Bartolett,i 81 it is possible bythe use of alumina-chromia or phosphate catalysts to obtain, under optimumconditions, a 67-70% yield of pyridine. These authors consider that theprobable sequence of reactions is amination, dehydration, and dehydrogen-ation, which presumably may take place by either of the routes shownon the next page.It is reported, however, that a mixture of pyridine and piperidine canbe obtained from dihydropyranYa0 a.nd, since dihydropyran is known to beformed when the vapour of tetrahydrofurfuryl alcohol is dehydrated overalumina, it is not improbable that it is an intermediate in the reaction underdiscussion.7 8 J.K. Juriev, V. A. Tronova, and Z. Y. Bukshpan, J. Gen. Chem. Russia, 1941,11,1128.79 G. Natt,a, G. Mattei, and E. Bartoletti, Chim. e l ' l n d . , 1942, 24, 81.80 J . Amer. Chenz. SOC., 1944, 66, 1710.81 Ital. P. 382,819; Chem. Zentr., 1942, I, 2930OWEN : FURANS. 171Pyridine is not formed from furfuryl alcohol *O or from furfurylamine 82under these conditions, but furfurylamine in the liquid phase readilyp 3 2 - - p 3 2 -IfH,O (yH -H,O /\!OJCH2*OH 3 CH2 YHoOH -+ ,NH, CH,*OH NHtakes up two mols. of hydrogen over Raney nickel a t 140", presumably togive the tetrahydro-amine, and a third mol.slowly a t 200". The productthen contains a small amount of pyridine, but the major constituent isN-tetrahydrofurfurylpiperidine (11).Piperidine can be obtained from furfuraldehyde or its hydrogenationproducts by treatment a t 200" under high pressure with ammonia and excessof hydrogen, in the presence of a cobalt catalyst.83Ring Fission.-One of the most promising applications of furan deriv-atives is to be found in the formation of various aliphatic compounds by ringfission, a few examples of which have already been mentioned. I n general,the method is most satisfactory when applied to tetrahydrofurans. Tetra-hydrofuran itself (I) was an important intermediate in the German industryduring the recent war,28 and was chiefly used for the preparation of adipicacid by reaction with carbon monoxide in the presence of a nickel carbonylcatalyst, and for the manufacture of butadiene by dehydration in the vapourphase over a phosphate catalyst.84 Similarly, 2-methyltetrahydrofuran hasbeen used for the preparation of penta-1 : 3-diene.85CH,:CH*CH:CH, , --? H02C*[CH2]4*CO,HCl*[CH,],*OH 7 91.1 Cl*[CH2],*OAcAliphatic 1 : 4-dibromides can be conveniently prepared by the actionof hydrogen bromide in acetic acid a t 130".86 A simpler method, which isrecommended by S.Fried and R. D. Kleene,s7 is to pass the anhydroushydrogen halide into the tetrahydrofuran until the calculated amount hasbeen absorbed, the temperature being allowed to rise to 150" during thestrongly exothermic reaction.According to these authors, the reactivityof the hydrogen halides is in the order HI>HBr>HCl, and with hydrogen82 C. L. Wilson, J . Amer. Chem. SOC., 1945, 67, 693.83 D.R.-P. 695,472. 84 B.P. 506,038. 85 U.S.P. 2,273,484.8% R. Paul, Bull. SOC. chim., 1938,5, 1053.81 J . A w r . Chem. Soc., 1940, 62, 3258; 1941, 63, 2691; compare G. B. Heisig,ibid., 1939, 61, 626172 ORGANIC CHEMISTRY.chloride it is advisable to carry out the reaction in the presence of zincchloride, the yield even then being rather poor. This is supported by theobservation that tetrahydrofuran-2 : 5-dicarboxylic acid fails to react withhydrogen chloride, but with hydrogen bromide a t 125" it gives a 60% yieldof meso-ad-dibromoadipic acid.57 It is evident that the initial product ofring fission is likely to be a halogenohydrin, which then reacts with a furthermol.of hydrogen halide to give the dihalide, and it has often been possibleto show the presence of the intermediate, and in some reactions to isolateit. Thus, when 2 : 5-bisacetoxymethyltetrahydrofuran (11) is treatedwith hydrogen bromide in acetic acid, the main product, 2 : 5-dibromo-1 : 6-&acetoxyhexane, is accompanied by a small amount of 2-bromo-1 : 5 : 6-triacetoxyhexane, evidently derived from the intermediate (111) , 57 whilstJ. K. Juriev, K. M. Minatschev, and K. A. Samurskaja 88 have obtained a55% yield of 4-chlorobutanol by the action of hydrogen chloride on tetra-AcO*CH,!OICH,OAc = AcO*CH,*CHBr * [ CH,] ,CH( OH)*CH,*OAc(111.) i (11.1A&H- .1"" H W ' O-AcO*CH,*CH( OAc) [ CH,] ,* CH (0 Ac) * CH,* 0 AcAcO*CH,*CHBr [CH,],*CHBr*CH,*OAchydrofuran (I).With an unsymmetrically substituted tetrahydrofuran,two intermediate products are possible, but C. L. Wilson4O has shownthat tetrahydrofurfuryl alcohol (IV), treated with hydrogen bromide a tloo", gives almost entirely 1 : 5-dibromopentan-2-01, with only traces of4 : 5Ac,O-ZnCI, [olCH2*OH (IV.)Cl*[CH,],*CH( OAc)*CH,*OAc + AcO*[CH,],*CHCl*CH,*OAc'4 /-> AcO* [CH,],*CH( OAC) *CH,*OAciibrom~pentanol.~~ This preferential fission of the 1 : 5-linkage is alsoshown if the reaction is carried out with hydrogen bromide in acetic acid,when the sole product is 1 : 5-dibromo-2-acetoxybutane; a further exampleis found in the formation of 6-bromo-a-bydroxyvaleric acid (VI) as an inter-s~ J .CTen. Chem. Russia, 1939, 9, 1710.89 See also R. Paul, Bull. SOC. chim., 1933, 53, 417; 1945, 12, 388OWEN : FURANS. 173mediate in the conversion of tetrahydrofuran-2-carboxylic acid or its ester(V) into a8-dibromovalericBr*[CH,],*CH(OH)*CO,Et --+ Br*[CH2I3-CHBr*COzEtAcC1 lo/CO,Et -> Cl*[CH,],*CH( OAc)*CO,Et(VII.)AcO*[CH2],*CH(OAc)*C02Et I=/J. B. Cloke and F. J. PilgrimQ1 have studied the action of acyl halideson tetrahydrofuran (I) and have shown that 8-halogenobutyl esters can beprepared by this method. Acetyl chloride gives a 50% yield of 6-chlorobutylacetate, accompanied by a considerable amount of higher-boiling ethers, butby the use of a minute quantity of zinc chloride in the reaction mixture theyield of the desired ester may be raised to 70% ; a large quantity of catalyst,however, is deleterious, and may reduce this figure to 10%.According toL. M. Smorgonskii and Y. L. G~ldfarb,~, acetyl bromide is particularlyeffective, and these authors have also extended the reaction to aracyl chlorides.Similar results have been obtained with 2 : 5-dimethyltetrahydrofuran.With unsymmetrically substituted compounds, it was at first thought 93that the scission took place preferentially in a manner similar to that en-countered with hydrogen halides, but more recently ,* it has been shown thattetrahydrofurfuryl alcohol (IV) with acetyl chloride gives a mixture ofproducts, in which the secondary chloride preponderates.Since the con-stituents cannot readily be separated, the estimation is based on hydrolysisto the mixed chloro-diols, and determination of 1 : 2-diol by titration withperiodic acid. Preferential scission of the 1 : 5-linkage does appear to occiir,however, with ethyl tetrahydrofuroate (V), which with acetyl chloride givesan 85 yo yield of ethyl 8-chloro- a-acetoxyvalerate (VII) Replacement of thehalogen atom in products of the above types by an acetyl group, followed bysaponification, gives triols or dihydroxy-acids, the overall yield being betterthan that obtained by direct hydration of the tetrahydrofurans.90393 Analternative procedure is to cleave the ring with acetic anhydride containingzinc c h l ~ r i d e .~ ~ , ~ ~ This gives good results when the nucleus carries a sidechain containing a hydroxyl or carboxyl group, e.g. (IV) and (V); other-wise, however, dehydration occurs during the reaction, and unsaturatedproducts are obtained. Acetolysis can also be effected by an acetic anhydride-acetic acid-sulphuric acid reagent,57 which has been used for the conversionof 2 : 5-bisacetoxymethyltetrahydrofuran (11) into 1 : 2 : 5 : 6-tetra-acetoxy-90 R. Paul, Compt. rend., 1941, 212, 398.0 1 J. Amer. Chem. Soc., 1939, 61, 2667.@a J. Gen. C h m . Rucl&a, 1940,10, 1113.94 Idem, Bull. SOC. chim., 1941, 8, 369, 911.R. Paul, Compt. rend., 1940,211, 645174 ORQANIU CHEMISTRY.hexane, and for the preparation of etjhgl 2 : 5 : 6-triacefoxyhexoate fromethyl ~-acetoxymethyltetrahydrofuran-2-carboxylate (VIII).