摘要:

Acetic AciidAcetic AnhydrideAeetoacetic BsterAcetoneAldolAlkyl OleatesAlkyl StearatesAmy1 AcetateButyl AcetateBatyl AlcoholButyl AldehydeCrotonaldebydeDiacetinDiacetone AlcoholDiethyl OxalateEthyl AcetateLactatesParaldehydePhthalatesTriacegin7hey are all manufwtured in England byBRITISH INDUSTRIAL SOLVENTS LTDWELBECK HOUSE, DOWNS SIDE, BELMONT, SURREYTELEPHONE : VIGILANT 0133iM E D I CA L EVANS PRODUCTSBacteriological Peptone(Evans)A new product developed in the research laboratories ofTHE EVANS BIOLOGICAL INSTITUTE, and released tomeet an urgent demand from bacteriologists and mediamakers for a reliable British-made Peptone.Bacteriological Peptone (Evans) i s a distinctive producthaving a variety of applications and exhibiting the followingfeatures :7 It i s free from contaminating organisms and toxic7 It contains no fermentable carbohydrate.7 Unlike many of the older peptones it provides insaline solution ALONE a complete medium for awide variety of bacteria.7 With the addition of dextrose it produces anexcellent growth of pneumococci, haemolyticstreptococci, and organisms of the coli-typhoidand welchii groups.7 No serum need be added when it i s used fortyping streptococci for the fermentation reactions.7 It conforms t o the requirements of the lndol test.7 Potent diphtheria, tetanus and other toxins canbe obtained with media containing this peptonaBacteriological Peptone (Evans) is issued in 5 and 8-oz.amber bottleswith bakelite screw caps.Home Medical Dept., Concert Street, Liverpool, I .by-products.Prices and an illustrated brochure will be sent on application toMade in England atThe Evans Biological Institute byEvans Sons Lescher & Webb Ltd.Liverpool and LondonvMICA, LEATHEROID,VULCANISED FIBREEMPIRE CLOTH AND TAPEB A K E L I T ESHEETS, TUBES, SPOOLS,etc.0B A K E L I T ER E S I N , V A R N I S HM O U L D I N G POWDERANDand aaE B O N I T EI Insulating Mater a forElectrical Man ufact u rersATTWATER & SONS, Ltd.ESTABLISHED 1868P R E S T O N E N G L A N DiBIOLOGICAL m M BE PREPARATIONSANTIPEOL ENcEoNu~ OINTMENTcontains sterile vaccine filtrates (antivirus) of all the common strains ofSTAPHYLOCOCCI, STREPTOCOCCI and B. PYOCYANEUS in a lanoline-zinc-ichthyol base.SPECIFIC AGAINST the micro-organisms causing abscesses, boils, eczema,dermatitis, erysipelas, hamorrhoids, impetigo, ulcers and all inflammatorycutaneous affections.The BURN THERAPY par excellence.ANTIPEOL LIQUID for infections of the ear, septic cavities and suppuratingwounds.OPHTHALMO-ANTIPEOLcontains, in a semi-fluid base, the sterile filtrates of STAPHYLOCOCCI,STREPTOCOCCI, B. PYOCYANEUS, PNEUMOCOCCI FRAENKEL andGONOCOCCI.INDICATIONS : Conjunctivitis, blepharitis, keratitis, dacriocystitis and allinflammatory conditions and lesions of the eye.RHI NO-ANTI PEOLa nasal immunising cream, contains Antipeol Liquid as well as the antivirusand autolysins of PNEUMOCOCCI, PNEUMO-BACILLI, ENTEROCOCCI,M. CATARRHALIS, B.PFEIFFER and calmative and decongestive ingredients.INDICATIONS : Coryza, rhinitis, hay fever, catarrh, influenza, colds and othernaso-pharyngeal infections.ENTEROFAGOSpolyvalent bacteriophages specific against 144 strains of micro-organismscommon to infections of the gastro-intestinal tract, kidneys and bladder.RAPIDLY EFFECTIVE RESULTS in enteritis, dysentery, colitis, diarrhoeas,B. coli infections, typhoid and para-typhoid fevers, and other intestinal andpara-intestinal infections. Oral Administration. No reaction. No shock.DETENSYLvegeto-polyhormonic hyportensor ensures gentle and regular reduction ofarteria tension.INDICATIONS : High blood pressure, arthritis, arterioscleyosis, palpitation,ocular and auditory troubles of hypertension.No contra-indications.CLINICAL SAMPLES AND LITERATURE FROM:MEDICO-BIOLOGICAL LABORATORIES LTD.,CARGREEN ROAD, SOUTH NORWOOD, LONDON, S.E.25C H I ” CLAYS (Kuolin)BALL CLAY (Argile)CORNWALL STONE (China Stone)FELSPARBENTONITE (and British Substitutes)TALC (and British Substitutes)for use in themanufacture of:PaperTextilesEarthenwarePorcelainRefractoriesCruciblesFurnace LiningsInsulatorsCementStonewareGlassPaintsEnamelsRubberUltramarineInksInsecticidesMatchesMedicinesPou It icesCosmeticsPolishesRoofing Feltsetc.0000THERE ARE MANY TYPES OF CHINA CLAY,BALL CLAY, etc., EACH HAVING ITS OWNPECULIARITIES. THEY DIFFER IN ANALYSES.PARTICLE SIZE, COLOUR, SUSPENSION, Etc.Let us knowwhat you are seekingand we will make recommendations and supplycopious samples for tests.Our British substitutes for foreign minerals arein daily increasing demand.A few of our SPECIAL KAOLINS supplied in thefinest powder form :STOCKALITE 0 DEVOLITE0 COLWHITE 0SPESWHITE 0 SUPREMETALC SUBSTITUTES :TALCOLIN TALCOLITE, Etc.EWGLISH CHINA CLAYS SALES GO., LTD.ST. AUSTELL CORNWALLAlso at London, Stoke, Manchester, Newcastle, Edinburgh, etc.Depots and Stores in many parts ofthe country.felegramr : Universal, St.Austell. ’Phone : 483 St. AurtrlSOUTH-WEST ESSEX TECHNICALCOLLEGE AND SCHOOL OF ART(Essex Education Committee)FOREST ROAD, WALTHAMSTOW, E.17H. LOWERY, M.Ed., Ph.D., D.Sc., F.1nst.P.Principal :SCIENCE DEPARTMENTHead of Department :R.W. JUKES, B.Sc., A.I.C., F.C.S.H. HOLNESS, B.Sc., A.I.C., M.I.T.J. BOR, M.Sc.Tech., F.1nst.P.W. B. CROW, Ph.D., D.Sc.Senior Lecturer in Chemistry :Senior Lecturer in Physics :. Senior Lecturer in Bio1og)L:Full-time Day Courses extending over 3 or 4 years areheld in preparation for the London University B.Sc.,General and Special Degrees in Chemistry, Physics, Botany,Zoology, and Mathematics ; also for A.I.C. and A.1nst.P.Many research investigations are in progress in theScience Department and students may become associatedwith any of these with a view to taking either the M.Sc.or Ph.D. Degree.Part-time Degree and National Certificate Courses arealso available in Science subjects.Special facilities are available for the study of industrialapplications of science, e.g., plastics, radiology, electronics,chemical engineering.The College was opened in 1938 and replaces the formerLeyt on and Walthamstow Technical Colleges.The buiIding is entirely new and the equipment up-to-date.In addition to excellent teaching accommodation,the College possesses ample social amenities, viz., concerthall with cinema and stage, swimming bath, two gymnasia,library, refectory, writing and private study rooms, andcommon rooms.Full particulars and Pros$ectw of classes may be obtainedon a$plication to the Principal.xvSCIENTIFIC BOOKS~~~ ~~~ H. K. LEWIS & Co. Ltd.A very large selection of new and standard works in everyThe Department for Scientific Books, English and Foreign, is onthe first floor.Foreign Books not in stock obtained under Licence.Orders and Inquiries by Post promptly attended to.branch of Science always available.Underground : Euston Square, Warren Street.'Buses : Euston Road and TottenhamCourt Road.SCIENTIFIC LENDING LIBRARYAnnual Subscription, Town or Country, from One GuineaThe LIBRARY is useful to SOCIETIES and INSTITUTIONS, andto those engaged on SPECIAL RESEARCH WORK, etc. TheLibrary includes all Recent and Standard Works in all branches ofMedical and General Science.Reading and Writing Room (first floor) open daily.New Books and New Editions are added to the Library and areavailable to Subscribers immediately on publication.Catalogue of the Library, revised to December 1937, with Supplement1938-1 941, containing Classified Index of Authors and Subjects, 20s.net(to Subscribers 10s. net). Supplement separately 4s. net (to Subscribers2s. net) postage 3d.Bi-monthly List of New Books and New Editions is issued free to allSubscribers and Bookbuyers regularly.Detailed Prospect us on Application.Every work is the latest edition.STATIONERY DEPARTMENT : Scient$c and General.Loose-Leaf Notebooks* Record Cards, Filing Cabinets,Slide Rules, Graph-papers, etc.SECONDMHAND BOOKS : 140 GOWER STREET.Large and varied stock.Books not in stock sought for, and reported when found.Hm Km LEWIS & Corn LtdmPUBLISHERS AND BOOKSELLERS136 GOWER STREET, LONDON, W.C.1l?&rm: Telephone :'PUBLICAVIT, WESTCENT, LONDON .' EUSton 4282 (5 lines)XviTHE POLYTECHNICRegent Street London, W.1DEPARTMENT OF CHEMISTRY &’ BIOLOGYHead of Department : H. LAMBOURNE, M.A., M.Sc., F.I.C.DAY COURSESB.Sc. Degree Special and General (External), LondonAssociateship of the Institute of Chemistry (A.I.C.)First Medical, Pre-Medical and Preliminary Scientific, University.ILDiploma.Courses in Chemistry, Biology and Physics.Prospectus free on application to theDirector of EducationPerkin and Kipping’sORGANIC CHEMISTRYRevised Edition. By F. STANLEY KIPPING, Ph.D., Sc.D., F.R.S.,formerly Professor of Chemistry, University College, Nottingham,and F. BARRY KIPPING, M.A., Ph.D., University Demonstrator inChemistry, Trinity College, Cambridge.Part I, 368 pages, 5s.6d. ; Part 11, 328 pages, 5s. 6d.The work has been brought thoroughly up to date in accordancewith the latest developments of the science.Part 111, Revised Edition, 428 pages, 10s. 6d.This part has been written as a continuation of Parts I. and II., andis intended mainly for the use of students working for an HonoursDegree Examination.Parts I. and 11. in One Volume. 10s. 6d.Parts I., 11. and 111. in One Volume. 20s.38 Soh0LONDON,Sq., w.l W. & R. CHAMBERS 11 T h W W, LTD. ED 1 W B U RGTHEROYAL INSTITUTE OF CHEMISTRYOF GREAT BRITAIN AND IRELANDThe Institute of Chemistry was established in 1877 to providethe Government and the public with the means of recognis-ing those who have been properly trained and proved to becompetent to practise chemistry as a profession.In 1885 theInstitute was granted a Royal Charter with authority to grantcertificates of competency, and to register persons qualifiedto practise. The aims of the Institute include the elevationof the profession of chemistry and the maintenance of theefficiency, integrity and usefulness of persons practising thesame, by compelling the observance of strict rules ofmembership, and by setting up a high standard of scientificand practical efficiency.Particulars of the Regulations and Examinations of theInstitute can be obtained (free) on application.All communications to be addressed to the Registrar,THE ROYAL INSTITUTE OF CHEMISTRY30 RUSSELL SQUARE, W.C.1UNIVERSITY OFST. ANDREWS (SCOTLAND)Chancellor: The Right Hon. The EARL BALDWIN OF BEWDLEY.Vice-ChnceUor and Princifial: SIR JAMES COLQUHOUN IRVINE.Dean ofthc Faculty of Science: Professor ROBERT JAMES DOUGLAS GRAHAM.The Unlversity confers the following De-, all open to men or womeri:B.Sc. (Ord. and Hons.), Ph.D., D.Sc.SESSION 1943-44 opens 12th October, 1943. The curriculum for Pure Sdencemay be taken in St. Andrew8 or Dundee, the curriculum for Engineering inDundee. Application for admission to the Faculty should be sent to the Dean notlater than 30th June.Residential Entrance Scholarshipe for Men and Entrance Bursary Competitions.-June annually. Entries due 3rd May. United College, St. Andrews.-ResidentialEntrance Scholarships-Seven of €100 per annum and one of 670 per annum tenablefor 3 or 4 years.Bursaries open to Science Students-Six of 650, two of €40, one of €30five of lesser amounts from L28 to LIZ, tenable for 3 or 4 years. University CollegiDundee-Bursaries open to Science Students-Ten of from k50 to l30, tenable for 3 0;PreUmhwy Examinstions.-March and August. Entries due 6th February and 23rdJuly.Post-Gmduate Study and Research in Chemistry, Mathematics, Astronomy, NaturalPhilosophy, Zoology, Botany, Geology, Anatomy, Physiology, Engineering, Provisionmade for duly qualified Research Workers. Students of the University are eligible forCarnegie Trust Scholarships and Fellowships and for other Scholarships, Fellowships andGrants.Mathwn Chemistry schoiarshfpe and B ~ r ~ a r i e ~ .- A Scholarship of kroo for entrantstudents for three years and a Bursary of 630 for students entering the Honours ChemistryClass, for one year, tenable at United College, St. Andrews.Chemistry Research Students should communicate with Professor Read, United College,or Professor Wynne-Jones, University College, Dundee.Residence Halls for Men Studenta at St. Andrews.-St. Salvator’s Hall, Dean’s Court,Swallowgate.Full information may be obtained from the Sec of the University, St. Andrews,or the Dean of the Faculty of Science, Bute=d Buildings~, St. Andrews.Rector: Air Vice-Marshal SIR DAVID MUNRO.4 years.xiCHEMICAL PLANTREBUILT & GUARANTEEDBYBARBERS’ENSURESCOMPLETESATISFACTIONOur comprehensive stock at Hayesincludes-STILLS fabricated In varlous matrlsMIXERS & KNEADERS,STEAM JACKETED MIXERS & PANS,CALORIFIERS, CONDENSERS,FRACTIONATING COLUMNS,HYDRO EXTRACTORS,VACUUM OVENS & PUMPSDEPHLEGMATORS & HEATEXCHANGERS,GLASS LINED EQUIPMENT,KESTNER FILM DRYERS,VACUUM DISTILLATION PLANT,TRIPLE ROLL MILLS,DISINTEGRATORS & BALL MILLS,SCREENING PLANT,CRYSTALLISERS. EVAPORATORS,CONFECTIONERY PLANT,GAS PRODUCERS, ACID PUMPS,DUST EXTRACTORS,MILLING & GRINDING PLANT,ALUMINIUM & COPPER VESSELS,SULPHONATORS, CENTRIFUGES,VULCANISERS,PAINT PLANT, DYEING PLANT,FUEL BRlQUElTlNG PLANTS,BELT CONVEYORS,ELEVATORS,WORM & STEEL TRAY CONVEYORS,ETC., ETC.FILTER PRESSES dl types & tlset.AUTOCLAVES 811 typm.BRICK-MAKING MACHINERY,WE HOLD COMPREHENSIVE STOCKSOF REBUILT & GUARANTEEDCHEMICAL PLANT BY LEADINGMAKERS.IMMEDIATE DELIVERY OF PLANTREQUIRED BY THE CHEMICAL &ALLIED TRADESC.BARBER LTD.CHEMICAL ENGINEERSandMA CHINE RY MERCHANTSSILVERDALE GARDENSHAYES, MIDDX.T r l c p h c : Telegrams :Hayes, Middx. “Barchem,”73s/6 Hayes, Wddx.FOYLES * BOOKSELLERS TOTHE WORLD*Large departments for Books onChemistry, Physics and all Branchesof Science. Nearly 3 million Vol-umes in Stock. Books Bought.JOIN THE SCIENTIFICBOOK CLUBMembers Obtain Books Cost-ing 5/-, 716, 1016 or 3511, forONLY 216LONDON, W.C.2119-125 CHARING CROSS ROAD,Open g a . m . 4 p.m., including Saturday.Tekphone: GERrard 5660 (16 lines).The fact that goodsmade of raw materialsin short supply owingto war conditions areadvertised in this Bookshould not be takenas an indication thatthey are necessarilyavailable for exportPURE HYDROGENPure Hydrogen (99’95%) for hydrogenationand metallurgical processes, together withoxygen (99’8%), are produced direct from thecell without the need for further treatment.SIMPLY PRODUCEDThe plant is arranged so that every part ofthe process is simplified and safe, withoutcornpHcation or the need for specially skilledlabour.AT LOYYCOSTAs a result, cost of operation and maintenanceare particularly low,renewnls are at aminimum,installations show a very long life. Plants ofall capacities have been supplied to mostcountries of the world.THE INTERNATIONALELECTROLYTIC PLANT CO. LTD.SANDYCROFT : CHESTERxxiPURIFICATIONOFWATERFOR ALL PURPOSESBOILER FEEDPROCESS WORKTEXTILE PURPOSESTOWN SUPPLY, Etc.IS THE SPECIALITY OFJOHN THOMPSON(KEMNICOTT WATER SOFTENERS) LTDHEAD OFFICE AND WORKSWO LV E RH A M PTO NEstablished half a centuryPLACE YOUR WATER PROBLEMSBEFORE US AND AVAIL YOUR-SELF OF OURUnrivalled ExperienceAlso Manufacturers of Continuous Blow-Down Apparatusfor BoilersTo face tillr page. sxvii

ISSN:0365-6217    

DOI:10.1039/AR94239BP001

出版商:RSC    

年代:1942

数据来源: RSC

摘要:

ANNUAL REPORTSON THEPROGRESS OF CHEMISTRY.GENERAL AND PHYSICAL CHEMISTRY.1. INTRODUCTION.THE nature of the Reports to follow has been altered to some considerableextent as compared with those of former years. In the past there has been atendency to produce reviews which often were annual reports in therestricted sense of the term. Within the last few years an effort has, however,been made to render these reports of more general interest by confining theReporter to one topic within the particular branch of physical chemistrybeing discussed. This year the process is being carried a stage further as awartime experiment and expedient in the hope that these Reports willappeal to all interested in physical chemistry whilst retaining their authori-tative and complete character.The first report is concerned with the physicochemical behaviour of rubbersolntions.From the thermodynamic point of view it is of great interest toknow something about the behaviour of solutions of high-molecular sub-stances because of their unusual properties ; but simultaneously that know-ledge proves to be extremely iiseful in learning a good deal about thestmcture of the rubber molecule itself.In the next section there is an account of the reactions of sodium atomswith a variety of organic and inorganic halides. These reactions have ingeneral a low activation energy and their study has only been made possibleby the extensive development of the technique of the so-called dilute“ flames.”Finally, the interesting developments in the, physical chemistry of theformation of the photographic latent image are recorded.In spite of mucheffort it is only recently that some reliable insight has been gained into theprecise nature of the reaction of radiation with photographic emulsions andthe manifold resultant effects that have been noted almost since thebeginning of photography. H. W. M.2. THE PHYSICAL CHEMISTRY OF RUBBER SOLUTIONS.THE physicochemical behaviour of rubber may conveniently be describedin terms of two main factors, (1) its long-chain character, and (2) i t s chemicalnature. It is frequently possible to neglect one of these ; e.g., E. H. Farme8 GENXRAL AND PHYSICAL CHEMISTRY.et a2.,1 have shown that much light can be thrown on the organic chemistry ofrubber from a study of structurally related, low-molecular hydrocarbons.In the following we shall, in the main, consider rubber as a typical memberof the class of long-chain compounds, and much of this section is thereforeof general application to linear high polymers.The Micellar Theory of Rubber Structure.In view of this line of approach, it is natural to consider first the evidencefor the size and shape of the rubber molecule.Two main points of view havebeen expressed. According. to the first,2 rubber consists of molecules havinga molecular weight of ca. 1000-1500, which in solid rubber are united in theform of “ micelles.” These larger units are supposed to have sufficientindividuality to resist dispersion by most solvents, so that, e.g., the osmoticpressure of a benzene solution measures the size, not of the molecule, but ofthe micelle.Only such solvents as camphor, menthol, and related materialsare considered to break down the rubber to give a molecular dispersion. Thealternative view 3 regards rubber as truly macromolecular, i.e., made up ofvery long chains of isoprene molecules, polymerised into single moleculeswhose size, measured osmotically, is of the order of 100,000. A recentcritical examination 4 seems to disprove the whole of the evidence advancedin support of the former view. The results of the earlier workers were shownto arise primarily from the rapid oxidative breakdown of rubber in moltencamphor. When this is prevented by the exclusion of air, camphor behavesvery similarly to benzene as a rubber solvent.The abnormally low molecularweights reported were obtained by the Rast method and are now shown to bemerely indicative of the “ non-ideal ” behaviour which is characteristic ofrubber solutions in general (see below). In default of any sound evidence forthe micellar theory, we therefore conclude that rubber is to be classed withthe synthetic polymers as a macromolecule.Osmotic Pressure Measurements.The molecular weight of rubber may be determined by three methods :(a) ultracentrifugal, (b) osmotic, and (c) viscosimetric. Of these, ( c ) is not anabsolute method, and ( a ) has been applied to a very limited extent, so that wemust of necessity rely mainly on ( b ) . Osmotic-pressure measurements ofrubber solutions have been reported by a series of workers,6 whose resultsshow considerable deviations in detail, but with general agreement in findingmolecular weights of the order of 1-4 x lo6.The calculation of molecularweight from osmotic data is by no means straightforward, since the U S U ~relationship between them is found not to hold. According to van’t Hoff’sJ., 1942, 121 ; Trans. Faraday Soc., 1942, 38, 340.R. Pummerer et al., Ber., 1927, 80, 2167; 1929, 82, 2628; Kautschuk, 1929, 5, 129.H. A. Staudinger, ‘‘ Die Hochpolymeren Organischen Verbindungen ” (JuliusG. Gee, Tram. Farday SOC., 1942, 38, 109.Reviewed by H. Staudinger and K. Fischer, J . p r . Chem., 1941, 157, 19; cf. alsoSpringer, 1932).refs. (19) and (29)GEE : THE PHYSICAL CHEMISTRY OB RUBBER SOLUTIONS. 9law, the osmotic pressure Tz.is related to the molecular weight M of the soluteand its concentration c by the equationII=cRT/M . . . . . . . (1)Experimentally it is found that n[/c is a function of c and the equation hastherefore been used in the formM=RTLim(c/IT) . . . . . (2)c+-oTwo difficulties are raised by this procedure: the experimental one ofperforming the extrapolation satisfactorily, and the need for some explanationof the departure from the simple law before any confidence can be placed inthe results.showed that the osmotic pressures of polymer solutionscould be represented satisfactorily by an equation of the formWo. Ostwald. . . . . I I = a c + b c n . ' (3)where a, by and n are constants.He identified a with RT/M, which isequivalent to assuming equation (2), and considered the first term as the trueosmotic pressure. The second term, bcn, in which n lay between 2 and 3, wasregarded as a " swelling pressure," by analogy with the work of E. Posnjak,'who had shown that if a piece of rubber was confined by a piston in a cylinderwhose base, permeable to solvent, was immersed in solvent, the degree ofswelling of the rubber was determined by the pressure applied to the piston.The pressure P corresponding to a given concentration c of rubber in theswollen mass was represented by an equation of the formP = bcn . . . . . . . . ( 4 )where the constants b and n were characteristic of the liquid, n lying usuallybetween 2 and 3.Ostwald regarded the pressure measured by an osmometeras composite of an osmotic pressure and a swelling pressure. Although itwill be shown later that this explanation must be rejected, equation (3) hasbeen widely used, especially by S. R. Carter and B. R. Record * and byK. H. Meyer and his co-w~rkers.~ The latter take n = 2, so that n/c becomesa linear function of c.An alternative explanation of the osmotic anomaly was proposed by0. Sackur 10 and applied to rubber by K. H. Meyer and H. Mark.11 Thedissolved particles were assumed to be solvated, so that the effective con-centration of the solution was increased. Sackur took the volume of solventremoved in this way per g. of solute, to be independent of the solute con-centration and wrote, instead of equation (1) :II = [c/(l - c$)]RT/M .. . . . . ( 5 )8 Kolloid-Z., 1929, 49, 60. Kolloid-Beih., 1912, 3, 417.J., 1939, 060. @ Helv. Chim. Acta, 1940, 23, 488.l1 Ber., 1928, 61, 1947. 10 2. physikal. Chem., 1910, 70, 47710 GENERAL AND PHTSICAL CEEMISTRY.where 8 was a constantt.representing s as a function of rI by the equationG. V. Schulz l2 has further modified thits equation byrrsv = k . . . . I . . * (6)which is again based on Posnj ak’s swelling-pressure equation (4).bination of (6) with ( 5 ) leads toCom-1M = CR!Z’/II(I - c(k/Il)’/”) . . . . . - (7)This equation has been employed very extensively by Schulz and has beenapplied to rubber by Staudinger and Fi~cher.~ It use in extrapolatingosmotic data requires the evaluation of the constants k and v, which is doneby the authors by a series of approximations.Where measurements areavailable for a series of M values, each II/c-c curve is extrapolated roughly,and the resulting M inserted in equation ( 5 ) to calculate s. The extrapol-ations are then adjusted until all the s values obey an equation of the form(6). A more direct method suggested by G. Gee l3 is to plot II/c againstdIT, which will be nearly linear if v has the expected value of ca. 2. Thedegree of solvation required by this theory is very large (ca. GO C.C. per g. ofrubber in a 1u/, solution of rubber in benzene I*) and cannot be held by any“ chemical ” bond. Even if we could regard this solvent as mechanicallyheld by a, spoiige-like molecule, equation (6) cannot hold down to infinitedilution, for a t this point s --+ cc.Both as a method of extrapolation andas an explanation of the osmotic behaviour of solutions, this treatment mustbe regarded as thoroughly unsatisfactory.An entirely different line of approach was suggested by R. E. Powell,C. R. Clark, and H. Eyring.15 Pointing out that in many of its propertiessolid rubber behaves as though its molecules were composed of a large numberof relatively short ‘‘ segments,” capable of more or less independent action,these authors apply the same idea to rubber solutions. Van’t Hoff’s law isassumed to hold provided we use, in place of the actual number of rubbermolecules present n,, the “ effective ” number n,*. These are related to thevolume fraction v, of rubber in the solution and the molecular weight M ,of the segments of solid rubber by the empirical equation :n,* -- n,[1 + (M/Ms -- l)v,l .. . . . ‘ (8)The Reporter can see no justification for the assumptions involved in writingthis equation. By use of approximations valid for dilute solutions thisleads towhere p7 is the density of the rubber. This is evidently of the same form as(3), and has been employed by the authors t o calculate values of M , from theslope of the IIlc-c curve. Results obtained vary between 900 and 4500,and are considerably larger than the segment sizes deduced in other ways.IT = RTcIM + RTc2/prMs . . . . . (9)l2 2. physikal. Chem., 1936, A , 176, 317.l4 Calculated from the data of Gee and Treloar, ref.(19).l6 J . Chm. PlbglBkB, 1941, 9, 268.l3 Trans. Faraday SOC., 1940, 36, 1162GEE: THE PHYSICAL CHEMISTRY OF RUBBER SOLUTIONS. 11Although this paper is of value in emphasising that the rubber moleculemay not be the effective unit in solution, we shall see later that a much moresatisfactory explanation can be given of the varying slopes of the II/c-ccurves obtained in different solvents. [If (9) were valid these should, ofcourse, all be equal.]Before these theories can be further appraised, it is desirable to examineboth the nature of osmotic pressure and the assumptions involved in thederivation of van’t Hoff’s equation. In considering the nature of the osmoticpressure of rubber solutions, it is necessary to abandon the kinetic picturefrequently advanced of a pressure produced by the solute molecules bom-barding the membrane.E. Guggenheim l6 has pointed out some of’ thedifficulties associated with this view, and- they are further enhanced if wecompare the osmotic pressure of a rubber solution with the swelling pressureof a rubber gel. Thermodynamically, these can both be defined as thehydrostatic pressure which has to be applied to the solution (or gel) in orderto increase the vapour pressure of the solvent in the solution until it becomesequal to that of the pure solvent. Under these conditions the solution (orgel) is in osmotic equilibrium with the solvent, and it is obvious that osmoticpressure and swelling pressure are actually synonymous terms. Both are,in fact, to be understood, not as pressures exerted by the solute, but aspressures applied by the observer. Their physical origin may be regarded asthe tendency of solvent molecules to flow from solvent to solution because ofthe increase of entropy which results.It can be shown that,17 for an idealsolutionwhere AS, = entropy of dilution and Vo = molar volume of liquid.In applying a pressure to the solution we compress it slightly, reducingthe mean intermolecular spacing, and producing (a) a decrease of entropyand ( b ) an increase of the energy content of the liquid. These factors combineto oppose the increase of entropy tending to cause osmosis, and at a certainpressure, osmotic equilibrium results. If a is the coefficient o f thermalexpansion of the liquid at constant pressure it may be shown that thedecrease of entropy is given by aVoP and the increase of energy byV,P(1 -- aT).Since CCN- for organic liquids, the second term isgenerally the more important. This way of considering the problem empha-sises the fallacy involved in comparing a solution with a perfect gas, forwhich a = 1/T, and the effect of pressure is solely to decrease the entropy.The increase of vapour pressure with applied hydrostatic pressure is alsoreadily understood on this basis. It is easy to derive from it the thermo-dynamic relationship between osmotic pressure and the vapour pressuresof the solvent (p,,~) and solution (porn) :rIVV,=TASo . . . . . . . (10)IIV, = R17hpo0/pom. . . . . . (11)l6 “ Modern Thermodynamics,” Methuen, 1933, pp.85 et aeq.I * G. N. Lewis and M. Randdl, “ Thermodynamics,” 1923, p. 133 (McGraw Hill).lo G. Gee and L. R. G . Treloar, Trans. Fnraduy Soc., 1942, 38, 147.E. Guggenheim, op. cit., p. 97; cf. ref. (20).A 12 UENERAL AND PHYSICAL CHEMISTRY.The theoretical basis of van’t Hoff’s law may be given in several W&YB :(1) The kinetic interpretation of osmotic presaure leads to it by analogy withgas pressure ; rejection of this interpretation necessarily invalidates this lineof argument. (2) If we assume Raoult’s law to hold, we may replace pom/pOoin equation (11) by No, the mo1.-fraction of solvent in the solution ; for dilutesolutiom, the expansion In No 21 1 - No leads readily to van’t Hoff’sequation-this method merely exchanges one assumption for another whichis in equal need of justification.(3) If the molecules of solvent and soluteare regarded as approximately equal spheres, it can be shown statistically 2othat the entropy of dilution is given by ASo = - R lnNo, which on sub-stitution into equation (10) gives the same result as above. This gives us amolecular basis for the law, and reveals at once the cause of its failure to applyto rubber solutions, for it is evident that the rubber molecule is not to beregarded as equal in size to the solvent molecule. The interpretation of theosmotic behaviour of rubber solutions thus reduces to the statistical calcul-ation of the entropy of dilution. K. H. Meyer 21 pointed out that this mightbe done approximately by computing the number of ways of arranging a longflexible chain on an array of lattice points.Each point may be occupiedeither by a solvent molecule or by a “ segment ” of the rubber chain,approximately equal in volume to the solvent molecule. The number ofconfigurations possible is, of course, restricted by the fact that successivesegments must occupy adjacent lattice points. Meyer’s suggestion wasworked out independently by P. J. Flory 22 and M. L. Huggins,= the entropyof mixing being calculated from the probability at a given concentration bymeans of Boltzmann’s equation. Flory’s result was :AS0 = - (RIP) [In 210 + (1 - l/x)vr] . . . (12)where x is the number of segments in the rubber molecule; vo and v, are thevolume fractions of solvent and rubber; and p is the number of solventmolecules which are replaceable by one segment.Both Meyer’s original statement and Huggins’s treatment assume p = 1.The latter’s analysis leads to a similar result :where--[hv, +go(1 - l/z)vr] .. (13)and 2’ is very nearly equal to the co-ordination number 2 of the lattice.Direct calculation of the number of configurations is difficult for long-chainmolecules, and an alternative method of approach has been used by A. R.(Camb. Univ. Press).20 R. H. Fowler and E. Guggenheim, “ Statistical Thermodynamics,” 1939, p. 16321 Helv. Chim. Acta, 1940, 28, 1063.la J . Chern. Physics, 1941, 9, 660; 1942, 10, 61.Ibi&., 1941, 9, 440; Ann. N.Y. Acad, Sci., 1942, 43, 1; J . Physical Chem., 1942,4 6 , l ; J .Amer. Chem. SOL, 1942,64,1712GEE : THE PHYSICAL CHEMISTRY OF RUBBER SOLUTIONS. 13Miller.% Improving on earlier analyses,25 he applied the Bethe approxim-ation to evaluate the entropy of mixing of solvent molecules with “ polymers ”which could be regarded as made up of two or three units similar in size tothe solvent. By a consideration of the way in which the results dependedon the number of units it was possible to extrapolate to very long chains.Miller’s final result 26 is -:)}I . . (15)In the concentration range in which osmotic methods are applied, all threeanalyses lead to an equation of the form :k being given by RT/2pg2V0 (Flory and Huggins) or RT(1/2 - l / x ) / p ~ V o(Miller) where pt is the density of the rubber. It is probable that thesecalculations of entropy can be refined, but even in their present state theyprovide a perfectly natural explanation for the departure of rubber solutionsfrom van’t Hoff’s law.It is now clear that this law could not be expectedto hold even in solutions in which the heat of dilution AHo is zero. Whenthe heat of dilution is not zero, a further complication arises, for we havethen l9Few measurements have been made of the heat of dilution of rubber solutions,and none of sufficient accuracy to determine the form of the relatiomhipbetween AHo and c. There is no reason, however, to expect this to differfrom that found for mixtures of two liquids. J. H. Hildebrand 27 has shownthat in the latter case AHo may be expressed as a power series in vr, in whichthe first term involves v,2.Now, for dilute solutions c a v, and we maywriteCombination of (16) and (18) gives for the reduced osmotic pressureThis equation will be seen to afford a justification for the use of equation (2)for the calculation of molecular weights from osmotic data. It is importantto notice that if the expression for AHo (18) had contained a term klc, theextrapolated value of Il/c would have been RT/N - El, which of course doesnot permit the use of equation (2). Evidence in favour of the assumptionk, = 0 is derived ( 1 ) from measured heats of mixing of simple liquids, and (2)from the fact that Lim (II/c) for polymers is independent of the nature of thesolvent, in the few cases for which data are available.a*The initial slope of the I@-c curve is given by (19) as k - k,, where k,24 PTOC.Camb. Phil. SOC., 1942, 38, 109.25 T . S . Chang, PTOC. Roy. SOC., 1939, A, 169, 512.26 PTOC. Camb. Phil. SOC. (in press).28 The best examples are given by (Mme.) A. Dobry’s data, J . Chim. phyeique, 1936,II/c=RT/M+kc . . . . . * (16)IIVo=TASo- AH0 . . . . . . (17)AHo = k2c2 + k3c3 + . . . . . (18)II/c = BT/M + ( k - k& - 4 c 2 . . . . (19)c 3 027 “ Solubility,” Reinhold (1936).32, 60; Kolloid-Z., 1937, 81, 19014 GENERAL AND PHYSICAL CHEMISTRY.depends on the nature of the solvent. The slope therefore depends on thesolvent, and it is evident that if a solvent can be found such that E , = k,it will form solutions which obey van’t HOB’S law.Such solutions are,however, properly regarded as being Eess and not more ‘( ideal ’’ than thosein which E , = 0.We are now in a position to discuss the evaluation of molecular weightsfrom the published osmotic data, and it is evident that the problem isessentially that of finding the best method of extrapolation. The dacultybecomes acute for high-~uolecular rubbers, since different methods of extra-polation, all equally consistent with the data, may lead to molecular weightsdiffering by a factor of 2 or rn01e.I~ Attempts to improve the extrapolationby measurements at higher dilution are of doubtful value on account of theincreasing experimental error. The theoretical considerations given abovesupport the method suggested by Meyer of a linear extrapolation ofTI/c-c. On the other hand, the most careful experimental work seems to bedefinitely opposed to this simple view.Gee and Treloar,le by means ofosmotic- and vapour-pressure data, have covered the whole range of con-centration for the system rubber-benzene, and find a continuous curvatureof the II/c-c plot. A set of points covering a limited concentration rangemight, it is true, be considered collinear within experimental error, but boththe slope and the extrapolated value depend on the particular set of pointschosen. It does not, in fact, appear possible to find with certainty thelimiting value of ( I I / C ) ~ from these measurements alone, a t any rate in thecase of the highest molecular-weight rubbers. A method of avoiding thisdifficulty was suggested by Gee.29 By using a mixture of solvent and non-solvent it is possible to increase AHo until the osmotic behaviour of thesolution approximates to van’t Hoff’s law at one temperature.For inetance,rubber solutions in benzene + 15%) methyl alcohol at 25” give values ofII/c almost independent of c from 8 to 20 g./l. This value of IIfc is consideredto give the best available estimate of RT/M, and is consistent with thegeneral curvature of the II/c-c curves for rubber in benzene. A series ofrubbers of different molecular weights were studied by Gee and Treloar, andtheir II/c-c curves in benzene form a band whose width diminishes withincreasing concentration, until at a concentration of ca. 30 g./1. the effect ofmolecular weight is indetectable.In Fig. 1 the curves for the highest andlowest members of the series are plotted, together with some older dataobtained by earlier workers.ll9 309 81 The curves have been drawn throughthe latter points in such a way as to fit them into the general results of Geeand Treloar. Although the curves are consistent with the data, it is obviousthat they are not derivable from the data alone. A similar family of curvesfor toluene solutions of rubber has been constructed by combining the data ofK. H. Meyer and C. G. Boissonnas 32 with those of Staudinger and Fischer29 Trans. Faraday Soc., 1940, 36, 1171; 1942,38, 108.W. A. Cespsri, J., 1914, 105, 2139.H. Kroepelin and W. Brumshagen, Be?., 1928, 61, 2441.82 Helv. Chim. Acta, 1940, 23, 43016 GENERAL AND PHYSICAL CHEMISTRY.molecular weights obtained from Figs.1 and 2 are compared with thosecalculated by the actual experimenters, whose methods of extrapolation arealso recorded.6I I0 10 20c, g.n.FIG. 2 .Ornotic Preasure of Rubber in Toluene.curve. S ymb 01. Material. Ref.1 X Purified rubber 52 0 Lightly milled crepe 32Sol rubber 5Al,O,-purified rubber 53 El4 AThis review of the situation suggests that most of the published osmoticmolecular weights of rubber may be subject to considerable uncertainty.The Reporter’s opinion is that the above method of interpreting the dataleads to the most probable molecular weights from existing data, and that,for future measurements, the best method is to find a solvent in which II/cis independent of c over the working range.For rubber such a solvent isbenzene + 15% of methyl alcohol. It should be pointed out that thisopinion is contrary to the views of other workers, of whom probably a majorityfavour linear extrapolation of n/c-c.33If the lI/c-c plot is truly non-linear, it appears to suggest that thecalculated entropy of dilution is not quite accurate. Before we can be sureof thb, it is necessary to have accurate data for the heat of dilution, andthese are lacking. Several attempts have been made to estimate AH,, fromthe temperature coefficient of osmotic pressure, but it has not hitherto beenpossible to obtain accurate data in this way. Meyer and Boiseonnas32 givedata for toluene solutions which are consistent with the expression AE,, =2 0 ~ 2 cals./mol., but the probable experimental error is & 50% and the form(Infersoienoe Publishers, 1940).** Mark accepts this as a fact in his book “ Physical Chemistry of High Polymers QEE : THE PHYSICAL CHEMISTRY OF RUBBER SOLUTIONS.17of the equation is obviously indeterminate. Gee and Treloar 19 give forbenzene solutions AHo = 55v: with a probable error of & 30%. Earlierdata of Kroepelin 34 lead to much larger heats, but are of extremely doubtfulaccuracy. All these workers agree in finding positive values of AHo, and thisis confirmed by direct calorimetric measurements of the integral heat ofswelling.35 The precise value found in this way depends somewhat on thehistory of the rubber sample, and is difficult to measure with precision, butvalues of the order of 2 cals./g.of rubber were found in the case of benzene,with a smaller value (ca. 0.5 cal./g. of rubber) for toluene. Chlorinatedsolvents gave an evolution of heat (CHCl,, 3 cals./g. ; CCl,, 2 cals./g.). Theonly data covering a wide concentration range are those of Gee and Treloar l9for benzene, calculated from vapour-pressure data. The accuracy is limitedboth by the smallness of the effect and by the relatively large correction forthe non-ideal behaviour of benzene vap0ur.3~ The data are not satis-factorily represented by the simple quadratic equation, and a modified formwas suggested, viz., AHo = 156v,2/(1*7 - 0 . 7 ~ ~ ) ~ . An alternative, based on(18), would be AHo = 55v: + 1OOv,3.In Fig. 3 the entropy of dilution forthe rubber-benzene system is compared with the theoretical results of Floryand Miller, calculated from equations (12) and (15). In each case onearbitrary constant has to be evaluated, and values of p = 1,Z = 6 have beenselected in the following way :For sufficiently small values of w,, equation (12) gives ASo+- (R/P) In wo. Now, in this region, the vapour pressure is given l9 bypOm/pOO = avo, where a = 4.4 at 25". This result requires that the limitingvalue of AS, shall be - R In v,, whence p = 1. A better fit over the mainpart of the curve results from taking p = 1.3, but a complete fit wouldrequire p to be made an empirical function of v,. Miller's equation (15) givesASo + - R In vo for all values of 2, and the value 2 = 6 has thereforebeen chosen to fit the main part of the curve.It is evident that either equation represents the main features of thecurve extremely well, and confirms the essential correctness of the method ofcalculation.Since no specific assumptions have been made regarding thechemical nature of either the solvent or the polymer, the same entropy ofdilution should be found for any other solvent-polymer system, providedonly that the polymer molecule is sufficiently flexible to justify the assumptionof very small segments. Experimental data by which this might be checkedare lacking, and the best we can do is to compare values of - AGo/T obtainedfrom vapour-pressure data,* with the theoretical curve for ASo. Three suchcomparisons are included in Fig.3 and conform very closely to the expectedbehaviour. If AH, > 0, we have -AG0/T < A&, and this is seen to be thecase for toluene and acetone. The toluene curve is almost identical with the34 Kolloid-Z., 1929, 47, 294.35 L. Hock and H. Schmidt, Rubber Chem. and Techn., 1934, 7 , 462; S. Bostrijm,36 G. Gee, Xrans. Paraday Soc., 1942, 38, 418.* Aff, is the Gibbs free energy of dilution, equal to AH, - TAS,,.Kolloid-Beih., 1928, 26, 43918 GENERAL AND PHYSICAL CHEMISTRYcorresponding curve for benzene, and the larger displacement of the acefoiiecurve agrees with the much greater heat of dilution to be anticipated. It isnoteworthy that the chloroform curve lies above the theoretical entropycurve : this is to be expected, since Hock and Schmidt 35 have shown thatAH,, < 0.FIG.3.Entropy of Dilution of Rubber Solutions.Curve. Source. Ref.1 - AQ,/T for CHC1, 372 Theoretical (Flory) 224 Experimental : benzene 19tj - AG,/T for acetone 37On the assumption of an entropy of dilution given by one of the theoreti-cal equations, it is possible to make considerable progress in understandingother properties of rubber solutions and gels. Before this work can bedescribed, it is necessary to consider another method of investigating the sizeand shape of rubber molecules, vk., the measurement of solution viscosities.s7 J. Lens, Rec. Tmv. chim., 1932, 51, 971.3 Theoretical (Miller) 265 - A.G,/T for toluene 3GEE : THE PHYSICAL CHEMISTRY OF RUBBER SOLUTIONS.19The Viscosity of Rubber Soldom.It is not proposed to discuss a t length the general problem of theviscosities of high-polymer solutions, since it number of recent articles havebeen devoted to this t0pic.3~ We shall be concerned with the problem odyin so far as it beam on the determination of the size and shape of rubbermolecules.Two general methods have been adopted in discussing the viscosity ofpolymer solutions. The first treats the suspended particle (molecule) as arigid ellipsoid, whose axial ratio is calculated from the experimental data.33uThis model is 80 far removed from our concept of the rubber molecule insolution that it does not seem profitable to discuss it futther in this Report.The other method attempts to relate the viscosity of a dilute solution with themolecular weight of the solute.The possibihy of doing $6 depends on theassumption of Staudinger’s law, or some modification of it. As origindlyproposed,% it took the formwhere qsp. is the specific viscosity of a solution of concentration c, and E, is aconstant. Sinoe it is found that GJC is always dependent on c, it is customaryto employ the value extrapolated to infinite dilution. Thisa quantity wastermed by W. D. Lansing and E. D, Kiaemer 4O the ‘‘ intrinsic visoosity ”[q], and by H. Staudinger, H. Berger, and K, Fkchertl the “visoositynumber ” ZT6 It has also been clearly rewgnised that (llc) In q+ ia muchless dependent on c than is -qBP./c,& 80 the most convenient form of equation isrsp.e 2 kmcM . . . . . . . (20)where -q7 = 1 + qsp. is the relative viscosity of the solution. Equation (21)might be justifiable in one o€ two ways : (a) by an acceptable theoreticalderivation, or (b) by comparing intrinsic viscosities with molecdar weightsmeasured in other ways.A very clear discussion of the factors involved in the shear of a liquidcontaining non-spherical particles is given by frawrence‘98b Rigid elongatedparticles increaae the viscosity of a solution when the two ends of a partideare aituated in layers of the liquid which, as a result of sheer, are moving atdiffepent rates. There is a tendency to stretch the particle and to rotate itinto the streamlinera. Orientation is, however, upset by Brownian motion,and a certain mean orientation results from the interplay of these two €orces.At high rates of shear, orientation becomes increasingly perfect, whereas arise of temperature favours disorientation. Furthermore, it is evident thatmaximum disturbance of flow will occur when the particle lies directly acros888 (a) J.M. Burgera, ‘‘ Second Report oh viscosity and Plasticity ” (Amsterdam,1938); ( b ) A. S. C . Lawrence, Ann. Reports, 1940, 37, 99; ( c ) F. Eirioh, Reports Progr.Physics, 1940, 7, 329.Staudinger, op cit. 40 J . Amer. Chem. Soc., 1935, 57, 1369.41 J . pr. Chern., 1942, 160, 95.4 2 A. R. Kemp and H. Peters, Ind. Eng. Chem., 1941, 33, 126320 GENERAL AND PHYSICAL CHEMISTRY.the streamlines, flow being quite undisturbed by thin particles if perfectlyoriented.These considerations provide a ready explanation of J. R.Robinson’s observations 43 on the viscosity of solutions of tobacco mosaicvirus. The viscosity falls rapidly as the rate of shear is increased, and in atypical capillary viscometer of the Ostwald type it is much smaller than in aCouette viscometer, in which the rate of shear can be made very small.The application of these ideas to rubber solutions requires us to know theform of the rubber molecule in solution. Tobacco mosaic virus is known toconsist of rigid rods, but Staudinger’s original suggestion 39 that all polymermolecules were similarly rigid has long been rejected. The chemical natureof rubber leads inevitably to the conclusion that the molecule must possess ahigh degree of freedom of rotation about the single bonds in its structure.As a result of this freedom, a fully extended configuration is extremelyimprobable, and in the absence ofany restrainingforce the molecule will assumea series of constantly changing, randomly kinked forms, in which the averagedistance between its ends is far less than the maximum possible.Astatistical analysis of the most probable molecular configuration has beenattempted by a number of workers 44 and an account of their work is given ina recent report.45 It will suffice here to point out that all treatments, inaddition to other approximations, involve the assumption that all config-urations formed by rotations. about a single bond are of equal energy.M. L. Huggins 46 has attempted to calculate the Viscosity of a solution oflong-chain molecules along the following lines.The molecule is treated as achain of “ submolecules,” whose distribution is given by equations analogousto those of Kuhn.PP Each submolecule is, in general, moving with a velocitydifferent from that of the layer of liquid with which it is in contact. Thisrelative velocity is calculated, and the work done on the liquid by thesubmolecule estimated by means of Stokes’s law. A summation over all thesubmolecules gives the total work done, and hence the contribution of themolecule to the viscosity of the solution.The result of this calculation is to give a relationship between viscosityand molecular weight which, for long molecules, reduces to the form ofequation (Zl), a definite value being given to the constant K in terms ofatomic parameters.This theoretical treatment is undoubtedly a greatadvance on any other attempt to solve the problem, and goes far to suggestthat Staudinger’s law should be approximately valid for rubber solutions.An evident source of error is the assumption that the molecular shapedistribution will be unaffected by suspending the molecule in a liquid under-going shear. Huggins has suggested that the correction is likely to be small,but a fuller analysis is desirable. Experimentally the problem could belargely solved by studying the viscosity of rubber solutions over a widerange of rates of shear. A more serious objection to this treatment is43 Proc. Roy. Soc., 1939, A, 170, 519.44 R.Kuhn, Kolloid-Z., 1934, 68, 2 ; E. Guth and H. Mark, Monatsh., 1934, 66,93.4 6 L. R. G. Treloar, Reports Progr. Physics, 1942, 9 (in press).46 J . Physical Chern., 1938,42,911; 1939,48,439; J. Appl. Physica, 1939,10,700GEE : THE PHYSICAL CHEMISTRY OF RUBBER SOLUTIONS. 21inherent in Kuhn’s distribution function. Since the molecule is idealised toa series of points, the effect of molecular size is neglected, and the calculateddistribution must be much more compact than the real one. It follows thatHuggins’s calculated viscosity must be too small, but provided that only thevalue of the constant K be thereby affected, the main conclusion of Huggins’sanalysis will still be valid.Theory thus gives us reason to think that equation (21) should beapproximately valid, and most workers would agree that the experimentalevidence leads t o the same conclusion.The use of viscosities for molecular-weight determination is, however, only possible if equation (21)--or someother-holds accurately. This can at present only be tested experimentally,and as we shall see, there is no consensus of opinion about the results. Beforegiving the experimental results, a number of other theoretical points must bementioned.In a rubber solution there exist molecules covering a range of molecularlengths, and the “ molecular weight ” of the rubber is therefore to be inter-preted as a mean value. If we have ni molecules of molecular weight Mi,the mean molecular weight M,,,., as ordinarily understood, is given byi iThis equation is valid for molecular weights determined by the osmoticmethod, but it can be readily shown 47 that, if Staudinger’s law holds for eachcomponent of a mixture, the mean viscosity molecular weight M , of themixture is given byM , = CniMi2/ZniMiiIt will be evident that M , > M,,m.and that the ratio Mv/Mosm. is a measureof the degree of homogeneity of the rubber. There are two ways of avoidingthis difficulty in order to test equation (21). One is to compare viscositieswith ultracentrifuge data, which can be employed to calculate M , the otheris to fractionate rubber so as to obtain materials as homogeneous as possible,and then to compare the viscosity with the osmotic molecular weight.A second difficulty arises from the possible existence of branches in themolecular chain of rubber.Although there is no evidence of this from othersources, it is difficult to point to any fact which conclusively disproves thepossibility. Now, it is clear that in general a molecule possessing a branchedstructure will take up a more compact form than a truly linear molecule, andwill therefore contribute less to the viscosity. It has therefore been suggested(see below) that viscosity measures the length of the main chain, and will onlybe comparable with the other molecular-weight methods in the case of linearmolecules. This argument is not very convincing, for although it may betrue that a short side chain contributes little to the intrinsic viscosity, yetthe effect of the long side chains, implied by the concept of branching,cannot be negligible.Qualitatively, however, it is clear that branchingshould reduce [q], and therefore invalidate the method.4 7 Lansing and Krwmer, ref. (40)22 GENERAL AND PHYSICAL CHEMISTRY,T4e last difficulty we shall discuss is that of the effect of solvent. Thetheory a8 formulated predicts that [ I ] should be the same in all solvents,which is well known to be untrue. A satisfactory theory must give someexplanation of this anomaly, and also suggest the best solvent for viscositymeasurements. It seems probable that the different values of [q] found invarious solvents are to be associated with different forms taken up by themolecule in solution.48 This would not be diflicult to understand, for it isonly in a solvent in which rubber forms ideal solutions (AHo = 0 ) that thevarious possible configurations of the molecule can have equal energy.Nowthis is a primary assumption in calculating the entropy of solution as describedearlier. If AH,, > 0, the total energy of the system (rubber and liquid) willbe reduced by the molecules taking up more compact configurations, so as toreduce the interface rubberlliquid. Despite the lower entropy of such astate, the net effect will be a reduction of free energy. It is clear, therefore,that the larger AHo becomes, the more compact will be thc mean molecularconfiguration, and the lower the solution viscosity. If, on the other hand,AHo < 0, extended configurations become favoured, and solutions of highviscosify are to be expected.Although, as already noted, few values ofAHo are available, we shall see later how they may be estimated, and the dataare in general agreement with the prediction that [q] should fall with AH,.Since the theory is worked out for AHo = 0, we have also R, criterion forselecting the best solvent, though in a later section it will be shown that otherconditions may in practice need to be considered.The first experimental data to be considered are those in which intrinsicviscosity has been compared with osmotic molecular weight. The evaluationof this evidence depends to a considerable extent on the method employed inextrapolating the osmotic data, and it is clear that the same method must beapplied to all the data.The figures given in Table I1 are based on themethod described above, the actual extrapolations for toluene being shown inFig. 2. Values in parentheses are those given by the originaJ authors, andthe effect of different methods of extrapolation is only too evident.TABLE 11.Comparison of Intrinsic Viscosity with Osmotic Holecular Weight.Material. Viscositysolvent. Ref.Intermediate latex fraction ...... Benzene 29Low latex fraction .................. 99 29Hydrocarbon from oxide fraction ,, 29p p e (acetone extracted) ......... Benzene 29Sol ” rubber ........................ 9 , 29 ........................ Toluene 5Lightly milled crepe ............... ,, 49( 1 ) Fractionated rubbers :(2) Unfrwtionated rubbers :Pu&ied dch Also8 ...............9 , 53.5 5.82.1 4.00-66 1.022.6 5.7,3.0 5-7,1.8, (2.4,) 3.83-6 (3.5) 8.41-6 (2.7) 3.74.65.24.9 (6.4)4.3 (4.2)4-3 (7.3)4 8 T. Alfrey, A. Bartovics, and H. Mark, J . Amer. Chein. Xoc., 1942, 64, 1667; cf.4g K. H. Meyer, Helv. Claim. Acta, 1941,24,217.ref. (22)GEE : THE PHYSICAL CHEMISTRY OF RUBBER SOLUTIONS. 23On the basis of the data for the fractions, Gee suggested that Sbaudinger’slam holds for rubber within the accuracy (ca. 10%) with which it can a tpresent be tested, at any rate over the molecular weight range 60,000-350,000. Subsequent work (see below) has led to the conclusion that thesefractions were much less homogeneous than was thought at the time, so thatme might expect the ratio M/[q] = 60, x lo4 to be somewhat too low.On:he other hand, the data available have now been congiderably extended, andLLonfirm the original conclusion that, for rubbers of high molecdlar weight,.q] cc M , with a constant of proportionality close to that previously found.The ratio M/[-q] for unfractionated rubbers is seen from Table I1 to be sorde-what lower than for the fractions, in agreement with their wider range ofrnolecular weights. (It will be noted that if the published osmotic maleoularweights are employed this last statement is untrue.)The other possible way of testing the Staudinger law in the high-molecular region is by comparison of intrinsic visoosity with ultracentfiftigbmolecular weights, which may be done for any rubber, whether homogeneousor not.No systematic work has been done on these lines, but Gaeiner 8ohas reported the results given in Table 111.TABLE 111.Comparison of Intrimic Viscosity with Ultracentrifuge MolecularWeight.solvent. (in ether). 171. 104M/[q].Sol A .................. Ether 4-00 1-85 21.5Benzene - 2.60 15.5Sol B .................. Ether 4.35 2-05 21Benzene - 3.74 11.5Low viscosity rubbers Ether 0.69 0.34, 20CHCl, 0.635 0.60, 12.5Material. Viscosity 10-=MIt is difficult to discuss these figures, since they were presented withoutexperimental details beyond a statement that the sedimentation equilibriummethod was employed. Considering the viscosity in ether, we find supportfor the Staudinger equation. The intrinsic viscosities measured in benzeneare not accurately proportional to M , and the ratios of M/[yj] are approx-imately twice as high as those found from osmotic data.This discrepwnoyrequires further investigation, but it is r‘elevant to point out that thedficulty of reconciling osmotic and ultracentrifuge data is not confhed torubber. R. Signer and H. Gross 51 have used the sedimentation equilibriummethod to estimate the molecular weights of some polystyrene fractions inchloroform, obtaining values which are proportional to [TI. They point outthat the calculation requires the assumption of van’t Hoff’s law, and give theapparent molecuhr weight of one fraction ( M = 3 x lo5) as a function ofLoncentration. The deviation from van’t Hoff’s law revealed by thesefigures may be expressed in terms of the calculated osmotic pressures.6” “ The Ultracentrifuge ” (Oxford, 1940).pa 423.L1 Hekv. Chim. Acta, 1934, 17, 33524 GENERAL AND PHYSICAL CHEMISTRY.lOQ/c is found to rise from 0.73 at infinite dilution to 2.9 a t a concentration of2.5 g./l. Schulz l2 gives osmotic data for a polystyrene fraction in toluene,which show 104JI/c increasing, over the same concentration range, from 0.73to 0.87. Although it is true that the slope of the rI/c-c curve is probablyhigher for chloroform than for toluene (AHo is probably negative forchloroform), yet there seems a definite discrepancy here. Almost the wholeof the work on the ultracentrifuge has been carried out with aqueoussolutions, and it is perhaps better to suspend judgment on the above resultsuntil more extensive experience has been gained of the technical difficultiesencountered with non-aqueous solutions.No other method is available for testing the applicability of Staudinger'sequation to high-molecular rubbers, but if we assume its validity; the constantmay be determined by reference to lower members of the series.Bycomparing intrinsic viscosities with molecular weights determined cryo-scopically, we can test the equation for low-molecular rubbers and, if it provesvalid in this region, employ the constant thus obtained for rubbers of muchhigher molecular weight, assuming the possibility of the enormous extrapol-ation involved. This is the method first employed by H. Staudinger andH.I?. B ~ n d y , ~ ~ and subsequently adopted and refined by A. R. Kemp andH. Peter~.~3 The former workers used crude degradation products obtainedby heating tetralin or xylene solutions of rubber in air; the latter startedfrom similar materials, but subsequently fractionated them, obtaining in thisway a range of products of low molecular weights, but containing variousamounts of oxygen (up to nearly 6%) and with rather low iodine values,Some of these materials became insoluble on heating in a vacuum to loo",probably owing to a form of cross linking or oxygen vulcanisation. Since alltheir products were treated in the same way, there is a considerable assump-tion involved in regarding them simply as low-molecular homologues ofrubber. Intrinsic viscosities in benzene were compared with cryoscopicmolecular weights determined in benzene or cyclohexane.The calculatedmolecular weight was substantially independent of concentration forM < 2000, but for the higher fractions deviations occurred which are, ofcourse, parallel to those found in the osmotic data. The authors reject allsuch results aa " unreliable," but in view of our conclusion that extrapolationto infinite dilution leads to the correct molecular weight, there is evidently nobasis for this selection. Table IV summarises their results, molecular weightsmarked * having been obtained by extrapolation. Data are included forsqualene, a hexamer of isoprene, of known constitution.Considering only the lower fractions to be reliable, Kemp and Peters takeM/[7jj = 2.2 x lo4, which they then use in equation (21) to calculate muchhigher molecular weights.The disagreement of the results with the osmoticdata is ascribed to the errors in the latter. From the point of view set out inthis Report, all the figures of Table IV are equally trustworthy. Since theyreveal an obvious trend of M/[q] with M , they must, in the Reporter'sopinion, lead to the conclusion that Staudinger's law does not hold accurately6a Ber., 1930, 63, 734. 63 Ind. Eng. Chem., 1941, 33, 1263, 1391GEE : THE PHYSICAL CHEMISTRY OF RUBBER SOLUTIONS. 25TABLE IV.Comparison of Intrinsic Viscosity in Benzene with Cryoscopic MolecularWeight,Material. 10-6M. EqI. 10-4MI[~ESqualene 64 ..............................0.00410 0.023, 1.8Fraction 3 ................................. 0.012 0.058 2.1Acetone-soluble, washed with alcohol 0.014, 0.065 2.2Fraction 4 ................................. 0-064 * 0.198 3.2Fraction 5 ................................. 0.090 * 0.262 3.4,for low-molecular materials. This behaviour is not confined to rubber, buthas been found in all series of materials whose viscosities have been studieddown to the low-molecular-weight region.65 Several types of explanationhave been proposed :( a ) Kemp and Peters reject cryoscopic and osmotic data for solutionswhich do not obey van’t Hoff’s law, and assume Staudinger’s law to hold.( b ) Staudinger and Fischer accept the osmotic data as correct, but alsoassume the validity of the viscosity law and derive a constant in the sameway as Kemp and Peters.Where the latter leads to molecular weights lowerthan those measured osmotically, the material is assumed to be branched.Arguing that viscosity is a measure of molecfilar length, these authorsdefine the degree of branching as N,,sm./Mv. Unless evidence of branching isforthcoming from other sources, it seems preferable to seek another explan-ation of the data.(c) Most workers now accept the evidence at its face value, and thusconclude that Staudinger’s law is not accurately obeyed, even by homogeneous,linear polymers. Two modifications of the equation have been suggested.Except for very low My the data forethe polyesters 55 obey an equation of theform : . . . . . M = K [ q ] + K ’ (22)A more radical modification has recently been proposed by R.Houwink : 56M=-K[vY . . . . . . . (23)P. J. Flory 57 has reported that this equation holds for polybutenes betweenA2 = 5 x lo3 and 5 x lo5, giving n = 1-65. We may seek to apply theseequations to rubber by combining the data of Tables I1 and IV. A fairlygood fit is given by equation (23) if we take K = 4.2 x lo4, n = 1.23, butthe lowest of the osmotic points is badly out. Equation (22) is very littlebetter than (21).Summarising, it is evident that the precise relationship between molecularweight and the intrinsic viscosity of rubber solutions over a wide range of Mremains to be determined. The principal experimental difEculty to be over-come is that of obtaining rubbers of medium and low molecular weight which64 Staudinger and H.P. Mojen, Kuutschuk, 1936, 12, 121.55 Polyesters : P. J . Flory and P. B. Stickney, J . Amer. Ghem. SOC., 1940, 6Q, 3032;W. 0. Baker, C. S. Fuller, and J. H. Heiss, ibid., 1941, 63, 2142. PoZyoxyethyZeneglycols : R. Fordyce and H. Hibbert, ibid., 1939, 61, 1912.66 J . pr. Chem., 1940, 167, 16. 6 7 Indiarubber World, 1942, 106, 68426 Q E X l W AND PHYSICAL CHEMISTRY.we can confidently assume to be linear and reasonably homogeneous. Untilthis can be done, the extension of the Staudinger equation beyond themolecular weight range 60,000--350,000 is dangerous. Within that range, theReporter’s opinion is that equation (21) may be applied to intrinsic viscositiesmeasured in benzene solution, using K = 6.0 x lo4, and that the molecularweights thus obtained are probably accurate to within 10 or 20%.We now discuss the use of solvents other than benzene for the measure-ment of intrinsic Viscosities. Gee49 has presented data which show that,within the molecular weight range 60,000--350,000, the intrinsic viscosities ina number of solvents are accurately proportional ; i.e., if the Staudinger lawi s valid for benzene, it is also valid for any of the other solvents examined,provided that the appropriate constant be used for each.A. R. Kemp andH. Peters 58 have recently shown that this is by no means true over a widerange of molecular weights, in the case of polyisobutylene. For instance,wit4 A2 = 1000, the intrinsic viscosities in cyclohexane, carbon tetrachloride,hexane, chloroform, and benzene are in the ratios 127, 118, BOO, 100, 85.With M = lo5 these values beoome 140, 120, 100, 100, 38, the intermediatevariation being newly linear with log M .With N < 1000 very large changesoccur in the ratios, and it is evident that if Staudinger’s law holds in thisregion for hexane solutions (taken as standard by Kemp and Peters) it cannotalso hold for the other solvents.No explanation has been offered of the curious behaviour of low-molecularrubbers in different solvents, though Burgers38 has poiated out that theviscosities of solutions of some of the parafis studied by K. H. Meyer andA. J. A. vtin der Wyk 59 are Zower than those given by A. Einstein’s equationfor spherical molecules.The intrinsic viscosities of high-molecular rubbersare readily interpreted on the basis of the theory already discussed. Althoughthe heats of mixing of rubber with various solvents are unknown, yet thereis reason to believe (see below) tbat the degrees of swelling, &, of a vulcanisedrubber in a series of solvents run closely parallel with their heats of dilution.We should therefore expect the intrinsic viscosities of a sample of rubber tobe a simple functim of the degrees of swelling. In Table V some swellingdata given by G. S. Whitby, A. B. A. Evans, and D. S. Pasternach 60 arecompared with intrinsic Viscosities found by Hemp and Peters 53 and Gee.29The parallelism is evidently not exact, particularly in the case of the morepolar liquids, but is sufficiently good to lend general support to the theory.It was suggested earlier that the ideal viscosity solvent would be one forwhich AH, = 0.Probably cyclohexane falls nearest to this condition, butis well known to disperse rubber with some dif6,culty (see below). Being asomewhat viscous liquid, it also leads to highly viscous solutions, which areevidently undesirable. Benzene lies next to cyclohexane in the table, and is avery satisfactory rubber solvent. It is therefore the Reporter’s first choiceas the standard liquid for the determination of intrinsic viscosities of rubbersolutions.68 Ind. Eng. Chern., 1942, 84, 1192.6o Tran8. -8?3r&y sot., 1942, 88, 269.HeZu. Chirn. Acta, 1936, 18, 1067GEE : THE PHYSICAL CHEMISTRY OF EUBBER SOLUTIONS.27TABbE v. ’Intrinsic Viswsites und Swelling Power.Liquid.Carbon tetrachlordeChloroform .........Toluene ...............Benzene ...............CycloHexwne -f ......Amy1 acetate .........Ethyl ether ............Hexane ...............Q liquid per Relative intrinsic viscosities.C.C. rubbe?). Kemp and Peters. Gee. - 6.0 * 1124.2 * 1074.2 103--(100)1041.8 - __ 581.6 I 641.2 883.9 t ’;;’ I* Unpublished data, quoted by kind permission of Prof‘. Whitby.1’ Q not measured, but almost certainly occupying this position in the table (byL omparison with swelling in other liquids).Xolubility and Fractionation.Two main lilies of attack on this problem may be distinguished : ( a ) ant-sperimental study of the behaviour of rubber towards various solvents, andt ) recent attempts to explain the main features of these results by atheoretical approach.It will be convenient in this Report to reverse thehistorical order, so as to be able to use the results of the calculations indiscussing the significance of the experimental work.J . N. Brmeted 61 was one of the first to suggest st plausible explanation(.If the unusual solubility relationships of high polymers. He argued %hat thedistribution of large polymers between two phases would be determinedJmost entirely by the difference of potential energy between them. Assumingthis to be proportional to the molecular weight of the polymer, he obtainedfor the relative concentrations c’ and c” of polymer in two phases inequilibriumn-here A is a constant, characteristic of the solvent and the series of polymersconsidered.If N is large it is readily seen that the ratio c’/c’’ may assumeextreme values. Considering A as continuously variable along a series ofliquids, it is clear that, as x decreases through zero and becomes negative, thedistribution undergoes a sudden change of character. With h > 0,‘ ’ c” --+ 0, andthe liquidis practicallya non-solvent ; with h < 0, c’/c’’ --+ myand the liquid is a perfect solvent. This prediction accords withRrmsted and Volquartz’s work on the solubility of high-molecular poly-ytyrenes 62 which shows a very sharp separation of liquids into non-solventsLind perfect solvents. Schulz 68 extended this concept to mixed liquids, by.iwming in this case that A is a linear function of the liquid composition.By considering a mixture of solvent and non-solvent, the proportion y~ c ’ / c ” = AM/RT .. , . + . (24)d l 2. physikal. Chern., 1931, Bodenstein Festband, p. 257; Compt. rend. Lab. Curls-62 Brransted and K. Volquhrtz, Trans. Paraday Soc., 1940, 38, 619.- r g , SBr. chim., 1938, 22, 99.2. phy&kal. Chem., 1937, A , 179, 32128 GENERAL AND PHYSICAL CHEMISTRY,of the latter in the critical mixture which is just a solvent was shown to begiven by an equation of the formy = A + B / M . . . . . ' (25)where A and B are constants. This equation has been shown to representsatisfactorily the data for a number of synthetic polymers.64Notwithstanding the apparent success of these theories, it has beenpointed out by Gee and Treloar l9 that their fundamental assumption isunsound.These authors showed experimentally that the molar entropy ofsolution of rubber in benzene is large compared with the heat of solution,and that it is proportional to the molecular weight of the rubber. Equations(24) and (25) therefore require another explanation; a possible one issuggested below.The equilibrium distribution of rubber between two phases has also beenexamined by 1310ry,22 H ~ g g i n s , ~ ~ and Gee.g5 The conditions for equilibriumbetween two phases ' and " may be put in the form "' == AG''} . . . . . . . (26) AGr' = A':'where the subscripts and refer respectively to liquid and to rubber.* Thefree energies are then expressed in terms of the heats and entropies ofmixing.The latter is taken to be that given by the calculations describedin the first section of this report, but a simple-power law is usually assumedfor the heat. Thus, Flory writes, for each phase :AGO = AH0 - TASO= kv,2 + RT{ln vo + v, (1 - Mo/M)}The assumption involved in writing down these equations is that the rubbermay be treated simply as a liquid, a point of view which finds justificationboth in its mechanical properties (which demonstrate the great freedom of themolecules to move relative to one another) and in the agreement with thistheory found experimentally by Gee and Treloar. If equation (27) is used togive the free energies in terms of composition, equations (26) give twosimultaneous equations from which v,1 and D;', the concentrations of rubberin the two phases, may be calculated. No simple algebraic solution ispossible, but Flory has given a graphical solution for several values of E.His conclusion is that, for any value of E, there exists a critical temperaturebove which rubber is miscible with the liquid in all proportions.Thecomposition of the critical mixture is given byAGT = EMvo2/M0 + RT(1n v, - vo (M/Mo - 1)) . . (27)(vr)crit, = 1/(2/M/Mo + 1) * - (28)64 R. A. Blease and R. F. Tuckett, Trms. Faraday SOL, 1941, 37, 671 ; cf. also refs.6 5 Ibid., 1942, 38, 276.* AGr' is the increase of Gibbs's free energy when 1 mole of rubber ia added to a(55), (63)-large bulk of phase 'GEE : THE PHYSICAL CHEMISTRY OF RUBBER SOLUTIONS.29Since, for rubber, M/Mo= 4 x lo4, (V,),it.E0*015. At a lower temperature,a two-phase system results, in which the dilute phase has a concentration< (v,),~~,. Indeed, except within a few degrees of the critical temperature,the dilute phase would be indistinguishable experimentally from pure solvent.The concentrated phase consists of a highly swollen gel, in which the cal-culated rubber content at a temperature 10" below the critical is only 10%.A similar analysis was carried through by Gee,65 using a more complexexpression for the heat of mixing, wix.,The essential features of the solution are the same as in that of Flory, butit is possible to have a two-phase system in which the two phases are con-siderably more concentrated than is permitted by Flory's analysis.It is of interest to try to find the relationship between this method ofanalysis and Bronsted's equation (24), since the latter has in practice provedso useful.If this equation were strictly true, it should be derivable from thesolution of equations (26) and (27). Although this cannot be done rigorously,a similar equation may be derived by the following approximate method.Combining the second equilibrium condition of equation (26) with (27) weobtainThe phase ' represents a dilute solution, for which we may write wo' N 1. Theresults of the complete analysis described above show that the compositionof the second phase is almost independent of M . This is confirmed byexperiment,63 so that wo" may be treated as a constant.Equation (30) thustakes the formwhere a = (1 - 2ro")/Mo ; b = k (1 - (vo")2)/Ho. The functions a and bwill not be strictly independent of T, and it is evident that equation (31)differs formally from (24) only in predicting a more complex dependence ontemperature. In fact, it is precisely in regard to temperature variation thatSchulz's equation fail~.~3 This analysis thus provides a reason for theformal success of the Bronsted-Schulz equations, while at the same timegiving a different, and more satisfactory, interpretation of the constants.The analysis described so far has considered the rubber as homogeneous,and we have now to enquire what modifications result when a mixture ofrubbers of different molecular weights is present.In the course of afractionation by any equilibrium method (see below) we obtain two phases,each containing a portion of the rubber. The problem is to calculate themolecular-weight distribution in the two fractions, given that of the totalrubber. Schulz examined this problem by assuming that the distribution ofeach molecular species between the two phases was given by a modificationof equation (24), which may be written :lnw,'/v,'' = M(a - b/RT) . . . . * (31)In ci'/ucit' = - AHi/RT . . . . . . (3230 GENERAL BND PBYSIUAL CHEMISTRY.E was treated as an empirical constant, characteristic of the solvent andpolymer and independent of Mi. By applying this equation to a range ofM i values, Sohula obtained graphically distribution curves for the fractionsresulthg from separating a rubber of given molecular distribution undervarious conditions.An important practical conclusion was that theefficiency of aepar3tion should be improved by keeping the conoentration ofthe dilute phase as low as possible. There is no obvious justification for thearbitrary introduction of the constant a, and it should be further pointed outthat this theory makes no attempt to examine the mutual effect of thedifferent rubbers on their several solubilities. In order to solve the problemcompletely, we require to h o w the entropy of mixing ASm of n, solventmolecules with ni rubber molecules of molecular weight .Mi, where is to begiven all values. Gee 65 suggested for this purpose an empirical extensionof Flory’s equation 22 to give the form :AXm = - k{no In v,, + Zni In vi> .. . . (33)On this basis, it was ahown that the distributions of two molecular speciesand between the two phases were related by the equationExtended to a wide distribution of rubbers, this analysis 67 confirms Schulz’sfindings regarding the effect of concentration on the efficiency of fraction-ation, and indeed leads to the conclusion that a “ fractionation ” in which theconcentration of ths dilute phases exceeded 1 yo would effect scarcely anyseparation at all, even of a complex mixture. The practical importance ofsuch a conclusion requires no emphasis.Before describing the experimental data, we may enquire how far theanalysis would be expected to apply to rubber.The first criticism to bc,made is a general one. We have found, in the viscous behaviour of rubbersolutions, evidence that the entropy of solution is less in a bad solvent( AHO large) than in a good one. It is clear, therefore, that an approximationis involved in assuming the entropy of mixing, whioh is calculated for anideal solvent, to apply to a liquid on the borderline between solvent andnon-solvent. Again, much of the work on solubility has been carried outwith mixtures of solvent and non-solvent. Hitherto, such a mixture hasbeen treated simply as a single liquid, whose properties are intermediatebetween those of its constituents. Actually, this is only valid to a firstapproximation, since in general the two liquids are differently distributedbetween two phases in equilibrium.62 No systematic study of such a systemhas yet been published.A final point is that, throughout the analysis, weare concerned either with a homogeneous material or with a homologousseries of similar polymers. This is doubtless a very good approximation inthe case of many synthetic polymers, but rubber, as normally handled, is far6 6 G. V. Schulz, 2. pliysikal. Chem., 1940, B, 46, 105, 137; 47, 155.13’ G. Gee, unpublislied calculationsGEE : THE PHYSICAL CHEMISTRY OF RUBBER SOLUTIONS. 31from being adequately described in this way. The proteikl present in latex ismost tenaciously retained by the eolid rubber, and mag pr‘ofoundly modify itssolubility behaviour. In many cases, as we shall see, its effect may well be solarge a8 to mask entirely the effect of molecular weight differences.Rubber has long been khown to be capable of separation, by treatmentwith solvent, into poi%ions of very different solubility.It was formerlyconsidered 6* that two sharply defined portiohs termed “ sol ” and “ gel ” werepresent, of which only the former wds soluble. The usual method ofseparating these consists in extraction with light petroleum, or ethyl ether,hut a careful study of the process 69 revealed that the boundary was illdefined, and that the sol portion was greatly diminished if care was taken toexclude oxygen. Kemp and Peters found that only &loo/, of some of therubber samples they examined was extracted by hexane after 3 days.Usingother extra~tants,~O they obtained much larger amounts of “ sol,” but evenchloroforni (the best liquid tried) extracted only 62% of latex film after 3days at 25”, without shaking. Actually, all samples of raw rubber examinedby the Reporter could be completely dispersed in solvent if sufficient time wasallowed, and agitation employed, although oxygen was rigorously excluded.The failure of rubber t o disperse easily and completely in hydrocarbonsolvents is very surprising, and quite inexplicable if we regard rubber merelyas a mixture of long straight-chain hydrocarbons. As we have noted above,cyclohexane should be a perfect solvent for rubber, but in fact even a purifiedrubber hydrocarbon disperses with some difficulty, giving “ ropy ” solutionsa t relatively low concentractions. Benzene is considerably superior in thisrespect, although the rubber is more easily precipitated from it by alcoholthen from cy~lohexane.~~ A still better solvent, from the point of view ofobtaining clear mobile solutions, is a mixture of benzene with several percent.of an alcohol. Such a mixture disperses “gel” rubbers which areinsoluble in benzene alone, while small additions of alcohol greatly increasethe solvent power of he~ane.~O It appears probable that we are dealinghere with two distinct factors. The lack of fluidity in the solvent which isbest thermodynamically (Le., for which AH,, = 0 ) arises from the greaterspace occupied by the average molecule in such a solvent (cf. discussion Ohviscosity, above), which leads to structural viscosity, Addition of alcoholreduces this effect, and a t the same time assists in separating the chains wherethey are held by the polar interactions between protein molecules.It is quite unnecessary to assume, with Kemp and Peters, that sol rubberis necessarily an oxidation product of gel rubber.It is well known thatoxidation promates degradation of the rubber chains to more soluble products,but there is every reason to believe-from the thermodynamic study of Geeand Treloar 19-that a linear, unoxidised rubber of any molecular weightg8 For a review, see G. S . Whitby, Tran8. Inst. Rubber Ind., 1929, 5 , 184.69 W. H. Smith and C. P. Saytor, J . Res. Nat. Bur. Stand., 1934, 13, 453; A. R.70 Idem, Ind.Eng. Chem., 1941, 33, 1391.7 1 H. Staudinger and H. P Mojen, Kautschzck, 1936, l2,121.Kemp and H. Peters, J . PTty&al Chem., 1939,439023, 106332 GENERAL AND PHYSICAL CHEMISTRY.would be completely soluble in benzene. The fact that complete solution ofrubber can be achieved experimentally also rules out the possibility ofinsolubility arising from chemical cross links between the molecules. Thefinal suggestion is that the protein remaining in association with the rubber isconnected with the difliculty of dispersing it. Direct evidence that this isthe true explanation is provided by the well-known fact 69 that the gel rubbercontains practically the whole of the nitrogen. The nature of the associationbetween rubber and protein is unknown, but its profound effect on thesolubility of rubber is evident when we examine the various attempts whichhave been made to fractionate rubber.Two general methods of fractionation have been adopted, based severallyon the greater speed of solution of the shorter molecules, and on their greatersolubility in critical mixtures of solvent and non-solvent.In almost allcases the initial material used contained appreciable percentages of nitrogen,and there is abundant evidence that this is a complicating, if not actually thecontrolling factor. This is shown very clearly in the early work of T.Midgley, A. L. Henne, and M. W. R e n ~ l l . ~ ~ Using a carefully controlledfractional precipitation method, they obtained a series of products whosenitrogen contents increased as their solubility decreased.Although theinitial material was completely soluble, they obtained finally an insolubleresidue containing 4.21% of nitrogen. Similar results were obtained in thevery careful work of G. F. Bloomfield and E. H. Farmer.73 By repeatedextractions with acetone-light petroleum mixtures of increasing petroleumcontent, they concentrated the bulk of the nitrogen in a residue of greatlyreduced solubility. Their earlier fractions were, however, substantiallynitrogen free, and it was shown 74 that the molecular weights of these fractionswere related to the composition of the extractant by an equation of the form(25). Kemp and Peters 69 employed the other method of separation, andobtained by diffusion into hexane a series of fractions of increasing molecularweight.Later ’* they were able to extract small quantities of very lowmolecular-weight material by using hexane-acetone mixtures as extractant.A further complication is revealed when the most soluble fractions of rubberare examined. K. C. Roberts 75 and Bloomfield and Farmer 73 obtainedmaterials of molecular weight 30,00040,000 which contained 1-3% ofoxygen. Roberts called his product “ caoutchol” and showed that itpossessed some extremely interesting physical properties, which have not yetbeen fully explained.It will be evident that the solubility behaviour of rubber is far morecomplex than would be explicable by the idealised picture of it which formsthe basis of the theoretical treatment, and little progress has yet been madein the application of these ideas to rubber.This is due in part to the experi-mental difficulties arising both from the protein content and from the readyoxidisability of rubber, and in part to the use of much too high concentrations72 J . Amer. Chem. Soc., 1931,58,2733; 1932,54,3343.Is Trans. Inst. Rubber Ind., 1940, 10, 69.I4 Gee and Treloar, aid., 1941, 17, 184. 76 J., 1938, 216; 1942, 223BEE : THE PHYSICAL CHEMISTRY OF RUBBER SOLUTIONS. 33in most of the attempted fractionations of rubber. It must be concludedthat nothing approaching a complete fractionation of rubber has yet beenachieved, and the evidence that a considerable proportion of rubber isreasonably homogeneous 729 78 is almost certainly spurious.Limited Swelling.When asoluble, high-molecular rubber is covered with solvent, it rapidly imbibes thesolvent, increasing in volume without (in the early stages) much change ofform.After a time, dispersion slowly takes place and a solution is produced.The phenomenon affords a visual indication of the difficulty of dispersinglong-chain materials : the rate of diffusion of solvent into the rubber greatlyexceeds that of the rubber outwards. If the liquid employed is a non-solvent,or the rubber is vulcanised, dissolution does not occur, but the rubber swellsto a more or less well-defined equilibrium size. In measurements carried outin air, a definite limit is seldom observed, an initial rapid swelling beingsucceeded by a relatively slow absorption which continues indefinitely.Thesecond stage, termed the "increment " by J. R. is absent whenoxygen is carefully excluded,77 and should evidently be neglected in cal-culating the true equilibrium swelling.The literature records an enormous number of swelling measurements,but their unsystematic nature renders many of them valueless from ascientific point of view. Whitby, Evans, and Pasternack 60 have provideddata for a large range of liquids on the same rubber, and their work forms auseful basis on which to build or test any theory of swelling. These authorshave drawn from their figures a number of interesting generalisations.Hydrocarbons in general are good swelling agents, the swelling being greatlydiminished by the presence of a polar group, especially OH or CN.Theeffect of a given group is greatly diminished by a long hydrocarbon chain,and is less in aromatic compounds than in aliphatic. Addition of chlorineto the molecule usually enhances swelling.A completely satisfactory theory of swelling should account quantitativelyfor these measurements : we are at present very far from such a state. Thefirst attempt at a quantitative theory was made by Wo. Ostwald,78 whoproposed a relationship between the swelling Q (c.c. of liquid per C.C. ofrubber) and the dielectric constant E of the liquid, 'uix.,Accumulation of more data soon revealed the inadequacy of this expression,and it was later suggested 79 that swelling was a function of p 2 / ~ and thesurface tension Q, p being the dipole moment.These two quantities havebeen combined by M. Takei 8o into the single parameter p2/ac. There seems,however, little justification, either theoretical or experimental, for thesesuggestions.Closely related to the problem of solubility is that of swelling.(35) . . . . . . Q = KE-2*16'6 Trans. I n s t . Rubber Ind., 1929, 5, 95.7 7 Gee, unpublished experiments. '' Oetwdd, J. Oil Cole C ~ W L ASBOC., 1939, 22, 31.'' KoZEoid-Z., 1921, 29, 100.'O KOUOid-Z,, 1942, 98, 31234 GENERAL AND PHYSICAL CHEMISTRY.The tihermodynamic oriterion for equilibrium swelling is given, of course,by AGO = 0, or 4iY0 = TA8,. For raw rubber we may express Atlo in termsof camposition by one of the theoretical equations, or may take the measuredvalues for benzene.The entropy of swelling of vulcanised rubber bas notbeen measured, but will clearly be lower than that of raw, since cross linkingreduces the number of possible configurations of the molecule. This effectwill depend on the degree of vuloanisation, and will probably be importantonly at fairly high degrees of swelling. A semi-empirical method of deducing' ._\.>\. ,2 4 .* .. 600Q, C.C. of /;quid/c.c. o f rubber. 'FIG. 4.Relatzon between Heat and Ezteni? of SweJling of Vzltoaraised Rubber.ASo, for the vuloanisate used by Whitby and his co-workers, was suggestedby If we then write AH, = k0v,2, the equilibrium condition becomesTAS&,2zk, . . . . . . (36)where the left-hand side is assumed to be the same for all liquids (thoughdependent an the state of vulcanisation of the rubber), and k, is a constantcharacteristic of the liquid. The function TAS,/vF becomes infinite atvr = 1, and decreaaes continuously to zero at V, = p , say, thereafter becomingnegative.The swelling Q is, of course, Q = (1 - v,)/vT, and is therefore afunction of E,, which we can easily evaluate from equation (36). The curveresulting from Gee'8 estimate of the entropy of swelling is shown in Pig. 4.We have now to consider whether E, can be related t o any other property oGEE : THE PHYSICAL CHEMISTRY OF RUBBER SOLUTIONS, 35the liquid, and for this purpose we assume the relationship derived for simpleliquids by J. H. Hildebrand *1 and G. Scatchard.82 If E, and E , are themolar cohesive energies of the liquids, and Vo and V , their molar volumes, wehave .. (37)Applied to rubber, this requires us to know the cohesive energy densityE J V , of rubber, and Gee has shown that a value of 66 cals./c.c. leads tovalues of k, consistent with Whitby's swelling measurements. Equation(37) does not admit the possibility of negative values of k,, and these areindeed to be ascribed always to specific interactions between the rubber andthe liquid. Such cmes being neglected, it is evident that ko = 0 whenEo/Vo = 66, i.e., rubber will swell to its maximum extent in a solvent whosecohesive energy density is 66 cals./c.c. T b is very close to the value (66.54)for cycZohexane.83 Values of E,/V, either higher or lower than 66 cals./c.c.will give positive values of ito and therefore lower values of &. Hence, if weplot & as a function of x = d ~ ( d E o / V o - 1/66}, we should obtain a curveapproximating in form to the error function, with its maximum at x = 0.Gee 36 has shown this to be roughly true, with certain well-defined exceptions.In the first place, aromatic and aliphatic liquids fall on separate curves,characterised by different values of a in equation (37) : this has not beensatisfactorily explained.Highly associated liquids (alcohols, acids) givemuch too large Q values, showing the heat of swelling to be smaller thancalculated. This is parallel to their behaviour in admixture with low-molecular hydrocarbons. 84 Chlorinated liquids also give unduly high valuesof &.This again is readily explicable by analogy with simple liquid systems,for their heats of mixing with hydrocarbonf3 are generally small and frequentlynegative.85 Although, then, the above theory is believed to give a soundbasis for the understanding of swelling, yet it is clear that exceptions to thegeneralisation will be frequent and may well be of great practical importance.Their elucidation, except in the above very crude way, has scarcely begun.This review of the physical chemistry of rubber solutions reveals a fieldof work burdened with a mass of data which until recently were almosttincorrelated and imperfectly understood. Considerable theoretical advanceshave been made in the last few years, and provide a sound foundation on whichto build, but we have a long way to go yet before anything approaching acomplete understanding is achieved.G.G.818283a485op. C i t . , p. 73.Chm. Reviews, 1931, 0, 321; J. Amp. Chem. Soc., 1934, 58, 995.H. Scatchard, S . E. Wood, and J. M. Mockel, J . Physical Chem., 1939, 43, 119.K. L. Wolf, H. Pahlke, and K. Wehage, 2. physikal. Chem., 1935, B, %,1.'' International Critical Tables," Vol. 5, pp. 150 et sep.REP.--VOL. XXXIX. 36 GENERAL AND PHYSICAL CHEMISTRY.3. THE ATOMIC REACTIONS OF THE ALKALI METALS.In the development of the theory of chemical reactivity the natural startingpoint was the investigation of the simplest type of reaction, that of a freeatom with a molecule. Such reactions require the rupture of the minimumnumber of bonds and on this account should be more readily amenable totheoretical treatment.Furthermore, a systematic experimental andtheoretical examination of atomic reactions would lead to the discovery of aclear-cut set of rules which would provide a basis for the understanding ofmore complex reactions. In this respect, the reaction of the alkali-metalatoms with the halogens, hydrogen halides, metallic halides, organic halogencompounds, and in a few cases, compounds containing oxygen and sulphur,which have been carried out almost entirely by M. Polanyi and his collabora-tors over the last 18 years, provides one of the most complete chapters inreaction kinetics. In the present Report the experimental and theoreticaldata relating to this work have been summarised.Since many atomic reactions of the alkali metals occur at far greatervelocities than can be measured by ordinary methods, it was found necessaryto devise new methods suitable for the direct measurements of the rate of fastreactions.Of these, the method of highly dilute flames (so-called because thereactions studied in this way were chemiluminescent) was suitable only forreactions which occur at almost every collision. The second method, called" the diffusion method,'' which eliminates the possibility of surface reaction,can be used for reactions occurring slower than every tenth (or possiblyfifth), or faster than every hundred-thousandth collision of the reactionpartners. The third, and most recently developed, method-" the life-period method ',-is of wide application and more reliable than the diffusionmethod. The majority of the investigations have been carried out by thefirst two meth0ds.l Since there is a fairly sharp distinction between thetype of reaction studied by these methods, the results have been summarisedaccording to the method of investigation employed.I.Highly Dilute Flames.This method was developed by H. Beutler and M. Polanyi2 for themeasurement of reactions occurring a t every collision. In this method, thevapour streams of the reacting substances, e.g., sodium and halogen, enter atopposite ends of an evacuated tube and a t a pressure ( lo3 mm.) such that themean free path is greater than the diameter of the tube. Under these condi-tions the reaction zone (" flame," since the reaction is accompanied byluminescence) is several cm.in length, and it is possible to me88w-e withaccuracy the density of the distribution both of the solid product depositedon the walls of the tube and of the chemiluminescence produced. The1 A detailed review of the work on '' Highly dilute flames " is given by G . Schay(Fortschr. Chem. Phyaik und physikal. Chem., 1931,21). An excellent general review ofboth the " highly dilute flame " and the " diffusion flame " method may be found in themonograph by M. Polanyi, "Atomic Reactions," Williams and Norg8t.q 1932.2 Ndurwi8a., 1926,13,711; Z. Physik, 1928,47,3?9 ; Z.physikd. Chem., 1928, B, 1,3BAWN: THE ATOMIC REACTIONS OF THE ALKALI METALS. 37extent of the reaction will be greatest a t the centre of the reaction zonewhere most penetration has occurred, and will decrease on both sides.If,therefore, no secondary reactions interfere, the precipitation curve will besymmetrically bell shaped. It is a t once apparent that the slower thereaction velocity the greater the zone of penetration and the broader thedeposit curve. By expressing the condition of the stationary state-namely,that the amount of reactant consumed by reaction in any element of volumeis equal to the quantity of that reactant accumulating there by diifuaion-in terms of a differential equation, and integrating, an expression wasobtained which gave the reaction velocity constant, k, directly in terms of themeasured half-breadth of the precipitation curve and the tube resistances ofthe reacting partners.This method was employed in investigating the reactions of sodium andpotassium with ~hlorine,~ bromine: and volatile halides ;s theresulting halides of the metals are non-volatile and readily retained by thewalls of the reaction vessel.All of these reactions were accompanied by astrong luminescence, the investigation of which proved to be extremelyimportant in the elucidation of the mechanisms of the reactions occurring.The luminescence in the case of sodium vapour reactions resulted from thechemical excitation of the D line.’ Haber and Zisch, who had previouslyobserved this emission in the reaction of sodium vapour with halogen andhalogen compounds in nitrogen a t atmospheric pressure, explained it asarising from the collisions of the molecules formed in the reaction andcarrying most of the energy liberated in the reaction with sodium atoms andexciting the latter. At low pressures the light was of much greater intensity,and it was this particular property that gave rise to the name “ highly diluteflames.”A detailed examination of the luminescence was carried out by St.v.Bogdandy and M. Polanyi * by the use of a “ nozzle flame.” The halogenwas introduced from a nozzle into an excess of sodium vapour, and theluminescence was thus rendered much more intense. It was observed thatthe light-distribution curve did not coincide with the precipitate-distributioncurve, and this was explained by the occurrence of the reaction in twostages, the primary reactionNa + C1, -+ NaCl + C1 + 40.7 kg.-cals..M. Polanyi and G. Schay, 2. physikal. Chem., 1928, B, 1, 30, 384.H. Ootuka, ibid., 1930, B, 7,406. H. Ootuka and G. Schay, ibid., 1928, B, 1,62.Idem, ibid., p. 68 (Na + HgC1,); see Ref. (4) (Na + HgBr,).F. Haber and W. Zisch, 2. Physik, 1922, 9, 302 ; H. Franz and H. Kallman, ibid.,1925, 34, 924; K. Liakilov and A. Terenin, Natumoiss., 1926, 14, 83; H. Beutler, St. v.Bogdandy, and M. Polanyi, 2. Physik, 1928, 47, 379; H. Beutler and B. Josephy,Saturwiss., 1927, 16, 540; R. L. Hasche, M. Polanyi, and E. Vogt, 2. Physik, 1927, 41,583; K. Kondratjew, ibid., 1927, 45,67; 1928,48,310.8 2. physikal. Chem., 1928, B, 1,21. The heats of reaction given in this Report werecalculated from the most recent values of the heats of dissociation of Na4, CI,, and NaCI,it., 17.5, 56.8, and 97-5 kg.-cals., respectively, and differ from the values in the originalpapers by 5-10 kg.-cals38 GENERAL AND PHYSICAL CHEMISTRY.which took place entirely in the gas phase, and a secondary reaction of thechlorine atoms formed with the portion of the sodium present as molecules :C1 -+ Na, --+ NaCl* + Na + 80 kg.-cals. .. . (2)Most of the chlorine atoms, however, disappeared by reaction on the wallsof the tube,Na + C1+ NaCl + 97.5 kg.-cals. (wall reaction) . . ( 3 )It was established that the secondary reaction involved Na, molecules by thefollowing observations : (a) The intensity of the light was diminished andalmost completely extinguished by raising the temperature of the reactionzone ; this was quantitatively explained by the increased dissociation of theNa, molecules a t the higher temperature.( b ) The light yield increased withsodium pressure more rapidly than would correspond to a linear relationship-These observations may be completely explained on the basis that the lightwas produced, not by direct excitation of the sodium atoms produced in (2),but by collision of the NaCl* molecules (carrying most of the energy of thereaction) with other sodium atoms which were thereby excited to emissioiiof the I) line :NaC1" +Na+NaCl + Na* (ZP) . . . . (4)Na*+Na+hv . . . . . * (5)This view was confirmed by quantitative measurements on the quenching ofthe luminescence by added nitr~gen.~ According to this mechanism, boththe secondary and the wall reaction contribute towards the total deposit ofhalide, and so the velocity constant could not be determined directly from thehalf-breadth of the precipitation curve.However, from a quantitativeanalysis of the precipitation- and light-distribution curves, it was possibleto resolve the deposit into its part reactions and so to determine accuratelythe velocity constants of the primary, secondary, and wall reactions for thereaction of sodium vapour with chlorine, bromine, and iodine, It was foundthat all the gas reactions occur without activation energy, some of the re-actions, in fact, occurring at a greater rate than that calculated on thenormal kinetic theory basis if every collision is assumed to be effective.It was shown by St.v. Bogdandy and M. Polanyi9 that when sodiumatoms are introduced into a mixture of chlorine and hydrogen, severalthousand molecules of hydrogen chloride are formed for each alkali atomreacting. This confirms the occurrence of the primary reaction (I), thesplit-off chlorine atom inducing the H,-Cl, chain. This inducing effect canalso be demonstrated in a mixture of chlorine and methane, e.g.,Na + C1, --+ NaCl + C1C1+ CH, -+ HC1+ CH,CH, + C1, + CH,CI + C1, etc.Similar investigations have been carried out with potassium vapour and2. Elektrochem., 1927, 53, 554; Naturwias., 1927, 15, 40; M. Polanyi, Trans.Faraday SOC., 1928, a-606BAWN : THE ATOMIC REACTIONS OF THE ALKALI METALS. 39the halogens.H. Ootuka 10 used potassium containing 7% of sodium inorder to permit visud observation of the luminescence, since the violetradiation of potassium is poorly visible, M. Krocsak and G. Schay l1 andF. Roth and G. Schay l2 worked with pure potassium and showedgthat thereactions occurred with zero activation energy by a similar mechanism to thatof sodium, but the chemiluminescence was much more complicated. Thereactions of both sodium and potassium with the hydrogen halides wereinvestigated by G. Schay,13 using the " nozzle flame method." In both casesthe reaction velocity increased from hydrogen chloride to hydrogen iodide,and this was in the order of the heats of reaction of the different series, Inthose cases where the heats of reaction were positive it was found that, toaccount for the observed velocities, enhanced collision diameters had t o beintroduced into the gas kinetic expression.The chemiluminescence of thesereactions probably occurs by a secondary reaction of hydrogen atomsproduced in the primary reaction, vix., Na + CIH --+ NaCl + H - 5000cals., although the actual fate o f the hydrogen atoms is not certain.A further group of reactions of which the mechanism has been investi-gated by the method of highly attenuated flames is that of the volatileinorganic halides of mercury,1* cadmium and zinc 15 with both sodium andpotassium vapour. The reaction mechanism was shown to be the following :Primary reaction : Na + HgCl,+ NaCl + HgCl + 25 kg.-cals.Secondary reaction : Na + HgCl + NaC1* + Hg + 65-3 kg.-cals.NaCl* + Na --+ Na' + NaCl c_, Na + h vThe sole source of the luminescence, which was much weaker than in the caseof the halogens, was found to be the secondary reaction.In contradistinctionto the reactions of the halogens, all the phenomena connected with thepresence of Na, molecules were absent, since no free halogen atom wasproduced in the primary reaction, All of these reactions occurred withoutinertia,M. Polanyi and G. Schay 16 observed that when sodium vapour reactswith stannic chloride, bromide, or iodide, the light emission is different fromthat observed in the previous reactions and shows a completely continuousspectrum. This luminescence is of a primary character and is emitted during,and not subsequent to, the primary reaction.The explanation offered wa8t,hat the SnCl, radicals formed by the reaction Na + SnCl, + NaCl +SnCl, lead to a subsequent luminescent reaction, SnCl, + SnCl, -+SnCl, + SnCI, 4- h v . R. A. Ogg and M. Polanyi l7 have explained thelo 2. physikal. Chem., 1930, B, 1, 422.l2 Ibid., 1935, B, 28, 323.l4 H. Beutler, St. v. Bogdandy, and M. PoIanyi, Natumuiss., 1926, 14, 164 ; also refs.(3) and( 8) ; H. Ootuka and G . Schay, ref, (6) ; K. Icondratjow, Physikal. 2. Bovietunion,1933, 4, 57; J. Berger and G. Schay, 2. physikal. Chem., 1935, B, 28, 332; H. Ootuka,ibid., 1930, B, 7, 406.l1 Ibid., 1932, B, 19, 344.1s Ibid., 1931, B, 11, 291.E. Horn, M. Polanyi, and H. Sattler, ibid., 1932, B, 17, 220.l6 2.Physik, 1925, 47, 814.l7 Trans. Faraday SOC., 1935, 31, 137540 GENERAL AND PHYSICAL CHEMISTRY.mechanism of this emission in terms of their theory of ionogenic reactions.They consider that the newly-formed SnIVCl, molecule passes over, byelectronic transition accompanied by the emission of light, into the stableSnWl, molecule. The continuous nature of the emission is explained byvariable quantities of energy being taken up by the translational motion ofthe two newly-formed molecules, which fly off in opposite directions (theinverse of the predissociation process).No luminescence was observed in the reaction of the organic monohalogencompounds, but when the number of halogen atoms in the organic moleculewas increased, an intense luminescence of the D line of sodium was produced.For instance, all the saturated dihalides l* so far examined have shown thischaracteristic luminescence.The light is produced homogeneously, and thereaction was shown to occur in two stages, corresponding to the successiveremoval of the halogen atoms :RX,+Na+NaX+*RX . . . (1)*RX+Na+(R)+NaX. . . . (2)the product R being a biradical, unsaturated molecule, or cyclic compound.Neither reaction (1) nor reaction (2), if R was a biradical, was sufficientlyexothermic to lead to the excitation of the D line. Experimental data alsoshowed that no reactions involving Na, molecules gave rise to sufficientenergy. It was concluded that the excitation reaction was the exothermicrearrangement of R.These reactions were of two types according towhether (a) R is an unsaturated molecule formed either by the closure of adouble bond or by migration of a hydrogen atom, or ( b ) R is a cyclic molecule.These classes of reaction are typified by ethylene dibromide, ethylidenedibromide, and trimethylene bromide :CH,Br*CH,- + Na -j- NaBr + CK,=CH,CH,*CHBr- + Na + NaBr + CH,=CH,CH,Br*CH2*CH2- + Na + NaBr + CH,-CH,\/CH,By analogy with the halogen reactions, it was assumed that the sodiumhalide formed possesses an internal energy greater than 50 kg.-cals., which istransferred to the sodium atom by collision, thus leading to luminescence.In the case of methylene bromide or iodide, the necessary energy forexcitation is supplied by the change in valency of the carbon atom in passingfrom the quadrivalent to the bivalent state in methylene.This change invalency of the carbon also explains the luminescence of the reactions of thepolyhalogen methane derivatives studied by Haresnape, Stevels, andWarhurst (p. 42).11. Diflwion Flames.The method of highly dilute flames is not suitable for reactions which donot occur at approximately every collision, since in such caBes the length ofC. E. H.,Bawn and W. J. Dunning, Trans. Paraday SOC., 1939,86,186BAWN: THE ATOMIC REACTIONS OF TRE ALKALI METALS. 41the tube needs to be very great. The method also suffers from the furtherdisadvantage that unless the reaction is accompanied by luminescence it isvery difficult to estimate the part played by wall reactions.These diffi-culties are avoided in the " diffusion flame " method developed by H. v.Hartel and M. P01anyi.l~ The principle of the method is similar to that ofthe nozzle flames in that it measures the distance one component can diffuseinto the other before being consumed by reaction, but by introducing bothreactants in a stream of inert gas, diffusion can be made so slow that thesodium vapour leaving the nozzle is used up before reaching the walls.Surface reactions are thereby eliminated. The distance to which the sodiumvapour penetrates is observed by means of a resonance lamp. By settingup the conditions of the stationary state and integrating, the following simpleformula for the velocity constant, k, may be deduced,lgk = (8/r2PH1) In (PT/POl2where po and pT are respectively the pressures of sodium at the nozzle andthe edge of the " sodium flame " m a d e visible by the resonance light, T isthe radius of the flame, 6 the diffusion constant of sodium, andpH, the pressureof halide in the reaction vessel.The derivation of this formula depends,however, on the validity of several assumptions which in the earlier workmay not have been strictly fulfilled. An elaborate investigation of theseassumptions was subsequently carried out by W. Heller,20 who determined theconditions under which the above formula holds good, and under which themethod may be employed to give quantitative determinations of reactionvelocities. By the use of the original method, v. Hartel and Polanyi l9and H.v. Hartel, N. Meer, and M. Polanyi21 carried out a systematicinvestigation of the reactivity of organic halides towards atomic sodium. Itis not possible in this review to give more than a general summary of theirresults, any reference to the particular organic halides investigated beingomitted. The authors showed that an increase in rate of reaction wasproduced by the following factors : (a) Variation of halogen from fluorine toiodine in the methyl, ethyl, and phenyl halides. - ( b ) Lengthening of thehydrocarbon chain, R, in the series RCI. (c) Passage from primary tosecondary to tertiary carbon. (d) Introduction of double bonds or carbonylgroups in neighbouring positions to the halogen. ( e ) Multiple substitution byhalogens; the greater effect was obtained when the chlorine atoms in themolecule were closer together, and a retardation of the reaction was observedwhen the chlorine atom was attached to a doubly-bound carbon atom.The laws deduced were explicit and simple, and N.Meer and M. Polanyi 22showed that there was a remarkable parallelism between the sodium atomreactions and organic reactions in which a single ion reacts with a molecule,e.g., I- + CH,Cl+ CH,I + C1-. The parallelism was not complete, andin some cmes an inverse effect was observed in the two group, e.g., the effectof increase in chain length. These differences, however, were explained on2. physikd. Chem., 1930, B, 11, 97.21 2.physika.Z. Chem., 1932, B, 19, 139.Trana. Paraday Soc., 1937, 88, 1556.a2 Ibid..p. 16442 GENERAL AND PIAYSICAL CHEMISTRY,lheoretical grounh. As a whole, the reeults achieved in this work must beregarded 8 most valuable contribution to the elucidation of the mechanismof organic reactions.More recently, J. N. Haremape, J. M. Stevels, and E. Warhurst 23 havecompared the reaction velocities of eleva of the fourteen possible methanederivatives containing chlorine or bromine or both. The authors conoludethat an increase in atomic refimtion runs pardel with a decrease in collisionnumber. W. Heller and M. Polanyia had previously pointed out theparallelism between decreasing force constante of the GCL linkage andincreasing rate of reaction in the seriea, CH&l--+ CCl,.The velocities of the reaction of sodium with the hydrogen halides has beendetermined by H.v. Ha~%el.~~ The same order of reactivity wa8 observedas that found by Schrty (Zoc. cit.), using the highly dilute flame technique.More recently, C . E. H. Bawn and A. G. Evans 26 have shown that the rates ofreactions of hydrogen and deuterium chloride are approximately the same,the activation energies being 6100 and 6400 cals., respectively. It is ofinterest that the full zero-point energy differences of IICl and DCl (- 1500cals.) do not appear in the differences of activation energy. The remom forthis have been discurssed theoretically by A. G. Evans and M. G. Evan~.~'The reaction of sodium with cyanogen gas l@ differs in a characteristicmanner from that of the halogens in that the reaction Na + (CN), +NaCN $- CN, although exothermic, ocours a t 1 in 16,000 collit~ions and thecollision yield is the same tct 300" and 600".The inertia is thus not the readtof &n aotivstion energy factor but ia probably due to sterio hindrance.The readion with cyanogen chloride and cyanogen bromide a* was followedby determining of the ratio of the products NaCl (or NaBr) and NaCN. Theratio of $hese products, formed in the primary reaction, was not greater thanI : 60 and deoreased with inorease in temperature. The results can beexplained by the occurrence of two primary readions :A N a C I + CNNa, $- 'ICN\ NaCN + C1whioh take place at different rates. The mechanism of the luminescence ofthese reactions is similar to that of the halogens.The reaotion velooities of a series of inorganic polyhalides (BCl,, BBr,,CCl,, SiC14, GeC14, SnCl,, TiCl,, PCI,, AsCl,, SCI,, S,Cl,, COCl,, POCI,,and CrO,Cla) have been determined, by meam of the diffusion-flametechnique, by W.Heller and M. Polanyi,29 who oonoluded, from these andprevious investigations on the halides of the second period, that the poly-halidea of the elements belonging to the middle groups of the periodic tableshow the highest inertia in their reactions with sodium vapour. With each48 Tram. Paraday SOC., 1040, 86, 466.2r Z.phyeikd. Chern., 1931, B, 11, 316.26 Trans. Paraday SOL, 1935, 31, 1392.28 Prans. Paraday SOL, 1936, 32, 633.C-t. rend., 1934, 199, 1118; W. Heller, iW., p. 1611.27 Ibid., p. 1400.J. Curry and M. Polasyi, 8.phyeika2. Chem., 1932, 23, 20, 276; see slso ref. (19)B A W : THE ATOMIO REAOTIONS OF TH.E ALKALI METALS. 43of the series, CCl, Lt SnCl,, CH,Cl+ CCl,, and CH3Cl-+ CHJ, thechemical inertia was shown to increase with increase of the restoring forceacting in the halogen linkage. Visible luminescenoes were obtained withall the above compounds with the exception of SiC1, and M I , . Theluminescences of COCI, and PC13 reactions were identified as the sodium Dline and probably arise by a mechanism similar to that of the halogens ormercuric halides (Zoc. cit.). The luminescence observed with GeC1, and theother compounds investigated, with the exceptions of COCl,, PCl,, CCl,,SiCl4, and AsCl,, showed a continuous speutrum similar to that with SnCla(loc.cit.) and may be accounted for by a mechanism similar to that given forthe latter. In the reactions of S2C1,, SCl,, or CS,, a further ser7es ofluminescenms was observed when oxygen was added, and although noexpbnation of this waB given, it should be noted that sulphur was present ineach of the molecules.The modXed procedure of W. Heller and M. Polanyi 29 has been UBBd byC. E. H. Barn &nd A. G. Evans for measurement of the velocity of reaofionof sodium atoms with oxygen, the oxides of nitrogen and sulphur, and withnitromethane, ethyl nitrate, and amyl nitrite. These studies present a newfeature in so far as the attack of the sodium was at a multivdent atom. Thereaction between sodium and oqgen was first inveatigsted by the diffusionflame method by F.Haber and B. Sachse,31 who showed that the velocityconstant varied linearly with the inert-gas pressure. They interpreted theseresults in terms of the termolecuhr reaction, Na + 0, + X __f NaO, + X,uhere the inert gas X acts a8 a third body to uauy off the energy of thereaction. Bawn and Evans confirmed these observations at low inert-gaspressures, but found that a t higher carrier-gas pressures the reaction changesover to a simple bimolecular association process. The following mechanismwas suggested :Na + O2 -+ Na02* (kl)Na02* 4 Na + 0, (k,)Na02* + XI_, NaO, + X (4)where X is the inert gas and the ashrisk indiclate8 that the molecule possesl3esthe heat of reaction as internal energy. By setting up the conditions of thestationary state it was shown fhrtt- dpNa/dt = bh3Po2 = ~ 1 ~ 3 1 ) m a .P 0 2 / ( k 2 + kflx)where k is the measured velocity constant, calculated from the expressiongiven on p. 41, whence it follows thatand at high pressures, when k, < k3, E = kl, Le., the reaction is bimolecular.The valuet3 of k,, k,, and k3 were determined from the experimental data forboth the reactions of oxygem and nitric oxide. The lifetime of the inter-mediate complex, z (= 1/k2), wm found to be about 8ec. for both NaO,*30 Tram. Paraday SOC., 1937, 83, 1680.*1 8. p & W . Chnc., Bodenstein Feetbd, 1031, 831.B44 GENERAL AND PHYSICAL CHEMISTRY.and NaNO*. The values of Ic, were lo2 to lo3 times smaller than thosecalculated by the kinetic theory expression, and this was shown to be due to atransition probability consequent on a change of multiplicity of state, whichappeared as a steric factor in the reaction velocity expreasion.Hydrogen sulphide was found to react with sodium a t about the samerate as nitric oxide, and preliminary observations showed that the ratedepended on carrier-gas pressure, as with oxygen and nitric oxide. W.Hellerand M. Polanyi29 observed that carbon disulphide reac€ed slowly withsodium atoms and interpreted their result in terms of a three-body collisionmechanism. Both sulphur di- and tri-oxide showed remarkable reactivitytowards sodium, the collision efficiency being about 1 /lo. Thermochemicalconsiderations indicate that none of the above reactions could proceedunder the conditions of the experiments by a metathesis reaction, andit was concluded therefore that all were association reactions.I n thereactions of the sulphur oxides, association occurred a t every collision, thelifetime of the intermediate complex being sufficiently long to ensure that itexperienced at least one collision with an inert-gas molecule during its life-time, Sodium atoms were found to be unreactive towards water vapour andcarbon dioxide at inert-gas pressures from 2 to 13 mm.The reactions of sodium with nitrous oxide and nitrogen dioxide 30 takeplace according to a bimolecular law, with collision numbers of 40 and 10respectively. Both reactions are accompanied by an intense luminescenceof the sodium D line, and this was explained by the occurrence of a series ofreactions analogous to those of the halogens, vix.,Na +N,O +N, +NaO .. . . . . . (1)NaO + Na, --+ Na20* + N a . . . . . . . (2)Na,O* + Na+ Na,O + Na* --+ Na+ hv . . (3)The mechanism was in agreement with the observations that ( a ) theluminescence decreased with increase in temperature of the reaction zone, ( b )the greatest luminescence was in the region of the nozzle, where the concen-tration of Na, molecules was highest, (c) neither (1) nor (4) was sufficientlyexothermic to allow of the production of luminescence in the primaryreaction.In contrast to the alkyl halides, the variation of structure from nitro-methane to ethyl nitrite to amyl nitrite 32 caused no appreciable change inthe activation energy, which in each case was about 2500 cals.Na + NO, -+ NaO + NO .. . . . . . . (4)111. Life-period Method.The " diffusion flame " method, although free from the difficulties of wallreaction, relied for the evaluation of the reaction velocity constant oncalculations of a somewhat doubtful degree of approximation, concerningthe stationary distribution of a diffusing and at the same time reacting gas.88 C. E. H. Bawn and A. G. Evans, Trane. Paraday Soc., 1937,33,1671BAWN: THE ATOMIC REACTIONS OF THE ALKALI METALS. 45Following a critical examination of the possible errors, W. Heller 2O hasascertained the experimental conditions under which the measured reactionvelocity should not differ by more than a factor of 2 from the absolute ratesfor reactions with collision numbers between 50 and 5000.By a furthermodification of the conditions, and using hydrogen as carrier gas, J. N.Haresnape, J. M. Stevels, and E. Warhurst 23 have extended the diffusionflame method to reactions with collision yields of 10 or possibly less, A newmethod has, however, been developed by L. Frommer and M. Polanyi33which embodies the main experimental feature of the diffusion method butis based upon a different set of measurements which leads to the evaluation ofthe velocity constant independently of any assumptions as to the spatialdistribution of the gases in the flame. The new method can be used over awider range of velocities and reacting substances than the diffusion method.The principle of the method of measurement is as follows : A known streamof n atoms of sodium vapour is introduced into the reaction chamber con-taining an excess concentration c of the halogen compound, and the numberN of sodium atoms in the stationary reaction zone (“ flame ”) is measured.The amount of reaction occurring in any element of volume, dv, per sec.ishcdv. If c is kept uniform over the reaction zone, the total reaction isEcJndv. TheJndv is the number of atoms, N , present in the reaction zone inthe stationary state. Since the total amount of reaction is equal to n, and allatoms entering the reaction zone are consumed before reaching the wall, itfollows that n = EcN or k = 1/m since T = N / n = average life of thesodium atom in the reaction vessel. Both n and c can be evaluated from themeasured rate of flow of sodium and halogen compound into the reactionvessel, and the number of sodium atoms in the uncombined state can bedetermined by measuring the absorption of resonance radiation of sodiumvapour passed through the flame. In the initial investigations, the latterwas determined by photographic measurements, but this laborious methodhas been replaced by a sensitive photoelectric photometer which measuresthe intensity of the light beam directly.34A comparison, in the case of methyl bromide, with the publishedmeasurements of Hartel and Polanyi,lg who used the diffusion method,showed a marked difference. There was, however, a close correspondencewith the results obtained by the revised diffusion method of W.Heller andM.Po1anyi.a This was further substantiated by the fairly extensivecomparison of the reactivities of the polyhalogenated methanes which havebeen measured by the life-period and the revised diff usion-flame method.23The new method has been employed by F. Pairbrother and E. Warhurst 35in the reactions of sodium vapour with chloro-, bromo-, and iodo-benzene.They find that the reaction rates increase in the above order, and that theeffect of change of halogen is much the same in the benzene series as in thecase of the corresponding methyl compounds. E. Warhurst 36 has recently33 Tram. Paraday SOL, 1934, 30, 519.34 F. Fsirbrother and F. L. Tuck, ibid., 1936, 32, 624.3 5 W d . , 1935, 31, 987. Ibid., 1939, 35, 67446 GENERAL AND PHYSICAL CHEMISTRY.carried out a more detailed investigation of the bromobenzene reaction inorder to define more closely the conditions under which the most reliablevalues for the velocity constant may be obtained.The Formation of Radicals by the Interaction of Sodium V u p w andOrganic BaZides.-Much evidence has accumulated to show that the primaryreaction of sodium atoms with organic halides, RX + Na --+ NaX + R-,results in the liberation of free radicals.For instance, by introducing theproducts of the reaction of sodium with methyl bromide into chlorine oriodine, E. Horn, M. Polanyi, and D. W. G. Style 37 showed that bromo- andiodo-methane were formed. Similar experiments were carried out withethyl 37 and phenyl 38 radicals. Additional support for this conclusion wasobtained when the CqBr-Na reaction was carried out in the presence ofhydrogen.Thus, H. v. Hartel and M. Polanyi 39 found that two molecules ofmethane were formed for each hydrogen molecule consumed. In the absenceof hydrogen, the product was ethane formed by the dimerisation of themethyl radicals. In the same way the products of the reaction of chloro-benzene,40 benzoyl chloride and acetonyl chloride 41 have been identified astheir dimers. A conclusive proof of the formation of free radicals in thereaction of sodium and bromomethane was provided by A. 0. Allen andC. E. H. who showed that mirrors of tellurium and antimony wereremoved when placed at a considerable distance from the reaction zone.With tellurium the methyl derivative of the metal was isolated.Theconcentration of the radical removing the mirror was shown to fall off withthe square of the distance from the mirror, showing that it must be formedin the reaction zone and diffuses outwards. The reaction of sodium with thealiphatic &halides 43* 44 has been used to prepare and study the reactions ofbiradicnls.IV. Theoretical.The reactions of the alkali-metal vapours with the halogens or halidesform ionically bound molecules and belong to a large group of reactionstermed by R. A. Ogg and Nl. Polanyi 45 “ ionogenic,” i.e., they involve thetransfer of electric charge-an atom bound initially by a homopolar linkagepasses into the ionic state. If the alkali vapour-halogen reactions proceededby means of a purely atomic process resulting in the formation of an atomic-ally bound halide molecule (which eventually made a transition, with theloss of energy, to the ionic state), then they would be highly endothermic inthe intermediate stage and hence would possess large activation energies.This is contrary to experimental fact, and hence the transition to the37 Trans. Faraday SOG., 1934, 30, 189.38 E.Horn and M. Polanyi, 2. physilcal. Chern., 1934, B, 25, 151.39 Ibid., 1930, B, 11,97; H. v. Hartel, Trans. Faraday Soc., 1934, 30, 187.40 Idem, ibid.42 Trans. Paraday SOC., 1938, 34, 463.43 C. E. H. Bawn and R. F. Hunter, ibid., p. 608.44 C. E. H. Bawn and J. Milsted, ibid., 1939, 35, 889.p5 M e m . Manchester Lit. Phil. Sac., 1933-34, 2’8, 41 ; Trana.Paraday Soc., 1935, 31,dl J. N. Haresnape, Thesis, University of Mancheater.1375BAWN: THE ATOMIC REACTIONS OF THE ALKALI METALS. 47ioniwlly bound halide molmule w-hich rend- tihe reaction enerrgetimllypossible mn& occur whilst the reacting system le still in the intermediateconfigumtion.A theory of these reactions w a ~ put forward by R. A. Ogg and M. Polanyi 45whioh explain8 many of their fundamenhi features. The general form of thetheory ean be illuatrakd by oonsidering a typimi system,Na +ClR-+Na+Cl- + R . . . . (1)On one side we have a homopolar state, and on the other an ionic stah, andthe method oonsists h calculebting two energy surfaoes corresponding ~ ~ p 8 6 -t i d y to thwe states. A linear oonfigurrttion of the three interraatkgprtides being msumed, both surfaces can be represenfed in terms of twoco-ordinates, wix., the separation (a) of mdiurn and chlorine atoms and (b) ofthe chlorine atom and the carbon atom of the radioal R.Both of themsnrfebces are oalculated by the use of appropriate potentid-energy functbns.The important featurw of these s w r f m s are the following : The t b - s t o m ,homopolar surfme (Na-Cl-R), forms two valleys, representing the initidand the final state, between which lies 8 seddle. The final-state valleydetermined in this way corresponds to the homopolar molecule sad the freeradical R, and w i l l not be the true final stwte. The latter is represented onthe surface (Na-Cl-R)~o~c, for which the NaCl is ionimlly bound. Theenergy of this vrtlley lies below the find-state valley (Na-Cl-R)h.Con-versely, there is a shallow valley representing the initial state of the ionicsurfwe (in which the sodium ion is free and the chlorine ion loosely bound toR). Reaction (1) thus occurs by a, transition from the initial-state valley ofthe surface (Na . . . Cl . . . RIh to the final-atate valley (Na . . . C1 . . . R)i.The two sur€mes crom -oh other, and the line of intersection represents allpossible transitions between the initial and the final state. The point oflowest energy value on thS intermtion line cornponds to the actual tmnai-tion state, and the dif€erence in energy between this state and the initits1state gives the upper limit of the activation energy. However, on account ofreamance between the two sta;tea, the surfaces never really cram but separabinto two continuous energy surfaoes.The question as to whether thereaction follows the lower of these two surfaoes (adiabtic meahanism) orwhether there is a probability of a transition acnm the gap between thesurfacm (diabatic mechanism) has been frequently discussed?* The generalconchaion has been reached that, for the above types of ionogenic reaction,the croesing point ig a good approximation to the transition state and thatthere is no restriction on the passage from one surface to the other.This theory has been applied to the calculation of the activation energiesfor the reaction of sodium vapour with (a) methyl and phenyl halides,P7M. G. Evans and E.Warhurst, ibid., 1939, 35, 593; H. London, 8. Physik, 1932,74, 143; R. A. Ogg and M. Polanyi, ref. (46); M. G. Evana and M. Polmyi, Trans.Paraday Soc., 1938, 34, 11; J. L. Magee, J . Ohem. Phyaics, 1940, 8, 687; E. Wigner,Tram. Paraday SOC., 1938, 84,29.M. G. Evans and E. Warhurat, ref. (46)48 GENERAL AND PHYSICAL CHEMISTRY.( b ) hydrogen and deuterium halides,48 and ( c ) the alkali metals.49 Evans andPolanyi 46 showed that in the reaction of sodium with the alkyl halides,repulsive forces were absent on one side of the reaction and that the sodiumatom could approach the homopolarly bound halogen atom as far as thenormal bonding distance of the ionic NaHal without incurring repulsion.This led to a very considerable simplication in the method of representing thechanges occurring in terms of energy surfaces, and the activation energycould now be found without plotting energy surfaces.It may be deducedfrom the theoretical calculations that in the series RCl and Li, Na, K, etc.,the change in activation energy should follow the heat of reaction, but in theseries, RCl + Na, RBr + Na, RI + Na, the changes in activation energywould not be simply related to the changes in heat of reaction. I n thereactions of sodium with hydrogen iodide and potassium with hydrogenbromide and iodide, the energy surfaces indicate clearly an increased prob-ability of the collided state (i.e., a small potential hollow in the configuration)and this explains the enhanced collision diameters (Zoc. cit.) which had to beintroduced into the kinetic expression to account for tha rates of these reac-tions.J. L. Magee46 has recently calculated the energy surfaces for thesodium vapour-chlorine reactions and concluded that the reaction Na, + Clcan occur by three possible mechanisms, two of which can lead directly toexcited sodium atoms and the third to an excited sodium chloride molecule.He also calculates the absolute rates of the reactions by use of the transition-state method. Although all of these calculations are semi-empirical, theactivation energies deduced are of the right order, and it may be justifiablyconcluded that the framework of the general scheme for the construction ofpotential-energy surfaces of the three-centre problem provides a reasonablepicture and explains the characteristics of this wide group of reactions.Luminescence and Quenching.-The potential-energy surfaces of thealkali-halogen reactioizs have also been useful in aiding the understanding ofthe chemiluminescence and quenching of the excited atoms.Evans andPolanyi49 have shown that in the particular case of the reaction Ns, +C1+ Na + NaC1, in which there is an absence of repulsion between thenewly formed particles in the final electronic state of the activated complex,the reaction path shows that practically the whole of the energy is transformedinto vibrations of the newly-formed molecule. Magee, on the other hand, byconsidering alternate mechanisms of the above reaction, concludes that apart of the luminescence may be produced by direct emission of radiation bythe excited atoms produced in the reaction.Qualitative energy diagramshave also been employed by Ogg and Polanyi45 to describe the primarychemiluminescence of the sodium-atannic chloride reaction (Zoc. cit.).The observations of K. 3’. Bonhoeffer 5O and F. L. M~hler,~l who reporteda bright luminescence when hydrogen atoms were introduced into sodium4a A. G. Evans and M. G. Evans, Trans. Paraday SOC., 1935, 31, 1400.4e M. G. Evans and M. Polanyi, ibid., 1939, 35, 178 ; J. L. Magee, ref. (46).61 Physical Rev., 1927, 29, 419.2. phyeikal. Chem., 1942,118, 199; 1926,116, 391BERG : THE PHYSICAL CHEMISTRY OF LATBNT-IMAGE FORMATION. 49vapour at 200°, a slight luminescence with potassium, and none withrubidium and czesium, have been discussed theoretically by J.L. Magee andTaikei Ri.E2 They show that the quenching reaction,Na*+2H+Na+H2 . . . . . (2)is adiabatic, and by means of a model energy surface constructed by semi-empirical methods, they give a qualitative explanation of the experimentalfacts. The above authors consider that the quenching of excited sodium byhydrogen molecules is determined by the primary formation of an Na*H2complex which undergoes an internal quenching and dissociation intoNa + H,. The theoretical evidence indicates that the Na*H, state involvesan attraction.There is much other evidence to show that quenching of sodium fluorea-cence is not merely a direct exchange of energy by collision, but thatchemical interaction between the excited atom and the gas-quenchingparticle takes place.V. Kondratjew and M. Siskin 53 have shown that thequenching efficiency of a number of molecules (O,, N,, CO, NO, I,) is relatedin a general way to the strength of the bond which the atom in the moleculemakes with sodium-a large quenching cross-section corresponds to anexothermic reaction of the type, Na* + AB + NaA + B. Bawn andEvans 3O have observed with oxygen, nitric oxide, and hydrogen sulphide thateffective quenching action corresponds to the direct chemical association,Na* + AB --+ NaAB*. Recently, K. 5. Laidler 54 has treated the generalquestion of quenching of excited sodium by the use of energy surfaces andconcludes that the quenching by atoms is normally an inefficient process,but that with molecules the effect of the additional atoms is to stabilise aquenched complex which finally decomposes to give a deactivated atom and avibrationally excited product.The pronounced ability of unsaturatedhydrocarbons in quenching sodium resonance radiation has been qualita-tively considered by R. G. W. Norrish and W. MacF. Smith 66 in terms ofthe crossing of energy surfaces. They conclude that the presence ofunsaturation in the molecule manifests itself in a relatively large trans-mission probability for the transition between the surfaces.C. E. H. B.4. THE PHYSICAL CHEMISTRY OF LATENT-IMAGE FORMATION INGELATINE-SILVER HALIDE EMULSIONS.The formation of the photographic latent image has never been discussedcoherently in these Reports, probably because its mode of formation was notfully understood and new theories were put forward at frequent intervals.Even now, ideas on detailed aspects are in a state of flux, but we are able todescribe the formation of the latent image in terms of movements of electronsand ions in the silver halide (most commonly, silver bromide) crystals52 J .Chem. Physic8, 1941, 9, 638.54 J . Chem. Physics, 1942, 1Q 34.53 Physical. 2. Sovietunion, 1936, 8, 644.56 Proc. Roy. Soc., 1940, A, 176, 29650 GENERAL AND PHYSICAL CHEMISTRY.containing a smdl amount of iodide which me the hght-sensitive unita inphotographic emulsions. The meohanism of latent-image formation suggwtedby R. W. Gurney and N. F. Mott may be said to have stood the test of timevery much better than any previously advanced theory, and has proved to bevaluable as a working hypothesis which has stimulated research considerably.It appears justifiable, therefore, to consider latent-image formation from theirpoint of view, not attempting to discuss the merits or otherwise of all previoustheories in detail.The main puzzle we have to solve is this: Absorption of quite a fewquanta of light by one of the small crystals or “ grains ” of silver halide in anemulsion renders this grain developable : treatment of the emulsion with asolution containing certain reducing agents will cause this grain to bereduced to metallic silver, whereas the unexposed grains are not so affected.We have to ask : (1) What is the change produced by the absorption oflight ? and (2) what is the mechanism of development and, in particular, howdoes a developer distinguish between exposed and unexposed grains 1There are many other subsidiary points to be considered aa well, whichare, however, all different aspects of the same two questions, and conneotedwith differences in latent-image formation aacording to conditions of exposure,development, e t a(1) Nature of the Photographic Latent Image.-When very large exposuresare given to a photographic emulsion, the emulsion “prints out,” &e.,darkens visibly: the emulsion grains are reduced partly or wholly tometallic silver and bromine.This process is helped considerably if bromineacceptors are present. From these facts, early photographic workersdeduced that the latent image itself is also a amall speok of metallic dverin more or less fwm contact with the parent cryshl.Other conceptions,e.g., the sub-halide theory of the latent image?, 4 are now of mainly historicalinterest. Chemical investigation of latent-image ailver is dBcult becauseof the minute amounts involved, and also because the silver in its intimatecontact with the silver halide behaves quite differently fkom fiee metallicsilver. This finding seemed to support the sub-halide theory, until E. Baur 5and W. Reinders showed that “ photo-halides ” are adsorption oomplexesof silver and silver halide. The best indication of the nature of the latentimage was obtained from the finding that on photolysis (i.e., decompositionby exposure to light) of silver bromide, silver and bromine were the onlyproducts, and that the reaction ourve of the bromination of silver has nokink,’ thus excluding the possibility of formation of sub-halides.Crystal-structure investigations by X-rays of exposed silver halide layers ahowedthat, apart from the silver halide, only silver wm present.* In order to carry1 Proc. Roy. SOC., 1988, A, 164,151.3 W. de W. Abney, “ Instructions in Photography,” London, 1882.4 M. Carey Lea, “ Colloides Silber und die Phofohaloide ” (Ed. by H. LUppo-Cramer,6 Z.phySih1. Chem., 1903,45,613.7 E. J. Hartung, J., 1924, 126, 2198.a I?. P. Kooh and H. J. Vogler, Ann. Physik, 1926,77,496.F. Guthrie, see note in (3).Dresden, 1918, Steinkopf).ti Ibid., 1911, 77, 213, 356, 677BERG : THE PHYSICB~~ CH~M~STRY OF LATXNT-IMAGE FORMATION.51out m y of these investigations, much heavier exposures had to be appliedthan are necessary to produce the latent image. To assume that latent-image substance is metallic silver is therefore an extrapolation, although noargument has been advanced as to why this extrapolation should not bemade.By far the most sensitive detector of the latent image in an emulsion isstiU the process of development, but in that process the original latent imageis lost. R. Hilsch and R. W. PohlO have demonstrated a discoloration oflarge silver halide crystals by exposures to light of the order of that whichwould produce a latent image in photographic emulsions. They argue thatthis discoloration is not visible in photographio emulsions only becauseemulsion layers are so thin. Single crystals about 1 mm.thick were used,and the ab.sorption of blue light by silver bromide crystals produced a newbroad absorption band with a maximum a t 6900 A. Absorption of light ofwave-lengths falling into this band caused a bleaching around the wave-lengthused, leaving parts of the band unaffected.lO* 11 The band is thus not due tohomogeneous particles, but to particles of a colloidal nature, the position ofthe absorption band depending upon the size of the particles. Absorption oflight leads to desttuction of the particles by a process to be discussed below.Here we note that red light can lead to a destruction also of the photo-graphio latent image (“ Herschel effect ”).The wave-length sensitivity ofthe Herschel effect corresponb closely to the absorption band in discolouredsilver halide crystals.l2. l3 All the experimental evidenoe available agreeswith the conception that the result of exposing silver halides to light--andincidentally to certain other influences-is the production of small specks ofcolloidal silver, and the liberation of bromine.( 2 ) Optical and Electrical Properties of Silver Halides.--An understandingof the optical and electrical behaviour of silver halide crystals is essential forany consideration of the mechanism of latent-image formation. Most of theinformation recorded here w a ~ obtained from work on large single crystals,but is often applicable t o the array of small crystals in a gelatine layerconstituting an emulsion.The main difference lies in the larger surface area,of the emulsion grains, which shows up in the spectral absorption.(a) Absorption of light. Ionic crystals as a rule exhibit well-defhedabsorption bands. The silver halides constitute a notable exception to thisrule. With these crystals, there is a “ tail ” on the long-wave-length sideof the main absorption bands, which can be measured almost right across thevisible spectrum.14 The maximum of the main band corresponds toabsorption of light in tightly bound ions in the body of the crystal ; the long-wave-length tail is thought to be due to ions on surfaces, internal cracks andI;. Physik, 1930, 64, 606.lo F.Lohle, Nach. Qes. Wiss. Qfittingela, Math.-Phys. Kl, 1933,II, p. 271.l1 J. H. de Boer, “ Electron Emission and Absorption Phenomena,” Cambridge,13 B. H. Carroll and E. M. IGretchman, J. Ree. Nat, Bur. Stand., 1933, 10, 449.14 J. Egged and M. Biltz, Trcms. Fm&y SOC., 1938, 84, 892.1936, p. 299.0. Bartelt and H. mug, 2. PhysB, 1934,89, 77952 GENERAL AND PHYSIUAL OHEMISTRY.imperfections.15 These ions must be more loosely bound than those in thebody of the crystal. Their binding energy, and their spectral absorption, isaffected considerably by any impurities that may be present.la Hence, thelong-wave-length absorption of silver halides varies for different specimensaccording to their mode of preparation and is therefore often more character-istic for surface impurities present than for the body of the material.Theseeffects are particularly pronounced for the silver halide emulsion grainsbecause of their relatively large surface areas. Emulsions are usuallyprepared in a surplus of bromide ions, so that the grains are covered withbromine ions in excess of the stoicheiometric proportion. This causes a shiftof the long-wave absorption limit towards longer wave-lengths; an evengreater shift is obtained if the crystals are prepared in the presence of asurplus of silver ions.17’ 1 8 s 19,The absorption of silver halides, and in many cases their photographicsensitivity, can be changed profoundly by the adsorption to the grainsurfaces of sensitising dyes.21 Photographic sensitivity can then be extendedeven into the infra-red.The adsorption process alters the absorptionspectrum of the dyes, and it has been shown that the photographic sensitivitycorresponds to that spectrum.22 Dyes to act as efficient photographicsensitisers must fulfil certain conditions, such as planar structure, possibilityof co-planar coupling of the electron transition in the silver halide and thatin the dye, and tendency to aggregate formation. This was reported uponby R. A. Morton last ~ e a r . ~ 3The mechanism of latent-image formation, apart from the absorption oflight, is identical in the dye-sensitised and the natural spectral sensitivityregions of silver halide grains. This is shown by the fact 24 that reciprocityfailure curves * for different wave-lengths of light are parallel, if exposure isplotted against time of exposure; no doubt the curves would be identical ifthey were plotted in terms of number of quanta absorbed rather than of(visual) intensity.Latent-image distribution as between the surface and theinterior of the grains was found to be identical, no matter whether light wasabsorbed in the natural or dye-sensitised sensitivity region.25One dye molecule can, be responsible for the production of as many as 50l5 See M. Ott, in Wien and Harms’s “ Handbuch der Experimentalphysik,” XII,l8 K. F. Herzfeld, 2. physikal. Chern., 1923, 105, 329.l7 W. Frankenburger, ibid., p. 273.18 K. Fajans, 2. Elektrochem., 1922, 28, 499.lQ W. Steiner, 2. physikal. Chem., 1927, 125, 275.K.Fajans and W. Steiner, ibid., p. 307.z2 J. A. Leermakers, B. H. Carroll, and C. J. Staud, J . Chem. Physics, 1937, 5 , 878;23 Ann. Reports, 1941, 38, 7 .zL J, 33. Webb, J , Opt. SOC. Amer., 1933, 23, 318.25 W. F. Berg, A. Marriage, and G. W, W. Stevens, Phot. J . , 1941, 81, 413.* A recsprooity failure curve is a plot of the logarithm of the relative exposure(intensity x exposure time) required to produce a certain density on processing again&,the logarithm of either $he inhensity or the t i m of the exposure (see Fig. 3).Part 1, Leipzig, 1932.21 H. Vogel, Ba., 1873, 6, 1302.J. A. Leermakers, ibid., p. 889BERG : THE PHYSICAL CHEMISTRY OF LATENT-IMAGE FORMATION. 53atoms of 27 which means, as will be seen below, that that number ofelectrons is produced by one molecule.This is accounted for by assumingeither (i) that an electron passes over from the dye to the crystal lattice andthat the dye regains an electron by a mechanism as yet not understood, or(ii) that the dye molecule does not actually lose an electron, but passes itsexcitation energy on to the nearest bromine ion.28 The energy deficits inboth cases would have to be supplied by thermal energy, a t the moment ofelectron transfer in case (ii), or afterwards in case (i). Supply of thermalenergy seems essential for dye sensitisation to work. With many dyes,sensitisation breaks down a t low temperatures, as would be expected,29* 30but this is not generally true.31 This aspect of dye sensitisation requiresfurther study.(b) Electronic proceaaea. (i) Photo-conductivity .Absorption of light inthe absorption tail of long wave-lengths produces an electronic conductivityin silver halide crystals.32 These crystals are insulators a t sufficiently lowtemperatures. The quantum equivalence of the production of electrons isof the order of 1 : for every quantum of light absorbed, one electron can bedrawn across the crystal to the anode if the electric field strength is ~ufficient.~~For smaller fields, the current across the crystal drops : some of the electronsare trapped a t places where electrons can occupy empty states of lower energy.The nature of these traps is not known. The electrons carrying the currentare considered to be liberated from the halide ions in the crystal and to passover to a neighbouring silver ion, thus producing a halogen and a silveratom.34 The electron is very loosely bound, and may be considered to bepractically free to move about in the crystal from one silver atom to thenext, its state of energy being referred to as “ conduction level.” Verysmall activation energies are involved in this movement, since photo-conductivity is not reduced until the temperature of the crystals is reduced tobelow The movement of the electrons in the conduction levelscorresponds to thermal energy, the velocity being of the order of 107 cm./sec.Because of its high speed and frequent collisions with elements of the crystalstructure, the electron gets near to any specified place in or on a silver halidegrain within a very short space of time.This is the reason why the presenceof even quite a small number of electron traps may be very effective inreducing photo-conductivity. Colloidal metal particles present in smallnumbers constitute very effective electron traps.35 This means that theconduction levels of metallic silver are below those of the silver halide.126 W. Leszynski, 2. wiss. Phot., 1926-27, 24, 261.2 7 H. Tollert, 2. physikal. Chem., 1929, A , 140, 355.2 8 See S. E. Sheppard, Atti X Cong. intern. Chim., 1938, 1, 235.29 S. E. Sheppard, E. P. Wightman, and R. F. Quirk, J . Physical Chem., 1934,88,817.30 G. Ungar, 2. Physik, 1937, 106, 322.3 1 J. H. Webb and G. H. Evans, J . Opt. SOC. Amer., 1938,28, 249.32 Wilson, Ann. Physik, 1907, 22, 107.33 W.Lehfeldt, Nach. Ges. Wiss. Citjttingen, Math.-Phys. K1, 1935, 11, p. 170.34 K. Fajans and K. von Beckerath, 2. physikal. Chem., 1921,93,478.35 R. W. Pohl, Proc. Roy. Soc., 1937, 49 (extra number), 1(ii) Halogen atoms. We have seen that abmrption of light of euibblewave-length produces silver and a halogen akom. Htllogen is known to beset free during photolysb of silver halides,’ and we must aak how it gets awayfkom the crystal. If absorption of light occur^ in a halogen ion on or nsar thesurface, the halogen can be taken up by the moisture or gelatine surroundingan emulaion g r h . If absorption oocurs on an internal imperfection, thehalogen atoms caa move by a replaoement procesa : an electron moving froma neighbouring halogen ion to the atom shifte the atom in effect to the placewhence that electron came.Because of the “wave-mechanical tunneleffect ” the activation energy for this process may be quite small, althoughthere is no experimental evidenoe on this The halogen atom maybetrapped, however, by polarising the surrounding crystal structure and thus‘ ‘ digging its own hole ” in the manner visudised by Mott.37 For fhis processto ooour, the atom has to remain stationary for about 10-8 e c . It has beensuggested that high mobility of halogen atoms would make latent-imageformation impossible, since they would follow the electrons and eventuallyrehalogenhe any silver formed.1 We must consider, however, that thehalogen would proceed by diffusion; thus it has a very high &awe offouohing the surface of a grain before it reaches the laht-bmge silver speck.On reaching the surface, it would react with the surrounds of the grainimmediately .(c) Ionic pocessw.At normal temperatura, silver halide orystals areeleotrolyhic conductors,38 the current being carried exclusively by ailver ions,whioh are smaller than the halogen ion^.^^*^ The silver ions move by twoWerent rnechmisma, which appear t o contribute approximately equalamounts to the observed conductivity.(i) Jhteratitial ions. In thermal equilibrium, a, oerhin number of silverions occupy “ h t m f i t i d ” positions in the centre of the unit oube of thestructure, which is distorted thereby. In these posifiom, the ions aremobile, requiring an activation energy of 8200 mh./mol.in silver bromideand 6500 cals./mol. in the chloride. The ooncenfration of these interstitidions is given by a, formula of the type C = Cog- and is always low at roomtemperature, of the order of for the ohloride,the energies of formation being 20,200 and 25,000 cals./mol. respectively.This means that, in an emulsion grain containing on the average lo@ ionpairs, there are only 1000 interstitial ions for silver bromide and 10 for silverchloride. It is possible, however, that owing t o impurities, the number ofinterstitial ions is higher than indicated by the formula; the curve ofconductivity plotted against temperature shows a kink at about room tem-for silver bromide andN. F. Mott and R. W. Gurney, “ Electronic Processes in Ionic Crystals,” Oxford,1940.s7 Proc.Physical SOC., 1838, 50, 186.See C. Tubandt, in Wien and Hmm’s “ Handbuch der Experixnentalphysik,” VII,E. Koch and C. Wagner, 2. p h y 4 d . Claem., 1937, B, 88,286.Part 2, Leipzig, 1928, p. 383 et seq.co W. Jost, Trans. Paraday SOC., 1938, 34,860BERG : THE PHYSICAL CHEMISTRY OF LATE”-IMAGE FORMATION. 65perature, indicating that the number of ions is higher than that oorrespondingto thermal equilibrium.=* 5 9 * a 41(ii) Holes. The hole left in the structure when a silver ion moves into aninterstitial position also has a certain mobility by a replacement mechaniam :a neighbouring ion jumps into the hole, leaving a, hole at the place whence itcame. The activation energy for this process is approximately the same asthat for the movement of interstitial i0ns.3~(3) Latent-Image Forrnation.--(a) Mechanism.How can the facts ofphoto-conductivity and ionic conductivity be used to account for themechanism by which the silver atoms, formed by absorption of light all overthe crystal, can collect in a small speck of metallic silver, situated on thesurface of the grains where the developing solution can reach it 1The clue to the answer was the discovery that photographic emulsions arerelatively insensitive to light, unless certain impurities are present duringemulsion making.&* These impurities were demonstrated to cause theproduction of silver sulphide specks on the grain surfaces [see Section 3, c].The specks were regarded as nucleating centres on which the photolyticallyproduced silver collects.The nucleating centres are now assumed to act bytrapping the eleotrons liberated by light absorption. The centres thusbecome negatively charged and attract the mobile interstitial ions, which aredischarged on the speck. In this way, a speck of metallic silver collects onthe silver sulphide speck,l as is illustrated diagrammatically in Fig. 1.The basic assumption, then, is that an electron in a silver sulphide speckcan occupy states of lower energy than it can in the conduction levels of thesilver halide. There is so far no experimental evidence supporting thisassumption. Silver selenides and tellurides can also act as nucleatingcentres44 and it is probable that other impurities, as well as mechanicalimperfections, can similarly act as electron traps.These last two types oftrap seem to come into action when the more efficient silver sulphide specksare overloaded or do not exist. This would seem to be clear from the factthat latent-image silver is also formed in the interior of photographicgrains,25* 46s 47* 48 where a normal developer does not reach and where silversulphide specks do notIt is commonly assumed that a silver speck hast o be above a, certain size to induce development, although its dimensionshave never been determined directly for obvious reasons. Recourse washad to calculation, the amount of silver produced for heavy exposures beingobserved, and extrapolated down to the small exposures used to form alatent image.For photolysis, a quantum efficiency of the order of 1 hasand are in fact not likely, to occur.(b) Size of Eat& image.41 W. F. Berg, Proc. Roy. SOC., 1940, A, 174, 559.4a T. Lsvedberg, Phot. J., 1932, 62, 310.43 S. E. Bheppard, Coll. Symp. Monograph, Wisconsin, 1923, p. 346.44 Idem, Phot. J., 1926, 65, 380.4 5 F. Kogelmann, “ Isolienrng der Substanz des latenten Bildes,” Graz, 1894.48 H. Llippo-Crmer, “ Kolloidchemie and Photographie,” 2nd Ed., Dresden, 1921.47 A. Kempf, 2. w k ? . Phot., 1937, 30, 236.(8 (Miss) G. Kornfeld, J . Opt. SOC. A m . , 1941, 81, 69856 GENERAL AND PHYSICAL CHEMISTRY.been found by the classical investigations of Eggert and Noddack.26D 49,m 51Since statistical considerations of necessity come in, it is only possible toobtain certain upper and lower limits in this way.Other calculations arebased on the shape of the toe of the characteristic 5 3 p 55* 56- 57containing, in many cases, tacit and unwarranted assumptions on the(31FIU. 1.Latent image formation according to Gurney and Mott.shaded area = sensitivity speck ; solid black area = latent image.The triangles represent a silver halide grain. 0 = interstitial ions; - = electrons;mechanism of latent-image formation and growth. Sheppard 58 has given acomprehensive review of the question and concludes that in certain cases (atthe threshold) * one, quite often a few, on the average (at a density of 1) a4D J. Eggert and W. Noddack, 2. Physik, 1923, 20, 299.50 E. Mutter, 2. wiss.Phot., 1928-29, 26, 193.5 1 H. Kieser, ibid., p. 1.53 T. Svedberg and H. Andersson, Phot. J., 1921,61, 325.54 L. Silberstein and A. P. H. Trivelli, J . Opt. SOC. Amer., 1938, 28, 441.5 5 L. Silberstein, ibid., 1941, 31, 343.5 8 Phot. J., 1931, 71, 331.* See Fig. 2-for an explanation of technical terms.s2 T. Svedberg, ibid., 1920-21, 20, 36.5B J. H. Webb, ibid., p. 569.J. C. M. Brentano and S. Baxter, Trans. Faraday Soc., 1940, 36, 581BERG : THE PHYSICAL CHEMISTRY OF LATENT-IMAGE FORMATION. 57few hundred, and sometimes several thousand quanta have to be absorbed bya grain. This does not mean that the size of the latent image varies withinthese limits. The most sensitive grains may already carry on their surfacesalmost all the silver required before exposure : the very insensitive grainsmay be situated a t the bottom of the emulsion layer, or the amount of silverformed may be distributed over several competing electron 6oLatent-image theory certainly has to account for the fact that certain grainscan be made developable on formation of quite a few silver atoms.I ncontradiction- to these conclusions, M. Savostjanova 6 1 claims that it ispossible to " see " the colloidal silver particles produced by photolysis ofsilver halides under the ultra-microscope. This would require a diameter€xposure = Intensity x timeContrast or y = tan ciZog exposureFIG. 2.Diagrammatic characteristic curve of a photographic material, to explain technical terms.of the order of 1000 A. It has been suggested62 that the particles arevisible by a fluorescence process : electrons are lifted in energy from thesilver to the silver halide conduction levels, and drop back emitting light,This suggestion has not been investigated experimentally.The recognition of the nature andinvestigation of the mode of formation of the sensitivity specks was one of thegreatest experimental advances connected with latent-image research. Itwas found that most photographic gelatines contain sulphur compoundswith groups S=-c< which occur in isothiocyanates and thiocarbamides( c ) Formation of sensitivity specks.59 S.E. Sheppard, Phot. J., 1928, 68, 397.60 S. E. Sheppard, A. P. H. Trivelli, and R. P. Loveland, J. Franklin Inst., 1926,61 9th Congrds International de Photographie, Paris, 1935, p.94.200, 51.W. F. Berg, Trans. Faraday SOC., 1938, 34, 88958 GENERAL AND PHY81Ufi UHEMXSTBY.and can form addition complexes with silver halides. These complexesdecompose on heat treatment, leaving speck8 of silver sulphide in intimatecontact with the silver halide graina. Sheppard 44 and his collaborators wereable to isolate the sensitiaing substances from the vasious by-products ingelatine manufacture. They followed up the various stag@ by testing forthe presence of semitisers by emulsion experiments, using photographicallyinert gehtines to which extracts of by-products were added. Sheppard andJ. H. Hudson 63 investigated the formation of addition complexes. Sheppardcollected data on the amount of silver sulphide present in an emulsion ofoptimum sensitivity. An average figure is 5 x g.of silver sulphide perg. of silver bromide.Continued heating of an emulsion leads to an increase in the amount ofsilver sulphide and presumably in the size of the existing specks. Heattreatment, known as digestion, is an integral part of emulsion making andhas to be controlled rigidly, since two disturbing photographic effects occur ifdigestion is over-done : the sensitivity of the emulsion drops, and the foglevel rises, Le., grains are developable without exposure.64 The drop insensitivity is explained by the formation of several sensitivity specks on onegrain competing with each other for the photolytic silver formed.59 Theincrease in fog is accounted for by the production of oversize specks, whichitself causes a drop in speed.The two effects are, however, distinct.64(4) Developnent.-(a) A few important data. A grain is either developedcompletely or not at all. It is only with great difficulty that one can findpartly developed grains 65s 66 by interrupting development at an early stage.Reduction of grains always begins at discrete points distributed over thesurfaces of the grain^.^^,^' As soon as a grain becomes visibly reduced,reduction proceeds at a fairly high rate ; different grains differ in the lengthof the induction period before visible reduction starts, according to theiroriginal sensitivity and the exposure they received. Large grains are on thewhole more sensitive to light than small ones.6oThe pracess of development shows many features of a reversible reaction,its kinetics following in a large measure the law of mass action; for instance,the rate of development is often proportional to the concentration of thedeveloping substance.65 Bromide ions in small amounts slow down the rateof development since they are a reaction product, thus reducing the rate ofdissolution of silver bromide.(In large amounts, bromide ions exert asolvent action on silver bromide.)Only in special cases, however, &s, e.g., with ferrous oxalate developers, isit permissible to regard development as a strictly reversible reaction. Sidereactions usually occur.88 2. W k 8 . PTbOt., 1927-28, 25, 113.64 See, e.g., B. H. Carroll and D. Hubbard, J.Res. Nat. Bur. Stand., 1931, 7, 219;86 S. E. Sheppmd and C. E. K. Mew, “Theory of the Photographic Proce88,”66 T. Svedberg, Phot. J., 1922, 62, 186.67 M. B. Hodgaon, J . Franklin Inst., 1917, 184,706.1933, 11, 753.London, 1907BERG: THE PEYSICAL CHHMlSTRY OF LATENT-IMAGE FORMATION. 59In thermodynamio equilibrium and therefore with a developer for whiohthe reaction is reversible, the oxidation-reduction potential of the solutionmust lie between two well-defined limits. If the potential is too low, d lsilver halide pins, including the unexposed ones, w i l l be reduced indis-criminately, if too high, the solution acts as an oxidiser, destroying the latentimage.65*6* With developers aa used in practiue, however, the redoxpotential is often considerably below even that required to reduce unexposedsilver bromide.Not all reducing substances can serve as developing agents.Sta,nniteions are often quoted as being in all circumstances incapable of dhoriminatingbetween exposed and unexposed grains.6a Unprotected precipitates of ailverhalides are reduced by normal developers whether they are expomd or not ;it is the presence of gelatine in conjunction with the electric double layerproduced when precipitating silver halides in the presence of surplus bromideions, which protects unexposed grains from redu~tion.~QSilver ions in solutions are adsorbed to metallic silver.mThe rate of development is influenced strongly by the surface conditionsof the grains, my, by adsorption of certain dyes.69The curve of density against time of development goes straight throughthe origin, and has its greatest slope at the origin for developers with neutralmolecules, like p-phenylenediamine and related compounds.70 For ioniseddeveloper molecules, the density-time curve has a region of zero slope nearthe origin, followed by a region of inoreasing slope (a ‘‘ toe ”).The developed silver was shown by photomicrographs 71 and the electronmicroscope T2 to be in the form of a tangled mass of thin threads or filaments.(b) #upersaturation theory of deuebpnent. For tt long time the super-saturation theory of development, due to R. Abegg 78s 74 and W. 0st~a,ld,7~ww held to be valid. The silver halide is supposed to dissolve in thedeveloping solution, the silver ions being reduced to atoms which thenpreoipitate on the latent image speck which acts as a nucleus.Manyobjeotiom have been advanced against this theory and it is now mainly ofhid0rica.l interest. The solubility of silver ha.lides is much too low to aocountfor rates of development as normally obtained, and unexposed grains are notdissolved.7s( c ) Development : a heterogenmw catalysed reaction. M. Volmer 76 firstindicated the view, now widely held, that the distinction between exposedand unexposed grains is due to a different rate of development. Thereaction is catalysed by the latent image and each grain is reduced &B anentity. This conception has been most succesBfu1 in accounting for the faotsW. Reinders, J . Phpical Chem., 1934, 88, 784.6D T.H. James and G. Kornfeld, Chem. Reviews, 1942, 30, 1.70 T. H. James, J . Phy&cul Chem., 1939,43, 701.71 F. E. C . Bcheffer, Brit. J . Phot., 1907, 64, 116, 271.7 2 C. E. Hall and A. L. Schoen, J . Opt. Xoc. Amer., 1941, 31, 281.73 Ann. Physik, 1897, 67, 426.7 6 “ Lehrb. der Allg. Chemie,” Leipzig, 1893, Vol. 2, part I.‘13 2. wk8. Phot., 1920-21, 20, 189.74 ATch. wiss. Phot., 1899, 1, 1560 GENERAL AND PHYSICAL CHEMISTRY.of development, but the way the catalysis works is still under discussion.Gurney and Mott 1 suggested that development functions by a mechanismanalogous to latent-image formation, electrons being handed over from thedeveloper into the latent-image speck. Interstitial silver ions then move upfrom the interior of the crystal and deposit on the latent image.Thisprovides a beautiful account of the threadlike shape of the developed silver,but serious objections have been raised,69 one of them being that themechanism does not provide for the escape of halide ions from the grain.The mechanism is not complete, since it does not account for the influence ofthe surface conditions of the grains on the rate of development. T. H. Jamesand (Miss) G . Kornfeld 69 consider that the reaction takes place on the tripleinterface of latent-image silver, silver halide crystal, and solution. Some ofthe silver ions of the crystal are considered to be " adsorbed " to the latentimage, which thus reduces the activation energy required for a developingmolecule to reduce a silver ion.The rate of development depends on therate at which developer molecules can reach the adsorption site and on thenumber of silver ions there. This accounts for the effect of surface layers onthe rate of development, both mechanical and electrical influences interferingwith the approach of developer molecules. The reaction would tend toproceed chiefly along surfaces and cracks and thus produce silver in fila-mentary form, an explanation which is, perhaps, the weakest point on anotherwise very satisfactory picture.(5) Photographic EfSects.-The theoretical picture of latent-image form-ation given above enables us to account for many of the more detailed featuresof the photographic process. On the other hand, we record here data whichmade it possible for several photographic effects to be accounted for interms of each other ; these data are of value regardless of any theories.The effects discussed now, and some others, are summarised in the charton pp.66, 67.The recognition that both electronic and ionic processes are involved inlatent-image formation gave a clue to the explanation of the effects known asreciprocity failure. The two processes have different time constants andtemperature coefficients ; this makes it clear why optimum sensitivity isoften found at a certain intensity level. Under certain conditions ofexposure, the processes can be separated in time and to some extent studiedseparately .(a) Reciprocity failure. The photochemical reciprocity law which holds forsingle-stage reactions states that the same product I x t of intensity of lightI and time of exposure t will produce the same effect, regardless of the valuesof I and t.77 For the double-stage photographic reaction this law cannot beexpected to hold : with most materials there is an optimum sensitivity a t acertain exposure time and intensity, and sensitivity falls off at both higherand lower intensities, so that we speak of high- or low-intensity reciprocityfailure (see Fig.3).7 7 R. Bunsen and H. E. Roscoe, Ann. Phpik, 1866, 96, 373; 1867, 100, 43, 481;1868, 101, 236; 1869, 108, 193; 1862, 117, 629BERG : THE PHYSICAL CHEMISTRY OF LATENT-IMAGE FORMATION. 61(i) Low-intensity teciprocity failare. C. E. Weinland78 and J.H. Webband C. H. Evans 79 have demonstrated that low-intensity reciprocity failure isdue to inefficiencies of latent-image formation in its first stages of growth.Stable, but not yet developable silver specks can be produced much moreefficiently a t shorter times of exposure and thus at higher intensities. Alow-intensity exposure following one at high intensity is as efficient as thelatter if it comprises approximately half the total exposure required. Tworeasons for the inefficiency of a low-intensity exposure have been advanced.The latent-image speck may not be able to grow unless a certain minimumI.42.72.451.4 3.0 <-6 3.2 3.8 2.4 r.0 7.6 0-2 0.8Log I (ergslcmZlsec.).FIU. 3.Thelog It necessary to produce a constant density is plotted against log I [Fig.77 ofMees’s book (see p. SS)].Reciprocity failure of a photographic material at different temperatures.concentration of electrons is reached in a grain.1 At low intensities this willnever be reached because electrons are lost all the time by recombination.Webb and Evans 79 assume that a speck of silver actually begins to form butdisintegrates by thermal agitation into electrons and interstitial silver ions.At the moment, a decision between these very similar mechanisms is notpossible; a process using up electrons in some other way is necessary t oaccount for low-intensity failure, but this has not been studied or understood.(ii) Hypersensitisation and latent-image intensi$cation. Weinland’s T8and Webb and Evans’s 79 results demonstrated that low-intensity failurecan be largely eliminated by giving a uniform pre-exposure a t higher intensi-ties, thus hypersensitising the photographic material.The reverse processJ . Opt. SOC. Arne?.., 1927, 16, 337; 1928, 16, 296. 70 Phot. J., 1940, 80, 18862 UENEkAL ABD PHYGICAL CEEMIBTRY.is cllso practised and is known as latent-image inbmifiocttion or '' latemi-fictlfion " ;ao~ *l an under-exposed picture obtained a t a, high intensity of lightcan be intensified by a uniform po&-exposure of long duration and lowintensity. These two effects are thus simply a manifestation of the existenceof low-intensity reciprocity failure.(iii) HigA-intensity reciprocity failure. High-intensity reciprooity failureseems to be fairly well understood.At high intensities the ionic3 process ig notsufliciently fast to neutralise the electrons colliding with the sensitivityspecks, whioh become charged to capacity and repel any further e1ectrons.lAs a consequence, latent image is formed elsewhere in and on the grain,25* 82presumably on electron traps which for normal intensities would not beeffective in holding electrons long enough for a latent image to be formed,or on places where a high momentary electron concentration causes thelattice to break down so as to form a silver speck. More latent image isformed in the interior of the grains, the higher the intensity of light; 25 thismeans that high-intensity reciprocity failure occurs only with latent imagesituated a t the surface of the grains, and will therefore vary according tothe amount of silver halide solvent in the developer and the duration ofdevelopment.25* 83 High-intensity reciprocity failure is thus characterisedby a change in distribution of the latent image; there is no evidence tosuggest that the fotal amount of silver formed decreases with increasingintemity.(iv) Veery high intensities.At very high intensities, a state is reachedwhere, as far as the ionic process is concerned, electrons are released practic-ally instantaneously : shortening the exposure time still further does notresult in a further change in the mode of latent-image formation, andreciprocity failure is absent.41 At an exposure time of about 4 x sec.the reciprocity failure curve (see footnote, p.52, and Fig. 3) of 4 materialstested bent over into the horizontal. The bend-over point shifts towards longtimes of exposure with decreasing temperature, depending on the mobilityof the ions, which varies with temperature as e-Elk27. This is because, belowroom temperatures, the number of interstitial ions is the con-ductivity varying as the mobility. In contradiction to this, Webb 84 hasfound from an investigation of the influence of temperature on reciprocityfailure, that the concentration of interstitial ions varies down to very lowtemperatures ; this contradiction has not been cleared up.(v) Effect of temperature on photographic semitivity . Webb 85 has clearedup a controversy of long standing by demonstrating that the temperaturecoefficient of photographic sensitivity depends largely on the intensity levelof the exposure. The change in reciprocity failure characteristics withtemperature as shown in Fig.3 can be accounted for, at least qualitatively,8o G. E. Moore, Phot. J., 1941, 81,27.81 Amer. Cincmcrtogr., 1940, 21, 499; Kinematograph Weekly, 1941, 295, 15.82 Phot. Cow., 1936, 72, 1.8s L. W. Strock, Norake Vidensk. Skr., 1938, I, Mat.-Naturv. Klasse, No. 10.@4 J . Opt. Boe. Amm., 1942, 33, 299. Ibid., 1936, 26, 4BERG : THE PHYSICAL OHEMISTRY or LATENT-IMAGE FORMATION. 63on the basis of the mechaniam described. The drop in ionic mobility atreduced temperature affects the branch of the reoiprocity failure ourve onthe high-intemity side of the optimum : at low temperatures, repulsion ofelectrons from a sensitivity speck occurs at an intensity where at normaltemperatures the mobility of the ions is quite sufficient to disoharge someof the electrons on the speck during the exposure.Thus the whole high-intensity branch shifts towards lower intensities at reduced, and towardahigher intensities at elevated temperatures.Low-intensity failure is affected in a similar fashion, but for a differentreason. Here, the stability of a small speck of silver depends on temperature,the probability of an electron being lost from the speck in a time intervaldepending on temperature as e-WIk*. The temperature ooefficient for theprocesses responsible for high- and low-intensity failure being different, theshape of the reciprocity failure curve alters somewhat with temperature ; themost important effect, however, is a bodily shift of the whole curve towardslower intensities as the temperature is deoremed.(vi) Intermittency effect.An intermittent exposure sometimes giveshigher, sometimes lower density than an uninterrupted equal exposure at theintensity of the flashes. Webb 86 hrta shown up the long suspected65conneotion between this effect and reciprocity failure. At a sufficiently highrate of interruption, a state is achieved when on the average a single quantumof light is received by each grain for each flash of light passed by a rotatingsector. For higher speeds of rotation of the sector, the incidence of quantaon m y one grain is not altered and corresponds to the average light intensitypassed by the sector.On varying the speed of the sector fkom stsndstill tohigh speed, at the ends of this range we are dealing with exposures at twowell-defined intensities a t which, according to the reciprooity characteristicsof the material, photographic sensitivity is Werent. According to whetherthe intensities fall on the high- or the low-intensity branch of the reoiprocityfailure curve, the continuous or the interrupted exposure will produce ahigher density. A seotor operating a t very high speeds is thus a convenientmeans of providing a scale of intensities for photometric use.(vii) Very low temperatures. At sufficiently low temperatures, the flat,high-intensity branch of the reoiprocity curve *1 ocoupies the whole range ofintensities and exposure times practicable : reciprocity failure disappears88 Here we have a complete separation of at least the first partof the electronio from the ionio process. In order to account for the faot thatphotographic sensitivity persists at very low temperatures, it is neoessltry toassume that eleotrons are trapped in the grains so as to prevent them fromrecombining; on warming, the ionic prooess occurs.It has been demon-strated that the latent image is in a fairly labile state until the material iswarmed : red light frees the electrons very effectively from their traps, sothat a latent density is easily bleached out; after the warming, this is no8% J . Opt. SOC. Amer., 1933, 23, 187.W. F. Berg md H.Mendelesohn, Proo. Roy. Soc., 1938, A, 168,168.C. H. Evans and E. Hirsohlaff, 3. Opt. 8oc. ATWP., 1989, 89, 16464 GENERAL AND PHYSICAL CHEMISTRY.longer pos~ible.3~9 89 The electron traps are of two kinds : the sensitivityspecks, which are of the order of one volt deep and hold the electrons tightly,and shallower, less well specifiable traps, from which electrons are releasedon warming. The small capacity of the sensitivity specks enables only themost sensitive ones to take up a sufficiently large number of electrons during asingle low-temperature exposure ; hence the loss in sensitivity. Repeatedexposures with intervening periods of warming increase the sensitivity to avalue approaching that of room tem~erature.~~ The loss in sensitivityaffects in the main the surface latent image, which is formed by the help ofsilver sulphide specks.The internal image sensitivity is but slightly affected,demonstrating that its mode of formation is different from that of the surfaceimage.25 Changes in latent-image distribution thus do not account for thedrop in practical speed at low temperatures. A drop in speed and contrastoccurs even between 90" and 20" K.,87 although at the higher temperatureionic conductivity is already zero. At 4" K. photographic sensitivity stillpersists.89 The changes occurring at these very low temperatures are notunderstood.(b) Herschel efSect. The Herschel effect is the bleaching out of a latentdensity (originally of a print-out density) by red or infra-red light. Ingeneral terms, it is accounted for by a reversal of latent-image formation :light can eject electrons from the silver speck, the electrons are trappedoutside the latent image, and silver ions leave the speck, going back into thesilver halide lattice as interstitial i0ns.l experimental evidence agreeswith this conception, except that we cannot account for the method ofelectron trapping.It has been suggested that the Herschel effect consists ina dispersal of the latent image, the total amount of silver present not beingaffected : this has not been proved convincingly.27* 91 The effect isstronger the higher the intensity of the first exposure : the numerous smallspecks of latent-image silver obtained a t high intensities are more readilydestroyed than the larger ones produced by a low intensity exposure.(c) Clayden efSect. In the following series of influences, viz., pressure,X-rays, very brief exposure to light, normal exposure to light, any of thepreceding ones desensitises a photographic material towards a subsequentone if given as a, pre-expos~re.~~ The combination of a brief followed by anormal exposure to light has been investigated more fully than the others,and the results are presumably valid for the whole series.As suggested byLcppo-&amer,82 the brief exposure produces internal latent-image specks,which compete with those on the surface for the latent image formed by thesubsequent exposure of normal duration.25 A similar effect occurs when alow-temperature exposure is followed by one at room temperature;89 we donot h o w where this fits into the series of influences quoted.Solarisation is the phenomenon in which, with certain (d) Solarisation.89 W.F. Berg, Trans. Paraday SOC., 1939,35,445.90 W. Leszynski, 2. wiss. Phot., 1926-27, 24, 275.91 A. P. H. Trivelli, J . Franklin Inst., 1929, 207, 765.92 R. W. Wood, Phil. Mag., 1903, 6, 577BERG : THE PHYSICAL CHEMISTRY OF LATENT-IMAGE FORMATION. 65photographic materials, continued increase in exposure eventually leads to adrop in developed density. This only occurs with the surface latent image 25and can be removed by treating the emulsion with a halogen acceptor beforeexposure or a silver halide solvent afterward~.~3 The effect is caused by thebromine atoms which are formed by the exposure; they leave the grains,attack the surface image, and form a coating of silver halide through whicha normal developer cannot act. Solarisation can be removed to some extentby prolonged development 25 because of the slight solvent action of theconstituents of a normal developer.Another effect connected with latent-image distri-bution and the formation of internal latent image is the Albertwhich is the finding that, if a material is treated with chromic acid after apreliminary exposure, a subsequent exposure will form a positive. Theeffect was at first accounted for by assuming that a combination of sensitivityspeck and latent image was destroyed the more readily the lerger the latent-image specks, so that the grains became desensitised more the larger the firstexposure had beeneg5 G. W. W. Stevens 96 has shown that the desensitisationis due to internal latent-image nuclei produced by the first exposure. Thisaccounts for the finding that the effect is largest for very brief first ex-posures, and that it does not occur if physical development after fixation isused.(f) Subatier reversal. A photographic material, if pre-exposed anddeveloped for a short time, gives a positive image with a second exposure.Optical screening by the density of the image developed after the firstexposure is to a large measure responsible for the rever~al,~' but it has alsobeen shown that a desensitisation of grains occurs during the first develop-ment,98 since chemical '' fogging agents " like sodium arsenite can producereversal. The mechanism of this desensitisation is not clear. It has beensuggested that silver is transferred from developing to neighbouring grainswhich are thereby desensitised. Silver transfer has, in fact, been demon-~trated,~* but why this should have a desensitising rather than a sensitisingaction is dficult to see, unless the silver is deposited in a very disperse state.The specks would have to be too small to act as latent-image specks, butlarge enough to compete with each other and any sensitivity specks presentfor any latent-image silver subsequently formed.98Many of the effects discussed so far are due tothe fact that a single quantum of light representing a small amount ofradiating energy is unable to make a grain developable. In many respectsthe photographic process is simplified when high-energy quanta, as of X-rays,or particles, as cc-rays, protons, or high-speed electrons, are used. Withthese, as a rule, no more than one quantum or particle is required to make a(e) Albert reversal.(8) High-energy quanta.93 J. E. Nafe and G. E. M . Jauncey, Physical Rev., 1940, 57, 1048.94 Albert, Arch. wiss. Phot., 1899, 1, 285.O 5 W. Clark, Phot. J., 1923, 63, 30; 1924, 64,91.98 G. W. W. Stevens and R. G. W, Norrish, ibid., 1938, 78, 613.IbicE., 1939, 79, 21. 97 A. Marriage, ibid., 1937, 77, 619llonanbeq(A ‘a ‘9) paonpax : amqwadmez‘1 ‘Iwnsodxa pu68 GENERAL AND PHYSICAL CHHIMISTRY.grain developable ; certain fast particles will even cause a string of grains to bemade developable. Exposures to X-rays have been studied very extensively,and, in confirmation of our conceptions on latent-image formation, it isfound that there is no reciprocity failure,g9 that therefore time- and intensity-scale curves can be used interchangeably, and that the characteristiccurves are of identical shape for all wave-lengths if referred to the numberof quanta absorbed.lW A density-exposure time or intensity curve isstraight at the origin through which it passes. Solarisation is fo,und withX-rays. 101(6) Outstanding &zcestions.-On the whole, the theory of the photographicprocess is reasonably coherent, but attention may be directed to a fewfeatures not sufficiently understood. Some of these difficulties have alreadybeen mentioned in the body of the Report. What are the basic reasons forthe big differences in photographic sensitivity between silver bromide,chloride, and iodide emulsions? What are the reasons which so far haveenabled gelatine emulsions to be made much more sensitive than those in anyother binding medium? What is the mechanism of desensitisation bycertain dyes, and of the reversal obtained, if an exposed material is desensitisedby a dye and exposed still further ? 11* lo2* lo3 What is the explanation of theWeigert effects,l04 the finding that polarised red light is capable of making aprint-out density dichroic with certain emulsions, and that a latent imageproduced by polarised light leads to a dichroic image on processing? Themechanism of development still requires a great deal of attention. Futurework will be followed with interest, not only by photographic technicians butby anybody concerned with the questions of catalysis and heterogeneousreactions, reactions involving solids, and photochemistry.As reviews should be noted :W. Meidinger, in Hay’s “ Handbuch der Photographie,” Vol. V, Springer,Wien, 1932.C. E. K. Mees, “ The Theory of the Photographic Process,” Macmillan,New York, 1942.N. F. Mott and R. W. Gurney, “Electronic Processes in Ionic Crystals,’’Oxford, Clarendon Press, 1940. (For a discussion of the fundamentalprocesses.)(Very full bibliography up to early 1932.)(Very comprehensive review.)99 W. Friedrich end P. P. Kech, Ann. Ptqsik, 1914, 45, 399.100 R. Glocker and W. Traub, Physikal. Z., 1921, 22, 345.lol K. Schaum, Arch. wise. Phot., 1904, 18, 77.102 J. Waterhouse, Proc. Roy. Soc., 1875, 29, 186.1-03 H. Luppo-Cramer, Eder’s “ Handbuch der photog rep hi^," Vol. 2, Part I (1923),104 See H. Freundlich, Phot. J., 1936, 76, 395.105 Phototechnique, Feb., 1940, p. 48.106 H. Luppo-Cramer, Phot. Korr., 1939, 75, 49.107 F. Weigert and J. Matulis, Koll-chem. Beihefte, 1933, 38, 384.109 J. Sterry, Phot. J., 1904, 44, 50.10) H. Liippo-Cramer, Phot. Ind., 1940, 38, 271.110 F. E. Ross, “The Phytics of the developed Photographic Image,” D. vanVol. 3, Pert I11 (1933), Wien.Nostrand, 1924, New YorkBERG : THE PHYSICAL CHEMISTRY OF LATENT-IMAGE FORMATION. 69S. E. Sheppard and C. E. K. Mees, “Investigations on the Theory of thePhotographic Process,” Longmans & Co., London, 1907. (For the basicstudies on development .)W. F. Berg, ‘‘Latent Image l?ormationyYy Trans. Farday Soc. (in thepress). W. E”. B.C. E. H. BAWN.W. I?. BERG.G. GEE.H. W. MELVILLE

ISSN:0365-6217    

DOI:10.1039/AR9423900007

出版商:RSC    

年代:1942

数据来源: RSC

摘要:

INORGANIC CHEMISTRY.1. GENERAL.THE first section of this Report deals with a wide range of publications whichhave appeared or become available during the past year. Topics forinclusion have been selected on the basis of their general interest rather thanof their value to the specialist : the arrangement follows the order of thegroups of the Periodic Table. The second and the third section of theReport are devoted to the Luminescence of Inorganic Solids and the Techniqueof Inorganic Chemistry, respectively. In discussing the first of these topicsit has been necessary to deal in an elementary manner with the underlyingphysical principles, but as far as possible emphasis has been placed on thoseaspects of the subject which are of interest in inorganic chemistry. The finalsection, dealing with technique, needs little introduction. The developmentof new technique or, more commonly, the application of physical and physico-chemical technique, has time and again paved the way for new andimportant developments.The summary given here, though far from com-plete, may, it is hoped, serve to stimulate further interest in the experimentalstudy of Inorganic Chemistry.Since the preparation of the last Annual Report there has been a sustainedinterest in the separation or enrichment of isotopes, and in applications ofenriched material in " tracer " experiments. Application of the thermaldiffusion method to the treatment of methane, for example, has given yieldsup to 0-3 g./48 hr. of methane in which the I3C : I2C isotope ratio is increasedto 11.5 times the normal value.1 The neon isotopes have been obtained in apure state by the same method.2 In this work a separation tube 29 m.longwas used. The isotopes 20Ne and 22Ne were obtained in quantity with apurity of 99.7-99.8~o , and a fraction was isolated in which the concentrationof the rare isotope 21Ne was enriched from the normal value of 0.27% to26y0. The same technique has been applied for separating from normalkrypton a mixture containing 22% of and 78% of 86Kr.3The efficiency of the electrolytic separation of the chlorine isotopes at aplatinum anode has been in~estigated.~ The electrolytic separation factorfor these isotopes in the range 25-40' was found to be 1.0060 & 0.0005, avalue which is about the same as that on Acheson graphite.The separationfactor for chlorine is about the same as that for the oxygen isotopes and isconsiderably less than that for the hydrogen isotopes. An interestingfeature of this work was the use of measurements of the temperature offlotation in ethylene dibromide of annealed sodium chloride crystals as aA. 0. Nier and J. Bardeen, J. Chem. Physics, 1941, 9, 690.2 G. Diekel and K. Clusius, 2. physikal. Chem., 1940, B, 48, 60.Idem, Naturwias., 1940, 28, 711.4 H. L. Johnstm and D. A. Hutchison, J . Chem. Physics, 1942, 10, 469EMEL~~US : URINECBAL. 71means of determining the isotopic composition of samples. Radioactiveisotopes of chlorine and bromine have been used as tracers in studying theexchange in carbon tetrachloride solution between the free halogens and thephosphorus tri- and penta-halides. In each case exchange was complete inless than 3 minutes, and the experiments were taken as an indication of theequivalence of the halogen atoms in the phosphorus halides.6The exchange of radioactive silver in solution with silver ions in silverchloride suspensions has been shown to occur.6 It is not confined to thecrystal surface, but is propagated into its bulk, probably by a process ofself-diffusion.It was found that adsorption of fluorescein, eosin, methylene-blue, or wood violet on the silver halide grains had very little, if any, influenceon the rate of attainment of equilibrium. The exchange was reversible in thesense that, when the precipitate contained the radioactive silver initially andthe solution was inactive, the activity of the solution increased up to a valuecorresponding with a homogeneous distribution of the radioactive silver.Sodium monothio-orthophosphate, Na,PO,S, has been prepared byheating together at 450-750" in a vacuum or an inert atmosphere the od-culated quantities of so$um metaphosphate and sodium sulphide.7 Theproduct was 70--87% pure, depending on the reaction temperature, but thepure hydrated compound Na,P03S,12H20, was readily isolated by re-crystallisation from warm water.This method of preparation is moreconvenient than those depending on the decomposition of thiophosphorylchloride with alkali or the fractional hydrolysis of phosphorua pentasulphidewith alkali.The compound, like trisodium orthophosphate, exists in twomodifications, the transition temperature being approximately 550'. Thesame authors have also shown * that when sodium phosphate is heated a ttemperatures up to 1100" with silica or sodium silicate, there is no evidencethat the sodium silicophoaphate, Na,[Si( PO,),], described by R. Schwarz:is formed.A lower oxide of boron with the empirical formula 330 is formed whenelementary boron ie heated with zirconium dioxide in a vacuum at 1800°.10The reactants were heated together in a tantalum boat in a high-frequencyfurnace. The boron oxide was volatile at the reaction temperature, andcondensed a8 a light brown, amorphous solid, which was obtained in sufficientquantities for analysis.The compound obtained by this thermal reductionmethod is almost certainly different from the oxide B,O, obtained by R. C.Ray and P. C. Sinha l1 from the product of hydrolysis of magnesium boride.Repeated extraction of magnesium boride with water yields a residue whichreacts with aqueous ammonia forming two salts, (NHJ ,B 2( OH) and(NH,) 2B406, which were separated by fractional crystallisation. These twoW. Koskoski and R. D. Fowler, J . Amer. Chm. Soc., 1942, 64, 860.A. hnger, J. C h . Phy8&~,2942,10,321.E. Zintl and W. Morawietz, 2. anorg. Chew&., 1940, 245, 16.Ibid., p. 12.lo E. Zintl, W. Morawietz, snd E. Gastinger, ibid., 1940, 24.6, 8.l1 J., 1941, 742.Ibid., 1928, 176, 23672 INORGANIC CHEMISTRY.salts when heated lost, in the first case, ammonia and hydrogen in equalvolumes, and in the second, ammonia : the solid residues had the compositionsB202 and B,O,, respectively.Both of these oxides were soluble in water andtheir solutions had reducing properties.A volatile gallium hydride, Ga2H6, has now been described.12 The firststage in the preparation of this important compound is the synthesis oftetramethyldigallane, Ga 2H2( CH,),, by passing a mixture of methylgalliumand hydrogen through a glow discharge. The methylgallane, which has anextrapolated b. p. of 172", decomposes a t 130" into methylgallium, gallium,and hydrogen. The gallium hydride is formed by allowing the methyl-gallane to react with triethylamine at a lower temperature. The followingreaction occurs :3Ga2H211\le, + 4NEt3 = Ga2H, + 4GaMe3*NEt3The molecular weight of the hydride corresponded with the formula Ga2H6.The b.p. (extrapolated) and m. p. were 139" and -21.4", respectively. Itunderwent decomposition into its elements at 130".A number of indium salts have been described by F. Ensslin and H.Dreyer.13 These include the halides of InIII, the solubilities of which inwater, expressed as g./100 g. of solution, are : InF3 7-83, InC1, 64.7, InBr,84-27, Id3 91-6. Among a number of complex salts described, indiumdiethylthiocarbamate, (NEt ,CSJ31n, and indium hexamminocobalticchloride, I~[CO(NH,)~]CI,, are insoluble in acid solution and may be usedfor the determination of indium.Amalgams of samarium, europium, and ytterbium have been described byJ.K. Marsh and H. N. McCoy.15 Marsh prepared them by shaking the rareearths in acid acetate solution with dilute sodium amalgam. It was foundthat whereas samarium, europium, and ytterbium form amalgams, thecerium-group elements all show a lower power of amalgam formation, whichdecreases in order of atomic number. I n the course of the experiments witheuropium amalgam evidence was obtained for the existence in solution ofeuropous hydroxide with a solubility of at least 1 g. /l. The use of the relativeease of amalgam formation exhibited by the rare-earth elements in effectingtheir separation was also studied.16 Samarium could be extracted byamalgam formation from neighbouring elements, particularly neodymiumand gadolinium, and fractional decomposition of the samarium amalgamgave a further means of purification, since this amalgam is relatively readilyattacked by water or dilute acids.McCoy, who was concerned primarilywith the preparation of amalgams, found that aqueous solutions of theacetates of europium and ytterbium in potassium citrate gave amalgams onelectrolysis or on treatment with potassium amalgam. The solid amalgamsHg,Eu and Hg,Eu, were prepared.Further important publications on carbonyl chemistry by W. Hieber andl2 E. Wiberg and T. Johannsen, Angew. Chem., 1942, S6, 38.lS 2. anorg. Chem., 1942, 249,119.l6 J . Amer. Chem. SOC., 1941, 63, 1622.1 4 J., 1942, 398.10 J. K. Marsh, J., 1942, 623EMEL~US : GENERAL. 73his co-workers have appeared during the period under review.17 Hieber hashimself published an important review article.ls The preparation ofrhenium carbonyl halides has been described by W.Hieber, R. Schuh, andH. E"~chs.1~ The iodo-compound, Re(CO),I, is formed when carbon mon-oxide at 1 atm. is passed over a mixture of potassium hexaiodorhenate,K,Re16, and copper powder heated to 200". The bromo-compound isformed analogously at 200-300" and with a carbon monoxide pressure of atleast 10 atm. To form rhenium carbonyl chloride a still higher pressure isneeded. These compounds are also formed in good yield by the reaction ofcarbon monoxide on metallic rhenium mixed with a dissociable halide,e.g. (at 210 atm./25Oo)Re + Cu(Hal), + 6CO = Re(CO),(Hal) + Cu(Ha1)COThe carbonyl halides of rhenium diminish in volatility from the iodo- to thechloro-compound, and this is the order of increasing stability, since theiodo- is readily converted into the bromo-compound, or this into fhe chloro-compound, by the action of the appropriate free halogen.These carbonyl halides are all very stable, and attempts to remove thehalogen by such obvious means as the action of metallic silver were unsuccess-ful.It was also found impossible to prepare rhenium carbonyls directly fromrhenium and carbon monoxide a t high pressure. Rhenium pentacarbonylwas, however, obtained in good yield by the action of carbon monoxide at200 atm. on the heptoxide, Re,O,, at 250".20 The carbonyl is a colourlesscrystalline substance which may be sublimed, and is soluble in organicsolvents.It is also formed directly from potassium per-rhenate.Cryoscopic measurements show rhenium carbonyl to be dimeric, thoughthe carbonyl halides of the element are monomeric. The carbonyl isdecomposed completely at 400". Some indications were obtained of theexistence of a carbonyl hydride of rhenium, but such a compound has notyet been isolated in a state of purity. Reaction of rhenium pentacarbonylwith amines was found to be comparable with that of the hexacarbonyls ofthe chromium group. In general, replacement of part of the carbon mon-oxide by co-ordinating amine groups occurs; e.g., pyridine reacts at 240"according to the equation[Re(CO),], + 4Pyr = 2Re(CO),Pyr, + 4COWith o-phenanthroline reaction occurs more readily and two molecules ofcarbon monoxide are replaced by one of the base.A similar replacementof part of the carbon monoxide occurs with the rhenium carbonyl halides,giving, e.g. ReCl(CO),Pyr, and ReCl(CO),Phen. R. Schuh 21 has publishedmeasurements of the extinction coefficients of rhenium carbonyl and of thecarbonyl halides in solution.Comparative data have been published on the formation of carbonyls ofiron, cobalt, and nickel from their respective halides by the action of carbon1' Cf. Ann. Reports, 1941, 38, 71.l8 Angew. Chem., 1942,65,7-11,24-28.2o W. Hieber and H. Fuchs, ibid., p. 266.l9 Z. anorg. Chem., 1941, 248,243,a l Ibid., p. 27674 INORGANIC CHEBISTRY.monoxide a t high pressure.22 T t is considered that in the synthesis ofcarbonyls by this method the ease of formation of the carbonyl increaseswith the ease of splitting off halogen from the metal, i.e., in the orderI > Br > C1. Intermediate formation of carbonyl halides also occurs, afact which explains the catalytic influence of iodine in the direct synthesis ofcarbonyle fkom metals and carbon monoxide.The r61e of copper or silverin these reactions is that of a halogen acceptor. Attempts to prepare ironcarbonyl hydride by high-pressure synthesis gave only negative results.=Important new mixed carbonyls containing cobalt have now beenproduced by the high-pressure method.% They are formed when eithercobalt or an anhydrous cobalt halide, mixed with the metal in question, istreated a t 200" with carbon monoxide at a pressure of 200 atm.Thecompounds so far prepared are of the following types : M[Co(CO),12, with M= Zn, Cd, Hg, Sn; M[Co(CO),],, withM = In, T1; M[Co(CO),], with M = T1.Negative results were obtained with lithium, beryllium, magnesium, calcium,silver, gold, germanium, antimony, and bismuth. With lead, there wassome indication of the formation of a mixed carbonyl.These compounds are all volatile solids, soluble in indifferent organicsolvents. The two thallium derivatives may be produced at will byemploying the appropriate proportion of thallium in relation to the cobalt orcobalt halide. The mixed carbonyls are more stable than the dimeric cobalttetracarbonyl. All are, however, decomposed by dilute acids to form cobaltcarbonyl hydride and a salt of the second metallic component.Theconstitution proposed is illustrated in the case of the mercury and thallousderivatives by the formulae (I) and (11).Organometallic derivatives of iron carbonyls form yet another importantnew group of compounds described during the pmt year. The resotion ofiron pentacarbonyl and iron carbonyl hydride with various organometalliccompounds was first studied by F. Hein and H. P ~ b l o t h . ~ ~ These experi-ments, which were made with derivatives of mercury, bismuth, thallium,and lead, gave for the most part very indecisive results, probably because ofthe instability of the products sought, A lead derivative of the formula( C2H5) ,PbFe( CO), was, however, obtained by the interaction of triethyl-leadhydroxide and the calcium derivative of iron carbonyl hydride in presence ofether.The lead compound wm a red crystalline substance which wassoluble in organic solvents, was readily oxidised, and was decomposed above130". Subsequently the corresponding methylmercury iron tetracarbonyl,(CH,Hg) 2Fe( C0),,26 was described. It was obtained by decomposingaqueous methylmercury hydroxide covered with a layer of light petroleumwith a solution of the cal&m derivative of iron carbonyl hydride. The2a W. Hieber, H. Behrens, andU. Teller, 2. anorg. Chem., 1942,249.26.23 W. Hieber and U. Teller, aid., p. 68.*6 Ibid., 1941, $348.84.a4 Idem, &id., p. 43.96 Idem, &id., 1942, W, 293EMEL~US : GEXERAL. 75compound was insoluble in water but soluble in a variety of organic solvents.It could be sublimed in a vacuum but decomposed above 100" into dimethyl-mercury and mercury iron carbonyl.Evidence bearing on the supposed tautomerism of hydrogen cyanide hasbeen obtained in an extension of earlier work on the synthesis of thiocyanatesby the reaction of sulphur with organic cyanide^.^' It was argued that ifhydrogen cyanide contained the tautomeric hydrogen isocyanide the lattershould react with sulphur to yield isothiocyanic acid, HNCS, which wouldrearrange to form HSCN.This, in turn, should form the known degradationpolymer H,C,N,S3. Sealed-tube experiments are recorded in which thereaction of liquid hydrogen cyanide with sulphur, selenium, and telluriumwas studied, but they did not substantiate this reasoning.In the courseof the work it was shown that the interaction of hydrogen cyanide with amixture of pyridine or quinoline and sulphur or selenium yielded thio-cyanates and selenocyanates of these bases. No tellurocyanates wereisolated.In a study of the cyanogen halides, P. Kailasam 28 found that bromineand dry cyanogen chloride react a t room temperature to form a crystallinemolecular compound, 3C3N,C1,C3N3Br. Cyanuric chloride reacts with 47 ?&hydrogen bromide at room temperature, forming the compound C3N3C1,Br,which is soluble in organic solvents and is decomposed by water a t 120°,giving cyanuric acid.Silicon monoxide, the existence of which has hitherto been uncertah ,has now been rein~estigated.~~ The oxide is best produced by heating silica,or silicates with elementary silicon in a, vacuum at 1450".It is volatileunder these conditions and, indeed, absorption bands associated with thevltpour of SiQ produced by reduction of silica with carbon a t above 1500"I Y ~ C observed by N. F. Bonhoeffer in 1928.30 Condensation of the vapourmay give a finely-divided, yellow material, which so far has not yielded acharacteristic X-ray diagram, or a glassy, brown solid. In either caseanalysis indicates the empirical formula SiO, but it is possible that a certainamount of disproportionation in the sense of the equation 2SiQ + SiQ, -+Mi takes place. The finely-divided, yellow condensate is pyrophoric. It,reacts with water vapour a t 500", forming silica and hydrogen, and is capableof reducing metallic oxides.An investigation of the iodides of silicon31 has shown that purifiedsilicon and iodine vapour react at 650", forming the solid tetraiodide.Whenthe latter is heated a t 280" with a twofold excess of finely divided silver,Si,16 and AgI result, but no indication is obtained of the existence of higheriodides. Disilicon hexaiodide when heated to 350-400" disproportionates27 C. R. McCrosky, F. W. Bergstrom, and G. Waitkins, J . Amer. Cliem. Xoc., 1940,62,2031 ; 1942,64,722.PTOC. Indian Acad. Sci., 1941,14 A , 165.2D E. Zintl, W. Briiuning, H. L. Grube, W. Krings, and W. Morawietz, 2. anorg. Chenz.,1940, 245, 1.2. physikal. C'henb., 131, 363.3 1 R. SSchwarz and A. Pfliigmacher, Ber., 1942, '75, 1062.c 76 INORGANIC CHEMISTRY.into silicon tetraiodide and a polymeric monoiodide (SiI),, which appears tobe similar to a monochloride described by R.Schwarz and U. Greg0r.3~ It isbelieved to be a chain-like polymer with alternate single and double bondsbetween the silicon atoms. Hydrolysis of the monoiodide yields an ivory-coloured polymer which has the empirical formula H4Si203 and is veryreadily oxidised. Thermal degradation of the latter in a vacuum leads toloss of hydrogen and the successive formation of the polymers (H3Si2O3),and (H2Si203),. The final degradation product is a mixture of silica andsilicon.H. H. Anderson, in an extension of earlier work 33 in which cyanates andisocyanates of silicon, phosphorus, and boron were prepared by the action ofsilver isocyanate on the corresponding halides, has now described analogousderivatives of phosphorus oxychloride, and arsenic and antimony trichlorides.In addition, the corresponding thio-compounds PO( SCN), and As( SCN),were isolated.34 The isocyanates PO(NCO),, As(NCO), and Sb( NCO), werefound to be volatile, but they isomerised when heated, forming the corres-ponding non-volatile cyanates.An ammonia addition product of nitrogen tribromide, NBr3,6NH,, hasbeen described by M.S~hrneisser.3~ It is best prepared by mixing a stream ofnibrogen carrying bromine vapour with an excess of ammonia in a vessel atloo", in which ammonium bromide is deposited. The issuing gas stream,when passed through a trap cooled to -95", deposits the red compoundNBr,,6NH3.It was impossible to prepare nitrogen tribromide from theammoniate because the latter exploded very readily. It was, however,shown that the ammoniate was soluble in methyl or ethyl alcohol, ether, orliquid ammonia, and that it could be recovered unchanged from the lastsolution.The chemistry of quinquevalent bismuth compounds has been investi-gated by various authors in recent years. Well-authenticated derivativesof this valency state are comparatively few. Both potassium bismuthoxyfluoride 36 and bismuth pentafluoride 37 are of this type, and the recentlyprepared sodium orthobismuthate 38 is yet another example. The ortho-bismuthate is prepared by heating bismuth trioxide and sodium monoxidein air or oxygen at 650" :Bi,03 + 3Na20 + 0, = 2Na3Bi04The product obtained with those proportions is free from sodium peroxide.The alternative method of oxidising bismuth salts to bismuthates in alkalinesolution has been reviewed critically by R.Scholder and H. Stobbe,39 whopoint out the great difficulty of obtaining all of the bismuth in the quinque-valent condition.The preparation of thionylimide, SONH, by the interaction of thionyl33 2. anorg. Chem., 1939, 241, 1.84 Idem, ibid., 1942, 64, 1767.9 6 0. Ruff', ibid., 1908, 57, 220.37 H. von Wartenberg, ibid., 1940, 244, 337.a8 E. Zintl and K. Scheiner, ibid., 1940, 245, 32.93 J . Amer. Chem. SOC., 1940, 62, 761.3 5 2. anorg. Chern., 1941, 246,284.Ibid., 1941, 247, 392EMELkTTS : GENERAL.77chloride and ammonia has been described by P. W. Schenk.40 Previously,F. Ephraim and H. Piotrowski,*l by dropping thionyl chloride on liquidammonia, had obtained a reaction which was represented by the equationA silver salt NAg(SO*NH,), was isolated, and the same salt was obtainedfrom the products of the interaction of sulphur dioxide and ammonja.Ephraim's work was repeated because it is known42 that primary amineareact with thionyl chloride according to the equation :SOCl, + NH,R = SONR + 2HC1It was considered that a comparable reaction would occur with ammoniaprovided it were not present in excess.The reaction was studied in the gas phase, the two reactants in stoicheio-metric proportion being allowed to mix and the products being then cooledas rapidly as possible to the temperature of liquid air.Under these condi-tions the product was a -colourless liquid, the anaJysis and molecular weightof which corresponded with the formula SONH. This compound melted at-85" and polymerised a t temperatures above -70" to a yellow-brown,glassy resin.Chromium tetrachloride has been prepared as an unstable brown solid bypassing a stream of chlorine over the trichloride a t 700" and cooling theissuing gas stream rapidly with a thimble containing solid carbon dioxide.The brown solid which condensed was analysed and had a compositionapproximating to CrC14. It has strong oxidising properties and was stableonly below -80°.43Chromic and cobaltic amides, Cr(NH,), and Co(NH,),, are formed in thereaction in liquid ammonia solution between potassium amide and thehexammine nitrates of the respective elements.44 Both the compounds arebelieved to be polymers.They react with ammonium salts in liquidammonia solution, forming salts with a polymeric cation, of which thefollowing were isolated :2SOC1, + 7NH3 = 4NH4C1 + NH(SO*NH,),IBr(NH,)Cr(NH3)3InBrn "O3(NH,)Cr(NH,),],(NO,),,[ WH,) zCO(NH3) 2In(NO&n INO3(NH,)CoNH&Both amides react with excess of potassium amide in liquid ammonia,thereby exhibiting their amphoteric character. The following amphotericamide derivatives were isolated :Gaseous chlorine dioxide and fluorine have been shown to underge avigorous but non-explosive reaction when they are mixed a t 0" and at lowpartial pressures in presence of a large excess of a diluent Thecompound C10,F was produced (m.p. - 115"; b. p. - 6"). Though theI0 Ber., 1942, 76, 94.42 A. Michaelis and 0. Storbeck, Annalen, 1893, 274, 187.H. von Wartenberg, 2. anorg. Chem., 1942, 250, 122.44 0. Schmitz-Dumont, J. Pilzecker, and H. F. Piepenbrink, ibid., 1941, 248, 175.45 H. Schmitz and H. J. Schumacher, ibid., 1942, 249, 238.I1 Ber., 1911, 44, 37978 INomANlc OHEMISTRY.reaction with water has not so far been eluoidated, C10,B may be regarded asa fluoride of chloric acid. It is known to be very sensitive to reaction withwater vapour, and is very much less liable to explode spontaneously than ischlorine dioxide. H. J. E.2. LTJMINESUENUE OF INORGANIC So~ms.The study of the phenomena of Auormnce and phosphorescence of in-organic solids has become so extensive that my attempt to review the fieldas a whole would be impossible here, but the increasing industrial applicationsand the advances in the theoretical interpretations of these phenomenawarrant a report of the recent work on selected aspects of it.The workbefore 1928, carried out mainly on alkaline-earth phosphors by Lenard andhis collaborators, has been recorded comprehemively,l and the luminescenceof incandescent solids was the subject of a report from the Carnegie Instituteof Washington in the same year.2 Neither of these types of phosphor will bereviewed here, as the more important advances in recent years have beenmade with other luminescent solids. Various aspects of the field werediscussed by the Faraday Society in 1937.3Luminescent inorganic solids can be broadly classified according as towhether the emission takes the form of (i) discrete or line spectra, or (ii)continuous spectra, where the broad bands persist as bands even a t lowtemperatures. F.R. Spedding bas recommended low-temperature investig-ation as a method of establishing the class. For the purposes of this Report,a chemical classiiication will be employed, broadly separating these materialsinto pure and impurity-activated compounds, and specifically classifyingthem according to chemical constitution. It is by no means possible toapply any theoretical interpretation applicable to one class of phosphors toanother class, and each case must be considered individually.Nechanism of Luminescence of Solids.-As far as has been investigated,luminescent properties appear to be extremely dependent on structure, andstrong luminescent properties are associated with defects in the crystal lattice.The more ideal crystals investigated do not appear to luminesce to the sameextent as those containing more lattice defects.H. W. Leverenz and F.Seitz consider it possible that truly ideal crystals are not luminescent underultra-violet or cathode-ray excitation. Pure crystalline compounds are ingeneral far from ideal and the defects may arise in several ways : (i) Ions ofeither type may be absent from the lattice. These vacancies are caused bydiffusion into the body of the lattice of holes present in the surface of thecrystal.Heating these crystals causes a considerable increase in the numberof these vacancies. (ii) In other lattices, particularly where the structure isopen, excess metal atoms or metal-impurity atoms may occupy interstitialpositions. (iii) In others, ions may diffuse from normal lattice positions into“ Handbuch der Experimentalphyaik,” 1928, 23.General Discussion on Lumhescenoe, 1939.Trans. E’amday Soc., 1939, %,66.a E. L. Nichols, H. L. Howes, andD. T. Willsins, Publ. No. 384,1928.ii J . Appt. Phpka, 1939, 10, 479REES : LUMIXESCENUE OF INORGANIC SOLIDS. 79interstitial positions, leaving vacancies. These crystals then have bothtypm of defect present in the lattioe. It can be seen that impurity atomsmay take up positions in any of these lattices, provided they be not preventedfrom doing 80 by their size.According to the modern theory of solids,s allowed energies of atoms andions are not single energy levels but are energy ranges or bands.Thismodifioation of the ionic levels is due to the presence of the crystal lattice,and it kj possible to trace back the bands to the energy levels of the free atomor ion. In ideal insulating crystals these bands are separated by forbiddenregions, the lower band of levelabeing Wed and the upper bandempty. The implications of this,vix., that the lowest-energy ultra-violet absorption band shouldcorrespond to excihtion of elec-trons from the filled band to theempty band, and the oonsequentphotoconduotivity due to ionh-ation, were not fulfilled by experi-ment.In the alkali halides,7for htanoe, the lowest-energyspectral absorption bands werenarrow, and photoconductivitydid not oocur on absorption.These bands oorrespond to 80-called ‘( excitation levels,” justbelow the d e d ionisation band.These levels have been asoribedto the energies of electrons whichPote%tkE-emrgy cumea for r a m l afid excitedstates in a crystal laztice.*have remained asaociated with their positive holes due to Coulomb a,ttraotionafter exoitation.Imperfect pure crystals urnally have extra, levels (“ centre ” levels) byvirtue of their lattice defeats. In certain alkali halide phosphors theseextra levels give rise to the F- or colour bands in the visible spectral region,these being assooiated with a stoichaiometrio excess of alkali metd in thelattice. Lattice defects caused by the presence of impurity atoms in inter-stitial positions ah0 result in similar energy levels between the lower filledand the upper empty energyThe luminescent prooess hw been illustrated by use of diagramsrepresenting the energy of the lattice m a function of the atomic displacementfrom equilibrium positions? The transition AA‘, corresponding to ctbsorp-F.Seitz, “ Modern Theory of Solids,” McGraw-Hill, 1940.R. Hilsch and R. PohI, 2. Phpih, 1928, 48,384.R. W. Gurney rand N. F. Mott, Trans. Faraday Soc., 1939,35,69.* N. Riehl, Ann. Phyeik, 1937, 20,663.* Reproduoed by permission from the Tran8cactio?w of the Faraday Society80 INORGANIC CHEMISTRY,tion, raises the lattice to an upper vibrational level of the higher electronicstate.Apart from resonance radiation by the transition A’A, the vibrationalenergy of the excited state is dissipated without optical emission, theequilibirum position $’ being attained. The transition B’B may then takeplace with the emission of light of lower energy, and therefore longer wave-length, than the exciting radiation. Excitation of phosphors in theabsorption bands due to the ‘‘ centre ” levels 5 is always more efficient in theproduction of luminescence than excitation in the strong fundamental band ;furthermore, excitation in either band causes the same luminescent emission.The explanation of this has been given by F. Seitz.lO Excitation at animpurity or lattice defect centre induces luminescence of the type discussedabove, whereas excitation in the fundamental band induces photoconduc-tivity, as the electrons have been excited into the ionisation bands.Theseelectrons transfer their energy to the impurity or defect centres, after whichradiation is emitted according to the above mechanism. Another conceptwhich is necessary for the comprehension of certain aspects of luminescenceis that of trapping levels, L e a , metastable states in which energy may betrapped for some time before thermal agitation brings the system into acondition for luminescent emission again. If the functions representing theenergy levels approach closely at some point C, then a probable mechanismfor the dissipation of energy is the thermal excitation to a vibrational level atC‘, a t which point the transition to the lower level and immediate dissipationof the $excess vibrational energy now possessed brings the lattice to theequilibrium state without optical radiation.Such a mechanism accountsfor the quantum efficiencies of luminescence less than unity and also fordecrease of luminescence at temperatures above normal.Certain luminescent materials show an after-glow lasting for minutes,hours, or even days in certain cases and microseconds in others. Thediatinction between the phenomenon of fluorescence and that of phosphores-cence is quite arbitrary in practice, as any fluorescent emission must takeplace at a finite time after the excitation process.P. Pringsheim l1 hasproposed that the term JEuoreScence be reserved for processes of emissioninvolving inner transition probabilities, whilst the term phosphorescenceshould be applied to processes in which metastable states are involved. Thedistinction between these processes can be made by considering the decaylaw which the after-glow follows, and the temperature-dependence of theprocess. This involves previous knowledge of the decay process, but isnevertheless the only logical way of defining the terms.In the previous section we have seen that absorption of energy inimpurity-activated phosphors may result in the excitation of electrons to theconduction band or to levels in the impurity centres without ionisation. Ineither process the subsequent emission is due to the transition from anexcited un-ionised state to the ground state in the impurity centre.Theprocess is, of course, accompanied by a loss of energy by thermal vibrationsin the lattice. In the case of phosphors not exhibiting photo-conductivitylo Tmns. Paraday SOC., 1939, 35, 74. 11 Ibid., p. 87REES : LUMINESCENCE OF INORGANIC SOLIDS. 81the decay law is simply I = I0e-=t, where I is the intensity of the phos-phorescence a t a time t after the removal of the excitation, when the intensitywas I , ; a is the decay constant, the order of which depends on the probabilityof the transition from the excited to the ground state. It is expected thata should have a small temperature dependence. In the case where photo-conductivity occurs, and even in certain cases of excitation to impuritylevels, however, the electron must recombine with one of the centres whichhas lost an electron and a bimolecular type of law results, vix., I =.I,/( 1 + at)2 ; a here exhibits a strong temperature dependence. These twolaws represent the simplest cases of the decay process, and most experi-mental determinations reveal a superposition or modification of theseprocesses.A.L. Reimann l2 found experimentally that the photoconductivity ofzinc sulphide phosphors a t - 196" decays slowly under the influence of infra-red radiation and after the excitation has been removed. The law governingthis decay is o = ao/(l + at), where Q is the conductivity after a time t andoo that a t t = 0.This is in agreement with the bimolecular decay law, as theconductivity is proportional to the number of free electrons.Preparative Methods.-In all preparations of luminescent materials theutmost purity is essential, and in the case of impurity-activated materialssmall and controlled amounts of particular elements must be added. Thereasons for this will be apparent in later discussion, as the presence ofcertain foreign ions causes marked diminution in the efficiency of theluminescence, even to the extent of its complete suppression. Thecharacteristics of the luminescence are strongly dependent on the activatorand its concentration. Furthermore, the emission characteristics aregoverned considerably by the conditions of preparation.In certain caseswhere the crystal exists in more than one modification the temperature ofpreparation and the conditions of quenching are important.In the subsequent matter, attention will be directed to the moreimportant aspects of the preparation and the spectral characteristics of themembers of the various classes of phosphors.Alkali Halide Phosphors.-Alkali halide crystals which are ideal do notluminesce to any extent, but when lattice defects are introduced, luminescencecan be induced by irradiation in certain ultra-violet regions. Alkali halidecrystals containing stoicheiometric excess of either alkali metal or halogenexhibit new absorption bands in the visible (F-bands) and the ultra-violet,respectively. l3 Irradiation in the F-bands is claimed to cause fluorescenceand phosphorescence.Sodium chloride containing F-centres, when irradi-ated with A4700 (the maximum of the F-band), emits fluorescent andphosphorescent light of the same spectral distribution as the F-band.14However, H. W. Leverenz and F. Seitz consider that the results areinconclusive. Investigations into those alkali halides in which thallium,bismuth, silver, copper, or lead have been incorporated into the vacancies in1* Nature, 1937, 140, 501.1 4 See E. Hirschlaff, " Fluorescence and Phosphorescence," Methum, 1938.l3 See R. Pohl, Phy&lcaE. Z., 1938, 59, 3682 INORQANIO CHEMISTRYthe alkali-metal ionic lattice, have been carried out mainly by R. Pohl,R. Hilsch, and their co-workers.15 Activators are added as chlorides ornitrates of the impurity element to the alkali halide to the extent of 10-3 orThey may be added to the melt or electrolytically diffusedinto the lattice at temperatures above 350".The presence of the activatorgives rise t o new absorption bands on the long-wave-length side of thecharacteristic absorption of the pure alkali halide.16 By far the mostsatisfactory case from experimental and theoretical points of view is that ofthe thallium-activated series, as here it has been conclusively shown thatthe thallium is monatomically dispersed throughout the lattice,l6 whereasthose activated by lead or dver, for instance, appear to occur in the halidelattice as complex molecular groups. The mechanism of the absorption andemission processes has been satisfactorily established by 3'.Seitz l7 for theformer case. The three absorption peaks due to the impurity atoms aresituated at A 1990, A 2160, and A 2550 and are invariant with the alkalihalide used. The absorption peaks have been associated with the electronictransitions lXo -+= 3P1, lh'o -+ 3P2, and %So -+ lP1 in the free thalliumions, modified by the perturbations introduced by the lattice field. Tran-sitions of this type preclude ionisation and hence photoconductivity, acondition which is supported by experiment. The two emission bands in thenear ultra-violet and visible in KC1-TI, for instance, are observed in allcases and appear to be due to the transitions lB0 + 3P1 and ISo +- 3P0from the lowest vibrational levels of 3P, and 3P0 to higher vibrational levelsof 18,.Excitation to the lPl level by absorption in the band at 1 1990 causesfluorescent emission of the same bands as excitation in either 3P1 or 3P2This is explained by Seitz in terms of a number of radiationless transitionsof the type described and illustrated in the figure. Phosphorescence,however, occurs only after the excitation + lPl, and the existence of ametastable state associated with thelPl band must be postulated. Vibrationalenergy misst be supplied before the thallium ion can jump to the triplet states,from which phosphorescent emission of the same nature as the fluorescentemission can occur. The metastable state occurs only when the absorbingthallium ion has another thallium ion among its nearest neighbaurs.Cathode-ray bombardment can also excite these phosphors to fluorescence,but they are unstable under such treatmenL5Zinc Xulphide and Related Phosphors.-Sulphides of zinc and cadmiumand their solid solutions have emission spectra which consist in general of one,or in some cases two broad bands in the visible region, the position of thebands being dependent on the activating impurity and to some extent on itsconcentration.These sulphides are liable to exhibit lattice defects in viewof the open type of lattice which they possess, foreign ions being capable ofhomogeneous penetration into the interstices of the lattice by diffusionmol. yo.l5 See R. Hilmh, Proc. Physical Soc., 1937, 49, (extra part) 40, for references t ol6 W.Koch, 2. Physib, 1929, 57,638.1' J. Chern. Physics, 1938,6,150; Trans. Faraday Soc., 1939,36,74.experimental workREES : LUMINESCENCE OF INORGANIC SOLIDS. 83processes a t elevated temperature&. N. Kiehl 1* has found that copper,silver, gold, and manganese could activate the lattice by taking up interstitialpositions, whereas larger atoms such as those of lead and bismuth, whichvannot penetrate the lattice, show no activation. Pure zinc sulphide may beactivated by heating to high temperatures, the zinc ions diffusing intointerstitial positions to produce lattice defects. The mobility of otheractivators such as copper appears to be much greater than that of the latticeions themselves, as copper can be incorporated by lattice diffusion, irrespw-tive of surface diffusion, at temperatures as low as 330°.19The preparative methods are merely means of facilitating these processes.Although the r61e of the flux employed is rather indefinite, particularly in thecase of the alksline-earth sulphides, its presence in the pure zinc mlphidehas been shown not to afiect the luminescence.20 A comparative investig-ation into the known preparative methods has been conducted by N.F.Zliirov,21 and W. H. Byler 22 has studied the effect of factors such as particlesize, time mid temperature of heating, and foreign ion content on theactivation by heat treatment, and has established a correlation with theluminescence. S. Makishima 23 directed attention to the quenching conditionsand the use of fluxes.The patent literature on preparative methods iuextensive. Those of interest are (i) explosion of powdered zinc and sulphurtogether with flux and activator in an atmosphere of nitrogen a t 300U0,24(ii) activation of commercial zinc sulphide by prolonged electron bombrtrd-~ n e n t , ~ ~ (iii) incorporation of the activator by cathode sputtering in a highvacuum .26The spectral absorption of zinc sulphide has been the subject of muchdiscussion. The fundamental band due to the zinc sulphide matrix appearsto have a maximum below A 2500 and a long-wave-length limit at h 3350, asmeasured by J. H. Gisolf2' and J. A. Kitchener.28 However, S e i t ~ , ~ ~quoting unpublished measurements by H. W. Leverenz, and F. A. Kroger 2Opoint out that activated zinc sulphide has its maximum excitation beyondthis limit a t h 3600, and that a less pronounced absorption is present in thisregion, ascribed by Seitz to the interstitial atom.The emission bands areuhwacteristic of the activator. With small activator concentrations, theband due to ZnS-Zn may appear together with that due to the foreignimpurity centres, unresolved in the case of ZnS-Ag, but noticeable as a'' hump " on the long-wave-length side:* and resolved in the case of ZnS-Cn.The appearance of the band depends on the activator concentration and onthe heat treatment in the preparation, for under certain conditions diffusionof foreign ions into interstitial positions may take plaoe without the move-'' Ann. Physik, 1937,29,640; Trans.Faraday Soc., 1939, %, 135.l9 E. Tiede and E. Weib, Ber., 1932, 65, 371.2o F. A. Krijger, Physim, 1940, 7, 1.21 J. Appl. Chem. U.S.S.R., 1934, 7, 343.23 J . SOC. Chem. Ind., Japan, 1938, 41, 202.26 B.P. 414,597.Ibid., p. 97.22 J . Amer. Chem. SOC., 1938, 60, 632.24 Fr. P. 686,768, 745,644. 25 D. R.-P. 605,969; B.P. 407,540.27 Trans. Faraday Soc., 1939, 36, 94.30 Ibid., p. 101. 29 Ibid., p. 9984 INORGANIC OHEMISTRY.ment of zinc ions within the 1attice.l8 This probably explains the appearanceof only one band at h 5400 in some cases of ZnS-Cu and two, at A 4600 andA 5400, in others. The emission band ascribed by C. J. Milner 30 to ZnS-Pbis in direct contradiction to experiments by N. Riehl l8 on the diffusion ofions in zinc sulphide.Milner's emission band has its maximum at A 4600and is no doubt due to ZnS-Zn. Furthermore, owing to the difficulty inexcluding minute activating traces during preparation, most zinc sulphidespossess weak bands due to these other impurities. This is particularly notice-able in specimens of ZnS-Ag, which invariably have a green component dueto copper. J. H. Gisolf and F. A. Kroger 31 have shown that this becomesparticularly apparent owing to differential behaviour with variation of tem-perature and intensity. Cadmium sulphide behaves'similarly to zinc sulphide,its fundamental absorption having a long-wave-length limit at A 5165, butthe emission band has its maximum in the infra-red, although this band doesextend into the red.Solid solutions of zinc and cadmium sulphides have intermediatecharacteristics depending almost linearly on the molar composition of themixture, and do not show two separate bands which vary in their relativeintensity according to the comp~sition.~~ F.A. Kroger 2O investigated thesystem of these two sulphides thoroughly, and found that a new group ofabsorption and emission bands appeared at -180", the position and separ-ation of these bands varying with the molar composition in the same way.He also studied the absorption and luminescence of ZnS-MnS and ZnS-CdS-MnS systems 33 and found in absorption a fundamental characteristicof the ZnS or ZnS-CdS and a band system characteristic of the MnS.The phosphorescence of these compounds decays according to a bimole-cular law, which is readily understood from the existence of photoconductivityunder excitation which also produces luminescence.M* 35 The existence ofmetastable states, or '' trapping " levels, rather complicates the process, butis nevertheless necessary to explain the long-duration phosphorescence.Aninvestigation by W. de Groot 34 on ZnS-Cu has led to some interestingconclusions. He found a rapid decay a t high intensities of exciting radiation,as predicted by the decay law, and also with radiation of shorter wave-length. The latter he ascribed to the fact that such radiation is absorbed ina thin layer of material, and the resultant intensity of radiation per unitvolume of luminescing material is consequently greater. J. W. Strange 36reports that zinc sulphide phosphors activated by silver and copper show twoshort-period decays of an exponential character when excited by cathoderays.At high current densities a component of about 1 microsecond decaywas observed, whereas at low current densities another component of about31 Phymkz, 1939, 6, 1101.33 Ibid., p. 779; 1940, 7, 92.s5 W. L. Levshin and W. W. Antonov-Romanovsky, Physikal. 2. Sovietunion, 1934,6, 769; W. W. Antonov-Romanovsky, ibid., 1935, 7, 3.86; Compt. r e d . Acad. Sci.U.R.S.S., 1936, 2, 97.sa J. H. Gisolf, ib;d.,p. 84.34 W. de Groot, ibid., 1939, 6, 275, 393.86 Tram. Faraday SOC., 1939,36,96REES : LUMINESCENCE OF INORGANIC SOLIDS. 8550 microseconds decay predominates. L. A. Levy and D. W. West 37record the quenching of the X-ray phosphorescence of zino sulphide by 1 partof nickel in lo6 without influence on the fluorescence.This result has led tosome interesting discussion as to its cause: but no completely satisfactoryexplanation has yet appeared.Silicate Phosphors.-Extensive publications have been made on theproperties of zinc orthosilicate, Zn,Si04, both pure and activated, and also ofa so-called zinc metasilicate, ZnSiO,. Reference to the ZnO-SiO, phasediagram 313 shows that the latter compound does not exist, the only compoundbeing the orthosilicate : papers purporting to deal with ZnSiO, are, in fact,considering Zn,SiO, with excess SiO ,. The pure orthosilicate, however,exists in three modifications, a, the normal modification formed by slowcooling of the melt, and the p- and the y-form obtained by rapid quenchingof the melt.The last two forms revert to the or-form on heating to 900"and 300" respe~tively.~~ Non-activated zinc silicate phosphors may beprepared in these three varieties-fluorescing green, yellow, and red,respectively. G. R. Fonda considers the yellow variety to be an amorphouscomplex of zinc oxide and silica in orthosilicate proportions, and the redphosphor a similar amorphous complex in the proportions ZnO,SiO,.Although Fonda claimed that the only crystalline phases of the yellowmodification were cristobalite and zinc oxide, H. P. Rooksby and A. H.McKeag41 and S. H. Linwood and W. A. Wey1,42 in agreement with A. Schleedeand A. G r ~ h l , , ~ have produced X-ray evidence for a second crystalline modific-ation of zinc orthosilicate, which appears to be derived from the cristobalitemodification of silica.The red phosphor, prepared by rapid quenching fromhigh temperatures, was reported by Schleede and Gruhl to be crystalline,although the X-ray evidence was by no means conclusive. Rooksby andMcKeag failed to obtain any evidence of a third crystalline modification, andLinwood and Weyl are of the opinion that it is a true glass.The a-orthosilicate absorbs in the range 11 2200-3000 43 and also appearsto be excited to luminescence by the neon resonance lines 1 736 and 1740."A. Ruttenauer 44 has demonstrated that A 3650 also produces luminescence.In an attempt to correlate these facts, F. A. Kroger45 has studied theabsorption, excitation, and emission spectra of zinc silicate phosphors.Theintroduction of manganese was not found to shift the absorption edge, butincreased the absorption on the long-wave-length side. A new series ofbands with a new edge at A 3000 appeared in compositions containing more37 Trans. Paraday SOC., 1939, 35, 128.8 8 E. N. Bunting, J . Res. Nat. Bur. Stand., 1930, 4, 131.3B 2. Elektrochem., 1923, 29, 411.41 Trans. Faraday SOC., 1941, 37, 308.43 C. G. Found, Trans. Ill. Eng. SOC., 1938, 33, 186; J. W. Marden, N. C. Beese,and G. Meister, &id., 1939, 34, 66.44 A. Ruttenauer, 2. tech. Physik, 1938, 14, 31; H. G. Jenkins and J. N. Bowtell,Trans. Faraday Soc., 1939, 35, 165; M. Schon, ibd., p. 162; G. R. Fonda andH.Huthsteiner, J. Opt. SOC. Amer., 1942, 32, 156.4O J . Physical Chem., 1940, 44, 851.42 J. Opt. Soc. Amer., 1942, 32, 443.4 5 Physica, 1939, 6, 76486 rNORQANfC aEMI8TBY.than 6% of manganese, In emission, pure oc-orthosilicate ha6 maximum atx 4150, minute additions of manganese developing a new band in the green at5230. With further increase above 6% of manganme E\, red band developedwith 8r maximum at X 6100.Taken in conjunction with J. T. Randall's work on manganese-ttctivatedphosphora,46 this work leads to one conclusion, viz., that these bands are dueto the bivalent manganese ion in different lattice environments. It is to benoted that substitution of some of the zinc by beryllium also leads to thedevelopment of the red band. S. W. Linwood and W.A. Weyl,a in theirexperiments on fluorescent glasses activated by manganese, found evidencethat the green band is assooiated wit4 manganese ions in the lattice positions,whereas the red band is attributed to transitions involving these ions ininterstitial positions. They consider that these conclusions may be appliedto the crystdline silicates also. The double silicates of zinc with titanium,zirconium, hafnium, and thorium produce emission bands in the violet. Inzinc beryllium silicates with additions of Group IV silicates the red band isretained also, the consequent phosphors corresponding to almost whiteThe phosphorescence of silicate phosphors has been studied by G. R.Fonda,47 N. C. BeesqZ8 and K. Birus and H. Zierold.49 The decay law for thegreen band in manganese-activated silicate phosphors was found t o becomplex, the first process being an exponential decay unaff eoted by tempera-ture, and the second a temperature- dependent bimoleoular process.A completely satisfactory mechanism for the luminesoence of silicates hasnot yet been advanoed, but any proposed mechanism must take into accountthe decay process and the photoconductivity.6oTzcngstates and MoZybcEates.-Pure tungstates and molybdates of bivalentmetals such as calcium, magnesium, zinc, and oadmium are luminescentunder cathode ray and ultra,-violet excita,tion.61 The emitted light is ELbroad band in the blue region of the spectrum, which appears to beindependent of any traoe elements which may be present, except that rare-earth ions, such as those of samarium and europium, may cause the appeaz-ance of new bands.52 Small concentrations of lead have been claimed asenhancing the luminescenoe of calcium tungstate although no appreciableshift occurs in the ~pectrurn.5~ No mechanism for the emission proaeas hasyet been advanced, as little physical investigation has been done on thesephosphors.There is no long afterglow in the oam of tung8tate phosphors,as in the phosphors already mentioned, but it is of the order of sec.,36 andthis makes any interpretation on the basis of mechanisms suggested for otherphosphors precarious.A, 170, 272.OO~OIWS. 5Tram. Faraday floe., 1939,85,11; Nature, 1988,142,113; Proc. Roy. doc., 1939,47 J . Appl. Physics, 1939, 10, 408.51 A.Schloemer, J . p. Ch., 1932,158,61.62 M. Servigne, Tech. Moderne, 1937,29,443; Compt. rend., 1936,200,2016.69 F. E. Swindellg J . Opt. SOC. Amer., 1933, 23, 129.J . Opt. Soo. Amer., 1939, 29, 26.N&&e., 1942, 3(f, 63. so R. Hof&adter, Physical Rev., 1938, 54, 864WELOH : MODERN TECHNIQUE IN INORGANIC CHEMI8TRY. 87M~Uameozcs hiwecent Muterials.-Numerous other compounds bwebeen reported as exhibiting visible luminescence &om time to time Itwould be impossible even to list them here, but some of the mom intereshmg;classes are: borates of cadmium and zinc activated by manganese (redluminescence) ; phosphates of zinc, activated by manganese (red) ;46magnesium titanate (red) ;54 beryllium nitride (blue) ;55 and the chromium-activated phosphors of al~mina.~6 A.L. G. R,3. SOME APPLICATIONS OF MODERN TECHNIQUE IN INORGANIC CHEMISTRY.A number of recent contributions to Inorganic Chemistry lend themselvesparticularly well to discussion from the viewpoint of experimental technique,and illustrate the important advances made possible by the development ofnew methods of investigation. It is generally recognised that the intro-duction of vacuum methods of manipulation by Stock and others was the keyto the chemistry of volatile hydrides, and many recent advances in InorganicChemistry have similarly been initiated by the development of new technique.A general trend of considerable importance is evident in the increasingamount of work being carried out on solid substances and their reaotions.This trend is attributable to the more widespread availability of X-raymethods for examination of crystalline materials.Since X-radiograms affordinformation on the actual lattice structure of a solid, they provide a powerfulmethod of general application for identifying and characteriaing crystallinesubstances. Although the more refined methods of X-ray crystallography,applied generally to single crystals, are needed for the elucidation of struc-tural details, powder X-radiograms obtained by the DebyeScherrer methodare eminently suitable for routine use in identifying a solid material orestablishing its homogeneity. It is important that the potentialities ofX-ray methods in ordinary research work on Inorganic Chemistry should bemore widely recognised, and for this reason 8ome prominence is given to suchmethods in this Report.A typical instance of the utility of X-ray powder analysis is afforded by arecent paper on the structures of heavy-metal iron cyanides,l in whichPrussian-blue and Turnbull’s blue, long thought to be distinct compounds,are shown to be chemically identical.A distinction between the formulaeFee[Fe( CN),], (ferric ferrocyanide) and Fe,[Fe( CN)& (ferrous ferricyanide)cannot be made from the X-ray data, but the latter formula is supported byoxidation-reduction potential measurements.In the study of different modifications of polymorphic solids, X-raymethods are invaluable, since they permit unequivocal identification of the64 E.Tiede and E. Villain, Bw., 1940, 73, 274.65 S . Sat&, Bull. Inst. Phys. Chem. Res., Japan, 1935, 14, 920.66 0. Deutschbein, Ann. Phyaik, 1932, 14, 712; 1939, 20, 828; Physikal. Z., 1932,1 H. B. Weiser, W. 0. Milligan, and J. B. Bates, J . Ph@al Chrm., 1942, 46, 99.33, 87488 INORUANTC CHEMISTRY.individual moWcations, even in specimens differing in their state of division,or containing impurities. The formation of different crystalline forms ofcalcium carbide in the electric arc furnace has recently been studied 2 withthe aid of X-ray powder technique. " CaC, 111," the form which reactsmost readily with nitrogen in the formation of calcium cyanamide, isobtained directly from calcium carbonate and carbon in the arc furnace ifnitrogen and sulphur are carefully excluded ; grinding of the product readilycauses transformation into another known form, " CaC, 11." A new form ofthe carbide, " CaC, IV," stable only at high temperatures, was identifiedthrough the use of a Debye-Scherrer camera designed for examination ofpowder samples held at temperatures up to 600".The use of X-ray techniquein this investigation has permitted a detailed and direct study of the inter-relationships of the different forms of calcium carbide.Non-stoicheiometric compounds have recently received considerableattention, and the applicability of X-ray analysis in this field is well illustratedby a paper on lower vanadium oxide^.^ A " VO-phase " is found to be stablein fhe composition range VO,,.,-VO,., at high temperatures ; on cooling,this phase decomposes into vanadium (or a solid solution of oxygen invanadium) and a higher oxide of undetermined constitution.The" V,O,-phase " has a lower limit of stability at the approximate compositionIn the examination of systems in which a number of solid phases mayoccur, X-ray methods have led to marked advances. Recent work includesst study of magnesium amalgams," in which the existence of compounds havingthe formuh Mg3Hg, Mg5Hg,, Mg,Hg, Mg,Hg,, MgHg, and MgHg, wasestablished. Reference may also be made to a detailed examination ofhydrated forms of cupric oxide, and of certain basic copper sulphates,5 inwhich previous conflicting conclusions were resolved by' X-ray analysis andby isothermal or isobaric dehydration experiments.Cu(OH), was found tobe the only compound formed in the system CuO-H,O, and in the portionsof the system CuO-S03-H,0 which were studied the stable solid phases werefound to be 4Cu0,S03,3H,0 and a new basic sulphate, 5Cu0,S0,,nH20.In further work on the system Na,O-Si0,-P,05 Debye-Schemer X-radio-grams have shown that anhydrous sodium silicophosphates do not occur asstable solid compounds.6Some attention has been given recently to techniques involvingsintering of powder mixtures, recent work on the theory of reactions betweensolids having led to renewed interest in processes of this type, which are ofconsiderable practical importance. The theoretical and experimental basisv01.35'2 M. A. Bredig, J . Physical Chem., 1942, 46, 801.4 G.Brauer and R. Rudolph, {bid., 1941, 248, 405.W. Klemm and L. Grimm, 2. anorg. Chem., 1942, 250,42.Fig. 1 on p. 406 of this papershows particularly well how the existence of the various solid phases can be deducedfrom the X-radiograms.H. B. Weiser, W. 0. Milligan, and E. L. Cook, J . Amer. Chem. Soc., 1942,64,503.E. Zintl and W. Morawietz, 2. anorg. Chem., 1940, 245, 12WELCH : MODERN TECHNIQUE IN INORGANIC CHEMISTRY. 89of the subject has been reviewed in detail by G. F. Huttig,' by whom much ofthe original work in this field has been carried out. W. Jander and hiscollaborators have also continued their studies on " intermediate states "in reactions between solid oxides.* The formation of these states is shownby changes in the catalytic activity, adsorptive power, rate of dissolution,and other properties of the oxide mixtures after controlled heat treatment,and their investigation has contributed considerably to our knowledge of thechanges involved in reactions between powdered materials.A brief outlineof Jander's views on these changes, in which Hiittig's slightly differentapproach is explained, has appeared re~ently.~ Several reactions betweenpowdered solids at temperatures well below their melting points, particularlyreactions involving arsenic trisulphide and metallic oxides, have beenstudied by E. Montignie.lo Other instances of sintering reactions includethe preparation of several alkali-alkaline earth phosphates from stoicheio-metric mixtures of the alkali carbonate with the alkaline earth monohydrogenphosphate (e.g., NaSrPO, is obtained from Na,C03 and SrHP04).11 Prepar-ation of a cadmium apatite, Cd,,(PO,),F,, and of certain mixed calciumcadmium apatites (the corresponding Ca,Cd, Ca,Cd,, and CaCd, compounds)by heating appropriate mixtures of calcium and cadmium phosphates andfluorides is also reported.l2In spite of the general trend towards the chemistry of ,solids, little workhas been published recently on reactions at temperatures above the limit ofapproximately 1000" attainable with common laboratory furnaces. Althoughthe use of temperatures much above this limit incurs difliculties with bothheating elements and refractory materials, the construction of suitablefurnaces is not outside the range of the average laboratory.l3 Some of themost interesting recent high-temperature studies are those of E.Zintl and hiscollaborators on the monoxides of boron and silicon,l4* 15 referred to else-where in this Report.1, An interesting development due to Zintl is the use ofKoEloid-Z., 1941,94,137 ; 95,258 ; 97,281 ; 1942,98,6, 263. Other recent papersby Huttig et al. on sintering in metallic powders are as follows : G. F. Huttig, C. Bittner,R. Fehser, H. Hannawald, W. Heinz, W. Hennig, E. Herrmann, 0. Hnevkovsky, and J.Pecher, 2. anorg. Chem., 1941,247,221 ; J. Hampel, 2. Elektrochem., 1942, 48,82 ; G. F.Huttig and K. Arnestad, 2. anorg. Chem., 1942, 250, 1 .W. Jander and G. Leuthner, ibid., 1939,241,57 (MgO + TiOz) ; W. Jander and K.Grob, ibid., 1940, 245, 67 (NiO + Al,O,); W.Jander and W. Wenzel, ibid., 1941, 246,67 (CuO + WO,); W. Jander and H. Riehl, ibid., p. 81 (ZnO + SiO,); W. Jender and(2. Lorenz, ibid., 1941, 248, 105 (MgO + V,O,); cf. also G. F. Huttig and H. Theimer,ibid., 1941, 246, 61 (ZnO + Cr,O,).10 Bull. SOC. chim., 1941, 8, 209.l 1 R. Klement and F. Steckenreiter, 2. anorg. Chem., 1940, 245, 236.12 R. Klement and F. Zureda, ibid., 1940, 245, 229.l3 See, for example, W. M. Cohn, Metal~?wirts., 1929, 8, 367, 599, 623; see also amonograph by W. Fehse, " Elektrische Ofen mit Heizkorpern &us Wolfram " (SammlungVieweg, Heft 90, 1928), for various types of high-temperature furnaces.W. Jander (with H. Herrmann), ibid., 1939, 241, 225.14 E. Zintl, W. Morawietz, and E.Gastinger, 2. anorg. Chem., 1940, 245, 8.15 E. Zintl [with W. BrHuning, H. L. Grube, W. Krings, and W. Morawietz], ibid., p. 1.lo Pp. 71, 7590 INORGANIC GHEWSTRY.elemental silicon as a reducing agent a t high temperatures. For instance,metallic niobium of 99% purity is obtained by heating the dioxide, NbO,,with the stoicheiometric quantity of silicon at 1800" ; the silicon and oxygenform silicon monoxide, which volatdises. Silicon monoxide itself appearsto have useful reducing properties; metallic magnesium distils from itmixture of calcined dolomite, lime, and silicon monoxide heated at 1350"(Mg0,CaO + CaO + SiO * Ca,siO, + Mg), and zinc is similarly obtainedfrFr0Dl zinc oxide or a calcined zinc ore at 1200-1300". In discussing the useof silicon ag a reducing agent, Zhtl points out that silicides formed asintermediate products in the reduction process are likely to be more readilydecomposed than corresponding carbides, so that silicon may be preferableto carbon as a reducing agent in certain cases.15A recent application of the aluminothermic process to the preparation ofborides of tungsten and molybdenum is reported;17 the known boride IVB,and a new compound, Mo,Al,B,, have been obtained. The existence ofsolid compounds of apparently complex composition, such as Mo,A1,B7, is aninteresting feature of modern work on solids which offers a wide field forfurther study.An improved method of preparing aluminium boride, AIB1,,h r the aluminothermic reaction has also been described.18Certain important aspects of reactions between solid materids and gaseshave been studied recently.In their studies on the reactions of solids, G. F.Huttig and his collaborators have included investigations on the degassing ofS O L ~ ~ S , ~ ~ and on the effects of foreign gases on solid-solid and gas-solidreactions.20* 21 The fact that reactions between solid substances may beaccelerated or retarded by the presence of gases which do not combine with thereactants is of some practical importance; for example, the reaction of cal-cinm sulphate with silica at 1100" (affording calcium silicate) proceeds best inan atmosphere of water vapour.21 The '' activation " of solids in reactions withgases has been further studied by R. Schenck and his co-workers,22 who haveshown that silver mixed with Mn,O,, Cr,O,, or Rh,O, combines rapidlywith oxygen at 400--800", a temperature range in which the pure metal doesnot react with oxygen at normal pressures.The effect of the added oxide isdue to formation of a stable compound (of type Ag2Mn20,) having a lowdissociation pressure of oxygen. Vanadium sesquioxide activates the silverin a similar manner after oxidation to the pentoxide, but alumina and ferricoxide have no effect.17 F. Halla and W. Thury, 2. anorg. Chem., 1942, 249, 229.lD G. F. Huttig, H. Theimer, and W. Breuer, 8. anorg. Chem., 1942, 249, 134;G. F. Huttig and W. Bludan, ibid., 1942, 250, 36.2o G. F. Huttig, E. Hermaam, W. Neugebauer, R. Rudisch, R. Wallouch, and 0.Hnevkovsky, Kolloid-Z., 1942,92,9; Ber. deut.keram. Ges., 1940,21,429; K. Sedlatschek,2. anorg. Chem., 1942, 250,23.F. Halla and R. Weil, 8. Krkt., 1939, 101,435.21 F. von Bischoff, ibid., p. 10.22 R. Schenck, A. Bathe, H. Keuth, and S. Suss, ibid., 1942, 249, 88. For earlierpape~s by R. Schenck et al. on this subject, see ibid., 1929,184,39; 1932, m6,273 ; 1934,220, 97WELCH : MODERN TECHNIQUE 1N INORGANIO OHEMISTRY. 91An iuteresting series of gas-solid reactions in whioh gaaeous oxygen is usedas an oxidising agent in presence of sodium oxide hm been investigated.Mixtures of sodium iodide and sodium oxide (or the peroxide or hydroxide)read with air or oxygen at 400-700" to form the periodate Na510,.43Similarly, the '' mixed oxide " Na,PrO,, containing quadrivalent prase-odymium, is formed by heating Pr,03 with sodium oxide a t 470", inand the orthobismuthate, Na,BiO,, is obtained from Bi,03, sodium oxide, andoxygen or air at 650".25 In these reactions the alkali oxide appears to have astabilising effect on the higher valency shown by the element undergoingoxidation; clearly, a number of other compounds of similar types may bereadily obtainable by analogous reactions.Reference has been made previously in these Reports to boron oxfluor-ide.26 It has now been shown 27 that boron trifiuoride rewts with aluminaat 450" according to the following equation : Al,O, + 2BF3 = 2MF3 +B,03 ; the boron trioxide formed in this primary reaction is volatilised as theoxyfluoride. At higher temperatures the reaction can proceed in thereverse direction. Boron trifiuoride also reacts with silica at 450", affordingsilicon tetrafluoride and boron oxfluoride ; titanium dioxide reacts similrtrly,but more slowly, whereas ferric oxide will not react at temperatures below500". Analogous reactions occur with compounds such as silicates andaluminosilicates, the electropositive constituents of these remaining as theirfluorides or borofluorides.These new reactions may find useful application,for instance in the treatment of silicate minerah.The study of hydrothermal reactions, involving liquid water at tem-peratures above loo", offers interesting possibilities, In three recent paperson this subject,28 W. Jander and his collaborators have diacussed theformation of magnesium hydrosilicates, and investigated the systemCaO-Si0,-H,O.Work in this field may have important applications,apart from its geochemical interest.No review of technique would be complete without some reference to themethods applied by W. Bdtz and his sohool in their studies on affinity;these continue to give successful results in investigations on metallicphc~phides.~~ The general method used is to heat a number of mixturescontaining different proportions of the combining elements, and examine theproducts by X-ray powder analysis, density determinations, and degradation23 E. Zintl and W. Morawietz, ibid., 1940, 246, 20. 21 Idem, ibid., p. 26.2b E. Zintl and I(. Scheiner, ibid., p. 32. 28 Ann. Reports ,1940, 37, 127.27 P.Baumgarten and W. Bruns, Ber., 1941, 74, 1232.28 W. Jander and R. Fett, 2. anorg. Chem., 1938, 235, 273; 1939, 242, 145;W. Jandcr and B. Franke, ibid., 1941, 247, 161.2s F. E. Faller, E. F. Strotzer, and W. Biltz, ibid., 1940, 244, 317 (rhodium phos-phides) ; M. Heimbrecht, M. Zumbusch, and W. Biltz, ibid., 1941, 245,391 (uraniumphosphides); M. Zumbusch and W. Biltz, ibid., 1941, 246, 35 (tantalum phosphides);F. E. Faller and W. Biltz., ibid., 1941, 248, 209 (tungsten, molybdenum, and chromiumphosphides); M. Zumbusch and W. Biltz, ibid., 1942, 249, 1 (vanadium phosphides);A. Reinecke, F. Wiechmann, M. Zumbusch, and W. Biltz, ibirE., p. 14 (niobium phos-phidee). See also a general artiole by W. Biltz, Angew. Chem., 1941, 64,32092 INOROANTC CHEMTSTRY.in a tensimeter at suitable temperatures.The formation of compoundrs isindicated by the appearance of new lines in the X-radiograms, and by breaksin the pressurecomposition curves given by the tensimetric measurements.These methods have served to clarify the chemistry of many simple binarysystems.Another useful method of studying equilibria in systems involvingsulphides had been applied by R. Schenck.m2 The equilibrium proportionsof hydrogen and hydrogen sulphide established in reactions of the typeNiS + H, = Ni + H,S are observed, and the composition of stable solidphases is deduced from breaks in the isotherms obtained by plotting theequilibrium percentage of hydrogen sulphide against the composition of thesolid mixture. In the system Ni-Sb-S this method has shown the existenceof the compounds NiS, Ni,S,, Ni3S,, NiSbS, NiSb, and NiSb,; 31 in thecorresponding system with cobalt the compounds are CoS, Co3S,, CosS8,CoSbS, and CO,,S~,S~,.~~Methods of thermal analysis applied to binary systems continue to yielduseful results. The system magnesium-germanium is shown to give acompound Mg ,Ge, corresponding with similar compounds of magnesium withsilicon, tin, and lead.33 In the system silicon-arsenic two compounds, SiAsand SiAs,, occur as stable solid phases.54 Special interest attaches to theapplication of thermal analysis at temperatures in the range -70" to -145",by which the existence of a compound BF3,S0, has been demonstrated; 36additive compounds of boron trifluoride with hydrogen chloride, methylchloride, or nitrous oxide could not be detected.An interesting contribution to the study of molten mixtures at hightemperatures is the introduction of coloured indicator8 to show changes inoxygen-ion activity in oxysalt melts.36 The colour change due to oxidationof tervalent chromium to the sexavalent state may be used for this purpose.The method has recently been applied to melts containing sodium oxide,boron trioxide, and one of several acidic or basic oxides, with the chromiumindicator.37 The proportions of sodium oxide and boron trioxide required togive a neutral tint to the indicator in the absence of added oxide are firstdetermined ; the third oxide is then added, and the new proportion of sodiumoxide to boron trioxide required to maintain " neutrality " is found.Theresult is a measure of the acidity or basicity of the added oxide.Little attention has been given here to vacuum manipulation of volatilesubstances, largely because recent issues of these Reports have mentioneda very large number of applications of vacuum technique. Reference must3O R. Schenck, I. Hoffmann, W. Knepper, andH. Vogler, 2. anorg. Chem., 1939,240,31 R. Schenck and P. von der Forst, ibid., 1939, 241, 145.32 Idem, ibid., 1942, 249, 76.33 W. Klemm and H. Westlinning, ibid., 1941, 245, 365.34 W. Klemm and P. Pirscher, ibid., 1941, 247, 211.3 5 H. S. Booth and D. R. Martin, J. Amer. Chern. SOG., 1942, 64, 2198.36 W. Stegmaier and A. Dietzel, Blastech. Ber., 1940, 18, 297, 353.'' H. Lux and E. Rogler, 2. unorg. Chem., 1942, S O , 159.173WELCH : MODERN TEUHNIQUE IN INORGANIC CHEMISTRF. 93be made, however, to the special methods required for the manipulation offluorine and its volatile derivatives; these have been discussed in a recentTilden lecture 38 and a review on the preparation of flu0rine.3~ Continuedinterest in fluorine chemistry is shown by a recent paper on chromiumfluorides.40 If fluorine is passed over chromium, chromic chloride, orchromic fluoride at 350-500", the tetra- and penta-fluorides are formed.The former is a brown solid which volatilises perceptibly at 150", and thelatter is a brilliant red compound which is liquid at 100". At temperaturesbelow 100" a small quantity of hexafluoride is obtained.Chromyl fluoride,CrO,F,, is obtained as a brown gas, condensing to brownish-red crystals with avapour pressure of 24 mm. at Oo, by treating chromyl chloride with fluorineat 200". The chromyl fluoride changes readily into a light-colouredpolymeride melting at about 200".Interest continues in reactions in non-aqueous solvents. A usefuldescription has recently been given of the apparatus and methods used toprepare cobalt and chromium amides by reactions between potassium amideand the cobalt or chromium hexammine nitrate in liquid ammonia solution.41Methods of carrying out certain reactions of the products are also described.A compound of univalent nickel, K,Ni(CN),,42 has recently been isolated asa red solid by the action of sodium or potassium (or calcium 43) on excess ofK$Ji(CN), in liquid ammonia solution.44 If excess of alkali metal or calciumis used in this reaction the product is not K,Ni(CN),, but K,Ni(CN),, acopper-coloured solid which is oxidised on momentary exposure to air.Other complex cyanides [K,Cd(CN)4, K3Cu(CN),, KAg( CN) ,, and K,Zn( CN),]give the free heavy metals on reduction with potassium or sodium in liquidammonia; the products (except zinc) are pyrophoric.a Pu?e NaNbF, andNa,NbF, have recently been prepared by crystallisation of solutions of thecomponent fluorides (NaF and NbF5) in anhydrous liquid hydrogenfluoride.45 The use of this solvent has obviated difficulties due to hydrolysisof the fluorides in aqueous media..The use of magnetic susceptibility measurements in the study of mole-cular constitution is now well e~tablished.~~ In the past there has been somespeculation regarding the constitution of the so-called '' even " polyhalides,such as CsI,.It has now been found that CsI, is diamagnetic, showing thatits molecule contains no unpaired electrons ;47 the simple formula CsI,38 H. J. Emelhs, J., 1942, 441.38 G. H. Cady, D. A. Rogers, and C. A. Carlson, Ind. Eng. Chem., 1942, 34,443.40 H. von Wartenberg, 2. anorg. Chem., 1941, a47, 135.41 0. Schmitz-Dumont, J. Pilzecker, and H. F. Piepenbrink, ibid., 1941, 248, 176.42 See I. Bellucci and R. M. Corelli, 2. anorg. Chem., 1914, 86, 88.43 J. W. Eastes and W. M. Burgess, J. Amer. Chem. Soc., 1942, 64,2716.44 Idem, ibid., p. 1187.4 5 A. W. Laubengayer and C. G. Polzer, ibid., 1941, 63, 3264.See also p. 77 of this Report.See W. Klernm, " Magnetochemie " (Physik und Chemie und ihre Anwendungen inEinzeldarstellungen, Vol. I), Leipzig, 1936 ; S. S. Bhatnagar and K. N. Mathur, " PhysicalPrinciples and Applications of Magnetochemistry," London, 1935.4 7 S. S. Hubard, J. Physical Chern., 1942, 46, 22794 TNOBUANIC UHBIMISTRY.requires one unpaired electron in the molecule, and the simplest formulapermitted by the magnetic data is Cs I The constitution Cs[I,-I,-I,]Cs,in which an iodine molecule is " shared between two tri-iodide groups, issuggested. It now remains for the structure to be determined by X-raymethods. The magnetic behaviour of a number of metallic complexes ofsalicylaldehyde imines has recently been examined.48 These complexes, manyof which contain quadridentate groups, afford interesting examples ofdifferent typea of bonding between the metal atoms and the organic groups,the magnetic data giving important information on these bond types. Themore rigorous application of magnetic measurements and X-ray data, indetermining the constitution of solids is typified by an investigation on thedihalides of titanium. and vanadium.49The usefulness of electron-diffraction observations in establishing thenature of surfaces is shown by recent work on the surface reaction of zincoxide with hydrogen ~elenide.~~ This reaction ceases when the productcontains about 6% of selenium. X-Ray examination indicates that itconsists chiefly of unreacted zinc oxide, but in electron-diffracttion photo-graphs the pattern due to zinc selenide largely predominates. It is deducedthat the selenide forms a coherent coating on the oxide particles, theelectron-diffraction method showing the structure of the surface rather thanthat of the interior of the particles.A new experimental tool with potential applications in InorganicChemistry is the electron microscope. The colloidal nature of molybdenum-blue (shown to have the formula Mo80B,nH20) is strikingly demonstrated byan electron photomicrogram giving a magnification of 62,000 diameters ;51the particle size is found to be of the order of 100 A.In conclusion, reference may be made to a recently published book dealkga t Borne length with reactions in electric discharges,62 a subjeot which meritscontinued interest ; recent new applications of electric discharge techniquesappear to be lacking.s,;A. J. E. W.'H. J. EMEL~ZUS.A. L. G. REES.A. J. E. WELUH.4 8 W. Klemm and K. H. Ra,ddatz, 2. anorg. Clzem., 1942, 260, 207.4B W. Klemm and L. Grimm, ibid., 1942, 249, 198, 209.50 M. L. Fuller and C. W. Siller, J . Appl. Physics, 1941, 12, 416.5 1 F. B. Schirmer, jun., L. F. Audrieth, S. T. GFO~S, D. S. McClellan, and L. J. Seppi,62 G. Glockler and S. C. Lind, " Tho Electrochemistry of Gases and other Dielectrics,"J. Anier. Chem. SOG., 1942, 64, 2543.New York, 1939

ISSN:0365-6217    

DOI:10.1039/AR9423900070

出版商:RSC    

年代:1942

数据来源: RSC

摘要:

CRYSTALLOGRAPHY.1. INTRODUOTION.THE more physical sections of last year’s Report dealt with the diffuseX-ray reflections from crystals, and with metal structures. Although avery large number of new papers on these subjects have appeared duringthe year, they are not further reviewed in the present Report. It is die-cult to deal adequately with the whole subject of modern X-ray crystal-lography each year, and improvements in the technique of analysis havenot been mentioned for some time. Hence a short section is now devotedto this subject, Since this was last reviewed1 the use of intensity dataand Fourier series methods has become more common in crystal analysis,and a number of improvements and suggested modifications in techniquehave been made. One of the most interesting developments is that of the“ X-Ray Microscope ” by Sir W..L. Bragg, which provides a direct opticalmethod for performing Pourier syntheses and giving projections of crystalstructures.In the field of inorganic chemistry a number of interesting structureshave been worked out in detail, including those of phosphorus pentachloride,which contains not PCI, molecules but a mixture of PClt and PCI; ions;and the hydrazinium ion, of interest because o f the like formal changes onadjacent atoms. The complex fluorides have been further examined andprovide more examples of 6- and 7-co-ordination complexes. The structureof hydrogen peroxide has been determined in an interesting manner, bya quantitative study of its addition compound with urea.The structure isneither cis- nor tram-, but the more stable intermediate form predictedsome time ago by quantum-mechanical calculation.Carbon compounds have received a lot of attention this year, and a newintermediate structure, in addition to diamond and graphite, has beenproposed for the element itself.dl-Ahnine and melamine (cyanuric triamide) are two new structureswhich have been very carefully determined, and both the inter- and theintra-molecular bond lengths are interesting. The structures of gutta-percha, rubber, and polychloroprene have received considerable attention,but no degree of flnality can be attached to these results. In connectionwith the rubber problem, however, the full and accurate determination ofthe structure of p-isoprenesulphone is of considerable interest.In protein structures, some interesting work on hzmoglobin hasappeared, careful comparative studies having been made on the crystals indifferent states of hydration.Some tentative oonclusions have beenreached regarding the dimensions and orientation of the molecule.Ann. Reports, 1938, $I$, 17496 CRYSTALLOGRAPHY.2. TECHNIQUE OF STRUCTURE ANALYSIS.As the trend of crystal analysis passes to the examination of morecomplex structures, and to the more accurate determination of atomicpositions in simple structures, it is inevitable that more and more attentionshould be paid to intensity data and their interpretation by Fourier seriesmethods. In most of the investigations now being made, the X-ray inten-sities are obtained by visual estimate from photographic records.If thisis done carefully and systematically a surprising degree of accuracy can bereached, sufficient indeed to determine the atomio positions to within verynarrow limits in most structures. This has been tested in an interestingmanner in a recent analysis carried out independently by two workers,starting from the same X-ray films. They finally obtained almost identicalvalues for the parameters of the structure.Unless adequate correction can be made for various factors, includingabsorption (one of the major sources of error), it may be a waste of timeto employ elaborate means of intensity measurement. However, if theobject of the work is to detect delicate variations in bond lengths ratherthan mere structural types, then quantitative intensity measurements arerequired. A number of integrating photometers have been speciallydesigned for crystal analysis and these have been reviewed in a recentarticle.2 Where the highest accuracy is not necessary, the more rapidmethod described by R.H. V. M. Dawton may be employed, which makesuse of positive prints from the X-ray negative. Another process, alsoemploying prints, has been mentioned by J. S. Lukesh.4 A method, whichis chiefly of use in correlating very strong and very weak reflections (therange covered may be several thousand to one) consists of the simple deviceof exposing a number of X-ray films simultaneously in series.5 Eachsuccessive film cuts down the intensity by a factor which is very constantfrom one film to another, and also constant over the surface of each film.This multiple film technique has been much used in the structure analysesreviewed in Sections 3 and 4.The principles involved in the construction of powder cameras andtheir calibration have been described in two recent articles,6 and a newmultiple exposure instrument for making accurate comparisons of spacingshas been constructed by A.R. Ubbelohde.’ E. Orowan has described aningenious method which gives directly on a stationary film the angularposition of the reflecting plane, in rotation or oscillation photographs. Afine grid of parallel wires is placed in front of the film and rotated in its ownplane in time with the crystal during the exposure.The wires cast a shadow1 See ref. (20), Section 3.5 J. J. de Lange, J. M. Robertson, and (M’iss) I. Woodward, Proc. Roy. SOC., 1939,6 A. J. Bradley, H. Lipson, and I?. J. Petch, J. Sci. Instr., 1941, 18, 216; A. J. C.J. M. Robertson, J. Sci. Instr., 1941, 18, 126.Proc. Physical soc., 1938, 50, 919. J . Chem. Physics, 1941, 9, 659.A , 171, 398, 404.Wilson and H. Lipaon, Proc. Physical SOC., 1941, 53, 245.J . Sci. Instr., 1939, 16, 165. Nature, 1942, 149, 355ROBERTSON : TECHNIQUE OF STRUCTURE ANALYSIS. 97on each spot and their inclination gives the precise angular setting of thecrystal a t the time when the re%eotion takes place-a result which can other-wise only be attained on a much more complicated moving-flm instrument.For the summation of Fourier series, it seems likely that accuratenumerical methods will always be necessary for a final expression of theresults ; but the Fourier series method is being employed more and moreas one of successive approximation, and in the earlier calculations, and inthe summing of Patterson series, rapid methods are very desirable.Inthis connection the most interesting development of recent years is SirF I G . 1 .Interjerenoe Pattern (Diopside).FIG. 2 .&ructure of Diopside, CaMg(SiO,),, Projected on (010).W. L. Bragg’s ‘‘ X-Ray Microscope.” In this optical method a projectionof the structure is obtained by means of a two-dimensional diffractiongrating, which may be constructed by drilling holes in a brass plate.Each.X-ray reflection from a given zone is represented by a hole of area propor-tional to the structure amplitude, and the holes are arranged in the positionsof the points in the corresponding section through the reciprocal lattice.With a monochromatic point source of light and appropriate lens systemthe diffraction caused by the holes results in a very realistic image of thecrystal structure being formed as shown in Fig. 1. The positions of theatoms in this structure are shown by black dots in Fig. 2. It is diflicult0 Nature, 1939,143,678; 1942, 149, 470; see also H. Boersch, 2. tech. Physik, 1938,337, mdM. J. Buerger, Proc. Nat. Acad. rSci., 1938, 26, 383; 1941, 27, 11798 (IRYSTAJLLOGRAPHY.to construct a smd-scale brass plate with the accuracy neceasary for acomplicated structure, but it has now been found possible to employ $Iphotographic method, whereby a large drawing of the reciprocal lattice,representing the crowgrating spectra, can be photographed on a muchreduced scale on a special type of plate.Optically true glass plates of courseare necesgary, as the paths of the wave from the various transparent holesmust be correct to a fraction of a wave-length. The crom grating for Fig. 1was produced photographically. The method haa definite possibilities as aquick routine way of effecting Fourier summations, particularly when deal-ing with Patterson series where the amplitudes are always positive.Sir W. L. Bragg's earlier photographic method of Fourier synthesis,lOwhereby successive exposures were made on a single film or paper of eachindividual term in the double Fourier series, has now been further developedand made faster and more automatic by M.L. Huggins 11 by means of aset of standard " hk masks," one for each term, and the use of special filmemulsions which reduce the high background intensity.A machine for the rapid numerical summation of Fourier series, con-structed from simple parts, has been described by D. Mrtcewan and C. A.Beevers.12 The information contained in the well-known Beevers-Lipsonstrips is actually built into this machine, so that no selection and sortingof cards or strips is needed. The operation of the machine is very speedy,Fourier terms being added in a t the rate of about one every two secondsand the results recorded on a set of sixteen counters.With regard to mathematical methods, a new type of synthesis of X-raydata has recently been suggested by S.H. Yu.lS Full details have not beengiven, but this method a p p m to be a modified form of the Pattersonfunction, in which less interference of the peaks (better resolution), com-bined with reduced diffraction effects, is olaimed. It has been pointed0ut,14 however, that all the advtmtagges claimed cannot be obtained fromthe X-ray data given by any actual arystd. There is natural physicallimit to the resolving power inherent in the finite size and thermal motionof atoms which compose the crystal.The application of a least-squares method to refining the trial para-meters in a structure determination has been described briefly by E.W.Hughes.16 It ie claimed that this method has several advantages over theFourier seriea method; e.g., all the observed structure factors need not beincluded in every case. Further details will be of hterest.The application of differential radial distribution OUIPVBB to the andyeesof g l m of related compositions has been desclribed by J. S. Lukesh,'6and the idea of suoh difference diagrams hm been applied to crystal andyshby M. J. Buerger.1' The simplest cme k3 that of iaomorphoue centm19 I;. Krist., 1929, 70, 476.la J . Sci. Instr., 1942, 19, 160.1 4 R. Lipson, aid., 1942, 150, 25; C. S. Lu, aid., p. 407.lC J . AM. Cbm. Soc., 1941,65, 1737.1. Proc. Nat. Amd. Sci., 1942, 20, 277.l1 J .Amr. Cham. SOC., 1941, 68, 66.l3 Nature, 1942,188, 638.l7 Ibid., p. 281ROBERTSON : WOBGANXJ STRUUTWRES. 99symmetrical structures with different atome on the symmetry centre, Inthis cme the difference diagram applied to the Patterson projections givesan undistorted picture of the actual cryshl struoture. Thh result, however,can be expressed physically in another way, and has been used, for example,to elucidate the struotures of the phthalocyanines.l* The more genemltreatment of the method may have interesting applications to solid solutione,3. INORUANIC STRUCTURES.There is less than the usual volume of work on inorganic structures toreport this year, although a number of accurate determinations of con-siderable interest have been carried out. Mention of partial or incompletedeterminations and work on complex systems and minerals has been largelyomitted on account of restrictions on space.With regard to elements, an important event is the proposal of a newstructure for oarbon,l in addition to the two well-known crptalline varieties,diamond and graphite.A third variety is now necessary in order fo explainthe appearance of certain extra diffraction lines which am given occasionallyby certain specimens of graphite and do not appear to be due to impuri-ties3 or to diffuse scattering effects.* In the new structure there arethought to be three differently disposed layers of carbon atoms in sequence,instead of two aa in graphite, before identical repetition is reached.As aresult the c axis is 14 times as long as in normal graphite. The structureis described crysbllographicdly m R h with a = 3.635 A., o! = 39*40°,and atoms at & (3, Q, Q). It appears to be one possible intermediate struc-ture between diamond and graphite, although not apparently existing asseparate single crystals.A discrepancy found in the carbon-carbon distance for the small graphitecrystallites which are formed during the carbonisation of cellulme, coal,e t ~ . , ~ has now been cleared up. These results led to certain carbon-carbondistances a,s amall as 1.37 A. for the carbon rings in microcrystalline graphiteas against a value of 1.415 A. in normal macrocrystalline graphite, a resultdiflicult to understand in view of the strong bonding in these net planes.A.Taylor 6 has now found that a correction should be applied to theseresults, based on a mathematical treatment of X-ray dif3fraction fromrandom layer lattices developed by 33. E. Warren,' which shows that thediffraction peaks are displaced from their position in the direction of increas-ing 8 by an amount which depends upon the crystal size. For the minutegraphite crystallites mentioned, this correction is important, and its applic-ation brings the results accurately into line with the normal carbon-carbongraphite distance of 1.415 A.J. M. Robertson, J., 1936, 1195.1 H. Lipson and A. R. Stokea, Nature, 1942, 149, 328.2 G. I. Finch and H. Wilmrtn, Proc. Roy. Soc., 1936, A, 155, 345.A. Taylor and D. Laidlew, Nature, 1940, 146, 130.4 (Mrs.) K.Lonsdale, (Miss) I. E. Knaggs, and H. Smith, ibid., p. 332.6 H. E. Blayden, H. L. Riley, and A. Taylor, J., 1939, 67..a Nature, 1942,160,462. 'I Physical Rev., I941, 69, 693.BEP.-VOL. XXXIX. 1 00 CRYSTALLOQRAPHY.X-Ray diffraction experiments on liquid argon8 and oxygeng havebeen described. In addition to the expected peak a t about 1-22 A. inthe atomic distribution curve for oxygen, due to diatomic molecules,there is a further fairly strong peak at about 2.2 A., probably due toozone, and third and fourth peaks indicating aggregates of 0, molecules, orpossibly 0,.Oxides of platinum are difEicult to prepare, but one such compound,formed on a platinum wire in use over a period of years, has now beenanalysed by X-ray methods lo and shown to consist of Pt,O, molecules.The cubic structure (a = 6.226 A., containing two molecules of Pt304) hasbeen analysed by Debye-Scherrer photographs and Fourier analysis, whichgive a Pt-0 distance of 2.2 A., and a Pt-Pt approach of 3.113 A.Silver arsenate l1 has a structure similar to that of silver phosphate,consisting of a cubic body-centred arrangement of arsenic atoms, eachsurrounded tetrahedrally by oxygen atoms, silver occupying positions onthe faces of the unit cube. Interatomic distances are As-0 = 1.75 0.05 A.,Ag-0 = 2.34 & 0.05 A. Potassium dihydrogen arsenate l1 is tetragonaland similar to the corresponding phosphate.In this case the arsenategroup is a slightly deformed tetrahedron with 0-As-0 angles of 104" and113".The signifhant distances are As-0 = 1.74 0-03 A. and 0-H . . -0 =2.54 & 0-03 A., the latter figure indicating a strong hydrogen bridge similarto that occurring in potassium dihydrogen phosphate. The influence ofrandomness in the dihydrogen arsenate structure is discussed.The nature of the hydrogen bridges in acid phosphates has been furtherstudied by A. R. Ubbelohde and (Miss) I. Woodward12 by means of theisotope effect. The expansion of the lattice which takes place on sub-stituting deuterium for hydrogen in (NH,)H,PO, is in the direction of thehydrogen bridges and of the same magnitude as in the isomorphous KH2P04,but the contraction a t right angles is considerably larger in (NH,)H,PO,than in KH,PO,. These isotope expansions for the acid phosphates areconsiderable and can be determined with an accuracy of about & 2%,but are not so large as those originally noted in oxalic acid dihydrate.13The positions of the lead and chromium atoms in the monoclinic crystalsof lead chromate (crocoite) l4 have been determined, but not those of theoxygen atoms, and so the exact structure is still in doubt.The complexstructure of tricalcium aluminate Ca3Al,0, l5 with 24 molecules in thecubic cell is similarly in doubt with regard to the arrangement of the oxygenatoms. Details have been given of the tetragonal lattice constants andA. Eisenstein and N. S. Gingrich, Physical Rev., 1942, 62, 261.P. C. Sharrah and N. S. Gingrich, J. Chem. Physics, 1942, 10, 604.lo E. G. Galloni and A.E. Roffo, ibicE., 1941, 9, 875.11 L. Helmholz and R. Levine, J. Amer. Chem. Soc., 1942, 64, 354.l2 Proc. Roy. SOC., 1942, A, 179, 399.l8 J. M. Robertson and A. R. Ubbelohde, PTOC. Roy. Soc., 1939, A, 170, 222.l4 S. B. Brody, J. Chern. Physik?, 1942, 10, 650.l6 H. F. McMurdie, J. Bee. Nat. Bur. S t a d . , 1941, 27, 499ROBERTSON : INORGANIC STRUCTURES. 101atomic positions in the disilicides of thorium,f6 niobium, tantalum, vanad-ium, and rhenium.17The structure of calcium cyanamide 1* is now reported to be analogousto that of sodium azide, NaN,, with C-N = 1.16 & 0.08 A. and Ca-N =2.49 & 0.04 A.A large number of heavy-metal iron cyanides, including Prussian blueand Turnbull’s blue, have been studied by means of X-ray diffractionpatterns obtained from the various iron cyanide gels.lg It is concludedthat they form an isomorphous series of compounds possessing face-centredcubic symmetry, and atomic positions have been proposed.Fig.1 gives a general view of the unit cell in the phosphorus penta-chloride structure, details of which have now been published by D. Clark,H. M. Powell, and A. F. Wells.20 The analysis has been carried out withFIU. 1.General View of Unit Cell. (Small circles, phoqhwus ; large circles, chlorine.)considerable accuracy, from visually estimated intensities, by two inde-pendent investigators, and the close concordance of their results is veryencouraging. All structures involving discrete Pc1, molecules, or sharedpolyhedra, have been eliminated, and it is found that the tetragonal unitcell contains separate tetrahedral PC14+ and octahedral PC1,- groups,arranged as shown in Fig.1. (This result is in keeping with the physicalproperties of the solid, which suggest an ionised form.) There is an appreci-able difference in the P-C1 distances in these two groups, the results averag-ing 2.06 A. in PCG and 1.98 A. in PCl:. This difference is in accord withthe change in electrical charge.21A rather interesting structure has now been proposed for the inter-l 6 G. Brauer and A. Mitius, 2. anorg. Chem., 1942, 249, 325.l 7 H. J. Wallbaum, 2. Metallk., 1941, 33, 378.19 H, B. Weiser, W. 0. Milligan, and J. B. Bates, J . Physical Chem., 1942, 46, 99.2o J., 1942, 642.21 L. Pauling and J. Sherman, 2.Krist., 1932, 81, 1.M. A. Bredig, J . Amer. Chem. SOC., 1942, 64, 17301 02 0RYSTBtI;OQRbpHY.rnediate-temperature modification of Ni(N03),,6NH, by S. H. Y u , ~ ~ It iscubic and consists of flat triangular NO, and octahedral Ni(NH,), groupsinterlocking into a calcium fluoride type of structure. The crystal appearsto be neither a single crystal in the usual sense nor a polycrystalline mass,but is what the author terms a double crystal aggregate. Apparently theNO, group can assume either one of two alternative configurations whichcorreepond to extreme positions of the oscillating NO, groups of the room-temperature modification.Hydrazinium difluoride (I) is of interest as one of those rare stablesubstances where the structure assigned to the ion places like formal chargeson adjacent atoms.The crystal structure has now been fully deterrni11ed.~3The rhombohedra1 cell, space group R%n, contains one molecule. Thelayer type of structure resembles that of cadmium chloride. Each hydra-zinium ion forms six hydrogen bridges to fluoride ions at the corners of anoctahedron elongated in the direction of the N-N axis, the hydrogen atomsin the ion thus having the trans- or staggered configuration. Each fluorideion in turn forms three hydrogen bridges to different NzH6++ ions. Thestructure as a whole is a very open one, obviously due to the stronglydirected nature of the hydrogen bridges. Their length, N-H . . . F =2.62 5 0.02 A., is the shortest yet found between nitrogen and fluorine.The N-N distance is 1.42 A.and the angle N-N' is 110". Ionic struc-tures, of the type (11), (111), (IV), etc., are thought to contribute to theextent of about 75% to the normal state..FHO'H\+ H~N-NYH +/H H\+ H7N-N-H /H I3-N-N': H\ H-N-N H\ -/H HfH H H H+ H+ H+ H+ H+(1. ) (11.) (111.) W.)Compbx F'lrnih.-The structures of the complex fluorides havereceived considerable attention in recent years.% Two distinct types of7-co-ordination complex are now established. First, there is the distortedOctahedron, where the seventh atom is acoommodated by spreading one ofthe faces and inserting the extra atom at its centre, as in the [ZrF,]s com-plex.25 The oxyhexafluoniobate ion [NboF6]= is now shown26 to haveessentially this structure, the oxygen being inserted along a three-fold axison one of the faoes, accompanied by some distortion.The symmetry dis-played is cubic holohedral, but the data for these structures are incom-patible with space-group theory and it is necessary to assume that theyare statisticd results for structures which display some randomness in theatomic arrangement. The second type of 7-co-ordination complex is thatla Nature, 1942, 150, 347.23 M. L. Kromberg and D. Harker, J . Chem. P h y s h , 1942, 10, 309.26 G. C. Hampson and L. Pauling, J . Amr. Chem. SOC., 1938, 60, 2702.2' M. B. Williams and J. L. Hoard, ib&i., 1942, 84, 1139.see Ann. RepO~t8, 1939, 88, 170; 1940, 87, 184; 1941, 88, 104ROBERTSON : INOlWAKIC STRWUTURES. 103of the [NbF,]= and [TaF,]= ~tructure~?7 which are based on the trigonsllprism with the seventh atom a t the oentre of one square face, followed bycertain distortions.The latest structure in this series to be examined by J.L. Hoard andM. B. Williams 28 is that of (NH4),SiF7, which forms crystals of tetragonalsymmetry (DL-P4/mbm). However, all possible structures containing7-co-ordination complexes, [SiF,]=, have been eliminated by the X-raydata, and instead, the atructure is shown to be an ordered aggregate ofammonium, octahedral hexafluosilicate, and " extra " fluoride ions, and thecompound should be described as ammonium hexafluosilicate-ammonium-fluoride, (NH,),SiF,,NH,F. In the structure, each ammonium ion is neareight fluorine atoms of neighbouring [SiF,]" groups and two fluoride ions.The Si-F distance is 1.71 A,, F-F = 2-42 A., and the N-F separations varyfrom 2.8 to 3.1 A.A Fourier projection on the (001) plane is given, alongthe two-fold axis of the octahedral [SiF,]" ion (crystallographic c axis).In all these analyses the X-ray work has been carried out with care, em-ploying oscillation, Laue, and powder photographs with Cu-Ka radiation,and visual estimates of intensities.Rubidium hexafluogermanate, Rb,GeF6, has been re-examined byW. B. Vincent and J. L. Hoard 29 and is now shown to be fully isomorphouswith the corresponding ammonium and potassium sa1ts.m It consists ofan aggregate of K+ and almost regular octahedral [GeF,]' ions. Thedimensions are almost identical with those of the ammonium salt, theeffective radii of rubidium and ammonium ions being about the same,except when the ammonium ion is restricted to a small co-ordinationnumber (4) by the formation of strong hydrogen bonds.Structure of Hydrogen Peroxide.-The direct crystal analysis of hydrogenperoxide is a matter of considerable experimental diffioulty, and hence thevery complete analysis of its addition compound with urea, [" hyperol,"CO(NH,),,H,OJ now made by C.S. Lu, E. W. Hughes, and P. A. Gigubre 31is of interest. The structure of urea is already accurately known.S2 Theaddition product has been analysed, moving-film methods and estimationof intensities by the multiple film technique33 being employed. Theorthorhombic crystal, a = 6.86, b = 4.83, c = 12.92 A., Df$-Pnca, con-tains 4 moleculeB of CO(NH,),,H,O,, and all the parameters have beendetermined and refined by Fourier series methods.It is found that eachnitrogen atom of urea makes two hydrogen bridges of 3-04 and 2.94 A. toperoxide oxygeas of two different peroxide molecules, and that the carbonyloxygen of urea makes two very strong hydrogen bridges (2.63 A.) to other2 7 J. L. Hoard, J . Amer. Ohem. SOC., 1939, 61, 1252.28 J. L. Hoard and M. B. Williams, ibid., 1942, 64,633.29 Ibid., p. 1233.O0 J. L. Hoard and W. B. Vincent, &id., 1939, 61, 2849.3 1 Ibid., 1941, 63, 1607.32 R. W. G. Wyckoff and R. B. Corey, 2. K&t., 1930, 76, 529; 1934, 89, 463.3s J. J. de Lange, J. M. Robertson, and (Miss) I. Woodward, Proc. Roy. SOC., 1939,A, 171, 398104 CRYSTALLOGRAPHY.peroxide molecules, as indicated diagrammatically in (V) and (VI).Theurea molecule is found to have the same configuration as in the ureaFIU. 2.A perspective viewof the hydrogen per-oxide molecule. Thehydrogen atoms (repre-sented by smallerspheres) are assumedto lie along hydrogenbonds, and the length ofthe 0-H bonds is takenas 0.96 A./"\H20,H ,O - -H ,N... ..,NH s - -H ,O . .(V.1--?Ha -FH, ,crystal.32 The hydrogen atoms of hydrogen peroxidebeing assumed to lie along the strong hydrogen bondsdirected towards the carbonyl oxygens, the structureshown in Fig. 2 is obtained for hydrogen peroxide.This structure is in very good agreement with thatpredicted in 1934 by W. G. Penney and G . B. B. M.Sutherland 34 for this-molecule by a quantum-mechanical calculation by themethod of electron pairs.They found that a model of this type, with theH-0-0 angle and the dihedral angle both about loo", was more stable thaneither the cis- or the trans-configuration. The structure is also in agreementwith dipole-moment determination^.^^ The 0-0 distance obtained in thepresent investigation is 1.46 & 0.03 A., and this is in agreement with electron-diffraction determinations.364. ORGANIC STRUCTURES.Quite a large number of quantitative studies of organic crystals andmolecules have been made during the year. Electron-diffraction studieshave given interesting results on methylenecyclobutane. There are nounusual bond distances to report, but this is the first four-membered carbonring compound to be measured, and it is found that the carbon atoms arecoplanar, lying a t the corners of a square, with 90" bond angles.Thefifth carbon atom, attached by a double bond, lies on the extension of oneof the diagonals. The bond lengths are normal, within the limits of experi-mental error (& 0.03 A.), In hexamethylethane 1 the central C-C bondTrans. Paraday Soc., 1934, 30, 898; J . Chem. Phyaics, 1934, 2,492.36 E. P. Linton and 0. Masss, Canadian J . Res., 1932, 7 , 81.36 P. A. Gigubre and V. Schomaker, unpublished preliminary results.1 S. H. Bauer and J. Y. Beach, J . Amer. Chem. Boo., 1942, 64, 1142ROBERTSON : ORGANIC STRUCTURES. 105appears to be somewhat stretched (1.58 & 0.03 A,) but the other distancesare normal.The experimental results in this case are scarcely capable ofdistinguishing between models involving free rotation, eclipsed or staggeredconfigurations for the methyl groups, but the observations favour thelatter slightly.In view of the interesting carbon-carbon distances found in methyl-acetylene,2 the propargyl halides have now been examined: and the carbon-carbon single bond is again found to have the value 147 & 0.02 A., as inmethylacetylene. The carbon-halogen distances are somewhat larger thannormal, being 1.83, 1.95, and 2.13 A., for C-C1, GBr, and C-I instead of1.76, 1-91, and 2.10 A,, as calculated from the covalent radii. It is thoughtthat this is possibly due to contributions from the extra ionic structure (I)in addition to the usual ionic structure (11).H-C=C=CH, H-C32--CH2 +- + .' ..(I.) : x :- :x:- (11.1 ....(Miss) I. E. Knaggs and (Mrs.) K. Lonsdale4 have begun an X-rayinvestigation of the esters of the interesting series of polyenedicarboxylicacids of general formula CO,H*[CH:CH],*CO,H. Data are given for the firsttwo members, trans- trans-met h yl fumarate , C6H804, and trans- trans- methylmuconate, C,H,,O,. The two crystals are structurally similar and belongto the triclinic pinacoidal class with one centrosymmetrical molecule perunit cell (space group Pi). X-Ray intensity data, " diffuse spot " pheno-mena, and magnetic properties combine to show that the chain-like mole-cules are nearly, if not quite, planar, and lie approximately in the (100)crystal plane, The structures are obviously suitable for detailed analyses,which should yield results of interest.Preliminary data have also been given for the crystal structure ofbiotin,5 where there are four asymmetric molecules of C1,,H1,O,N,S in anorthorhombic unit cell, with a = 5-25, b = 10.35, c = 21.0 A., space groupdl-AZanine.-A very full and accurate account of the structure of dE-alanine has now been published by H.A. Levy and R. B. Corey.6 Theorthorhombic cell (C;,-Pm) contains four asymmetric molecules ofCH,*CH(NH,)*CO,H, and the difficulty of the analysis is increased by thefact that this complicated three-dimensional structure yields no clear pro-jection in which the positions of the atoms may be resolved by the usualFourier series methods.Nevertheless, the parameters have been carefullyrefined, and excellent agreements obtained between the calculated andobserved values of the structure factors. The experimental data have beenobtained by visual estimation of the intensities, using the Weissenbergcamera and multiple-film technique.'See Ann. Reports, 1939, 86, 175.L. Peuling, W. Gordy, and J. H. Saylor, J . Amer. Chem. Soc., 1942, 64, 1753.J . , 1942, 417.Ibid., 1941, 63, 2096.P2,2,2 1.6 I. Fankuchen, J . Amer. Chem. SOC., 1942, 04, 1742.See ref. ( 6 ) , Section 2106 URY STALLOQRAPHY,It is found that the molecules are linked together by a three-dimensionalframework of moderately atrong hydrogen bridges (about 2.8 A. in length)and this appears to be responsible for an abnormdly close approach (3.64 A.)of the methyl groups of adjacent molecules. Thearrangement is shown diagrammatically in (111).The close approach of the methyl groups imposedby the strong intermolecular forces probablyprevents these groups from undergoing anyrotational motion of the kind thought t o exist inother crystals when the methyl-methyl contactdistances are larger.8 One interesting feature of* (111.) .the present work is that probable positions havebeen assigned to all the hydrogen atoms in thestructure and their scattering contributions have been included in thecalculation of the structure factors. A small but significant improvement issaid to result from this procedure.The alanine molecule has more or less the expected configuration. Thecc-carbon atom and the carboxyl group are coplanar to within the limits ofexperimenta1 errors (for interatomic distances about 0.03 A., and for bondangles about 3").The G O distances in the carboxyl group are also sub-stantially equal, at 1-23 and 1.25 A., and are in excellent agreement withthose proposed for the oxalic acid dihydrate structure." The C-C distancesare normal at 1.54 A., but the C-N distance of 1-42 A., although similar tothat found in glycine (1.39 A.) and in diketopiperazine 10 (1.41 A,), is muchless than the value found for C-N single bonds in most structures, or thatcomputed from the oovalent radii.ll The reason for this contraction is notyet fully understood, but it is one of the significant experimental resultswhich should be considered when attempts are made to arrive at the con-figuration of polypeptide chains in protein structures.Melamine (Cyanuric Triamide) .-A preliminary account of this structureby (Miss) I.E. Knaggs and (Mrs.) K. Lonsdale was mentioned last yeax,12and now a full analysis has been given by E. W. Hughes.13 The mono-clinic unit cell (P2Ja) contains four molecules of C3N6H6, and all the para-meters have been determined and refined by several Fourier projections.The final result is shown in Fig. 1, which gives the molecular dimensions,The bond lengths are estimated not to be in error by more than 0.05 A.,and the molecule is planar. The significant feature is the approximateequa1it.y of all the C-N links, which indicate almost complete resonance ofthe three double bonds about the three carbon atoms, as in (1V)-(VI).The calculated bond length for equal contributions from these structures is1.343 A., which agrees very closely with the average value found (1.346 A.).I I2 .g ~ ..sac) .? %78 a'i.64 --o -----NH~,~CH~ t-.Q A0 . . . . . . . .8 L. Pauling, Physical Rev., 1930, 36, 430.J. M. Robertson, Trans. E'araduy SOC., 1940, 36, 917.V. Schoma,ker and D. P. Stevenson, J . Amr. Cbm. Sw., 1941, 63, 37.10 See Ann. Reports, 1939, 36, 179.l2 Ann.. Reporta, 1041, 38, 107. I t J . Amer. Chem. am., 1941,62, 1737ROBERTSON : ORaANIC STRUCTURES. 107Thia result is in agreement with the bond lengths about the guanidinecarbon atom in dioyansrnide,14 end those of the inner ring system of theYH, +#HaCs;sHa-N/ \N-I I I r//"\n Ilq.&/c\NA&rH, N*p\\N/C\NH, $&/b\N[c\&H,(W.1 (V.1 w.1phthalocyanine molecule,15 but in the ring of cyanuric triazide there appearto be alternate long and short bonds of 1.38 and 1.31 g.16FIU.1.The mshmirae mohulc. Smdl cir&8, carbons ; large circles, nitrogen ; hydrogemnot shopun. The molecule tk coplanar. Bond lengths are in A?agstr&n unit8.The packing of the molecules in the crystal is governed by the form-ation of hydrogen bridges between the ring nitrogen and the amino-groupof neighbouring molecules. These seem, however, to be of a rather weaktype, four pairs of distances varying from 3.00 to 3.10 A. being recorded.Melamine monohydrochloride hemihydrate is also briefly examined inthe same paper.13 The orthorhombic unit cell contains eight molecules,and there is no obvious resemblance to the melamine structure.Organic Xelenium Compounds.--J. D.McCullough and G. Hamburgerhave completed accurate studies of diphenylselenium dibromide and di-chloride.17 The compounds are of interest in connection with the effectof the unshared electron pair on the bond orientation and interatomicdistances. (In the simpler but similar compounds, tellurium tetrachloride l8and potassium fluoroiodate,lg the structures are probably of the trigonal1 4 E. W. Hughes, J . Amer. Chem. Soc., 1940, 62, 1268; Ann. Repork, 1940, 87, 192.lB J. M. Robertson, J., 1936, 1196.16 (Miss) I. E. b g g s , Proc.Roy. Soc., 1935, A, 150, 676.17 J . Amer. Chem. Soc., 1941, 63, 803; 1942, 64, 508.** D. P. Stevenson and V. Schomaker, W., 1940, 62, 2267.lB L. Helmholtz and M. T. Rogers, &id., p. 1637108 CRYSTALLOGRAPHY.bipyramidal type with the unshared pair in one of the equatorial positions.)The heavy atoms in the diphenylselenium dihalides have simplified theanalysis and made it possible to proceed by accurate Patterson and Fourierprojections. The molecules are found to have a two-fold axis of symmetry,and to approximate closely to the trigonal bipyramidal structure withselenium at the centre, halogen a t the apices, and the equatorial positionsoccupied by the phenyl groups and the unshared pair. Thus, for the di-bromide, Br-Se-Br = 180" 5 3", C-Se-C = 110" &'lo", Se-Br = 2.52 -j=0.01 A., Se-C = 1.91 & 0.03 A.The dichloride is not isomorphous withthe dibromide, the orthorhombic unit cell containing eight molecules ofSe(C,HJ2C1, instead of four in the case of the dibromide, but the molecularstructure is essentially the same, with Se-C1 = 2.30 & 0.05 A. The selen-ium-halogen distances are thus greater than the sum of/\,He\/\ the single-bond radii (Calc. : 2-31 for Se-Br and 2.16 forSe-Cl), the lengthening effect being due to the unsharedIn selanthren2* (VII) measurements have now beenmade, by means of a Fourier projection, of the angle offold about the line joining the selenium atoms, which give a result of 127".The C-Se distance is 1.96 A., and the selenium valency angle is about 96".Linear Polyesters.-A large number of trimethylene glycol polyesters ofhigh molecular weight have been prepared and their fibre patterns examinedby C.S. Fuller, C . J. Frosch, and N. R. Pape; 21 X-ray diagrams from boththe stretched and the unstretched fibres have been examined, and thepatterns obtained depend to some extent on the manner in which thesamples are prepared. It is concluded that the chain molecules are essen-tially of the planar zigzag type, but that the chains are tilted to the fibreaxis at an angle of about 30", rather than being coiled or kinked parallel tothis axis. This unexpected result appears to be connected with the layersof ester group dipoles present in the structure. Stretching the fibre causesa decrease in the angle of inclination of the chains to the fibre axis, witha re-orientation of the dipole layers.Gutta-percha, Rubber, and Po1ychloroprene.-A series of papers on themolecular structure and properties of these polymers has been publishedduring the year.The crystal-structure determinations are mainly in agree-ment with earlier work,22 but C. W. BunnB now gives an exhaustive dis-cussion and investigation of the permissible configurations of the isopreneunit in these structures, and attempts to correlate the results with thephysical properties of the substances. The X-ray data, obtained by Cu-Kecradiation from normal fibre photographs, consist of 24 observed reflectionsO \ s e / \ ' I pair.(VII.)20 R. G. Wood and G. Williams, Nature, 1942, 150, 321.J .Amer. Chem. Soc., 1942, 64, 164. Earlier papers; C. S. Fuller and C. L.Erickson, ibid., 1937, 59, 344; C. S. Fuller and C. J. Frosch, ibid., 1939, 61, 2576;J. Physical Chem., 1939, 43, 323.22 K. H. Meyer and H. Mark, Ber., 1928, 61, 1939.as Proc. Roy. Soc., 1942, A , 180, 40, 67, 82ROBERTSON : ORGANIC STRUCTURES. 109from p-gutta-percha, 40 from rubber, and 18 from polychloroprene; butonly about 10 of these reflections can be separately indexed for each struc-ture, the remainder being attributable to small groups of planes of closelysimilar spacings. Visual estimates of the intensities are given. Thestructures are descfibed in terms of the following units : p-gutta-percha,[-CH2-C(CH,)=CH-CH2-],, orthorhombic, a = 7.78, b = 11-78, c = 4.72 A,,space group P2,2,2, with four long-chain molecules parallel to the c axis;rubber, [-CH2-C(CH,)XH-CH2-In, monoclinic, a = 12.46, b = 8.89,c = 8.10, p = 92", space group P2,/a, with four long-chain moleculesparallel to the c axis ; polychloroprene, [-CH2-CCl=CH-CH2-],, completelyanalogous to p-gutta-percha, but with a = 8.84, b = 10.24, c = 4.79 A.The fundamental difference between gutta-percha and rubber is attri-buted (in accord with earlier suggestions) to the isomerism which arises fromthe presence of double bonds in the chains of 1 : 4-isoprene polymers.Thetrans-form (VIII) (with the chain bonds 1, 2, and 3, 4 on opposite sides ofthe double bond) has a repetition period of 5.0 A., normal interatomicdistances and bond angles being aasumed, while the cis-form (IX) (withchain bonds on the same side of the double bond) has a period of 9-1 A., if/(VIII; trans-.)all the chaincompoundeda one-period9.1 A.(IX; cis-.)carbons lie in a plane.Many different types of chain can befrom these isoprene units, but only the trammodel can giveunit, of about 5 A., and there appear to be only two types ofcis-model which give a two-period unit. The possibilities are thus con-siderably narrowed, and it is generally assumed that @-gutta-percha andchloroprene represent the tram-configuration (c = 4.7 A.) while rubberrepresents the cis-configuration (c = 8-1 A,). The author discusses theprecise form of these chains in considerable detail. In p-gutta-percha thechain carbons of each isoprene unit are thought to lie in a plane, as requiredby stereochemical considerations. To explain the observed periodicity(4.7 A.) and the X-ray intensities, however, it must be assumed that adjacentisoprenes are not coplanar, bond 1 ' 4 being moved out of the plane byappropriate rotation around bonds 3-4 and 1'-2', the isoprene units them-selves being kept parallel.In addition, the methyl group ( 5 ) is consider-ably displaced (by 0.65 A.) out of its ideal position in plane 1-2-3-4, th110 CIRYBTALLOORAPHY.main oame of this being repulsion by the adjoining -CH, group. In rubber,there am further deviations from the ideal structure, and the two isopreneunits which make up an identity period are probably not identical in con-figuration.[A weak (001) refleation is observed.] There are other distor-tions, including a certain non-planarity in the icloprene miits, which differfrom those of the P-gutta-percha structure. The polychloroprene structureis very similar to that of p-gutta-percha, except that the chlorine atomsuffers a somewhat greater displacement than that of the correspondingmethyl group in p-gutta-percha.The atomic positions arrived at in the above structures are set out indetail, and appear to afford a reasonable interpretation of the X-ray inten-sities; but owing to the limited nature of the X-ray data they can only beregarded ae approximate, and may have to be considerably modified oreven perhaps drastically altered if more precise data beoome available.This point &I mentioned, but does not appear to be sufEciently stressed, iRthese papers.It is, however, empbsised by the appearance of anotherpaper by C. J. B. Clews24 dealing with the structure of polychloroprene.In this paper, not only is there lack of agreement with Bunn regardingthe molecular codguration, but even the unit cell is assigned quite differentdimensions (u = 8.90, c = 12.21, b = 4-70 A,). It is difficult to reconcilethese two investigations. The peat need in this field at present is obviouslymore precise experimental work, rather than further speculation.Rubber can also be made to crystalhe by a homogeneous two-dimen-sional dilatation, and this state has been examined by means of an inter-esting X-ray study carried out on an inflated rubber balloon by A.Schalla-r n a ~ h . ~ ~ Using a monoclinic which differs somewhat from thatemployed by Bunn, Schallarnach shows that there is a tendency for thecrystallites to arrange themselves with the b c plane in the plane of theskin, the a; axis (8.5 A.) perpendicular, an& with a random distributionaround a.p-~~o~reizes~Z~ho.ne.-One promising method whereby further progressmay be achieved in the rubber problem lies in the accurate and systematicstudy of bond lengths and valeacy angles in relatively simple compoundscontaining isoprene units or related atomic groupings. Such preliminaryreconnaieaance should obvioudy precede any serious frontal attack on thepolyisoprene structuree themselves, and a good start in this direction hasbeen made by E.G. Cox and G. A. Jeffrey 27 in their analysis of P-isoprene-sulphone. This work has been carried out very carefully by quantitativeintensity measurements on the crystal, and three-dimensional Pattersonand Fourier syntheses, and owing to the presence of the relatively heavysulphur atom the results me largely independent of any purely chemicalevidence.The two moleoules of C5H80,S in the monoclinic cell (P2Jm) appear to24 Proc. Roy. SOC., 1942, A, 180, 100.*a W. Lotmar and K. H. Meyer, Mmatsh., 1936, 69, 115.S T Trans. Faraday SOC., 1942, 88, 241.2s Nature, 1942, 149, 112ROBERTSON : ORGANIC SITCUCTTJRES. 111exhibit a plane of symmetry. They have a heterocyclic struoture, andthe interatomic distances (accurate to 0.02 A.) show that resonanoe mustoccur between the carbon-carbon bonds in the C,S ring.The C-C distancesin the ring (compare X) are dl 1-41 A., the methyl group bond length isnormal ah 1.54 A., C-S is 1-75 A,, and S-0, 1.44 A. The distribution of thefour bonds about the sulphur atom is approximately tetrahedral, the0-S-0 angle being biaected by the plane of the carbon atoms. It is(X.) (XI.)suggested that the normal state involves resonance between (X) and theionised forms (XI) and (XII), the three structures making approximatelyequal contributions. This would involve 33 yo of double-bond characterin the C-C links and lead to a distance very nearly equal to the observedvalue. Another poasibility is that of '' hyperconjugation " as discussedrecently by R.S. Mullken, C. H. Riehe, and W. G. Brown,28 but the observednormal value of the methyl group bond is against this explanation.The intermolecular approaches are of the approximately normal van derWaals type, vix., 3-45-4.0 A. The successful elucidation of this structurewhich has now been achieved should have important bearings in connectionwith the study of vulcanisation processes and rubber chemistry generally.Protein Xtrluctzlres.-Recent X-ray work on keratin and myosin wasmentioned last year, and a more complete account has now been pub-li~hed.2~ X-Ray studies of some iodinated amino-acids and proteins30show that in some cases iodination appears to cause a structural rearrange-ment.Some extremely interesting work is now being carried out on the crystalstructure of haemoglobin by M. F. Perutz,3I based on the earlier work ofJ. D. Bernal, I. Fankuchen, and M. P. Perutz.32 When wet crystals ofhorse methaemoglobin (a = 109 & 0-5, b = 63.8, c = 55.1 A., p = 111.1")are dried they contract by over 30% of their wet volume, and the con-traction is found to take place wholly in the b plane, the length of the baxis remaining unchanged. Further, the contraction can be made to takeplace slowly and in stages, and the change in the X-ray diffraction patterncan be followed. It is found that there are large changes in the intensitiesof the (h01) reflections in the different states of hydration. Now, by com-puting these changes in terms of Patterson-Pourier projections, it shouldbe possible to distinguish those features of the structure which are definitelyof intramolecular origin from those which are due merely to intermoleculareffects. Thus, if the vector maps prepared at different shrinkage stages26 J. Amr. Chem. SOC., 1941, 08, 41.ao M. Spiegel-AdoIf, R. H. Hamilton, and G. C. Henny, Biochem. J., 1942, 30, 825.19 W. T. Astbury, J., 1942, 337.Nature, 1939,148,731; 1942,140,491. m Ibid., 1938,141, 623112 CRYSTALT-tOURAPHY.show peaks which coincide and are of closely similar shape, then it isprobable that they belong to the intramolecular type.By proceeding in this way, vector maps for different stages have beenprepared and examined. It is concluded from this and other evidence thatthe haemoglobin molecules (two per cell) form coherent sheets parallel tothe c plane (OOl), with layers of water and possibly ammonium sulphatelying between the protein sheets. On drying, the protein layers movetogether and slip over one another, thus increasing the monoclinic angle.Each molecule lies on a two-fold symmetry axis and consists either of twosheets 18 A. thick with water and possibly salt between, or of a single rigidsheet 36 A. thick possessing two-fold symmetry. More recent data, and acomparison with the crystal structure of horse oxyh~rnogl~bin,~~ lead tothe view that the molecule is a platelet with a roughly elliptical base, theaxis of the ellipse being about 48 by 64 A., and the thickness of the platelet36 A. This general method of analysis promises to yield further importantresults.J, M. ROBERTSON.83 Nature, 1942, 150, 324

ISSN:0365-6217    

DOI:10.1039/AR9423900095

出版商:RSC    

年代:1942

数据来源: RSC

摘要:

ORGANIC CHEMISTRY.1. INTRODUUTION.PROGRESS in stereochemistry has in the main continued on existing linesand no striking new developments have been noted. The boundaries ofstereochemistry, never very precise, continue to become more fluid. On theone hand, stereochemical work becomes more and more affected by quantita-tive developments of a physical or mathematical kind, and on the otherhand, the bearing of stereochemical considerations on other branches ofchemistry becomes increasingly evident. In this Report, no detailed dis-cussion is attempted of recent progress in particular branches of stereo-chemistry ; these have mostly been dealt with in recent years in other report8to which references are given.Instead, an account is given of the general trend of stereochemical work.The study of atomic and molecular dimensions by physical methods, thequantum mechanical treatment of valency problems and the development ofmathematical theories of optical rotatory power are now giving the subject aquantitative basis.On the more qualitative side, the study of various formsof isomerism continues. The use of stereochemical methods in the study ofreaction mechanism, particularly in replacement reactions (Walden inver-sion) and molecular rearrangements, is referred to and a brief survey isattempted of the somewhat related topics of asymmetric transformation,induced optical activity and asymmetric synthesis.The selection of material is particularly difficult in a report dealing withdevelopments in general methods. In the compilation of this section of thepresent Report, in spite of both the reduced amount of published matter andthe inaccessibility of various journals, this difficulty was ever present; aconsiderable selection of material was necessary even within the three topicschosen for consideration.The reporter has limited himself to a discussionof, first, boron fluoride catalysis, secondly, some metal enolate condensationsand fmally a pot pourri of generally unrelated reactions chosen because ofgeneral interest and general applicability.Evidence that lignin is largely a phenylpropane derivative has accumu-lated in recent years as the result of the efforts of H. Hibbert, K. Freudenbergand others, and the Report on lignin deals largely with this theme. After abrief account of extraction methods and such physical evidence asis avail-able, such as studies on the absorption spectrum, the general evidence as tothe predominantly aromatic nature of lignin is reviewed.This is followed byan account of the theoretical speculations of K. Freudenberg as to thestructure and building up of the lignin molecule as chains of substitutedpropylphenol residues containing in some instances chroman and in othersfuran rings, and the experimental evidence adduced in support of this theory.H. Erdtman’s idea that lignin is related to dehydroisoeugenol and th114 ORGAN10 CHEMISTRY.experimental evidence in favour of this view is then discussed, and the workof H. Hibbert and his school is summarised.This includes the breakdownof lignin by oxidative alkaline degradation to such products as vanillin andsyringic aldehyde, and the isolation by ethanolysis of products like a-ethoxypropiovanillone and syringoyl methyl ketone, which with the isolationof propylcyclohexanols by the hydrogenation of wood and lignins givesstrong support €or the phenylpropane theory of lignin structure. Hibbert’stheory that lignin is essentially a product, derived from plant respiratorycatalysts is also discussed. The report concludes with a summary of recentattempts to defhe the type of union between lignin and carbohydrates inwood, if such exists.The aotive study of polycyclic aromatic compounds in the last decade hasbeen stimulated by the intrinsic interest of their physico-chemical behaviour,by the cancer-producing activity of many of them (which can now be corre-lated in general.terms with structure), by their relation to (mainly hydro-aromatic) natural products, and by their importance in the dye indnstry.Reinvestigation of coal tar has disolosed the presence of many methyl homo-logues of naphthalene and phenanthrene, and a variety of polycyclic substances;and destructive hydrogenation of ma1 or tar oils, followed by dehydrogen-ation, provides a useful source of pyrene and chvene. Synthetic methodsinclude, in addition to significant variations and extensions of standardanthracene and phenanthrene syntheses and of the Diels-Alder reaction,numerous processes in which the ends of a four-carbon chain are attachedto the o-positions of an aromatic nucleus.Difficulties in the preparationand striking abnormalities in the behaviour of 4 : 5-diaubatituted phenan-threnes and similar compounds are attributed to spatial interference of thesubstituents.In linearly but not in angdarly fused systems, increase in the number ofrings remlts in enhanced chemical reactivity and depth of colour and inincreasing stability of the mmo-dihydro-compounds relatively to the parenthydrocarbons, which are known up to the deep green heptscene, C,,H,,.Various reactions suggest a degree of fixation of double bonds in these sub-stances. The revemible photochemical addition of oxygen, giving tb meso-peroxide, is a speoific property of anthraoene systems, whose meso-reactivityis further exhibited in their behaviour as dimes in the Diels-Alder reactionwith a great variety of partners, and in their ready substitution, notably byreactive diazo-oompounds.Much work has been done on substitution andorientation in, among others, phenanthrene, pyrene and ohrysene, and it isinteresting that carcinogenic hydrocarbons are hydroxylated in the animalbody in positions which are not those of normal substitution.In the heterocyclio series new or improved methods are reported for thesynthesis of derivatives of quinoline, thiophen and selenophen, piperidineand tetrahydro(thio)pyran, and of bicyclic syatema with a nitrogen atom asbridge-head, The selective hydrogenation of polycyclio systems containingnitrogen rings, the mode of condenmtion of thioxwphthenquinone8 withthioindoxyls to indigoid dyes, and the tranaform&tions of addones, espeoiallBALFE AND KENYON : STEREOCHEMISTRY.115to intensely chemilumineswnt derivatives of 5 : 6’-diacridyl, have beencarefully studied. G. Heller’s formula for indophenine is confirmed.The systematio investigation of plant constituents hm disclosed minorstructural novelties in flavan derivatives and their glycosides, and in bemyl-isoquinoline alkaloids ; more notable are the Osage orange pigments whichcontain an isoflavone system joined to a terpene residue, and the fusedindole and isoquinoline structure assigned to Erythrina basee. The constitu-tion of retronecine, the commonest basic moiety of the Selnecio alkaloids, isnow largely elucidated.A remarkable toxic polypeptide from the fungusAmanita phalloides contains a diastemisorneride of the common hydroxy-proline together with a, hydroxytryptophan.Pa. P. BALBE.J. w. COOK.J. KENYON.E. G. V. PERCIVAL.F. S. SPRING.T. S. STEVENR.2. STBREOOHEMISTRY.Stereochernidxy was laat included in these Reports itl 1939.1 In recentyears there have been important developments in the quantitative study ofstereochemistry (atomic and molecular dimensions, bond lengths and inter-valenoy angles) as was indicated in 1932 by the inclusion of Stereochemistryin the General and Phyaical Chemistry Section of these Reports.2 In 1937a treatment of the topic on these lines was included in the Organic ChemistrySection? In the Pedler lecture, 1942, W.H. Mills4 gave a, very welaomeoutline of the theoretioal bmis of stereochemistry-the relation between theelectronic structures of atoms and their valency configurations-in a simpli-fied form, for the encouragement of the organic stereochemist. N. V. Sidg-wick and H. M. Powell give a survey of the spatial arrangements of thecovalenoies of multivalent atoms and their relation to the electronic groupsoccupied by the valency electrons : they also list the various methods usedin the study of atomic structure and molecular configuration. Prominentamongst these methods is X-ray crystallography,6 which, incidentally, hasbeen used to distinguish between racemic mixtures and meso-compounds inthe ap-disthyldibenzyl series.’ W. T.Astbury 8 surveys the application ofX-ray and other methods to the study of the proteins, one of the numerousapplications of stereochemiatry.There is little scope for the organic chemist in these quantitative investi-1 P. Maitland, Ann. Report%, 1939,36, 247.2 N. V. Sidgwick, ibid., 1932, 29, 64.4 J., 1942, 467.5 Proc. Roy. Soc., 1940, A, 176, 153.6 J. M. Robertson, Ann. Reports, 1940,37,188; 1941,38,108; G. C. Hampson, ibid.,7 C. H. Carlisle and (Miss) D. Crowfoot, J., 1941, 7.I,. 0. Brookway and T. W. J. Taylor, ibid., 1937, 34, 196.1940,57,179; H. M. Powell, ibi&., 1941, $8,99.J., 1942, 337116 ORUANIC CHEMISTRY.gations, though A. Liittringhaus and collaborators use a chemical methodin the investigation of intervalency angles. They determine the number ofcarbon atoms necessary in chains attached to the two valencies, to permitring-closure between the two chains.If in a ring of type X<(cH2 (CH ) m>CH,,2)nthe number of carbon atoms (m + n + 1) and the carbon valency angle areknown, the valency angle of X< can be calculated. When X = CH,, thevalency angle >CH2 can be calculated from the number of atoms (m + n + 2)in the ring.Though, as the authors observe, the results may be affected by the deform-ability as well as the magnitude of the angles, they obtain values (CH2<112", S< 110") which are in harmony with those obtained by other methods.For _\S they obtain a value of 75", which, in comparison with the valuefor S< , they suggest indicates repulsion between the two negatively chargedoxygen atoms.Theories of Optical Rotatory Power.The relation between chemical constitution and optical rotatory power,and the effect of solvents on optical rotatory power, were for many yearsdomains of the organic chemist, whose treatment was necessarily somewhatempirical. Treatments of a more theoretical character are developing.Theories of optical rotatory power have been reviewed by W.J. Kauzmann,J. E. Walter, and H. Eyring,l* who discuss the interpretation of the effects oftemperature, solvents and molecular structure on rotatory power, and byW. J. Kauzmann and H. Eyring.ll C. 0. Beckmann and H. C. Marks l2give an equation for the influence of dipolar solvents on the rotatory powers ofoptically active compounds, which is in accord with observations on certainderivatives of (+)-tartaric acid and (-)-menthol, in a variety of mixedsolvents.Meanwhile, the organic chemists proceed with the provision ofdata, for eventual digestion by theory. B. K. Singh l3 and M. and A. Singh l4continue the examination of the relation between chemical constitution andoptical rotatory power, and an addition has been made l5 to the somewhatrestricted list of compounds to the rotatory dispersion of which a two-termDrude equation (see below) has been fitted. P. A. Levene and A. Rothen l6discuss the effect of structure on the rotatory dispersion and absorptionspectra of substances which contain phenyl or cyclohexyl groups.The examination of the effect of solvents on rotatory power has continued.-O/ O +YA.Luttringhaus and K. Buchholz, Ber., 1940, 73, 134; A. Liittringhaus andK. Hauschild, ibid., p. 145.lo Chep Reviews, 1940, 26, 339.12 Ibid., 1940, 8, 827, 831.lS B. K. Singh and A. B. La], PTOC. Indian Acad. Sci., 1940, 12, A , 157.14 J . Indian Chern. SOC., 1941, 18, 89.l5 R. S. Airs, M. P. Balfe, 5. M. Irwin, and J. Kenyon, J., 1942, 531.l6 J . Chem Physics, 1939, 7 , 975.l1 J. Chem. Physics, 1941, 9, 41BALFE AND KENYON : STEREOCHEMISTRY. 117A. W. H. Pryde and H. G. Rule l7 report that there is no simple relationbetween the optical activities and refractive indices of solutions of certainhydrocarbons [( -)-dimenthyl, (-)-dibornyl and (+)- and (-)- woc&m-phanes], giving references to other workers who have suggested that suchrelationships should exist.They observe that the same conclusion can bedrawn from the measurements by J. Kenyon and B. C. Platt l8 of therotatory powers of solutions of ( +)-y-methyl-n-heptane in various solvents.The latter authors point out that their results show the existence of solventeffects on rotatory power, even when solvent and solute are both of non-polar type.As T. M. Lowry emphasised in his text-book,ls in investigations of whichthe main object is the recording of rotatory powers, it is highly desirable thatmeasurements should be made over a wide range of wave-lengths. It is to behoped that the practice of making observations up to the limit of ultra-violet transmission will develop, and w i l l be supported by the observationof absorption spectra, at least in those cases where the rotatory dispersionshows complexity.Unfortunately, facilities for the measurement of rotatorypower outside the visible spectrum are at present available in only a fewlaboratories.Stereoisomerism.The application of adsorption methods to stereochemical problems hasreceived some attention. (Miss) M. M. Jamison and E. E. Turner20 havedescribed the partial separation of the diastereoisomerides of ( -)-menthy1 &Z-mandelate by preferential adsorption on alumina. Separations of opticallyactive forms from racemic mixtures by selective adsorption on opticallyactive 'solids were referred to in 1939 (G. Karagunis and G. Coumou10s,21triethylenediaminochromitrichloride, Cr en3C13, on quartz ; and G.M.Henderson and H. G. Rule,22 p-phenylenebisiminocamphor on lactose).R. P. Linstead and collaborators 23 discuss the stereochemistry of catalytichydrogenation. Using derivatives of diphenic acid and of phenanthrene assubstrates, hydrogenation being conducted principally in acetic acid solutionin which platinum oxide catalyst is suspended, they conclude that thehydrogen atoms add to one side of the molecule only (&-hydrogenation).trans-Hydrogenation is impossible, since the other side of the molecule is pro-tected through adsorption on the surface of the catalyst. This work, thoughnot the first, is probably the most important contribution to the topic.F. E. Ray and S. Palinchak 24 report the existence of the anion of 9-aci-nitro-2-benzoylfluorene in an optically active form, through the separationof its diastereoisomeric brucine salts.It was believed that the centre ofasymmetry in such ions is a tercovalent, negatively charged carbon atomlinked to the nitrogen atom (as shown in I), but on the whole this structurel 7 J., 1940, 345.lo " Optical Rotatory Power," London, 1935.* l Nature, 1938, 142, 162.24 Ibid., 1940, 62, 2109.** J . , 1939, 633.2o J., 1942, 611.22 J . , 1939, 1568.J . Amer. Chem. Soc., 1942, 64, 1985, 1991, 2003, 2006, 2009, 2014, 2022118 ORGANIC CHEMISTRY.is improbable (see T. W. J. Taylor and W. Baker 25). F. E. Ray and S. Palin-chak,2P finding +hat the anion remains optically active only when the saltcontains a moleoule of alcohol of crystdimtion, have sdopted a suggestion ofR.L. Shriner and J. H. Young 26 that the configuration of the asymmetriccmbon atom is stabilised by the formation of a fourth (co-ordinate) link tothe hydroxylic group of the alcohol molecule (11). An alternative, andequally plausible, explanation was put forward by T. W. J. Taylor and W.Baker,25 zlix., that the salt is not formed by substitution, but is an additioncompound (111). In the formation of both (11) and (111), the four valencies ofthe carbon atom need not be disturbed, so the optical activity of the uci-nitro-compounds can no longer be regarded as relevant to optical stability of terco-.&O?"O Na,H*OR(111.)valent carbanions. It is indeed by no means obvious that a structure such~ E I (I) would retain the optical aotivity of the CF atam.Ten electrons areavailable for formation of the three bonds of the nitrogen atom. Each bond(including the carbon-nitrogen bond) will, therefore, approximate moreolosely to a double bond than to single bond, and the effect of this on thetetrahedral structure of the carbon atom is open to question.J. B. Kass and S. 33. Radlove describe the four diastereoisomeric9 : 10 : 12-trihydroxyatearic acids. R. Adams, C. M. Smith, and S. Lowe,@from (+)-3-methylcyclohexanone (previously described by 0. Wdaoh 29 ; eeealso T. M. Lowry, D. M. Simpson, and C. B. Allsopp m), and the (-)-enan-tiomorph [obtained by resolution of the dl-form via the (-)-menthyl-hydrazone], have prepared the (+)- and the (-)-form of 1-hydroxy-6 : 6 : 9-trimethyl-3-n-amyl-7 : 8 : 9 : lO-tefrahydro-6-dibenzop~an (IV), which isrelated to the active principles of mmihuana or b h i s h .They find that the(-)-form of (IV) has a higher narcotic activity than the (+)-form. 0. Leaf,A. R. Todd, and S. Wilkinson 31 also have prepared (+)-(IV) from (+)-3-methylcyclohexanone and found its narcotic activity to be less thm thwt ofthe &form.Q - p 5 H l l (n) CH,--V-CH@ HOHY2H6Me2-0(IV. ) (V.)26 N. V. Sidgwick's " Organic Chemistry of Nitrogen," Oxford, 1937, p. 237.26 J. Amer. Chem. Soc., 1930,52, 3332.28 Ibid., p. 2087.so Prm. Roy. SOC., 1937, A, 162, 483.a T Ibid., 1942,64, 2263.119 Annalun, 1896, a@, 840.s1 J., 1942, 186BALFE AND KENYON : STEREOCHEMISTBY.119J. Chrtt and F. G. Mam 82 find confirmatory evidenm for the tetrahedralconfiguration of the 3-covalent arsenio atom, in the isolation of two forms(due to folding along the As-As axis) of 5 : 10-dihydroarsanthren. J. F.Kincaid and F. C. Henriques,a on the basis of calculations of the energyrequired for inversion of the molecules, conclude that NR‘R”R”’ is irre-solvable, on account of ease of racemisation, unless the nitrogen atom formspart of a ring; PR’R”R”’ should be resolvable if chemical difficulties can beovercome ; S@R’R‘‘R”’ is optically stable and racemises by decomposition ;CRR‘R”R”‘ can only be racemised by mechanisms which involve the break-ing of bonds. These conclusions are in general agreement with experimentalevidence.The search for optical activity due to symmetrically placed hydrogenand deuterium, reviewed in 1939,l has continued, though the general trendof the recent papers is to provide confirmatory evidenoe that optical activitycannot be observed in such compounds.The most unequivowl method isthat of H. C. Brown and C. Groot,3* namely, the introduction of deuteriuminto an optically active compound in such a way that two of the groupsattached to the wymmetric carbon atom become structurally identical butisotopically distinct. Starting fkom optically active amyl rtlcohol, theyprepared (V); it will be observed that disturbance of the bonds of theasymmetric carbon atom does not oocur, 80 racemisation cannot be due tothis cause.The observed rotatory power of their produot, in a 1 dm. oolumn,was less than 0.005” and probably less than 0.002°. Other oommunicationson this topic are by G. R. Clemo and G. A. Swann 85 and by H. Erlenmeyerand 0. B i t t e r l b ~ . ~ ~Optical activity due to the molecular dissymmetry which ariaes fromrestrioted rotation1 continues to receive attention. The my-first of theexhauetive studies of R. Adams and collaborators 37 deals with substituteddiphenic acids in which the 5- and the 5’-position are joined by bridges con-taining from 6 to 10 oarbon atoms; the optically active forms of these oom-pounds have half-life periods of some 20 minutes. The authors also deal withrestricted rotation in arylamines 38 (e.g., VI),Me Me in which free rotation about the bond attach- Iing the nitrogen atom to the ring is pre-MeQMe vented by interference between the sub-stituents on the nitrogen atom and the twomethyl groups which are o- to that bond;the compound thus exists in two enantiomorphous forms, in one of whichthe CO*[CH],*CO,H group is above the plane of the benzene ring and in the/ -N*CO*[CH,],*CO,Hm.1sa J., 1940, 1184.33 J .Amw. Chem. Soc., 1940,62,1474; other papers which approach the subject in 8similar way were reviewed in 1939 (ref. 1).84 J. Amer. Chem. SOC., 1942, 64, 2663.36 J . , 1942, 370.37 R. Adams and N. Kornblum, J . Amer. Chem. SOC., 1941, 63, 188,38 Ibid., 1940, 02, 2191 ; 1941, 03, 2589; 1942, 64, 1475.36 Helv. Chim. Aota, 1940, 23, 207120 ORGANIC CHEMISTRY.other, below it.The compounds of this type have half-life periods of some4 to 30 hours (for other work on this type of restricted rotation, seeP. Maitland l). R. Adams and collaborators 39 also describe the resolutionof compounds of the type (VII), which have half-life periods ranging from9 minutes to 70 hours. G. Wittig, A. Oppermann, and K. Faber40 haveresolved (VIII), which has a half-life period of 43 minutes.Me MeThe Walden Inverszion.The Walden inversion, first observed in 1895, was for many years regardedas a stereochemical mystery, for which, its discoverer observed 34 years later,41no satisfactory solution had been found. I n the early examples investigated,complicating factors existed, such as the presence of carboxylate groups inthe reacting molecules, or heterogeneous reaction conditions (use of silveroxide for hydroxylation of halides), and to these complications many of theearly difiiculties may be ascribed.Subsequent developments 42 have clari-fied the general principles which underlie retention of configuration? race-misation or inversion of configuration during aliphatic substitution reactions,and the stereochemical outcome of these reactions can now be used, alongwith other methods, to elucidate their mechanism.At the same time, the detailed study of what were once called " Waldeninversion reactions " continues. W. Hiickel and H. Pietrzok 43 have studiedthe reaction of phosphorus pentachloride with (-)-menthol, in the presenceand in the absence of tertiary bases.P. A. Levene and A. Rothen 44 find thata t temperatures below - 40" reaction of methyl- and ethyl-phenylcarbinolswith hydrogen bromide occurs without change in sign of rotation, or inversionof configuration, possibly owing to formation of an addition complex (IX),which decomposes into the bromide with retention of con-R H -8' figuration (this is analogous to the mode of decompositionPh>'< O<g of certain sulphinoxy chlorides and chloroformates sug-gested by W. A. Cowdrey, E. D. Hughes, C. K. Ingold,S. Masterman, and A. D. Scott 45 as an explanation ofretention of configuration). P. A. Levene and A. Rothen 44 are of the opinionthat certain of their results invalidate some of the conclusions of E. D.J . Arner. Chem. SOC., 1940, 62, 53; 1941, 63, 1589, 2773; 1942, 64, 1786, 1791,(IX,)1795.40 J .pr. Chem., 1941, [ii], 158, 61.41 P. Walden, " Salts, Acids and Bases," London, 1929.42 See reviews by H. B. Watson, Ann, Reports, 1938, 55, 218, and E. D. Hughes,Trans. Faraday SOC., 1938, 34, 202.43 Annalen, 1939, 540, 250.I 4 J . Biol. Chem., 1939,127,237 ; see also P. G. Stevens and N. L. McNiven, J . Arne?-.4 6 J . , 1937, 1252.Chem. SOC., 1939, 61, 1295BALFE AND KENYON : STEREOCHEMISTRY. 121Hughes, C. K. Ingold and their collab0rators,4~ but this is probably due totheir adopting a too limited point of view. As E. D. Hughes, C. K. Ingold,and I. C. Whitfield 46 point out, the effect of solvent and other experimentalcircumstances on the stereochemical outcome of any given reaction is acornbination of the effects on three possibly concurrent reaction mechanisms,which result in inversion of configuration, racemisation, or retention ofconfiguration.It is therefore to be expected that, when the conditions arevaried, complicated but not incomprehensible changes in the course of thereaction may occur.P. D. Bartlett and collaborators 47 show that compounds (X) and (XI)Hwill not undergo certain replacement r'eactions, such as halogenation of thealcohols (X = OH) or hydrolysis of the halides (X = Cl). This lack ofreactivity is attributed to steric factors. Inversion of configuration of thecarbon atom (C*) is imp,ossible, because it forms the junction of two rings.The racemisation reaction, which involves separation of a carbonium kation,cannot occur because the carbon atom (C*) cannot adopt the planar con-figuration which is essential in its kationic state.P. D. Bartlett and L. V.Rosen4* also ascribe the lack of reactivity of neopentyl halides to sterichindrance of the methyl groups in the neopentyl radical, which prevents theclose approach of a reacting ion.S. Winstein and collaborators 49 report a study of the r6le of neighbouringgroups in replacement reactions. They conclude that in the reaction of2 : 3-dibromobutanes or 1 : 2-dibromocyclohexanes with silver acetate (inacetic acid), removal of one bromine atom from the dibromide (XII) isaccompanied by formation of the intermediate (XIII), which has a (presum-ably mesomeric) ring structure.The formation of this intermediate andits reaction to give the bromo-acetate (XIV) are both accompanied byinversion of configuration on C, ; these two consecutive inversions necessarily,/BP.,\ 8' Rr >A=?< O z>c-c< O >>C--?<Br OAc(XII.) (XIII.) (XIV.)result in retention of configuration. Similarly the reaction of the bromo-acetate (XIV) with silver acetate proceeds through a ring-structuredintermediate (XV), which is formed and opened with inversion of configur-413 Nature, 1941, 147, 206. 4 7 J . Amer. Chem.Soc., 1939,61, 3184; 1940,62, 1183.4 8 Ibid., 1942, 84, 643; see also A. G. Evans and M. Polanyi, Nature, 1942, 149,49 J . Amer. Chern. SOC., 1942, 84, 2780, 2787, 2791, 2792, 2796.608, 665122 ORGANIC CHEMISTRY.ation, 80 that the acetate (XVI) again hrcs the same configuration &B thebromide from which it is derived.3 O\/O ____ O +O\/O - > L K X--KOAc c\ v+(XVI.)CH3 C%(XIV.) (XV.)When the solvent contain8 a smdl amount of water, a large proportion ofthe reaction product is monoacetate of inverted codguration.To explainthis, it is suggested that, by addition of water, the intermediate (XV) isconverted into (XVII), which loses a proton and then, by ring opening asshown by the dotted line in (XVIII), gives the monoaoefate (XIX).(XVII.) (XVIII. ) ( X X . ) h (XV.1In (XIX), C , has the inverted configuration because it has undergone oneinversion reaction [formation of the ring (XV)] and one reaction with reten-tion of configuration (XVIII + XIX).Since in any of the ringopeningreactions, either the bond to C, or that to C, may be broken, mixture.s ofisomers may result if the compounds have the necessary lack of symmetry.The above mechanisms apply also to the reactions of tram-2-acetylcyclo-hexyl p-toluenesulphonate with silver acetate, and the hypothesis of inter-mediate ring (XIII) formation is used by the authors to explain retention ofconfiguration in the bromination of 3- bromo-2-butanol with phosphorus tri-bromide. The assumption that the neighbouring group always intervenes,and that the resulting intermediate is both formed and reacts with inver-sion of configuration is also used to explain the formation of the trans-formof 1 : 2-dibromocyclohexane by a variety of reactions, viz., the reaction ofbromine with cyclohexene, of hydrogen bromide with the cis- and the trans-diacetate, with cyclohexene oxide, and with the 1-bromo-2-p-toluenesul-phonate, and the bromohydrins.Two examples, starting fiom the trans-diacetate (XX) and cydohexene oxide (XXV) must suffice to show theapplication of the mechanism :ixxv.)93 Br(XXVI. )o,,(XXVII a )H>C-Br 1 Br>C-H(XXrnI.BALFE AND KENYON : STEREOCHEMISTRY. 123Neighbouring groups play a particularly important part in substitutionreactions in the sugar series.50 In the alkaline hydrolysis of sugar p-toluene-sulphonates, an intermediate of ethylene oxide type ia formed, and reach,with the same configurational and isomeric changes described above, if theneighbouring group is tram- to the toluenesulphonyl group, and is convertedinto >CH*Oe by the alkali.Such groups are OH, O*COR, and O-SO,R. Ifthe neighbouring group is resistant to the alkali, or cis- to the toluenesul-phony1 group, the reactivity of the latter is much reduced, and it reacts, ifat all, without inversion. G. J. Robertson and W. for example,have found that when 3-p-toluenesulphonyl-4-awtyl-6-triphenyhethyl-Z-methyl a-methylaltroside is treated with dilute alkali solution, the acetylgroup is removed but the toluenesulphonyl group remains. With less dilutealkali, decompositions occur, but in no caBe is the 3 : 4-anhydro-derivra$iveformed, because there is not, adjacent to the p-toluenesulphonyl group, atrans-substituent which is converted into >CH*Oe by reaction with the alkali(the acetyl group is cia- to the 3-toluenesulphonyl group, and the tram-2-substituent, methoxyl, is resistant to a&&).W.E. Grigaby, J. Hind, J. Chanley, and I?. H. Westheimer 52 show thatinversion of configuration occurs during the opening of the ethylene oxidering in cyclopentene oxide by reaction with sodiomalonic ester (XXIX ---+XXX).H H HH H OHMolecular Rearrangements.Molecular rearrangements form another topic in which the study ofstereochemical changes can be used as a mode of investigation. It isestablished that, when the migrating radical (specifically, the carbon atomtherein which breaks one bond itnd forms another) is o p t i d y active, theoptical activity is lost in intermolecular rearrangements and retained inintramolecular rearrangements.The occurrence or not of Walden inversionduring the intramolecular rearrangement depends on whether the carbon inquestion carries a sextet or an octet of electrons during its migration. Sub-sequent to the review of the subject in 1941,53 the paper by H. I. Bernsteinand E. S. Wallis54 may be noted, wherein are discussed some molecularrearrangements of optically active truxillamic a d s .R. S. Airs, M. P. Balfe, and J. Kenyon 65 desoribe the rearrangement ofSee review by S . Pea9 Ann. Reports, 1939, 36, 258 ; also R. S. Isbell, Ann. Rev.Biochem., 1940, 9, 65.51 J., 1940, 319.53 H. B. Watson, Ann. Reporta, 1941, 38, 121.li2 J. Amer. Chem. SOC., 1942, 64, 2606.J.Org. Chem., 1942,7, 261.5 6 J., 1942, 18; see also M. P. Bslfe and J. Kenyon, Trans. Farday Soc., 1941, 37,721124 ORGANIC CHEMISTRY.y-methyl-a-ethylally1 alcohol to a-methyl-y-ethylallyl alcohol (XXXI --+XXXII) , which is analogous to the rearrangement of a-phenyl-ymethylallylalcohol and its hydrogen phthalate to the corresponding y-phenyl-a-methyl-Et \C/CH\\C/MeH/\OH \H(XXXI.) Et\c+/Me (XXXII.)H/ HO \Hally1 derivative^.^^ A n interesting feature of these intramolecular reactionsis that in the product of the rearrangement, the asymmetric carbon atom isnot the asymmetric atom of the original compound, yet a high degree ofoptical activity is retained during the rearrangement.Configurational Relationship.Since the relative configurations of members of series of compounds ”b>c<g, in which X is varied, cannot be deduced directly from the signs oftheir rotatory powers, the experimental investigation of relative configura-tions has long been a feature of stereochemical In the period underreview, P.A. Levene and M. Kuna 57 have deduced the following relations :(i) propionic series, (- )-CH,*CHR*CO,H and (+)-CH,*CHR-CH,*NH, havethe same relative configurations ; (ii) butyric series, ( +)-CH,-CH,*CHR*CO,Hand ( +)-CR,*CH,-CHR*CH,*NH, have the same relative configurations (Rvaries from C,H5 to CGH13) ; (iii) (-) C5H11 CH*[CH,],*CO,H, (+) C H 2 -C~!&>CH.CK2*C02H, ( - ) C5f!&>CH*C0,H have the same relativeconfiguration.Thoughthis convention was suggested in 1898,5* it has not been universally accepted.The present position is that those authors who use (+) and (-) apply themto sign of rotation, and d and I , if used in conjunction with (+) and ( -),refer to relative configuration.When d and 1 are used alone, however, itwill generally be found that they apply to sign of rotation.It is to be observed 59 that relative configurations can only be assignedinside series in which only one of the groups attached to the asymmetriccarbon atom is variable. If two, or more, groups vary, as in i>C<i(XXXIII) and E>w; (XXXIV), then (XXXIII) may be converted into(XXXIV) by two routes (c -+ e and d --+ f, or c -+ f and d _c, e)and the assignment of relative configuration between the two compounds willdepend on the arbitrary choice of one or other of these routes.In this Report, (+) and (-) are used for sign of rotatory power.6 6 J.Kenyon, S. M. Partridge, and H. Phillips, J., 1937, 207.5 7 J . Biol. Chem., 1941, 140, 225, 259.6 8 H. Landolt, “ Optisches Drehungsvermogen,” 2nd ed., 1898; see also 2ndK. Freudenberg, “ Stereochemie,” Leipzig, p. 062 ; see also S. M. Partridge, J.,English ed., “ Optical Rotatory Power of Organic Substances,” 1902.1939, 1201BALFE AND KENYON : STEREOCHEMISTRY. 125Asymmetric Tramformation.When in a pair of diastereoisomerides (+)A(-)B and (+)A(+)B, thecomponent B is optically unstable, the two forms will differ in stability, sothat at equilibrium they will be present in unequal amounts in solution.The subject was reviewed in 1939,l largely on the basis of some interestingdevelopments by M.M. Jamison and E. E. Turner,GO whose work has beencontinued.61 In their latest paper,62 these authors record a case [the brucinesalt of 2’4 a-hydroxyisopropyl)diphenyl-2-carboxylic acid] where theyhave been able to observe the ‘‘ first order ” and ‘‘ second order ” trans-formations (R. KuhnG3). The “ first order ” effect is the establishment ofthe equilibrium between the diastereoisomerides in solution, and the ‘‘ secondorder ” effect refers to the case where one isomeride separates from thesolution. This development opens up the possibility of further study of the‘‘ van’t Hoff-Dimroth rule,” which relates the relative stability and solu-bility of a pair of interconvertible isomerides in a given solvent.Induced Optical Activity.Rotatory dispersion (variation of rotatory power with wave-leugth oflight) can usually be expressed by a one-term or a two-term Drude equation,lg[cc)~ = kl/(h2 - A:) or [a12 = k J ( ~ 2 - A?) - k,/(~2 - A:)in which the constants ’A, and ’A, are frequencies characteristic of certainbands in the absorption spectrum of the substance.These “opticallyactive ’ ’ absorption bands are those which exhibit circular dichroism (unequalabsorption of right and left circularly polarised light), and although theremay be more than one such anisotropic absorption band associated with eachoptically active centre in the molecule, bands in the visible or near ultra-violet regions have a dominant effect.lo Thus it may be stated, thoughperhaps with a certain degree of over-simplification, that there is usually oneterm in the Drude equation for each optically active centre in the molecule,though, if there is overlapping between the anisotropic bands associatedwith different centres of asymmetry, their resultant may be covered by asingle term.In some substances, however, a two-term equation is requiredfor the rotatory dispersion, though there is only one formal centre of asym-metry. Since such substances frequently contain groups of the type >C = 0,>C = C<, it has been suggested (see, e.g., T. M. Lowry 65) that under theinfluence of the fixed asymmetric centre, a condition of“ induced asymmetry”arises in these unsaturated groups, which tend towards a semi-polar character6o J., 1938, 1646.62 Idem., J., 1942, 437.61 (Miss) M.M. Jamison and E. E. Turner, J., 1940, 264.Ber., 1932,65,49; as M. M. Jamison and E. E. Turner (ref. 62) point out, the termsare unfortunately chosen, as the phenomena bear no necessary relation to first- andsecond-order kinetics.*‘ 0. Dimroth, AnnaZen, 1910, 377, 127; 1913, 390, 91 ; see also A. Smits, “ Theoryof Allotropy,” London, 1922, p. 69; T. M. Lowry, C. A. H. MacConkey, and H. BurgessJ . , 1928, 1333.Op. oit. (ref. 19), pp. 374, 389, 411126 OR(3AMC OHEMISTRY.@ 9 Q 9 >C - 0, >C - C<. Though this view may in some cases be valid, theourrent trend of opinion is awBy from it. For example, E. Erlenmeyer’sclaims to have obtained cinnamic acid in optically active forms (in whichthe “ induced aBymmetry ” haa become permanent) are now regarded asresting on an insecure experimental basis. F.Eisenlohr and G. Meier 67ascribe the optical activity to impurities; other work pointing in the samedirection is reviewed by P. D. Ritchie.68 Further, there are cases, of which(+)-tetrahydrofurfury1 alcohol l5 (XXXV) is an example, in which nosuch induced asymmetry can be postulated. Thealternative hypothesis 89 that, in the presence of a$\C-CH,.OH fixed centre of asymmetry, anisotropy may be in-duced in absorption bands associated with other parts(XXXV.) of the molecule, is more generally applicable. Theoptical activity derived from the induced anisotropy may, in any givenregion of the spectrum, be an important or even the main contribution tothe rotatory power, according to the relative magnitudes of the constants,El, k,, A,, A,, and so long as the fixed centre of asymmetry remains, the“ induced optical aotivity ” remains constmt.The hypothesis of “ induced anisotropy,” though not based on the simpleconcept of “ induced dissymmetry,” still depends on the structure of certaingroups in the molecule.For example, W. C. Price 70 observes that the pro-perties of the x: molecular orbital (which is occupied by two of the electronsof a double bond) give rise to the optical activity of the >C=O bond incerfain ketones. T. M. Lowry, D. M. Simpson, and C. B. Allsopp 30 suggestthat the induced optical activity of the > G O groups should be attributedto the non-bonding (unshared) electrons of the oxygen atom, but that it maystill be described as induced in the double bond, because “ the act of rotationof the plane of pohrised light takes place when the electron is raised to itsexcited state, the orbital of which is located in the double bond.”H-c H-crnc-H ‘0’Asymmetric Synthesis.Absolute asymmetric ayntheais 68 is the synthesis of optically activesubstances from materiais which do not contain 8ources of asymmetry.T.L. Davies and R. Heggie 71 b d that the oombination of bromine withtrinitrostilbene, in a beam of right circularly polarised light of A 3600-4 5 0 0 ~ . gives a dextrorotrttory product. A solution of 0.043 g, of trinitro-stilbene in 5 c.c, of glaoial acetio acid had, after brominrttion in this way, arotatory power of 0.04” in a 1.5 cm.column. The rotatory power is notpermanent, possibly owing to the chemical instability of the trinitrostilbenedibromide which is presumed to be the optically active produot. Although6 6 See, e.g., Biochem. Z., 1912, 43, 446. 67 Ber., 1938, 71, 1005.68 “ Asymmetrio Synthesis and Asymmetric Induction,” Oxford, 1933.69 W. Kuhn, Ber., 1930, 6% 190; W. Kuhn, K. Freudenberg, and J. Wolf, ibid.,70 Ann. Reporta, 1939, 36, 62.p. 2370.7 1 J. Amer. Chem. BOG., 1936, 57, 377, 1622BALFE AND KENYON : STEREOCHEMISTRY. 127the observed rotatory power is low, it is of the order of magnitude predictedby an expression 72 based on the circular dichroism and rotatory power of thesubstance.Partial asymmetric synthesis 1n 68* 73 was defined by W.Marckwald 74 as asynthesk which produces optically active compounds from substances ofsymmetrical constitution, optically active substances being used as inter-mediates, but without any process of fraotionation. Processes based ondifferences between diastereoisomerides (e.g., differences in rate of formation,rate of decomposition, stability or solubility) are mostly excluded by thisdefinition.have made an extensive studyof asymmetric synthesis based on a-ketonic esters. For( -)-menthy1 benzoylformate (XXXVI), by interaotion with a Grignardreagent, gives the ester of atrolactinic acid (XXXVII) and on removal ofthe (-)-menthol, it is found that the acid (XXXVIII) is optically active.A.McKenzie and collaborators 689 75*(XXXVI.) (XXXvII.) (XXXVIII. )A. McKenzie has suggested 77 that the formation of the (-)-acid(XXXVIII) may be primarily due t o the existence of (XXXVI) as an un-equal mixture of diastereoisomerides, in which the ketonic group exhibitsinduced dissymmetry, the mutarotation of (XXXVI) which occurs in dco-holic solvents being connected with the attainment of equilibrium betweenthe two forms. M. M. Jamison and E. E. Turner 78 have, however, concludedthat the mutarotation is due to solvation or hemiacetal formation. It ismore in harmony with current views to ascribe asymmetric synthesis to thesame causes as induced anisotropy. The electronic disturbance which causesanisotropy of an absorption band associated with the unsaturated centre mayresult in an unsymmetrical opening of the double bond during the additionof the entering atom (or ion) and so result in the formation in unequalamounts of the two possible stereoisomeric configurations.The following reactions may be classed as asymmetric syntheses : form-ation of only one of the diastereoisomeric glycols from (-)-benzoin by inter-\dOH0H(XXXIX.) H/ \cJ/ --+ H/ \CJ--Ph (XL.)Ph \C/OH* Ph\Ph \Meaction with a Grignard reagent 79 (XXXIX --+ XL) ; formation of opticallyactive methyl ap-dibromo- p-phenylethyl ketone from ( +)-y-phenyl-cr-methyl-72 W.Kuhn and E. Knopf, 2,physikaZ. Chem., 1930, B, 7, 292.7 1 E. E. Turner, Ann. Repo~t8, 1936, 23, 234.76 Ergebn. Enzymforxh., 1936, 5, 68.7 7 See, e.g., A.MoKenzie and A. D. Wood, J., 1939, 1330.7 8 J., 1941, 638.7@ A. McKenzie and H. Wren, J., 1910,97, 473.7 p Ber., 1904, 37, 1368.76 A. McKenzie, J., 1904, 85, 1249128 ORGANIC CHEMISTRY.ally1 alcohol (XLI -+ XLII) ** ; and the anionotropic rearrangementsof type (XXXI) + (XXXII).55* 56 There are differences between theseMe*CH( 0H)GH:CHPh 4 Me*CH( OH)*CHBr*CHBrPh --+(XLI.)Me*CO*CHBr*CHBrPh(XLII. )four types of reaction. In (XXXVI) -+ (XXXVIII), the original centre ofasymmetry is recovered unchanged when the (-)-menthol is removed fromthe hydroxy-acid by hydrolysis; in (XXXIX) --+ (XL), the originaland the new centre of asymmetry both remain in the final product; in(XLI) --+ (XLII), the original centre of asymmetry is destroyed so thatthe optical activity of the new centre can be observed without interference;in (XXXI) --+ (XXXII), the asymmetry of the original centre is lost atthe moment when that of the new centre is formed.These, however, arepoints of detail, and should not obscure the main fact that in each of the fourcases an asymmetric centre is formed in an optically active condition, underthe influence of an optically active centre already existing in the molecule,M. P. B.J. K.3. GENERAL METHODS.Boron Fluoride Catalysis.The catalytic activity of boron fluoride has been demonstrated for avariety of reaction types during the last twelve years mainly as a result ofthe researches of J. A. Nieuwland, G. F. Hennion and their collaborators.In general, boron fluoride compares with aluminium chloride in versatilityand in many cases it is superior to the latter in giving higher yields, cleanerproducts, and increased velocity of reaction.There are, moreover, anumber of reactions which are catalysed by boron fluoride but not to anappreciable extent by aluminium chloride.Addition CompZexes.--H. Meerwein has shown that boron fluoride uniteswith water to give a dihydrate, BF3,2H,0, which can be distilled withoutdecomposition and is characterised by the formation of crystalline salts withsome ethers; the dihydrate adds another mole of boron fluoride to givea monohydrate, BF3,He0, which, however, decomposes on distillation.Similarly with alcohols boron fluoride gives two series of addition complexes.The series BF3,2ROH can be distilled without decomposition, but the seriesBF3,ROH decompose on distillation.These complexes are extremelystrong acids. In the case of methyl alcohol, one mole of boron fluoride isabsorbed, giving the unstable complex, BF3,MeOH, which is probablymethoxyfluoboric acid; it has been characterised as its mercuric salt,Hg[BF3,MeO] 2. Distillation of methoxyfluoboric acid gives a mixture of1 H. Meerwein and W. Pannwitz, J . pr. Chem., 1934, 141, 123; H. Meerwein, Ber.,2 V. Gasselin, Ann. Chim. Phys., 1894, [vii], 3, 5.80 J. Kenyon and S. M. Partridge, J . , 1936, 1313.1933, 66, 411SPRING : UENERAL METHODS. 129BF2*OMe and the complex BF3,Me20, which is probably the methyl ester ofmethoxyfluoboric acid.3* This is a colourless liquid, b.p. 128", most readilyobtained by mixing dimethyl ether and boron fluoride. Ethyl alcohol anddiethyl ether behave in the same way; the complex BF,,Et,O, a liquid, b. p.126-127", is widely employed as a, catalyst, since it affords a convenientmethod for storing boron fluoride and it has not the inherent disadvantagesassociated with the handling of a toxic gas.Boron ffuoride forms addition complexes with acids, RC02H,BF3 and(RoCO~H)~,BF~, esters, R*C02R,BF3, and with amides, R*CO*NH,,BF3.5*With acetic anhydride, boron fluoride yields a complex of diacetoaceticanh~dride.~ These are fuming liquids, or solids, and they can replace boronfluoride as catalyst for many reactions, a remark which also applies todihydroxyfluoboric acid, HOB( OH),F,.'Esteri$cation.-The complex (CH,*CO,H),,BF, is a very efficient catalyst forthe preparation of esters from carboxylic acids and alcohols and the complexCH,*CO*NH2,BF3 is a good acetylating agent, reacting vigorously withalcohols 9 and with phenols : l oCH3-CO*NH2,BF, + R*OH -+ CH,*CO,R + BF3,NH3Boron fluoride has been employed as a catalyst in the preparation of celluloaeesters.llEsters can readily be obtained by treatment of an acid with an olefin inthe presence of a boron fluoride catalyst under mild reaction conditions, themethod being applicable to both aliphatic and aromatic carboxylic acids.12Camphene and acetic acid give isobornyl acetate l3 and cyclohexene and thesame acid give cydohexyl acetate.14 The complex BF3,Et20 is a particularlyactive catalyst for this type of e~terificati0n.l~Interesting observations on the simultaneous nuclear alkylation ofaromatic acids during the boron fluoride catalysed esterfication with olefinshave been made.Propylene quantitatively esterifies benzoic acid with theL. A. O'Leary and H. A. Wenzke, J. Amer. Chem. SOC., 1933,55,2117.S . Sugden and M. Waloff, J., 1932, 1492.H. Meerwin and D. Vossen, J. pr. Chem., 1934, 141, 149.H. Bowlus and J. A. Nieuwland,:J. Amer. Chem. SOC., 1931,53, 3835; H. Meerwein,Ber., 1933, 66, 411.T. €3. Dorris, F. J. Sowa, and J. A. Nieuwland, J . Amer. Chem. SOC., 1938, 60, 656;J. W. Kroeger, F. J. Sowa, and J. A. Nieuwland, ibid., 1937, 59, 996; U.S.P. 2,192,015.8 H. D.Hinton and J. A. Nieuwland, J . Amer. Chem. SOC., 1932, 54, 2017; F. J.Sowa and J. A. Nieuwland, ibid., 1937, 68, 271; T. B. Dorris, F. J. Sow&, and J. A.Nieuwland, ibid., 1934, 56, 2689; D. M. Smith, W. M. D. Bryant, and J. Mitchell, ibid.,1940, 62, 1, 4, 608.9 H. D. Hinton and J. A. Nieuwland, ibid., 1933, 55, 5052.10 U.S.P. 2,036,353. U.S.P. 2,113,293.12 T. J3. Dorris, F. J. Sow&, and J. A. Nieuwland, J. Amer. Chem. Soc., 1934, 56,2689 ; U.S.P. 2,065,540.l3 D.R.-P. 589,779.14 L. Brunel, Ann. Chim. Phys., 1905, [viii], 6, 216; H. L. Wunderley and F. J.16 8. V. Zavgorodnii, Acta Univ. Voromgiensis, 1938, 10, [2], 41.Sow&, J . Amer. Chem. SOC., 1937, 59, 1010formation of iaopmpyl benzoate, nuclear alkylation not occurring ; this esteris stable in the presence of boron fluoride.12 With salioylic aoid, on the otherhand, propylene reaota in the presence of boron fluoride to give ikopropylsaJioylate, which when gently warmed in the presence of the catalystmmnges to 2-hydroxy-3-~opropylbenzoic mid.Continued treatment ofaalicylio acid with propylene in the presenoe of boron fluoride gives a theo-retical yield of isopropyl2-hydroxy-3 : 5-diisopropylbe~0ttte.1~O l e h will esterify p-nitro- and o-chloro-bemoic aoids, phenylacetic andfuroic acids, but o- and p-aminobenzoic acids do not react. Prupylene, thebntylenes and the amylenes all give either 8ec.- or tert.-a&yl derivativee-never primary-and cyckpropane yields n-propyl esters.1' With the com-plex BF,,Et,O as catalyst, carboxylic acids can be esterified by meam3 ofethers, the catalyst effecting cleavage of the ether : l8BF EtO R*CO,H + Et,O 4 R*CO,Et + EtOHA similar ether cleavage by means of boron fluoride in the presence of aceticanhydride has been described by H.Meerwein and H. Maier-Huser : 19Yet another noteworthy catalysed synthesis of an erJfer ha8 been describedby J. F. M c K ~ M ~ and F. J. Sowa 2o in which a nitrile is treated with andcohol, the ortho-ester being an intermediate :R-CN + 3R'mOH 3 R*C(OR'), 5 R*CO,R' + R',OBoron fluoride also catalpes the acidolysis of esters.21Syntheses of Acids and Esters.-Employkg a boron fluoride catalyst,attractive methods for the preparation of acids and esters have been de-veloped by E. I. du Pont de Nemours.Carboxylic acids are obtained directlyfrom alcohols and carbon monoxide at 125-180°/ca. 500 atms., the catalystconsisting of boron fluoride and a limited amount of water. Methyl alcohol 22and ethyl alcohol 23 give acetic and propionic acids respectively. The method,which is applicable to polyhydric alcohols, ethylene glycol yielding succinicacid, can also be applied t o olefins, ethylene when treated with carbonmonoxide under similar reaction conditions giving propionic acid.% Astriking development of this method is the preparation of esters by theinteraction of an ether and carbon monoxide, dimethyl and diethyl ethersgiving methyl acetate and ethyl propionate resp~tively.~5Synthesis of Nit~iles a d Szlbstitutd Ami&es.-Acid amides are convertedl6 W.J. Croxall, F. J. Sowa, and J. A. Nieuwland, J . Amer. Chsm. SOC., 1934, 66,1' T. 33. Dorris and 3'. J. Sowa, J . Amer. Chem. SOC., 1938,00,358.Is G. F. Hennion, H. D. Hinton, aad J. A. Nieuwlaad, a i d . , 1933,55,2867.lB J . pr. Chem., 1932, 134, 51.21 F. J. Sowa, ibid., p. 654.2Q U.S.P. 2,170,825; 2,162,469; 2135,464; B.P. 636,422; 516,477.24 U.S.P. 2,135,459.25 U.S.P. 2,136,450; 2,136,449; 2,136,447.2054; J . Org. Chem., 1937, 2, 253.2o J . Amer. Chem. Soo., 1938, 00, 124.22 U.S.P. 2,135,448; 2,135,451-2 and -3SPRING : GENERAL METRODS, 131into nitriles when treated with boron fluoride. Thus treEtfment of acetamid8with the complex CH3*CO*NH2,BF3 in the presence of a small quantity of acarboxylic acid yields acetonitrile :CH3*CO*NH2,BP3 + CH3-CO*NH2AcOH Me*CN 4- Me*C02H + BF3,NH3The complex Me*CO*NH,,BF3 can also be used for preparing subatitutodacetarnidea 26 by intermtion with batm :RoNHR' + Me*CO*NH2,BF3 --+ NRR'-CO*Me + BF3,NH,Mercury-catalysed Additive Reactions of Acetylenic Hydrocarbons.-Thecomplexes MeOH,BF, (methoxyhoboric acid) and Et,0,BF3 are muchsuperior to sulphuric acid or any other acid in catalysing interaction betweenacetylene and its derivatives and hydroxyl-containing compounds in thepresence of mercuric oxide.These complexes are stable, readily dissolvemercuric oxide, and are effective in very low concentration^.^^ In thepresence of this mixed catalyst, methyl alcohol readily reacts with acetyleneand monoalkylacetylenes, yielding dimethylscetal 28 and dimethylketds 3Orespectively :BF,,E40 R-CiCH + MeOH R*C(OMe),*CH,The same catalyst being used, ethylene glycol reacts with acetylene withproduction of a dioxole derivative : 31The addition of monohydric alcohols other than methanol cannot besuccessfully effected by this wtalyst, but the applicability of the method isconsiderably increased if the catalyst contains a small amount of trichloro-acetic acid.With this modified catalyst, straight-chain primary alcohols(e.g., EtOH, n-PrOH, n-BuOH, rt-pentanol and n-hexanol) readily add toalkylacetylenes with formation of the corresponding ketal, but the additionof branched-chain or iso-alcohols t o alkylacetylenes cannot be achieved.32The trichloroacetic scid-promoted catalyst is also of value in the condensationof polyhydric alcohols and a-hydroxy-acids with alkylacetylenes ; thus2' F.J. Sowa and J. A. Nieuwlmd, J. Amer. Chem. SOC., 1937,59, 1202.5. A. Nieuwland, R. R. Vogt, and W. L. Foohey, ibid., 1930, 52, 1018; T. H.Vaughn, H. Bowlus, and J. A. Nieuwland, Proc. Indiana Acad. Sci., 1931, 40, 203.H. D. Hinton and J. A. Nieuwlasd, J . Amer. Chem. SOC., 1930,52, 2892.29 D. B. Killian, G. F. Hennion, and J. A. Nieuwland, ibid., 1934, 56, 1384.3O G. F. Hennion, D. B. Killian, T. H. Vaughn, and J. A. Nieuwland, &id., p.31 T. H. Vaughn, Proc. IndicCrccz Acad. Sci., 1933, 42, 127; F. G. Hennion, D. 33.32 D. B. Killian, G. F. Hennion, and J. A. Nieuwlmd, J. Amer. Chem. SOC., 1936,58,33 Idem, ibid., p. 1658.1130.Killian, T.H. Vaughn, and J. A. Nieuwland, J. Amer. Chem Soc., 1934, 56, 1130.80.REP.-VOL. XXXIX. I132 ORGANIU CHEMISTRY,~~-hydroxyisobutyric acid and butylacetylene give 2 : 5 : 5-trimethyl-2-butyl-1 : 3-dioxol-4-one :Alcohols also add to acetylenic alcohols in the presence of a trichloroaceticacid-promoted BF3,Et,0/Hg0 catalyst; thus dimethylethynylcarbinolyields 3 : 3-dimethoxy-2-methyl-2-butanol, hydrolysis of which with dilutemineral acid gives 2-hydroxy-2-methyl-3-butanone : 34OH*CMe,*CiCH + MeOH -+ CMe,( OH)*CMe(OMe), + CMe,(OH)*COMeA general mechanism for mercury-boron fluoride catalysis has beensuggested by G. F. Hennion, R. R. Vogt, and J. A. Nieuwland;S6 oneimportant experimental observation in this connection is the ease of additionof methyl alcohol to dialkylacetylenes, e.g.,BF8 /Hgo C,Hll*CiCMe C,H,,*C( OMe) ,*CH,Mefrom which it is concluded that the intermediate formation of a mercuryacetylide is not a prerequisite of reaction.36The catalysed addition of methyl alcohol to vinylacetylene (I) is par-ticularly interesting. In the presence of the BF3,Et,0/Hg0 catalyst (bestresults being obtained in the presence of trichloroacetic acid promoter) threemoles of methyl alcohol react with formation of a methoxy-dimethylketal37(111).The reaction probably involves a primary 1 : 4-addition7 followed byrearrangement of the dene to the acetylene (11), which reacts further togive the methoxyketal (111) (2 : 2 : 4-trimethoxybutane) : 38MeOH(I.) CH,:CH*CiCH ---+ [MeO*CH,*CH:C:CH,] +MeOH(11.) MeO*CH,*CiCMe ----+ MeO*CH,*CH,*CMe( OMe), (111.)The last step has been experimentally reali~ed,3~ and the mechanism receivesconsiderable support from the facts that ally1 alkylacetylenes (IV) react withmethanol in the presence of catalyst to give unsaturated ketals (V), theethylenic linkage remaining unchanged,*O and that methyl vinyl ketone (VI)readily adds methanol in the presence of the same catalyst to yield methylp-methoxyethyl ketone (VII) (4-methoxy-2-butanone) .41 The method ofpreparation of the boron fluoride-mercuric oxide catalyst is of particularimportance in this type of addition to a vinylacetylene.A BF,,Et,O/HgOcatalyst promoted by trichloroacetic acid being used, the method is limited34 J.F. Froning and G. F. Hennion, J . Amer. Chem. SOC., 1940,02,653 ; H. Scheiblerand A. Fischer, Ber., 1922, 55, 2903.36 J . Org. Chem., 1936, 1, 159.36 G. F. Hennion and J. A. Nieuwland, J . Amer. Chem. SOC., 1935,57,2006.37 D. 33. Killian, G. F. Hennion, and J. A. Nieuwland, ibid., 1934, 56, 1786.38 D. B. Killian, Q. F. Hennion, and J. A. Nieuwland, ibid., 1936, 58, 892.38 G. F. Hennion and J. A. Nieuwland, ibid., 1936, 57, 2006.40 D. B. Killian, G. F. Hennion, and J. A. Nieuwland, ibid,, 1936,58, 1668.*l U.S.P. 2,010,828SPRING : GENERAL METRODS. 133to the addition of methyl alcohol. If the catalyst, however, is prepared withmethyl alcohol [BF3,Et,0 is dissolved in methyl alcohol and then warmedwith mercuric oxide, the mercuric salt of methoxyfluoboric acid,Hg(MeO*BP3),, being formed] and promoted with trichloroacetic acid, thehigher alcohols can be successfully added to vinylacetylene (and its homo-logues) to give 2 : 2 : 4-trialko~ybutanes.~~Acetylene and alkylacetylenes also add carboxylic acids in the presenceof a boron fluoride-mercuric oxide ~atalyst,~3 the substituted vinyl ester(IV.) CH,:CH*CH,*CiCR + 2MeOH -+ CH,:CH*CH,*CH,*C( OMe),R (v.1(v1.1 CH,:CH*COMe + MeOH + MeO*CH,*CH,-COMe (VII.)(VIII) being first formed.A second addition occurs probably throughthe ester (IX), leading to a methyl ketone and an acid anh~dride.4~ Thereaction can be controlled to give good yields of either the vinyl ester or theR C O H R*CO,H + R’*CiCH + R*CO*O*CR’:CH, -4(VIII.)(IX.) [(RCO*O),CR’Me] --+ (R*CO),O + R’COMeketone-anhydride mixture.In the case of the vinyl ester, boron fluoride inthe presence of hydrogen fluoride is an effective promoter of the mercurycatalyst .46Carboxylic acids are also added to vinylacetylenes in the presence of atrichloroacetic acid-promoted BF,,Et,O/HgO catalyst ; e.g., vinylacetyleneand acetic acid yield the acetate of A1:3-butadien-2-01.~7Polymerisation of Ethylenic Hydrocarbons.-The catalytic polymerisationof ethylenic hydrocarbons was first observed by A. Butlerow and W.Gorianow 48 in 1873, but more than fifty years elapsed before this catalyticactivity attracted attention. In 1927, M. Otto 49 showed that boron fluoridecatalyses the polymerisation of both ethylene and propylene 50 and that thispolymerisation is promoted by the addition of finely divided metals such asnickel.Polymerisation of isobutene occurs in the presence of a varietyof acidic catalysts, the most effective being boron fluoride, aluminiumchloride and titanium ~hloride.~l The reaction with boron fluoride takesplace with almost explosive violence and is among the most rapidly completedof organic reactions. Under suitable conditions a practically quantitativeyield of a rubber-like polymer is produced in a fraction of a second ; a polymer42 R. 0. Norris, J. J. Verbane, and G. F. Hennion, J. Amer. Chem. SOL, 1939,61,887.43 G. F. Hennion, D. B. Killian, T. H. Vaughn, and J. A. Nieuwland, ibid., 1934, 56,4 4 G. F. Hennion and J. A. Nieuwland, ibid., p. 1802.4 5 D.R.-P.590,237.4 7 J. H. Wemtz, J . Amer. Chem. SOC., 1936, 57, 204.4 a Annalen, 1873, 169, 147.6o U.S.P. 2,183,603; B.P. 293,487.61 M. Otto and M. Mueller-Cunradi, U.S.P. 2,084,501; F. A. Howard, U.S.P.2,049,062; P.K.Frolich,U.S.P.2,109,772; D.R.-P. 278,486; H. C.ZimmernandE. W.Carlson, U.S.P. 2,074,093; H. G. Schneider, U.S.P. 2,131,196.1130.4 6 U.S.P. 1,912,608.49 Brennstoff-Chern., 1927, 8, 321334 OROIANIO OHEFMISTRP.of moleoular weight greater than 400,000 can be obtained.62 This reactionhaa an added theoretical interest in that the velocity of reaction is notapparently decreased by lowering the temperature and the molecular weightof the polymer is greater the lower the reaction temperature. The presenceof very small quantities of diimbutylene decreases both the yield and themolecular weight of the polymer, indicating that the normal dimer cannotbe present either during or preceding the polymerisation process.Triiso-butylene has a similar effect and neither it nor the dimer can be catalyticallypolymerised to high molecular weight products. The boron fluoride catalystis very readily poisoned by small amounts of hydrogen halides, an effectwhich is probably due to the formation of dimer and consequent inhibitionof the true polymerisation.To obtain efficient temperature control, an inert solvent can be employedand it is observed that the rnoleclslar weight of the polymer produced inoreweswith the amount of diluent s4 up to a limiting concentration of approximately80% diZuent by volume.Further dilution leads to a dramatic decrease inthe molecular weight. R. M. Thomas and his associates 53 believe that thepolymer has a head-tail structure with one terminal ethylenic linkage :....The estimation of unsaturation (based upon the Cverage molecular weight)agrees with this hypothesis.Concerning the mechanism of polymerisation it has been suggested 55that the catalytic activity of boron fluoride involves the activation of theethylenic link by association with the catalyst. A natural consequence ofthis mechanism is that boron fluoride should establish an equilibrium betweensuitable cis-trans isomers. It has been shown that boron fluoride does infact convert cis-stilbene into an equilibrium mixture containing 92.2% ofthe tram isomer : 56+ BF3 Ph-!?H+BF3 4 9HPh + H - r P hPh-C-H CHPh - Ph-C-H + -BF3AZkyZ~tion.~~-A solution of boron fluoride in sulphuric acid catalysess2 R.M. Thomas, J. C. Zimmern, L. B. Turner, R. Rosen, and P. K. Frolich, Ind.s3 R. M. Thomas, W. J. Sparks, P. I<. Frolich, M. Otto, and M. Mueller-Cunradi, J.64 U.S.P. 2,176,194.G 6 C. C. Price and J. M. Ciskowski, J . Amer. Chem. SOC., 1938, 80, 2499.O 7 See also Ann. Report8, 1941, 88, 118.Eng. Chem., 1940, 32, 299.Amer. Chem. SOC., 1940, 62, 276.C. C. Price and M. Meister, ibid., 1939, 61, 1696SPRmu : UENERAL METHODS. 135the alkylation of aromatic hydrocarbons by ~lefine.~* With propylene,isopropylbenzenes are obtained, the diisopropylbenzene fraction being almogtentirely the p-isomer, whereas in the corresponding alkylation using alumin-ium chloride, the m-isomer is the major dialkyl-pr0duct.5~ Aromatichydrocarbons can also be alkylated by means of alcohols in the presence ofboron fluoride ; since n-propyl alcohol and isopropyl alcohol both yieldisopropylbenzenes and 12-butyl alcohol and sec.-butyl alcohol both yieldset.-butylbenzenes, J. F.McKenna and F. J. conclude that thereaction proceeds by dehydration of the alcohol to an olefin and addition ofbenzene to the latter according to the Markownikoff rule. Later, N. F.Touasant and G. F. Hennion 60 showed that such alkylations are facilita;tedby the presence of dehydrating agents such as phosphoric oxide, sdphuricacid, and benzenesulphonic acid, the effect of which is most marked in thecase of primary alcohols.Boron fluoride likewise catalyses the alkylation ofnaphthalene by means of alcohols, the products being p-alkylnaphthaleneswith the exception of the case of benzyl alcohol, which yields mainly a-benzylnaphthalene.61 The dehydration mechanism is contested by C. C.Price and M. L ~ n d , * ~ who show that alkylation of benzene with d-sec.-butylalcohol in the presence of boron fluoride yields sec.-butylbenzene which islaevorotatory, and also by C. C . Price and J. M. Ciskowski.61 The latterauthors show that cycbhexanol is unaffected by boron fluoride under con-ditions under which cyclohexylation of naphhhalene by means of this alcoholproceeds rapidly.Aromatic hydrocarbons can also be alkylated by means of ethers 63 andesters 64 in the presence of boron fluoride.Phenols are readily alkylated by means of ethylenic hydrocarbons in thepresence of boron fluoride.Propylene and phenol give phenyl isopropylether,65 which on standing in oontact with the catalyst rearranges to 2-is0-propylphenol; at 40°, the major product is 2 : 4 : 6-triisopropylphenyl 60-propyl ether.66 Phenols can also be alkylated by means of alcohols in thepresence of boron fluoride,67 and by means of ethers with BF3,Et,0 as cata-lyst;68 methylation of phenol with BF3,Me,0 can be enforced to givepentamethylanisole in high 635 8 S. J. Slanina, F. J. Sowa,, and J. A. Nieuwland, J . Amer. Chem. Soc., 1935,57,1574 ;H . L. Wunderley, F. J. Sowa, and J. A. Nieuwland, ibid., 1937,&8,271; V.N. Ipatieffand A. V. Grosse, &id., 1936, 59, 2339; V. N. Ipatieff, B. B. Corson, and H. Pines, ibid.,p. 919.5Q T. M. Berry and E. E. Reid, ibid., 1927, 49, 3142.6o Ibid., 1940, 62, 1146. 61 Ibid., 1938, 60, 2499.63 M. J. O’Connor and F. J. Sowa, ibid., 1938, 60, 125.64 J. F. McKenna and F. J. Sow&, ibid., 1937,519, 1204.6 5 T. B. Dorris, F. J. Sowa, and J. A. Nieuwland, ibid., 1934,56,2689.6 6 F. J. Sowa, H. D. Hinton, and J. A. Nieuwland, a i d . , 1932, 5% 2019; 1933, 55,F. J. &low&, G. F. Hennion, and J. A. Nieuwland, ibid., 1935, 57, 709; J. h’.c 8 G. F. Hennion, H. D. Hinton, and J. A. Nieuwland, ibid., 1933, 55, 2857.6o A. J. Kolka aad R. R. Vogt, ibid., 1939,61,1465.tiOa Ibid., 1937, 59, 470.82 Ibid., 1040, 68, 3105.3402.McKenna and F.J. Sowa, ibid., 1938, 80, 124136 ORGAKIC CHEMISTRY.V. N. Ipatieff and A. V. Grosse 70 have shown that in the presence of waterand finely divided nickel, boron fluoride catalyses a reaction between olefinsand paraffins whereby higher paraffis are produced :a reaction which is more conveniently effected by means of sulphuric acid.71Olefins have also been used in the presence of boron fluoride to alkylatealicyclic hydrocarbons. 72Miscellaneozls Reactions.-In the presence of a boron fluoride catalystalcohols react with cyclic oxides to give a monoether of a diol; ketones andaldehydes also react with cyclic oxides in the presence of boron fluoride,yielding ketals and acetals respectively. 73 Boron fluoride also catalyses areaction between an olefin and a glycol to give a monoether of the glyco1,74and in the presence of the same catalyst methyl alcohol adds to trimethyl-ethylene thus : 75CHMeXMe, + MeOH --+ CH,Me*CMe,*OMeA preparation of amino-ketones from Schiff bases has been described byH.R. Snyder, H. A. Kornberg, and J. R. R ~ r n i g . ~ ~ Thus benzylideneanilinereacts with BF,,Et,O to yield a complex, which rapidly reacts with a methylketone, forming an amino-ketone :Ph*CH:NPh -> Ph*yH*NHPhJ. CH,*CORRSCONeBF3In the presence of boron fluoride, acetic anhydride condenses withacetophenone and with acetone to give benzoylacetone and acetylacetonerespectively.5An interesting observation has been made by D. S. Breslow and C.R.Hauser,?' who show that boron fluoride can replace basic condensing reagentsin Claisen, aldol and Michael type condensations.Boron fluoride is a very effective catalyst for the sulphonation andnitration of aromatic hydrocarbons and their derivatives.78Metal Enolate Condensations.SodizLm Triphenylmethyl .-Esters of the type CHR,*CO,R' fail to undergothe normal Claisen condensation in the presence of sodium ethoxide ;7O using70 J . Amer. Chem. SOC., 1935, 57, 1616.71 A. E. Dunstan and S. F. Birch, Trans. Faraday SOC., 1939, 35, 1013; A. E. Dun-72 V. N. Ipatieff, V. M. Komarewsky, and A. V. Grosse, J . Amer. Chem. SOC., 1936,7s A. A. Petrow, ,T. Qen. Chem. Russia, 1940, 981.74 U.S.P. 2,198,046.'13 J . Amer. Chem. SOC., 1939, 61, 3556.78 R.J. Thomas, W. F. Anzilotti, and G. F. Hennion, Ind. Eng. Chem., 1940, 32,79 S . M. McElvain, J . Amer. Chem. SOC., 1929, Sl, 3127.stan, Nature, 1940, 46, 186.57, 1722.76 U.S.P. 2,197,023.7 7 Ibid., 1940, 62, 2385, 2389, 2611.408BPRING : CIEXERAL METHODS. 137sodium triphenylmethyl, however, W. Schlenk, H. Hillemann, and 1.Rodloff 8oprepared the sodium enolate of methyl diphenylacetate and showed that,by treatment of the product with acid chlorides or with alkyl halides, thecorresponding acyl- or alkyl-diphenylacetic ester was produced. C. R.Hauser and W. B. Renfrow 81 find that this reaction is generally applicableto disubstituted acetic esters; for example, in the presence of sodiumtrip henylmet h yl , ethyl isobut yr at e yields ethyl is0 bu t yr yliso but p a t e :CHMe,*CO,Et + CPh,*Na + [CMe2C02Et]Na (1.) + CHPh,[CMe,*CO,Et]Na + CHMe,*COX -+ CHMe,*CO*CMe,*CO,Et + NaX(11.)The sodium enolate (I) will also condense with isobutyryl chloride to yieldthe same ester (II).82 The ester (11) yields a sodium enolate when treatedwith sodium triphenylmethyl and this on treatment with acetyl and Go-but yryl chlorides gives the diketo- esters Me* CO *CMe ,*CO*CMe2-C0 ,Et (111)and CHMe,*COGMe,*CO*CMe,*CO,Et (IV) respectively.83 (111) is notchanged by treatment with sodium triphenylmethyl, but (IV) is cyclised tohexamethylphloroglucinol, a reaction not accomplished by the weaker base,sodium ethoxide.Sodium triphenylmethyl has been widely employed in the preparation ofthe sodium enolates of esters and anhydrides.84 Attention is directed totwo general methods that have been developed by C.R. Hauser and hisassociates. First, trisubstituted acetic acids or their esters can be obtainedby treatment of a, disubstituted acetic ester with sodium triphenylmethyl,followed by treatment of the sodium enolate with an alkyl halide :85CHR,*CO,Et + CPh,Na + [CR,*CO,Et]Na + CHPh,Secondly, the ethyl acetoacetate ketone synthesis has been extended byB. E. Hudson and C. R. Hauser 88 to allow of the preparation of ketones ofthe type R’COCHR,. A disubstituted acetic ester is converted into itssodium enolate by means of sodium triphenylmethyl ; treatment of theenolate with an acid chloride gives an aa-disubstituted p-keto-ester, hydro-lysis of which gives the required ketone :[CR,*CO,Et]Na + R’*COCl+ R’*COCR,*CO,Et 3 R’COGHR,AZLyZ Carbonates .T h e carbalkoxylation of ethyl acetate by means ofethyl carbonate had been attempted unsuccessfully by W. Wislicen~s.~~Annalen, 1931, 48’7, 135.J . Amer. Chem. Soc., 1937, 59, 1823; 1938, 80, 463; C. R. Hauser, ibid., 1938,[CR,*CO,Et]Na + R’X -+ CR,R‘*CO,Et + NaX60, 1957; B. E. Hudson, R. H. Dick, and C. R. Hauser, ibid., p. 1960.82 Organic Syntheses, 1939, 19, 44.83 B. E. Hudson and C. R. Hauser, J . Amer. Chern. SOC., 1939,61,3567.E. Muller, Annalen, 1931, 491, 251 ; E. Muller, H. Gawlick, and W. Kreutzmann,ibid., 1934, 515, 97 ; B. E. Hudson, R. H. Dick, and C. R. Hauser, J . Amer. Chem. SOC.,1938, 60, 1960; C. R. Hauser and D.S. Breslow, ibid., 1939, 61, 793.85 R. E. Hudson and C. R. Hauser, J . Amer. Ohm. Soc., 1940, 62,2467.88 Ibid., 1941, 68, 3156, 3163.a 7 Ber., 1887, 20,2930; 1894, 27, 795; Annalen, 1888, 246,313138 ORQANIC CHEMISTRY.Later, by a modification of the reaction conditiom and using powderedsodium as condensing agent, H. Lux 88 obtained the required ethyl mdonatein 18% yield, and a similarly low yield of ethyl phenylmalonate was obtainedfrom ethyl phenylacetate.89 More recently, better yields have been obtainedin this type of carbalkoxylation reaction by using potassium or alloys ofpotassium and sodium.90 A. C. Cope and S. M. McElvain 91 and R. Connor 92showed that malonic esters undergo alcoholysis when heated in the presenceof sodium ethoxide; e.g.,which is a reversal of the carbalkoxylation reaction.Recognising thisfact, V. H. Wdlingford and his co-workers 93 have developed a fairly generalmethod for the preparation of substituted malonic esters from monocarb-oxylic esters. The essential featurea of the new procedure are that the alcoholformed is removed from the reaotion mixture by distillation 94 and that anexoess of alkyl carbonate is employed as solvent. Carbdkoxylation of ethylacetate with ethyl carbonate gives mainly trioarbethoxymethane. On theother hand, C. R. Hauser, B. Abramovitch, and J. T. Adams 95 have shownthat, if the sodium enolates of tert.-butyl acetate and tert.-butyl propionate(prepared by means of sodium triphenylmethyl) are treated with ethylcarbonate, the mixed esters ethyl tert.-butyl malonab and ethyl tert.-butylmethylmalonate are obtained in good yield.In general, carbalkoxylation iseffected most readily in the case of aryl-substituted acetic esters and appearsto be limited to esters carrying two &-hydrogens. If the oarbon atom p-with respect to the ester group is tertiary, as in ethyl tmt.-butylacetate, theyield is extremely low when sodium ethoxide is used;9a this is not the case,however, if sodium triphenylmethyl is used as the Condensing agent .95Esters carrying only one a-hydrogen atom do not react with ethyl carbonate,but in such cases carbalkoxylation can be effected by interaction of the sodiumenolate of the ester (prepared by using sodium triphenylmethyl) and ethylchlorocarbonate.96Curbulhxylution of X&nes.-The condensation of ketones with alkylcarbonates with formation of P-keto-esters has been described by a numberof workers.Using acetophenone and ethyl carbonate with sodium ethoxideas condensing agent, L. ClaisenQ7 obtained a poor yield of ethyl benzoyl-acetate. 0. Schroeter98 and H. Lux99 have described the preparation ofNaOEt Ph*CH(CO,Et), + EtOH -> Ph*CH,*COzEt + CO(0Et)ZBer., 1929, 62, 1827.G. S. Skinner, ibid., 1937, 59, 322.89 W. L. Nelson and L. H. Cretcher, J . Amer. Chem. SOC., 1928, 50, 2758.91 Ibid., 1932, 54, 4319.O3 V. H. Wdlingford, A. H. Homeyer, and D. H. Jones, W., 1941,68,2066.O4 This priociple had previously been applied to ester condensation : U.S.P. 1,425,626(1922), 1,472,324 (1923), 1,805,281 (1931) ; S.M. McElvain, J . A w r . Chem. Soc., 1929,61, 3124; R. R. Briese and S. M. McElvain, ibid., 1933, 55, 1697.95 J . Amer. Chem. Soc., 1942, 64,2714.g6 R. E. HudsonandC. R. Rauser, W., 1941, 68, 3156.O 7 Ber., 1887, 20, 656.Ibid., 1933, 55, 4697.Ibid., 1916, 49, 2712. 99 IN., 1 9 2 9 , a 1826SPXING : GENERAL METHODS. 139p-keto-esters from ketones and alkyl carbonates with sodium as condensingagent; a remarkably successful application of this reaction is olaimed byN. A. Preobrashenski, M. N. Schtschukina, and R. A. Lapina,l who obtained70-80 74 yields of tropinonecarboxylio esters by condensation of tropinonewith alkyl carbonates in the presence of sodium or potassium. Applyinggeneral principles established in the cam of the carbalkoxylation of esters,)*i,z., use of a large excess of alkyl carbonate and removal of the alcoholintroduced with the sodium ethoxide and that produced by the reaction,I-.H. Vallingford, A. H. Homeyer, and D. M. Jones2 have shown that thecarbalkosylation of ketones by means of alkyl carbonates with sodiumethoside as condensing agent can be effected in reasonable yield in a, largenumber of cases. Certain limitations to the use of the method are to beiioted ; ketones which readily polymerise or undergo self-condensation giveu:isatisfactory yields. Again, in the cases of propiophenone and butyro-phenone, the required keto-ester is accompanied by the mixed carbonate ofthe enolic form of the ketone, ArC(O*CO,R):CHR'.r'nrbalkoxylation of Nitriles.4arbalkoxylation of various aryl aceto-nitriles by means of alkyl carbonates in the presence of a condensing agenthas been described :3Ar*CR,*CN + CO(OR), --+ Ar*CH(CN)CO,R + ROHF.Adickes and G. Hinderer * used alcohol-free sodium ethoxide and enforcedthe condensation by removal of the alcohol produced during the reaction.ln this way they obtained a 63% yield of ethyl a-cyanophenylacetate fromphenylacetonitrile and ethyl carbonate. V. H. Wallingford, D. M. Jones,and A. H. Homeyer 5 have shown that the method is not limited to aryl-substituted acetonitriles but is of general applicability ; the essential condi-tions are the use of a large excess of dkyl carbonate and the continuousremoval of alcohol from the reaction mixture.Miswllamous Reactions.An analytical method for the estimation of water, alcohols, carboxylicacids and acid anhydrides has been developed by W.M. I). Bryant, J. Mit-chell, and D. 'M. Smith,6 using the Karl Fischer 7* reagent. This reagentis a solution of iodine, sulphur dioxide and pyridine in methanol; it isextremely sensitive to water, the reagent being its own indicator. Alcoholsare esterified with an excess of acetic acid, boron fluoride being used ascatalyst; after decomposition of the catalyst, the water produced iEl esti-mated by titration with the reagent.J . Arner. Chern. SOC., 1941, 63, 2252.Hessler, ibid., 1940, 32, 19; W. L. Nelson and L. H. Cretcher, ibid., 1928, 50,2758; P. Ruggli, E. Caspar, and B. Hegedus, Helv. Chim.Acta, 1937, 20, 250; J. B.Niederl, R. T. Roth, and A. A. Plentl, J. Amer. Chem. SOC., 1937, 59, 1901.1 Ber., 1936, 69, 1615.4 J . pr. Chern., 1938, [ii], S O , 89. J . Amer. Chem. SOC., 1942, 64, 576.Angew. Chem., 1935, 443, 394. Ibid., 1940, 62, 1, 4, 608.D. M. Smith, W. M. D. Byrant, and J. Mitchell, J . Amer. Chem. SOC., 1939, 61,2409.8 140 ORGANIC CHEMISTRY.Alternative methods for the esterification of sterically hindered acidshave been described recently. V. Prelog and M. Plantanides9 have de-veloped the earlier observation of A. T. Lawson and N. Collie lo that thetetramethylammonium salts of acids readily decompose on heating to yieldthe methyl ester of the acid and trimethylatmine. With oleanolic andacetyloleanolic acids: 2 : 4 : 6-trimethyl- and 2 : 4 : 6-triethyl-benzoic acids,llhigh yields of the corresponding methyl ester are obtained.M. S. Newman 1,has shown that esterification of a sterically hindered acid is readily achievedby dissolving it in sulphuric acid and pouring the solution into an excess ofan alcohol. This attractive method has the advantage of giving a varietyof eBters in high yield. Equally attractive is the method of hydrolysis ofesters of sterically hindered acids described by the same author. The esteris dissolved in sulphuric acid, and the solution poured into water,A new optically active reagent for carbonyl compounds, E-menthyl N -aminocarbamate,13 which is stable and yields crystalline derivatives with alarge number of ketones and aldehydes has been described; a successfulresolution of dl-camphor has been achieved by its means.A remarkable " carboxylation " of p a r a 5 or cycloparaf€in hydrocarbonsis described by M.S. Kharasch and H. C. Brown.l* If, for example, cyclo-hexane is treated with oxalyl chloride and the solution irradiated or refluxedwith a catalytic quantity of benzoyl peroxide, hexahydrobenzoyl chloride isobtained in 66% yield :A chain mechanism is postulated for the reaction. With unsaturated hydro-carbons containing a highly polar double bond, oxalyl chloride reacts in theabsence of light and peroxides, as shown by the following examples :Ph*CH:CH, + (COCI), + Ph*CH:CH*COCI + CO + HClCPh,:CH, + (COCl), --+ CPh,:CH*COCl + CO + HC1This type of reaction probably proceeds by a polar and not a chainmechanism .l5A method for the introduction of cyanoethyl groups into compoundscontaining an active methylene group has been described by H. A. Bruson.laFluorene reacts with acrylonitrile in the presence of a catalytic amount oftrimethylbenzylammonium hydroxide to give (I, R = CH,*CH,*CN) . Inthe w e of cyclopentadiene, in absence of catalyst, normal 1 : 4-addition ofsorylonitrile occurs to give (11); in the presence of the catalyst, however,0 Z.physio1. Chem., 1936, 244, 56.10 J., 1888, 63, 031.11 R. C. Fuson, J. Corm, and E. C. Homing, J . Amer. Chem. Soc., 1039, 61, 1920.l2 Ibid., 1941, 68, 2431.18 R. B. Woodward, T. P. Kohmann, and G. C. Harris, ibid., g . 121.14 Ibid., 1942, 64, 329; 1940, 68, 464.15 M.S. Khtlrasch, S. S. Kane, and H. C. Brown, ibid., 1942, 64, 333.l8 Ibid., p. 2647SPRING : GENERAL METHODS. 141this reaction is suppressed, and each of the six available hydrogen atoms isreplaced to give the hexa-cyanoethyl derivative (111, R = CH,*CH,*CN).The reaction whereby C-methylation of pyrrole derivatives is accomplishedby methanolic sodium methoxide 1’ has been successfully applied to theindole series by R. H. Cornforth and (Sir) R. Robinson,l* and the scope ofthe reaction investigated. The authors show that indole and indole-2-carboxylic acids are converted into 3-alkylindoles when heated with a;iooholicsodium alkoxide. Thus skatole is obtained from either indole (71%) orindole-2-carboxylic acid (63%) by heating with methyl-alcoholic sodiummethoxide ; the reaction succeeds with aliphatic primary alcohols (and thecorresponding sodium alkoxide) , but alkylation of indole-2-carboxylic acid isnot accomplished by secondary alcohols.J. W. Cornforth, ( h a . ) R. H. Corn-forth, and (Sir) R. Robinson l9 show that C-methylation of reactive phenolssuch as @-naphthol and resorcinol can also be accomplished by heating withalcoholic sodium methoxide. The yield is increased if the phenol. is firstconverted into a rnethylene-bis-derivative or into a N-piperidylmethylderivative. Thus 2 : 2’-dihydroxydi-cr-naphthylrnethane gives a 75% yieldof 1 -methyl-2-naphthol.A simple procedure for the preparation of volatile acid chlorides has beendescribed ; it consists in the distillation of the required acid chloride from amixture of benzoyl chloride and an organic acid.Although phthalic, maleic and succinic acids, when treated with thionylchloride, give the corresponding anhydrides 21 (and not the acid chloride),L. P.Kyrides22 has shown that, when the reaction is carried out in thepresence of zinc chloride, the corresponding acid chloride is obtained in goodyield (maleic acid giving fumaryl chloride). Developing an earlier observa-tion of van Dorp and van DorpYB it is shown that phthalyl chloride is avaluable reagent for the nearly quantitative conversion of acids and theiranhydrides into the corresponding acid chloride.A valuable method is described by R. Q. Brewster and J. A. P ~ j e , , ~who show that a diazonium salt can be converted into the corresponding1 7 Fischer-Orth, “ Die Chemie des Pyrrols,” Leipzig, 1934, I, 33, 287.l 8 J ., 1942, 680.2o H. C. Brown, J . Amer. Chem. Soc., 1938, 60, 1325.2 1 H. Meyer, Monatsh., 1901, 22, 437; L. McMaster and F. F. Ahmann, J . Amer.Chem. SOC., 1928, 50, 145; P. Carr6 and D. Libermann, Compt. rend., 1934, 199,1422.Is Ibid., p. 682.22 J . Amer. Chem. Soc., 1937, 59, 206.23 Rec. Trav. chim., 1906, 25, 96.24 J . Amer. Chem. Soc., 1939, 61, 2418142 ORGANIC CHEMIBTRY.“ hydrowbon ” by treatment with alkaline formaldehyde. Yields of 60-’80% of the damhated derivative are obtained from a variety of bmes. Themethod is a valuable supplement to the alcohol method of deaminrttion,since beat yields of “ hydrocarbon ” are obtained in cases where the alcoholmethod gives poor yields of hydrocarbon and high yields of ether.F. S.S.4. THE LIGNIN PROBLEM.Since the last Report on this subject 1 much work on lignin has appearedand lignin chemistry now rests on sounder foundations than appears fromearlier reviews.2m 3* Since the review by H. Freudenberg? valuable accountsof ligajll chemistry hsve been contributed by H. Hibbert 5, * in more thanseventy papers in the last twelve years as representing the Canadian, andby H. Erdtman representing the Swedish school.AJl the above authors acknowledge the value of the suggestion of P.Klason,* that lignin is related to coniferyl alcohol[I), which is present as the glucoside coniferin inHO/-\CH:CH*CH2*OH all young plant tissue, and in the cambial sapof the spruce, and his later suggestions9 thatconiferyl aldehyde or mixed coniferyl aldehyde-coniferyl alcohol types are fundamental to the lignin structure.PTotolQpin and Bxtracted Lignirm-The term protolignin has been sug-gestedl* to denote the virgin lignins present in plant tissues and it hasbeen shownll that these substances are extremely sensitive to acids andalkalis, being transformed into dark amorphous polymers.K. Freuden-berg4 indeed recommends where possible the carrying out of rmctions onlignh in situ in the ground wood itself, provided the cellulosic conatituentseither do not react or give easily separable products.Many attempts have been made to isolate a pure unchanged ‘‘ native ”lignin from various sources.In earlier work the “acid” lignins weremuch used, involving extraction with concentrated hydrochloric l2 orMeO-\-/(1.)Ann. Reporte, 1939, 36, 380.A. G. Norman, “Biochemistry of Cellulose, the Polyuronides, Lignin, etc.”Ann. Rev. Bwchem., 1939, 8, 88.Paper Trade J., 1941, 118, TAPPI Sect. 39.Ann. Rev. Biochem., 1942, 11, 183.Svensk Papperstidn., 1941, 44, 243.Svenek Kern. Tidskr., 1897, 9, 136.Ber., 1923, 56, 300; 1930, 63, 792.* M. Phillip, Chem. ReVdew~, 1934, 14, 103.Oxford, 1937.l* J. Konig and E. Rump, “ Chemie und Structure der Pflanzen-Zell-membran,”l1 A. B. Crctmer, M. J. Hunter, and H. Hibbert, J . AmeT. Ohem. Soc., 1939,l* R. Willsthtter and L. Zechmeiater, Ber., 3913, 46, 2401.Berlin, 1914, 86.61, 609PERUIVAL: THE LIUNIN PROBLEM.143sulphuric acid,= mixturw of hydrochloric and phosphoric acids,14 andliquid hydrogen fluoride.15 With the possible exception of the ligninobtained by the last process the products were without doubt highly modified.Much of K. Freudenberg's work has been carried out on " cuproxam "l i e l6 obtained by the extraction of wood meal with dilute sulphuricacid and cuprammonium hydroxide alternately, the latter dissolving thecellulosic portions after decomposition of the supposed lie-carbohydratecomplexes.Another method of attack, introduced by P. Klason18 and used byE. Hiigglund l9 and others, has been exploited with considerable successby H. Hibbert and his ~o-workers.~ This involves extraction with anhydrousalcohols, glycol, glycerol, etc., oontaining small amounts of hydrogenchloride.The alcohol residua combine with the lignin to give suohproducts as methanol-lignin and glyool-lignin.21B. Holmberg22 claims that lignins from the most varied botanicalsouroe8 are ementially of uniform type, but it is 010- that l i g n b fromdifferent source8 are of different composition, and H. Hibbert and his co-workers have suggested this as a, meam of differentiating between angio-sperms and gymnosperms,a3 since all plants universally recognked asgymnosperms (e.g., spruce) yield lignins which on oxidation with nitro-benzene under alkaline conditions give vanillin only, whereas angiosperms(rye, bamboo, mrtple, sassafras, aspen, jute) give both v d l i n and syring-aldehyde.% As the result of his oxidation experiments with chlorinedioxide P.B. Stklrkar concludes that jute lignin ia less highly polymerisedthan wood lignins.In all probability lignins are mixtures of compounds of similar typethough not of identical composition, and the proportions may vary withthe age of the plant and conditions of growth. Af present, however, it isnecessary to regard lignins extracted fkom the same ~ources under the sameconditions as essentially identical.Physical Studies and the Aromatic Nature of .Lignin.--If the lignin iscarefully prepared, the morphology of the cell ia retained. The doublerefraction observed under the polrariSing microsoope becorn- weaker inl3 P. -on, C e l l h a d a m k , 1923, 4, 81.l4 K.Freudenberg and H. Urban, ibid., 1926, 7, 73.l6 K. Freudenberg, M. Harder, and L. Markert, Ber., 1928, 81, 1760.17 K. Freudenberg, A. Jmson, E. Knopf, and A. Haag, Ber., 1936, 69, 1415.18 Tekn. Tkhkr. Am?. Kemi, 1893, 28, 66.113 E. Hiigglund end T. Rosenqvist, Biochem. Z., 1926, 179, 376.20 F. Brauns and H. EIibbert, Cunadian J . &8., 1036,18, B, 28.21 K. R. Grey, E. G. King, F. Brauna, and H. Hibbert, ibid., p. 35.22 Svensk Papperstidft., 1931, 84, 216.a3 E. West, A. €3. MaaInnes, J. L. McCsrthy, and H. Hibbert, J . Amw. Chem. ~Yoc.,24 R. H. Creighton, J. L. McCarthy, and H. Hibbert, ibid., 1941, 68, 3049.2 b J . Indian Chm. Soc., 1935,12, 470.K. Wiechert, aid., 1940, 18, 67.1939, 61, 2666; 1940, 62, 2803144 ORGANIC CHEMISTRY.suspensions in liquids of increasing refractive index and disappears entirelyin iodobenzene (n, 1-62), reappearing with liquids of higher refractive index.The refractive index of lignin is thus estimated to be 1.61 in agreement withits aromatic nature.Z6X-Ray analysis reveals no crystalline or pseudocrystalline structure inlignin, its tensile strength is very low, and its mechanical properties areequal in all directions.The thickness of the layers of lignin azobenzene-sulphonic acid was determined to be 20 A., which is taken to indicate athree-dimensional structure.4From the position and intensity of the spectral absorption bands oflignin sulphonic acids, R. 0. Herzog 27 characterised it as an aromatic sub-stance with an ultra-violet absorption spectrum similar to that of isoeugenoland claimed to show therefrom that the basal unit of lignin is a di- or tri-hydroxyphenol, partly or wholly etherified, having a side chain of aboutthree carbon atoms containing no double bond or carboxyl group con-jugated to the benzene ring.Confirmation of these results is supplied byE. Virasoro 28 for quebracho and red willow lignins.From recent determinations of ultra-violet absorption spectra of l i g n h ~ , ~ ~the very persistent absorption band at 2810 A. was related to theassumption that each native lignin building unit contains two pyranrings as proposed by K. Freudenberg.4 It was also pointed out thatthe carbonyl group shown by F. E. Brauns30 to be present in nativelignin could be responsible for the band at 2520 A.in the absorptionspectrum of the derived phenylhydrazone, and the absence of this bandin the lignins examined is held to indicate the enolisation of thisgroup.J. F. Hechtmann31 has used the molecular still to attack the problem.At 260°/1 p a distillate (4%) was obtained whioh is related to lignin asshown by its absorption spectrum and is held to be partly depolymerised.The residue is considered to be more highly polymerised than the startingmaterial.From dielectric-constant measurements 32 it is concluded that themolecular weight of lignin is ca. 3900 and that normal lignin is fairlyhomogeneous.General Analytical Evidence.-Spruce lignins contain C, 65-66% ; H,6%; 0, 28-29%; lignins from deciduous woods contain ca.5% less C .The oxygen content of lignin is distributed among different groups, sprucelignin containing OMe, 15-16% and hardwood lignins OMe, 21-22%.Hydroxyl groups occur to the extent of ca. 9% in spruce lignin, of which26 K. Freudenberg, H. Zucker, and W. Durr, Ber., 1929, 62, 1814.27 R. 0. Herzog and H. Hilmer, Ber., 1927, 60, 366; 1931, 84, 1288.2 8 E. Virasoro, A w l . Asoc. Quim. Argentina, 1942, 20, 64.29 R. E. Glsding, Paper Trade J., 1940, 111, TAPPI Sect., 288.5o J . Amer. Chem. SOC., 1939, 61, 2120.Paper Trade J., 1942, 114, TAPPI Sect., 259.32 W. P. Conner, J . Chem. Physics, 1941, 9, 691PERCIVAL: THE LIGNIN PROBLEM. 1456-7% are secondary and the remainder K. Freudenberg et aEF4claim a proportion of methylenedioxy-residues (4%), but H.Hibbert 35makes other suggestions as to the probable source of the formaldehyde onwhich this conclusion is baaed and points out that piperonyl derivativeshave never been isolated by the degradation of lignin and that but 8 minuteyield (o.03y0) of formaldehyde results from sassafras lignin, although itwould be natural to expect methylenedioxy-groups to be present in thiscase, since the plant is the source of safiole.36The remainder of the oxygen, ca. 74%, is considered to be ether oxygen.Estimation of the acetic acid liberated by oxidation with chromic acida7shows that at least 2*7y0 of methyl groups bound to carbon are present,and, since this is approximately equal to the amount of tertiary hydroxyl,the presence of the >CMe*OH group is inferred.Free phenolic groupsappear to exist in small amount approximately equivalent to the tertiaryhydroxyl group content.Evidence of the aromatic nature of lignin is provided by the formationof mercury substitution products by treatment with mercuric acetate:*the mercury being replaceable by iodine. Bromination 39 and nitration 40show that up to two hydrogen atoms in each aromatic nucleus can besubstituted.The Views of Karl Freudenberg.Freudenberg4 points out that, if all the oxygen atoms, methoxyl, andmethylenedioxy-residues in lignin are replaced by hydrogen, a hydrocarbonis obtained best expressed by C9HI0, i.e., phenylpro-@-\ CH,-YH*CH:, pane minus two hydrogen atoms (11), and suggests\-/- that this is the fundamental type-unit of the lignin/- (11.) molecule, the central idea being that “ lignin iscomposed of similar units which unite with each other like the amino-acidsin proteins or the monoses in polysaccharides. We assume that the unitsare connected through an ether linkage between phenolic hydroxyl and thecarbinol group in the side chain.”Starting with guaiacylglycerol and acetylguaiacylcarbinol, this workervisualises condensation to such products as (111) and (IV) and an extensionof this principle to form chains of such units, (V) and (VI).No phenylpropane derivatives, however, have been isolated by thisworker, his conclusions being based on c6-4, fragments.Credit for theexperimental confirmation of this theory, without which these speculationsare of little value, must be given to Harold Hibbert and his school (p.15033 K. Freudenberg, F. Sohn, and A. Jason, Annalen, 1935, 618, 62.34 K. Freudenberg, F. Klink, E. Flickinger, and E. Sobek, Ber., 1939, 72, 217.35 M. J. Hunter, G. F. Wright, and H. Hibbert, Ber., 1938, 71, 734.36 M. J. Hunter and H. Hibbert, J . Amer. Chem. Soc., 1939, 61, 2196.37 K. Freudenberg and F. Sohns, Ber., 1933, 66, 262.38 K. Freudenberg and H. F. Miiller, Ber., 1938, 71, 2500.3D K. Freudenberg, W. Beltz, and C. Niemann, Ber., 1939, 72, 1664.40 K. Freudenberg and W. Durr, Bw., 1930, 63,2713et q.). The assumption is dm made that oondenaation to form chroman(VII) or furan.(VIII) ring8 may ocour, although the uncondensed etherMeV0QHOHc3-(VII.)MeC/OH P-b1(VIII.) I 1systems may be present also.The main fea;tures of these formula maybe represented by (IX) and (X).I/\J% JfMe A C0,H *.../'.d' OMe I IIOMeHO*CB:\/I ?&€!(XI.) F"- CH(=.I (X.PEROIVAL: TEE LXGNlX PWBLEM. 147Freudenberg consideras that the few end groups in lignin are present inthe form shown by (111) and (IV), because methylation, followed by oxid-ation, yields only a relatively small amount of veratric acid (XI). Hestates also that the groups which split off formaldehyde on treatment withacid are intermediate members within the chains, but admits that there isas yet no proof of this.From the isolation of the pyrogallol component of the oxidation productsof beech lignin as the symmetrical dimethyl ether, it is concluded thatunits like (XII) can occur, and the same is assumed to be true for thepyrogallol component of spruce lignin.I/\I(XIII.)MeO(,!lOMed (XII.)MeCO\/CH I ITo sum up, Freudenberg considers that five out of eight units in sprucelignin belong to types like (IX) or (X), two units contain methylenedioxy-groups and one is of the pyrogdol type (XII), the average molecular weightper unit being 185.Condensation to form two- or three-dimensional unitsis also en~isaged.~On the basis of the isolation of vanillin-5-carboxylic acid from sprucelignin, the above views have been somewhat modified41 and it is now statedthat at least half of the basal units in apruce are represented by (XIII).The experimental evidence on which the above theories rest is summazisedbelow.Potash fusion of apruce lignin yields protocatechuic acid (lo%),isolated as veratric acid (XI), this yield being comparable with that obtainedfrom eugenol or polymerised coderyl alcohol on similar treatment,= andfrom beech lignin, gallic acid, as trimethylgallic acid, and protocatechuicacid (3-3.5%) axe 0btained.l7*~~ If apruce meal is treated with diazo-methane, and the product oxidised with permanganate, veratric acid (4%,based on the lignin content) is obtained,& from which it may be concludedthat every sixteenth unit, on the average, possesses a free guaiacyl end groupand that lignin in wood contains approximately 0-6% of free phenolicgroups. The yield of veratric acid from isolated spruce lignins is lower(1-2%), ao most of the veratric acid may originate from lignin precursorsof low molecular weight,The main evidence adduced in favour of the above theoretical sohemesis aecured by treatment of spruce wood-meal with diazomethane, followedby potassium hydroxide solution (70%) at 165-170", methylation with'1 K.Freudenberg and F. Klink, Be?., 1940, 73, 1369.K. Freudenberg and H. l?. Miiller, Ber., 1938, 71, 1821.K. Freudenberg, M. Meister, and E. Flickinger, Ber., 1937, 70, 600148 ORGANIC CHEMISTRY.methyl sulphate, and oxidation with permanganate to yield veratric acid(XI) (20-21~0), isohemipinic acid (XIV) (6-12%), dehydroveratric acid(XV) (2-3y0), and trimethylgallic acid (XVI) (in traces) Lower valuesare obtained for isolated 1ignins.l'. 43C0,H C0,H C0,H CO,HA A /\ AH0,Ck)OMe MeOd)lOMeOMe OMe OMe OMe(XIV.) P V .1 (XVI.)By substituting ethylation for methylation, evidence was obtained, bythe isolation of (XVII) and (XVIII), albeit in poor yield, that the etherbonds which open are as indicated and are here occupied by ethyl groups.CO,H CO,H CO,HA A /\\/ \/ Me01 IlOMe HO,Cl IlOMeOEt\/ I IlOMeOEt OEt(XVII. ) (XVIII.) (XIX.)By examining the yields recoverable by treating (XI) , (XIV), (XV),(XVI), and piperonylic acid with the above reagents,a and applying cor-rection factors to the yields obtained from spruce lignin, the true yield44of (XIV) is calculated to be 80% and of (XI) and (XV) combined, 32%.The high correction factor ( x 9) for (XIV) is unfortunate, as is also the factthat this method gives no practical evidence for the presence of methylene-dioxy-groups, the presence of which is denied by H.Hibbert.35 H. Erdt-man considers that the drastic alkaline conditions introduce a very uncer-tain factor and that a large portion of the isolated (XIV) may be formedby secondary condensation reactions between aromatic nuclei and sidechains. On account of the difficulty of separating (XVI) similar experi-ments with beechwood lignin are less satisfactory, the corrected yieldsbeing (XI) 7%, (XIV) 13.5%, and (XVI) 13%. By the ethylation tech-nique beech lignh gives, in addition to the ethyl ether of vanillic acid(XVII), the ethyl ether of syringic acid (XIX) in harmony with the factthat symmetrical dimethylpyrogallol derivatives are found in the distillationproducts of beech lignin.Freudenberg supposes that the formation of /I\ lignin sulphonic acids by reaction with sulphites isdue to the rupture of ether linkages (X) in every)\)IoMe third unit with the appearance of new phenolicOHoMeC bH residues (XX), since a higher yield of (XI) is obtainedby methylation, followed by oxidation, than from ligninbH*S03H itself.Methylenedioxy-groups appear to be largelyFX.) eliminated by the sulphite cooking process.44 K. Freudenberg, K. Engler, E. Flickinger, E. Sobek, and F. Klink, Ber., 1938,71, 1810PERCIVAL : THE LIGNIN PROBLEM. 149A similar explanation is also applied to B. Holmberg's reaction4 withthioglycollic acid, S*CH2*C02H replacing SO,H in (XX).The Views of Holger Erdtrnan.points out that structures such as (V) and (VI) would beexpected to show special reactivity in the CH residue para to the methoxylgroup and considers that the probable direction of the condensation shouldinvolve the hydroxyl group of the side chain with an adjacent molecule inthe p-position to the methoxyl group.Furthermore, the transformationof (V) into (VII) appears improbable, since condensations of this typegenerally require alcohols having hydroxyl groups on carbon atoms directlyattached to the benzene nucleus, e.g., benzyl alcohol.Erdtman also draws attention to the mode of attachment of naturallyoccurring dimeric forms of phenylpropane units in substances such asmagnolol 45 (XXI), dehydroeugenol 46 (XXII), and egonol 47 (XXIII).Erdtman/O-CH,(XXIII.) I A A - r L - 0\/--Me0 0CH,( OH)*CH2*CH2-l IIH \=/The lignans, products of great importance in wood chemistry, such aspinore~inol,~~ lariciresinol 49 (XXIV), and conidendrin 50 (XXV) alwayshave the middle carbon atom in the side chain as the connecting atom, asin (VIII) but not as in the Freudenberg formula (VII).(XXV.)(XXIV.) OH46 Y. Sugii, Chem. Zentr., 1930, 11, 263.46 H. Cousin and H. HBrissey, Compt. rend., 1937,146,1413; H. Erdtmrsn, Biochem.4' S . Kawai and N. S~iyama, Ber., 1939, 72, 369.4 8 H. Erdtmm, Annalen, 1933, 508, 283.40 R. D. Haworth and W. J. Kelly, J., 1937, 384; R. D. Haworth and D. Woodcock,60 R. D. Haworth, T. Richardson, and G. Sheldrick, J., 1936, 1676.Z., 1933, 268, 172.J., 1939, 1064150 ORUANIO CHE)1\6ESTRY.In 1933 Erdtm&n,4O on the bask of the production of dehydrodiiso-eugenol (XXVI) by the action of ferric chloride on, or the enzymatic dehydro-genation of isoeugenol, suggested that lignin might be a high-molecular-weight dehydrogenation product ofMe0 0 OMe phenylpropane derivatives and K.as a type model for many lignin reac-(XXVI.) Me tions.By methylation of (XXVI),followed by cautious oxidation, Erdt-man 46 obtained an acid (XXVII), from which Freudenberg and his co-workers obtained (XI) (21% instead of 53%) and (XIV) (6% instead of66%), a result not very different from that obtained for lignin itself underPreudenberg et aE.43 have used (XXVI)(XXVII.)OMe T03H OMe/\OH H--, /-\oMe / @\ OM* HO,CH\ OMeHO,d I' JHMe = H02d \/ IlC02H (,,!lOMe(XW.) (XI.)similar conditions [(XI), 14%; (XIV) 4%].Under the conditions of thesulphite treatment 51 an acid (XXVIII) was obtained which gave onmethylation and oxidation (XI) (17% instead of 39%) and (XIV) (4%instead of 40%); compare ligninsulphonic acid, whioh gives (XI) 3% andSimilar analogies between (XXVII) and lignin me ehown in its reactions\/-( XXVIII . )( X I V ) 343%.with thioglycollic acid, on methanolysfs, and with hydrazine.4The Views of Harold Hibbert.Much of the value of the work of Hibbert and his school lies in thecautious way in which he has interpreted his results, the full value of whichis now clearly discernible. The development of the alcoholysis methodfor the isolation of lignin 52 yielded much useful information.A morefruithl field has proved to be his studies of the action of alkali on the lignin-sulphonic acids. Spruce ligninsulphonic acid gave vanillin ( 6-7y0),53and this is not only a commercial outlet for waste sulphite liquor but showsthe importance of the guaiacyl nucleua in the lignin molecule. Soon after-w&r& small amounts of acetovanillone (XXIX) 54 and guaiacol 65s 66 were51 I(. Freudenberg, Papierfabrilcant, 1938, 36, 34.G3 H. Hibbert and H. J. Rowley, Canadian J. Res., 1930, 2, 357 et seq.83 G. H. Tomlinson and H. Hibbert, J . Amer. Chern. SOC., 1936,58,345.64 I. H. Buckland, G. H. Tomlinson, and H. Hibbmt, ibid., 1937, 59, 597.66 F. Leger and H.Hibbert, Canadian J. Ree., 1938, 16, B, 68.6u Idem, $bid., p. 151 ; J . AmeT. Chrn. Boc., 1938,60,565PERCIVAL: TRB: LIWIN PBOBLEM. 162isolated together with the vanillin. Oak and maple ligninsulphonic acidsgave, in addition, syringaldehyde 57 (XXX), aoetosyringone 58 (XXXI),and 1 : 3-0-dimethylpyrogallol 56 (XXXII).MeCO CHO MeCOA A A A.\/ I IlOMe Me01 IIOMe MeOk,!lOMe MeO!\)lOMeOH OH\/OH OH(XXIX.) (XXX.) (XXXI.) (XXXII.)The isolation of (XXIX) and (XXXI) indicated for the first time theexistence of side chains of two carbon atoms in a t least some of the ligninbuilding stones. K. Freudenberg et ~ 1 . ~ ~ later obtained, by the action ofnitrobenzene and alkali on spruce lig~hsulphonic acid, a higher yield ofvanillin (26%) : this was confirmed 59 and the mme technique applied toaspen and maple woods gave a- mixture of ttadlin md gyringic aldehyde(4045%).59 These results prove beyond doubt the aromatic characterof lignin and dispose of the theory of its carbohydrate nature.60 Thefailure to isolate piperonyl derivatives, as already mentioned, throws doubton K.Freudenberg’s view that spruce lignin contains 25% of these nuclei.The most,important advance in lignin chemistry has been made as theresult of Hibbert’s experiments on the treatment of various plant materials,such as maple wood, with ethanolic hydrogen chloride (2%); ethoxyl-containing ethanol-lignins of high molecular weight (ca. 60% of the KlasonIignin content), together with soluble distillable fractions (30%) of lowmolecular weight,B1 were thus isolated.The products so far identified (ca.13%) in this mixture are : for spruce, vanillin, ~-hydroxy-ol-(4-hydoxy-3-methoxypheny1)- a-propanone (XXXIII) , 62 a- hydroxy- a- (4- hyltroxy-3-meth-oxypheny1)- @- pr opanone ( XrUrrV), G4 a- (4 - hydroxy- 3 -methoxyphenyl) - ap -VHMe-OH COMe YOMe VOMeCO CH-OH co( XXXIII . ) (XXXIV.) (XXXV.) (XXXVI.)propanedione (XXXV) ,63 and a- (4-hydroxy-3-methoxyphenyl) - p-propanone(XXXVI); 64 for maple wood, both these and the corresponding syringyl57 A. Bell, W. L. Hewkins, G. F. Wright, and H. Hibbert, ibid., 1937, 59, 698.68 K. Freudenberg, W. Lautsch, and K. Engler, Ber., 1940, 73, 167.SB R. H. J. Creighton, J. L. McCarthy, and H. Hibbert, J . Amer. Ohern. SOC., 1941,60 R.S. Elpert, Cellulosechernie, 1936, 17, 26.6 1 L. Brickman, J. J. Pyle, J. L. McCrtrthy, and H. Hibbert, J. Amer. Chem. SOC.,62 A. S. MacIMBs, E. West, J. L. McCarthy, and H. Hibbert, ibid., 1940, 62, 2803.64 R. Hibbert, private communication.08,312.1939, 81, 868.L. Briokmm, W. L. Hawkins, andH. Hibbert, ibid., p. 2149152 ORGANIC CHEMISTRY.derivatives.a* 66* 66 The products (XXXIII) and (XXXIV) and theirsyringyl analogues were isolated in the form of their ethyl ethers as a resultof the ethanolysis.The importance of the characterisation of these derivatives is that allare derivatives of phenylpropane and thus fit in with the theoretical specu-lations of K. Freudenberg. It appears that there is a positive relationshipbetween lignin and the above ethanolysis products, for ethanol-lignin ontreatment in this way yields a proportion of these simple products.67 Atthe same time, however, irreversible polymerisation reactions proceed, thuslimiting the yield of distillable oils.the similarity between the structure of theunits formed by the ethanolysis of lignin and the ene-diol oxidase systemof A.Szent-Gyorgyi.68 The hydroxy-compounds corresponding to (XXXVI)and its syringyl analogue are benzoins, compounds which are well known toresinify in the presence of mineral acid and also undergo ene-diol dismut-R-C-hHMe += R*-bMe R*CO*COMeH. Hibbert points outOH -2H+ 2H9 OH11RGH( 0H)COMeR = guaiacyl or spingylation, and Hibbert considers that the ene-diols may act as reductants ofene-diol-1 : 2-diketone systems, the oxidants being the vanilloyl (XXXV)and syringoyl methyl ketones isolated above.The suggestion is madethat, analogous to the animal respiratory catalyst system of A. Szent-Gyorgyi (C, system), there exists a system of plant respiratory catalysts( C6-C, system), made up of mono- and di-hydroxyconiferyl alcohols inequilibrium with their keto-forms, in which coniferyl or syringyl alcoholappears as the analogue of fumaric acid in the C4 systems, and that ligninis essentially a product derived from these plant respiratory Catalysts. Inthis connection the changesR*CH,*CO*CH,-OH =+ RCH:C( OH)*CH,*OH + R*CH( OH)*C( 0H):CH11R*CO*CH( OH)*CH, =+= ROC( OH):C( OH)*CH, + R*CH( OH)*CO*CH, + 2Hand R*CH( OH)*CO*CH, -% R*CO*COCH, - 2H(R = guaiacyl) have recently been realised experimentally in vitro byH.Hibbert and his co-w~rkers.~~E. E. Harris, J. D’Ianni, and H. Adkins 69 studied the catalytic hydro-genation of a methanol aspen lignin and isolated 4-n-propylcyclohexanol65 M. Kulka, W. L. Hawkins, and H. Hibbert, J. Amer. Ch.em. SOC., 1941, 63, 2371.69 M. J. Hunter, A. B. Cramer, and H. Hibbert, ibid., 1939, 61, 516.6 7 W. B. Hewson, J. L. McCarthy, and H. Hibbert, ibid., 1941, 63, 3061.6 8 Ber., 1939, 72, 53 A. J . Amer. Chem. Sm., 1938, 80, 1467PERCIVAL : THE LICfNIN PROBLEM. 153(XXXVII), 4-n-propylcyclohexane-1 : 2-diol (XXXVIII), and y-(4-hydroxy-cycZohexy1)propanol (XXXIX), all of which were synthesised.69* 7O H. Hib-CHPr“ CHPra CH*[CH2],*OH/ \ / \ / \H2Q p 2 H2Q QH2 H2Q p 3 2\ / \ / \ /H,C CH, HC*OH CH, H2C CH,CH*OH CH-OHbert and his co-~orkers,~~ by the hydrogenation of spruce and maplewood,isolated (XXXVII) and (XXXIX) in yields of 19.5y0 and 58y0 respectively.A re-investigation showed the (XXXVII) fraction to be a mixture of thatsubstance with y-cyclohexylpropanol in the ratio of 1 : l .V 2Y . Hatihama et u Z . , ~ ~ using less active catalysts, hydrogenated hydro-chloric acid lignin and obtained n-propylguaiacol , and similar results havebeen reported by K. Freudenberg and his co-worker~.~~These hydrogenation results are explained by the rupture of the etherlinkages and lend further support to the phenylpropane theory of ligninstructure. That oxygen is attached to the terminal carbon atom of thepropyl side chain in protolignin is shown by the fact that about half ofthe propylcyclohexane derivatives from wood 71 and certain lignins 69* 75have oxygen in that position (XXXIX).A comparison of the hydro-genation of maple ethanol-lignin and maple wood shows that & terminalprimary alcohol or methylene ether group is present in the protoligninbecause of the higher yield of (XXXIX) obtained from the latter.5 Theexamination of hydrogenation products of fractions of a maple ethanol-lignin 75 shows that “ for a given series of such fractions, there is a, parallelbetween increasing solubility and susceptibility to depolymerisation intosimple propylphenol units, and the increasing yield of the water-insolublepropylcyclohexanol degradation products.” The conclusion is drawn thatthose readily soluble lignin units which are easily depolymerised by ethan-olysis and hydrogenolysis contain large numbers of C-0-C.bonds, whereasin the lignin fractions which exhibit these properties to a less degree C-C-Cbonds are more numerous. A somewhat similar conclusion is drawn byH. Adkins and his co-~orkers,~~. 77 who show that “soda lignin ” frommixed hardwoods contains one oxygen per 13.6 carbon atoms and methanol-lignin (from fresh aspen) one oxygen atom for six carbon atoms and concludethat in the treatment of lignin with soda further cyclisation has occurred.CH*OH(XXXVII.) (XXXVIII. ) (XXXIX.)70 E. Bowden and H. Adkins, J . Amer. Chem. Soc., 1940, 62, 2422.71 H.P. Godard, J. L. McCarthy, and H. Hibbert, ibid., 1941, 68, 3061.72 J. R. Bower and H. Hibbert, unpublished results.73 J . Soc. Chem. Ind. Japan, 1940, Suppl. 43, 127.74 Ber., 1941, 74, 171.’i5 L. M. Cooke, J. L. McCarthy, and H. Hibbert, J.-Amer. Chem. SOC., 1941, 63,7 6 E. E. Harris and H. Adkins, Paper Trade J., 1938, 107, TAPPI Sect., 38.77 H. -4dkins, R. L. Frank, and E. 8. Bloom, J . Amer. Chem. Soc., 1941,68, 3041.3056164 ORGANIC CHEMISTRY.The same also applies to alkali and sulphurio acid lignim3, which appearto be more complex than alcohol lignins or protolignin.Further light has been thrown on the nature of the terminal group inthe protolignin building unit by oxidation with chromic acid and estim-ation of the liberated acetic acid,64 spruce and maple wood, various ethanol-lignins, and thirteen aubstances, such 8s vanilloyl methyl ketone, thoughtto be closely related to the lignin progenitors, being used.The results showthe absence of terminal methyl groups in the three carbon side chainsattached to the aromatic nuclei in the “native lignin” present in spruceand maple wood, and in the various insoluble maple ethanol-lignins. Theamount of acetic acid obtained fiom the spruce ethanol-lignin is equivalentto one terminal methyl group in four or five phenylpropane units, which isthe same as that obtained for a apruce “ cuproxam ” lignh3’ The amountof aoetic acid liberated from ethanol-lignins depolymerised by furtheretbnolysis WBB doubled, showing that terminal methyl groups are exposedby this treatment.Taking st broad view of all the results and views expressed above, thesuggestion of R.D. Haworfh 78 that lignin represents the polyterpene ofthe C,-C, metabolism and his further suggestion79 “that the wide dia-tribution of fhe n-propylbenzenes and the lignans in unrelated plant familiesis a strong indication of their aesociation with a general metabolic process,a d in spite of the inconclusive state of knowledge concerning the con-stitution of lignin, the relationships are sufficiently striking to indioate thatthe .la-propylbenzenw and the lignans are conneoted with the lignificationprocess ” are olearl y justified.The Natwe of the Union between Lignin and Other Plant Constituents.Since the extent to which polymerisation of the protolignin takes placeduring extraction is unknown, the problem of deciding whether the ligninprecursom or protolipin are attached to the carbohydrates of wood byglycosidic or ether linkages is difficult to solve.The fact that the C6-C3units, whafever they may be, exist almost certainly in the early stages ofgrowth its simple glycosides, e.g., coniferin, does nof necessarily mean thatthe carbohydrate-4pi.n complex which is generally assumed to exist con-tains glycosidic linkages also.A. J. Bailey 80 has studied the extraction of butanol-win by butanol-water and butaaol-water-alkali and has shown that all the lignin of mpene m be removed at 160” in 7 hours with aqueous butanol previously bufferedto neutrality.E’urthermore, the extraction of 93% of the lignin content ofpoplar has been achieved *1 by heating with a concentrated solution ofsodium cymenesulphonate a t 105” for 30 hours, and thwe experimentsseem to point to a relatively weak carbohydrate-lignin union, if indeed76 Nature, 1W1, 147, 255. ’* R. D. Haworth, Tilden Leoture, J., 1942, 448.8o Paper Xrcrde J., 1940, UO, No. 2, 20; No. 6, 27; No. 7, 27; 111, No. 9, 86.*I Pehpetz, Ph.D. Theeia, Columbia University, 1937COOK : POLYCJYoItIC AROMATIC COMPOUNDS. 155any exists in these cases. The enzymic degrrtdp,tion experiments ofT. Ploetz 82 seem to show that some form of combination does exist andA. G. Norman3 holds the view that hemicellulose-lignin complexes existin wood.H. Hibbert and co-workers,B from their experiments on the hydrolysisof glycosides such as acetovanillone- p-cellobioside and the correspondingglucoside and xyloside, a- hydroxypropiovanillone- and a- hydroxysyringone-p-d-xylosides and a-hydroxypropioveratrone-p-d-glucoaide, conclude thatthe aliphatic type of glycoside linkage as found in the last is unlikely andthat a phenolic glycoside linkage “ is a plausible type for the lignin-carbo-hydrate linkage of wood if such exists.” Much further work remains tobe carried out before this problem can be decided.E. G. V. P.5. POLYCYCLIC AROMATIC COMPOUNDS.The polycyclic aromatic hydrocarbons were last reviewed in a generalarticle in these Reports 1 in 1933 ; since then a large volume of material hasbeen published dealing with work in this field.Reports have alreadyappeared on the rubenes and on naturally occurring polycyclic quinones,3and the scope of the present article will be limited by omitting references tothese and to polycyclic hydrocltrbons resulting from the dehydrogenation ofnatural products. Only compounds containing more than two condensedrings are discussed, and, a8 the chemistry of dyes and their intermediatesis regularly reviewed in the Annual &ports of the Progress of AppliedChemistry, the polycyclic compounds which come in this category receivelittle attention. Much of the recent progress has been concerned with thedevelopment of synthetic methods. Methods which are directed primarilyto the production of hydroaromatic structures have already been reviewed.4*The ease with which aromatic structures may be obtained fiom the hydro-aromatic compounds by dehydrogenation often renders these methods verysuitable for the production of purely aromatic polycyclic compounds.Evenwith the limitations imposed by the omission or curtailment of these sectionsof the subject the field for review remains too large for adequate treatment,and the selection of topics within this field has been determined largely bythe personal interests of the Reporter.Hydrocarbons of Coal Tar.At the time of the appearance of A, E. Everest’s book 6 it might reasonablyhave been assumed that knowledge of the constituents of coal tar wasapproaching finality. That this was not so has been shown by the large8a Ber., 1940, 73, 790.J.H. Fisher, W. L. Hawkins, and H. Hibbert, J. AmeT. Chern. Boc., 194 1, 68,G. A. R. Kon, Ann. Reports, 1932, 29, 163.a R. D. Haworth, ibid., 1937, 34, 389.R. P. Lhstead, &id., 1936,33,312.“ The Higher Coal Tar Hydrocarbons,” London, 1927.3031.3 A. R. Todd, ibid., 1941, 88, 205.6 H. D. Springall, ibid., 1939,36,286166 ORGAN10 CHEMISTRY.number of polycyclic aromatic compounds isolated from the higher-boilingfractions of coal tar during recent years. For the most part these werealready known as synthetic compounds, so their identification has presentedlittle difliculty. Much of this new work on high-boiling tar fractions hasbeen carried out by 0. Kruber, who has recently summarised the results.'Thus, 7 of the 10 possible dimethylnaphthalenes have been isolated 8 as wellas 2 : 3 : 6- and 1 : 3 : 7-trimeth~lnaphthalene.~ Polycyclic hydrocarbonsand their derivatives shown to be present in coal tar include 1-, 3-, and9 -meth ylphenant hrene, lo 2 - hydroxyphenant hrene, l1 4 : 5 -methylenephen-anthrene," l2 triphenylene (LXXVII),13 2- and 3-methyIfiuorene,l4 2-hydroxyAuorene,15 1 : 2- and 2 : 3-benzfluorene,l6 cyanofluorenes lo and1 : 2-ben~anthracene.~'~ lS The two theoretically possible benzpyrenes alsohave been isolated from coal tar l7 and one of these, the 3 : 4-compound(IV), was found to be a very potent cancer-producing hydrocarbon.According to an estimate quoted by Kruber,' 132 kg.of ordinary coal tarpitch contains 1 g. of carcinogenic 3 : 4-benzpyrene.A much higher value isgiven by A. Winterstein,lg who used the chromatographic adsorptiontechnique for the isolation of 3 : 4-benzpyreneFO Winterstein obtained2.5 g. of almost pure benzpyrene from 50 kg. of tar, which contained 3 kg. ofmaterial boiling above 450". Chromatography has also been used for theseparation of mixtures of polycyclic aromatic hydrocarbons 21 and for theisolation and purification of other higher coal tar hydrocarbons.223 : 4-Benzyprene appears to be a feebly coloured (yellow) hydrocarbon.Two deep orange hydrocarbons have been isolated from coal tar. Theseare naphthacene (2 : 3-benzanthracene) (XIII) 20* 22 and perylene (XLV).17The former is the " chrysogen " which imparts the yellow colour to incom-pletely purified coal-tar anthracene.a In addition to the polycycltc hydro-carbons and their derivatives a considerable series of heterocyclic compoundsAngew.Chem., 1940, 53,69.Compare 0. Kruber and A. Marx, Ber., 1939, 72, 1970.0. Kruber, ibid., p. 1972.lo 0. Kruber and A. Marx, Ber., 1938, 71, 2478.l 1 0. Kruber, Ber., 1936, 09, 246.l3 H. Kaffer, Ber., 1935, 68, 1812.l6 Idem, Ber., 1936, 69, 107.l 7 J. W. Cook, C. L. Hewett, and I. Hieger, J., 1933, 396.l 8 0. Kruber, Ber., 1941, 74, 1688.l9 Festschrift Ernil Barell, p. 320 ; Bade, 1936.2o A. Winterstein and K. Schlin, Naturwiss., 1934, 22, 237.21 Idem, 2. physiol. Chem., 1934, 230, 146.22 A. Winterstein, K. Schon, and H. Vetter, ibid, p. 158.23 Compare J. W. Cook et al., Proc. Roy.Soc., 1932, B, 111, 469.* This tetracyclic hydrocarbon was subsequently synthesised by W. E. Bachmannand J. C. Sheehan ( J . Amer. Chem. SOC., 1941,63,204).f- Note on Nomenclature.-Throughout this Report the Richter system of numberingis usually employed where numbering is necessary (but see p. 168). In the Americanliterature other systems, including the Patterson system, are often used. In thisconnection attention is called to " The Ring Index " by A. M. Patterson and L. T.Capell (New York, 1940).l2 Idem, Ber., 1934,67,1000.l4 0. Kruber, Ber., 1932, 65, 1382.l6 Idem, Ber., 1937, 70, 1566COOK POLYCYCLIC AROMATIC UOMPOUNDS. 157(with cyclic oxygen, nitrogen or sulphur) has been isolated in recent yearsfrom the higher-boiling fractions of coal tar.24 Phenanthrene has beenisolated from Roumanian petroleum by I.Gaviit and I. I r i m e ~ c u . ~ ~An interesting new source of higher aromatic hydrocarbons has beenfound in the products of destructive catalytic hydrogenation of coal or taroils or pitch.26 The crude products are submitted to dehydrogenation andto selective extraction with solvents, and in this way such coal tar hydro-carbons as pyrene (XXVIII) (and its homologues), fluorene (XXVI), andchrysene (CXXXIV) may be obtained. Pyrene and chrysene appear to beparticularly abundant, and yields of pyrene have been claimed amounting to1% on the coal used or 8% on the total heavy oil. In view of the difficultyof isolating these two hydrocarbons in quantity from coal tar 27 this newsource should offer considerable possibilities of commercial development.From the hydrogenated oils have also been isolated complex hydrocarbonsnot so far found in coal tar, namely, 1 : 12-benzperylene (I) and coronene(II)?/\A A/\I II I\/ AA/\/I II I\/\/I l l /==\/\ I \/\ \-/I I ( ,II I I II\A \/\/\/I \i II\/\/(111.) (11.) (1.1Ultra-violet absorption data of coronene are given by J.W. Patterson,%and a new synthesis from 7-methyl-1 -tetralone has been described.30Another new source of polycyclic aromatic hydrocarbons has been foundin the products of thermal decomposition of natural gas. Methane, its chiefconstituent, is largely decomposed to carbon black and hydrogen whenpassed over strongly heated refractory material.By cooling the emittedgas, a waxy solid is obtained, and from this A. W. Campbell, N. H. Cromwell,and J. J. Hager31 isolated acenaphthylene and pyrene. By solventextraction of the carbon black prepared by this process, J. Rehner 32 obtainedfluoranthene (111) (3-5 g. from 5 kg. of carbon black).Carcimgenic Hydrocarbons.The cancer-producing action of many of the polycyclic aromatic hydro-carbons has caused much attention t o be given to the synthesis and24 0. Kruber, Angew. Chem., 1940, 53, 09; Ber., 1940, 73, 1184; 1941, 74, 1688.25 Ber., 1942, 75, 820; cf. T. Coaciug, Ann. sci. Univ. Jassy, Sect. I , 1940, 26, 415.26 I.G., Brit. Pat. 453,264; 470,338 ; 493,307; 493,447; 493,508; 497,089;510,736; compare K. Zerbe and K. G. Grosakopf, Brenstoff.