--AoO*CH,(0)CO,Et 4 AcO*CH,*CH( OAc) [CH,],*CH( OAc) *CO,Et(vr 11.)The h ydrogenolysis of tetrahydrofurfuryl alcohol (IV) over copperchromite gives only pentane- 1 : 5-dio1, whilst from furfuraldehyde a mixtureof the 1 : 2-, 1 : 4-, and 1 : 5-diols is obtained.56 The claim that furfurylalcohol under these conditions gives the 1 : 2- and 1 : 4-diols is, however,not in accord with the observations of H. Adkins and R. Connor,95 and ofL. E. Schniepp and H. H. Geller,96 who regard the product as a mixture of1 : 2- and 1 : 5-diols.Hydrogenolysis of halides, by treatment in dry ether with an alkalimetal, results in the formation of unsaturated alcohols. 3-Bromotetra-hydrofuran (IX) , for example, gives allyl~arbinol,~~ whilst pent-4-en-1-01 canbe obtained in 90% yield from tetrahydrofurfuryl chloride (X).97An interesting method for the preparation of long chain di-, tri-, andtetra-ketones has been described by K.Alder and C. H. Schmidt.982-Methylfuran condenses readily with methyl vinyl ketone to give 2-methyl-5-7-ketobutyIfuran (XI), which undergoes the expected scission on boilingwith acid alcohol to give nonane-2 : 5 : 8-trione. Hydrogenolysis of (XI)gives a mixture of nonane-2 : 5- and -2 : 8-diones :MJl II*[CH,],*CO*Me (XI.)f ‘O’\Me*(CO*[CH,],),*CO*Me Me C 0 [ CH ,] gC 0 Me +Me*[CH,],*CO*[CH,],*CO~MeFuran itself condenses with methyl vinyl ketone’ to give 2 : 5-di-y-ketobutyl-furan (XII), which by acid scission yields dodecane-2 : 5 : 8 : ll-tetraone,whilst on hydrogenolysis dodecane-2 : 5 : ll-trione is obtained :--Me*CO*[CH,] 2./!,0)l*[CH,]2*CO*Me (XII.) y\Me*(CO*[CH,],),*CO*Me Me*CO*[CH,],*CO*[CH,],*CO*MeL. N. 0.9 5 J. Amer. Chem. SOC., 1931, 53, 1091.9 7 R. Paul and H. Norrnant, Bull. SOC. chim., 1943,10,484.g6 Ibid., 1945, 67, 54.@ * Ber., 1943,76, 183LYTHGOE : CHEMISTRY OF ADENINE NUCLEOTIDE COENZYMES. 1756. CHEMISTRY OF ADENINE NUCLEOTIDE COENZYMES.~Certain adenosine-5’-phosphoric acid derivatives are active biologicallyas coenzymes of phosphate transfer, and adenine dinucleotides as coenzymesof hydrogen transfer. Phosphate esters seem particularly suited to act ascoenzymes; for example cocarboxylase is known to be aneurin pyrophos-phate and an Z-tyrosine codecarboxylase is probably identical with pyridoxalph~sphate.~Adenosine-Ei’-phosphoric Acid Derivatives Active in Phosphate Transfer.as phos-phate donors by virtue of their labile phosphate content, removal of which isaccompanied by liberation of large amounts of free energy, available as energysource for cellular activities ; the adenosine derivatives of lesser phosphoricacid content act as phosphate acceptors for resynthesis of the polgphosphates.Muscle AdenyEic Acid.-G. Emden and M.Zimmermann in 1927isolated from muscle extracts an acid recognised by the following propertiesas differing from the isomeric yeast adenylic acid (adenosine-3’-phosphate)in the location of the phosphoryl residue : (a) Yeast adenylic acid is notdeaminated by G.Schmidt’s muscle adenylic deaminase. ( b ) Acidicdephosphorylation of the muscle acid takes place more slowly than with theyeast acid and the yield of furfural obtained in W. S. Hoffman’s method *for the determination of pentose is less; these differences are paralleled bythe characteristics of the corresponding d-ribose-3- and -5-phosphoric acids.9( c ) The muscle acid forms a deep blue soluble copper complex (Klimek-Parnas test) lo and increases the acidity of boric acid in the Boesekentest ; these properties indicate the presence of a cis-1 : 2-glycol grouping.Muscle adenylic acid has the ultra-violet absorption spectrum of an adenine-9-glycoside ; it can be dephosphorylated to adenosine enzymically anddeamiiiated to inosinic acid enzymically or by use of nitrous acid; it istherefore adenosine-5’-phosphoric acid (I, R = NH,).The constitutionI n biological systems the adenosine polyphosphates fiinctionf Cf. C. Lutwak-Mann, Biol. Rev. Camb. Phil. Xoc., 1939, 14, 399.2 K. Lohmann and P. Schuster, Naturwiss., 1937, 25, 26.3 J. C. Gunsalus, W. D. Bellamy, and W. W. Umbreit, J. Biol. Chem., 1944, 155,685 ; J. Baddiley and E. F. Gale, Natzwe, 1045,155, 727.4 For recent reviews on biochemical aspects, see €1. M. Kalckrtr, Biol. Rev. Camb.Phil. Xoc., 1942,17, 28; F. Lipmann, Advances in Enzymology, New York, 1941,1, 100;D. M. Needham, Ann. Reports, 1941,38,241.2. physiol. Chem., 1927,167, 114.G. Emden and G. Schmidt,, ibid., 1929,181, 130.l b i d . , 1928,179 243.8 J .Biol. Chern., 1927, 73, 15.9 P. A. Levene and S. A. Harris, ibid., 1933,101,419.10 Biochem. Z., 1932, 252, 392; 2. physiol. Chem., 1933, 217, 75.l 1 Ber., 1913,46,2612.12 J. M. Gulland and E. R. Holiday, J., 1936, 765176 ORGANIC CHEMISTRY.of the inosinic acid (I, R = OH) suggested earlierlater work.has been confirmed byI H H HH02C~-!$-((-CH,*O*P03H2 H H HH H H 1“’“ I CHN\/\N/Inosinic acid gives on acid hydrolysis a ribosephosphoric acid, oxidisable toa phosphoribonic acid (11) which forms only a y-lactone ; l4 additional evid-ence that the ribosephosphoric acid is a 5-phosphate is given by P. A.Levene and E. T. Stiller’s synthesis : l5 &Ribose -+ 2 : S-mono-R (1.1 (11.)MeOH-Me,CO- -_ flu1 POCI, acetone methylribofuranoside - -+ 2 : 3-monoacetone methylribofur-anoside-5-phosphat e ---1-+ d-ribose-5 -p hosphoric acid.Partial synthesisof inosinic acid by a route demonstrating the position of the phos-Dil acid pmdlnephoryl group has been effected.16 Partial syntheses of muscle adenylicacid have been carried out by T. Jachimowicz l7 who phosphorylatedadenosine without protecting the hydroxyl groups a t C, and CX; by P. A.Levene and R. S. Tipson l8 starting from 2’ : 3’-monoacetone adenosine;and by the same authors l8 and by H. Bredereck, E. Berger, and J. Ehren-berg starting from 2’ : 3’-diacetyladenosine; the last workers used di-phenylphosphoryl chloride as phosphorylating agent. The yield of muscleadenylic acid from all these syntheses was very low; they serve as con-firmations of structure rather than as methods of preparation. Hydrolysisof adenosine triphosphate occurring in muscle,20, 21 or obtained enzymicallyfrom adenosine,22 remains the most convenient way so far published ofobtaining muscle adenylic acid.Crystalline acridine salts of this acid havebeen described by T. Wagner-Jauregg 23 and by R. S. T i p ~ o n . ~ ~Adenosine Di- and Tri-phosphates.-Adenosine triphosphate (ATP) wasisolated from muscle extracts in 1929 by K. Lohmann25 and by C. H.Fiske and Y. Subbarow; 26 according to K. Lohmann 27 this nucleotideis the parent of the adenylic acid from muscle, which is an artefact formedduring the isolation process and does not occur free in that source. ATPl3 P.A. Levene and W. A. Jacobs, Ber., 1911,44,746.l4 P. A. Levene and T. Mori, J . Biol. Chem., 1929,81,215.l5 Ibid., 1934,104, 299.l6 P. A. Levene and R. S. Tipson, ibicl., 1935,111, 313.l7 Biochem. Z., 1937, 292, 356. 18 J . Biol. Chem., 1937, 121, 131.2O S. E. Kerr, J . Biol. Chem., 1941,139, 131.21 M. V. Buell, ibid., 1943,150, 389.22 P. Ostern, T. Baranowski, arid J. Terszakowed, 2. physiol. Chem., 1938, 251,23 Ibid., 1936, 239, 188. t4 J . Biol. Chem., 1937,120,621.25 Naturwiss., 1929,17, 624. 26 Science, 1939,70, 381.2’ Biochem. Z., 1931,233, 460,Ber., 1940, 73, 269.268LYTHGOE : CHEMISTRY OB ADENINE NUCLEOTIDE COENZYMES. 177(Cl,,H16013N5P3) gives on hydrolysis with dilute alkali muscle adenylicacid and inorganic pyrophosphate; with dilute acid 1 mol.of adenine and2 mols, of orthophosphoric acid are liberated rapidly, the third phosphorusatom appearing as d-ribo~e-5-phosphate.~~ Whereas dephosphorylationof the latter occurs only slowly, the first 2 mols. of phosphoric acid are liber-ated from ATP a t about the same speed as from inorganic pyrophosphate(7-15 mins. in N-HCl a t 100°),287 2o and determination in this way of the) which is thus 2 : 1 for ATP forms acid-labile phosphorusacid-stable phosphorus ratio P,/P~ (=- -one of the most important ways of characterising this and other adenosinepolyphosphates,28 characteristic melting points being in general lacking inthis series.The structure of muscle adenylic acid being known from earlier work,it remained to determine the location of the acid-labile phosphoryl residues.Evidence (although not conclusive in so complex a molecule) was obtainedindicating that the cis-1 : 2-glycol grouping is not involved, ATP giving apositive response to the Klimek-Parnas and Boeseken tests just as doesadenylic acid.28329 Linkage to the adenine residue as suggested by H.K.Barrenscheen and W. Filz 30 seems excluded, since contrary to the findingsof these authors ATP can be deaminated by nitrous acid to an inosine tri-phosphate (ITP) 28, 313 32 and thus the only functional grouping of the aden-ine residue, vix., the amino-group, must be unsubstituted. Basicity deter-minations with phenolphthalein as indicator, and by electrometric titra-tion,28933 show ATP to possess three primary and one secondary acidicgroupings, which led K.Lohrnann to propose for it the structure (111).o---/ OH OH OH 1 H H I l lCH-?-~--/ --CH,. o -P- o -P- o *P*O H I H H H + + + 0 0 0(111.)CH-?z?!;- 1---O-- -CH,*O-P*O*P*OH 9" 0"1 H H H + + NH20 0I tK. Lohmann 34 later showed that in absence of Mg" ions (washed) crab-muscle pulp removes from ATP only one of the acid-labile phosphoryl28 K. Lohmann, Biochem. Z., 1932,254,381.29 H. K. Barrenscheen and T. Jachimowicz, ibid., 1937, 292, 350.30 Ibid., 1932,250,281 ; H. K. Barremcheen, K. Brsun, and W. Filz, ibid., 1933,265,31 W. Kiessling, ibid., 1934, 273, 103.32 A. Kleinzeller, Bwchem. J., 1942, 36, 729.33 K. Makino, Biocham,. Z . , 1935,278, 161.141.34 Ibicl., 1935, 282, 104, 120178 ORGANIC CHEMISTRY.residues.The product, adenosine diphosphate (ADP), has the compositionC,,H,,010N,P2 ; it resembles ATP in that it can be deaminated to an inosine-diphosphate (IDP), or hydrolysed by acid to adenine, d-ribose-5-phosphoricacid, and phosphoric acid. ADP also is active as a cophosphorylase; ithas a ratio PL/Ps = 1 : 1. ADP contains 2 primary and 1 secondary acidicgroupings and must therefore be represented by (IV); this supports struc-ture (111) for ATP. Other workers 2 9 y 3 5 have, however, claimed that it ispossible by use of certain enzyme preparations to remove the acid-stable(5’) phosphoryl residue from ATP, leaving the acid-labile grouping intact.36This claim is incompatible with the Lohmann formula and, if substantiated,would indicate that the acid-labile phosphoryl groupings are attached toone of the secondary hydroxyls of the ribofuranoside residue as in (V).35OH OH,Po0 *POOHI I/ J .J.0NH2 (V.)Structures of this type have recently been shown to be highly improbableby the use of both enzymatic 37 and chemical 38 methods. Earlier work,39 inwhich it had been shown that the structures of certain nucleosides andnucleotides could be determined by titration with sodium metaperiodate,was extended to the study of ATP; the consumption by the latter of 1 mol.of metaperiodate proved conclusively the presence of an unsubstitutedu-glycol grouping in the molecule as required by t,he Lohmann stricture(111). It therefore seems certain that ADP and ATP are correctly describedas adenosine-Ej’-di- and -tri-phosphates.Improved methods have been described recently for the preparation ofATP from rabbit muscle 2o and by enzymatic synthesis from adenosine,22and for the preparation of ADP40 and IDP32 by the action of purifiedmyosin preparations on the respective triphosphates.The composition ofATP has been verified by recent analytical data 2o and by the preparationof crystalline acridine salts,23 and its mononucleotide nature has been estab-35 T. Satoh, J. Biochem. Japan, 1936,21, 19.3 6 Cf. W. L. Liebknecht, B9:ochern. Z . , 1939,303,96.3 7 31. Dainty, A. Kleinzeller, A. X. C. Lawrence, M. Miall, D. M. Needham, and S.Shen, J. Gen. Physiol., 1944, 27, 355; J. M. Gulland and E. O’F. Walsh, J., 1945,169.38 B.Lythgoe and A. R. Todd, hrature, 1945,155, 695.39 I d e m , J., 1944, 592.40 M. N. Ljubimova and D. Pevsner, Biochimia, 1941, 6, 178; K. Bailey, Biochem.J . , 1942, 36, 121LYTHGOE : CHEMISTRY OF ADENINE NUCLEOTIDE COENZYMES. 179O--:C,H,N5*CH-?L I--lished by molecular weight determination^.^^ On the other hand, the exist-ence of a diadenosine pentaphosphate with a ratio PL/P, = 3 : 2 postulatedby certain workers 42 seems doubtful; it is 41, 34 probably a mixture of ADPand ATP formed as a result of autolytic decomposition.Di(adenosine-5’-phosphoric Acid) .-Isolated from yeast by W. Kiesslingand 0. M e y e r h ~ f , ~ ~ this adenine dinucleotide is active as a cophosphorylase ;it can accept 2 mols. of phosphoric acid from phosphopyruvic acid to give adiadenosine tetraphosphate.A diadenosine triphosphoric acid is formedby removal of one of the two acid-labile phosphoryl residues from thetetraphosphate by washed crab muscle. Diadenosine tetraphosphate hasthe ratio PL/P, = 1 : 1 and contains 4 primary and 1 secondary acidicgroupings. It is readily split by very dilute alkali giving ATP and a com-pound, probably muscle adenylic acid. This suggests a phosphate linkagebetween the two nucleotides of the same type as the internucleotidic linksin nucleic acids, and structure (VI) was suggested for the compound.0OH OH OH--i-(.H,.O.P*O*P***P*OH I l lAdenine Nucleotides Active as Coenzymes of Xydrogen Transport.46The chemistry of these coenzymes, present in minute concentrationbut of primary importance in the metabolism of the living cell, has become4l K.Lohmann and P. Schuster, Biochem. Z . , 1937,294, 183.42 P. Ostern, ibid., 1934, 270, 1; P. Ostern and T. Baranowski, ibid., 1935, 281,43 Biochem. Z., 1938, 296,410. 44 J. Biol. Chem., 1943, 148, 255.45 2. physiol. Chem., 1941,267,264.4 6 For reviews dealing with biochemical aspects, see C. A. Bnumann and F. J.Stare, Physiol. Rev., 1939, 19, 363; D. E. Green, “Mechanisms of Biological Oxid-ations,” Cambridge, 1940, p. 36 ; F. Schlenk, “ Symposium on Respiratory Enzymes,”Wisconsin, 1941, p. 104.157; F. Beattie, T. H. Milroy, and R. W. F. Strain, Biochem. J . , 1934,28, 84180 ORGANIC CHEMISTRY.known during the last decade matii1l;y owing to the investigations of threegroups of workers; those of IT.von Euler in Stockholm, 0. Warburg inBerlin, and P. Karrer in Zurich. As a result, the structure of cozymase(codehydrogenase I ; diphosphopyridine nucleotide, DPN) may be con-sidered as fairly well established; the evidence on which this structureis based will be reviewed here, aiid the present position regarding the struc-ture of codehydrogenase I1 (triphosphopyridine nucleotide, TPN) and ofriboflavin-adenine dinucleotide will be summarised. Features of the out-standing researches here recorded have been the contributions made by thestudies of synthetic model substances, the successful employment of ultra-violet absorption spectroscopy and of manometric micromethods in inter-preting the reactions of the coenzymes, and the extensive use of biologicaltest methods, especially in determining the nature of fission productsobtained from the coenzymes by degradation reactions.Di- and Tri-phosphop yridine NucZeotides.-Chemical investigation ofhydrogen transport coenzymes commenced with the work of H.von Eulerand K. M y r b a ~ k . ~ ~ These workers purified extensively the constituent ofHarden and Young’s 48 complex “ coenzyme of alcoholic fermentmation ’’ nowknown as DPN. As test method in the purification process they usedmeasurement of the stimulation produced in the fermentation of glucoseby washed yeast. Hydrolysis of their most active preparations from yeastgave adenine, a pentose, and phosphoric acid, and up to 1935 the coenzymewas regarded as a mononucleotide. While this work was in progress 0.Warburg and W.Christian 49 were investigating a different 50 though closelyrelated coenzyme (TPN), present in horse erythrocytes, which was necessaryfor the oxidation of glucose-6-phosphate to 6-phosphogluconic acid by mole-cular oxygen in a system containing the appropriate specific protein (apo-enzyme; “ Zwischenferment ”) and 0. Warburg’s 51 “ old flavoprotein.”In 1935 they succeeded (with A. Griese 52) in isolating this coenzyme pureand establishing its mode of action. TPN has the compositionCZ1Hz8O1,N7P3 ; on acid hydrolysis its molecule yields 1 molecule of adenineand 1 molecule of nicotinamide; 3 phosphoryl residues are liberated as in-organic phosphate on treatment with alkali ; and 2 pentose units are probablypresent (furfural estimations).The amino-group of the adenine residue isfree, undergoing deamination under van Slyke conditions to which nicotin-amide is inert. Investigation of themechanism of coenzyme action of TPN by separate study of the constituentreactions manometrically and by ultra-violet absorption methods was neces-4 7 Z.physio1. Chem., 1931,198, 236; 203,143 ; Naturwiss., 1929,17,291; K. Myrbiickand H. Hellstrom, 2. physiol. Chem., 1932, 212, 7 ; K. Myrback, ibid., 1935, 233, 95.4 * A. Harden and W. J. Young, Proc., 1905, 21, 189.