-Chem., 1938, 19, 61; E.Bed, H.Biebesheimer, and W. Koerber, Ind. Eng. Chem., 1941, 33, 672.27 Compare E. A. Coulson, Chem. and Ind., 1941, 60, 699, and 0. Kruber, Ber.,1931,64,84.2a See, c.g., I.G., Brit. Pat. 470,338; 497,089.28 J . Amer. Chem. SOC., 1942, 64, 1485.31 Ibid., 1936, 68, 1061.3O M. S. Newman, ibid., 1940, 62, 1683.Iba., 1940, 62, 2243chemistry of compounds of this class. Comprehensive reviews have beenpublished dealing with the structural relationships of the carcinogeniocompounds and their biological eff e ~ t s . ~ ~ These carcinogenic hydrocarbonsfall into two main groups. The major group is related to 1 : 2-benzanthra-cene, and a second group is related to 3 : 4-benzphenanthrene (cf. CXX). Itis possible to state in general terms the nature and positions of substitutionlikely to lead to pronounced carcinogenic activity; this property is Borne-times very sensitive to modification of molecular structure, whereas in othercases it survives quite gross struotural changes.This point may be illustratedby reference to derivatives of 3 : 4-benzpyrene (IV), the active carcinogeniohydrocarbon of coal tar. Its 4’- and 6-methyl derivatives are much slowerin action, and its 2’- and 3’-methyl derivatives appear to be without carcino-genic activity.34 On the other hand, high activity is still retained in1 : 2 : 3 : 4- and 3 : 4 : 8 : 9-diben~pyrene.~~1 : 2 : 5 : 6-Dibenzanthracene (V) is noteworthy as the Erst polycyclichydrocarbon found to have carcinogenic activity, and 20-methylcholanth-rene * (VI) is of interest as a highly carcinogenic hydrocarbon, related to thesterols and bile acids; it may, in fact, be prepared from cholic and deoxy-cholic acid, and from cholesterol.9 : 10-Dimethyl-l : 2-benzsmthracene (VII)is outstanding as the most rapidly acting cancer-producing hydrocarbon 80far found, and it is remarkable that a similar high order of activity isshown by the structurally analogous 4 : 9-dimethyl-5 : 6-benzthiophanthrenWI.) (VIII.) (IX.)An analogous compound, 9 : 10-dimethyl-1 : 242’ : 3’-thiopheno)anthracene33 J. W. Cook, E. L. Kennaway, et al., Amer. J . Cancer, 1937,29,219; 1938,33, 50;1940,39,381,521; L. F. Fieser, aid., 1938, 34, 37.34 L. F. Fieser and H. Heymann, J . Amer. Chem. SOC., 1941, 68,2333.35 W.E. Bachmann et al., Proc. Roy. Soc., 1937, B, l28, 343.36 R. B. Sandin and L. F. Fieser, J . Amer. Chem. SOC., 1940,62,3098; C. E. Dunlap* The name cholmnthrene is derived from the ciEoEane (bile acid) group of naturaland S. Warren, Cancer Research, 1941, 1, 953.products, and the sterol eystem of numbering is used for cholanthrene derivativesCOOK : POLYOYOLT.0 BROMBTIO COMPOUNDS. I59(IX), has been prepared by E. B. Hershberg and L. F. Fieser 37 by a methodwhich could be adapted to the formation of the corresponding compoundcontaining radioactive sulphur. This compound with labelled sulphurwould be of value in biochemical investigations.In the 1 : 2-benzanthracene series it has often been found that a condensedbenzene ring is equivalent to two o-methyl groups in its influence in promotingcarcinogenic activity.The extension of this principle to the 3 : 4-benzphenanthrene series by C. L. Hewett has led to the discovery of simplehomologues of chrysene and phenanthrene possessing weak carcinogenicactivity. These are 1 : 2-dimethylchrysene (XI) 38 and 1 : 2 : 3 : 4-tetra-methylphenanthrene (XII)F9 which are related in the manner indicated tothe moderately potently carcinogenic 1 : 2 : 3 : 4-dibenzphenanthrene (X).40Moreover, it may be observed that substitution of a condensed benzene ringfor the methyl groups at positions 2 and 3 in 1 : 2 : 3 : 4-tetramethylphen-anthrene leads to the structure of the very active carcinogenic hydrocarbon9 : 10-dimethyl-1 : 2-benzanthracene (VII).(XII.)When carcinogenic hydrocarbons are introduced into the animal body,they become hydroxylated and excreted.The compound obtained thusfrom 3 : 4-benzpyrene appears to be a monohydroxy-derivative, which hasnot yet been identified.41 E. Boyland and his collaborators42 isolated adihydroxy-derivative of 1 : 2 : 5 : 6-dibenzanthracene from the urine of rabbitsreceiving a-diet which contained the hydrocarbon. In the case of mice andrats an isomeric dihydroxy-derivative was obtained ;& this was identical with4' : 8'-dihydroxy-l : 2 : 5 : 6-dibenzanthracene (compare V) synthesised byJ. Cason and 1;. F. Fieser.& An interesting feature of these biochemicaloxidations is that oxidation takes place a t positions which are not thosenormally attacked by chemical oxidising agents.This is also the case withanthracene, which was found by E. Boyland and A. A. Levi 45 to give variousoxidised products in which the meso-ring had not been attacked. Possiblythe reactive positions normally attacked in chemical oxidation are protectedby some kind of conjugation with the enzymes on which the hydrocarbonsare adsorbed during biochemical oxidation. In connection with this it may$ 7 ,T. Arner. Chem. SOC., 1941, 63, 2561.3* C. L. Hewett end R. H. Martin, ibid., p. 1396.J. G . Chalmers and D. Crowfoot, Biochem. J., 1941, 35, 1270.42 E. Boyland, A. A. Levi, E. H. Mawson, and E. Roe, ibi&., p. 184.43 K. Dobriner, G. I. Lavin, and C. P. Rhoads, Cancer Reeearch, 1942, 2,44 .J.Amer. Chem. SOC., 1940, 62, 2681.46 Biochem. J . , 1936, 29, 2670; 1936, 30, 728.38 C. L. Hewett, J.,4O C. L. Hewett, J.,1940, 293.1938, 193.79160 ORGAN10 CHEMISTRY.be pointed out that positions 4' and 8' of the 1 : 2 : 5 : 6-dibenzanthracenemolecule are the positions of substitution when the 9 : 10-quinone issulphonated .44Structure.Study of a series of derivatives of naphthacene (2 : 3-benzanthracene)(XIII) by L. F. Fieser 46 showed clearly that attachment of the fourth benzene.ring increased the stability of the meao-dihydro-structure of anthracene anddecreased that of the completely aromatic structure. These differences areenhanced in pentacene (2 : 3 : 6 : 7-dibenzanthracene) (XIV), a deep blue,highly reactive hydrocarbon to which E.CIar and Fr. John *7 assigned thestructure of a diradical. That the diradical structure is inadmissible wasshown by measurements of magnetic susceptibility carried out by E. Mullerand I. Muller-Rodloff ,** who concluded that the solid hydrocarbon cannotcontain more than 1% of a diradical. The series of linear benzologues ofanthraceiie has been extended by the synthesis of hexacene (XVII), a deepgreen hydrocarbon, by E. Fusion of a mixture of phthalic anhydrideand 1 : 5-dihydroxynaphthalene with aluminium chloride-sodium chlorideat 200" gives the dihydroxy-diquinone (XV),50 which on reduction by Clar'selegant method of fusion with zinc chloride, sodium chloride and zinc dust 51passes into the orange-red dihydrohexacene (XVI), a typical naphthacenederivative.This is dehydrogenated to hexacene (XVII) by sublimationwith copper powder. Green solutions of hexacene are sensitive to light andair, and are immediately decolorised by maleic anhydride (cf. p. 190). Theorange-yellow solution of dihydrohexacene (XVI) becomes appreciablypaler when boiled, and this is attributed to isomerisation to (XVIII).C . Marschalk 49 obtained the orange dihydrohexacene (XVI) by reductionof a tetrahydroxyhexacenequinone which he prepared by condensation of4 6 J . Amer. Ghem. SOC., 1931, 53, 2329.4 7 Ber., 1930, 63, 2967.4g Ber., 1939, 72, 1817; see also ibid., 1942, 76, 1283; cf. C. Marschalk, BulZ. SOC.60 E. Clar, U.S.P. 2,210,396.4 8 Annalen, 1935, 517, 134.chim., 1939, 6, 1112.61 Ber., 1939, 72, 1645COOK : POLYUYULIC AROMATIC UOMPOUNDS.161Zezcco-quinizarin with naphthalene-2 : 3-dicarboxylic anhydride or of Zezcco-naphthaquinizarin with phthalic anhydride. By similar condensation ofZeuco-quinizarin with anthraoene-2 : 3-dicarboxylic anhydride he obtained 52a tetrahydroxyheptacenequinone, and with Zeuco-naph t haquinizarin theproduct was a tetrahydroxyoctacenequinone. The parent hydrocarbons,heptacene * and octacene, have not been described, although Marschalkprepared colourless hexa- and tetra-hydrides of heptacene, as well as its blue-violet dihydride (probably XIX).H,By condensation of the dianhydride of anthracene-2 : 3 : 6 : 7-tetracarboxylicacid with Zew-quinizarin, Marschalk 53 obtained a diquinone (XX) in whichthere are no fewer than 11 Zinear condensed rings.In their deep colour and high reactivity these linear benzologues ofanthracene are in striking contrast to the colourless and comparatively inertangzllar benzologues (e.g., 1 : 2-benz- and 1 : 2 : 5 : 6-dibenz-anthracene).The reaction of naphthacene, pentacene and hexacene, but not of anthracene,with sulphur to give products which no longer have the characteristics of theparent hydrocarbons is a further illustration of the reactivity of these linearbenzologues of anthra~ene.~~ Thus, pentacene reacts with sulphur in boilingtrichlorobenzene to give a microcrystalline green pigment containing approxi-mately six atoms of sulphur combined in a molecule of pentacene.These differences in reactivity of the benzologues of anthracene aredoubtless related to their bond structure, which E.Clar 55 has attempted tocorrelate with their absorption spectra. He points out 56 that the quinonecorresponding to a very reactive hydrocarbon has very little reactivity, andvice versa. Naphthacenequinone (XXXIII) is converted into a vat (i.e.,an alkaline solution of the quinol) only with diiliculty; pentacenequinonedoes not give a vat. The reactivity of a quinone, i.e., its tendency to attracthydrogen, is determined by its reduction potential. Clar has developednumerical relationships from which he claims that it is possible to predict theabsorption spectrum (k., the colour) and reactivity of one of these hydro-carbons, and also the reduction potential of the quinone, even when thes2 Bull.SOC. chirn., 1941,8,364. 62 Idem, ibid., 1939,6,1122.s6 Ber., 1936, 69, 607; 1940, 73, 81, 596.* The synthesis of heptacene, an " ultra-green " hydrocarbon, has been described ina paper which became available after this Report was written (E. Clar, Ber., 1942, 75,1330). The compound has the expected high reactivity, and in a vacuum at 320" itdisproportionates to give a dihydride of heptacene.63 Ibid., 1942,9,400.c 6 Ber., 1940, 73, 104162 ORUmCY UHEMISTRY.hydrocarbon and its quinone are both unknown. For details of theserelationships the original papers should be consulted. It will be noted thatwhen more than two rings are fused in a &war condensation it becomesimpossible to represent them all by Kekul6 formula, and a " polyene " type ofsystem develops.This condition does not regult from angular fusion of rings.The bond-structures of polycyclic aromtic hydrocarbons and theirquinones have been investigated by both chemical and physical methods.The chemical reactivity of anthracene is higher than that of phenanthrene,and this is reflected in their respective heats of combustion; the value foranthracene is some 7 kg.-cals. greater than that for ~henanthrene.~7 Clearlythis is related to the fact that in the arnguEar structure, but not the linearstructure, it is possible for all three rings to be completely benzenoid incharacter. In terms of Kekul6 formulz the bond-structures (XXT) and(XXII) for anthracene are open for consideration. Of these, K.Fries, R.BrWalter, and K. Schilling 57 preferred the first (XXI), as there is a closeraverage approach to the stable condition of an isolated benzene ring. Totest this, they studied the bromination of 2 : 6-dihydroxyanthracene andobtained the 1 : 5-dibromo-compound (XXIII). This was regarded assupporting formula (XXI) ; the disposition of double bonds in the alternativeformula would favour 3 : 5-substitution. As is. justly pointed out by L. F.E'ieser and W. C. Lothrop,58 this evidence is inconclusive, as the two bond-structures may be tautomeric with one of them predominating or morereadily substituted. In an attempt to resolve this uncertainty the Americanworkers applied a type of method which they had previously used to studythe bond-structure of na~hthalene.~~ They found that neither the 1 : 5-dimethyl nor the 1 : 5-diallyl derivative of 2 : 6-dihydroxyanthracene wouldcouple with diazotised sulphanilic acid or p-nitroaniline and concluded thattautomerism is negligible and that these anthracene derivatives have thebond-structure shown in formula (XXIII).If the tetrasubstituted anthra-cenes were derived from the structure (XXII), then diazo-coupling a t a freep-position would be expected. It is assumed that the parent hydrocarbonlikewise has the bond-structure (XXI). This conception of fixed doublebonds will hardly find general acceptance, but the growing body of evidencethat condensed-ring polycyclic compounds do, in fact, behave as if theirmolecules contained fixed double bonds cannot be discounted.In the case ofanthrmne the position is perhaps expressed by the statement that thebond-structure (=I) is the one which makes the predominant contributiont o the resonance atate of the molecule.c7 Annabn, 1935, 516,248.6e L. F. Fieser and W. C. Lothrop, ibitE., 1936, 6'4, 1469.ti* J. Amer. Chcm. SOC., 1936, 68, 749COOK : POLYUYOLIO AROMdTfO (30MFOUNDS. 163An interesting type of tautomerism in the anthracene series has beendescribed by E. Bergmann and (Mrs.) 0. Blum-Bergmann,w who found thatthe meso-&chlorides (XXIV) of 9 : 10-diphenyl- and 9 : 10-di-a-naphthyl-anthracene lose hydrogen chloride t o give the 2.chloro-9 : lO-diarylctnfhracene(XXV). This is interpreted &e follows :(XXrV.) (XXV.)It is concluded from dipole measurements that the dichloride of 9 : 10-diphenylanthracene (XXIV) has the cis-configuration (cf.p. 174).The bond-structure of fluorene (XXVI) was studied by W. C. Lothrop,61who considered that this problem should be susceptible to the methods ofinvestigation used by Fieser and Lothrop 62 for other polycyclio structures ;namely, rearrangement of ally1 ethers, and diazo-coupling of hydroxy-derivatives. It was concluded that there is little or no fixation of doublebonds, in which respect fluorene resembles benzene, tetralin and indane,rather than naphthalene, anthrltoene and phenmthrene. W. C. Lothrop andJ. A. C o h a n 68 failed to find any evidence that the Mills-Nixon effectwas operating to cause strain in the anhydride ring of fluorenone-2: 3-dicclrboxylic anhydride (XXVII).A comprehensive investigation of the chemistry of pyrene was madeby H.Vollmann, H. Becker, M. Corell, and H. S t r e e ~ k . ~ ~ The bond-structures (XXVIII) and (XXIX) are consistent with the chemical reactivityof pyrene, including its degradation by ozone.65 The failure of 3-hydroxy-pyrene to undergo diazo-coupling suggests bond-fixation, in which case thehydroxyl group would occupy one of the asterisked positions.Incidentally these are the positions which are attacked when two substituentsare introduced into the pyrene molecule; mixtures of 3 : 8- and 3 : 10-compounds are invariably formed.The mode of interaction of carbonyl compounds with organometallicoompounds has also been used to determine the effective bond-structures of60 J .Amer. Chm. Soc., 1937, 59, 1439.63 IM., 1941, 63, 2564.e5 See also L. F. Fieser and F. C. Novello, J . Amw. Uhcm. rSoc., 1940,68,1855.REP.-VOL. XXXIX, B61 Ibid., 1939, 61, 2116.13' Anmlen, 1937, 581, 1.Ibid., 1936, 58,749, 2050; 1937, 59,945164 OROANIO OHEMISTRY.polycyclic aromatic compounds. The observation of G. Charrier and E.Ghigi 66 that addition of methylmagnesium iodide to the cmbonyl group ofmeso-benzanthrone is accompanied by 1 : 4-addition has been extended byC. F. H. Allen and S. C. O~erbsugh,~~ who showed that both phenylmag-nesium bromide and benzylmagnesium chloride, as well as n-heptyl-magnesium bromide and cyclohexylmagnesium chloride, react with meso-bemanthrone (XXXI) to give 4-substituted derivatives (XXX).Similarresults were obtained 68 in the interaction of Grignard reagents with1 '-phenylmesobenzanthrone. In many cases high yields of the 4-derivativesare obtained and Allen and Overbaugh consider that this indicates thatmeso-benzanthrone must have the bond-structure shown in formula (XXX)or (XXXI) .(XXX.) (XXXI.) (XXXII. )The &st of these accords with the usual bond-structure of naphthalene, and isalso compatible with the failure of meso- benzanthrone to combine with maleicanhydride.Anthraquinone reacts with Grignard reagents exclusively by addition tothe carbonyl group, which indicates the bond-structure (XXXII). In thecase of naphthacenequinone the presence of a naphthalene system disturbsthe normal anthraquinone bond structure, and in agreement with theformulation (XXXIII), C.F. H. Allen and L. Gilman 69 found that naphtha-cenequinone adds on phenylmagnesium bromide (phenyl groups at thepositions asterisked) to give a tetrahydroquinone which is oxidised by airin alkaline suspension to a diphenylnaphthacenequinone. Further study ofthis Grignard condensation 70 showed that the normal diol (XXXIV) is alsoformed but in the presence of magnesium this diol is reduced (apparently bythe Gomberg reagent, Mg + MgBr,) 71 to diphenylnaphthacene. Thisreduction of diols is peculiar to the products from naphthacenequinone anda tetrahydronaphthacenequinone, and is associated with the abnormal bond-structure.II' I\ \/(XXXIII.) (XXXIV.) (XXXV.)(R = H, OMe or OEt.)66 Cfazzetta, 1932, 63, 928.6 8 Idem, ibid., p.1322.70 C. F. H. Allen and A. Bell, ibid., 1940, 62, 2408.7 1 Compare M. (Xomberg and W. E. Bac-, ibid., 1927,49,236.J . Anzer. Chem. SOC., 1935, 57, 740.Q9 Ibid., 1936, 58, 937COOK : POLYCYCLIC AROMATIC COMPOUNDS. 1651 : 4-Addition of Grignard reagents is also shown by 2 : 3 : 6 : 7-dibenzanthraquinone (pentacenequinone) and its derivatives, 72 althoughphenyl-lithium reacts normally, and by perinaphthenone * derivatives ofthe type (XXXV), which react with phenylmagnesium bromide with intro-duction of a phenyl group at the asterisked position.73 In the case of a3 : 4-benzpyrene derivative related to meso-benzanthrone (XXXVI) evenmethyl-lithium leads to 1 : 4-addition as well as addition to the carbonylgroup, so that a methyl group may be introduced into position 6.'*I1 I I//\/\//\/I II I IIOH CO NH*C,H,Me/\/'\AI I1 II IThe structure of some anthraquinone acid dyes has been the subject of aninteresting series of papers by C.F. H. Allen and his collaborator^.^^ Thesewere concerned primarily with toluidine-blue and toluidine-green, little-useddyes which had long been on the market but their constitution had not beendisclosed. Toluidine- blue has certain absorption spectral characteristicswhich distinguish it from other blue dyes. Elementary analysia of thepurified blue dye gave the formula C2,H2,010N2S,Na, ; it forms a yellow vat ;the zinc dust fusion reduction method of Clar 51 gave anthracene, a sulphide,and a purple solid, C2,H2,04N~.This indicated that the dye is a sulphonatedanthraquinone derivative. Oxidative hydrolysis " with concentratednitric acid gave 1 : 4 : 5 : 8-tetrahydroxyanthraquinone, the structure of whichwas pro;ed by its synthesis from 3 : 6-dimethoxyphthalic anhydride andquinol. The purple product, C2,H2,04N2, was shown by its synthesis from4 : 8-dichloroanthrarufin and ptoluidine to be 1 : 5-di-p-toluidino-4 : 8-dihydroxyanthraquinone (XXXVII) , and from it toluidine- blue was obtainedby sulphonation. Toluidine-green was similarly shown to be the corres-ponding 1 : 4-di-p-toluidino-5 : 8-dihydroxy-compound. Finally, " reductivehydrolysis " of toluidine-blue by tin and hydrochloric acid gave Zeuco-1 : 4 : 5 : 8-tetrahydroxyanthraquinone and 4-aminotoluene-3-sulphonic acid,by which the positions of the two sulphonic acid groups were established.Comparison of the absorption spectra of these two dyes with others ofsomewhat similar structure enabled further conclusions to be drawn.a-Hydroxyl groups have a much greater influence on the absorption spectrum72 C. F. H. Allen andp. Bell, J . Amer. Chem. SOC., 1942, 64, 1253.73 C. F. Koelsch and R. H. Rosenwald, ibid., 1937, 59,2166; J. Org. Chem., 1938, 3,72 L. F. Fieser and E. B. Hershberg, J . Amer. Chem. SOC., 1938, 80, 2542.76 C. F. H. Allen, G. F. Frame, and C. V. Wilson, J . Org. Chem., 1941, 6, 732; 1942,7, 63; C. F. H. Allen, C. V. Wilson, and G. F. Frame, ibid., 1942,7, 68, 169.* The nomenclature of this ring system was discussed by J.W. Cook and C. L.Hewett (J., 1934, 368). In the present report the system proposed by L. F. Fieser andE. B. Hershberg ( J . Amer. Chem. SOC., 1938, 60, 1659) is used, whereby 1 : 8-trimethyl-enenaphthalene is termed " perinaphthane."462; C. F. Koelsch and J. A. Anthes, ibid., 1941, 6, 558106 OWANIO CHEMISTRY.than p-hydroxyl groups, an effect which is attributed to chelation with theadjacent carbonyl group. In the case of the sulphonated dyes a sodiumsulphonate group at position 2’ has a markedly different effect from that ofthe same group at position 3’ or 4’. This is regarded as indicating a hydrogenbond involving the imino-hydrogen, so that the complete structure(XXXVIII) is assigned to the anion of toluidine-blue.A n analogousstructure may be given to toluidine-green. It is suggested that in theisomeric 3’- and 4’-sulphonates the imino-hydrogen may be bonded to thecarbonyl oxygen, as in (XXXIX) or (XL).1 : 5-Di-~-toluidinoanthraquinone baa 6 smooth absorption curve for themain band in the visible region of the Bpectrum, whereas the 1 : 4-compoundshows two maxima on this band. This double head appears only when bothpositions 1 and 4 have substituents which are able to furnish electrons by amesomerio shift (XLI). Sulphonation at positions 3’ and 4’ restrains thismesomeria shift on account of the strong inductive effeot of the positivesulphur atom, transmitted to nitrogen through the aromatio ring. If thesulphonic acid group is at position 2’, hydrogen bond formation (XXXVIII)will result in a tendency for the hydrogen to release its electrom shared withnitrogen, thus neutralising the inductive effect of the sdphonic acid group.Allen and hia collaborator8 give a considerable body of spectroscopic datawhioh are in acoord with these theoretical views.The strwtures of some metallia ‘‘ hkee ” of alizarin have been studiedby W.I?. Beech and H. D. IC. D r e ~ . 7 ~Mesomerism i s invoked by R. Soboll to account for the marked differencesin colour between two series of salts of hydroxyanthraquinones with ammoniaand amines 77 and a re-interpretation of the structures of the variouslyrtoloured anthraquinol -a-carboxylic acid laotones and their salts is also gi~en.7~Space does not permit of a summary of thia work.The unsymmetrical struoture (XLII) for pioenequinone was establishedby J.W. Cook 79 who showed that its degradation product, picylene ketone,J., 1940, 603.7 7 R. BchollaadP. J. Dabll, Ber., 1941,74,1129; R. Saboll, ibid., p. 1171.78 R. Scholl, I(. Meyer, and C. Seer, ;bid., p. 1182. w J., 1941, 686COOE : POLYOY~IO AROMATIO COIKPOUNDS. 167fs identical with a synthetic specimen of 2’ : 1’-naphtha-1 : 2-fluorenone(XLIII) .( XLII . )Molecular Compounds.The problem of the structures of the highly coloured molecular compoundsof polycyclic aromatic hydrocarbons with polynitro-compounds has longattracted attention.sl J. Weiss 82 has recently suggested that these mole-cular compounds are ionic in character, the coloured positive ion beingformed by transfer of an electron from the hydrocarbon (donor) molecule tothe molecule of the polynitro-compound. This simple theory has much tocommend it, but lacks experimental verification.In an earlier publication 83Weiss described a dark brown perchlorate of anthracene which he formulatedas [CI4Hlo]*[C1O4]-, the formation of the positive ion being attributed here toloss of an electron by oxidation. There is no evidence that oxidation isconcerned in the formation of this complex, and in the opinion of theReporter the compound probably contains perchloric acid, rather than theperchlorate ion.Highly coloured addition compounds of polycyclic compounds withbromine and iodine have been described. The halogen in these is looselybound and may be removed by treatment with sodium thiosulphate.Thesecomplexes are probably of the same type as those formed with polynitro-compounds. Substances which have been found to give these unstablecoloured hdogenides include meso-bemanthrone ( XXXI),84 perinaphthenone(XXXV; R = H),85 perylene * (XLV),86 anthanthrene (XLIV) 87 and10 : lO’-dihydro-9 : g’dphenanthrylidene (XLVI) .88 Highly coloured mole-cular oompounds of polycyclic bydrocarbons [l : 2- and 2 : 3-benzanthrscene,9 : B’-dianthranyl, 1 : 12-benzperylene (I), 2 : 3 : 10 : ll-dibenzperylene] withantimony penhohloride and (in 8ome cases) stannic chloride have been8o J. W. Cook, C. L. Hewett, W. V. Mayneord and E. Roe, J., 1934, 1727.81 See Ann. Reports, 1931, 28, 134.J., 1942, 245.K.Brass and E. Clar, Ber., 1938,60,090.83 Nature, 1941, 147, 512.85 Idem, Ber., 1939,72,1882; A.M. Lukin, Compt. rend. Acacl. Sci., U.R.S.S., 1940,86 K. Brass and E. Clar, Ber., 1932, 65, 1660; A. Zinke, H. Troger, and E. Posch,8 7 K. Brass and E. Clm, Bey., 1939,72, 1882.6 8 E. Berpann and F. Bergmatnn, J. Arne*. Chem. SOC., 1937, 59, 1449,* In the case of perylene (XLV) cornplfoations are caused by addition of brornineto “ unsaturated ” centres, followed by elimination of hydrogen bromide, so thatsubstitution of bromine for hydrogen accompanies molecular compound formation.% , G O .Ber., 1941, 74, 107 ; M. Pestemer and E, Treiber, ibid., p. 964168 ORGANIC CHEMISTRY.described by K. Brass and K. Fanta.89 Possibly these have structures of thesame type as those of the complexes of polynitro-compounds. The Reporterhas observed that pyrene and 1 : 2-benzanthracene v readily form crystallinecomplexes with antimony trichloride ; the complex from pyrene is colourless.H.Beyer and J. Richter have described a hexacyclic hydrocarbon whichgives a colour with antimony trichloride. Perinaphthenone (XXXV ;R = H), in contrast to meso-benzanthrone (XXXI), has marked basicproperties 91 and forms a series of coloured salts with mineral acids.92In connection with the structure of these coloured molecular compoundsthe choleic acids which some of the polycyclic hydrocarbons form whencrystallised with deoxycholic acid are of interest.93 It by no means followsthat these curious compounds, which contain from 2 to 4 molecules ofdeoxycholic acid combined with 1 molecule of hydrocarbon, possess the sametype of structure as the complexes of hydrocarbons with polynitro-compounds.It is significant that, unlike picrates, which are completely dissociated byalkali, the choleic acids are sufficiently stable to form water-soluble sodiumsalts.94 W.Marx and H. Sobotka 95 adduce evidence that choleic acids arecompletely dissociated in solution.Sterwchemiatry and Xteric Factors.The characteristic ultra-violet absorption spectrum of diphenyl isassociated with a planar molecule, for M. Pestemer and E. Mayer-Pitsch 96and M. T. O’Shaughnessy and W. H. Rodebush 97 found that in o-substituteddiphenyl derivatives in which rotation is restricted so that the rings are notco-planar, the diphenyl spectrum is no longer exhibited and the two halvesof the molecule behave as independent resonators. These observations havebeen used by R.N. Jones 98 in a spectroscopic study of the configuration ofsome condensed ring systems. The spectrum of 9 : 10-dihydrophenanthrene(XLVII) is very similar to that of diphenyl, from which the inference isdrawn that the two aromatic rings approximate to a co-planar configuration.8g Ber., 1936, 69, 1.9 1 J. W. Cook and C. L. Hewett, J., 1934, 369.92 A. M. Lukin, Bull. Acad. Sci. U.R.S.S., CZasse sci. chim., 1941, 411.g3 L. F. Fieser and M. S. Newman, J . Amer. Chem. SOC., 1935, 67, 1602.94 See dso,A. Winterstein and H. Vetter, 2. physiol. Chem., 1934, 230, 169; L.F.g6 J . Org. Chem., 1936, 1, 276.97 J . Amer. Chem. Soc., 1940, 63, 2906.Ber., 1940, 72, 1319.Fieser and A. M. Seligman, J . Amer. Chem. SOC., 1936, !j8,2480.s6 Monatsh., 1937, 70, 104.Ibid., 1941, 08, 1668COOK : POLYCYCLIC AROMATIC COMPOUNDS. 169This is likewise the case with 4 : 5-methylene-9 : 10-dihydrophenanthrene(XLVIII) ; hence any distortion from the planar configuration which thestrain in the 5-membered ring might tend to produce is effectively preventedby an opposing influence of the second bridge. Jones found that theabsorption spectrum of anthracene is practically unaltered by aryl sub-stitution in the meso-positions (9-phenyl- and 9 : 10-diphenyl-anthracene ;9 : 9’dianthranyl). He therefore concludes that the planes of the merro-arylgroups are inclined to that of the anthracene system, as indicated in (XLIX)for 9-phenylanthracene.The same considerations hold for rubrene (L), thespectrum of which closely resembles that of the unsubstituted hydrocarbon,naphtha~ene.~~ A non-planar configuration of rubrene isX-ray crystallographic evidence.1(XLVII.)supported by(XLVIII.)Ph CO,HA steric influence of meso-aryl groups on substituents in the lateral rings ofthe anthracene system is shown in the case of 9 : lo-diphenylanthracene-1 : 5-dicarboxylic acid (LI), which cannot be esterified by alcohol and acid.2The ester, obtained through the silver salt, is extremely resistant tohydrolysis, 90% being recovered unchanged after 30 hours’ boiling with 25%methyl-alcoholic potash.A striking example of the influence of spatial factors on chemicalreactivity is found in the Diels-Alder addition product (LII) of 9-bromo-anthracene and maleic anhydride.This adduct was prepared by E. de B.Barnett, N. F. Goodway, A. G. Higgins, and C. A. Lawrence * 3 and foundby them to give no ionised halogen after 40 minutes’ boiling with alcoholicpotash. The adduct of 9 : 10-dichloroanthracene showed similar resistance,although E. Clar had found that the chlorine atoms are displaced by aryl99 C. Dufraisse and R. Horclois, Bull. SOC. chim., 1936, 3, 1880.E. Bergmann and E. Herlinger, J. Chem. Physics, 1936, 4, 532.R. Scholl, H. K. Meyer, and W. Winkler, Annalen, 1932, 494,201.J., 1934, 1224. Ber., 1931, 04, 2194.* These authors prepared the adduct in boiling o-dichlorobenzene, but do not mentionW.E. Brachmann and M. C. Kloetzel ( J . Org. Chem., 1938, 3, 55) obtained the yield.an almost theoretical yield in boiling xylene170 OBGANIa QBEMISTRY.groups in the Friedel-Crafts reaction. P. D. Bartlett and S. G. Cohen tirebveatigated the behaviaur of (LII) and found that treatment for 16-18hours with boiling 15-30% potassium hydroxide solution had no other effectthan to open the anhydride ring and iaomerise the cis-dibasic; acid into thetrans-aoid.6 In fact, the bromine atom in this compound W&B found to be a tleast a million times less reactive than the bromine atom in 9-bromo-9-methylfluorene. This is in marked contrast to the well-known labilitynormally shown by bromine atoms ih the meso-positions of 9 : 10-dihydro-anthracene.Chemioal inertness is shown also by the adducts of 9-nitro- and 9-metamido-anthracene with mdeia anhydride ; for example, the nitro-compound is not reduced by stannous chloride.Bartlett and Cohen pointout that these adducts are among the few compounds which are structurallyincapable of replacement reactions accompanied by Walden inversion. Thiscircumstance may account for the inertness of the bromine in (LII), but canhardly explain the resistance to reduction shown by the corresponding nitro-compound. I?. Bergmann ' found that the adduct of anihracene with(LIII. )maleio anhydride does not undergo the dohydrogenation with nitrobenzenecharacteristic of many other hydroaromatic structures formed in dienesyntheses .BAn interesting hydrocarbon with a multiplrtnar moleoule (LIII) has beendescribed recently by P.D. Bartlett, M. J. Ryan, and S. G. C0hei1.~ Thiswas obtained by reduction, through a series of intermediates, of the Diels-Alder adduct of anthracene and p-benzoquinone. It wa8 given the nametriptycene (from triptych), and it was found that the meso-hydrogen atomsshow none of the reactivity of the similarly linked hydrogen of triphenyl-methane. There was no exchange with phenylisopropylpotasaium, and nochlorination by sulphuryl chloride in presence of benzoyl peroxide. Oxid-ation with chromic acid gave only anthraquinone and carbon dioxide.Intermediate hydroxy-compounds analogous to triphenylcarbinol were notformed, The anomalous behaviour is attributed to a damping of resonanceby the structural rigidity, with a preference for the bond Structure shown in(LIII) .The question of the configuration of the fluorene molecule was revived byJ .Amer. Chern. Soc., 1940, 63, 1183.Compare 0. Diels and K. Alder, Annalen, 1931, 486, 191.J . Amer. Ohem. rSoc., 1942, 04, 176.Compare E. Clar, BeT., 1936, 69, 1686; E. Bergmann, L. Haskelberg, and F.J . Amsr. Chem. SOC., 1942, 64, 2649.Bergmann, J. Org. Chem., 1942, '7, 303COOK : POLYUYCLIU AROMATIC COMPOUNDS. 171J. Iball's examina,tion lo of the crystal structure of fluorene by X-ray anaIpis.The data which he obtained accord with a structure in which the aroma.ticnuclei &re inclined to one mother at an angle of about 20", but are Mcultto reconcile with a planar moleoule.Thus in the solid state fluorene a p p mto have 8 non-planar conf?guration, and the stereochemical implications ofthis were disaussed by J. W. Cook and J. Iball,ll who reviewed the literatureof attempts to prepare stereoieomeric fluoreine derivativea dependent on sucha non-planar configuration. The dipole moments of fluorene, fluorenone anda number of derivativea were mectsured by E. D. Hughes, (Mrs.) C. G. Le FBvre,and R. J. W. Le F&vre,12 who were able to exclude a number of possibleconfigurations, but their results (obtained with solutions) neither supportednor excluded a non-planar configuration. A non-planar configuration wouldtend to relieve the strain in the 5-membered ring, but would imply anunprecedented distortion of the vdency directions of benzene out of the planeof the aromatic ring.Spatial influences have been encountered in investigations relating t o thesynthesis and reactions of a number of polycyclic aromatic compounds. Thusin the application of the Pschorr phenanthrene synthesis to the acid (LIV)ring-closure takes place mainly at the p-position of the naphthalene ring andonly t o a lesser extent at the more reactive a-position.Accordingly, the1 : 2-benzanthracene structure is formed in addition to the 3 : 4-benz-phenanthne structure.13 The cyclisation of y-3-phenanthrylbutyric acid(LV) takes place almost exdusively a t position 2, with the formation of akefotetrahydro-1 : 2-ben~anthrracene.~~ That failure t o cyclise at position 4is not due to chemical inertness at this position is shown by the fact thatposition 4 is attacked when 3-sminophenanthrene is submitted to the Skraupreaction, the product being represented by (LVI).15 Also W.E. Bachmannl6 found that cyclisation of the chloride of f1-3-phenan-occurred at position 4 exclusively.and M. C. .KZoetzelthrylpropionic acidOther examples of the phenomenon just discussed were furnished byy-8-methyl-2-naphthylbutyrie acid (LVII) l7 and 7-5 : 6 : 7 : 8-tetramethyl-2-naphthylbutyric acid (LVIII),lS which undergo oyclisation at position 3 to10 2. Krist., 1936, A , 94, 397.l2 J., 1937, 202.l4 R. D. Hsworth and C. R. Mavin, J., 1933, 1012.l6 E. Mosettig and J.?V. Kreuger, J. Org. Chem., 1938, 8, 317.l6 J . Amer. Chem. SOC., 1937, 59, 2207.18 C. L. Hewett, J . , 1940, 293.l1 Chern. and Ind., 1936,55,467.la J. d. Cook, J., 1931, 2524.R. D. Haworth and G. Sheldrick, J., 1934, 1950172 O&GANIC CHEMISTRY.give an anthracene derivative, whereas the parent y-2-naphthylbutyric acidundergoes ring-closure a t position 1 to giveThe experiment with (LVII) was made in the course of an unsuccessfulattempt to synthesise 4 : 5-dimethylphenanthe (LIX). Introduction oftwo methyl groups in these 4 : 5-positions would impose considerable strainon the molecule and could not be effected without distortion. Even thePschorr reaction failed to give this type of structure, for the amino-acid (LX)did not undergo ring-closure under the usual conditions.la The chemicalbehaviour of the few known 4 : 5-substituted phenanthrenes is governed bythese considerations of spatial configuration.Thus, 4-phenanthraldehyde-5-carboxylic acid (LXI), obtained by ozoniaation of pyrene,20 is stable toalkaline oxidation, and is readily transformed back into pyrene derivatives.Treatment with alcoholic potttsh or potassium cyanide gives 1 : 2-pyrene-quinone (LXII) , whereas hydrazine hydrate gives 1 - hydroxypyrene (LXV) .phenanthrene derivative.Me(LVIII. )co(LXI.) (LXII. j\Unsuccessful attempts have been made to synthesise 1’ : 9-dimethyl-1 : 2-benzanthracene (LXIII), of analogous structure to 4 : 5-dimethylphenan-threne.21 M. S. Newman e2 has succeeded, however, in synthesising a hydro-carbon which is believed to be the similarly constituted 6 : 7-dimethyl-chrysene (LXIV) .* Its ultra-violet absorption spectrum was examined by1s E.E. Lewis and R. C. Elderfield, J . Org. Chern., 1940, 6, 290,20 H. Vollmann, H. Becker, M. Corell, and H. Streeck, Annulen, 1937, 531, 66;21 L. F. Fieser and A. M. Seligman, ibid., 1938, 60, 170; 1939, 61, 136.2a Ibid., 1940, 62, 2295.* In the original paper the Patterson system of numbering is used, so that thecompare L. F. Fieser and F. C. Novello, J . Amer. Chem. Soc., 1940, 62, 1855.compound is described &e 4 : 6-dimethylchryseneCOOK : POLYCYCLIC AROMATIC COMPOUNDS. 173R. N. Jones,23 who found that it had the chrysene type of spectrum, which,nevertheless, differs in some respects from those of other methylchrysenes.Clearly the molecule of this dimethylchrysene must be distorted, either bymodification of the angles of the aromatic rings, or by accommodating themethyl groups in a non-planar configuration. If the latter be the case, then acompound of this type may exist in optically active forms, a possibility towhich Newman drew attention.Similarly situated methyl groups in morecomplex ring systems were already known.24 Doubtless in the more complexmolecules their greater flexibility enables the strain to be distributed through-out the ring system.Somewhat similar considerations should apply in the case of 3 : 4 : 5 : 6-dibenzphenanthrene (LXVI). In an undistorted molecule there would beno room for the hydrogen atoms at the asterisked positions, for the distancebetween the two carbon atoms would be the ordinary interatomic distanceof the carbon atoms of a benzene ring.Consequently this hydrocarbon shouldundergo very ready dehydrogenation to 1 : 12-benzperylene (I). Although3 : 4 : 5 : 6-dibenzphenanthrene is this conversion has not beenrecorded. That such a tendency does exist was shown, however, in the caseof a carboxylic acid (LXVII) derived from a tetrahydroxy-3 : 4 : 5 : 6-dibenzphenanthrene ; this acid on dehydrogenation with sulphur lost sixhydrogen atoms to give a 1 : 12-benzperylenecarboxylic acid (LXVIII) .26Of interest in this connection is Newman's synthesis 30 @* 16') of coronene(11) from a 3 : 4 : 5 : 6-dibenzphenanthrene derivative (LXIX) by heatingwith fused potassium hydroxide.It will be noted that the hypotheticalintermediate 1 : 10-benzperylene derivative contains methyl groups inpositions equivalent to the 4 : 5-positions of phenanthrene. These methylgroups suffer loss of hydrogen to form the final ring of coronene./\/\I II II II I\/\/\/y\ I1/\A/(LXVI.)/\A I II I\A/\ I II/\/\/ w (LXVII.)CO,H -+/\/\(LXVIII.)HvxC123 J . Amer. Chem. SOC., 1941, 68, 313. *' R. Scholl and C. Tiinzer, Annalen, 1923, 433, 172; D.R.-P. 458,710; compareR. Scholl and K. Meyer, Ber., 1934, 07, 1232.26 J. W. Cook, J., 1933, 1692.26 C. L. Hewett, J., 1938, 1286; compare H. A. Weidlich, Ber., 1938, 71, 1203174 OBaANIO UHDMTSTRY.It is now recognised that addition of halogens to a, double bond is a two-stage process and that the entering halogen atoms arrange themselves asremotely its possible from each other (tram-addition).Consequently it is ofinterest that measurement of the dipole moments of the dichlorides (LXX)has shown that additions of chlorine to 1 : 6-diohloro- and 9 : 10-diphenyl-anthracene are pure ~is-reactions.~~ The suggestion ia made that this maybe interpreted as due to primary addition of a ohlorine rnokcule rather thanchlorine atoms, a8 in the addition of oxygen or mdeic anhydride to anthraceneand its derivatives (see p. 189). In the oase of 1 : 8-dichloroanthrsoene,however, addition of chlorine is a normal trans-reaction, for the dipolemoment of the dichloride is smaller than that of the parent oompound.Dipole moment measurements of the stereoisomeric diols (LXX ; X = OH)formed by hydrolysis of the 1 : 5odichloroanthracene dihalides me in agree-ment with the configurations assigned to these diols by E.de B. Barnett,J. W. Cook, and M. A. Matthewa.a8The stereochemistry of the Psohorr phenanthrene synthesis has beenstudied by P. R~ggli.8~ This synthesis depends for its auocess on theoiroumsfance that the carboxyl group of o-nitro-a-phenyloinnamio mid(LXXI) and its derivatives induces a cis-configuration of the two aromaticnuclei. trawo-Aminostilbene fails to give phenanthrene, whereas thecis-isomeride does so. In faot, the ring-olosure of atilbene derivatives tophenanthrenes serves as criterion for the cis-configuration.SimilarlyI?. ‘Bergmann so found that both of the stereoisomeric a-(g-phenanthryl)-stilbenes (LXXII) give lithium compounds; but only one of these resotswith carbon dioxide to give 10-phenyl-1 : 2 : 3 : 4-dibenzphenanthrene(LXXIII) ; the other gives olp-diphenyl-a-(9-pen~nthryl)succinio anhydride.(LXXI.) (LXXIII. )Synthesis.So much of the work on polycyclic aromatio compounds published duringthe period under review has been devoted to their synthesis that it isimpossible to give anything approaching a, complete account of this in acomparatively short Report. Much of the earlier work has been summarisedby L. F. Fie~er,~l and in the present Report treatment will be confined largely27 E. Bergmann and A. Weizmann, J . Amer. Chem.SOC., 1938, 60, 1801.28 Rec. Trav. chim., 1925, 44, 728.P. Ruggli and A. Staub, Helv. Chim. Acta, 1936, 19, 1288; 1937, 20, 37;P. Ruggli and A. Dinger, ibid., 1941, 24, 173.J . Amer. Chem. SOC., 1942, 64, 69.31 “ The Chemistry of Natural Products related to Phenanthrene,” New York, 1936COOK : POLYOYCYLIC hBOMATIC 00lWOUNDS. 175to the more recent publications, and especially to methods and techniqueswhioh show novel features.Numerous refinements and modifications have been made to well-knownmethods for the synthesis of anthracene and phenrtnthrene derivatives. Theinteraction of arylmagnesium halides with dicarboxylio anhydrides (e.g.,phthalic or sucoinic anhydride) gives phthrtloylio acids or p-aroylpropionicacids in good yield32 and this reaction has been extensively used for thepreparation of intermediates in the synthesis of anthracene or phenanthrenederivatives, For instance, o-anisylmagnesium bromide waB condensed withphthalic anhydride 33 or o-tolylmagnesium bromide with naphthalene-1 : 2-dicarboxylic anhydride.34 The latter unsymmetrical anhydride gave mainlythe keto-acid (LXXrV), from which homologues of 1 : 2-benzanthraceneare obtainable.Hydroaromatic keto-acids are likewise formed from phenyl-magnesium bromide and A4-tetrahydrophthdic anhydride or from cia-hexa-hydrophthalic anhydride by the Friedel-Crafts reaction.35 Other examplesof interaction of Grignard reagents with phthalio anhydrides are given by R.Bousset,36 L. F. F'ieser and E. B. Herahberg:' and by L. F.Ewer and A. M.Seligman.38 From a-naphthylmagnesium bromide and cycibpentane-1 : 2-dicarboxylic anhydride WM obtained a keto-acid (LXXV) which was con-verted into 2 : 3-cycbpentenophenanthrene (LXXVI) .89 Similarly, condens-ations between 9-phenanthrylmagnesium bromide and succinic anhydride 40or ap-dimethylsuccinic anhydridepl and between a- or p-naphthylmagnesiumbromide and +dimethylsucch.ic anhydride gave keto-acids which wereconverted into homologues of triphenylene (LXXVII) 42 or phenanthrene.(LXXIV.)Ar CO*fiHCO2HCH( LXXVIII . )32 C. Weizmann, E.(LXXVII.)GMe P H 2 \ Ar-CO-QHC02H*CH CMe\CH2/(LXXIX.) (L=.)Bergmann, md F. Bergmann, J., 1935, 1367; C. Weizmann,0. Blum-Bergmann, and F. Bergmarm, ibid., p. 1370; C. Weizmann and E.Bergmenn,J., 1936, 567.33 B. P. Geyer, J. Arner. Ch.ern. Soc., 1942,64,2226.34 L. F. Fieser and M. S. Newman, ibid., 1936, 58, 2376.36 L. F. Fieser and F. C. Novello, ibid., 1942, 64, 802.36 Bull. Soc. chim., 1936, 2, 2182.3 i J . Amer. Chem. Soc., 1937, 59, 1028.39 E. Bergmaan and 0. Blum-Bergmann, ibid., 1937, 59, 1572.4u I d e m , &id., p. 1441.42 See also L. F. Fieser and L. M. Joahel, .ibid., 1939, 61, 2968.38 Ibid., 1938, 60, 170.41 L. F. Fieser and W. H. Daudt, ibid., 1941, 63, 782176 ORUANIC CHEMISTRY.An interesting route to aroylbenzoic acids has been provided by the dehydro-genation with sulphur of compounds of type (LXXX) formed by the Diels-Alder addition of dimethylbutadiene (LXXIX) to P-aroylacrylic acids.43The p-aroylacrylic acids (LXXVIII) were obtained from maleic anhydrideand aromatic hydrocarbons by the Friedel-Crafts reaction.The cyclisation of o-benzoylbenzoic acids to anthraquinone derivatives *is often difficult to effect and better results may then be obtained by pre-liminary reduction to o-benzylbenzoic acids.For this reduction zinc dustand alkali may frequently be successfully used, but in other cases (e.g.,o-l-naphthoylbenzoic acid") the yields are poor. In such cases high-pressure hydrogenation over a copper chromite catalyst has been found veryeffe~tive.4~ The ready dehydration of o-benzylbenzoic acid to anthrone 46is complicated in other cases by sulphonation or by oxidation of the unstableanthranol. In such cases dehydration to anthranyl acetates by aceticanhydride containing hydriodic acid 47 or by acetic acid and acetic anhydridecontaining zinc chloride 48 gives excellent results.Another procedure which has been used for the formation of compoundsof o-benzylbenzoic acid type is illustrated by a synthesis described by L.3'.Fieser and J. C a ~ o n . ~ ~ o-Chlorophenylmagnesium bromide was condensedwith acenaphthenone to give a carbinol (LXXXI), which was converted bydehydration and hydrogenation into o-chlorophenylacenaphthene. Thechlorine in this was replaced by the cyano-group by means of cuprous cyanide,and the product (LXXXII) hydrolysed to the acid (LXXXIII).(LXXXI.) (LXXXII.) (LXXXIII.)In a variant of this, 4-methyl- l-naphthylmagnesium bromide was condensedwith o-chloroacetophenone, and the product transformed through similar43 L.F. Fieser and M. Fieser, J . Amer. Chem. Soc., 1935, 57, 1679.44 R. Scholl, C. Seer, and A. Zinke, Monatsh., 1920, 41, 601.L. F. Fieser and E. B. Hershberg, J. Amer. Chem. SOC., 1937, 59, 1028, 2331;L. F. Fieser and H. Heymann, ibid., 1941, 63, 2333; 1942, 64, 376.46 E. de B. Barnett, J. W. Cook, and I. G. Nixon, J., 1927, 608.4 7 R. Scholl and K. Meyer, Ber., 1932, 66, 1398; R. Scholl, G. von Hornuff, and4 8 L. F. Fieser and E. B. Hershberg, J. Amer. Chem. SOC., 1937, 59, 1028; compare48 J . Amer. Chem. SOC., 1940, 62, 432.* The mechanism of this dehydration has been studied by M. S. Newman (J. Amer.Chem. SOC., 1942, 64, 2324), who suggests that the 6rst stage is the formation of apositive carbonium ion :H.K. Meyer, Ber., 1936, 69, 707.F. F. Blicke and R. J. Warzynski, ibid., 1940, 62, 3191.Q-FCOOK : POLYCYCLIO AROMATIO COMPOUNDS. 177stages into the acid (LXXXIV).50 Compounds of this type, which may beconverted into meso-substituted anthracenes by cyclo-dehydration and thenreduction, have been obtained frequently by reduction of the phthalides(type LXXXV) formed by the interaction of Grignard reagents witho-benzoylbenzoic acids.51In suitable cases moderately complex structures may be obtained byquite simple procedures. Phthalic acids frequently condense directly witharomatic compounds to give anthraquinones when fused with aluminiumchloride-sodium chloride.52 An alternative condensing agent is boric acid,which effects smooth condensation of 1 : 4-dihydroxynaphthalene withphthalic anhydride or naphthalene-2 : 3-dicarboxylic anhydride to give a di-hydroxynaphthacenequinone (LXXXVI) or a dihydroxypentacenequinone(LXXXVII) .53 A monohydroxynaphthacenequinone was obtained fromo-( l-hydroxy-2-naphthoy1)benzoic acid by the well-known cyclisation withboric acid and sulphuric acid.54 3 : 7-Dimethoxy-1 : 2 : 5 : 6-dibenzanthra-quinone (LXXXVIII) was obtained by J. Cason and L. F. Fieser 55 by theaction of aluminium chloride on 4-methoxy-2-naphthoyl chloride.(LXXXVI. ) (LXXXVII. ) (LXXXVIII.)In the dehydration of benzoylbenzoic acids to anthraquinones by acidchlorides at high temperatures,ss the addition of a few drops of concentratedsulphuric acid is recommended by H.Waldma~m.~'The Elbs pyrolysis reaction for the synthesis of anthracene hydrocarbonshas been used for the production of wide variety of these.31 It consists in50 L. F. Fieser and A. M. Seligman, J . Amer. Chem. Soc., 1939, 81, 136.61 L. F. Fieser and M. S. Newman, ibk?., 1936, 68, 2376; M. S. Newman, ibid.,1937, 59, 1003; L. F. Fieser and G. W. Kilmer, ibid., 1939, 61, 862; J. W. Cook,A. M. Robinson, and F. Goulden, J., 1937, 393; A. T. Marchevskii and M. I. Ushakov,J . Ben. Chem. U.S.S.R., 1940, 10, 1369; B. M. Mikhailov and A. N. Blokhina, ibid.,p. 1793.s2 See, e.g., F. Mayer, 0. Stark, 8ndK. Schgn, Ber., 1932, 66, 1336.6s C. Weizmann, L. Hsskelberg, and T. Berlin, J., 1939, 398; compare 0. Dirnroth64 I. Y. Postovskii and L.N. Golyrev, J . Ben. Chem. U.S.S.R., 1941, 11,429.66 J . Amer. Chem. SOC., 1941, 63, 1256.'* E.g., D.R.-P. 590,579.and R. Fick, Annalen, 1916, 411,325.s 7 J . p r . Chem., 1938,160, 121178 ORQLWM CREMfSTRY,the thermal dehydration of derivatives of o-methylbenzophenone. Amodification of this hats been introduced suctcessfully by Heser and hiscollaborators for the synthesis of 20-methyloholanthrene (VI) and com-pounds of analogous structure. Thus, the ketone (LXXXIX), obtainedfrom 7-cyano-4-methylindane and a-naphthylmagnesium bromide, gave20-methylcholanthrene in 50% yield when heated at 405410" for 40rnin~tes.~8 Cholanthrene itself 59 and other of its derivatives and analogues 60have been similarly prepared. The methylene group of a six-membered ringhas also been used in place of the methyl group of the earlier examples of theElbs reaction.61E.Bergmann G2 noted the formation of anthracene in the acid hydrolysisof the acetal of o-benzylbemaldehyde, and this observation was extended byC. K. B r a d ~ h e r , ~ ~ who obtained 9-alkylanthracenes (XCI) in 70-80% yieldsby prolonged heating of o-benzylphenyl ketones (XC) with hydrogen bromidein acetio acid. This reaction bears considerable resemblance t o the Elbspyrolysis. A somewhat similar method was used for the synthesis of 9- and9 : 10-substituted phenanthrenes.64 For example, 9-methylphenanthrenewas obtained from the carbinol (XCII) by refluxing with hydrogen bromidein acetic acid. The formation of 2 : 3 : 6 : 7-tetramethoxy-9 : 10-dialkyl-anthracenes by condensation of veratrole with aliphatic aldehydes is des-cribed by A.MiilIer, M. Raltschewa, and M. Papp;65 the condensation isaccompanied by dehydrogenahion of the intermediate dihydroan thracenederivative. V. I. Khmelevskii and G. I. Fedorov 66 obtained 9 : 10-diphenyl-idem, ibid., 1935, 67, 228, 942.63,301.176, 2255; 1940, 68,2103.68 L. F. Fieser and A. M. Seligman, J . Amer. Chem. SOC., 1936, 58, 2482; compare68 L. F. Fieser and A. M. Seligman, Cbid., 1936, 57, 2174; W. F. Bruce, ibid., 1941,L. F. Fieser et al., ibid., 1935, 57, 1681; 1937, 59, 394, 883, 2561; 1938, 60,61 L. F. Fieser and A. M. Seligman, ibid., 1936, 68,478.62 J. Org. Chem., 1939, 4, 1.63 J. Arner. Chem. SOC., 1940, 62, 486, 1077.61 C.K. Bradsher et al., ibid., 1938, 60,2960; 1939, 61, 1524, 2184.6 6 Ber., 1942, 75, 692. J. Ben. Chem. U.S.S.R., 1939, 9, 1423COOK : PoLYoycLro AROMATIC) COMPOUNDS. 179anthcene in 20% yield by heating benzophenone with oalcium; similartreatment of fluorenone gave the red hydrocarbon, rubitme (XCIII).The action of Grignmd reagents on anthraquinone furnishes a convenientmethod of introducing alkyl or aryl substituents into both meso-positions ofanthracene and its derivatives. On account of the lack of solubility of themquinones in ethyl ether the use of butyl ether hw been suggested.*7 Withalkyl substituents, direct reduotion of the 9 : 10-diols ie troublesome; thedifficulty has been overcome in a very ingenious manner by W. E.Bachmtlnnand J. M. Chemerda,68 who converted the diol from 1 : 2-benzanthraquinoneand methylmagnesium iodide into its dimethyl ether (XCIV), fkom whioh thehydrocarbon (XCrC’I) wm obtained by shaking with powdered sodium inether-benzene. They interpret the reaction 69 as involving replacement ofone methoxy-group by sodium, followed by transannular loss of sodiummethoxide from the product (XCV) :MeOaneah’arMe0 % e--+WW.) (XCv.) (XCVI.)This method gives excellent results and has been used for the synthesisof a number of 9 : IO-dialkylanthracene~.~~ An alternative procedure wasdescribed by R. B. Sandin and L. F. Fie~er,~l who obtained an iodomethylcompound (XCIX) (also prepared directly from 9-methyl-1 : 2-benzan-thracene) by addition of hydrogen iodide to the magnesium complex (XCVII)from 1 : 2-benzanthraquinone and methylmagnesium iodide.The formationof the iodomethyl compound is believed to take place by isomerisation of theintermediate (XCVIII) and its reduction with stannous chloride was statedto give 9 : 10-dimethyl-1 : 2-benzanthracene (XCVI) in 99% yield !(XCVII. ) (XCVIII.)-+MeCH21(XCIX.)9 : 10-Dialkyl-9 : 10-dihydroanthracenea are formed by the action ofalkyl halides on the disodio-addition compound of anthracene 72 and may bedehydrogenated to dialkylanthracenes.73C. F. H. Allen and R. W. MoGibbon, Camdian J. Bee., 1938,16, B, 35.68 J . Amer. Chem. Soc., 1938, 00, 1023.70 Idem, J . Org. Chem., 1939, 4, 583; 1941, 0, 36; G. M. Badger, J. W. Cook, and7 1 J .Amer. Ohem. SOC., 1940, 62,3098.is M. Lerer, Ann. Off. nut. U o d . liq., 1933, 8, 681.93 Compare G. Huge1 land M. Lerer, Bull. SOC. chim., 1933, S3, 1497.Ibid., 1939, 61, 2358.F. Goulden, J., 1940, 16180 ORGANIC CHEMISTRY.One of the most generally useful methods for the synthesis of polycyclichydrocarbons is that whereby a four-carbon chain is attached, in stages, byits terminal carbon atoms, to ortho-positions of an aromatic ring. This leadsto fusion of a new six-membered ring, which may be rendered aromatic bydehydrogenation. The simplest procedure for effecting this is the Friedel-Crafts reaction with succinic anhydride, followed by reduction of thep-aroylpropionic acid to a y-arylbutyric acid and then cyclisation of this.This method was used by R.D. Haworth 74 * for the synthesis of phen-anthrene homologues and by J. W. Cook and C. L. Hewett 75 to establish thestructure of 3 : 4-benzpyrene (IV), the potent carcinogenic hydrocarbonisolated from coal tar l 7 (p. 166). The latter synthesis has been improved inseveral stages by subsequent workers. Reduction of pyrenoylpropionicacid (C) to pyrenylbutyric acid (CI) is suitably carried out by zinc dust andalkali under pre~sure.7~ Cyclisation of (CI) to ketotetrahydrobenzpyene(CII) was the least satisfactory stage of the original synthesis. The bestmethod of effecting this is Bachmann’s modification77 of the method ofF i e ~ e r . ~ ~ This consists in the action of stannic chloride on the chloride of theacid (CI) in cold benzene solution.The pentacyclic ketone is readilyobtained pure in almost theoretical yield and the same procedure has givenexcellent results in the cyclisation of many other arylbutyric acids (Fieser,Bachmann) and also in such cases as the cyclisation of the acid (CIII) to thecholanthrene derivative (CIV).79CH2*C0,H H,&-COReduction of the carbonyl group of (CII) is necessary before dehydro-genation to benzpyrene, if satisfactory yields are to be obtained, and this74 J . , 1932, 1125.76 H. Vollmann et al., Annalen, 1937, 531, 128.7 7 W. E. Bachmann, M. Carmack, and S . R. Safir, J . Amer. Ch.em. SOC., 1941, 63,78 L. F. Fieser and F. C. Novello, ibid., 1940, 02, 1858.78 W. E. Bachmann, J . Org. Chem., 1938, 3, 434.* E. Mosettig and H.M. Duvall (J. Amer. Chern. SOC., 1937, 69, 367) found a con-venient route to 1- and 4-phenanthrol in the palladium dehydrogenation of Haworth’ske t o te t r ah ydr ophsnanthrenea ,76 J., 1933, 398.1684COOK : POLYCYCLIC AROMATIC COMPOUNDS. 181may be effected suitably by high-pressure hydrogenation with copperchromite (which gives tetrahydrobenzpyrene) 80 or by aluminium iso-pr~poxide.~~ The latter method gives the carbinol, which is readilydehydrated to dihydrobenzpyrene. Many other examples have been givenof the reduction of ketones of type (CII) by aluminium isopropoxide.Another excellent reagent for bringing about the cyclisation of acids oftype (CI) is anhydrous hydrogen fluoride, which is commercially available inthe United States and has been extensively used by Fieser and his school,not only for cyclisations but also as a condensing agent in Friedel-Craftsreactions,81 for which, however, the method is not of general application.J. H.Simons, S. Archer, et had already carried out a series of alkylationswith hydrogen fluoride," which was also used by W. S. Calcott, J. M. Tinker,and V. Weinmayr 83 for the synthesis of perylene (XLV) from phenanthreneand acraldehyde, and of 4 : 5-benzpyrene (CXII) from 9 : 10-dihydro-anthracene and acraldehyde.The use of hydrogen fluoride sometimes alters the course of a reaction.Acenaphthene and acetic acid reacted with hydrogen fluoride to give 94% ofa mixture of acetylacenaphthenes, from which 25% was easily isolated as thehitherto relatively inaccessible 1 -acetylacenaphthene (CV) .81 Moreover, inthe cyclisation of y(2-phenanthry1)butyric acid (CVI) L. F. Fieser andW. S. Johnson 84 obtained a 78% yield of the ketotetrahydrobenzanthracene(CVII) with hydrogen fluoride, whereas the ketotetrahydrochrysene (CVIII)was mainly found when the acid (CVI) was cyclised with S5y0 sulphuric acid. l4(CVI.)(CVII. ) (CVIII.)Cyclisation of y- (3-phenanthryl) butyric acid (LV) with hydrogen fluoridedoes not appear to have been described. It would be of interest to know ifthis occurs in position 4 (cf. p. 171).The succinic anhydride condensation (by the Friedel-Crafts or theGrignard reaction) is by no means the only method by which the four-carbon chain may be introduced in syntheses of the type discussed above.An acetyl or other acyl substituent may be brominated, and the productcondensed with malonic ester derivatives ; alternatively the carbonylgroup may be condensed with ethyl succinate, or reduced to carbinol with80 L.F. Fieser and M. Fieser, J . Amer. Chem. Soc., 1936, 67, 782.81 See, e.g., L. F. Fieser and E. B. Hershberg, ibid., 1939, 61, 1272; 1940, 62, 49.83 Ibid., 1938, 60, 986, 2952, 2963, 2965, 2956.83 Ibid., 1939, 81, 949. * J. H. Simons (Ind. Eng. Chem., 1940,32,178) gives a review of the uses of hydrogenfluoride in organic chemistry.Ibid., p. 1647182 ORGANIU CWMIBTRY.aluminium i~opropoxide,~~ then the hydroxyl replaced by chlorine, and thechloro-compound oondemed with mdonic ester derivatives ; the latterprooedure gives a three-carbon chain terminating in carboxyl, and this maybe extended by the Amdt-Eistert reaction.In all these ways suitablysubstituted four-carbon chains have been attached to aromatic molecules,and further substituents may be introduced by the action of Grignardreagents on aroylpropionic acids or the ketotetrahydo-compounds obtainedby cyclis&tion of arylbutyrio acids. Reactions such as these have been usedfor the synthesis of a wide variety of derivatives of phenanthrene, chrysene(CXXXIV), 1 : 2-benzanthracene, triphenylene (LXXVII), and otherpol ycyclio aromatic compounds.The positions of attachment of the new ring may often be modiiied byusing partially hydrogenated aromatic compounds for the initial condens-ations. Thus, although phenanthrene is substituted mainly in position 3 andto a lesser extent in position 2, in Friedel-Crafts reactions carried out innitrobenzene solutions, if 9 : 10-dihydrophwthene (CIX) is used, sub-stitution takes place almost exclu~vely in position 2.86 As dihydrophen-anthrene is readily obtained from phenanthrene by hydrogenation withoopper ~hrornite,~' the scope of this synthetic method becomes very muchenlctrged.In Friedel-Crafts reactions with 1 : 2 : 3 : 4-tetrahydrophen-anthrene (CX) the acyl &;roup enters the 9-positionss and the succinicanhydride oondensation with s-hexahydropyrene (CXI) was used by Cooksynthesis of 4 : 5-benzpyrene (CXII). and-Hewett 78 for the/\ fa(CIX.) \tA new synthesis of 3 : 4-benzpyrene end its derivatives was developed byL.3'. E'ieser and E. B. Her~hberg,*~ who obtained the keto-compound86 Compare H. Lund, Ber., 1937, 70, 1520.A. Burger and E. Mosettig, J. Amer. Chem. SOC., 1935, 57, 2731; 1936, 58, 1857;e 7 Idem, ibicl., 1935, 57, 2731; J. R. Durland and H. Adkins, &id., 1837,59,136;1937,69,1302.L. F. Fieser and W. S. Johnson, ibid., 1939, 61, 169.W. E. Bechmasn and W. S. Struve, J . Org. Chem., 1939, 4,472.J. Amer. Chem. SOC., 1938, 60, 1658COOK : POLYUYCLIU AROMA!l'ItY COMPOUNDS. 183(CXIV) by aluminium chloride-sodium chloride fusion of 3-benzoyl-perinaphthane (CXIII). The latter was the product of the Friedel-Cra,ftsreaction between benzoyl ohloride and perinaphthaae, which was formed bycopper ohrornite hydrogenation of perinaphthenone (XXXV ; R = H).9-Alkylphenanthrenes (CXVI) were obtained by C.K. Bradsher andS. T. Amore by the action of hydrogen bromide in acetic acid on oxides ofthe type (CXV), synthesised from o-iododiphenyl. J. W. Cornforth and( S i r ) R. Robinson 91 obtained 2 : 7-dimethoxy-9 : 10-dihydrophenanfhrene(CXVIII) by heating 6 : 6'-di-iodo-3 : 3'-dimethoxydibenzyl (CXVII) withcopper bronze. This dibenzyl derivative (CXVII) was readily synthesisedfrom m-methoxybenzyl chloride. T. Hasselstrom 92 found a suitable methodfor the preparation of l-methylphenanthrene in quantity, in the thermaldegradation of retene (1-methyl-7-isopropylphenanthrene) by 9 hours'boiling in contact with fuller's earth.OMeOMe(CXVII.) (CXvrU,)The Psohorr phenanthrene synthesis continues to provide a valuablemethod in many instanom, and within reoent years has been used for thepreparation of various homologues and derivatives of phenmthrene 98 andof 4 : 5-methylenedioxyohrysene,8* 1 : 2-dimethylohrysene (XI),96 1 : 2 : 7 : 8-dibenzanthracene,96 3 : 4 : 6 : 6-dibenzphenmthrene (LXVI)?s 1 : 2 : 3 : 4-dibenephenanthrene ( X),Q7 picene?a and ~holanthrene.~~The synthesis of derivatives of 3 : 4-benzphenanthrene has reoeivedattention on amount of the oaroinogenio properties of some members of thisgroup. The most generally useful method is that of C.L. Hewett.1 ThisJ . Arner. Chem. Soc., 1941,63,493; compare {bid., 1939,61,31!31; 1940,62,2806.Q1 J., 1942, 684. @a J . Arner. Chem. SOC., 1941, 63, 1104.83 M.T. Bogert and (3.8. Stamatoff, Rec. Trav. cMm., 1933,6a, 684; J. T. Cassadayand M. T. Bogert, J. Amer. Chem. Soc., 1939, 61, 2461, 3055; R. A. Konovalova, S.Junusov, and A. P. Orekhov, J. Gen. Ghem. rJ.rS.S.R., 1939, 9, 1607; Bull. SOC. chim.,1939, 6, 1479; E. E. Lewis and R. C. Elderfield, J. Org. Chem., 1940, 6, 290; T. M.Sharp, J., 1936,1234; P. Hill and W. F. Short, J., 1937, 260; A. Higginbottom, P. Hill,and W. F. Short, ibid., p. 285; C. L. Hewett and R. H. Martin, J., 1940, 1396.94 L. H. Briggs and J. M. Wilson, J., 1941, 500.94 C. L. Hewett, J., 1940, 293.97 C. L. Hewett, J., 1938, 193.98 H. Waldmann and G. Pitschak, AnnuZen, 1937, 527, 183.99 L. F. Fieaer and a. W. Kilmer, J. Amel.. Chem. SOC., 1940, 63, 1364.e6 J.W. Cook, J . , 1932, 1472.1 J . , 1938, 1286; 1940, 293, 1159184 ORGANIC CHEMISTRY.is based on a patented method for synthesising phenanthrene derivatives.2Elimination of hydrogen bromide from the acid (CXIX) by fusion withpotash leads to 3 : 4-benz-l-phenanthroic acid (CXX) in 60% yield. Themethod lends itself to considerable variation. Other benzphenantlpnesyntheses of less general application have also been described.3 Thebenzretene obtained from retene by the succinic anhydride synthesis byD. E. Adelson and M. T. Bogert4 was originally thought to be a 3 : 4-benzphenanthrene derivative, but was afterwards shown to be a homologueof 1 : 2-ben~anthracene.~(CXIX. ) (CXX . ) (CXXI.)The Diels-Alder diene synthesis continues to be used to build up poly-cyclic structures.* The formation of hydroanthraquinones from conjugateddienes and p-benzoquinone or a-naphthaquinone was among the earliestexamples of this reaction; more recently, H.J. Backer, J. Strating, andL. H. H. Huisman obtained octamethylanthraquinone (CXXI) by aerialoxidation of the adduct of apy8-tetramethylbutadiene and p-benzoquinone.Other fully substituted anthraquinones were similarly prepared, as well as thecorresponding anthracene hydrocarbons. From hexatriene and a-naphtha-quinone, 1 -vinylanthraquinone was prepared. By addition of maleicanhydride to dienes of the type (CXXII), H. A. Weidlichs obtainedanhydrides such as (CXXIII), from which were prepared by decarboxylationand dehydrogenation, 1 : 2 : 5 : 6- and 3 : 4 : 5 : 6-dibenzphenanthrene (LXVI),1 : 12-benzperylene (I) and picene.E. Clar prepared a series of polycycliccompounds by addition of methyleneanthrone (CXXIV) to maleic anhydride,cinnamic acid and various 1 : 4-quinones. A new type of diene synthesiswas described by K. Alder and H. F. Rickert,lo who found that under forcedconditions, and in the presence of an inhibitor of polymerisation, styrene willcombine with butadiene and other conjugated dienes. Similarly, indeneI.G., Brit. Pat. 469,633.See, e.g., C. L. Hewett, J., 1936, 596; M. S. Newman and L. M. Joshel, J . Amer.Chem. Soc., 1938, 60, 485; 1940, 62, 972; W. E. Bachmann and R. 0. Edgerton, ibid.,p. 2970; N. C. Ganguly, Science and Culture, 1941, 7, 320.4 J . Arner. Chem. SOC., 1937, 59, 1776.6 L.F. Fieser and R. C. Clapp, ibid., 1941, 63, 319.6 Rec. Trav. chim., 1939, 58, 761.7 L. W. Butz, E. W. J. Butz, and A. M. Gaddis, J. Org. Chem., 1940, 5, 171.* Ber., 1938, 71, 1203.Ber., 1936, 69, 1686; see also C. F. H. Allen et al., J. Amer. Chern. SOC., 1940, 62,656.lo Ber., 1938, 71, 379.* For a general review of the Diels-Alder diene synthesis, see J. A. Norton, Chem.Reviews, 1942, 31, 319COOK POLYCYCLIC AROMATIC COMPOUNDS. 185(CXXV) and butadiene combine to give a tetrahydrofluorene (CXXVI), fromwhich fluorene may be obtained by dehydrogenation with selenium, Indenehas also been condensed with methyleneanthrone (CXXIV) to give a productwhich was dehydrogenated to the meso-benzanthrone derivative (CXXVII) .ll(CXXII.) (CXXIII.)/\-/\(CXXVI. )riH2I II It I\A /\///\A/\co(CXXIV. )/--\I--A/\/\ I II I II\+II f\/\p(CXXVII.) /\Fluorene derivatives have also been prepared from a-hydrindone by theMannich reaction l2 and the structural relationships of the dibenzfluoreneshave been clarified by the synthetic work of R. H. Martin l3 and G. Swain andA. R. T0dd.1~ The chemistry of fluorene and its derivatives has beenreviewed by G. Rieveschl and F. E. Ray.15 New methods of passage fromthe fluorene series to the phenanthrene series have been described. R. F.Schultz, E. D. Schultz, and J. Cochran l6 obtained 9-phenanthrol and (princi-pally) its methyl ether by ring-enlargement of fluorenone with diazomethane,and W. G. Brown and B. Bluesteinl’ obtained phenanthrene in almostquantitative yield by dehydration of 9-fluorenylcarbinol.F. G. Baddar 18converted 3 : 4-benzfluorenone (CXXVIII) into meso-benzanthrone (CXXIX)by fusion with aluminium chloride-sodium ohloride.New syntheses of chysene derivatives have been described.l9l1 G. Swain and A. R. Todd, J., 1942, 626.l2 R. H. Harradence and F. Lions, J . Proc. Roy. SOC. N.S.W., 1939, 72, 284.l3 J . , 1941, 679.l6 Chem. Reviews, 1938, 23, 287.l7 Ibid., p. 3266.lB M. S. Newman, J . Amer. Chem. SOC., 1938, 60,2947; 1940, 62, 870, 2295; C. K.Bradsher and A. S. Burhans, ibid., 1940, 82, 3140; L. M. Joshel, L. W. Butz, and J.Feldman, ibid., 1941, 88, 3348; L. W. Butz and L. M. Joshel, ibid., 1942, 64, 1311.l4 Ibid., p. 674.l6 J .A m r . Chem. SOC., 2940, 62, 2902.1* J . , 1941, 310186 ORGANIC CHEMISTRY.Reactiom.This section will be devoted ohiefly to the reactions of the anthraoenegroup, which show many features of unusual interest. Reactions of otherclasses of polycyclic oornpounds have been extensively studied, but are forthe most part the normal reactions of aromatic compounds. Their interestlies partly in the products which may be obtained thereby, and in thepositions at which substitution occurs.Substituenta may be introduced directly into four of the five positions ofphenanthrene by suitable procedures. The 1-, 2-, 3- and 9-aldehydes haveall been obtained from the carboxylic acids.20 The oximes of 1-, 2-, 3- and9 -ace t yl p henanthrenes give almost 8x clusively the acet amido - compoundsby the Beckmann rearrangement ; the oximes of the benzoyl-phenanthrenesgive in addition anilides of the corresponding acids (these are the chiefproducts with the 1- and the 9-isomeride).Bachmann and 130atner,20 whocarried out these transformations, diazotised the 1-, 2- and 3-minophen-anthrenes by the procedure of C. de Milt and G . van Zandt21 for thediazotisation of weakly basic and insoluble amines. They prepared by thismeans the 1-, 2- and 3-chloro-, -bromo- and -iodo-phenanthrenes. Thekinetics of the addition of bromine to phenanthrene have been studied byC. C. Price.22 A. Jeanes asd R. Adamss have shown that alkali metalsadd on to phenmthrene only at the 9 : 10 which are ah0 thepositions attacked in the oxidation of phenmthrene by osmium t e t r ~ x i d e .~ ~A cumprehensive survey of the chemistry of pyrene (XXVIII) by H.Vollmam, H. Beeker, M. Corell, and I-f. Streeck 64@-lS0) has elucidated thepositions of substitution in numerous reactions and has disclosed methodsfor the introduction of substituents into the positions (1, 1 : 2, 1 : 6) notdirectly attacked. Oxidation of pyrene wifh chromic acid gave it mixture of3 : 8- and 3 : 10-quinones (CXXX and CXXXI), which were separated, andtheir structures determined. Oxidation of 3 : 4-benzpyreae lik&ise gave(CXXX.) (CXXXI . ) (CXXXII.) (CXXXIII . )amixture of quinones (CXXXII and CXXXIII), which Vollmann separatedby crystallisation of the products of reductive acetylation. Other reactions555; W.E. Brtchmann and C. H. Boatner, &id., 1936, 58,2097.2o C. W. Shoppee, J., 1933, 37; W. E. Bmhrnmn, J. Amer. Chem. Em., 1936, 57,2 1 Ibid., 1936, 58, 2044.22 Ibid., 1936, 58, 1834, 2101; compare L. F. Fieser and C. C. Price, ibid., p. 1838,and M. S. Kharasch, P. C. White, and F. R. Mayo, J . Or3. Chem., 1938,2,574.z3 J . Amer. Chern. Soc., 1937, 59,2608.24 Compare W. Huckel: and H. Bmtschneider, Annulen, 1939,540, 157.2 6 R. Criegee, €3. Marchaad, and H. WannoWius, Anden, 1942, 550, 99COOK : POLYOYawO AROMATIC OOMPODNDS. 187with 3 : 4-benzpyrene led to monosubstitution at position 6 in all cases(nitration, chlorination, oxidation with lead tetra-acetate, diazo-coupling,condensation with N-rnethylf'ormanilide) exmpt in the Friedel-Craftsreaction with acetyl chloride, which gave mainly 10-acetyl-3 : 4-benzpyrene.26Progress has also been madein the orientation of chrysene derivatives.In monosubstitution the group normally enters the 2-position (CXXXW).This is the case with chlorination and bromination, nitration, sulphonation,and in the Friedel-Crafts reaotion with rtcetyl and benzoyl chlorides.Thevariow derivatives have been correlated with one another, and with synthetic2-chrysenol of proved ~tructure.~' A second product of the E'riedel-Craftereaction with acetyl chloride was shown not to be the l-acetyl compound,aa it was reduced to an ethylchrysene which differed from synthetic 1-ethyl~hrysene.~' Diaubstitution of chrysene is generally assumed to givesymmetrical (2 : 8-) derivatives, but this does not appear to have beenrigidly proved.2* The increasing availability of chrysene, pyrene andfluoranthene (111) has been responsible for a steady increase in the number o f- .(CXXXIV.) (CXXXV. ) (CXXXVI. )patent specifications dealing with substitution produots of these hydro-carbons. R. R. Pritchard and J. I;. Simonsen *9 showed that aulphonation ofmmo-benzanthrone (XXXI) occurs mainly in position 6, although othersubstituents chiefly enter the 1'-position. Sulphonation of 1 : 2-benzanthra-quinone gives the 4'-sulphonic acid (CXXXV), the structure of which wasshown by alkaline fusion to 5-hydroxy-2-naphthoic acid.30 The oxidation ofpyrene and 1 : 2-benzanthracene to quinones by hydrogen peroxide has beenrecorded,31 whereas fluorenone was oxidised by hydrogen peroxide in etherto a peroxide to which formula (CXXXVI) is ascribed.32 Fluorenone reactswith formamide to give the formyl derivative of 9-aminofluorene.=Probably the most outstanding reaction which has been found in thea* A.Windaus and S. Rennhak, 2. phyaiol. Chem., 1937, 249, 256; A. Windausand H. Xaiohle, AnnuZen, 1939, 537, 161 ; I;. F. Fieser and E. B. Hershberg, J. Amer.Chem. Soc., 1938,60,2542; 1939, 61, 1665; H. J. Eckliardt, Ber., 1940, 73,15.K. Funke and E. Muller, J . pr. Chem., 1936, 144, 242; M. S. Newman and J. A.Cathcart, J. Org. Chem., 1940, 5, 618.2 8 See K. Funke, E. Muller, and L. Vadssz, J. p r . Chern., 1936,144,265; K. Funkeand J. Ristio, ibid., 146, 309; 148, 161,ZB J., 1938, 2047.30 A.Sernpronj, Gazzetta, 1939, 69,448; J. Cason and L, F. Fieser, J . Amer. Chem.SOD., 1940, 62, 2681.31 R. T. m o l d and R. Larson, J , Org. Chem., 1940, 6,260.39 G. Wittig and G. Pieper, Ber., 1940, 78, 295.33 B. Schiedt, J. pr. Chm., 1941,167,203188 ORQANIO CHEMISTRY.anthracene group is the photo-chemical oxidation to transannular peroxides.This appears to be a general reaction of anthracene derivatives, and is notshown by any other group of polycyclic aromatic compounds. The recogni-tion of this resulted from the elucidation of the naphthacene structure of therubrenes, which were well known to haw the capacity to form photo-oxides ( p a 166). The formation and properties of these " photo-oxides " havebeen reviewed by C.Dufraisse 3 and also by W. Bergmann and M. McLean 35in a section of a general article on transannular peroxides.Anthracene itself undergoes photo-oxidation, a fact which eluded dis-covery until 1935.36>38 This is remarkable in view of the large number ofinvestigations which had been made on the photo-polymerisation of anthra-cene. The failure to observe the phenomenon earlier is due in part to theinstability of this photo-oxide (CXXXVII), which is completely destroyed byprolonged irradiation, and in part to the circumstances that the photo-polymerisation has been studied mostly in benzene solution, which favourspolymerisation but not oxidation. Carbon disulphide is the solvent mostfavourable for photo-~xidation.~~ Solutions of anthracene in carbondisulphide are non-fluorescent, and loss of fluorescence is concomitant with anincrease in the velocity of photo-oxidation. The photo-polymeride ofanthracene is not an intermediate in the photo-oxidation, although there aregrounds for the belief that partial photo-dimerisation may be involved.Atrimeric formula has been suggested for the photo-polymeride of anthra~ene,~8the formation of which has been studied from the magnetic standp0int.3~Anthracene photo-oxide (CXXXVII) liberates iodine from potassium iodide,and is transformed by hydrogen chloride into chloroanthrone, from which themore stable methoxyanthrone (CXXXVIII) was prepared for identification.38(CXXXVII.) (cxxxVII1.) (CXXXIX. )A feature of the rubrene photo-oxides is their ability to undergo thermaldissociation with liberation of oxygen.Similar behaviour is shown by thephoto-oxides of 9 : 10-diarylanthracenes, which in favourable cases mayliberate up to 95% of their combined oxygen.40 10-Aryl-9-alkylanthracenephoto-oxides liberated less than 50% of their oxygen on thermal decomposi-34 Bull. SOC. chim., 1939, 8, 422.36 Chem. Reviews, 1941, 28, 367.36 C. Dufraisse and M. Gerard, Compt. rend., 1936, 201, 428; 1936, 202, 1859.37 C. Dufraisse and M. Badoche, ibid., 1935, 200, 1103.36 C. Dufraisse and M. Gerard, Bull. SOC. chim., 1937, 4, 2062.3Q S. S. Bhatnagar, P. L. Kapur, and G. Kaur, Proc. Indian Acad. Sci., 1939, 10, A ,40 C. Dufraisse and A. Compt. rend., 1936, 201, 280; C. Dufraisse and J. le468; compare J.Farquarson and M. V. C. Swtri, Current Science, 1940, 9, 135.Bras, Bull. SOC. chim., 1937, 4, 349; D. Duveen and A. Willemart, J., 1939, 116COOK : POLYCYCLIC AROMATIC COMPOUNDS. 189tion,41 whereas 9-alkyl- and 9 : 10-dialkyl-anthracene photo-oxides did notliberate oxygen.42meso-Substituents favour the formation of photo-oxides, and althoughthe experiments of Dufraisse and his collaborators were usually carried out insunlight, they state that, if the solvent is favourable, the ordinary illuminationof the laboratory is sufficient. J. W. Cook and R. H. Martin 43 carried outthe photo-oxidation of a number of meso-substituted 1 : 2-benzanthraceneswith the aid of an ordinary gas-filled electric lamp. The ease with which9 : 10-dimethyl-1 : 2-benzanthracene is photo-oxidised is shown by the factthat Sandin and Fieser 71 (P- 17’) isolated its photo-oxide when they attemptedto purify the hydrocarbon by chromatographic means.9 : 10-Dimethoxy-anthracene was the most easily photo-oxidised substance studied byDufraisse; 44 its solutions are oxidised exceedingly rapidly when exposed tolight, but are unaffected in the dark. The sodium salt of anthraquinol isoxidised in the dark, but gives no peroxide. The photo-oxides differconsiderably in stability, and the compound from 1 : 4-dimethoxy-9 : 10-diphenylanthracene liberates oxygen even in the cold.45 The influence ofmethoxy-groups in various positions was studied by C. Dufraisse andL. Vell~z.~6The ready elimination of oxygen from these photo-oxides would suggestthat they may be molecular compounds.The heat of formation is too greatfor this, however, although the heat of formation of the photo-oxide ofrubrene (L), 22.6 kg.-cals., is small in comparison with that associated withthe formation of a stable oxide.47 The structure of these compounds asmeso-peroxides is fairly conclusively established by their hydrogenation toderivatives of 9 : 10-dihydroxy-9 : 10-dihydroanthracene (CXXXIX).48s 83The mechanism of photo-oxidation was discussed by Dufraisse,a whorejects the view that the chief function of light is to induce dissociation to afree radical. If this were so, all meso-additions to anthracene derivativeswould be activated by light, which is not the case. C. K. Ingold and P.G.Marshall 49 observed that 9 : 10-diarylanthracenes undergo reversibledeepening in colour when heated* and attributed this to di-radicalformation. C. Dufraisse and J. Houpillart disputed this interpretationand showed that the increase in colour is due to the well-known effect ofincrease in absorption power with rise in temperature. All attempts tobring about the formation of photo-peroxides in the phenanthrene, naphtha-4 1 A. Willemart, Cornpt. rend., 1936, 203, 1372; Bull. SOC. chim., 1939, 6, 204.42 A. Willemart, Cornpt. rend., 1937, 205, 866; Bull. SOC. chirn., 1938, 5 , 556.43 J., 1940, 1126.44 C. Dufraisse and R. Priou, Bull. SOC. chirn., 1939, 6, 1649.4 s C. Dufraisse, L. Velluz, and L. Vellux, C m p t . rend., 1939, 208, 1822; 209, 616.4 6 Ibid., 1941, 212, 270.4 7 C.Dufraisse and L. Enderlin, ibid., 1930, 191, 1321.4 8 C. Dufraisse and J. Houpillart, ibid., 1937, 205, 740.49 J., 1926, 3080.* Some 9 : 10-dialkylrtnthracenes behave in the same way.-Reporter.BuU. SOC. chirn., 1938, 5, 1628190 OMANIU CIHBMfSTltY.lene and aoridine series have been unsucces~fu1,51 and the phenomenonappears specific t o the anthracene series. A. von Rebay and H. Fettback 52obtained a peroxide of cymene by the direct action of oxygen, but this is anacidic peroxide, and therefore is not analogous to the anthrecene photo-oxides.A reaotion of anthracenes whioh has many points of resemblanoe la photo-oxidation is the DieleAlder reaotion, in which anthraoene derivativesparticipate a~ diene components, undergoing transannular addition of re-agents such m maleic anhydride and its derivativerj,53 p-benzoquinone,"acetylenedic&rboxylic ester,65 diazoaoetio ester,66 a~raldehyde,~~ and even,at somewhat high temperature& of ally1 chloride 68 and vinylThe reaction with mdeic anhydride has been most widely studied. Theaddition compounds (type CXL) undergo dirrsociation at high temperatures,and W.E, Beohmann and M. C. Kloetzel6O demonatrratied conclusively thatthe addition of maleic anhydride is an equilibrium reactiop. In the caam ofanthraoene and its 9-methyl and 9 : lo-dimethyl derivatives the equilibriumwas very strongly in favour of the adduct (98-99% formed) ; in other casescomparatively little combination took place when molecular proportionswere used, but satisfactory yields of adduct were formed by using 30 mole-cular proportions oE maleic anhydride.In this way the yields of productwere increased t~ follows : 9 : 10-diphenylanthracene (16% to 78%) ;1 : 2 : 6 : 0-dibenzanthracene (30% to 91%) ; 20-methyloholanthrene (22%to 83%). The rates of reaotion varied greatly; 9 : 10-dimethylanthraoenereadad rapidly at room temperature, whereas with 9 : 10-diphenylanthracenethe reaction wm incomplete after d a p of boiling equimoleoular proportionsof the reactants in benzene solution. For the sterio effects shown by some ofthese addition compounds, me p. 169.The high chemical reactivity of many anthracene hydrocarbons has beendemonstrated by several substitution reactions investigated in detail by61 C.Dufraisse and R. Priou, BulZ. SOC. chim., 1938, 5, 611; C. Dufraisse andJ . Houpillwt, aid., p. 626.62 Ber., 1939, 72, 1643.b3 0. Diels and K. Alder, Annalen, 1931, 480, 191 ; E. Clar, Ber., 1931, 04, 2104.64 E. Clar, Ber., 1931, 64, 1676.b6 0. Diels and K. Alder, Eoc. cit.; 0. Diels and W. Friedriohsen, AnnccEen, 1934,5 0 0. Diels, S. Schmidt, and W. Witte, Ber., 1938,71, 1186.5 7 A. G. Slobodskoi and V. I. KhmelevskiI, J. Qen. Chem. U.S.S. R., 1940, 10, 1 199.K. Alder and E. Windemuth, Ber., 1938, 71, 1942.5y I(. Alder and H. F. Rickert, Annalen, 1939, 543, 1.513, 145; 0. Dielsand W. E. Thiele, Ber., 1938, 71, 1173.60 J . AWM. chsm. SOC., 193a,w, 481STEVENS : HETEROCYOLTC OOMPOUNDS. 191L. F.Fieser and his collaborators. The most striking of these is diazo-coupling, which occurs very readily in certain cases.61 Both 3 : 4-benxpyrene(IV) and 20-methylcholanthrene (VI) gave deep red colours with p-nitro-benzenediazonium chloride, and in the case of benzpyrene the crystallineazo-compound (CXLI) was isolated and its orientation determined.62 Thealdehyde group was directly introduced into the 9- and the 10-position ofanthracene and 1 : 2-benzanthracene respectively, by means of N-methyl-formanilide and phosphorus oxychloride. This reaction is not specific toanthracene derivatives, for it is shown also by pyrene 64 (*- w) and, slowly,by acenaphthene.83 3 : 4-Benapyrene, 20-methylcholanthrene and oertainother anthracene hydrocarbons undergo direct thiooyanation,a and they arer e d y oxidised by lead tetra-acetate.In the cam of benzpyrene the 5-acetate (CXLII) is formed in high yield, whereas 20-methylcholanthrene isoxidised in the five-rnembered ring, giving a mixture of the 15-acetate(CXLIII) and the corresponding 16-keto-c0mpound.7~ @.le6) With 10-methyl-1 : 2-benzanthracene (CXLN) oxidation takes place in the methyl group in(CXLII.)~H,-~H-OAC(CXLIII.) (CXLIV.)spite of the presence of an unsubstituted (but hindered) meso-position. Thereaction of chloromethylation, while by no means specific to the anthracenegroup, takes place very easily with 1 : 2-benzanthracene 66 and gives the10-chloromethyl derivative.Despite the length of this Report some of the topics have been treatedvery superficially, and it has been necessary to omit reference to muchimportant work which come6 within the scope and period of this review.J.W. C.6. HETEROOYCLIC COMPOUNDS.Oxygen Bing Cmpoud8.Flavan Derivatives.-The Simonis synthesis of flavanonea from polyhydricphenols and cinnamoyl chloride has been improved ; and with cold alcoholic61 L. F. Fieser and W. I>. Campbell, J . Amer. Ohem. SOC., 1938, 60, 1142.62 L. F. Fieser and E. €3. Hershberg, ibid., 1939, 61, 1665.L. F. Fieser and J. E. Jonea, ibid., 1942, 64,1666.J. L. Wood and L. F. Fieaer, ibicl., 1941,63, 2333.6 5 G. M. Badger esd J. W. Cook, J., 1989, 802.S. Huzise and H, Tatsita, Ber., 1941, 74, 276192 ORGANIC CHEMISTRY.sodium ethoxide the initial isomerisation proceeds quantitatively 2 in theflavone synthesis :/COMe4(=IO*COPhHBr --+co,,( )Ph0\/ \R.L. Shriner and R. B. Moffett3 find that the same flavylium salt (I)results from the two syntheses shown, and in the ozonolysis of a related casehave isolated products corresponding to both positions of the olefinic bond.They regard such substances as carbenium salts with the ionic charge on C,or C,, and the independent existence of the two types as possible.CHR, y \COR,/\ /\0 0 +0,SalicylaldehydeAlpinetin, from AZpinia chinensis, has been identified as 5-hydroxy-7-methoxyflavanone.4 The colouring matter of Butea frondosa is shown bymethylation and hydrolysis to be the 3'-7-diglucoside of the flavanone butin,unusual in having a sugar residue in the unfused benzene ring.5 AmpelopsisnzeZicefoZia affords ampelopsin (11), one of the few natural flavanolones; itsderivatives can be dehydrogenated to those of myricetin ; and alkali convertshexamethylampelopsin into the chalkone (111), identified by synthesis .6The simple 5-hydroxyflavone occurs in the primrose ; and Z.Horii 8 hsasynthesised nobiletin (IV), a pentahydroxybenzene derivative :Me0/\/coM:2s,o,;MeO(i\,, --+ ~e,so,Me02 V. V. Ullal, R. C. Shah, and T. S. Wheeler, J., 1940, 1499.J . Amer. Chern. Soc., 1940, 62, 2711; 1941, 63, 1694.Y. Kimura, J . PharmSoc. Japan, 1940, 60, 87.5 P. S. Rao and T. R. Seshadri, Proc. Indian Amd. Sci., 1941, A , 14, 29.6 M. Kotake and T. Kubota, Annalen, 1940, 544, 253.P. Karrer and G.Schwab, Helv. Chim. Acta, 1941, 24, 297.J . Pharm. SOC. Japan, 1940,60, 246STEVENS : HETEROCYCLIO COMPOUNDS. 193The Linaria vulgaris glycosides linarin and pectolinarin have been synthesisedfrom the aglycones and acetobromorutinose ; a third glycoside, neolinarin,is probably a crystalline form of the amorphous pe~tolinarin.~The fruits of Sophoru juponicu 10 contain sophoricoside, identified bymethylation and hydrolysis as genistein 4’-glucoside, and sophorabioside,formulated as genisteinsglucose-lrhamnose. As well as these isoflavonesthere has been isolated sophoraflavonoloside, which appears to be cam-pherol=glucos&lglucose. Campherol also occurs, as its 7-diglucosideequisetrin, in Equisetum arveme.11In Persicaria hydropiper isorhamnetin is present as the potassium salt ofits 3-sulphuric ester (persicarin) .I2Hibiscus cannabinus flowers afford cannabiscitrin, a glucoside of theflavonol cannabiscetin, which, being oxidised to a quinone by p-benzoquinone,has hydroxyl groups in positions 5 and 8.Since the completely methylatedglucoside yields on degradation 3 : 4-dimethylgallic acid, cannabiscitrin is3 : 5 : 8 : 3‘ : 4’-pentahydroxy-5’-glucosidoxyflavone, with the sugar group inthe benzene moiety.13 Quercetagetin occurs in Tagetes erecta as the mono-glucoside quercetagitrin, l4 which does not behave as a 3-glucoside. Methyl-ation and hydrolysis give pentamethylquercetagetin, which is oxidised toveratric acid but distinct from 5-hydroxy-3 : 6 : 7 : 3’ : 4’-pentamethoxy-flavone, and in which the 7-position, o r t h to the vacant position 8, for the freehydroxyl group is indicated by the smooth Claisen rearrangement of its ally1ether.The furoflavone karanjin has been synthesised from karanjic acid via themethoxy methyl ketone.l5The fruits of the Osage orange, Maclura pomifera, yield two pigments,osajin and pomiferin,ls containing respectively two and three hydroxylgroups, one of which is methylated with difhulty. Tetrahydro-osajindimethyl ether is readily converted by alkali into the ketone (V) withpotassium formate, and then into homoanisic acid; it can also be degradedto (VI) together with anisic acid, while tetrahydropomiferin trimethyl etherby the same process gives (VI) with veratric acid. The pigments are thusisoflavones (VII) evidently with an attached terpene skeleton containingone ring and two double bonds; these are unconjugated, one linking an@ G.Zemplhn and R. Bognar, Ber., 1941, 74, 1818; G. Zemplhn, R. Bognar, andlo G. Zemplh and R. BognOr, ibid., p. 482 ; J. Rabate and J. Dussy, Compt. rend.,l1 H. Nakamura and G. Hukuti, J. Pharm. SOC. Japan, 1940, 60, 179.l2 R. Kawaguchi and K. W. Kim, ibid., 1937,57, 180; 1940,60, 174.l3 P. S. Rao, T. R. Seshadri, and K. Neelakantam, Proc. Indian Awd. Sci., 1941, A ,l4 Idem, ibid., p. 289.l5 T. R. Seshadri and V. Venkateswarlu, ibid., 1941, A, 13, 404; cf. Ann. Reports,l8 M. L. Wolfrom et al., J . Amer. Chern. Soc., 1938, 60, 674; 1939, 61, 2832; 1940,L. Mester, Ber., 1942, 75, 489.1936, 202, 1117; 1937, 205, 1431; J. Rabat6, Bull.SOC. chirn., 1940, 7 , 665.14, 105; P. S. Rao and T. R. Seshadri, ibid., p. 265.1939, 36, 316.62, 651, 1484; 1941,63, 422, 1248, 1263, 3366; 1942, 64, 308, 311impropylidene group, and one so placed aa to interact with a hydroxyl groupunder the influence of sulphuric acid.Polycyclic Chromans.-The reduction and other transformations ofelliptone have been studied, and its constitution confirmed through thesynthesis of dehydrotetrahydroelliptone (I) by the method previously usedfor dehydrotetrahydrosurn~tr01.l~ Derris malaccensia resin contains malac-~ 0 1 , ~ ~ s and a new phenol, possibly (11), which yields formic acid on alkalinehydrolysis, and resembles in this respect and in ite colour reactions koflavonessyntheaised €kom derritol methyl ether and related deo~ybenzoins.1~~Malaccol closely resembles aumatrol, and is believed to be 15-hydroxyelliptoneMe0 coA new synthesis of the brazilin skeleton has been devised : 21Ph*CH, KCN ; NHCO\CO -+PhO CH ,/EOH Ph'CH2\C/NH2 --++ / \1' S.H. Harper, J., 1942, 587, 598; cf. T. 8. Kenny, A. Robertson, and S. W.18 T. M. Meijer rand D. R. KooIhaats, Bee. Trav. chim., 1939, 58, 207.1Q S. H. Harper, J., 1940, 309, 1178. 2o Idem, J., 1942, 595.a1 P. R d e r and H. Epler, Annakn, 1940, W, 263; P. PfeBer and R. Simons, J .George, J., 1939, 1601.pr, Chem., 1942, 160, 83STEVENS : EB!CEROOYClIXO OO?KPOuNDS. 196The acid (IV, C02H for Ph), prepctred similarly, ww converted by nitrous midinto the corresponding hydroxy-compound, dl-brazilic acid, with rammisationif the optically active amino-acid were used.=Thiophen Xeries.-w.Steinkopf and M. Boetius 23 have prepared tetra-deuterothiophen from tetrakischloromercurithiophen and deuterium chloride.The three possible selenophthens have been isolated from the products ofinteraction of acetylene and selenium, and oriented by determining theirelectric moments : */ NP'O < 11 > p = 1.52 D. / I b e p = 1.07D. \\ I1 Se/ \\ I-/H. J. Backer and W. Stevens25 have improved the &berg synthesis offurans and thiophens and extended the method to selenophens :Se- Se-So se-= I1R-R NaOMe RC0.CO.R + + Me0,dBy heating 2-iodothiophen with copper, and by other related methods, aseries of a-polythienyls was obtained, which were shown to be less saturatedthan the polyphenyls or polypyridyls by their increasing visual colour fromterthienyl upwards and by their halochrmy in sdphuric acid.26The results of a detailed X-ray analysis of p-isoprene sulphone are notreconcilable with the accepted formula (I), and resonance with structuressuch as (11) is invoked to account for the equality in length of the C-C bondsin thebO,Me (X = 0, S, Se) \/MeO,C*CH,~X~CH,*CO,Me X" CMe=CHso,-Nitrogen Ring C'mpounds.Pyrrole Group.-Infia-red and Raman spectra are recorded for deutero-pyrroles, effectively prepared as follows : 28Pyrrolidine is prepared by treating N-chloro-N-acyl-lz-butylamines withsulphuric acid ; 29 and formo-o-t;oluidide, with potassium tert.-butoxide a tne P.Pfeiffer and E. Heinrich, J. pr. Chem., 1940,156, 261.ar B. Ttu.namusbi, H. Akiyama, and S. Umezawa, Bull. Chem. SOC. Japan, 1939, 14,85 Rec. Trav. c h h . , 1940, 59, 423, 899.B6 W. Steinkopf, R. Leitsmann, and K. H. Hofmann, Antzabn, 1941,546, 180.28 F. A. Miller, J . Amr. Chem. Soc., 1942, 64, 1643; R. C. Lord and F. A. Miller,2B G. H. Coleman, C. C. Schulze, and H. A. Hoppens, Proc. Iowa Amd. Sci., 1940,47,Annalen, 1941, 646, 208.318; S. Umezawa, ibid., p. 363.E. G. Cox and G. A. Jeffrey, Trans. Paraday SOC., 1942,8$, 241.J . Chem. Physice, 1942,10, 328.264; G. H. Coleman and G. Alliger, aid.? 1941, 48,246.REP.-VOL. XXXIX. 196 ORGANIC CHEMISTRY.350-360", yields 46% of ind01e.~O The carbazole synthesis by heating1 -arylbenztriazoles, now extended to nitro- 31 and aza- 32 derivatives, has alsobeen effected across the peri-positions of naphthalene giving4 : 5-benzacridans :H.Adkins and H. L. Coonradt find that nickel is an unsuitable catalystfor the selective reduction of a pyrrole ring attached to or fused with abenzene nucleus, unless that ring is made more sensitive by N-carbethoxyl-ation. With copper chromite the heterocyclic ring only is attacked, exceptin the case of 2-phenylindole, which yields cyclbhexyl-indole and -indoline.By a method familiar in the pyrrole series, primary and secondary, but nottertiary, radicals are introduced into the 3-position of indole by the sodiumalkoxide at 210-220°.35 W. Borsche and H. Groth have acetylated manyindoles in the 2- or 3-position with acetyl chloride or acetic anhydride, and afew 2 : 3-disubstituted indoles in the benzene nucleus by the Friedel-Crafts method.36Oxindoles have been prepared in good yields by the reactions : 37Pyrolysis of 3-diazoacetyl-2-aminopyridine (from diazomethane and nico-tinoyl chloride) affords 7-azaoxindole : 38 the related isatin and indigo aredescribed.Indophenines and Indigoid Colouring Matters.-G.Heller's formula (I)30 F. T. Tyson, J. Amer. Chem. SOC., 1941, 68, 2024.31 R. W. G. Preston, S. H. Tucker, and J. M. L. CBmeron, J., 1942, 500.a2 E. Spiith and K. Eiter, Ber., 1940, 73, 719; E. Koenigs and P.-L. Nantkrt, Ber.,1941, 74, 215.33 H. Waldmann and S. Back, Annalen, 1940, 645, 62; H. Waldmann and K.-G.Hindenburg, J.pr. Chem., 1940, 156, 157.34 J . Amm. Chem. SOC., 1941, 83, 1563.35 (Mrs.) R. H. Cornforth and (Sir) R. Robinson, J., 1942, 680.36 Annalen, 1941, 649, 238.37 G. Hahn and H. J. Schulz, Ber., 1939,72,1308; G. Hahn and M. R. Tulus, Be?.,K. Miescher and H. Kagi, Helv. Chirn. Acta, 1941,24,1471; H. Kiigi, ibid., p. 141E.1941, 74,600STEVENS : HETEROCYCLIC COMPOUNDS. 197for indophenine is confirmed : isatin and a-thienylmagnesium bromide affordp-(a-thienyl)dioxindole, which can be dehydrated to indophenine ; and therelated pigments prepared from thiophen with benzil, or mesoxalic orbenzoylformic esters, are degraded by alkali to derivatives of cccc’-dithienyl.39P. Pratesi,40 finding that pyrrole derivatives give the indopheninereaction if the N - and one a-position are free, proposed the structure (11) forthe pigments.W. Steinkopf and H. Wilhelm 41 have prepared severalpyrrole-blues in the same way as indophenine, from pyrryldioxindoles, andassign to them an analogous constitution, supported by molecular weightdeterminations. The production of a pigment from 2 : 4-dimethyl-3-ethylpyrrole is explained by displacement of the 2-methyl group to thenitrogen atom.J. Harley-Mason and F. G. Mann 42 have studied the course of an extemiveseries of condensations of thionaphthenquinones with thioindoxyls :(111.) w.1The results were curious : 4-substituted quinones always gave dyes of type(111) ; 7-derivatives gave different types with different thioindoxyls, butnever mixtures; and all other quinones gave type (IV) products.Indoxyl,however, condensed with the p- and oxindole with the a-carbonyl group of allquinones.Phthalonitrile and its analogues are converted by ammonium sulphideinto dithio-p-isoindigos (V), the reactions of which have been studied indetail. By hydrolysis of the S-dimethyl derivative, (V) yields p-isoindigo(VI), which is also obtained by heating monothiaphthalimide with copper.43NH co- (VIII.) 039 W. Steinkopf and W. Hanske, Annalen, 1939, 541, 238.40 Ibid., 1933, 504, 258.41 Ber., 1937, 70, 2233; Awmlen, 1941, 546, 211.43 H. D. K. Drew and D. B. Kelly, J., 1941, 625, 630, 637; J. C. Porter, (Sir) R.u J., 1942, 404.Robinson, and M. Wyler, &id., p. 620J. van Alphen relrsfes depth of colour in indogenides to extensive resonancewith o-quinonoid structures, as in the violet base (VII).That participation ofsimilar o-quinonoid structures is of importance for the colour of indigotin issuggested by the merely yellow colour of its isomeride (VIII), obtained byfusing 2-methylbenziminazole with phthalic anhydride.uTryptophan, Derivatives.-The Adamkiewicz-Hopkins reaction dependson condensation with an aldehyde (glyoxylic acid), yielding a tetrahydro-cmboline, which is oxidised to the colouring matter by impurities in thesulphuric acid; the test, in the absence of aldehyde, is diagnostic45 fortetrahydrocarboline-y-carboxylic acids (I, from tryptophan and RCHO).CH,One of the poisonous constituents of the fungus Amanita phalloida isthe crystalline polypeptide phalloidine,46 which contains no free amino- orhydroxyl groups and is formulated: 2 alanine + cysteine + 2 Z-hydroxy-proline (b) + hydroxytryptophan - 68,O.The hydroxyproline is dia-stereoisomeric with the commonly occurring Z-hydroxyproline (a). Thestruoture (11) for hydroxytryptophan is supported by the developmentof a diazotisable amino-group on alkaline hydrolysis, and the negativeAdamkiewicz-Hopkins and positive oxindole colour reactions.A Q=CH--YH*CO2R A CO*CH,*YH*CO,H W.)(/\C0213 NH, or m2NH2 NH,Kynurenine, an intermediate stage in the metabolic transformation oftryptophan into kynurenic (4-hydroxyquinoline-2-carboxylic) acid, has beenregarded as (111), but A. Butenandt, W. Weidel, and W.von Derjugin now 47conclude that it is a monobasic acid (IV) more obviously transformable intokynurenic acid, and have synthesised the racemic substance from o-bromo-o-nitroacetophenone and ethyl phthalimidomalonate. Bacteria producefrom tryptophan a, compound of kynurenine and sucrose which can functionas the v+ hormone in DrosophiZ~.~~ Tryptophan can be converted in theorganism of the rat into another quinoline derivative, xanthurenic acid,recognized as 4 : 8-dihydroxyquinoline-2-carboxylic acid by synthesis fromo-anisidine and ethyl 0xaloacetate.4~44 Rec. Trav. chim., 1940,59,289; 1941, 80, 138.46 D. G. Harvey, E. J. Miller, and W. Robson, J., 1941, 153.46 3’. Lynen and U. Wieland, AnmaZen, 1937, 533, 93; H. Wieland and B. Witkop,47 Naturwiss., 1942, 80, 61; Y.Kotake and J. Twao, 2. phylsiol. Chem., 1931, 195,4* E. L. Tatum and G. W. Beadle, ,Y&nce, 1940, 91, 468; E. L. Tatum and A. L.*r~ L. Musajo and (Signa.) M, Minchilli, Bw., 1941, 74, 1839.ibid., 1940, 543, 171.139.Haagen-Smit, J . Bid. Chem., 1941, 140, 576STEVENS : HETEROOYCLTd COMPOUNDS. 199Six-membered Rings.-Di-( p-chloroethy1)methylamine condenses inpresence of aodamide with phenylawtodtde and other reactive methylenecompounds-fluor ene , N - m e th yloxindole , phenylmet hylsulp hone- yielding(IIa.) (IIb.)piperidines (I); pyrans and thiopyrans have been made sirnilm1y.w Twostable, readily interconvertible substances 'are regarded as the hydroxy-pyridine (IIa) and the pyridone (IIb) ; only the former gives a coloration withferric chloride.51 The activity of the side chain in a-picoline and the like hasbeen exploited in condensation with such reactive ketones as ethyl mesoxalate,alloxan, or bend to give carbinols ; S2 with cinnamaldehyde, yielding carbinolsand then yellow phenylpyridylbutadienes ; 63 and with formddehyde andsulphanilamide to produce 2- ( p-sulphanilamidoethyl)pyridine.a2 : 3-Disubstituted quinolines (111) are effectively prepared by the inter-action of arylamines and formyl (hydroxymethylene) ketones, the alternative,CH:NPh ,CHO /NH,Ph R*VH ~ _ _ 3 R*VHNZnO1,-EtOH R'CO R'CO(m.)3 :4-compounds not being formed.55 R, H.F. Mamke, L. Marion, andF. Leger have prepared by unambiguous methods and fully characterbed theBeven mono- and 21 di-methylquinolines, as reference compounds foralkaloidal work.56 The Friedel-Crafts reaction has limited application toquinolines and other condensed pyridine~.~6 Hydrogenation of acridinetakes the following course : 34[C13Hi7N -k %Hi7N -k %&LiNI- NihCr,O,Ni or -f (SYmm.) (wvmm.)C13H9N z u z c13HllN -Cl3H1,N (ms~mm.) C13H23NtThe water-soluble compounds of phosphorus oxychloride with acridoneor its N-derivatives are formulated as (IV), because the Grignard reagent,which with the free oxychloride gives triphenylphosphine oxide, converts50611093.62536455580. Eisleb, Ber., 1941, 74, 1433.J.R. Stevens, R. H. Beutel, and E. Wmberlin, J. Amer. Chem. Soc., 1942, 64,S. M. McElvain and H. G. Johnson, &id., 1941,62,2213.E.Spath, G. Kubiczek, and E. Dubensky, Ber., 1941,74, 873.(Signa.) L. Monti and L. Felioi, Uazzetta, 1940,70,375.V. A. Petrow, J., 1942, 693.C a d h n J . Re&, ,1942, 20, B, 133200 ORGANIC CHEMISTRY.these compounds into diphenylphosphinic acids and diacridyl or diacriden(V), and because (IV, R = H) is hydrolysed to 5-~hloroacridine.~7 Aniline,sodium hydrogen sulphide or thiosulphate, and potassium thioselenate ,respectively convert the compounds (IV) into acridoneanils, thioacridones,and selenoacridones. Reduction with zinc dust readily affords the diacridens,which, except the 4 : 4'-disubstituted compounds, are strongly fluorescentand exhibit chemiluminescence in their atmospheric or peroxide oxidation.The diacridylium salts (VI = " lucigenin "), also obtained by reducingacridones with magnesium-magnesium iodide, fluoresce in neutral or acidsolution, and are oxidised in alkaline media with most striking chemilumin-escence, suppressed in this case also by 4 : 4'-substitution.The emission oflight appears to be associated with the reduction of an intermediate peroxide ;i t is suggested that an earlier stage involves diradicals having the compositionof diacridens, which are, however, diamagnetic. By heating with sulphurNN'-dimethyldiacriden is split into methylthioacridone ; or, more readily,into methylacridone by successive action of thionyl chloride and water.Dixant h ylene, dif€av ylene, and their t hio - analogues, but not tetra -arylet hyl-enes, undergo similar fissions.58bicyclo-Axa-aEEanes.59-Progress has been made with the synthesis ofthese ring systems, which occur in several groups of alkaloids. Compoundswith a nitrogen atom as bridge-head (I) are advantageously prepared fromdihalogeno-amines ;so the substance (I ; y = 0, x = x = 5), with fused seven-i5' K. Gleu and s. Nitzsche, J . pr. Chem., 1939, 153, 225, 233; K. Gleu, s. Nitzsche,and A. Schubert, Ber., 1939, 72, 1093 ; K. Gleu and R. Schaarschmidt, ibid., pp. 1246,1404; 1940, 73, 909; K. Gleu and A. Schubert, ibid., p. 805; B. Tamamushi andH. Akiyama, Trans. Paraday Soc., 1939, 36,491,68 A. Schonberg and (Miss) W. Asker, J., 1942, 272, 725.69 Ann. Reports, 1939, 36, 321.8o V. Prelog et al., Annalen, 1940, 545, 229, 231, 243, 247, 266STEVENS : HETEROCYULIC COMPOUNDS.201rings, has been so obtained, and 2- and 3-alkylquinuclidines have beenproduced as follows (R = tetrahydropyranyl) :/T\ //p \Br NH, Br \A/’’R*CH,*CO 2HQH2 QH2 QH2 YH2 QH2 QH2( 3 3 2 QHMep2 CH, CHMeCH,RCHlCOMe .1RCH ,*CHMe*NH,R*COEtCH,Br *CO ,E t } -+ R*CEt:CH*CO,Et + R*CHEt*CH,*CO,H -%R-CHE t *CH ,-NH, a 3 - eth ylquinuclidineWhen heated with hydrobromic acid, p p-diallylethylamine,( C,H5),CH*CH2*NH2gave the dimethyl derivative of (I ; y = 1, x = x = 2).61 F. Lions andA. M. WillisonG2 have effectively synthesised octahydropyrrocolines, by avariant of the Robinson tropinone synthesis, from ethyl acetonedicarboxylate,y-aminobutaldehyde (as acetal), and a second aldehyde :CH2-CHO vH,*CO t CH,-YH*QH*CO,EtR-CHO CH,*CO,Et R-hH*CH*CO,EtCH2<CH,-NH2 QO 4 CH2<CH2-N QOS. Sugasawa and co-workers 63 have prepared fused hydroisoquinolinesystems by applying the Bischler synthesis to N - P-arylethyl derivatives ofactams (n = 2, 3,4) :In the same way they have prepared a benzpyridocoline derivative as apossible starting point for the synthesis of emetine :CH2 CH,/ \EtO,C*E VHMe/ \EtO,C*(IE€ VHMeCHO C02Et CH CO\ /6l R.Paul and H. Cottin, Bull. Soc. chim., 1940, 7 , 626.e 2 J . Proc. Roy. 8oc. New South Wales, 1940, 73, 240.6s J . Pharm. Soc. Japan, 1937, 57, 296; Ber., 1938,71, 1860; 1939,72, 980; 1940,73,782; 1941, 74,465,459, 537; PTOC. Iwy. Acad. Tokyo, 1939, lS, 82202 ORGAXIC CHEMISTRY.and derivatives of several of the possible dibenzpyridocolines, some of whiohcould be dehydrogenated to quaternary compounds in which all four ringsare aromatic.Alkaloids.Simple Pyridine Bases.-Inactive pelletierine acetal has been synthesised 64and converted into precisely identifiable derivatives of the unstable alkaloid :H a d2 ‘yH2 p( OEt),NH CH,c H 2CH(OW2 A C)H(OEt)2 (jhle + vH2 -- b\ I -+ CH, CH CH,CH2 \ / \ // BrCH, NThe hydroxyl group in +-conhydrine is finally shown to be in position 5 byoxidation of dihydro-#-conhydrinemethine t o n-hexyl dimethylaminomethylketone. 65MeO,C*CH,*COCH,*CO ,Me 1 --+ MeO,C*CE-CO-CH*CO,Me NHaMeP. S.Ugriumov 68 has synthesised arecaidine (111) by the route :MeOCHO NH2Me 0:CHIMe M e * p m e * v e 2CHsO;(1.)Me*(7H--NMe-~HMe~ $?H,-CO-pCO,Me I.educe;> $Z12--CH=~C0,Hdehydmte CH *NMe.CH Me02C*~CO-$N202Me CH,*NMe*CH, 2 3CH,*NMe*CH, W.) (111.)and obtained stereoisomeric homologues by treating (I) in the same way asSenecio A E k ; ~ l o i d s .~ ~ - 8 ~ ~ ~ i o and related species afford alkaloids whichare esters of (usually) monoterpene acids of incompletely determinedconstitution, with amino-alcohols (“ necinea ”), frequently retronecine.This mono-olefinic, dihydroxylic base haa been converted into heliotridanunder conditions unlikely to cause rearrangements; and the synthesis ofdl-dihydroheliotridanmethine (I) completes a. P. Menschikov’s 67 demon-stration that heliotridan is 1 -methylpyrrolizidine (II).68 The isomeric2-methylpyrrolizidine has also been synthe~ised.6~PrCH-CHMe EHO *CH-(fE€--C*CH ,*OH pMek AH, AH, N bH\ /CH2(1.1 (11.1 PI.)(11)\1/ \3/ CH2 CH,64 M.A. Spielmann, S. Swedesh, and C. W. Mortenson, J . Org. Chem., 1941, 6, 780;65 E. Spath end R. Lorenz, Ber., 1941, 74, 699.66 C q t . r e d . Acad. Sci. U.R.S.S., 1940, 29,48; J . Ben. Chem. Russia, 1941, 11,J. P. Wibaut and M. G. J. Beets, Rec. Trau. chim., 1940, 59, 663.829.Ann. Reports, 1930, 86, 328.a. R. Clemo and T. A. Melro~e, J., 1942, 424.@8 R. Adams and E. F. Rogers, J . AM. Chem. Soc., 1941,8a, 228STEVENS : HETEfL0UYOI;IU COMPOUNDS. 203Of the two hydroxyl groups in retronecine, one (p) is more readilyesterified than the other (a) ; by hydrogenation, errpeoia;lly of the e&m, theformer group ia usually replaced by hydrogen before the double bond isattacked; but free retronecine is reduced over nickel to a, saturated dihy-droxylic base inert to further hydrogenation, which is identical withplatynecine, the basic moiety of platyphylline.These and other trans-formations are summarised :Retronecine H,-Ptos Deoxyrdxonecine H9-ptOg Retronecanol ketoneOHa, OHP, = onester OHa, = OHa - C,Hl,NOH,-Ni I.c,Platynecine B5C1; chloro-ester Hn-BijF isoRetronecano1 acidOHa, OHp soO1.1 Cla, OBzp hydrol. OHP - C,H1a*C02HThe p-hydroxyl group ie thus primary and in the side chain, C-methyldeterminations in this series being positive only in its absence, and its readyhydrogenolysis shows that the double bond is in the 1 : 2 or 1: 7-position.That retronecine is not a vinyhmine derivative follows from ita stability toacid and alkali, and from the observed increase in basic strength of retro-necine and deoxyretronecine on hydrogenation, in contrast to the behaviourof authentic tertiary cyclic vinylamines.These results establish the formula(111) for retronecine, apart from the position of OHa, which is chosen toaccommodate the formation of a cyclic ether stable to acid when retronecanolis treated with cyanogen bromide and then with alkali.70Further Xenecio species have furnished among others the following newbases : integerrimix~e,~~ longilobine,'l and pter~phine,~~ each of which is theester of retronecine with a specific acid ; r0smarinine,7~ saponified to romar-inecine and senecic acid ; ~tosenine,~~ which gives, probably, jaconecine, witha mmcyclic base; and i~atidine,~~ which is doubly abnormal in yieldingmonocyclic isatinecine with, apparently, a per-mid.The relahd speoiesErechtites hieracifolia and Crotalaria Crantianu afford hieracifoline 7 1 andgrantianine,74 esters of retronecine.isoQuinuline AZlcaZoids.-Hunnemanine (I), from Iiunnemannia furnar-icefolia, is the first phenolic alkaloid of the ten-membered ring seriea. Itsconstitution follows from its methylation to allocryptopine and the degrad-ation of its ethyl ether to a known methoxyethoxy-o-toluic acid.75 Corydalisophiocarpa offers a more striking novelty in the diisoquinoline derivativeophiocarpine (11), hydroxylated in the position corresponding to the' 0 R. Adams and E.F. Rogers, J . Amer. Chem. Xoc., 1941, 63, 537; R. A d m s andJ. E. Mahan, ibid., 1942, 64,2588; R. A h a , M. Carmack, and J. E. Mahan, ibid., p.2693; R. Adam8 and I(. E. Hmlin, ibid., p. 2597.71 R. H. F. M d e , Canadian J . Rea., 1939, 17, B, 1, 8.H. L. de Wad, Nature, 1940,148, 777; J . S. AfTiCcn Chem. Inst., 1941. 24, 29;J. 5. Bbckie, Pharm. J., 1937, 188, 102.73 E. S. Schdanovitsoh and G. P. Menschikov, J . Gen. Chem. Rwaicr, 1941,11,836.74 R. Adams, M. Carmack, and E. F. Rogers, J . A ~ T . Chm. Sac., 1942, 64, 571.75 R. H. F. Manske, L. Marion, and A. E. Ledingham, ibid., p. 1659.a 204 ORGANIC CHEMISTRY.potential hydroxyl group of the phthalide-isoquinoline bases.76 Ophiocar-pine has been converted into tetrahydroberberine, is not a pseudo-base, andgives on oxidation the lactam (111; RR’ = CH202, R” = H).CH, CH,CH, qH2Papaver armeniacum 77 yields armepavine, a methyl ether (IV) ofcoclaurine oxidisable to p-hydroxybenzoic acid and the lactam (I11 ; R =R’ = OMe, R” = Me).Dimerising dehydrogenation of coclaurine wouldlead to magnoline (V), isolated from Magnolia f ~ s c a t a , ~ ~ which gives byoxidation of its triethyl ether the lactam (I11 ; R = OMe, R’ = OEt, R” =Me) with 2-ethoxy-5 : 4’-dicarboxydiphenyl ether.The constitution assigned 79 to bebeerine (chondrodendrine), in which a pairof coclaurine structures are joined “ head to tail ” by two similar dehydrogen-ations, depended on its degradation to the acid (VI), which was decarboxyl-‘11.)OMe7 6 It.H. F. Menske, Canadian J . Res., 1939, 17, B, 61; 1942, 20, B, 57.7 7 S. Junusov, R. A. Konovalova, and A. P. OrBkhov, J . Gen. Chem. Russia, 1940,7 8 N. F. Proskurnina and A. P. Or6khov, ibid., p. 707.10, 641.Ann. Reports, 1937, 54, 360STEVENS : HETEROCYCLIC COMPOUNDS. 205ated and oxidised to a monobasic acid of established structure. The tribasicacid obtained by oxidation of (VI), and also of bebeerilene, the final productof Hofmann degradation of bebeerine, has now been unambiguously syn-thesised by F. Faltis, L. Holzinger, P. Ita, and R. Schwarz.*O These authorsemphasise the possible biogenesis of such alkaloids by successive dehydro-genations and methylations of coclaurine, and propose the modified formula,(VII) for trilobine.Erythrina and Lycoris AZkaZoids.-Erythrina species afford, as well ashypaphorine (tryptophan methylbetaine), a series of '' free " alkaloids,*lisolated by the usual procedures and assigned names beginning with theelement " erythr-," and also a series of " liberated " bases *2 extractable onlyafter hydrolysis ( ? of water-soluble glycosides) and distinguished by theprefix " eryso- ." Erythraline, C,,H,,N(OMe)( O,CH,), is a tertiary baselacking NMe and CMe groups ; it yields indole on potash fusion, oxidation ofthe methohydroxide affords 4 : 5-methylenedioxyphthalic acid, and theabsorption spectrum of tetrahydroerythraline closely resembles that of6 : 7-methylenedioxytetrahydroisoquinoline. It is suggested that tetra-hydroerythraline, which is evidently tetracyclic with the nitrogen atomcommon to two rings, has the modified benzylisoquinoline structure (I).Erythramine is dihydroerythraline, and erythratine a closely related basewith an additional alcoholic hydroxyl group.The following " liberated "bases are phenolic, the first-named probably a catechol derivative : erysopineand erysonine, C,,H,,O,N( OMe) ; erysodine and erysovine, C,,H150N(OMe),.CLycorenine, C,,H,,( OH),( OMe),(NMe), yields a monoxime, and givesformaldehyde on ozonolysis, as does the nitrogen-free product of its Hofmanndegradation. The latter also affords 3 : 4-dimethoxy-6 : 3'-dialdehydodi-phenyl, the related acid having been synthesised.83 Lycorenine is regardedas (11), with a hydroxyl group and a double bond in ring 3. Oxidation bylead tetra-acetate shows that the hydroxyl groups in dihydrolycorinone, andtherefore in lycorine, are attached to adjacent carbon atoms.8q*O Ber., 1941, 74, 79.81 K. Folkers and R. T. Major, J . Amer. Chem. Soc., 1937, 59, 1580; K. Folkere sndF. Koniuszy, ibid., 1939,61, 1232, 3063; 1940, 62,436, 1673; K. Folkers, F. Koniuszy,and J. Shavel, ibid., 1942, 64, 2146.8a Idem, ibid., 1941, 63, 1544; K. Folkers and F. Koniuszy, ibid., 1940, 68, 1677 ;K. Folkers and J. Shavel, ibid., 1942, 64, 1892; R. A. Gentile and R. Labriola, J . Org.Chem., 1942, 7, 136.83 H. Konda and T. Ikeda, Ber., 1940, 73, 867.84 H. Kondo and K. Katsura, ibid., p. 112ORQANIC) CHEMISTRY. 206C i m h Alkdokh.-Remova1 of halogen aoid fiom halogenodihydro-quinidine (I) gave a, secondary base, niquidine, now separated into twocomponents. These are regarded as geometrical isomerides (II),85 rrhceformaldehyde is formed in the reaction, and the niquidines afford on oxidationacetaldehyde and on reduction the same dihydroniquidine, oxidised byhydrogen peroxide to p-propylglutaric (111, R' = H) and quininic acids.The same reagent oxidises dihydroquinidine to the acid (111, R' = C0,R).Niquine, derived from quinine, undergoes degradations analogous to thoseof the niquidines; and dihydroniquine, when boiled with acetic acid,yields dihydroniquidine and epi-C,-dihydroniquidine by inversion at C, andThe preparation of epiquinine and epiquinidine by epimerisation ofquinine or quinidine has been impr~ved,~' and all four vinyl-free stereo-isomerides synthesised.88Aconite Alkaloid8.-A secondary alcoholic group in mesaconitine andrelated bases has been oxidised to carbonyl, and the transformations of theketones suggest that mesaconithe contains the structure-CH2*CH(OH)*CH,*b(OMe)*CH2*NMe*.8gR. A. Konovalova and A. P. Or6khov,90 among others, consider that theaconitines may be derived from a fundamental nucleus C19H,,NH, anddevelop formula as follows :C,.86Aconitine C ~ 9 H ~ ~ E t ( o H ) ~ ( o M e ) ~ ( o A c ) ( ~ B z )Mesaconitine Cl,H1,NMe( OH),( OMe),( OAc) (OBz)Hypaconitine C,,H2,,NMe(OH),(OMe),(OAc)(03Bz)Pseudaconitine ClgH2JWt( OH),( OMe),( OAc)( OVeratroyl)Indaconitine C,,H,,,NEt(OH),( OMe),( OAc)( OBz)Bikhaconitine ClgH2 ,NEt( OH) (OMe),( OAc)( OBz)Lappaconitine C 19H ,,We( OH) ,( OMe),( O*CO*C,H,*NHAc)The much less toxic atisines are unesterified hydroxylic bams, of which the=me authors have isolated four new examplee from Aconitzcm talussiczlm;8 5 E. M. Gibbs and T. A. Henry, J., 1939,240,1294.86 W. Solomon, J., 1941, 77.6' P. Rabe and H. Hater, J . pr. C-., 1939, 164, 66.88 P. Rabe and G. Hagen, Ber., 1941, 74, 636; V. Prelog et al., i b i d . , p. 647.R. Majima and K. Tarn-, Annah, 1940,545,l.Bull. Soc. chim., 1940, 7, 95: cf. W. Freudenberg, Ber., 1936, 69, 1962STEVENS : HETEROCYCLIC COMPOUNDS. 207these last may be referred to the same nucleus as the aconitines if theN-substitutions (not yet investigated) are assumed to be a8 shown :Tal atisine C1BH,,NMe(OH), -at. ; dextrorotatory.Talatisamine C,,H,,NH(OH) ( OMe), inactive.Talatisidine C,,H,,NEt( OH),( OMe), lmvorotatory.isoTalatisidine 9) inactive.The original atisine from A . hderophyllzlrn, now shown to be C21H,,-,NMe(OH),with two double bonds and at lea& one CMe group, gives l-methylphenan-threne and another phenanthrene homologue C17H1, on dehydrogenation withselenium. The same plant affords the mono-olefinic hetisine, C20H2,03N,and the saturated heteratisine, Cl,H2,NMe( OMe)(OH),(*CO*O*), which maybe related to the bctonic Sterraona bases?l Napelline, C,,H,,NMe(OR),,which occurs with aconitine in A . napellus, yields on dehydrogenationphenanthrene homologues, including possibly the C,,H,, hydrocarbon froma t i ~ i n e . ~ ~ The tertiary base kobusine, C,,H,,N(OH),, oceur8 in A.SachalineneeSabadilla and €Wlebore AlkaEoids.-These fall into two classes : tho%, suchas jervine, which are fkee hydroxyhted bases not found in ssbadilla, seeds;and esters, like cevadine (tiglylcevine) ; it is suspected that the free andesterified alkamines are chemically related. Dehydrogenation of cevine gpwith selenium has afforded, as well as cevanthrol and cevanthridine (whichmay be C,,H,,N), 4 : 5-benzhydrindene, and p-picoline with its a'-ethyl and-hydroxyethyl derivatives. Other hydrocarbons separated chromato-graphically are regarded from their absorption spectra as derived fromcycbpentenofluorene. By oxidation cevine gave an acid C6H1,(C0,H),,which yielded a dianhydride and thence a keto-anhydride, and appeared tohave two " primary " and two " tertiary " carboxyl groups, indicatingunambiguously the formula (I). A further product was the lactone of amonocyclic acid ClOHl5( OH)(CO,H),, convertible into decevinic acid, thestructure proposed for which is now withdrawn. The skeleton (11) issuggested for cevine.HO,C*CH,*~Me*CO,HHO 2CCH,*CMe*C0 2H(1.)91 A. Lawson and J. E. C. Topps, J., 1937, 1640; W. A. Jscobe and L. C. Craig, J.9% Idem, aid., p. 611 ; W. Freudenberg and E. F. Rogers, J . Amer. Chmn. SOC., 1937,Og H. Suginome and H. Simamouti, Annakn, 1940, 545, 220.O4 L. C. Craig and W. A. Jacobs, J . Biol. Ckm., 1941,139,263,293; 141,263; idem,with G. I. Lavin, aid., 18g, 277; cf. Ann. Report8, 1940, 37, 379.Biol. Chem., 1942, 148, 689, 006.59, 2672208 ORGANIC CHEMISTRY.By dehydrogenation, protoveratrine gave several of the products obtainedfrom cevine.95 JervineJ96 which is accompanied in white hellebore by thenew mono-olefinic base veratramine, C,,H,O(NH)( OH):’ gave on dehydro-genation 3 -met hyl-6 - e th ylp yridine with (probably ) its 5 - h ydrox y -derivative ,and hydrocarbons regarded on spectroscopic grounds as homologues of4 : 5-benzhydrindene and 2 : 3-benzfluoreneJ but no base corresponding tocevanthridine .from 8. aviculare , is identical withsolasonine , which is gg formulated : rhamnose - galactose - glucose - solaso-dine. Further reactions confirm the view that solasodine is a steroidderivative unsaturated a t C, - C, and hydroxylated at C,, and the behaviouron alkylation , hydrogenation, and bromination suggests a pseudo- basicsystem (111). 8. auriculatum affords, with solasonine, the very similarsolauricine , hydrolysed to rhamnoae , galactose, glucose, and so1auricidine.lErythrophleum AZkaZoicE8.-The constituents of E. Guineense arevariable, some samples containing only erythrophleine and others onlyca,ssaine and its allies. These bases, with coumingine and coumingidine fromE. Couminga,9 are alkamine esters (see table) of diterpenoid acids derivedfrom the tricyclic cassanic acid C2,H,,O2, which yields 1 : 7 : 8-trimethyl-phenanthrene on dehydrogenation ; caasaic acid is a ketodihydroxycassanicacid, and erythrophleic acid may be a methoxycassaic acid.Solanum AZkaEoids.-PurapurineErythrophleine --+ Erythrophleic acidCassaine + Cassaic acid + NMe,*CH,*CH,*OH--+ Dihydroxycassenic acid + NMe,*CH,*CH,-OHCoumingine --+ Cassaic acid + + NMe,*CH,*CH,mOHHO-CMe,-CH,CO,HCoumingidine --+ Cassaic acid (inter aZia) + NHMe*CH,CH2*OH+ NHMe*CH ,CH,-OH*CassaidineT. S. S.85 L. C. Craig and W. A. Jacobs, J . Biol. Chem., 1942, 143,427.B’ K. Saito, Bull. Chem. Sm. Japan, I940, 16, 22.O * R. C. Bell and L. H. Briggs, J., 1942, 1.*@ L. H. Briggs, R. P. Newbold, and H. E. Stace, J., 1942, 3; cf. Ann. Reports, 1940,37, 378.L. H. Briggs, J. J. Carroll, and R. C. Bell, J . , 1942, 12; L. H. Briggs and J. J.Carroll, J., 1942, 17.F. Fcsltis and L. Holzinger, Ber., 1939, 72, 1443; B. K. Blount, H. T. Openshew,and *4. R. Todd, J . , 1940, 286; G. Dalma, Helv. Chim. Acta, 1939,22,1497; L. Ruzickeand G. Delma, ibid., p. 1616; 1940,23,753.* L. Ruzicka, G. Ddma, and W. E. Scott, i b d , , 1941,24,63, 179E; idem and B. G .Engel, ibid., p. 1449; E. Schlittler, ibid., p. 319E.W. A. Jacobs, L. C. Craig, and G. I. Lavin, ibi&., 1941, 141,51.,, Identical with “ norcastmidine.