‘9 Biochern. Z., 1931,242,206; 1933,266,377; 1934,274, 119; 1935,275,464.6O H. von Euler, E. Adler, F. Schlenk, and G. Gather, 2. physiol. Chem., 1935,51 0.Warburg and W. Christian, Biochem. Z . , 1932, 254, 438; 1933, 263, 228;5 2 Ibid., 1935, 282, 157.TPN is therefore a dinucleotide.233, 120.1936,287, 291, 440LYTHGOE : CHEMISTRY OF ADENINE NUCLEOTIDE COENZYMES. 181sary before the manner in which the degradation fragments are united in themolecule was revealed. Equivalent amounts of TPN and glucose-6-phos-phate react anaerobically in presence of apoenzyme according to the equationTPN + R-COH + H20 = R*CO,H + TPN-H, (1). The dihydrocoen-zyme (isolated from the reaction product as calcium salt) can also be obtainedfrom TPN by reduction with sodium hyposulphite in presence of bicarbonate :TPN + Na,S,O, + 2H20 = 2NaHS0, + TPN-H,The dihydrocoenzyme is not autoxidisable, hence if in reaction (l), conductedin presence of molecular oxygen, less than 1 equivalent of coenzyme is used,the substrate is only partially oxidised, reaction ceasing when all TPN isconverted into dihydro-TPN.Dihydro-TPN, however, can react with flavo-protein anaerobically, the alloxazine ring of the latter undergoing reduction :Since the dihydroflavoprotein is autoxidisable :FH2 + 0, = F + H202a small quantity of coenzyme is able to oxidise large quantities of substratein the complete Warburg system, being continually regenerated from itsdihydro-derivative; L e . , TPN owes its coenzyme activi'y to its redoxproperties.TPN is very unstable to alkaline hydrolysis, but relatively stable towardsacid. It shows an absorption band at 260 mp to which both adenine and nico-tinamide residues contribute ; on catalytic hydrogenation the pyridinenucleus is reduced, giving a hexahydro-derivative, devoid of catalyticactivity (" irreversible hydrogenation "), in which the 260 mp band hasdiminished in intensity to the adenine value." Reversible hydrogenation "of TPN to dihydro-TPN alters the intensity of the 260 mp band only slightly,but a further characteristic band at 346 mp appears ; this change is reversedon addition of flavoprotein (regeneration of TPN). The pH-stabiIityproperties of dihydro-TPN are the reverse of those of the oxidised form;it is relatively stable to alkali, but decomposes immediately on acidification(irreversible addition of acid to the pyridine nucleus 53) ; the 345 mp bandvanishes and is replaced by another a t 295 mp.On catalytic hydrogenationthe dihydro-TPN takes up 2 mols. of hydrogen giving the " irreversible "hexahydro-derivative ; this stepwise formation of the latter shows that di-hydro-TPN owes its production to a partial hydrogenation of the pyridinering; hence the centre of biological activity in the coenzyme molecule isnicotinamide, a fact of importance in connection with the vitamin activityof the latter.54The properties of TPN and dihydro-TPN described above are character-istic of a group of compounds containing nicotinamide linked in a particularmanner; in order to elucidate the nature of this linkage, search for modelsubstances with a redox behaviour similar to that of the coenzyme was com-1938, 21, 223.TPN-H, + F = FH, + TPN63 p.Karrer, F. W. Kahnt, R. Epstein, W. Jaffi?, arid T. Ishii, Helv. Chim. Acta,6 4 Ann. Rev. Biochem., 1941, 10, 352; 1943, 12, 326182 ORGANIC CIXEMISTRY.menced by P. Karrer and 0. Warburg.55, 56 Nicotinamide itself is not hydro-genated by sodium hyposulphite .(Na,S,O,) ; trigonelline (VII) , however,forms a dihydro-derivative with this reagent. It was then found 57 thatnicotinamide methiodide (VIII), which shows ultra-violet absorption closelysimilar to that of TPN, takes up 1 mol. of hydrogen in presence of sodiumhyposulphite,C,H,0N2+I- + Na,S,O, + 2H,O = C7H,,0N, + HI + 2NaHS0,the reduced form showing analogous properties to those of dihydro-TPN(failure to autoxidise ; oxidisation by flavoprotein ; spectroscopic properties ;pH-stability).Other derivatives (IX, X) in which linkage of nicotinamideis effected by either of its other two functional groups (0 or NH, of the carb-oxyamide group) proved ineffective as redox models; 55 in the coenzymenicotinamide is evidently bound as a quaternary pyridinium compound.Me (VII.) Me (VIII.) (IX.1 (X.1The equations of reduction of coenzyme and model (VIII) show liberation of3 mols. of acid; the third, liberated from the model substance as hydrogeniodide, must be set free from the coenzyme as the hydroxyl of a phosphorylgroup, hence TPN is a quaternary pyridinium p h ~ s p h a t e . ~ ~ Facile reduc-tions of quaternary pyridinium compounds were already well known, andcomparison with dihydropyridines of established structure showed that re-duction of (VIII) gave mainly the 1 : 2- or 1 : 6-dihydro-derivative (XI orXII); clear cut decision between these two alternatives has not yet beeneffected.57, 58h e(XII.)A //IICO*NH,H N’IR(XIII.)Reversible reduction of the coenzyme is therefore represented by, e.g.:I OH + 2H ___, +- R-O-P-OR’ I P- R-O*P-OR’ + J.0 065 Biochem. Z., 1936, 285, 297.5 6 0. Warburg and W. Christian, ibid., 1936,287,291.6’ P. Karrer, G. Schwarzenbach, F. Benz, and U. Solmsen, HeZv. Chim. Acta,6 8 P. Karrer, G. Schwartzenbach, and G. Utzinger, ibid., 1937,20, 720.1936, 19, 811LYTHGOE : CHEMISTRY OF ADENINE NUCLEOTIDE COENZYMES. 183These reductions have been shown 57, 59 to proceed by addition of 1 atomof hydrogen at a time via a radical of tshe semiquinone type with a strongnegative potential (XIII).I n further experiments, P.Karrer and co-workers 6o found, in N-glucosido-1 : 2-dihydronicotinamide (XV, R = H) and tetra-acetyl-N-glucosido-nicotinamide bromide (XIV), subst,ances which in respect of redox behaviourand spectroscopic and pH-stability properties were the closest hithertoavailable models for the behaviour of TPN. (XIV) was prepared by con-densing acetobromoglucose with nicctinamide in dioxan ; reduction of (XIV)with sodium hyposulphite a t pH 7-5 gave the dihydro-derivative (XV,R = Ac) which was deacetylated to (XV, R = H)I l / / o n H o qCH-~-~-~--CH,*OAc !/'ZL*Gq cH-i--'-- I - -CH,*ORH OAcH H H O R H H(XIT'.) P V . 1Attempts to prepare N-pentosidonicotinamide derivatives analogous to(XIV) and (XV) were only partially successful, as the compounds were difficultto purify.P. Karrer had been led to the hypothesis that a n N-pentosido-pyridinium phosphate residue exists in the coenzyme molecule by thelability to alkali of already known N-glycosidopyridinium compounds,61whereas AT-glycosides, which would arise on partial reduction, are more,stable to alkali generally.62 The model experiments described, togetherwith the isolation of nicotinamide nucleoside from DPN described below,have demonstrated the correctness of this hypothesis.The important progress made on DPN in the meantime indicated a closerelationship of this coenzyme with TPN. Reinvestigation of the productsof hydrolysis of DPN undertaken with purer material than hitherto availableshowed the presence of nicotinamide 5 6 3 639 643 65 as well as adenine; thatDPN is active in hydrogen transport, recognised by H.von Euler et aE.66and by 0. Warburg and W. Christian 5 6 3 67 independently, became under-standable in the light of the results obtained with TPN; by virtue of itsj9 E. Adler, H. Hellstrom, and H. von Euler, 2. physiol. Chem., 1936, 242, 225.6o P. Karrer, B. H. Ringier, J. Buchi, H. Fritzsche, and U. Solmssen, I-lelv. Chim.61 P. Karrer, A. Widmer, and J. Staub, ibid., 1924,7, 519.6 2 Alkali-labile N-glycosides have, however, been described by R. Kuhn andActa, 1937, 20, 55.R. Strobele, Ber., 1937, 70, 747, 753.64 Idem, ibid., 1936, 240, 113.6 6 0.Warburg and W. Christian, Biochem. Z., 1935,275,464.'66 H. von Euler, E. Adler, and H. Hellstrom, Svensk Kem. Tidskr., 1936, 47, 290;13' Bioclwn. Z., 1936, 268, 81.H. von Euler, H. Albers, and F. Schlenk, 8. physiol. Chem., 1935,257, 1.H. von Euler and E. Adler, 2. physiol. Chem., 1936, 238, 233184 ORGANIU CHEMISTRY.nicotinamide content DPN transfers hydrogen in H. von Euler’s enzymesystem from diphosyhoglyceraldehyde to acetaldehyde, the former beingoxidised to phosphoglyceric acid and the latter reduced to ethylPure DPN, C,1H,7014N7P2,69, 70 contains 1 adenine and 1 nicotinamideunit ; by a modified Bial’s reaction 71 the presence of 2 pentose units can bedetected.72 The amino-group of the adenine is free; a deaminocozymasehas been described, resulting from the action of nitrous acid on DPN.73DPN is therefore a dinucleotide, differing from TPN only in apoenzymespecificity and in having 1 phosphoryl group less.Dihydro-DPN 749 759 76has been obtained analytically pure as the sodium salt by P. Ohlmeyer; 77both reduced and oxidised foriiis of DPN show redox behaviour and spectro-scopic and pH-stability characteristics identical with those of the TPNanalogues; 563 599 74 hence, as in TPN, the nicotinamide fragment of DPNmust be linked as a pentosidopyridinfum phosphate. Titration with alkali 7Oand cataphoretic experiments 78 show DPN to be a monobasic acid (with avery weakly basic grouping due to the adenine part); dihydro-DPN is di-basic.77 These results, as well as showing the presence of the pyridiniumphosphate zwitterion, indicate a pyrophosphate linkage in the molecule.The presence of such a link, originally suggested by E.Adler, H. Hellstrom,and H. von E ~ l e r , ~ ~ was confirmed by later experiments of H. von Eulerand co-worker~.~~ Brief treatment of DPN with hot alkali destroys thecoenzyme activity completely, liberating nicotinamide and a fragmentcontaining acid-labile phosphorus in the same sense as ADP or ATP, and likethese showing cophosphorylase activity in biological systems. This frag-ment, isolated pure as the barium salt, was shown by R. Vestin, F. Schlenk,and H. von Euler *O to be identical with ADP. (IV) the structure of whichwas already known. On the bases of the foregoing evidence, H. von Eulerand F.Schlenk 70 proposed structure (XVI) for DPN.One of the pentose units of DPN must from the foregoing evidence be&ribose; H. von Euler, P. Karrer, and E. Usteri have demonstrated thepresence of this in DPN by isolating the phenylosazone after hydrolysisof the coenzyme successively with dilute acid and a phosphatase preparation.The yield, however, was very small and would not exclude the presence of adifferent sugar as the second pentose unit. Conclusive evidence that6 8 0. Warburg and W. Christian, Biochem. Z . , 1939, 303, 40.69 H. von Euler and F. Schlenk, Svensk Kern. Tidskr., 1936,48, 135.70 Idem, 2. physiol. Chenz., 1937, 246, 64.7 2 F. Schlenk, J . Biol. Chem., 1942,146, 619.7 3 F. Schlenk, H. Hellstrom, and H. von Euler, Ber., 1938, 71, 1471.7 4 H.von Euler, E. Adler, and H. Hellstrom, 2. physiol. Chem., 1936, 241, 239.75 E. Adler and H. von Euler, Svensk Vet. Akad. Arkiv Kemi, 1937,12B, 36.7 6 D. L. Drabkin, J . Biol. Chem., 1945,157, 563.7 8 0. Meyerhof and W. Mohle, ibid., 1937,294,249.79 R. Vestin and H. von Euler, 2. physiol. Chem., 1936, 245, 1 ; F. Schlenk, H. vonEuler, H. Heiwinkel, W. Gleim, and H. Nystrom, ibid., 1937, 247, 23; R. Vestin andH. von Euler, ibid., p. 43.‘1 W. Mejbaum, ibid., 1939, 258, 117.7 7 Biochem. Z., 1938,297,66.Bey., 1937, 70, 1369.81 Helv. China. Acta, 1942, 25, 323LYTHGOE : CHEMISTRY OF ADENINE NUCLEOTIDE COENZYMES. 185/--0- although devoid of coenzyme activity, can bed-ribose was present in the nicotinamide part of the molecule as ~ l ! as in thendenylic acid part was obtained in 1942 when F.Schlenk 72 isolated from th186 ORGANIC CHEMISTRY.activity be isolated ; 86, 87 this is evidence against structure (XVIII). More-over, the reversible enzymatic transformation TPN JI DPN describedby various worlcers,881 g9 and the claim that DPN can be converted to TPNby phosphoryl chloride in ether,gO further emphasise that structurally TPNdiffers from DPN in no way except in the possession of a third phosphorylresidue, and also show that the structure (XVIII) should in all probabilitybe modified in regard to the location of the latter. The third phosphorylgroup may, for example, be linked to the adenylic acid part of the moleculein the same way as the phosphcryl group in yeast adenylic acid givingstructure (XIX).87 To settle this point a redetermination of the basicityof TPN seems desirable, the existing data 52, 91 on this point not being satis-factory ; isolation of larger fission fragments might also clarify the problem.It would be of interest to apply the metaperiodate oxidation method 38, 39to TPN; since (XVIII) should consume 2 mols.but (XIX) only 1 mol. ofthe reagent, decision between these two alternatives should be possible bythis method.0- OHOH0 0 0NH20- OH(XIX.)H H H0 0PO3H2Flavin-adenine Din~cleotide.~~-This coenzyme which is not, strictlyspeaking, a dinucleotide forms the prosthetic group of a wide variety of8 6 H. von Euler, F. Schlenk, H. Heiwinkel, and B. Hogberg, 2. physiol.Chem.,1938,256,208.8 7 F. Schlenk, B. Hogbcrg, and S. Tingstam, Svensk Vet. Akad. Arkiv Kemi,1939,13A, 11.8 8 H. von Euler and E. Adler, 2. physiol. Chem., 1938, 252, 41 ; H. von Euler andE. Bauer, Ber., 1938, 71, 411.E. Adler, S. Elliott, and L. Elliott, Enzymologia, 1940, 8, 80.F. Schlenk, h’aturwiss., 1937, 25, 668. O1 H. Theorell, ibid., 1934, 275, 19.O 2 For reviews dealing with biochemical aspects, see D. E. Green, “Mechanismsof Biological Oxidations,” Cambridge, 1940, p. 74; T. R. Hogness in “ Symposium onRespiratory Enzymes,” Wisconsin, 1941, p. 134LYTHGOE : CHEMISTRY OF ADENINE NUCLEOTIDE COENZYMES. 187flavoproteins; it is active in hydrogen transport in biological systems byvirtue of the redox properties of the riboflavin residue.Chemical investig-ation of FAD dates from the isolation from liver, kidney, and yeast by 0.Warburg and W. Christiang3 of the pure barium salt, C2,H3,0,,H,P,Ba,of the coenzyme. These workers, having effected separation of the apo-enzyme of the d-alanine oxidase 94 from the coenzyme by acidification withhydrochloric acid in presence of ammonium sulphate, were able to use theoxidase system as test method in their purification process; the equationsof the oxidation are :CH3*CH(NH2)*CO2H + FAD + H2O 1 CH,*CO*CO,H + NH, + FAD-Hp,FAD-H2 + 0, = FAD + H202Reduction of FAD to the dihydro-derivative can be effected by sodiumhyposulphite at pH 7.5, and like the similar reductions of the nicotinamidecoenzymes probably proceeds stepwise via a radical of the semiquinonetype.95 Unlike the dihydronicotinamide coenzymes, dihydro-FAD is aut-oxidisable. The evidence a t present available indicates a structure (XX).Hydrolysis of FAD gives 1 mol.of adenine; this must have been combinedin such a way that the amino-group is unsubstituted (dearnination). Theultra-violet absorption spectrum of FAD follows that of riboflavin closely,OH OHI I H H HJ . J / QH OH OH 10 0O*P*O*P*O*CH,-l --l--/-CH,Iand on alkaline photolysis lumiflavin is liberated from the coenzyme a t thesame rate as from riboflavin itself. E. P. Abraham 96 showed that by care-fully controlled hydrolysis with dilute alkali, resulting in complete loss ofd-amino-oxidase activity, FAD gave adenosine-5’-phosphoric acid, identifiedbiologically by its cophosphorylase activity, whilst after similar treatmentof FAD with dilute acid riboflavin-5’-phosphoric acid could be detected byvirtue of its activity as the prosthetic group of 0.Warburg’s “ old flavo-protein.” P. Karrer and H. Frank9’ have shown that the phosphorylresidues occupy terminal positions on the sugar chains since no formalde-hyde is produced on oxidation of PAD with periodic acid. The ready fissionof FAD into the two constituent nucleotides, which probably proceedsenzymatically 93 as well as by action of acid and alkali, has raised the question93 Biochem. Z . , 1938, 298, 150.94 H. A. Krebs, Biochem. J . , 1935, 29, 1620, 1951; N. B. Das, ibid., 1936, 30, 1080,96 Cf. E. Haas, Biochem. Z . , 1937, 290, 291.96 Riochem.J . , 1939, 33, 543.1617; F. B. Straub, Nature, 1938,141,603.97 Helv. Chini. Acta, 1940, 23, 948188 ORUANIC CHEMISTRY.of the status of riboflavin-5'-phosphate 98 as coenzyme of hydrogen transfer.Whilst it seems likely that its isolation from some enzyme systems is a resultof the breakdown of P A D , present evidence indicates that in others, e.g.,cytochrome-c red~ctase:~ ribofiavin-5'-phosphate is the true prostheticgroup and not merely an artefact.B. L.7. PYRAZINE AND ITS DERIVATIVES.Pyrazine bases occur in fuse1 oil ; the hornologues which have been isolatcdtetramethyl-,2 2 : 5-diethyl-,2 from this source include 2 : 5-dimethyl-,l~and methyltriethyl 3-pyrazine.Synthesis of Pyrazine Homologues.A. Reactions Involving the Self-condensation of u- Amino-carbon91 Com-pounds.