ISSN:0365-6217    

DOI:10.1039/AR9423900113

出版商:RSC    

年代:1942

数据来源: RSC

摘要:

BIOCHEMISTRY.1. NUTRITION.FOR many years past it has been the habit-of Yuccessive authors of thesection on Biochemistry to begin the review of the year’s work with asystematic account of the continual growth of knowledge about vitamins.We are breaking with tradition by skipping the theme on this occasion-notwithstanding its undiminished interest and importance-and devotingattention to two topics which seem ripe for review. The first is of particularinterest during the war, namely, the evolution of new methods for thedehydration of foodstuffs by which their nutritive value may be left largelyunimpaired. This, incidentally and in spite of our opening remark, willnecessarily bring into the picture one aspect of applied vitamin chemistry.The second topic has to do with the other kind of “ accessory food factors,”as distinct from vitamins, namely, those inorganic “ trace elements ”essential for animal and plant health.Dehydrated Foods.A recent conference of The Nutrition Society was devoted to a descriptionof the new developments in the preparation of dehydrated foods, and theoriginal records may be consulted for the complete detai1s.l The informationreleased at this meeting is the fruit of several years’ co-operative investigationscarried out under the auspices of Government research departments, (Sir)E.Appleton,2 Secretary of the Department of Scientific and IndustrialResearch, explained how the Low Temperature Research Station (FoodInvestigation Board) under F. Kidd at Cambridge had been mainly respons-ible for the experimental work on processing, and the Dunn NutritionalLaboratory (Medical Research Council) under L.J. Harris, also at Cambridge,for ensuring the preservation of nutritive values.Dried milk and dried eggs have, it is true, long been familiar articles ofcommerce, and perhaps the most significant new departure, therefore, is thatit is now possible for the first time to produce dried vegetables, which, afterreconstitution and cooking, are not inferior to cooked fresh vegetables inrespect of their principal dietetic virtue, i.e., antiscorbutic potency, as wellas in their palatability and appearance. It may be recalled that driedvegetables were not unknown in the last war, but they were then ineffectivein preventing outbreaks of scurvy among the troops and were unattractivein colour and taste.Another aspect, which scarcely needs stressing inwartime, is the immense sawing in shipping space which is effected when foodsare imported in the dried state. It has been calculated that over 3000 tonsof water were imported annually in our food before the war.2 With dried1 Proc. Nutrit. Soc., July 25th, 1942 (vol. 1, 1943, in press) ; for abridged report, seeChem. and Ind., 1942, 61, 342; Lancet, 1942, 2, 192, 186; Brit. Med. J . , 1942, 2, 254;Faod Manufacture, 1942, 17, 291.a Proc. Nutrit. Sac,, {hid210 BIOCHEMISTRY.meat there is the additional aaving of refrigeration space, not to mention theroom occupied by the bones and the other inedible parts of the carcase.Moreover, the dried products are sometimes more simple to prepare and cookthan are the fresh, and it is an added convenience, for example, for armies inthe field, to have their rations available in a compact, ready-for-use andnon-perishable form.In one respect the dehydrated foods may even beregarded as nutritionally superior to the fresh : in the preliminary processingof vegetables to which they are subjected prior to the operation of dehy-dration, an enzyme, ascorbic oxidase, is inactivated ; in untreated foods thisenzyme is able under some conditions t o destroy the vitamin.We may cite, as a typical example, the procedure which is adopted in thedehydration of ~abbage,~ for retaining its vitamin C and its culinary'' quality "-these two properties frequently, but not invariably, go together.The cabbage, cut into wide strips, is scalded in water containing a trace ofsulphite.This step is necessary in order to protect ascorbic acid fromsubsequent oxidation and to prevent deterioration in appearance. Toreduce the destruction of ascorbic acid, this scalding is done in the minimumamount of water, and for the least time necessary to inactivate the enzyme.Also the same water is used repeatedly to prevent undue loss of ascorbic acidby diffusion. Conditions of humidity, temperature, reaction, etc., areimportant during dehydration as well as during subsequent storage. Forinatance, dried cabbage is packed in an inert gas, in air-tight tins and at lowrelative humidity to preserve the vitamin.Such cabbage wm reconstitutedand cooked and tested in the R.A.F. : it wits judged more palatable thanordinary cooked cabbage and found to contain twice as much vitamin C.4By procemes differing in points of detail it has been possible to producedried carrots containing their full 'complement of carotene;3 dried meatsactive in vitamin B,, nicotinic acid and riboflavin and with protein of highbiological value;5 and dried egg8 with vitamins A and I> and protein-valueunimparied.5 Work on the drying of Jean fish is also promising.*After the war, we can expect that the use of dehydrated foods will haveimportant economic consequences. It will make a wider choice of highquality commodities available to aJl sections of the community, and glutswill be more readily dealt with.' The dried foods will be needed in post-warrelief; and ultimately we may expect that their use will affect the wholemachinery of world distribution, storage and consumption of foodstuff s.8Trace Elements.Another of the last year's conferences of The Nutrition Society wasdevoted to a consideration of " Trace Elements in Relation to Health." 9 Afairly comprehensive review on "The Significance of Trace Elements inR.J. L. Allen, J. Barker, and L. W. Mapson, Food Manufacture, 1942,17,291.T. F. Macra0, aid.E. M. Cruickshsnk, J. (3. Sharp, and E. 0. Bate-Smith, ibid.G. A. Reay, a i d .Proc. Nutpit. Soc., Oct. 17th, 1942 (vol. 1, 1943, in press); abridged report in Chem.7 (Sir) J. Barcroft, ibM. 8 J. Hemmond, aid.and Id., 1942, 61, 403HARRIS: TmmrnON. 21 1Nutrition ” is also available.10 Our object here will be to attempt a briefsumming-up of the present position.De$nition-s.--In the view of H.H. Green: we should use the term “ traceelement” to include all those minerd elements normally or occasionallypresent in living tissues in small traces, no matter whether they are essentialto health, or toxic, or apparently inert. The term has, however, frequentlybeen restricted to such of these elements a8 are known to be essential to life,or has been sometimes still further limited to those which are essential toanimal life (e.g., by E. J. Underwood 10). It would probably be better tokeep to the first-mentioned dehition. Then we can subdivide the traceelements into three classes.The first are the essentiaE trace elements,namely, those which are needed for the maintenance of health, and are thenormal constituents of cells, such, for example, aa Cu and Zn. As C. R.Harington9 has pointed out, we can only feel fully secure in admitting aclaimant to this category when its biological function has been dearlydefined: Cu and Zn do fidfll this test. The second class is that of theabnormal or toxic trace elements which may appear in some soils, pastures ordrinking water and so cause damage to plant or animal health, e.g., Se andMo. The third class is that of the trace elements consistently occurring inliving matter but for which no function has yet been traced : some elementsin this clasa may yet prove to need promotion to Class 1 or Class 2.Further-more, some trace elements appear to be needed by plants and not by animals,and the reverse may be true of others.T r w Elements for Plants and Animab.-For example, boron is essentialfor plants, but apparently not for animals. Tungsten is said to atimulateearly but not late growth in plants, but is likewise not known to be needed byanimals. On the other side of the picture, Zn, Cu and Mn appear to be ofrelatively greater importance to the animal than to the p1ant.lEnumeration of Trace Elemenh.-For animals the following appear to bethe essential trace elements (apart from Fe, which is needed in more than“ traces,” and I, whose functions had long previously been recognised) :Co, Cu, Mn, and Zn.Some of the deficiency diseases caused by the absenceof essential trace elements (e.g., Cu and Co) occur spontaneously in animals incertain areas, one a t least has only been produced experimentally (Zn).Elements which commonly oocur in the tissues of higher animals but forwhich no function has yet been clearly defined include Mi, F, Si, As, Al, Rb,Br, Ba. Toxic trace elements, occurring naturally in soils, pastures ordrinking water, and responsible for spontaneous disease in man or animalsare F, Mo, and Se. Among the trace elements which in some circumstancesmay be necessary for, or may stimulate, the growth of plants are includedCu, Zn, W, and Mn. Some of these essential trace elements, e.g., Cu, becometoxic to plants when present in concentrations relatively higher althoughstill so small absolutely as to merit the term ‘‘ trace ”; and the same is truefor animals.10~ 11lo E.J. Underwood, Nzctrit. Abetv. Rev., 1940, 8, 616.l1 See review by H. H. Green, Pvoc. Nuhit. SOC., 1942, Zoo. cit212 BIOOHEMISTRY.Physiological Function and Enzyme Systems.-The rale of a trace elementcannot be clearly understood until it has been identified as a constituent of acompound or system possessing some definite biochemical action. We nowhave knowledge in these terms of the biochemical significance of two t’raceelements, Cu and Zn.12 Although Cu has long been known to be present inhaemocyanin, the respiratory pigment of crustacea, and in turacin, a pigmentin birds’ feathers, it is only recently that T.Mann and D. Keilin 13 haveisolated it from mammalian tissues in the form of crystalline Cu-proteincompounds, hzemocuprein and hepatocuprein, from red blood cells and fromliver respectively. Cu has also been proved to be a constituent of variouspolyphenol oxidase systems, e.g., the polyphenol oxidase of potatoes la-responsible for their darkening when cut-that of rnushr~oms,~~ and1a~case.l~ Other enzymesls which appear to be Cu-protein compounds areascorbic oxidasel’ and indophenol 0xidase.1~ Not less important is thediscovery that zinc forms a part of the enzyme carbonic anhydrase, whichis concerned in the transport and excretion of carbon dioxide and in main-taining the acid-base equilibrium of the body.12, SoDeJiciency of Co and Cu in Farm Animals.-A disease occurring spon-taneously among sheep and cattle in some parts of Western -4ustralia, knownas “ enzootic marasmus,” has been traced to deficiency of C O .~ ~ Stock cannow be kept free from the disease by the administration of a fraction of amilligram daily of a cobalt salt.22 Similarly in Southern Australia the“ coast disease ” of sheep has been proved to be due to a combined deficiencyof cobalt and copper.23” Diseases of cattle or sheep likewise due todeficiency of Co are the “ bush sickness ” of New Zealand,25 the “ pine ” ofScotland,26 the “ salt-sick ” of Florida,27 the “ nakuruitis ” of Kenya, and al2 See D. Keilin and T. Mann, Proc. Nutrit. SOC., 1942, and earlier literature there cited.l3 Proc. Roy. Soc., 1938, By 126, 303.lP F.Kubowitz, Biochem. Z . , 1937, 292, 221; 1938, 296, 32.l5 F. Keilin and T. Mann, Proc. Roy. SOC., 1938, B, 125, 187.l6 Idem, Nature, 1939, 143, 333.l7 E. Stotz, C. J. Hailer, and C. G. King, J . Biol. Chem., 1937, 119, proc. xcv.Purified catalase contains Fe, but not Cu as for a time supposed by K. Agner,D. Keilin and E. F. Hartree, Proc. Roy. SOC., 1938, B . 125, 171.Biochem. J., 1938, 32, 1702.2o D. Keilin and T. Mann, Biochem. J., 1940, 34, 1163.21 J. F. Filmer, Austral. Vet. J . , 1933, 9, 163; E. J. Underwood and J. F. Filmer,ibid., 1935, 11, 84; J. F. Filmer and E. J. Underwood, ibid., 1937, 13, 57; E. J. Under-wood and R. J. Harvey, ibid., 1938, 14, 183.23 J. F. Filmer and E. J. Underwood, ibid., 1937, 13, 57.Z s E.W. Lines, J . Counc. Sci. Indust. Res. Austral., 1935,8, 117; H. R. Marston, ibid.,p. 111 ; H. R. Marston and I. W. McDonald, Commonwealth Austral. Counc. Sci. Indust.Re8. BUZZ., 1938, No. 113, 72 and 79.24 For a fascinating account of “Cobalt and other trace elements in relation todisease in Australia,” see review by Sir C. J. Martin, Proc. Nutrit. BOO., 1942, ZOC. cit.26 R. E. R. Grimmett and F. B. Shorland, N.Z. J . Agric., 1935, 50,367; H. 0. Askewand J, K. Dixon, N.Z. J . Technol., 1936, 18, 73; E. M. Wall, ibid., 1937, 18, 642.26 13. H. Corner and A. M. Smith, Biochem. J., 1938,32, 1800.47 W. M. Neal and C. F. Ahmsnn, J . DaiTy Sci., 1937,20,741HARRIS : NUTRITION. 213corresponding condition in Western Canada.28 Also, several workers havehad indications that a deficiency of cobalt may be one of the factors whichlower resistance to helminthic infestation in Great Britain.29A deficiency of copper has been shown to be responsible for diseases infarm animals occurring spontaneously in various restricted areas, e.g., for" Lecksucht " in cattle in Holland,30 " enzootic ataxia " of newborn lambsin parts of Western Australia,31 or " swayback " in newborn lambs inEngland.32~ 33 " Swayback " is characterised by a macrocytic hyper-chromic anaemia.The pathology has been studied by J. R. M. I n n e ~ . ~ ~The disease is prevented by administering Cu to the pregnant ewe or by usingCu salts as a top dressing on the deficient pa~tures.3~Experimental Dejiciency of Co and Cu.-Long prior to the discovery of thespontaneously occurring Co-deficiency disease of farm animals just described,G.Bertrand and his collaborators 35 had proved that the element could bedetected in animals' organs; and tests on mice36 seemed to indicate itsprobably essential nature. Further work on rats:' however, was incon-clusive, and thus it was left to the veterinarians to settle the issue, asrecorded above.With copper it was the other way, since the element had already beenproved t o be necessary for hcemoglobin formation in the rat.38 Its functionapparently is in the conversion of absorbed Fe into hcemoglobin and it aidsrapid haematopoiesis.Mn as an Essential Trace Element.-The presence of traces of Mn inanimal and vetegable tissues was first proved by G.Bertrand and F.Medigre~eanu.~~ In rats kept on a diet devoid of Mn, an experimentaldeficiency-disease is produced, characterised by abnormalities in reproduc-t i ~ n . ~ l Lately, the question of Mn deficiency has emerged as of practical aswell as theoretical importance, since diseases occurring spontaneously infowls, known as " perosis " (or " slipped tendon ") and " chondrodystrophy,"28 J. E. Bowstead and J. P. Sackville, Canadian J . Res., 1939,17, 15.29 W. L. Stewart; E. L. Taylor, Proc. Nutrit. SOC., 1942, loc. cit.3O B. Sjollema, Biochent. Z., 1933, 267, 151; E. Brouwer, A. M. Frens, P. Reitsma,and C. Kalisvaart, Landbouwk. Onderz. Rij'kslandbouwprasfstation, Hoorn, 1938, No. 44(4) C, 267.31 H. W. Bennetts and F. E. Chapman, Austral.Vet. J., 1937, 18, 138; L. B. Bull,H. R. Marston, D. Murnane, and E. W. L. Lines, Commonwealth Austral. Counc. Sci.Indust. Res. Bull., 1938, No. 113, 23.32 G. Dunlop and H. E. Wells, Vet. Rec., 1938, 50, 1175 ; J. R. M. Innes, Proc. Nutrit.SOC., 1942, loc. cit.; T. Dalling, ibid.a 3 Rep, Inst. Animal Pathol., Cambridge, 1934, 4, 227.34 H. W. Bennetts and A. B. Beck, Austral. Vet. J., 1939.3 5 G. Bertrand and M. MBcheboeuf, Compt. rend., 1925, 180, 1380, 1993.36 G. Bertrand and H. Nakamura, ibid., 1927, 185, 321.37 F. J. Stare and C. A. Elvehjem, J . Biol. Chem., 1933, 99,473.38 E. B. Hart, H. Steenbock, J. Waddell, and C. A. Elvehjem, ibid., 1928, 77, 797.39 See review by C. A. Elvehjem, PhysioE. Rev., 1935, 15, 471.40 Ann. Inst.Pasteur, 1913, 27, 1, 282.41 E. R. Orent and E. V. McCollum, J . Biol. Chem., 1932,98,101214 BIOCHBMISTR~.have been traced fo a lack of the element. The former disorder involvesa malformation and crippling of the birds’ legs, and the latter a deformityin the embry0s.aExperimerztal DeJiciency of 2n.-No instance is known of deficiencydisease due to absence of Zn occurring spontaneously in men or animals.Nevertheless, the experiments of G. Bertrand and B. Benzon 43 on mice gavesome preliminary suggestion that it was probably an essential element, andthis was eventually proved more definitely by W. R. Todd, C. A. Elvehjem,and E. B. Hart:&, in their tests on rats, although the only symptoms so farrecorded are diminished growth and poor fur.Toxic Trace Elements and Farm Animak-Two examples of diseaseswhich occur spontaneously in farm animals in England, as a result of theabsorption of traces of toxic elements from the soil of the pasturage, may becited, namely, those caused by fluorine and molybdenum.Molybdenosis isseen in sheep and cattle in Somerset and elsewhere and is characterised bydiarrhoea (“ scouring ”). The severity of the dhease goes parallel with theamount of Mo in the pastures. It can be produced experimentally byadministration of Mo salts. Treatment is by medication with coppersulphate, by changes in the pasture grown on the affected soil, or by makingthe soil more acid.44Fluorosis in farm animals 45 is sometimes an “ industrial disease,” sinceit is seen near brick works, glass works, aluminium factories, or superphos-phate factories, the fluorine being discharged into the atmosphere by thefactory smoke and settling on the land.The animals affected show lame-ness, enlargement of bones and cachexia. An account of industrial fluorosisin England ia given by F. Blake~uore.~~51-In the human, on the other hand,fluorosis, although sometimes an ‘‘ industrid ” disease,51 seems to beaasociated more often with the use of drinking water, particularly well water,contsining an undue amount of fluorine (uix., more than about one part permillion). Its symptoms are mottling of the enamel of the teeth, and, inmore severe cases, osteosclerosis of the bones, especially the spine. It isBeen not only in India, and in parts of America, but also in some meas inEngland,48* 49* 50 there being an appazent relation between the fluorine contentof the rocks or a03 and the incidence of theFluorosis and Dental Disease in Hamarts.-It is known that fluorine is anormal constituent of the teeth, and it has been stated that there may bemore F in sound teeth than in carious teeth.It is not surprising to learn,F1uorosia in Ht.~rmzns.~~* 489 49942 For literature, see review by Underwood, Eoc. cit.43 Compt. rend., 1922, 175, 200.4 4 W. S . Ferguson, Proc. Nzstrit. SOC., 1942, Eoc. cit.46 See review by A. W. Peiroe, Nutrit. Abstr. Rev., 1939, 9, 263.4u PTOC. Nutrit. Soc., 1942, loo. cit.‘7 M. C . Smith, Amr. J . Publ. Health, 1936, 25, 696.4a M. M. Murray, Proc. Nut&. SOC., 1942, Eoc.d t .48 D. C. Wilson, &id.61 H. A. Krebs, &id.p90 AWW. J . Phykol., 1934, 107, 146.F. H. Kemp, ibid.sa C. N. Bmmehead, ibidANDERSON : IMMVNOUEEMI8TRY. 215therefore, that it has been maintained in America, and evidence fromEngland seems to support the claimJs3 that teeth which have been mottledby fluorosis are less liable to decay than are unmottled teeth. If this is so,we have another instance of a trace element which can be both " essential "and '' toxic "; this would explain what, at first sight, appears to be theparadox of one disease (fluorosis) preventing another (oaries).L. J. H.2. IMMUNOCHEMISTRY.It is obviously impossible in the space available to review the whole ofthe great output of papers on this subject which has appeared since the lastReport, for 1940.Accordingly a somewhat arbitrary selection bets had tobe m%de.Antigens.BacteriaZ.-An antigenic polysaccharide built up of equimolecularproportions of acetylated d-glucosamine and galactose (but containingneither uronic acid nor pentose) was obtained from virulent and avirulentB. anthracis by G. 1vanovica.lBuct. dysenterice Shiga has been shown to contain a protein neurotoxin 2and a polysaccharide-protein-phospholipin somatic antigen,3 the former beingproduced by both " rough " and " smooth " strains and the latter by" smooth " strains only. S. M. Partridge and W. T. J. Morgan * have shownthat the somatic antigen complex c&n be dissociated by treatment withformamide into the non-antigenic phospholipin and a polysaccharide-protein moiety which has the somatic antigen properties of the " smooth "organisms.Treatment with trypsin gives the feebly antigenic phospholipin-polysaccharide. The free polysaccharide behaves as a hapten, reacting withanti-sera, but being non-antigenic. The protein-polysaccharide complexcan be split, by solution in 90% aqueous phenol and dialysis, to give thepolysaccharide hapten and the antigenic protein, which, however, no longercarries the specific somatic antigenic properties of the complex.5 Thecomplex can also be degraded by boiling with 1 yo acetic acid ; an almost non-antigenic protein is then obtained, which can be further dissociated bysolution in phenol, probably with loss of a prosthetic group.The poly-sacoharide and the protein can be recombined by solution in formamide andprecipitation with alcohol. If the conjugated protein is used, an antigenwith the properties of the somatic antigen of Bad. dysenterice is produced;the simple protein, however, give8 rise t o a non-antigenic complex, suggestingthat the prosthetic group is essential for antigenicity.63 M. M. Murray, ref. 48 ; but cf. J. D. King, Proc. Nutrit. Sac., 1942.1 2. Imun., 1940,97,402.2 A. Boivin and A. Delaunay, Comlpt. rend. SOC. Biol., 1940,133,376; R. Prigge andL. Kicksoh, 2. Hyg., 1941,123,417; T. Wagner-Jauregg and E. Helmert, Angew. Chem.,1942, 55, 21.3 W. T. J. Morgan and 5. M. Partridge, Biochsnz. J., 1940, 84, 169.4 Brit. J . Exp. Path., 1940, 21, 180.6 W.T. J. Morgm end S. M. Partridge, Biochem. J,, 1941, 86,1140216 BIOCHEMISTRY.A similar antigenic complex has been isolated from Bact. typhosum,0 901, by E. Soru and C. Combiesco and by W. T. J. Morgan and S. M.Partridge by extraction with diethylene glycol. The latter showed thatit is dissociated by boiling with 1% acetic acid into an ether-soluble phos-pholipin, a water-soluble polysaccharide, and an insoluble protein. Thepurified polysaccharide, 1.2yo of organic phosphorus, less than 0.1 yo ofN, [a]5461 + 128", gave a precipitin reaction with anti-0 serum. It wasnon-toxic to mice and not antigenic. The purified protein, 1l05Y0 N,o.47y0 P, [a15461 -55', soluble in alkali, but insoluble in acid, has propertiesvery near to those of the corresponding protein from Bact.dysenterim Shiga,having the same absorption spectrum and giving rise to antibodies whichreact with the Shiga protein. The proteins can replace one another incombination with the polysaccharide of either organism to give antigenswhose specificity corresponds to that of the polysaccharide. A similarprotein, 10°7y0 N, 1.1% P, [a]6461 -50°, was also isolated from Bact.dysenterice Flexner 88. prepared anantigenic substance from Bact. typhosum, by extraction with 85% phenol,which conferred immunity to living organisms when injected into mice.J. H. Orr and G. B. Reed obtained antigens having the characters of apolysaccharide from CZ. welchii by precipitation with alcohol of an extractmade by boiling the cultures with acetic acid.The substance gaveprecipitates with homologous antisera.S. C. Wong and T. Tung lo and S. C. Wong 11 obtained protein andpolysaccharide fractions from the various types of C. diphtherim and showedthat the protein fractions contain at least a labile type-specific protein and aheat-stable protein common to the species. The specific antigen was lostwith loss of virulence of the organism. The polysaccharides were very poorantigens and did not give precipitation reactions with antisera, although theygave complement fixation. It is claimed by L. Hoyle 12 that three lipoidantigens can be obtained from the gravis, mitis and intermedius types ofC. diphtherice and from C. hofmannii by extraction with alcohol. By crossabsorption experiments he showed them to be (1) a specific antigen, h,present only in C.hofmannii, (2) a specific antigen, d, characteristic ofC. diphtherice mitis but probably also present in small amounts in gravis andintermedius strains, and (3) a group antigen, G, present in large amounts inC . diphtherice gravis and intermedius and in C. hofmannii and in smallamounts in C. diphtherie mitis. It seems possible that the antigens are notreally lipoids but either proteins or polysaccharides extracted with the lipoid.Leuconostoc mesenteroides antisera give precipitin reactions with thepolysaccharide, dextran, derived from broth cultures of the 0rgani~rn.l~J. W. Palmer and T. D. GerloughCcnnpt. rend. SOC. Biol., 1940, 133, 498.Science, 1940, 92, 155.7 Brit.J . Exp Path., 1940, 23, 151.lo PTOC. Soc. Exp. Biol. Med., 1940, 43, 749.l 1 Ibid., 1940, 45, 860.l3 T. H. Evans, U'. L. Hawkins, and H. Hibbert, J . Exp. Med., 1941, 74, 511.J . Bact., 1940, 40, 441.l2 J . Hyg., 1942, 42, 416ANDERSON : 1MMUNOUH.EMISTRY. 217E. J. Hehre l4 and E. J. Hehre and J. Y. Sugg l5 have shown that dextrancan be synthesised from sucrose by an enzyme in sterile filtrates fromL. mesenteroides and that the polysaccharide gives precipitin reactions withantisera to the homologous organism and to Types 11, XI1 and XXpneumococci but not with antisera to Types I and I11 pneumococci. Thepolysaccharide is synthesised from no sugars other than sucrose and raffinose.D. G. Evans l6 has shown that a heat-labile toxin can be obtained fromHcemophilus pertussis by freezing suspensions a t low temperature andthawing, which will provoke antibodies in rabbits. The antisera neutralisethe toxin in uivo and confer passive immunity when injected into animals.The toxin can be converted into a toxoid by the action of formaldehyde.Rabbits, actively immunised to H .pertussis toxin, are also protected againstthe toxins of H . parapertussis and Brucella bronchiseptica. H . pertzlssiaantitoxin neutralises all three toxins equally. E. W. Flosdorf, A. Bondi, andT. F. Dozois l7 report that H . pertussis in Phase I contains heat-stable andheat-labile toxins, which are also present in the other phases of H . pertussisand in H . parapertussis but in much smaller amounts. Polysaccharidefractions have been isolated from Br.bronchiseptica, H . pertwsis andH . purapertussis by extraction with hydrochloric acid or trichloroacetic acidor by tryptic digestion, followed by precipitation with alcohol ;IS that fromBr. bronchiseptica is toxic to mice and rabbits, but the other two are not.They are antigenic and confer protection when injected into mice. Crossprotection was found between the different strains.have elucidated the structure of thesoluble specific polysaccharide of Type 111 pneumococcus as being composedof glucose linked to C, of glucuronic acid and this by p-linkage to C , of asecond glucose molecule :R. E. Reeves and W. F. GoebelCO,HGroup specific proteins and type specific proteins and polysaccharides(the latter corresponding to those isolated by R.E. Hoffstadt and W. M.Clark 20 and by L. A. Julianelle and C. W. Wieghard have been isolatedby fractionation of the serological types of Staphylococcus aureus by W. F.Verwey .21The group specific polysaccharide of haemolytic streptococci has beenshown by C. A. Zittle and T. N. Harris 22 to have 1.72% of nitrogen, 0.70% of14 Science, 1941, 93, 237.l o J . Path. Bact., 1940, 51, 49 ; Lancet, 1942, 242, 529.1 7 J . Bact., 1942, 43, 265.18 G. Eldering, Amer. J . Hyg., 1942, 36, 294.2o J . Infect. Dia., 1935, 66, 250.21 Ibid., 1940, 71,636.J . Exp. Med., 1942, 75, 339.l9 J. BioZ. Chem., 1941, 139, 511.205 J . Ezp. Med., 1935, 62, 11, 31.2a J . Biol. Chm., 1942, 142, 823218 BIO~HXMISTBY.phosphorus, and [.ID -71.5' and to give positive orcinol and glucosaminetests.The heat-labile, L, and heat-stable, 8, antigens of vaccinia virus areprobably carried by a single protein molecule, since both are associated witha single electrophoretically homogeneous component, and are precipitated inequal titre by antibodies to either of them.The active parts are differentlyaffected by heat and by heat and alkali.% A nucleoprotein antigen has alsobeen found in alkaline extracts of the elementary bodies of vaccinia.%R. W. Linton 26 has given a good review of the chemistry and serology ofthe Vibrio cholerce group showing that the immunological grouping dependson the presence of specific protein and polysaccharide components. A.Damboviceanu and C.Barber 26 have described the " complete antigens "(O-antigens) of various types of V . cholera and P. B. White 27 has continuedhis work on the antigenic fractions of S, R and p forms of the organisms.A. J. WeilZ8 has reviewed the work on the Wassermann antigen andrelated " alcohol-soluble " antigens. H. Brown and J. A. Kolmer, 29 from astudy of the precipitate from " antigen " and syphilitic sera, came to theconclusion that the active substance is " an unknown, non-nitrogenous,phosphorus-containing substance " equally adsorbed on a mixture oflecithin and cephalin. J. Furth and E. A. Kabat3O showed that theWassermann hapten is associated with materials sedimented a t high speedin the ultra-centrifuge. M. C. Pangbornsl has isolated a phosphatide,cardiolipin, from ox heart, which, with lecithin and cholesterol, behaves a,sthe Wassermann antigen.Cardiolipin contains 4-11 yo of phosphorus and2.72% of sodium but no nitrogen, calcium, magnesium or sulphur. Theratio of Na to P is 1 : 1. On saponification it gives 62*2y0 of fatty acids and30-4y0 of a water-soluble fraction which consists of non-reducing carbohydrateand has 13.4% of phosphorus. There is apparently no glycerophosphoricacid present.Artijicial Antigens.-l?ollowing their work on the somatic antigens ofBact. dysenterh Shiga and Bact. typhosum, Morgan and Partridge 32coupled the conjugated protein fractions of these organisms with the non-antigenic polysaccharides agar, gum acacia and cherry gum by solution informamide and precipitation with alcohol, and obtained antigens whichprovoked antibodies in rabbits having sharp specificity for the polysaccharidesconcerned, Gum acacia gave no precipitate with cherry gum antiserum,but cherry &;urn gave weak precipitation with gum acacia antiserum.Kanten,gum tragacanth, hyaluronic acid, the specific polysaccharides of Types I*a J. E. Smadel and T. M. Rivers, J . Eq. Ned., 1942, 75, 161; T. Shedlovsky andJ. E. Smadel, W., p. 165.24 J. E. Smedel, T. M. Rivers, and C. L. Hoaglmd, Arch. Path., 1942, 34, 275.26 Bad. Rev., 1940, 4, 261.21 J . Path. Bact., 1940, 50, 160, 166; 61, 446, 447, 449.2* Bmt. Rev., 1941, 5, 293.29 J . Bhl. Chem., 1941,137,626.al Proc. Boc. Exp. Biol. Me&., 1941,48,484; J . Biol. Oh., 1942, 148, 247.s2 Chem.and I d . , 1941,60,722; Brit. J . Ewp. Path., 1942, %, 84.aa Compt. rend. SOC. Biol., 1940, 188, 601.ao Science, 1941, 94,46ANDERSON : ZMMUNOCHEMISTFLY. 219and I1 pneumococci and blood group A specific polysaccharide, isolated frommucin or peptone, gave no precipitates with the two gum antisera. Neitherglucuronic nor galacturonic acid inhibited the reaction between gum acaciaor cherry gum and the homologous antisera. W. T. J. Morgan 33 isolated anon-antigenic polysaccharide from commercial pepsin, peptone or gastricmucin, by extraction with 90% phenol and fractional precipitation, andshowed that it could be coupled to the conjugated protein of Bact. dyaenterice!Shiga to give a powerful antigen, which on intravenous injection into rabbitsgave rise to specific anti-" blood group A" agglutinins.W.F. Goebel 34 showed that artificial antigens containing hexuronicacids as determinant groups reacted with various antipneumococcal horsesera, but that their injection into animals conferred no protection againstpneumococci. The injection of an antigen made by coupling synthetiodiazotised p - aminobenzyl cellobiuronide (6- p-glucuronosidoglucose) withH OH CH,*OHCO,H H OHhorse serum globulin, however, led to the production of antibodies which gaveprecipitates with Type I11 polysaccharide, agglutinated with Type I11pneumococci, and conferred passive immunity on mice against virulentTypes 11, I11 and VIII pneumococci. The corresponding antigen containinggentiobiuronic acid (4- p-glucuronosidoglucose) gave rise to antisera whichH OHCO,H H OHhad no protective effect in mice against Types TI1 and VIII pneumococcibut conferred immunity against Type I1 organisms, Antigens containingcellobiose or gentiobiose instead of the corresponding aldobionic acids gaveantibodies devoid of protective effect against pneumococci. Antigens con-taining the specific polysaccharides gave rise to antibodies with sharp typespecificity, and antigens containing the unit pattern (aldobionic acid) yieldedantibodies with a wider specificity covering all those types (e.g., 11,111 andVIII) with specific polysaccharides built up from the same unit pattern.Artificial antigens containing sulphonamide~,~~ ~trychnine,~~ 1 : 2 : 6 : 6-dibenzanthracene and similar substances 37 and oxazolones 38 have been33 Chm.and Ind., 1941, 60,722.34 Nature, 1939, 145, 77; Science, 1940, 91, 20; J . E q . Med., 1940, 78, 33.35 A. G. Weddum, PTOC. Soc. Exp. Biol. Med., 1940, 45, 218.36 S. B. Hooker and W. C. Boyd, J. Immun., 1940, 38, 479.s' H. J. Creech md R. N. Jones, J . Amer. Chem. SOC., 1941,65, 1661, 1670.s* W. LettrB, K. Buoholz, and M. E. Fernhole, 2,phySiol. Ohm., 1941,887,108220 BIOOHEMISTRY.prepared. Hooker and Boyd 36 could not produce antibodies to an antigencontaining morphine.It has been shown by F. C. Bawden and A. Kleczkowski39 that suchantigens as tomato bushy stunt virus, or human serum globulin, when heatedwith non-specific proteins, such as rabbit serum albumin, form complexeswhich retain their original specific antigenicity but are no longer able toprecipitate with homologous antisera, behaving like haptens. Complementfixation occurs, showing that reaction between the complex and the antibodyhas taken place.Since peptides with molecular weights between 600 and 1000 consistingof 8 to 12 amino-acid residues, obtained by the hydrolysis of silk, inhibit theprecipitin reaction between silk fibroin and antibody, K.Landsteiner 40infers that silk fibroin contains determinant groups not larger than suchpeptides and suggests that similar relations probably hold for other proteinantigens .D. Lackmann, S. Mudd, M. G. Sevag, J. Smolens, and M. Wiener41report that the specific precipitation between the nucleic acids of yeast,thymus and streptococci and certain antibacterial sera, particularly horseantipneumococcus antisera, is inhibited by purine nucleotides, purinenucleosides and purine bases, weakly inhibited by pyrimidine bases and notat all inhibited by pentoses and phosphate.Antibodies.Production.-An excellent monograph by F.M. Burnet 42 on the produc-tion of antibodies, reviewing the known facts and suggesting possibilitiesappeared during 1942. L. Pauling43 in an extremely interesting papersuggests that all antibodies have the same polypeptide chains as normal serumglobulins but differ from them in the configuration or folding of the chain,particularly a t the ends. The modifications in configuration are introducedduring the synthesis of the globulin from amino-acids in presence of theantigen, which impresses the effect of its determinant groups on the stillmobile ends of the polypeptide globulin chain.Pauling suggests that theantibodies are bivalent in the sense that they have two reactive sites, one ateach end of the chain, and that this is in conformity with the “ lattice ”hypothesis of antigen-antibody reactions. He adduces a considerable bodyof evidence in support of his theory, ranging from combining ratios to theeffectiveness of antigens in provoking antibody production. As a result ofthis theory L. Pauling and D. H. Campbell 44 were led to the prediction that,if globulin were placed under mild denaturing conditions, such as heating at50-60°, or solution in urea or alkali, and the condition then removed in the39 Nature, 1941, 149, 593; Brit.J . Exp. Path., 1941, 22, 208; 1942, 23, 169.40 J . Exp. Med., 1942, 75, 269.42 “ The Production of Antibodies,” Monographs from the Walter and Eliza HallMacmillan and Co. Ltd.,41 J . Immun., 1941,40, 1.Institute of Research in Pathology and Medicine, No. 1.Melbourne, 1942.Is J . Amer. Chem. SOC., 1940, 62, 2643.44 Science, 1942, 95, 440; J . Exp, Med., 1942, 76, 211ANDERSON : IMMUNOUHEMISTRY. 221presence of an antigen, the polypeptide chains would unfold and then refoldin a manner complementary to the antigen. By such treatment theysucceeded in producing in vitro antibodies to 1 : S-dihydroxybenzene-2 : 4 : 6-tri-p-azophenylarsonic acid, methyl-blue and the specific poly-saccharide of Type I11 pneumococcus.P.R. Cannon 45 has pointed out that antibody production depends on thesame factors as globulin synthesis and that a good response can only beexpected if the amino-acid or protein intake of the animal is adequate.Lack of protein tends to lowered resistance to infection. R. Schoenheimeret aZ.,46 by feeding amino-acids containing isotopic nitrogen to rats, haveshown that serum globulin and antibodies are concerned in metabolicprocesses involving dietary nitrogen and that the half-life period of an anti-body molecule is about two weeks. Antibodies introduced by passiveimmunisation do not undergo such changes, suggesting that antibodies arestill being produced during active immunisation, even when the amount incirculation is diminishing, until an equilibrium is reached.Passively intro-duced antibody, on the other hand, disappears more rapidly and is notreplaced by synthesis,Puri$cation.-Further work on the purification of antibodies by theaction of proteolytic enzymes has been reported. F. Modern and G. Ruff 47increased the purity of antitetanus serum from 7000 units per gram of proteinin the original serum to 22,500 units per gram of protein by the consecutiveaction of pepsin and papain and precipitation with 22% sodium sulphatesolution, but with a loss of about 50% of the antitoxin. A. L. Petermannand A. M. Pappenheimer48 digested the euglobulin fractions of horseantisera to Types I and I1 pneumococci with pepsin and obtained antibodywhich was homogeneous in the ultra- centrifuge, with a molecular weight lessthan 100,000;.the purified antibody combined with twice as much specificpolysaccharide per mg. of nitrogen as did the original antibody, and wascomparable with normal rabbit antipneumococcus antibody in size andcombination with polysaccharide. M. L. Petermann 49 split horse diphtheriaantitoxin, by the action of papain, into equal fragments (as determined by theultra-centrifuge), only one of which flocculated with toxin ; it was soluble in5% sodium chloride solution a t 58" and pH 4.2. These " half " moleculescan be further broken down to " quarter " molecules. J. H. Northrop 50obtained crystalline diphtheria antitoxin by the digestion of toxin-antitoxinfloccules with trypsin a t pH 3.7 and fractionation with ammonium sulphate.The product was homogeneous in the ultra-centrifuge and electrophoresisapparatus, had a molecular weight about 80,000, and contained 700,000 tolo6 units per gram of nitrogen.By dialysis of rabbit antipneumococcus sera S.Rafael, C. F. Pait, and46 J . Immun., 1942, 44, 107.46 R. Schoenheimer, S. Rrttner, D. Rittenberg, and M. Heidelberger, J . Biol. Chem.,47 Compt. rend. SOC. Biol., 1940, 133, 158.1942, 144, 641, 545.4a Science, 1941, 93, 458.J . Biol. Chem., 1942, 144, 607. 60 Science, 1941, 98, 92222 BIO-MISTRY.M. C. Terry showed that the antibody is associated with pseudo- andeu-globulins in different proportions depending on the intensity of immuni-sation. The less avid antibody produced in the early stages of immunieationis associated with the euglobulin.With progressive immunisation theantibody becomes more avid and more of it is associated with the pseudo-globulin and less with the euglobulin. The ratio of pseudoglobulin to totalglobulin increases with prolonged immunisation. R. A. Kekwick and B. R.by electrophoretic analysis, also found that the distribution ofdiphtheria antitoxin between p- and y-globulin altered with the progress ofimrnunisation. The first response is the production of antibody associatedwith y-globulin, which combines more rapidly with toxin but affords a some-what unstable complex expressed by the formula TA,. With furtherinjections the amount of y-globulin antitoxin remains about constant butp-globulin antitoxin is formed ; this has a longer flocculating time and gives amore stable complex having the formula TA,.A.Kleczkowski53 and F. C. Bawden and A. KleczkowskiW offer anexplanation of the apparently greater heat stability of H-type antibodies ascompared with O-antibodies. They state that the antibodies are equallysusceptible to heat but that the antibodies to O-antigens form complexes withnon-specific proteins (e.g., albumin) more readily than do antibodies toH-antigens ; the complexes combine with antigen, as shown by complementfixation experimenta, but do not precipitate (compare the similar findingswith heated protein anfigen~).~gAntigen-Antibody Reactions.Interest in this field has mainly centred on the “ valence ” of antigens andantibodies and the development of theories describing quantitatively thereactions between antigens and antibodies.S. B.Hooker and W. C . Boyd s5s 66 by using a number of haptens of knownconstitution, tested the prediction from the “ lattice ” or ‘‘ alternation ”hypothesis that bivalent haptens and the corresponding antibodies shouldform precipitates. They came to the conclusion that the “ lattice ” theory isprobably not true, since precipitation did not occur even with tervaJenthaptens (that is, haptens containing three active groups ; for example,R OHH O C ) R , where R is -N=N(I)As03H2 orR-OH/-\N=N( )OH-lN=N(-)CO2H) or even in one case with auhapten, although combination between the hapten and antibody had occurred,since inhibition of subsequent precipitation with the homologous antigen wassexavalentJ .Immun., 1940, 39, 317, 337, 349. 6 2 Brit. J . Exp. Path., 1941, 22, 29.64 Ibid., 1942, B, 178.66 W. C. Boyd, J . Exp. Mad., 1942, 75,407.53 lbid., p. 188.55 J. Imun., 1941, 42,419ANDERSON : IMXUNOOHEMISTBY. 223observed. It is suggested that, if the haptens carry many polar or " solu-bilising " groups, combination with the antibody is not adequate to renderthe complex insoluble. This is borne out by the fact that acetylation orbenzoylation of such polar groups converts non-precipitable haptens intoprecipitable ones. Prom this basis Boyd 56 developed his " occlusion "theory of precipitin reactions, according to which precipitation depends onthe reduction of the solubility of the complex below the point a t whichit can remain in solution by mutual neutralisation of the polar groups ofthe antibody and antigen or hapten and by the blocking off, or occlusion, ofthe polar groups of closely neighbouring antibody molecules.If the activegroups of a hapten molecule are close together, there may not be room fortwo or more molecules of antibody (steric hindrance), 80 their polar groupscannot be occluded and precipitation does not occur (compare the examplequoted above). If, however, the hapten molecule is larger, for example whenR of the above formula is -N=N<>N=N<>AaO,H,, the polargroups are further apart, more antibody molecules may be able to react,and a larger number of polar groups can be occluded with consequentprecipitation.The deoisive factors seem to be the number of polar groupsof the antigen or hapten left free after combination with the antibody, andthe distance separating the active groups, which determines the amount ofsteric hindrance exerted by one antibody molecule on another.By immunising rabbits with arsenil-sheep globulin, containing the twodeterminant factors phenylarsonic acid and the species speoific sheep protein,and fractional precipitation of the antisera, F. Haurowitz and P. Schwerin 57concluded that the antibody molecules are mainly univalent, that is, eachcorresponds to one determinant group only. The antigen-antibody precipi-tate is considered to consist of aggregates of complex particles containinga multivalent antigen molecule to which are attached several univalentantibody molecules.S. B. Hooker and W. C. Boyd 58 review the evidenceon the " valence " (in the sense of the number of active groups) of antigensand antibodies and conclude that antigenic proteins, having a molecularweight of about 35,000, seem to have a minimal functional valence of five withan upper limit less than thirty. Larger antigenic molecules probably have ahigher valence, which is likely to be proportional to their surface area.Although the evidence in the case of antibodies is still inadequate, the bulk ofit favours the view that they are univalent or, less probably, bivalent.A. D. Hershey 59 has developed a quantitative theory of the antigen-antibody reaction based on the " lattice " theory and the assumption ofmultivalent antigen and antibody molecules.The dissociation of thecomplexes being taken into account, the quantitative relations o thecomposition of the precipitate, velocity of flocculation and optimal propor-tions are developed. The valence of antibody is considered to be not morethan 2.67 B d . J . Exp. Path., 1942,aS, 146. ba J . I m w . , 1942, 46,127.In Ibi& 1941,&, 456, 486, 615; 1942,&, 30224 BIOCHEMISTRY.The solubility of antigen-antibody precipitates in excess of antigen isconsidered by W. C. who also demonstrated 61 that most antiseracan be grouped into (a) an H type comprising horse sera and some rabbitsera, which give optima by both " constant antibody " and " constantantigen " flocculation procedures, and (b) an R type, to which most rabbitsera belong, which give optima only by the " constant antibody " method.The differences are accounted for by the chemical and physical properties ofthe antisera.The velocity of combination of the pneumococcus poly-saccharides with antibody is dealt with by M. Mayer and M. Heidelberger,62and S. C. Liu and H. Wu 63 describe the acid and alkaline dissociation of suchprecipitates.Very interesting demonstrations of the combination of antigen andantibody have been given by the use of the electron-microscope, for example,in the case of tobacco mosaic virus and its antiserum.64Toxins.A number of papers dealing with the purification and chemistry of toxinshave appeared.A. M. Pappenheimer 65 has dealt with diphtheria toxin,which has the properties of a protein. W. McD. Hammons6 found thatstaphylococcus enterotoxin behaved as a large complex carbohydratemolecule, and I. A. Parfentjew, F. L. Clapp, and A. Waldschmidt 67 showedthat staphylococcus toxin can be toxoided, without loss of antigenicity orspecificity, by peptic digestion a t pH 4-6-5.2. Scarlatina1 toxin has theproperties of a protein of low molecular weight and is probably not conjugated,according to E. S. G. Barron, G. F. Dick, and C. M. Lyman.68 W. L.Koerbner and W. E. B ~ n n e y , ~ ~ on the other hand, claim to have isolated aprotein-free toxin. G. A. Hottle and A. M. Pappenheimer 70 and A. H.Stock 7 l and his co-workers 72 describe the toxin as a protein which is relativelyresistant to pepsin, papain and trypsin and is heat-coagulable. The oxygen-labile hzcmolysin, streptolysin-0, appears to be a protein whose activitydepends on the presence of SH groups, which become reversibly convertedinto the dithio-group, S-S, upon inactivation of the toxin by oxygen, accordingt o C.V. Smythe and T. N. Harris 73 and to E. W. Todd.74 The latterworker 7 4 n 7 5 has shown that streptolysin-0 is neutralised by antisera toCZ. welchii 8-toxin, which is also oxygen-labile, but not by antisera to thestable a-toxin. 8-Toxin is neutralised by high titre streptolysin-0 antisera.All the known oxygen-labile hamolysins which are reactivated on reduction61 J . Exp. Med., 1941, 74, 369.63 Proc.SOC. Exp. Biol. Med., 1940, 43, 747.6 6 Amer. J . Publ. Health, 1941, 31, 1191.70 J. Exp. Med., 1941, 74, 545.6o J . Immun., 1940, 38, 143.a2 J . Biol. Chern., 1942,' 143, 567.64 T. F. Anderson and W. M. Stanley, J . Biol. Chem., 1941,139,339.6 5 J . Bact., 1942, 43, 273.6 7 J . Immun., 1941, 40, 189.69 J . Immun., 1941, 40, 459.7 1 J . Biol. Chem., 1942, 142, 777.c2 L. E. Krejci, A. H. Stock, E. B. Sanigar, and E. 0. Kraemer, ibid., p. 785.73 J . Immun., 1940, 38, 283.7 6 Brit. J . Exp. Path., 1941, 22, 172.J . Biol. Chem., 1941, 187, 267.Biochem. J., 1941, 3!j, 1124ANDERSON : IMMUNOCHEMISTRY, 226(from streptococci, pneumococci, CZ. tetuni and CZ. wekhii) are closely relatedserologically but are not identical. The hzemolysins of staphylococci, Cl.septicum and CZ.cedematiens, which are irreversibly destroyed by oxygen,are not neutralised by antitoxins to toxins of the first group.The opalescence produced in normal human sera by the a-toxin ofCl. welchii (the Nagler phenomenon) was shown by R. G. Macfarlane, C. L.Oakley, and C. G. Anderson 76 to be due to the liberation of free fat and torequire the presence of calcium ions. A similar but more rapid and moresensitive liberation of fat is produced by the action of a-toxin on lecitho-vitellin. It has beenshown by M. G. Macfarlane and B. C. J. G. Knight 77 that the a-toxin acts as alecithinase which causes splitting of lecithin to phosphorylcholine and adiglyceride. It is suggested that the lecithin acts as a, stabiliser in the serumor lecitho-vitellin and that the fat can aggregate when the lecithin isdestroyed.E. M. Crook 78 showed that the Nagler reaction was given byfowl serum but not by horse, pig, sheep or rabbit sera. Cl. septitum, CZ.histolyticurn, Cl. tetani, and C1. botulinum do not give active toxins, whereasCl. ademutiens, C1. sordellii, Cl. chauvmi, Cl. sprogenes, Cl. centrosporogenes,CZ. tertium, and C1. bifermentans produce more or less reaction. The reactionswere inhibited only by the homologous antisera. Psezcdomonas pyocyaneaalso gave a reaction with lecitho-vitellin, which was inhibited by Cl. welchiiantitoxin. W. E. van Heyningen79 and E. F. Gale and W. E. vanHeyningen 80 have described the preparation and partial purification of thea-toxin of CZ.welchii.Complement.In a series of papers Ecker and Pillemer and their co-workers 81-93 havedescribed the preparation and properties of the various components of comple-ment and elucidated their r61e in the process of specific complement fixation,To summarise briefly, they found that a t 1" components C1 (mid-piece,insoluble in CO,; may contain also C3 and C4), C2 (end-piece, soluble inCO,; may contain also C3 and C4) and C4 (fourth component, inactivatedby ammonia) combine with sensitised sheep red blood corpuscles, but C3The reaction can be used for the titration of a-toxin.76 J . Path. Bact., 1941, 52, 99.i 8 Brit. J . Exp. Path., 1942, 23, 37.8o Ibid., 1942, 36, 624.81 E. E. Ecker, L. Pillemer, C. B. Jones, and S.Seifter, J. Biol. Chem., 1940, 135, 347.82 E. E. Ecker, L. Pillemer, and A. 0. Kuehn, R o c . SOC. Ezp. Biol. Med., 1940,45,115.83 L. Pillemer, S. Seifter, and E. E. Ecker, ibid., p. 130.84 L. Pillemer and E. E. Ecker, J . Biol. Chem., 1941, 137, 139.85 Idem, Science, 1941,94,437.88 E. E. Ecker and L. Pillemer, J. Immun., 1941, 40, 73.E. E. Ecker, C. B. Jones, and A. 0. Kuehn, ibid., p. 81.8 8 L. Pillemer, S. Seifter, and E. E. Ecker, ibid., pp. 89, 97.89 L. Pillemer and E. E. Ecker, ibid., p. 101.9 1 L. Pillemer, S. Seiffer, and E. E. Ecker, ibid., 1942, 75,421.9a L. Pillemer, S. Seifter, F. Chu, and E. E. Ecker, ibid., 1942, 76, 93.7 7 Biochem. J., 1941, 35, 884.79 Biochem. J., 1941, 35, 1246, 1257.L. Pillemer, E. E. Ecker, J. L. Oncley, andE.J. Cohn, J. Exp. Med., 1941,74,297.L. Pillemer, F. Chu, S. Seifter and E. E. Ecker, J . Immun., 1942, 46, 51220 BIOUHEMISTRY.(third component, inactivated by zymin or by an insoluble carbohydratefrom yeast 8Q) does not. C1, although combining with sensitised cells in theabsence of C4, is hamolytically inert unless C4 combines previously orsimultaneously. C4 does not combine in the absence of C1. Although C3is not fixed by antibody-red-cell aggregates, it is essential for haemolysis,acting on the sensitised cells after fixation of C1, C2 and C4 and behaving asthough it were a catalyst or enzyme.M. Heidelbergerg4 developed a method of estimating the amount ofcomplement fixed by antigen-antibody complexes, which was used byPillemer and co-workersB3 to estimate the amounts of the various com-ponents fixed.J.Gordon and W. R. Atkin 96 showed that the inhibition of complementaction by sodium hexemetaphosphate (calgon) was due to combination with3erum protein and not to removal of calcium.C. G. A.3. PROTEOLYTIC ENZYMES.Recent advance in our understanding of proteolytic enzymes hasdepended primarily on the work of the Northrop schoo1,l who purified andcrystallised the protein-digesting enzymes of the stomach and pancreas.Pepsin, trypsin, chymotrypsin and their inactive precursors have all beenprepared in this way, usually from more than one species (pig, ox, sheep,chicken, and salmon2), and a beginning made in the study of pepsin asregards chemical composition, the nature of the groups connected with theiractivities, and the changes which occur when the inactive zymogen isconverted autocatalytically into active pepsin.The other emymes, like themore recently crystallised pancreatic carboxypeptidase of M. L. Anson,s haveso far received scant chemical attention, and little work has been done onspecies differences in composition, though it has been shown that swine andsalmon pepsin have different tyrosine contents.l* The position is complicatedby the recent discovery4 that crystalline swine pepsin, originally believedhomogeneous, can be fractionated into a t least two components of differentproteolytic activity.Enquiries into the presence of any non-protein prosthetic group particu-larly responsible for the enzymatic activity have so far yielded negativeresults : this stands in contrast to work on intestinal peptidases, which whenpartially purified are found to require manganous, cobalt, zinc or magnesiumions for a~tivation.~ In the case of pepsin, the early work of R.Herriottand J. H. Northrop,l studying the effect of acetylation with keten, and iodin-s4 Science, 1940, 92, 684. O6 Brit. J. Exp. Path., 1941, 22, 226.1 J. H. Northrop, “ Ckysttllline Enzymes,” Columbia University Pmm, 1937.3 E. R. Norris and D. W. Elm, J. Biol. Chem., 1940,134,443.4 R. Herriott, V. Desreux, and J. H. Northrop, J . Gem. Pkysiol., 1939,28,439; 1940,6 Review by M. J. Johnson and J. Berger, dduancee in Entymology, 1942, 2.69.Ergebn. Enqpqforsch., 1938, 7, 118.24,213CJRANMER AND NEUBERQER : PROTEOLYTICI ENZYMES.227rttion, on the autivity, pointed to the phenolic hydroxyl of the tyroeineresidues (of which there is 12% lin mammalian pepsin), togethsr with themany fiee carboxyl groups of the protein, as primarily responsible. J. St. C.Philpot and A. Small,6 investigating the iodination of tyrosine in more detail,later claimed that some unknown reactive group must be present, sinceiodination and inactivation proceed fmter than the deorease in reactivitywith Folin phenol raagent, which marks the formation of di-iodotyrosylgroups. Recently, R. M. Herriott showed that this result was due to theintermediate formation of monoiodotyrosine, which has a stronger reactionwith the Folin reagent than tyrosine itself.The speoscity of theee enzymes hrts been investigated by Bergmann andhis collaborators, using a wider rmge of synthetic substrates than previoualyavailable, prepared by his own ca,rbobenzyloxy-method.* He began byshowing that, like the peptidases, the proteinases also would attack simplepeptides, but always provided that the terminal amino- andlor oarboxylgroups of the peptide chain were blocked.Trgpsin, for example, splitsbexlzoylglycyl-lysineamide, but not glycinelysinectmide or benzoglglycyl-lyaine ; pepsin splite carbobenzyloxyglutamyltyro&ne but not glufaJlayl-tyrosine. Bergmann has explained this in terms of the undesirability of apositive or negative charge on the substrate close to the link to be broken,but it must be noted that carbobenzyloxylation introduces another peptidebond, and it seem8 possible that the enzyme requires a t least one otherpeptide bond to combine with its substrate, in addition t o that actuallybroken. This is made probable by the fact that pepsin also splits glycyl-glutaznyltyrosine a t a reasonable rete.Conversion of the carboxyl Q~OUPSinto amid0 makeg the substrate resistant to peptic hydrolysis, but theoonclwion that a negative (carboxylate ion) charge adjacent to the attaokedlink is essential is unwarranted, since ~rbobe~y~oxyglu~myltyro~ylgly oineis also split, though rather Dore slowly. On the ofher hand, iodination ofthe tyrosine, and probably conversion of the y-carboxyl into amide, makesthe substrate completely resistant to P8pSin. If tyrosine is replaced byphenylalanine in the peptide, the bond is still broken, but the rate is slower,which may mean that the enzyme substrate affinity is smaller.9 Tryptophan-oonta,ining peptide8 hawe not been studied.In a, similar way, alterations in the typical subsfrates for other enzymeschange their digestibility.Trypain, for example, does not &ttack benzoyl-glycyl-lysineamide if the c-amino-group of the lysine is carbobenzyloxylated :quite apart from charge effects, steric faotors may also enter here. Thatspatirtl disposition is important is shown by the faot that if any of the amino-acid8 in the peptide are not of the bvo-configuration, the peptide becomesresistant to attack. Bergmsnn lo explains this as meaning theLt the enre;gmemust actually attach itself to the peptide bond under attack, and the fact6 Proo.Roy. SOC., 1939, A, 170,62.8 Reviews by M. Bergmann, Advances in Enzymology, 1941,1,63; 1942,2,49.* J. S. Fruton and M. Bergmann, J. BWZ. Chm., 1939, W, 627.10 M. Bergmasln ct d., ibid., 1836,109,826; 1987,117, 189.J . &n. PhysioZ., 1941, 26, 185.RlEP.-VOL. XXXIX. 228 BIOUHEMISTRY.that methylation of the imino-group of the bond again prevents its breakageperhaps offers some support for this concept.Proteinases, then, attack even so small a protein fragment as a tripeptide,provided it contains the particular amino-acids to which the proteinase isadapted : pepsin must have an aromatic, and preferably a dicarboxylicamino-acid adjacent ; chymotrypsin, a pancreatic enzyme with optimumpH 7.6, also demands an aromatic amino-acid component, but splits thebond on the other side of it, i.e., bond 2 inCarbobenzyloxyglu tamyll ltyrosyll 2glycineamidewhere pepsin splits bond 1 of the same substrate.ll This specificity is tosome extent bound up with the charge character of the substrate, and thishelps to explain the existence of a pH optimum.The fact that the optimumfor pepsin with protein is in the neighbourhood of 2.5 may perhaps bereconciled with the value of 4.1 obtained with synthetic substrates by adifference of surface pH and bulk pH.12With the knowledge gained from the study of pure enzymes, theBergmann school have turned to the investigation of the enzyme mixturesof papaya latex, and later, with marked success, to the intracellularproteinases and peptidases of spleen and kidney.By comparing the activitytowards various synthetic substrates of preparations after various activationtreatments, with cyanide, hydrogen sulphide, etc., it has been possible toidentify in beef spleen, a pepsinase, a trypsinase, a carboxypeptidase, and aleucine aminopeptidase, which are different from, but similar in hydrolyticspecificity to, the digestive enzymes after which they are named, and to thesimilar enzymes present in kidney of ox and pig.13 On the basis of resultsobtained with papain preparations, two of which have been in crystallineform,l4 Bergmann has proposed that these intracellular proteinases, formerlycollectively termed cathepsin, play a synthetic rdle in vivo.He finds thatglycyl-leucine is not attacked by cystine-activated papain, unless acetyl-phenylalanylglycine is simultaneously present, as a ‘ I co-substrate.” Thesequence of reactions is first synthesis of acetylphenylalanyl-glycyl-leucine ;from this leucine is split off, then glycine is liberated, and so the co-substrateis regenerated and the peptide split. This means that an enzyme will notdisplay its true specificity in presence of a peptide mixture, for example, apartial protein hydrolysate, and also shows that synthesis, like hydrolysis, isspecific and dependent on the peptide mixture with which the enzyme has todeal : it suggests in fact a way in which foreign proteins (viruses, antigens)may deflect the normal course of protein synthesis.It should be noted inthis connection that the equilibrium of these enzymatic reactions is very faron the side of hydrolysis, so that unless the product is removed from thel1 M. Bergmann and J. S. Fruton, {bid., 1937, 118, 405.l2 G. S. Hartley and J. W. Roe, Trans. Paraday SOC., 1940, 36, 106; J. L. Danielli,l3 See Advances in Enzymology, 1942, 2, 57.14 A. K. Balls and H, Lineweaver, J . Bid. Chem., 1939, lW, 669,Biochem. J., 1941, 35,470NORRIS: SOl\IE PLANT PRODUUTS AND ENZYMES, 229system, or energy is supplied in some as yet unknown way, any measurabledegree of synthesis of peptides by these proteinases is exceedingly improbable.In the short space available it has been impossible to include any separatediscussion of peptidases (see ref.5 ) , which in any case are still in a moreundeveloped state, nor can we do more here than refer to the meagrephysicochemical work as yet done on proteolytic enzpes.15J. L. C.A. N..4. SOME PLANT PRODUCTS AND ENZYMES.Owing to wartime preoccupations the output of published papers showsa continued decline ; nevertheless, although there appears to be no outstandingachievement during the year under review, a steady advance is being main-tained. The subjects dealt with in this Section are similar to those in previousyears. The problem of starch is still engaging the attention of many workers ;knowledge of growth factors for both higher and lower plants accumulates ;improvements in the technique of protein investigation continue to be made.Of reviews on the subject of the chemistry of plant products which haveappeared recently, two may be mentioned here.A review of the lipoidconstituents of algs, with special reference to the carotenoids and sterols, isgiven by I. M. Heilbron,l and an account of recent progress in the chemistryof pectic materials and plant gums is contributed by E. L. Hirst.2 It mayalso be mentioned that an account of polysaccharides, particularly in theirstructural aspects, was given by S. Peat in last year’s Annual Report.3 Someof the subject matter of the review by Hirst and the Report by Peat is againreferred to briefly in this Report in order to maintain some slight degree ofcontinuity.Growth ,Substances.-The close association of auxin with leaf proteins isindicated by experiments of S.G. Wildman and S. A. G ~ r d o n , ~ who haveisolated the leaf proteins of spinach and obtained from them by the action ofproteolytic enzymes a separation of auxin. The most active enzyme in thisrespect was the mixture contained in a commercial tryptic extract, but otherproteolytic enzymes were also effective, including trypsin and papain.Auxin was obtained in this manner from both cytoplasmic and chloroplasticproteins, and in the former case two fractions of protein were prepared by anisoelectric method, auxin being obtained from each fraction, but in differingamounts. The authors concluded from diffusion rates that the leaf auxinhad a lower molecular weight than indolylacetic acid.Wide variations in the auxin content of some alga were discovered byJ.van O~erbeek.~ The auxin present appears to be indolylacetic acid or oneof its homologues; auxin-a or -b was apparently not present. The highestauxin concentration was found in the young blades. The same author hagstudied the occurrence of auxin and its precursor in coleoptiles of Zea andl J . , 1942, 79. Ibid., p. 70. Ann. Reports, 1941, 38, 150.4 Proc. Nat. Acad. Sci., 1942, 28, 217.1 5 J. A. V. Butler, J . Amer. Chern, SOC., 1941, 63, 2968, 2971.Plant Physiol., 1940, 16, 291230 BIOOHEMISTRY.Avena seedling0 He finds that larger amounte of auxin can be removedfrom the coleoptile tips, wherein it is formed, by diffusion into agar than byether extraction. After all the auxin obtainable by diffusion has passed out,there still remains a residue extractable by ether.A differentiation is madebetween “active ” auxin extractable by ether, and “ potential ” auxin,obtained by diffusion plus that obtained by ether extraction, and the differencebetween these two represents the auxin precursor. It is considered that thetemporarily delayed growth observable on decapitation of the growing tip isdue not to lack of precursor of the auxin, but to lack of activation of suchprecursor. This activation occurs a t the apical surface and is delayed bydamage to the apical cells.A new method of preparation of pure biotin has been described by D. B.Melville, K. Hofmann, E. Hague, and V. du Vigneaud,’ who employ a biotinconcentrate prepered oommercitllly from milk.The crude biotin obtainedfrom this source was purified by ohromatographic adsorption applied to themethyl eater, the adsorbents employed being Deoalso and activated alumina,From the purified methyl ester, the crystalline biotin was obtained aftersaponification. The method should be applicable to the preparation ofrelatively large amounts of biotin, Additional proof that biotin is a cyclicurea derivative is adduced by D. B. Melville, K. Hofmann, and V. du Vigneaud,8p who find that the diamino- carboxylic acid C,Hr802N2S preparedfrom biotin is reconverted into biotin by the action of oarbonyl chloride. Ina further contribution to the study of the structure of biotin, the sameauthors 10 describe a number of derivatives of biotin and suggest severalpossible struotures ; but finality has not been reached.A microbiological method for assay of biotin is described by G.M. Schull,€3. L. Hutchings, and W. H. Peterson.ll The organism employed is Lacto-bacdhs casei e and the method utilises the principle that under properlystandardised conditions the titratable acidity produced by the organism is afunction of the quantity of biotin present in the medium. Advantagesclaimed for the method are that it is independent of the effects of colour orturbidity in the solutions employed, that its accuracy is within lo%, andthat, since the same organism is employed for assay of pantothenic acid 12and riboflavin,13 it is not necessary to maintain separate stock cultures.The biotin contents of a number of widely differing animal and plant sourceshave been determined by the method, and the results tabulated,Further information is now available with regard to the nature of anantibiotin factor first noticed by R.E. Eakin, E. E. Snell, and R. J. Wil-liams,1*, 15 who named their material avidin. This substance occurs in egg-white, and active concentrates were prepared and shown to combine stoi-Q Amer. J . Bot., 1941, 28, 1.a Science, 1941, 94, 308.10 J . A m r . Chem. SOL, 1942, 64, 188.12 D. Pennington, E. E. Snell, and R. J. Williams, ibid., 1940, 135, 213.13 E. E. Snell and F. M. Strong, Ifid. Eng. C’hem. (AnaE.), 1939, 11, 346.l4 J . Biot. Chem., 1940,186, 801.J . Biol. Ohem., 1942, 142, 616.Ann. Reports, 1941, 88, 250.l1 J .Bid. Chem., 1942, 142,9131&id., 1941, 140, 635NORRIS: SOME pUS!C PRODUCPTS BND ENZYMES. 231oheiometrically with biotin, rendering it unavailable to yeast. D. W. Wooleyand L. G. Long~iworth~~ had also noticed that eggwhite rendered biotinineffeotive for Cbstridizlm bwtgZiczlm, and commenced work on the oonoen-tration and attempted isolation of the substance responsible. The work ofthe last-named confirms that of Eakin et al. and extends the latter's obaerv-ations. They were able to prepare substance fkom egg-white which wag15,000 times more effective against biotin than the original egg-white. Thisproduct appears to be a basic protein of isoelectric point pH 10; it was notobvioualy crystalline but was homogeneous under electrophoresis and in theultra-centrifuge.It is conoluded that it is a pure substance and a molecularweight of 70,000 is ascribed to it, although the true figure may be slightlylower.Additions to the group of bios substances may stiU be expeoted. VitaminB, is generally recognised as a constituent necessary for many micro-organismsand vitamin B, also has been added to the list. C. Marchant 1' finds thatSaccharomyces hanseniaspora ualbyensis requires the known bios constituents,including vitamin B,, which can be supplied in plaoe of a " bios VII " solu-tion. In the case of S. gaEactosus, vitamin BB cannot replace the " bios VII "solution, and it is conoluded that this must contain an additional unknownfactor or factors.In last year's Report1* the use of Proteus morgunii as an organism suitablefor assay of pantothenic acid was described.In 8 more recent paper 1@ anenquiry into the part played by pantothenic acid in the metabolism of theorganism leads to the suggestion that the acid is concerned in some mannerwith the metabolkm of pyruvic acid. The mechanisms involved are a tpresent unknown, but the evidence largely points to the agency of panto-thenic acid in converting pyruvic acid into acetio acid. It may be, however,that pantothenic acid is concerned in some other stage involving pyruvicacid or some intermediate derived from pyruvic acid.In the search for rtnaloguea of pantothenic acid which might have bio-logical acitivity of the kind amociated with the acid, J. W. Barnett and F. A.Robinson 2O have prepared a number of products, including two new lactones,by condensing P-alanine with five Werent lactones, and by condensinga-hydroxy- PP-dimethylbutyrolactone with four different amino-acids.Noneof the substances synthesised could replace pantothenic acid as a growthstimulator. It would Beem that, if either the a-hydroxy-group is lost, OF thecarbon chain lengthened, the resulting produot has no biological activity.On the other hand, the same authors found21 that certain analogues ofpantothenic acid had an inhibitory effect on Strelptococczls hmolyticw, andCorynebacterium diphtheria. Such analogues were obtained by conden-&ion of cc-hydroxy- pp-dimethylbutyrolactone with taurine and taurineamide and by condensation of P-hydroxy-yy-dimethylvalerolactone andl6 J .Boil. Chem., 1042, 142, 285.1 8 Ann. Reports, 1941, 38, 251.19 A Dorfman, S. Berkman, and S. A. Kosor, J . Biol. Chem., 1942, 142, 393.Biochem. J., 1942, 36, 367.1' Canadian J . Res., 1942, 20, B, 21.21 Ibid., p. 364232 BIOCHEMISTRY.taurine. In these cases the inhibition was reversed by pantothenic acid;but with other analogues the inhibition was not reversible. In all thecompounds dealt with in this manner, the a-hydroxy-group of pantothenicacid is absent.Xturch and Amylases.-Differences between the enzymically synthesisedstarch of C. S. Hanes,22 and naturally occurring starch appear to be reflectedin their constitutions. W. Z. Hassid and R. M. McCready 23 have approachedthe problem on recognised lines, and submitted the synthetic starch tomethylation and hydrolysis.They were unable to isolate any tetramethylglucose, the sole product appearing to consist of 2 : 3 : 6-trimethyl glucose.In the apparent absence of an end-group they suggest that the chains mayexist as continuous loops of glucopyranose residues. It is possible, however,that the small quantity of synthetic starch available to these workers renderedthe isolation of the small amount of tetramethyl glucose impossible, since aninvestigation of the enzymically synthesised starch by W. N. Haworth,R. L. Heath, and S. Peat 24 has shown that end-group assay is possible andthat on methylation the starch gives a product, containing 44.5% of methoxyl,which resembles in general characters the product obtained similarly frompotato starch, amylose, and amylopectin.They conclude that the unitchain-length is not less than 80-90 glucose units, the linkages being of thenormal 1 : 4-~-glucosidic type. The length of this chain distinguishes thesynthetic from the natural starch, and the length estimated in this manneragrees with that suggested by Hanes on the basis of copper-reducing power.The enzyme concerned in the production of synthetic starch, the phos-phorylase of potato, has been examined and purified by D. E. Green andP. K. S t ~ m p f , ~ ~ who, by successive ammonium sulphate fractionations, haveobtained a preparation some 370 times more concentrated than the original.They find that adenylic acid is not a component of the system, as is the casewith animal phosphorylase, and that catalytic amounts of starch, dextrin orglycogen are required for starch formation from glucose l-phosphate.Noinhibition was observed when heavy metals, oxidising or reducing agentswere present.Although our knowledge of the structure of starch has been placed on amuch firmer basis in recent years, none would suggest that the last word onthe subject has yet been written. The presence of linkages other than thecommonly accepted 1 : 4-glucosidic linkage is frequently suggested in theliterature, and the problem is discussed by R. W. Ken and N. 3'. Schink,26who have carried out experiments with particular reference to fermentabilityof syrups prepared by diastatic action on maize starch.In discussingstructural problems, these authors stress especially the non-homogeneousnature of starch and claim the existence of at least two fundamentallydifferent configurations in maize starch, probably only one of which is com-posed of the usual 1 : 4-glucoside or maltose type of linkage. A hint of other22 Ann. Reports, 1940, 37, 419.28 J . Amer. Chem. Soc., 1941, 63, 2171.z 6 J . Biol. Ohem., 1942, 142, 366.24 J., 1942, 65.26 Ind. Eng. Chem., 1941, 88,1418NORRIS : SOME PLANT PRODUOTS AND ENZYMES. 233possible linkages is contained in the claim of Y. Nakamura 27 to have isolateda new disaccharide, “ amylolyose,” by the action of diastase on starch. Thenew sugar contains 1 : 5-linkages and it is suggested that, although thepreponderant linkage between the glucose units in starch may be thenormal 1 : 4 type, 1 : 5- and 1 : 3-linkages may also be present.Increasing attention is being given to bacterial amylases, especially intheir industrial aspects, and a comparison of the properties of a pure bacterialamylase and the cr-amylase of malt is the subject of papers by R.H. Hopkinsand D. Kulka 28 and R. H. Hopkins, D. E. Dolby, and E. G. S t ~ p h e r . ~ ~ Itwas found that the properties of the two amylases are very closely similar,although the bacterial enzyme is able to function at higher temperaturesthan the a-enzyme, and acts more powerfully, particularly in its ability toliquefy starch paste. The points of resemblance between the two enzymesare the rapidity of liquefaction of starch paste and the formation of dextrins,maltodextrin and a little maltose.This stage of the reaction is complete a tabout 35% of total hydrolysis in terms of maltose, after which reactionproceeds very slowly until it ceases at about 90% of total hydrolysis. Veryvarying dextrins are produced in the first stage of the reaction, the malto-dextrin obtained having R 33 (maltose, R = loo), and another dextrinformed having R 15 or less.The well-known production of the non-reducing, crystalline dextrins fromstarch under the action of BaciZZzls maceruns affords another example ofbacterial amylase action, and the enzyme has been the subject of recentstudies by E. B. Tilden, M. Adams, and C . S. 31 By an elaboratemethod of purification, involving adsorption on alumina and subsequentelution with phosphate, an enzyme preparation has been obtained which issome 140 times as active as the original.The crystalline dextrins themselveshave been the object of study by D. French and R. E. R ~ n d l e , ~ ~ who havedetermined the molecular weights of the Schardinger a- and p-dextrins ofpotato starch. X-Ray diffraction experiments, combined with crystal densitymeasurements, led the authors to the conclusion that in the first case therewere six glucose units in the molecule, and in the latter seven.Hemicellubses, &.-The outstanding difficulties in the isolation of hemi-celluloses still remain those of pre-treatment of material, and subsequentextraction, The products obtained are almost invariably mixtures resultingfrom varying degrees of att.ack on the encrusting hemicelluloses themselvesand on the cellulose fraction. The subject is again discussed by I.A. P r e e ~ e , ~ ~who confirms the loss of hemicellulose material resulting from pre-treatmentwith boiling alkali. Hot alkaline extraction favours the dissolution of non-pentosan, whereas cold extraction favours the dissolution of xylan. Therelative proportions of these types in the mixture obtained vary with theraw material and the conditions of its treatment.Hemicelluloses conforming to the general type of those found in hard-2 7 J . Agric. Chem. SOC. Japan, 1941, 17, 779.2 8 J. Insf. Brew., 1942, 48, 170.30 J . Bact., 1942, 43, 627.3a J . Arner. Chern. rSoc., 1942, 64, 1661.29 Ibid., p.174.s1 J . Arner. Chem. Soc., 1942, 64, 1432.83 Biochem. J., 1941, 96, 669234 BIOOHEMISTRY.woods have been isolated by E. Anderson, R. B. Haster, and M. G. Seeley 34from cotton-wood, Populus Macdozlgali. They comprise a mixture of mole-cules of a methoxyuronic acid combined with a chain of q l a n units. Thechain in this case appears to be smaller than is usually the case, consisting aait does of only 7-9 xylan units. The presence of starch was indicated by theiodine coloration in the hemicelluloses extracted prior to chlorination, but itwas absent from the products obtained after chlorination. It is suggested thatthe composition of these hemicelluloses indicates a possible origin by partialoxidation or decarboxylation of starch or dextrin.The presence of pecticsubstances in the wood was confirmed by isolation of a product probablyidentical with pectic acid. Another wood product, an arabogalectan oflarch, is described and discussed by E. V. White.35,3s Similar fractions ofthe product are obtained by water extraction of the sawdust, and subsequentprecipitation with alcohol. A study of the products of acetylation andmethylation leads to the conclusion that the polysaccharide consists of ahighly branched chain of galactose units. Terminal groups of arabo-furanoae and galactopyranose are attached to carbon atom 6 of the units ofthe galactose chain.In addition to starch, which comprises the bulk of the seed, oat seedshave been shown to contain lawans, and amongst other possible poly-saccharides, D.L. Morrisa7 has isolated and examined lichenin and anaraban. The author revives the hypothesis of Karrer that lichenin may bewidespread in nature as a “reserve cellulose,” and shows that this is thepolysaccharide responsible for the cupric chloride crystallisation pattern.These crystal patterns are not without interest. It was shown by D. L.Morris and C. T. Morris 38 that the crystal form of cupric chloride was un-affected by inorganic salts, simple sugars and glycine, but that solutions ofpolysaccharides such as starch and glycogen produced patterns which werecharacteristic of the added substance. A substance in oats gave a charac-teristic pattern, shown above to be due to lichenin. Cupric chloride crystalpatterns were employed by the authors39 in the identification of a cmbo-hydrate in Zea mays shown to be glycogen, athough it was noticed that somesubstance appeared to be present which rendered the patterns for the Zeaextracts slightly different from those of other samples of glycogen.Theisolated glycogen gave patterns identical with those for animal glycogen.The substance responsible for the differences in the extracts was found4()to be a protein or proteins, and the effect of the protein appears to be depen-dent on the total amount of protein present rather than on the ratio of proteinto polysaccharide. The specificity of the patterns is due to the polysaccharideand the action of the protein is entirely non-specific.Other polysaccharides recently investigated include an insoluble poly-saccharide of yeast, the mucilage of the seed of Indian wheat, and a new34 J .Biol. Chem., 1942, 144, 767.36 Ibid., 1942, 64, 302.3 8 J. P h y e h l Chern., 1939, 43, 623.40 Ibid., 1941, 141, 516.35 J . Amer. C k m . BOG., 1941, 63, 2871.37 J . Biol. Ohem., 1942, 142, 881.39 J . Biol. Chem., 1939, 130, 535NORRIS: SOME PLANT W D U W AND ENZYMES. 236rhamnosm. The struoture of the first of these, the so-mlled yeast cellulose,has been investigated by W. Z. Hassid, M. A. Joslyn, and R. M. McCready,qlwho confirm the findings of L. Zechmeister and G. T0th,4~ that the glucosidiclinkages in the molecule are of an unusual ohmacter. After methylationand hydrolysis the sole product was 2 : 4 : 0-trimethyl gluoose ; no traceof tetramethyl gluoose could be found, and hence it follows that the moleculeconsists most probably of a closed chain whose glucopyranose units arecombined through oarbon atoms 1 and 3, and not 1 m d 4 as is most generallythe case.Evidence wm also adduced to show that the glucosidic linkageswere predominantly of the p-type, and viscosity measurements indicated amolecular weight of about 6500. The mucilage of the seed of Indian Wheat,a native of the arid regions of the south-western United States, is similar incomposition to that of Psyllium seed,@ and WM obtained in about 19% yield.On the basis of hydrolysis and isolation of the sugars formed, and of deter-mination of pentosan and uronic aoid, E, Anderson, L.A. Gillette, and M. G.Seeley 44 oonclude that the mucilage oonsists of aldobionic acids oontaininggalacturonic acid and arabinose, combined with xylose, and a, small fraotionwhich resists solution on hydrolysis. More precise evidence of structure ca,nonly be established by further examination. A preliminary note on a newrhamnose-containing polysscoharide from one of the Chlorophyceae, Ulvalactzcca, is contributed by M. M. T. Plant and E. D. Johnson.46 It is weaklyacidic and non-reducing, and, like other algal polysacchitrides, it contains asulphuric ester grouping, but, unlike some others of this type, it does notcontain uronic acid residues. Rhamnose was identified as a product ofhydrolysis and further examination of the polysaocharide is to be under-taken.Protein8 and Amim-ao&.-A crystalline protein preparation of some-what unusual properties is described by A.K. Bttu8,46 and by this author withW. S. Hale and T. H. Harri~3.4' It was derived from a light petroleum extractof flour, and was precipitated therefkom, after removal of sterols, by etherand alcoholic hydrochloric mid. It could be crystallised fiom 75% alcohol,and examination showed it probably to be the hydrochloride of an oxihedfragment of a lipo-protein, containing cysteine, in the original flour. Thechief amino-acids present, to about two-thirds of the total, were arginine,cystine and tyrosine. It was found to be toxic to yeasts and to some animals ;it reversibly inhibits chymopapain and protects carotene from oxidation bycarotene oxidase.It is attacked by proteolytic enzymes such as trypsin.The effect of the conditions of hydrolysis of zein on the products obtainedhas been examined by R. Borchers and C. P. Berg,*s since it had been observedthat zein hydrolysates obtained by sulphuric acid hydrolysis in the autoclavea t 165" gave anomalous results in growth experiments with young rats. It4 1 J . Amer. Chem. Soc., 1941, 68, 296.42 Biochem. Z., 1934, 270, 309; 1936, 284, 133.43 E. Anderson and M. Firema, J . Biol. Chem., 1936, 109, 437.4 4 Ibid., 1941, 140, 669.4 6 J . Washbingtm Aoad. SC~., 1942, 32, 132.4 8 J . Biol. Chem., 1942, 142, 693.4 5 Nature, 1941, 147, 390.47 Cereal Chem., 1942, 19, 279.H236 BIOCHEMISTRY.was thought that racemisation or destruction of essential constituents mighthave occurred under the conditions of hydrolysis, since the optical rotationof the hydrolysate under autoclaving was lower than that obtained by re-fluxing. Using acid of concentration between 14 and 33% by volume in aseries of experiments it was found that no racemisation or destruction occurredeither in refluxing or in autoclaving a t 120-180".The refluxing could beconsiderably prolonged with little or no harmful effect, but continuationunder the autoclave beyond the time necessary for hydrolysis induced bothracemisation and destruction. It was considered that sulphuric acid of aslow a concentration as 8% was unsuitable for complete and satisfactoryhydrolysis of zein.In a further communication, R.Borchers, J. R. Totter, and C. P. Berg a9have traced the failure of autoclave hydrolysates of zein to support growth ofrats on otherwise suitable diets, in part, a t least, to the loss of threonine,whose structure renders it specially liable to racemisation. The loss ofthreonine is not great under reflux or mild autoclave treatment in hydrolysis,but becomes increasingly marked as autoclave conditions become moredrastic.A brief discussion of some of the methods now being employed in theisolation of protein hydrolysis products was included in last year's AnnualReport,so and to these may be added investigations by R. L. M. Synge and co-workers, The investigations had as their object the examination of hydro-lysates from animal proteins, but inasmuch as the methods are of generalapplication, and in part involve new physico-chemical applications, they maybe included here. In a series of papers published in 1939, R.L. M. Synge s1describes experiments to determine the partition coefficients of a number ofacetamido-acids between immiscible solvents, notably chloroform and water,and also ether and ethyl acetate-water systems. The preparation of anumber of acetamido-acids is described, including some not previously pre-pared. If such technique is to be of service in separating amino-acids fromthe mixture in an hydrolysate, it would be necessary to acetylate the amino-acids of the hydrolysate and to obtain the acetyl compounds in pure form.Separation on the basis of their different partition coefficients betweenchloroform and water would follow. A known mixture of amino-acids beingused, it was shown that those susceptible to fractional extraction in this waywere recovered in high yield.The acetylhydroxy-amino-acids are almostentirely extracted from aqueous solution by chloroform. A means of iso-lation of these acids was sought via the acetylbenzoyl derivatives. A methodfor preparing the serine and hydroxyproline derivative was evolved, and itwas found that such derivatives were debenzoylated by alkali a t roomtemperature and deacetylated by boiling dilute acid. On the basis of theseproperties a method of isolation of the hydroxyamino-acids in protein hydro-lysates was worked out. In the last paper of this particular series the methylesters of the acetylhydroxyamino-acids were prepared, their partition betweenJ .Boil. Ohem., 1942,142, 697. Ann. Reports, 1941, 38, 266.51 Biochem. J., 1939, 83, 1913, 1918, 1924, 1931NORRIS: SOME PLANT PRODUCTS AND ENZYMES. 237water and chloroform measured, and the results applied to the separation ofhydroxyamino-acids in protein hydrolysates. In a later communication byA. J. P. Martin and R. L. M. Synge 52 the extractional fractionation of amino-acids in protein hydrolysates as acetyl derivatives is carried a considerablestep forward. The design, construction and operation of a 40-unit counter-current liquid-liquid extraction apparatus is described, and by its means asatisfactory procedure for the isolation and determination of methionine,valine, proline, leucine and phenylalanine was elaborated.On test withknown mixtures of the amino-acids the results showed that the method hadsome advantages over Dakin’s butyl alcohol extraction, followed by Fischer’sester distillation.A renewed attack on the hydroxyamino-acid fraction is made by the sameauthors,53 who studied next the applications of the reaction with periodicacid. Threonine yields acetaldehyde on treatment in neutral solution withperiodate, and this may be removed by simple aeration. The other hydroxy-amino-acids do not yield acetaldehyde under these conditions, and thereaction is adapted to the purposes of a micro-determination of threonine.The other amino-acids yield formaldehyde under periodate treatment, but thedetermination of this by means of dimedone, for example, was unsatisfactory.It was found, however, that threonine and serine and periodate in presence of50% potassium carbonate gave definite yields of ammonia, and this findingwas applied to the determination of the amino-acids in complete proteinhydrolysates. In further application of the acetylation-benzoylationtechnique previously mentioned, low recoveries of threonine in a woolhydrolysate were obtained ; serine, however, was obtained in good yield.The same procedure applied to gelatin and isinglass yielded a substancesimilar to Van Slyke’s hydroxylysine, which, like the latter, yielded form-aldehyde and ammonia with periodic acid.At a later date, A. J. P. Martinand R.I;. M. Synge 54 have enlisted the aid of chromatography in the solutionof the problem of the determination of the higher amino-acids in proteinhydrolysates. A novel form of chromatogram is described, and a generaltheory of chromatography developed. The form employed depends not onadsorption on a solid phase but on partition of solutes between two liquidphases, and the visual detection of the colourless acids in the chromatogramis achieved by the use of an indicator added to one of the phases. The newchromatogram is successfully applied to a micro-determination of the amino-acids phenylalanine, leucine, proline, valine and methionine in artificialmixtures and in hydrolysates from wool.Pigments.-In isolating the lipoid pigments from ‘‘ Sherbro ” palm oil,R.F. Hunter and A. D. Scott 55 consider that in the light of the physicalproperties, they have obtained purer specimens of a- and @-carotenes thanany hitherto described in the literature. In addition to these they obtainedy-carotene, neo-y-carotene, lycopene, neolycopene and neolutein. The chro-matogram also revealed the presence of a new carotenoid adsorbed at a posi-62 Biochem. J., 1941, 35, 91.54 Ibid., p. 1358.63 Ibid., p. 295.6 K Ibid., p. 31238 BIOUHEMISTRY.tion intermediate between those of y- and (3-carotenee. The authors suggeetthat in the biogenesis of the carotenes, a common intermediate complex is thesource of lycopene, 13-carotene and y-carotene, and that a-carotene may laterarise from isomerisation of the p-compound. The carotenoid compositionof the unsstponifiable matter of a West African plantation oil was shown bythe aame authors, with J. R. E d i ~ b u r y , ~ ~ to be qualitatively similar to thatof the Sherbro oil, and again the new carotenoid intermediate on the chro-matogram between y- and p-carotene was observed, Ergosterol was alsoreported. In a dkoussion of the isomerisation of the osrotenoids the authorsdraw an analogy between the reversible isomerhation of the carotenoids andthe mobile cis-trans isomerism of azobenzene.Although it has been established that some varieties of Irls owe theircolour to members of the anthocyanin group, it was shown by van Wisselingh6'and has recently been confirmed by W. F. O'Connor and P. J. D r u m 58that the water-flag, Iris peudacoru8, owe8 its colour ta carotenoids, of whichp-carotene, violaxanthin and lutein have been isolated and characterised.There is evidence that zeaxanthin is present, and a red wax-like substancehas alao been noted. It is possible, on the evidence of adsorption bands incarbon disulphide solution, that this substance is closely identifiable with asimilar substance obaerved by R. Kuhn and A. Winterstein 59 in the caro-tenoids of VioZa tricolor.The view has been very generally held that secondary produds of plantmetabolism once formed are then defhitely eliminated from the metabolicprocesses of the plant. Examples am gradually accumulating, however,which indicate that iome modification of this view is necessary, in particularwhere the normal metabolism of the plant is disturbed by etiolation. It hasbeen found that when seedlings of Salvia oflcinuZis are etiolated there is aresorption of essential oils.6o Alkrt1oid.a have also been utilised by etiolatedplants as shown in the cases of hordenine,61 caffeine and theobromine,a2 andthe alkaloids of Lupinus Zuteu~.~~ Recent experiments by A. Frey-Wyaslingand F. Blank G4 have shown that anthocyanins are utiliaed similarly. Seed-lings of red radishes or red cabbage were found to develop anthocyanins veryearly on germination, but when the seedlings were subjected to the metabolicdisturbances consequent on etiolation, the anthocyanins disappeared in afew days only a t 95' F., and slowly during several weeks a t 50' F.The pigment present in some flowers, and notably the Compositte, whichgives an intense red coloration with alkali has been shown to be the chalkonebutein. J. R. Price 65 has identified it in Dahlia variubilis and shown thatit may be formulated as ahown (I). It probably occurs both as suoh andas an easily hydrolysable glycoside. The pigment has also been isolated5 6 Biochem. J., 1942, 38, 097.5' Nature, 1941, 147, 58.60 A. Frey-Wysaling and F. Blank, Vtrh. Schweiz. Natwf. Qes., 1940,61 Y. Raoul, Compt. rend., 1937, 205, 450.82 T. Weevers, Arch. Nderl. Sci. naturelles, 1930, 4, l l l / B .63 J. C. J. Wallebrook, Rec. Truv. bot. Nderl., 1940, 87.64 Nature, 1941, 147, 148,Plwa, 107, N.S. 7, 371.Ber., 1931, 64, 332.* 6 J., 1989, 1017NORRIS : SOME PLANT PRODUCTS AND ENZYMES. 239by T. A. Geissmann66 from Coreopsis Douglmii, a native of SouthernCalifornia, and in this case also it is evident that both glycoside andaglycone exist in the flower. TheOH OH term anthochlor pigment is applied toOH(-_)-COCH:cH-(-)OH this type of polyhydroxychalkone andothers of the group are likely soon tobe isolated. For instance, a pigmentof Cormpi8 grundiflora is under investigation and is probably a penta-hydroxychalkone. The interest of theae pigments lies not solely in theirchemistry and struoture, since they are equally important to thegeneticist. Genetical data have shown that anthocyanins, flavones andpigments such as butein are derived from some common intermediate;increasing knowledge of the pigments derived from this intermediate willlead in turn to an understanding of the nature of the intermediate, andthe stages in biogenesis of large groups of plant products may be revealed.F. W. N.C. G. ANDERSON.J. I;. CRANMER.L. J. HARRIS.A. NEUBERUER.F. W. NOREIS.-(1.16 6 J . Amer. Cham. Soc., 1941, 83, 656