-The most widely employed method for the synthesis of pyrazinehomologues is by the self-condensation of a-amino-carbonyl compoundsusually in the presence of an oxidising agent.(a) a-Amino-ketones are relatively easily prepared in the form of theirhydrochlorides either by the method developed by V.Meyer which consistsin the reduction of isonitrosoketoiies in acid solution or by the phthalimido-ketone method of S. Gabriel.5 A typical pyrazine synthesis is that of 2 : 5-di-2Me*CO*CH:N*OH \ SnCl,/HCI'(111.)mefhylpyrazine (11) ; successive treatment of aminoacetone hydrochloride(I) with sodium hydroxide and mercuric chloride gives 2 : B-dimethyl-9s P. Karrer, Helv. Chim. Acta, 1935, 18, 69, 426; H.Theorell, Biochem. Z . , 1934,9x1 E. Haas, E. L. Horecker, and T. R. Hogness, J. Biol. Chem., 1940, 136, 747.272, 155; R. Kuhn, H. Rudy, and F. Weygand, Ber., 1936,69,2034.1 E. C. Morin, Compt. rend., 1888,106,360; P. Brandes and C. Stoehr, J. p r . Chem.,2 A. C. Chapman and F. A. Hatch, J . SOC. Chem. Ind., 1929, 98; T. Taira, J. Agric.P. Schorigin, W. Issaguljanz, \I7. Below, and S. Alexandrowa, Ber., 1033, 66,S. Gabriel and G. Pinkus, ibid., 1893, 26, 2197; H. Gutknecht, ibid., 1879, 12,S. Gabriel and J. Colman, ibid., 1902, 35, 3805; S. Gabriel and T. Posner, ibid.,1896,54,481; E. Bamberger and A. Einhorn, Ber., 1897,30,224.Chem Xoc. Japan, 1936, 12, 576.1087.2290.1894, 27, 1141; S. Gabriel, ibid., 1908, 41, 1127NEWBOLD AND SPRING : PYRAZINE AND ITS DERIVATIVES.189pyrazine (11). The mechanism of this general preparative method is notclearly established ; treatment of aminoacetone hydrochloride with alkaligives a product, C,Hl,N,, which when heated with hydrochloric acid is re-converted into the parent aminoacetone hydrochloride. This product doesnot appear to be the dihydropyrazine (111) since it cannot be oxidised to2 : 5-dimethylpyrazine nor can it be reduced to 2 : 5-dimethylpipera~ine.~The method has been used to prepare 2 : 5-diisopropylpyrazineY7 2 : 5-di-ethylpyrazine (the requisite a-amino-ketone salt being obtained by thephthalimido-ketone route),8 tetrarnethylpyra~ine~~ 2 : 5-dimethyl-3 : 6-di-propylpyrazine,g~ 10 2 : 5-dimethyl-3 : 6-diisobutylpyrazine, l1 2 : 5-dimethyl-3 : 6-diamylpyrazine,12 2 : 5-diphenylpyrazine,13 and 3 : 6-diaryl-2 : 5 4 -methylpyrazines.14 Tetraphenylpyrazine has been obtained by thereduction of benzil monoxime 15 (and benzil dioxime Is) with sodium amal-gam, and tetramethylpyrazine has been obtained by the reduction of iso-nitrosolaevulic acid.A variant of this method has been employed by F.B. Ahrens and G.Meissner who obt’ain 2 : 5-dimethylpyrazine in poor yield by the electrolyticreduction of isonitrosoacetone in sulphuric acid followed by treatment of thesolution with alkali and mercuric chloride. Catalytic reduction of isonitroso-acetophenone in neutral solution gives 2 : 5-diphenylpyrazine, the finaloxidation being effected by atmospheric oxygen,lg and catalytic reductionof bend monoxime and dioxime gives tetrapheiiylpyrazine.20 The form-ation of 3 : 6-di-p-phenylethyl-2 : 5-dimethylpyrazine as one of the productsof a catalytic reduction of benzylidenedimetyl monoxime has been observedby 0.Diels and W. Poetsch.21 Other variants have been described.22 Tetra-methylpyrazine is obtained in very high yield by the reduction of the iso-nitroso-derivative of methyl ethyl ketone with zinc dust and alkali,23 amethod also employed by R. Campbell, R. D. Haworth, and W. K. Perkh2*(b) A remarkable rearrangement of oximes leading to their conversion’ M. Conrad and K. Hock, ibid., 1899, 32, 1199.’ H. Kiinne, ibid., 1895, 28, 2158; H. Gutknecht, ibid., 1879, 12, 2290.lo F. P. Treadwell, ibid., 1881, 14, 2036; H.Kunne, Zoc. cit.; 8. Gabriel andl1 E. Lang, ibid., 1885,18, 1364.l3 S. Gabriel, ibicE., 1908,41, 1127 ; E. Braun and V. Meyer, ibid., 1888, 21, 19.lo M. Tiffeneau, J. LBvy, and E. Ditz, BUZZ. SOC. chim., 1935, 2, 1845.l6 E. Braun and V. Meyer, Ber., 1888, 21, 1269.l 7 K. Thal, ibid., 1892, 25, 1718.l9 W. H. Hartung, J. Amer. Chenz. SOC., 1931, 53, 2248; W. H. Hartung, J. C.2o C. F. Winans and H. Adkins, ibid., 1933, 55, 2051.21 Ber., 1921, 54, 1585.22 F. Knoop, F. Ditt, W. Heckstedon, J. Maier, W. Merz, and R. Hiiile, 2. physiol.23 0. Wallach, Nach. Ges. Wiss. Gottinyen, 1927, 238.S. Gabriel, Ber., 1908, 41, 1127.E. Kolehorn, ibid., 1904, 3’7, 2474.T. Posncr, ibid., 1894, 27, 1037; G. Kalischer, ibid., 1895, 28, 1513.12 L. Behr-Bregowski, ibid., 1897, SO, 1515.N.Polonowska, ibid., 1888, 21, 488.Ibid., 1897, 30, 532.Munch, W. A. Deckert, and F. Crossley, ibid., 1930, 52, 3317.Chern., 1936, 234, 30.24 J . , 1926, 32190 ORGANIC CHEMISTRY.into a-amino-ketones (and thence into pyrazines) is described by P. W. Neberand co-worker~.~~ The oxime is converted into its toluenesulphonate whichwhen treated with potassium ethoxide in alcohol gives an amino-ketal whichis readily converted into the corresponding amino-ketone hydrochloride thusaffording a pyrazine synthesis :R*CH2*C*Me R*CH*C( OEt),*Me R*CH*CO*Me - I NH,,HCl II - IN*OH NH2(c) Certain pyrazine bases can be obtained by the action of ammoniaupon a-halogenated ketones. 3-Chlorobutan-2-one gives a good yield oftetramethylpyrazine 26 as does p-bromolaevulic acid 27 (and p-hydroxy-levulic acid), whilst o-bromoacetophenone gives the dihydro-derivativeof 2 : 5-diphenyIpyrazine, diphenacyIamine,28 and 2 : 6-diphenylpyrazineand its dihydro-derivative.29 Treatment of bromoacetaldehyde withammonia gives a poor yield of pyrazine.30( d ) a-Amino-acids or their esters can be used as starting materialsfor pyrazine syntheses.Thus A. Neuberg31 reduced alanine ester withsodium amalgam in the presence of hydrochloric acid and treated the reactionproduct (the hydrochloride of a-aminopropaldehyde 32) with alkali andmercuric chloride to obtain 2 : 5-dirnethylpyrazine. A more valuable methodis that developed by H. D. Dakin and R. West 33 in which an cc-amino-acidis treated with acetic anhydride and pyridine34 to yield the acetamido-ketone (IV) which on hydrolysis with mineral acid followed by treatmentwith alkali and mercuric chloride yields the pyrazine (V).Using this(IV.1 w.1method alanine is converted into tetramethylpyrazine, and phenylalanineand tyrosine yield (V, R = Ph-CH,) and (V, R = p-OH*C6H,*CH2)respectively.( e ) The most efficient synthesis of the parent pyrazine is that describedby L. Wolff and R. M a r b ~ r g . ~ ~ Treatment of chloroacetal with ammonia26 P. W. Neber and A. V. Friedelsheim, Annulen, 1926, 449, 109; P. W. Neber andH. Uber, ibid., 1928,467, 52; P. W. Neber and A. Burgard, ibid., 1932, 493, 281; P. W.Neber and G . Huh, ibid., 1935, 515, 283; P. W. Neber, A. Burgard, and W.Thier, ibid.,1936, 528, 277.26 M. Dbmhtre-Vladesco, Bull. SOC. chim., 1891, 6, 820.2 7 L. Wolff, Ber., 1887, 20, 425.2Q F. Tutin, J., 1910, 97, 2495; S. Gabriel, Ber., 1913, 46, 3861.30 A. E. Tschitschibabin and M. N. Schtschukina, ibid., 1929, 62, 1075.31 Ibid., 1908, 41, 956.32 E. Fischer, ibid., p. 1019; E. Fischer and Kametaka, Annalen, 1909, 365, 10.33 J . Biol. Chem., 1928, 78, 745, 757.34 Cf. P. A. Levene and R. E. Steiger, ibid., 1927, 74, 689; 1928, 79, 95.35 Annalen, 1908, 363, 169.28 S. Gabriel, ibid., 1908,41, 1130NEWBOLD AND SPRING: PYRAZINE AND ITS DERIVATIVES. 191yields diacetalylamine (VI) which on heating with hydrochloric acid gives2 : 6-dihydroxymorpholine (VII) which when treated with hydroxylaminehydrochloride yields pyrazine (VIII) .Acid hydrolysis of aminoacetalNHNH / \(OEt),CH CH( OEt), \ /CH2 VH2HOCH CH*OH / \ 4 HzQ 7%(VIII.) 0 (VII.)followed by treatment of the product with alkali and an oxidising agentgives p y r a ~ i n e , ~ ~ which is also obtained in small yield by the catalyticdehydrogenation of ethanolamine.37B. Other Methods.-Pyrazine bases can be obtained by decarboxylationof pyrazinecarboxylic acids.38 Although condensation of 1 : 2-diamineswith 1 : 2-dicarbonyl compounds has found little application in the pyrazineseries it is claimed that diacetyl and ethylenediamine react to give a dihydro-pyrazine which on oxidation with Fehling's solution yields 2 : 3-dimethyl-p y r a ~ i n e . ~ ~ Condensation of ethylenediamine with benzil' also gives adihydropyrazine derivative which on distillation is converted into 2 : 3-di-phenylpyrazine.40 Treatment of glucose with ammonia gives a mixture ofbases including pyrazine, 2-methyl-, 2 : 5-dimethyl-, and 2 : 6-dimethyl-p y r a ~ i n e s , ~ ~ and 2 : 5-bistetrahydroxybutylpyrazine is obtained by theaction of ammonia upon fructose.42 2 : 5-Dimethylpyrazine is said to bereadily obtained by distillation of glycerol with certain ammonium salts.43According to R.Leuclcart,44 treatment of benzoin with ammonium formategives a nearly quantitative yield of tetraphenylpyrazine. Using form-amide instead of ammonium formate, A. Novelli 45 obtained 4 : 5-diphenyl-glyoxaline as major product together with a small amount of tetraphenyl-pyrazine; this reaction is said to be general for aromatic acyloins.Whenbenzoin is heated with ammonium acetate in acetic acid solution, a mixtureof tetraphenylpyrazine and 4 : 5-diphenyl-2-methylglyoxaline is obtained.4636 S. Gabriel and G. Pinkus, Ber., 1893, 26, 2207; 1908, 41, 960; L. Wolff, ibid.,37 J. G. Aston, T. E. Peterson, and J. H'olowchak, J . Amer. Chent. SOC., 1934,56, 153.38 S. Gabriel and A. Sonn, Ber., 1907, 40, 4850; C. Stoehr, J . p r . Chem., 1894,38 E'. Jorre, Diss., Kiel, 1897.4 0 A. T. Mason, J., 1889, 55, 99; 1893, 63, 1297.41 P. Brandes and C . Stoehr, J . p r . Chem., 1896,54, 481 ; C. Tanret, Bull. SOC. c h h . ,1885, 44, 103; 1897, 17, 801; Compt. rend., 1885,100, 1540.42 Lobry de Bruyn, Rec. Trav. chim., 1899, 18, 72, 81 ; R.Stolte, Biochem. Z . , 1908,4 3 C . Stoehr, Ber., 1891, 24, 4105; J . p r . Chem., 1893, 47, 439; 1895, 51, 449;D.R-P. 73,704; 75,298; A. Etard, Compt. rend., 1881, 92, 460; M. Dennstedt, Ber.,1892, 25, 259.1893, 26, 1830; 1888, 21, 1483.49, 392.12, 499.44 J . p r . Chem., 1890,41, 330; J . Org. Chem., 1938,2,328.4b Anal. Asoc. Quim. Argentina, 1939, 27, 161.46 D. Davidson, M. Weiss, and M. Jelling, J . Org. Chem., 1937, 2, 328192 ORGANIC CHEMISTRY.Only a very small yield of tetraphenylpyrazine is obtained by the action ofliquid ammonia upon ben~il,*'?~~ but K. Bulow48 has shown that thispyrazine is obtained in reasoiiable yield by the action of formamide uponbenzaldehyde. Treatment of a-benzil dioxime with potassium ferrocyanideyields tetrapheiiylpyrazine dioxide which on reduction with zinc dustand acetic acid gives tetraphenylpyrazine.49Properties of Pyraxine and its Homobgues.-Pyrazine and its homologuesare monoacidic bases, neutral to litmus.The lower homologues are volatilecompounds boiling without decomposition and are miscible in all proportionswith water. They are insoluble in alkaline solutions and can be precipitatedfrom aqueous solution by the addition of alkali. They are hygroscopicand readily form crystalline hydrates. The parent pyrazine has b. p.115"/730 mm. and separates from a concentrated aqueous solution as prisms,m. p. 55°.50 Very characteristic of pyrazine and its lower homologues aretheir great volatility and tendency to sublime even at room temperature inclosed vessels.Pyrazines possess a characteristic odour comparable withthat of the higher pyridine bases. They are weaker bases than pyridine,the introduction of aromatic substituents decreasing the basic strength ;thus, 2 : 5-diphenylpyrazine is soluble in concentrated hydrochloric acidbut it is precipitated from this solution on dilution with water.51 2 : 5-Di-methylpyrazine, probably the most fully examined me-mber of the series, hasbeen characterised by the preparation of a monohydrochloride, a picrate,an aurichloride, and a mon~methiodide.~~ When suitably reduced, thepyrazines are converted into the corresponding piperazines ; 53* 52 conversely,piperazines can be oxidised to the corresponding pyyrazines.61 Towardsoxidising agents the pyrazine nucleus resembles that of pyridine ; pyrazinehomologues are readily oxidised to pyrazinecarboxylic acids.2 : 5-Di-methylpyrazine can be oxidised stepwise t o 2-methylpyrazine-5-carboxylic ,acid and thence to pyrazine-2 : 5-dicarboxylic a ~ i d , 5 ~ and oxidation ofquinoxaline gives pyrazine-2 : 3-dicarboxylic acid.55 Methyl groups attachedto the pyrazine ring are reactive and condense with aldehydes to yieldstyryl derivatives ; 56 thus 2 : 5-dimethylpyrazine condenses with benz-aldehyde in the presence of zinc chloride to give the styryl derivatives (IX)4 7 W. B. Leslie and G. W. Watt, J . Org. C'he?it., 1942, '7, 73.4 8 Ber., 1893, 26, 1830.49 E. Durio and M. Bissi, Gazzetta, 1930, 60, 899 ; K. v . Auwers and V. Meyer, Ber.,1888, 21, 806.L.Wolff, ibid., 1893, 26, 721.61 C. Stoehr, J . pr. Chem., 1893,47, 439.5 2 Idem, Ber., 1891, 24, 4105.63 L. Wolff, ibid., 1893, 26, 722, 725; M. Godchot and M. Mousseron, Bull. SOC.chim., 1932, 51, 349; Compt. rend., 1930, 190, 798; F. B. Kipping, J., 1929,2889.54 C. Stoehr, J. pr. Chem., 1893, 47, 447, 476; 1894, 49, 397; 1895,51, 463; 1896,54, 490.55 S. Gabriel and A. Sonn, Ber., 1907, 40, 4852 ; J. W. Sausville and p. E. SPoerri,J. Amer. Chem. SOC., 1941, 63, 3153.66 R. Franke, Ber., 1905, 38, 3724NEWBOLD AND SPRING : PYRAZINE AND ITS DERIVATIVES. 193and (X). 2 : 6-Dimethylpyrazine has zero dipole moment; 57 electrondiffraction studies on pyrazine indicate a C-N link of 1.35 A.58Pyrazinecarboxylic acids can be obtained by oxidation of pyrazinehomologues or of quinoxalines.Pyrazine-2 : 3-dicarboxylic acid (and its5 : 6-disubstituted derivatives) can be obtained by condensation of hydrogencyanide tetramer with 1 : 2-dicarbonyl compounds followed by hydrolysisMe( I p = C H * P h Ph*CH=CH( N/ \CH=CH*PhN(X.)of the intermediate dicyanide (XI) .59obtained by reduction of the isonitroso-derivatives of @-keto-esters.Pyrazinecarboxylic esters have beenThus,N f & (XI.)H,N-CH-CN CHOHN=C-CN--+-I- bHO \N/Ireduction of the isonitroso-derivative of ethyl acetoacetate (XII) followedby treatment with alkali and an oxidising agent gives ethyl 2 : Fi-dimethyl-pyrazhe-3 : 6-dicarboxylate (XIII, R = Et). The reduction has beenaccomplished by stannous chloride and hydrochloric acid 60 and by catalyticmethods.s1 C.Gastaldi has developed a method for the preparation of the(XII.) (XIII.)acid (XIII, R = H) in which the bisulphite compound of isonitrosoacetoneis treated successively with potassium cyanide and hydrochloric acid to give3 : 6-dicyano-2 : 5-dimethylpyrazine (XIV) .62 Alkaline hydrolysis of thedicyanide does not convert it into the corresponding dicarboxylic acid but1 (XV.) (XVI.)yields 2-hydroxy-3 : 6-dimethylpyrazine-5-carboxylic acid (XV). Con-version of the dicyanide (XIV) into the corresponding acid has been achieved67 A. E. van Arkel and J. L. Snoek, Rec. Trav. chirn., 1933, 52, 719, 1013.68 V. Schomaker and L. Pauling, J. Amer. Chem. SOC., 1939,81, 1769.m Grhchkevitsch-Trochimovski, Rocz.Chem., 1928, 8, 165; L. E. Hinkel, G. 0.Richards, and 0. Thomas, J., 1937, 1432; R. P. Linstead, E. G. Noble, and J. M.Wright, ibid., p. 911.6o S. Wleugel, Ber., 1882, 15, 1050; S. Gabriel and T. Posner, ibid., 1894, 27, 1141;V. Cerchez and C. Colesui, Bull. SOC. chim., 1931, 49, 1291.O1 H. AdkinsandE. W. Reeve, J. Amer. Chem. Xoc., 1938, 60, 1328.62 Gazzetta, 1921, 51, 233.REP.-VOL. XLII. 194 ORGANIC CHEMISTRY.by C. Gastaldi and G. Princivalle 63 by acid hydrolysis to the diamide (XVI)which with nitrous acid gave the required acid. Pyrazine-2 : 5-dicarboxylicacid has been obtained by the action of ammonia upon dihydroxymaleicacid.&4 H. E. Fierz-David and E. Ziegler 65 have shown that, after couplingacetoacetanilide with a diazonium salt solution, reduction of the productwith alkaline hydrosulphite gives the anilide corresponding to (XIII).Properties of Pyraxinecarboxylic Acids.-Pyrazinecarboxylic acids givered-violet colourations when treated with aqueous ferrous sulphate.Theionisation constants of pyrazinemonocarboxylic acid and pyrazine-2 : 3 4 -carboxylic acid are 1.2 x 103 and 1.7 x (first ionisation constant)respectively.66 Pyrazine-2 : 3-dicarboxylic acid has been converted intophthalein-like compounds. Thus with resorcinol, (XVII) is obtained ;this dissolves in alkali to give a blood-red-coloured solution which on dilutionwith water shows an orange-green fluore~cence.~~ Pyrazine-2 : 3-dicarboxy-amide (XVIII) when treated with two molecular proportions of alkalinehypobromite solution undergoes an intramolecular rearrangement and giveslumazine (XIX).68 2 : 3-Dicyanopyrazine has been converted into phthalo-cyanine-like compounds when heated with suitable metallic reagents.69(XVIII.) (XIX.)Although antipellagra activity without the vasodilator effect of nicotinicacid is claimed for pyrazinecarboxylic acid and pyrazine-2 : 3-dicarboxylicacid,70 both acids were ineffective in the treatment of black tongue.'l Bothacids also appear to act as growth factors for Proteus vulgaris and Strepto-bacterium plantarium, but they act only in much greater concentration thandoes nicotinic acid.72 A series of alkyl-substituted pyrazine-carboxyamides63 Bazzetfa, 1928, 58, 412.135 Helv. Chim. Acta, 1928, 11, 776.G 6 J. W. Sausville and P. E. Spoarri, J . Amer. Chem. SOC., 1941, 83, 3153.6 7 S. C. De and P. C. Dutta, Ber., 1931,64, 2606.68 S. Gabriel and A. Sonn, ibid., 1907, 40, 4855; R. A. Baxter and F. S. Spring,64 H. J. H. Fenton, J., 1905, 87, 806.J., 1945, 229.R. P. Linstead, E. G. Noble, and J. M. Wright, J . , 1937, 911.70 C . E. Bills, F. G. McDonald, and T. D. Spies, Sth. Med. J., 1939, 32, 793.71 W. J. Dann, H. I. Kohn, and P. Handler, J . Nutrition, 1940, 20, 477.7 2 E. F. Moller and L. Birkofer, Ber., 1942, 75, 1108NEWBOLD AND SPRING: PYRAZINE AND ITS DERIVATIVES. 195and -hydrazides have been prepared as potential analeptics 73 and the pre-paration of NN’-dibenzyl- and NN’-diaryl-pyrazine-2 : 3-dicarboxyamidesis described with a view to their employment as antispasm0dics.~4 Pyrazinoylderivatives of sulphanilamide 75 have also been described.Amino-pyraxines.-Treatment of pyrazine-2 : 3-dicarboxyamide withone molecular proportion of alkaline hypobromite solution gives Z-amino-pyrazine-3-carboxylic acid which on heating yields aminopyrazine ; 68the latter has also been obtained by a normal Hofma.nn reaction uponpyrazine~arboxyarnide.~~ The dihydrazide of pyrazine-2 : 5-dicarboxylicacid has been degraded to the corresponding diisocyanate, but this compoundproved remarkably resistant to hydrolysis as did the corresponding diurethaneobtained via the dia~ide.~’ Aminopyrazine was not obtained by treatmentof pyrazine with potassium amide in liquid but 2 : Ei-dimethyl-pyrazine has been aminated using sodamide to give 2-amino-3 : 6-dimethyl-p y r a ~ i n e . ~ ~ Lumazine, which can be obtained either as described abovefrom pyrazinedicarboxyamide or by treatment of glyoxal with 4 : 5-diamino-2 : 6-dihydro~ypyrimidine,~~ and substituted lumazines (XX) can behydrolysed using either acid or alkaline conditions t o give aminopyrazines(XXI) and 2-aminopyrazine-3-carboxylic acids (XXII) respectively : 81(=.) (XXI.) (XXII.)The amino-acids (XXII) are smoothly decarboxylated when heated togive the corresponding aminopyrazines (XXI). Sulphanilyl derivativesof various aminopyrazines have been much investigated particularly in theUnited States. Single large doses of sulphapyrazine (XXITI) 66,82 given tomice infected with p-hemolytic streptococci are claimed to be considerablymore effective than similar doses of sulphathiazole, sulphapyridine, or sulph-anilamide and equally effective as a similar dose of s~lphadiazine.~~ The73 D.R.-P. 632,257; Canad. P. 378,818; U.S.P. 2,149,279.74 J. H. Billman and J. L. Rendall, J . Amer. Chem. SOC., 1944,66, 540.‘5 T. C. Daniels and H. Iwamoto, ibid., 1941, 63, 257.76 5. A. Hall and P. E. Spoerri, ibid., 1940, 82, 664.77 P. E. Spoerri and A. Erickson, ibid., 1938, 60, 400.79 A. E. Tschitschibabin and M. N. Schtschukins, J . Russ. Phys. Chem. SOC.. 1930,F. W. Bergstrom and R. A. Ogg, ibid., 1931, 53, 245.62, 1189; R. R. Joiner and P. E. Spoerri, J . Amer. Chem. SOC., 1941,63,1929.R. Kuhn and A. H. Cook, Ber., 1937,70, 761.J Weijlard, M. Tishlor, and A. E. Erickson, J. Amer. Chern. SOC., 1945,67,802.82 G. W. Raiziss, L. W. Clemence, and M. Freifelder, ibid., 1941, 63, 2739; R. C.Ellblgson, ibid., p. 2524; H. J. Robinson, H. Siegel, and 0. Graessle, J . Pharmacol.,1943, 79, 354; G. I. Trevett, Bull. Johns Hopkins Hosp., 1944, 74, 299.83 W. H. Schmidt and C. L. Sesler, J. Phamacol., 1943, 77, 277196 ORGANIC CHEMISTRY.absorption, distribution, and excretion of sulphapyrazine in man is de-scribed,84 and a comparative study of its efficiency in the treatment ofpneumococcal infections is made.86$ 86~y~roxy-~yraxines.-2-Hydroxy-3 : 6-dimethylpyrazine was obtained byC . Gastaldi 62 by decarboxylation of 2-hydroxy-3 : 6-dimethylpyrazine-5-carboxylic acid (XV), and a similar synthesis of 2-hydroxy-3 : 6-diphenyl-pyrazine is recorded. A synthesis of 5 : 6-di- and 3 : 5 : 6-tri-substituted-2-hydroxypyrazines has been developed by Y. A. Tota and R. E. Elderfield.87It consists in the condensation of an a-amino-ketone with an a-halo-acid halidefollowed by treatment of the+ P H 2 0 B rMeCOMe.AH*NH,,HCl CO*Brproduct with ammonia, thus :BrCauo, MeCO - I Me*CHTishler, and A. E. Erickson have shown that More recently, J. Weijlard, Mdrastic alkaline hydrolysis of lumazine or 2-aminopyrazine-3-carboxylic acidgives 2-hydroxypyrazine-3-carboxylic acid, decarboxylation of which yieldshydroxypyrazine :/N\/NH\p IPhenylglyoxal or functionally related compounds can cyclise, given anappropriate source of nitrogen, to give either 2-hydroxy-3 : 6-diphenyl-pyrazine or 2-benzoyl-5-phenylglyoxaline.88 Hydroxypyrazines are ampho-teric. 2-Hydroxy-3 : 6-dimethylpyrazine couples with diazonium salts toyield crystalline 5-arylazo-derivatives, and 2-hydroxy-3 : 6-dimethyl-pyrazine-5-carboxylic acid also couples with loss of carbon dioxide to givethe s$me azo-derivatives. 89 The formation of certain quaternary ammoniumsalts from 2-hydroxy-3 : 6-dimethylpyrazine and their conversion intocyarline dyes is claimed by C. Gastaldi and E. Princivalle.goG. T. N.F. S. S.84 M. Hamburger, J. M. Ruegsegger, N. L. Brookens, and E. Eakin, Amer. J.Med.85 L. H. Schmidt, J. M. Ruegsegger, C. L. Sesler, and M. Hamburger, J . Pharmacol.,86 For other sulphanilyl derivatives of aminopyrazines see R. C. Ellingson, R. L.8 7 J . Org. Chem., 1942, 7 , 313.88 H. Muller and H. v. Pechmann, Ber., 1889, 22, 2557; A. Pinner, ibid., 1906,38, 1531 ; C. Engler and E. Hassenkamp, ibid., 1885,18, 2240; S. Minovivi, ibid., 1899,32, 2206; F. R. Japp and N. H. J. Miller, J., 1887, 30; F. R. Japp and J. Knox, ibid.,1905, 701 ; M. Busch and W. Foerst, J. pr. Chem., 1928,119, 287.8Q C. Gastaldi and E. Princivalle, Cfazxetta, 1928, 58, 679; E. Princivalle, ibid., 1930,80, 298.90 Ibid., p. 412; Annuli Chim. Appl., 1936, 26, 450.Sci., 1942, 204, 186.1941, 73, 468; Amer. J . med. Sci., 1941, 202, 432.Henry, and F. G. McDonald, J. Amer. Chem. SOG., 1945, 67, 1711