ISSN:0365-6217    

DOI:10.1039/AR9423900209

出版商:RSC    

年代:1942

数据来源: RSC

摘要:

INDEX OF AUTHORS’ NAMES.ABEUG, R., 59.Abney, W. de W., 50.Abramovitch, B., 138.Adams, J. T., 138.Adams, M., 233.Adams, R., 118, 119, 120,186, 202, 203.Adelson, D. E., 184.Adickes, F., 139.Adkins, H., 152, 153, 182,Agner, K., 212.Ahmann, C. F., 212.Ahmann, F. F., 141.Airs, R. S., 116, 123.Akiyama, H., 195, 200.Albert, 65.Alder, K., 170, 184, 190.Alfrey, T., 22.Allen, A. O., 46.Allen, C. F. H., 164, 165,179, 184.Allen, R. J. L., 210.Alliger, G., 195.Allsopp, C. B., 118, 126.Alphen, J. van, 198.Amore, S. T., 183.Anderson, C. G., 225.Anderson, E., 234, 235.Anderson, H. H., 76.Anderson, T. F., 224.Anderason, H., 56.Anson, M. L., 226.Anthes, J. A., 165.Antonov-Romanovsky, W.Anzilotti, W. F., 136.Appleton, (Sir) E., 209.Archer, S., 181.Arnestad, K., 89.Arnold, R.T., 187.Asker, (Miss) W., 200.Askew, H. O., 212.Astbury, W. T., 111, 115.Atkin, W. R., 226.Audrieth, L. F., 94.Bachmann, W. E., 158,164,171, 179, 180, 182, 184,186, 190.Back, S., 196.Backer, H. J., 184, 195.Baddar, F. G., 185.Badger, G. M., 179, 191.Badoche, M., 188.Bailey, A. J., 154.Baker, W., 118.Baker, W. O., 25.196.W., 84.Balfe, M. P., 116, 123.Balls, A. K., 228, 235.Barber, C., 218.Barcroft, (Sir) J., 210.Bardeen, J., 70.Barker, J., 210.Barnett, E. de B., 169, 174,Barnett, J. W., 231.Barron, E. G. S., 224.Bartelt, O., 51.Bartlett, P. D., 121, 170.Bartovics, A., 22.Bate-Smith, E. C., 210.Bates, J. B., 87, 101.Bathe, A., 90.Bauer, S.H., 104.Baumgarten, P., 91.Baur, E., 50.Bawden, F. C., 220, 222.Bawn, C. E. H., 40, 42, 43,44, 46, 49.Baxter, S., 66.Beach, J. Y., 104.Beadle, G. W., 198.Beck, A. B., 213.Becker, H., 163, 172, 186.Beckerath, K. von, 53.Beckmann, C. O., 116.Beech, W. F., 166.Beese, N. C., 85, 86.Beets, M. G. J., 202.Beevers, C. A., 98.Behrens, H., 74.Bell, A., 151, 164, 165.Bell, R. C., 208.Bellucci, I., 93.Beltz, W., 145.Bennetts, H. W., 213.Benzon, B., 214.Berg, C. P., 235, 236.Berg, W. F., 52, 55, 57,63, 64, 69.Berger, H., 19.Berger, J., 39, 226.Bergmann, E., 163, 167,169, 170, 174, 175, 178.Bergmann, F., 167,170,174,175.Bergmann, M., 227, 228.Bergmann, W., 188.Bergstrom, F. W., 75.Berkmen, S., 231.Berl, E., 157.Berlin, T., 177.Bernal, J.D., 111.Bernstein, H. I., 123.Berry, T. M., 135.176.240Bertrand, G., 213, 214.Beutel, R. H., 199.Beutler, H., 36, 37, 39.Beyer, H., 168.Bhatnagar, S. S., 93, 188.Biebesheimer, H., 157.Biltz, M., 51.Biltz, W., 91.Birch, S. F., 136.Birus, K., 86.Bischoff, F. von, 90.Bitterlin, O., 119.Bittner, C., 89.Blackie, J. J., 203.Blakemore, F., 214.Blank, F., 238.Blayden, H. E., 99.Blease, R. A., 28.Blicke, F. F., 176.Blokhina, A. N., 177.Bloom, E. S., 153.Bloomfield, G. F., 32.Blount, B. K., 208.Bludan, W., 90.Bluestein, B., 185.Blum-Bergmann, (Mrs.) O.,Boatner, C. H., 186.Boersch, H., 97.Boetius, M., 195.Bogdandy, S. von, 37: 38,39.Bogert, M. T., 183, 184.BognAr, R., 193.Boissonnas, C.G., 14, 16.Boivin, A., 215.Bondi, A., 217.Bondy, H. F., 24.Bonhoeffer, K. F., 48, 75.Booth, H. S., 92.Borchers, R., 235, 236.Bostrom, S., 17.Bousset, R., 175.Bowden, E., 153.Bower, J. R., 153.Bowlus, H., 129, 131.Bowstead, J. E., 213.Bowtell, J. N., 85.Boyd, W. C., 219, 220, 222,Boyland, E., 159.Bradley, A. J., 96.Bradsher, C. K., 178, 183,Brhuning, W., 75, 89.Bragg, (Sir) W. L., 97, 98.Bras, J. le, 188.Brass, K., 167, 168.Brauer, G., 88, 101.163, 175.223, 224.185INDEX OF AUTHORS’ NAMES. 241Brauns, F., 143.Brauns, F. E., 144.Bredig, M. A., 88, 101.Brentano, J. C. M., 56.Breslow, D. S., 136, 137.Bretschneider, H., 186.Breuer, W., 90.Brewster, R. Q., 141.Brickman, L., 151.Briese, R.R., 138.Briggs, L. H., 183, 208.Brockway, L. O., 115.Brody, S. B., 100.Brransted, J. N., 27, 29.Bromehead, C. N., 214.Brouwer, E., 213.Brown, H., 218.Brown, H. C., 119,140,141.Brown, W. G., 111, 185.Bruce, W. F., 178.Brumshagen, W., 14.Brunel, L., 129.Bruns, W., 91.Bruson, H. A., 140Bryant, W. M. D., 129, 139.Buchholz, K., 116, 219.Buckland, I. H., 150.Buerger, M. J., 97, 98.Bull, L. B., 213.Bunn, C. W., 108.Bunney, W. E., 224.Bunsen, R., 60.Bunting, E. N., 85.Burger, A., 182.Burgers, J. M., 19, 26.Burgess, H., 125.Burgess, W. M., 93.Burhans, A. S., 185.Burnet, F. M., 220.Butenandt, A., 198.Butler, J. A. V., 229.Butlerow, A., 133.Butz, E. J. W., 184.Butz, L. W., 184, 185.Byler, W.H., 83.Cady, G. H., 93.Calcott, W. S., 187.Cameron, J. M. L., 196.Campbell, A. W., 157.Campbell, D. H., 220.Campbell, W. P., 191.Cannon, P. R., 221.Carlisle,, C. H., 115.Carlson, C. A., 93.Carlson, E. W., 133.Carmack, M., 180, 203.CarrB, P., 141.Carroll, B. H., 51, 52, 58.Carroll, J. J., 208.Carter, S. R., 9.Cason, J., 159,176, 177, 187.Caspar, E., 139.Caspari, W. A., 14.Casseday, J. T., 183.Cathcart, J. A., 187.Chalmers, J. G., 159.Chamberlin, E., 199.Chang, T. S., 13.Chanley, J., 123.Chapman, F. E., 213.Charrier, G., 164.Chatt, J., 119.Chemerda, J. M., 179.Chu, F., 225.Ciskowski, J. M., 134,Claisen L., 138.Clapp, F. L., 224.Clapp, R. C., 184.Glar, E., 160, 161, 167,170, 184, 190.Clark, D., 101.Clark, W., 65.Clark.W. M.. 217.35.169,Clemo, G. R.; 119, 202.Clews, C. J. B., 110.Clusius, K., 70.Cochran, J., 185.Coffman, J. A,, 163.Cohen, S. G., 170.Cohn, E. J., 225.Cohn, W. M., 89.Coleman, G. H., 195.Collie, N., 140.Combieeco, C., 216.Conner, W. P., 144.Connor, R., 138.Cook, E. L., 88.Zook, J. W., 156, 158, 166,167, 168, 171, 173, 174,176, 177, 179, 180, 182,183, 189, 191.Zooke, L. M., 153.Zoonradt, H. L., 196.Zope, A. C., 138.Zorell, M., 163, 173, 186.Zorelli, R. M., 93.Zorey, R. B., 103, 105.Zorner, H. H., 212.Zornforth, J. W., 141, 183.>omforth, (Mrs.) R. H.,141, 196.Zorse, J., 140.;orson, B. B., 135.>osciug, T., 157.=ottin, H., 201.>oulson, E. A., 157.2oumoulos, G., 117.?ousin, H., 149.Yowdrey, W.A., 120.?ox, E. G., 110, 195.h i g , L. C., 207, 208.h m e r , A. B., 142, 152.h e c h , H. J., 219.highton, R. H. J., 143,h t c h e r , L. H., 138, 139.Xegee, R., 186.3romwel1, N. H., 157.:rook, E. M., 225.>rowfoot, (Miss) D., 115,151.159.CroxaM, W. J., 130.Cruickshank, E. M., 210.Curry, J., 42.Dahll, P. J., 166.Dalling, T., 213.Dalma, G., 208.Damboviceanu, A., 218.Danielli, J. L., 228.Daudt, W. H., 175.Davies, T. L., 126.Dawton, R. H. V. M., 96.De Boer, J. H., 61.De Groot, W., 84.De Lange, J. J., 96, 103.Delaunay, A., 216.Derjugin, W. von, 198.Desreux, V., 226.Deutschbein, O., 87.De Wad, H. L., 203.D’Ianni, J., 152.Dick, G. F., 224.Dick, R. H., 137.Diekel, G., 70.Diels, O., 170, 190.Dietzel, A., 92.Dimroth, O., 125, 177.Dinger, A,, 174.Dixon, J.K., 212.Dobriner, K., 159.Dobry, (Mme.) A., 13.Dolby, D. E., 233.Dorfman, A., 231.Dorp, van, 141.Dorris, T. B., 129, 130, 135.Dozois, T. F., 217.Drew, H. D. K., 166, 197.Dreyer, H., 72.Drude, 125.Drumm, P. J., 238.Dubensky, E., 199.Diirr, W., 144, 145.Dufraisse, C., 169, 188,Dunlap, C. E., 158.Dunlop, G., 213.Dunning, W. J., 40.Dunstan, A. E., 136.Durland, J. R., 182.Dussy, J., 193.Duveen, D., 188.Du Vigneaud, V., 230.Eakin, R. E., 230.Eastes, J. W., 93.Ecker, E. E., 225.Eckhardt, H. J., 187.Edgerton, R. O., 184.Edisbury, J. R., 238.Eggert, J., 51, 56.Eirich, F., 19.Eisenlohr, F., 126.Eisenstein, A., 100.Eisleb, O., 199.Eiter, K., 196.Elam, D.W., 226.189, 190242 INDBX OF AUTHORS' NAME8.Elderfield, R. C., 172, 183.Eldering, G., 217.Elveh'em, C. A., 213, 214.Erneldus, H. J., 93.Enderlin, L., 189.Engel, B. G., 208.Engler, K., 148, 151.Ensslin, F., 72.Ephraim, F., 77.Epler, H., 194.Erdtman, H., 142, 148,149, 150.Erickson, C. L., 108.Erlenmeyer, H., 119, 126.lbkmne, A., 188.Evans, A. B. A., 26, 33.Evms, A. G., 42, 43, 44,Evans, C. H., 61, 63.Evans, D. G., 217.Evans, G. H., 53.Evans, M. G., 42, 47, 48,Evans, T. H., 216.Everest, A. E., 155.Eyring, H., 10, 116.Faber, K., 120.Fairbrother, F., 45.Fajans, K., 52, 53.Faller, F. E., 91.Faltis, F., 205, 208.Fankuchen, I., 105, 11 1.Fanta, K., 168.Farmer, E.H., 8, 32.Farquarson, J., 188.Fedorov, G. I., 178.Fehse, W., 89.Fehser, R., 89.Feldman, J., 185.Felici, L., 199.Ferguson, W. S., 214.Fernholz, M. E., 219.Fett, R., 91.Fettback, H., 190.Fick, R., 177.Fieser, L. F., 188, 159, 160,162, 163, 165, 168, 172,174, 176, 176, 177, 178,179, 180, 181, 182, 183,184, 186, 187, 189, 191.48, 121.49.Fieser, M., 176, 181.Filmer, J. K., 212.Finch, G. I., 99.Fireman, M., 235.Fischer, A., 132.Fischer, H., 141.Fischer, K., 8, 10, 14, 19,Fisher, J. H., 155.Fliokinger, E., 145, 147,Florey, M. E., 238.Flory, P. J., 12, 25, 28, 30.Flosdorf, E. W., 217.Folkers, K., 206.25, 139.148.Fonda, 0. R., 85.Foohey, W. L., 131.Fordyce, R., 26.Forst, P. von der, 92.Found, C.G., 85, 86.Fowler, R. D., 71.Fowler, R. H., 12.Frame, G. F., 165.Frank, R. L., 163.Franke, B., 91.Frankenburger, W., S2.Franz, H., 37.French, D., 233.Frena, A. M., 213.Friedrich, W., 68.Friedrichsen, W., 190.Fries, K., 162.Freudenberg, K., 124, 126,142, 143, 144, 145, 147,148, 150, 161, 163.Freudenberg, W., 206, 207.Freundlich, H., 68.Frey-Wyssling, A., 238.Frolich, P. K., 133, 134.Frommer, L., 45.Froning, J. F., 132.Frosch, C. J., 108.Fruton, J. S., 227, 228.Fuchs, H., 73.Fuller, C. S., 25, 108.Fuller, M. L., 94.Funke, K., 187.Furth, J., 218.Fuson, R. C., 140.Caddis, A. M., 184.Gale, E. F., 225.Galloni, E. G., 100.Gaaguly, N. C., 184.Gasselh, V., 128.Gastinger, E., 71, 89.Gaviit, I., 157.Gawlick, H., 137.Qee, G., 8, 10, 11, 14, 17,23, 26, 28, 29, 30, 31, 32,Geissmann, T.A., 239.Gentile, R. A., 206.George, S. W., 194.GBrard, M., 188.Gerlough, T. D., 216.Geyer, B. P., 175.Ghigi, E., 164.Gibbs, E. M., 206.GiguBre, P. A., 103, 104.Gillette, L. A., 235.Gilman, L., 164.Gingrich, N. S., 100.GisOlf, J. H., 83, 84.Glading, R. E., 144.Gleu, K., 200.Clocker, R., 68.Glockler, G., 94.Godard, H. P., 153.Goebel, W. F., 217, 219.G o l p v , L. N., 177.33, 34, 35.Gomberg, M., 164.Goodway, N. F., 169.Gordon, J., 226.Gordon, S. A., 229.Gordy, W., 105.Gorianow, W., 133.Goulden, F., 177, 179.Gray, K. R., 143.Green, D. E., 232.Green, H. H., 211.Gregor, U., 76.Grigsby, W. E., 123.Grimm, L., 88, 04.Grimmett, R.E. R., 212.Grob, K., 89.Groot, C., 119.Gross, H., 23.Gross, S. T., 94.Grosee, A. V., 135, 136.Grosskoff, K. G., 157.Grube, H. L., 75, 89.Gruhl, A., 85.Guggenheim, E., 11, 12.Gurney, R. W., 60, 64, 60,Guth, E., 20.Guthrie, F., 60.Haag, A,, 143.Haagen-Smit, A. L., 198.Haber, F., 37, 143.Hilgglund, E., 143.Hagen, Gt., 206.Hager, J. J., 157.Hague, E., 230.Hah, G., 196.Hale, W. S., 235.Hall, C. E., 59.Halla, F., 90.Hamburger, ct., 107.Hamilton, R. H., 11 1.Harnlin, K. E., 203.Hammon, W. MOD., 224.Hammond, J., 210.Hampel, J., 89.Hampson, G. C., 102, 115.Hams, C. S., 232.Hannawdd, H., 89.Hanske, W., 197.Harder, M., 143.Hareanape, J. N., 42, 45,Harington, C. R., 211.Harker, D., 102.Harley-Mason, J., 197.Harper, 8.H., 194.Harradence, R. H., 186.Harrer, C. J., 212.Harris, E. E., 152, 153.Harris, G. C., 140.Harris, T. H., 235.Harris, T, N., 217, 224.Hart, E. B., 213, 214.Hartel, H. von, 41, 42, 46,Hartley, G. 8., 228.68, 79.46.46INDEX 02 AUTEORS' NANBS.Haptree, E. F.. 212.243Hieeer. I.. 156.Hartung, E. J:, 50.Harvey, D. G., 198.Harvey, R. J., 212.Hasche, R. L., 37.Haskelberg, L., 170, 177.Hasselstrom, T., 183.Hassid, W. Z., 232, 285.Hatihama, Y., 153.Haurowitz, F., 223.Hauschild, K., 116.Hauser, C. R., 136, 137,138.Hawkins, W. L., 161, 152,155, 216.Haworth, R. D., 149, 154,155, 171, 180.Haworth, W. Nu, 232.Heath, R. L., 232.Heohtmann, 5. F., 144.Hegedus, B., 139.Heggie, R., 126.Hehre, E.J., 217.Heidelberger, M., 221, 224,Heilbron, I. M., 229.Heimbrecht, M., 91.Hein, F., 74.Heinrich, E., 196.Heinz, W., 89.Heiss, J. H., 25.Hettsche, H. O., 236,Heller, W., 40, 42, 43, 44,Helmert, E., 216.Helmholz, L., 100, 107.Henderson, G. M., 117.Heme, A. L., 32.Hennig, W,, 89.Hennion, G. F., 130, 131,132, 133, 136, 136.H e ~ y , G. C., 111,Henriques, F. C., 119.Henry, T. A., 206.HBrissey, H., 149.Herligman, E., 169.Herriott, R., 226, 227.Herrmmn, E., 89, 90.Herrmann, H., 89.Hershberg, E. B., 159, 165,176, 176, 181, 187, 191.Hershey, A. D., 223.Herzfeld, K. I?., 52.Herzog, R. O., 144.Hessler, 139.Hewett, C. L., 156, 169,167, 168, 171, 173, 180,182, 183, 184,Hewson, W.B., 162.Heymann, H., 158, 176.Heyningen, W. E. van,225.Hibbert, EL, 25, 142, 143,145, 148, 150, 161, 162,153, 106, 216.Hieber, W., 73, 74.226.45.Higiinbottom, A., 183.Higgins, A. G., 169.Hildebrand, J. H., 13, 35.Kill, P., 183.Hillemann, H., 137.Kilmer, H., 144.Kilpert, R. S., 151.Ailsoh, R., 51, 70, 82.Kind, J., 123.Kindenbur , K&., 196.Hinderer, 8., 139.Einton, H. D., 129, 130,Hirschlaff, E., 63, 81.Kirst, E. L., 229.Hnevkosky, O., 89, 90.Hoagland, C. L., 218.Hoard, J. L., 102, 103.Hock, L., 17, 18.Hodgson, M. B., 58.Hoter, H., 206.Hoffmann, I., 92.Hoffstadt, R. E., 217.Hofmann, K., 230.Hofmann, K. H., 196.Hofstadter, R., 86.Holmberg, B., 143, 149.Holzinger, L., 208, 208.Horneyer, A.H,, 138, 139.Hooker, S, B., 219, 220,Hopkins, R. H., 233.Hoppene, H. A., 190.Horclois, R., 169.Horii, Z,, 192.Horn, E., 39, 46.Horning, E. C., 140.HornulT, G. von, 176.Hottle, a. A,, 224.Houpillart, J., 189, lQ0.Houwink, R., 26.Howard, F. A., 133.Howes, H. L., 78.Hoyle, L., 216,Hubard, S. S., 93.Hubbard, D., 58.Hudson, B. E., 137, 138.Hudson, C. S., 233.Hudson, J. H., 58.Huckel, W., 120, 186,Hiittig, G. F., 89, 90.Hugel, G., 179.Huggins, M. L., 12, 20, 28,Hughes, E. D., 120,121,171.Hughes, E. W., 98, 103,Huisman, I;. H. H., 184.Hukuti, G., 193.Hunter, M. J., 142, 145,Hunter, R. F., 46, 237.Hutohings, B. L., 230,Hutchison, D. A., 70.Huthsteiner, H., 86.131, 136,222, 223.98.106, 107.152.Kuzise, S., 19.[ball, J., 17 1.Ckeda, T., 205.Cngold, C.K., 120, 121, 189.Cnnes, J. R. M., 213.Cpatieff, V. N., 136, 136.Crimescu, I., 157.Crwin, S. M., 116.tsbell, E. S., 123.Its, P., 205.Cvanovics, G., 215.[was, J., 198.Jaoobs, W. A., 207, 208.James, T. H., 59, 60.Jamison, (Miss) M. M.,117, 125, 127.Jander, W., 89, 91.Janson, A., 143, 146.Jauncey, G. E. M., 65.Jeanes, A,, 186.Jeffrey, G. A., 110, 195.Jenkins, H. G., 86.Jervine, 208.Johsnnsen, T., 72.John, Fr., 160.Johnson, E. D., 236.Johnson, H. G., 199.Johnson, M. J., 226.Johnson, W. S., 181, 182.Johnston, H. L., 70.Jones, C. B., 225.Jones, D. M., 138, 139.Jones, J. E., 191.Jones, R, N,, 168, 173, 219.Josephy, B., 37.Joshel, L. M., 176, 184, 186.Joslyn, M.A., 235.Jost, W., 64.Julianelle, L. A,, 217.Junu~ov, S., 183, 204.Rabat, E. A., 218.Kebgi, H., 196.Kaffer, H., 156.Kailasam, P., 76.Kalisvaart, C., 213.Kallman, H., 37.Kane, S. S., 140.Kapur, P. L., 188.Karagunis, G., 117.Karrer, P., 192.Kass, J. B., 118.Kaster, R. B., 234.Katsura, K., 205.Kaur, Q., 188.Kau~mam, W. J., 116.Kawaguchi, R., 193.Kawai, S., 149.Keilin, D., 212.Kekwick, R. A,, 222.Kelly, D. B., 197.Kelly, W. J., 149.Kemp, A. R., 19, 24,26, 31, 32.25244 INDEX OF AUTHORS’ NAMES.Kemp, F. H., 214.Kempf, A., 55.Kennaway, E. L., 158.Kenny, T. S., 194.Kenyon, J., 116, 117, 123,124, 128.Kerr, R. W., 232.Keuth, H., 90.Kharasch, M. S., 140, 186.Khmelevskii, V.I., 178,Kicksch, L., 215.Kieser, H., 56.Killian, D. B., 131, 132, 133.Kilmer, G. W., 177, 183.Kim, K. W., 193.Kimura, Y., 192.Kincaid, J. F., 119.King, C. G., 212.King, E. G., 143.King, J. D., 215.Kitchener, J. A., 83.Klason, P., 142, 143.Kleczkowski, A., 220, 222.Klement, R., 89.Klemm, W., 88, 92, 93, 94.Klink, F., 145, 147, 148.Kloetzel, M. C., 171, 190.Klug, H., 51.Knaggs, (Miss) I. E., 99,105, 106, 107.Knepper, W., 92.Knight, B. C. J. G., 225.Knopf, E., 127, 143.Koch, E., 54.Koch, P. P., 50, 68.Koch, W., 82.Koelsch, C. F., 165.Konig, J., 142.Koenigs, E., 196.Koerber, W., 157.Koerbner, W. L., 224.Kogelmann, F., 55.Kohmann, T. P., 140.Kolka, A. J., 135.Kolmer, J. A., 218.Komarewsky, V. M., 136.Kon, G.A. R., 155.Kondo, H., 205.Kondratjew, K., 37, 39,Koniuszy, F., 205.Konovalova, R. A., 183,Koolhaas, D. R., 194.Kornberg, H. A., 136.Kornblum, N., 119.Kornfeld, (Miss) G., 55, 59,Koser, S. A., 231.Koskoski, W., 71.Kotake, M., 192.Kotake, Y., 198.Kraemer, E. O., 19, 21, 23,Krebs, H. A., 214,190.49.204, 206.60.224.Krejci, L. E., 224.Kretchman, E. M., 51.Kreuger, J. W., 171.Kreutzmann, W., 137.Krings, W., 75, 89.Krocsak, M., 39.Kroger, F. A,, 83, 84, 85.Kroeger, J. W., 129.Kroepelin, H., 14, 17.Kromberg, M. L., 102.Kruber, O., 156, 157.Kubiczek, G., 199.Kubota, T., 192.Kubowitz, F., 212.Kuehn, A. O., 225.Kuhn, R., 20, 125, 238.Kuhn, W., 126, 127.Kulka, D., 233.Kulka, M., 152.Kuna, M., 124.Kyrides, L.P., 141.Labriola, R., 205.Lackmann, D., 220.Laidlaw, D., 99.Laidler, K. J., 49.Lsl, A. B., 116.Landolt, H., 124.Landsteiner, K., 220.Langer, A., 71.Lmsing, W. D., 19, 21.Lapin&, R. A., 139.Larson, R., 187.Laubengayer, A. W., 93.Lautsch, W., 151.Lavin, G. I., 159, 207, 208.Lawrence, A. 8. C., 19.Lawrence, C. A., 169.Lawson, A., 207.Lawson, A. T., 140.Lea, W. C., 50.Leaf, G., 118.Ledingham, A. E., 203.Leermakers, J. A., 52.Leger, F., 150, 199.Lehfeldt, W., 53.Leitsmann, R., 195.Le Fbvre, (Mrs.) C. G., 171.Le Fbvre, R. J. W., 171.Lenard, P., 78.Lens, J., 18.Lerer, M., 179.Leszynski, W., 53, 64.LettrB, H., 219.Leuthner, G., 89.Levene,P.A., 116,120,124.Leverenz, H. W., 78, 81,Levi, A.A,, 159.Levine, R., 100.Levshin, W. L., 84.Levy, H. A., 105.Levy, L. A., 85.Lewis, E. E., 172, 183.Lewis, G. N., 11.Liakilov, K., 37,83.Libermann, D., 141.Lind, S. C., 94.Lines, E. W. L., 212, 213.Lineweaver, H., 228.Linstead, R. P., 117, 155.Linton, E. P., 104.Linton, R. W., 218.Linwood, S. H., 85, 86.Lions, F., 185, 201.Lipson, H., 96, 98, 99.Liu, S. C., 224.Lohle, F., 51.London, H., 47.Longsworth, L. G., 231.Lonsdale, (Mrs.) K., 99,105, 106.Lord, R. C., 195.Lorenz, G., 89.Lorenz, R., 202.Lothrop, W. C., 162, 163.Lotmar, W., 110.Loveland, R. P., 57.Lowe, S., 118.Lowry, T.M., 117, 118,125,Lu, C. S., 98, 103.Liippo-Cramer, H., 55, 64,Luttringhaus, A., 116.Lukesh, J. S., 96, 98.Lukin, A.M., 167, 168.Lund, H., 182.Lund, M., 135.Lux, H., 92, 138.Lyman, C. M., 224.Lynen, F., 198.Maass, O., 104.McCarthy, J. L., 143, 151,McClellan, D. S., 94.McCollum, E. V., 213.MacConkey, C. A. H., 125.McCoy, H. N., 72.McCready, R. M., 232, 235.McCrosky, C. R., 75.McCullough, J. D., 107.McDonald, I. W., 212.McElvain, S. M., 136, 138,Macewan, D., 98.Macfarlane, M. G., 225.Macfarlane, R. G., 225.McGibbon, R. W., 179.MBchebceuf, M., 213.MacInnes, A. S., 143, 151.McKeag, A. H., 85.McKenna, J. F., 130, 135.McKenzie, A., 127.McKenzie, J., 127.McLean, M., 188.McMaster, L., 141.McMurdie, H. F., 100.McNiven, N. L., 120.Macrae, T. F., 210.Magee, J. L., 47, 48, 49.126.68.152, 153.199INDEX OF AUTHORS’ NAMES. 246Mahan, J.E., 203.Maier-Huser, H., 130.Maitland, P., 115, 120.Majima, R., 206.Major, R. T., 205.Makishima, S., 83.Mann, F. G., 119, 197.Mann, T., 212.Manske, R. H. F., 199, 203,Mapson, L. W., 210.Marchand, B., 186.Marchant, C., 231.Marchevskii, A. T., 177.Marckwald, W., 127.Marden, J. W., 85.Marion, L., 199, 203.Mark, H., 9, 16, 20, 22, 108.Markert, L., 143.Marks, H. C., 116.Marriage, A., 52.Marschalk, C., 160, 161.Marsh, J. K., 72.Marshall, P. G., 189.Marston, H. R., 212, 213.Martin, A. J. P., 237.Martin, (Sir) C. J., 212.Martin, D. R., 92.Martin, R. H., 159, 183,185, 189.Marx, A., 156.Marx, W., 168.Masterman, S., 120.Mathur, K. N., 93.Matthews, M. A., 174.Matulis, J., 68.Mavin, C.R., 171.Mawson, E. H., 159.Mayer, F., 177.Mayer, M., 224.Mayer-Pitsch, E., 168.Mayneord, W. V., 167.Mayo, F. R., 186.Medigreceanu, F . , 2 13.Meer, N., 41.Meerwein, H., 128, 129, 130.Mees, C. E. K., 58, 68,Meidinger, W., 68.Meier, G., 126.Meijer, T. M., 194.Meister, G., 85.Meister, M., 134, 147.Melrose, T. A., 202.Melville, D. B., 230.Mendelssohn, K., 63.Menschikov, G. P., 202,Mester, L., 193.Meyer, H., 141.Meyer, H. K., 169, 176.Meyer, K., 166, 173, 176.Meyer, K. H., 9, 12, 14, 16,Michaelis, A., 77.Midgley, T., 32.204.69.203.22, 26, 108, 110.Miescher, K., 196.Mikhailov, B. M., 177.Miller, A. R., 13.Miller, E. J., 198.Miller, F. A., 195.Milligan, W. O., 87, 88, 101.Mills, W. H., 115.Milner, C.J., 84.Milsted, J., 46.Milt, C. de, 186.Minchilli, (Signa.) M., 198.Mitchell, J., 129, 139.Mitius, A., 101.Mockel, J. M., 35.Modern, F., 221.Moffett, R. B., 192.Mohler, F. L., 48.Mojen, H. P., 25, 31.Monti, (Signa.) L., 199.Montignie, E., 89.Moore, G. E., 62.Morawietz, W., 71, 75, 88,Morgan, W. T. J., 216, 216,Morris, C. T., 234.Morris, D. L., 234.Mortenson, C. W., 202.Morton, R. A., 52.Mosettig, E., 171, 182.Mott, N. F., 50, 54, 60, 68,Mudd, S., 220.Muller, A,, 178.Muller, E., 137, 160, 187.Muller, H. F., 145, 147.Mueller-Cunrsdi, M., 133,Muller-Rodloff, I., 160.Mulliken, R. S., 111.Murnane, D., 213.Murray, M. M., 214, 215.Musajo, L., 198.Mutter, E., 56.Nafe, G. E., 65.Nakamura, H., 193, 213.Nakamura, Y., 233.Nantka, P.-L., 196.Neal, W.M., 212.Neelakantam, K., 193.Nelson, W. L., 138, 139.Neugebauer, W., 90.Newbold, R. P., 208.Newman, M. S., 140, 157,168, 172, 173, 175, 177,184, 185, 187.Nichols, E. L., 78.Niederl, J. B., 139.Niemann, C., 145.Nier, A. O., 70.Nieuwland, J. A., 129, 130,Nitzsche, S., 200.Nixon, I. G., 176.Noddaok, W., 56.89, 91.218, 219.79.134.131, 132, 133, 135.Norman, A. G., 142, 155.Norris, E. R., 226.Norris, R.. O., 133.Norrish, R. G. W., 49, 66.Northrop, J. H., 221, 226.Novello, F. C., 163, 172,175. 180.Oakley, C. L., 225.O’Connor, M. J., 135.O’Connor, W. F., 238.Ogg, R. A., 39, 46, 47, 48.O’Leary, L. A., 129.Oncley, J. L., 225.Ootuka, H., 37, 39.Openshaw, H.T., 208.Oppermann, A., 120.Orbkhov, A. P., 183, 204,Orent, E. R., 213.Orowan, E., 96.Orr, J. H., 216.Orth, H., 141.O’Shaughnessy, M. T., 168.Ostwald, Wo., 9, 23, 59.Otto, M., 52, 133, 134.Overbaugh, S. C., 164.Overbeek, J. van, 229.206.Pahlke, H., 35.Pait, C. F., 221.Palinchak, S., 117, 118.Palmer, J. W., 216.Pangborn, M. C., 218.Pannwitz, W., 128.Pape, N. R., 108.Papp, M., 178.Pappenheimer, A. M., 221,Parfentjew, I. A., 224.Partridge, S. M., 124, 128,Pasternack, D. S., 26, 33.Patterson, J. W., 157.Paul, R., 201.Pauling, L., 101, 102, 105,106, 220.Peat, S., 123, 229, 232.Pecher, J., 89.Peirce, A. W., 214,Pelipetz, M. G., 154.Penney, W. G., 104.Pennington, D., 230.Perutz, M. F., 111.Pestemer, M., 167, 168.Petch, N.J., 96.Petermann, M. L., 221.224.215, 216, 218.Peters, H., 19, 24, 25, 26,Peterson. W. H.. 230.31, 32.Petrow, A. A., 136.Petrow, V. A., 199.Pfeiffer, P., 194, 195.Pflugmacher, A., 75.Phillips, H., 124.Phillips, M., 142246 INDEX OF AUTHORS’ NAMES.Philpot, J. St. C., 227.Piepenbrink, H. F., 77, 93.Pieper, U., 187.Pietrzok, H., 120.Pillemer, L., 226.Pilzecker, J., 77, 93.Pines, H., 135.Piotrowski, H., 77.Pirscher, I?., 92.Pitschclk, c., 183.Plant, M. M. T., 235.Plentanides, M., 140.Platt, B. C., 117.Plentl, A. A., 139.Ploetz, T., 163.Pobloth, H., 74.Pohl, R., 79, 81, 82.Pohl, R. W., 61, 53.Poje, J. A., 141.Polanyi, M., 36, 37, 38, 39,41, 42, 43, 44, 45, 46, 47,48, 121.Polzer, C.G., 93.Porter, J. C., 197.Posch, E., 167.Posnjak, E., 9.Postovskii, I. Y., 177.Powell, H. M., 101, 116.Powell, R. E., 10.Pratesi, P., 197.Preece, I. A., 233.Prelog, V., 140, 200.Preobrrashenski, N. A,, 139.Preston, R. W. G., 196,Price, C. C., 134, 136, 186.Price, J. R., 238.Price, W. C., 126.Prigge, R., 216.Pringsheim, P., 80.Priou, R., 189, 190.Pritchard, R. R., 187.Proskurnim, N. F., 204.Pryde, A. W. H., 117.Pummerer, R., 8.Pyle, J. J., 161.Quirk, R. F., 53.Rabat& J., 183.Rabe, P., 206.Raddatz, K. H., 94.Radlove, S. B., 118.Rafael, S., 221.Raichle, K., 187.Randall, J. T., 86.Randall, M., 11.Rao, P. S., 192, 193.Raoul, Y., 238.Rstner, S., 221.Rettschewa, M., 178.Ray, F.E., 117, 118, 186.Ray, R. C., 71.Reay, Q. A., 210.Rebay, A. von, 190.Record, B. R., 9, 222.Reed, G. B., 210.Reeves, R. E., 227.Rehner, J., 157.Reid, E. E., 136.Reimann, A. L., 81.Reinders, W., 50, 58.Reinecke, A., 91.Reitsma, P., 213.Renfiow, W. B., 137.Rennhak, S., 187.Renoll, M. W., 32.Rhoads, C. P., 159.Ri, T,, 49.Richardson, T., 149.Richter, J., 168.Rickert, H. F., 184, 190.Riehe, C. H., 11 1.Riehl, N., 79, 83, 84, 89.Rieveschl, G., 185.Riley, H. L., 99.Ristic, J., 187,Ritchie, P. D., 126.Rittenberg, D., 221.Rivers, T. M., 218.Roberts, K. C., 32.Robertson, A., 194.Robertson, G. J., 123.Robertson, J. M., 96, 99,100, 103, 106, 107, 116.Robinson, A, M,, 177.Robinson, F. A., 231.Robinson, J.R., 20.Robinson, (Sir) R., 141,183, 196, 107.Robson, W., 198.Rodebush, W. Ex., 168.Rodloff, I., 137.Roe, E., 169, 167.Roe, J. W., 228.Roffo, A. E., 100.Rogers, D. A., 93,Rogers, E. F., 202,208,207.Rogers, M. T., 107.Rogler, E., 92.Romig, J. R., 136.Rooksby, H. P., 85.Roscoe, H. E., 60.Rosen, L. V., 121.Rosen, R., 134.Rosenqviet, T., 143.Rosenwald, R. H., 165.Ross, F. E., 68.Roth, F., 39.Roth, R. T., 139.Rothen, A., 116, 120.Rowley, H. J., 160.Rudiach, R., 90.Rudolph, R., 88.Ruttenauer, A,, 86.Ruff, G., 221.RUE, O., 76.Ruggli, P., 139, 174.Rule, H. G., 117.Rump, E., 142.Rundle, R. E., 233.Ruzich, L., 208.Ryan, M. J., 170.Sachse, H., 43.S a c k , Q., 9.Sackville, J. P,, 213.Safi, S.R., 180.Saito, K., 208.Sandin, R. B., 158, 179,Sanigar, E. B., 224.Sarkar, P, B., 145.Sastri, M. V. C., 188.Sat6, S., 87.Sattler, H,, 39.SavostjanovB, M., 57.Saylar, J. E,, 106.Saytor, C. P., 31.Scatchard, G., 36.Schaarschmidt, R., 200.Schallamach, A., 110.Schaum, K,, 68.Schay, G,! 36, 37, 39.Schdanovitsch, E. $., 203.Scbffer, F. E. C., 59.Scheibler, H., 152,Scheiner, K., 76, 91.Schenk, P, W., 77.Schenk, R., 90, 02.Schiedt, B,, 187.Schilling, K., 162.Schink, N. F., 232.Schirmer, F. B., jun., 94.Schleede, A., 86.Schlenk, W., 137.Schlittler, E., 208.Schloemer, A,, 86.Schmeisser, M., 76.Schmidt, H., 17, 18.Schmidt, B., 190.Schmitz, EL, 77.Schmitz-Dumont, O., 77,93.Schneider, H. G,, 133.Schoea, A. L., 59.Schon, K., 156, 177.Schon, M., 86.Schonberg, A., 200.Schoenheimer, R., 221.&holder, R., 76.8ohol1, R., 166, 169, 173,Schomaker, V., 104, 106,Schroeter, O., 138.Schtschukina, M.N., 139.Schubert, A., 200.Schuh, R., 73.Schull, G. M., 230.Schulte, E. D., 186.Schultz, R. F., 186.Schulz, G. V., 10, 24, 27,Schulz, H. J., 196.Schulze, C. C., 196.Schumclker, H. J., 77.Schwab, G., 192.Schwmz, R., 71, 76, 76,Schwerin, P., 223.189.176.107.29, 30.205INDEX OF AU!PECOB8’ NAlKfS. 247Scott, A. D., 120, 237.Scott, 5. R., 33.Scott, W. E., 208.Sedlatschek, K., 90.Seeley, M. G., 234, 235.Seer, C., 166, 176.Seifter, S., 225.Seitz, F., 78, 79, 80, 81, 82,83.Seligman, A. M., 168, 172,175, 177, 178.Sempronj, A., 187.Seppi, L.J., 94.Servigne, M., 86.Seshadri, T. R., 192, 193.Sevag, M. G., 220.Shah, R. C., 192.Sharp, J. G., 210.Sharp,’T. M., 183.Sharrah, P. C., 100.Shavel, J., 205.Sheldrick, G., 149, 171.Sheppard, S. E., 53, 55,57, 58, 69.Sherman, J., 101.Shoppee, C. W., 186.Shorland, F. B., 212.Short, T. M., 183.Short, W. F., 183.Shriner, R. L., 118, 192.Sidgwick, N. V., 115, 118.Signer, R., 23.Silberstein, L., 56.Siller, G . W., 94.Simamouti, H., 207.Simons, E., 194.Simons, J. H., 181.Simonsen, J. L., 187.Simpson, D. M., 118, 126.Singh, A., 116.Singh, B. H., 116.Singh, M., 116.Sinha, .P. C., 71.Siskin, M., 49.Sjollema, B., 213.Skinner, G. S., 138.Slanina, S. J., 135.Slobodskoi, A.Q., 190.Smadel, J. E., 218.Small, A., 227.Smith, A. M., 212.Smith, C. M., 118.Smith, D. M., 129, 139.Smith, H., 99.Smith, M. C., 214.Smith, W. H., 31.Smith, W. MacF., 49.Smits, A., 125.Smolens, J., 220.Smythe, C. V., 224.Snell, E. E., 230.Snyder, H. R., 136.Sobek, E., 145, 148.Sobotka, H., 168.Sohns, F., 145.Solomon, W., 206,Soru, E., 216.Sowa, F. J., 129, 130, 131,Spiith, E., 190, 199, 202.Sparks, W. J., 134.Spedding, F. H., 78.Spiegel-Adolf, M., 11 1.Spielmmn, M. A., 202.Springall, H. D., 165.Stace, H. A., 208.Stamatoff, C. S., 183.Stanley, W, M., 224.Stare, F. J., 213.Stark, O., 177.Staub, A., 174.Staud, C. J., 52.Staudinger, H. A., 8, 10,14, 19, 20, 24, 25, 31.Steckenreiter, F., 89.Steenbock, H., 213.Stegmaier, W., 92.Steiner, W., 52.Steinkopf, W., 195, 197.Sterry, J., 68.Stevels, 5.M., 42, 45.Stevens, G. W. W., 62, 65.Stevens, J. R., 199.Stevens, P. Q., 120.Stevens, W., 199.Steve~tlson, D. P., 106, 107.Stewart, W. L., 213.Stickney, P. B., 25.Stobbe, H., 76.Stock, A. H., 224.Stokes, A. R., 99.Stopher, E. Q., 233.Storbeck, O., 77.Stotz, E., 212.Strange, J. W., 84.Strating, J., 184.Streeck, B., 163, 172, 186.Strong, F. M., 230.Strotzer, E. F., 91.Struve, W. S., 182.Stumpf, P. K., 232.Style, D. W. G., 46.suss, s., 90.Sugasawa, S., 201.Sugden, B., 129.Sugg, J. Y., 217.Sugii, Y., 149.Suginome, H., 207.Sujiyama, N., 149.Sutherland, G. B. B. M.,Svedberg, T., 55, 56, 58.Svedlovsky, T., 218.Swadesh, S., 202.Swain, G., 185.Swann, G.A., 119.Swindells, F. E., 86.Synge, R. L. M., 236, 237.Szent-GyBrgyi, A., 152.Tiinzer, C., 173.Takei, M., 33.135.104.Tamamushi, B., 195, 200.Tamura, K., 206.Tatsita, H., 191.Tatum, E. L., 198.Taylor, A., 99.Taylor, E. L., 213.Ta lor, T. W. J., 115, 118.Teger, U., 74.Terenin, A., 37.Terry, M. C., 222.Theimer, H., 89, 90.Thiele, W. E., 190.Thomas, R. J., 136.Thomas, R. M., 134.Thury, W., 90.Tiede, E., 83, 87.Tilden, E. B., 233.Tinker, J. M., 181.Todd, A. R., 118, 166, 186,Todd, 33. W., 224.Todd, W. R., 214.Tollert, H., 53.Tomlinson, G. H., 150.Topps, J. E. C., 207.Toth, G., 235.Totter, J. R., 236.Toussaint, N. F., 136.Traub, W., 68.Treiber, E., 167.Treloar, L.R. G., 10, 11, 14,Trivelli, A. P. H., 56, 57,Troger, H., 107.Tubandt, C., 64.Tuck, F. L., 46.Tucker, S. H., 196.Tuckett, R. F., 28.Tulus, M. R., 196.Turner, E. E., 117, 136,Turner, L. B., 134.Tyson, F. T., 196.Ubbelohde, A. R., 96, 100.Ugriumov, P. 8., 202.Ullal, R. V., 192.Umezawa, S., 195.Underwood, E. J., 211,218,Ungar, G., 53.Urban, H., 143.Ushakov, M. I., 177.Vadasz, L., 187.Van der Wyk, A. J. A.,Van Zandt, G., 186.Vaughn, T. H., 131, 133.Velluz, L., 189.Venkateswarlu, V., 193.Verbane, J. J., 133.Verwey, W. F., 217.Vetter, H., 156, 168.Villain, E., 87.208.17, 20, 28, 31, 32.64.127.214.26248 INDEX OF AUTHOBS’ NAMES.Vincent, W. B., 103.Virasoro, E., 144.Vogler, H., 92.Vogel, H., 52.Vogler, H.J., 50.Vogt, E., 37.Vogt, R. R., 131, 132, 135.Vollmann, H., 163, 172,180, 186.Volmer, M., 59.Volquartz, K., 27.Vossen, D., 129.Waddell, J., 213.Wagner, C., 54.Wagner-Jauregg, T., 215.Waitkins, G., 75.Walden, P., 120.Waldmann, H., 177, 183,Waldschmidt, A., 224.Wall, E. M., 212.Wallach, O., 118.Wdlbaum, H. J., 101.Wallebrook, J. C. J., 238.Wallingford, V. H., 138,Wallis, E. S., 123.Wallouch, R., 90.Waloff, M., 129.Welter, J. E., 116.Walter, R., 162.Wannowius, H., 186.Werhurst, E., 42, 45, 47.Warren, B. E., 99.Warren, S., 158.Wartenberg, H. von, 76, 77,Warzynski, R. J., 176.Waterhouse, J., 68.Watson, H. B., 120, 123.Webb, J. H., 52, 53, 56, 61,Weddum, A. G., 219.Weevers, T., 238.Wehage, K., 35.Weib, E., 83.Weidel, W., 198.Weidlich, H.A., 173, 184.Weigert, F., 68.196.139.93.62, 63.Weil, A. J., 218.Weil, R., 90.Weinland, C. E., 61.Weinmayr, V., 181.Weiser,.H. B., 87, 88, 101.Weiss, J., 167.Weizmann, A., 174.Weizmann, C., 175, 177.Wells, A. F., 101.Wells, H. E., 213.Wenzel, W., 89.Wenzke, H. A., 129.Werntz, J. H., 133.West, D. W., 85.West, E., 143, 161.Westheimer, F. H., 123.Westlinning, H., 92.Weyl, W. A., 85, 86.Wheeler, T. S., 192.Whitby, G. S., 26, 31, 33.White, E. V., 234.White, P. B., 218.White, P. C., 186.Whitehead, W., 123.Whitfield, I. C., 121.Wibaut, J. P., 202.Wiberg, E., 72.Wiechert, K., 143.Wiechmann, F., 91.Wieghard, C. W., 217.Wieland, H., 198.Wiener, M., 220.Wightman, E. P., 53.Wigner, E., 47.Wildman, S. G., 229.Wilhelm, H., 197.Wilkins, D. T., 78.Wilkinson, S., 118.Willemart, A., 188, 189.Williams, G.; 108.Williams, M. B., 102, 103.Williams, R. J., 230.Willison, A. M., 201.Willstiitter, R., 142.Wilman, H., 99.Wilson, 53.Wilson, A. J. C., 96.Wilson, C. V., 165.Wilson, D. C., 214.Wilson, J. M., 183.Windaus, A., 187.Windemuth, E., 190.Winkler, W., 169.Winstein, S., 121.Winterstein, A., 156, 168,Wislicenus, W., 137.Wisselingh, van, 238.Witkop, B., 198.Witte, W., 190.Wittig, G., 120, 187.Wolf, J., 126.Wolf, K. L., 35.Wolfrom, M. L., 193.Wong, S. C., 216.Wood, A. D., 127.Wood, J. L., 191.Wood, R. G., 108.Wood, R. W., 64.Wood, S. E., 35.Woodcock, D., 149.Woodward, (Miss) I., 96,Woodward, R. B., 140.Wooley, D. W., 231.Wren. H.. 127.238.100, 103.Wright, G. F., 145, 151.Wu, H., 224.Wunderley, H. L., 129, 135.Wyckoff, R. W. G., 103.Wyler, M., 197.Young, J. H., 118.Yu, S. H., 98, 102.Zavgorodnii, S. V., 129.Zechmeister, L., 142, 235.Zemplh, G., 193.Zerbe, K., 157.Zhirov, N. F., 83.Zierold, H., 86.Zimmer, J. C., 133, 134.Zinke, A., 167, 176.Zintl, E., 71, 75, 76, 88, 89,Zisch, W., 37.Zittle, C. A., 217.Zucker, H., 144.Zumbusch, M., 91.Zureda, F., 89.91

ISSN:0365-6217    

DOI:10.1039/AR9423900240

出版商:RSC    

年代:1942

数据来源: RSC

摘要:

INDEX OF SUBJECTS.Acenaphthenone, condensation of, with o-Acenaphthylene from natural gas, 157.Acetic acid, ethyl ester, carbalkoxylation of,Acetylation, catalytic, 129.Acetylenes, catalytic reactions of, 131.trans-2-Acetylcyclohexyl p-toluenesulphon-ate, reaction of, with silver acetate, 122.Acetyloleanolic acid, esterification of, 140.synthesis of, with boron fluoride catalyst,Acid chlorides, volatile, preparation of, 141.Aconite alkaloids, 206.Aconitine, 206.$-Aconitine, 206.Aconitum heterophyllum, akaloids from, 207.Awnitum napdlus, alkaloids from, 207.Awnitum smhalinense, alkaloid from, 207.Awnitum talassicum, alkaloids of, 206.Acridine, hydrogenation of, 199.Acridone, compounds of, with phosphorusAcrylonitrile, reaction of, with fluorene anddE-Alanine, structure of, 105.Algs, auxin content of, 229.lipoid constituents of, 229.Alizarin, metallic lakes of, 166.Alkali metals, atomic reactions of, 36.Alkaloids, aconite, 206.cinchona, 206.Erythrinu, 205.ErythropMeum, 208.hellebore, 207.Lycoris, 205.pyridine, 202.isoquinoline, 203.sabadilla, 207.Benecio, 202.Sohnum, 208.utilisation of, by etiolated plants, 238.chlorophen ylmagnesium bromide, 176,137, 138.130.oxychloride, 199.with cyclopentadiene, 140.Alkylation, catalytic, 134.Alkylguinuclidines, 201.Alpinetin, constitution of, 192.Alpinia chinensis, 192.Anzanitu dm?loides.noisonous constituents * A of, 19%.Amides. substituted, svnthesis of, with boronfluoride catalyst, 130.coefficients of, 236.bases, 136.fluoride, structure of, 103.Amino-acids, acetyl derivatives, partitionAmino-ketones, preparation of, from SchiffAmmonium hmfluosilicate-ammoniumAmpelopsin, 192.iodinated, structure of, 111.Ampelopsis mdimfolia, 192.Amylase, bacterial and malt, 233.Amylolyose, 233.Angiosperms, differentiation of, from gymno-sperms, 143.Anhydrides, sodium enolates of, 137.Animals, trace elements for, 211.Anisotropy, induced, hypothesis of, 126.Anthanthrene, halogenides of, 167.Anthocyanins, utilisation of, by plants, 238.Anthracene, perchlorate, 167.derivatives, synthesis of, 175.group, reactions in, 186.maleic anhydride compound of, 170.photo-oxide, 188.reaction of, with maleic anhydride, 190.reactivity of benzologues of, 161.spectrum and structure of, 169.structure of, 162.Anthracene, 9-amino-, acetyl derivative, and9-nitro-, maleic anhydride compoundsof, 170.9-bromo-, maleic anhydride compound of,169.9 : 10-dichloro-, maleic anhydride com-pound of, 169.synthesis of, 177.agents on, 179.dyes, acid, structure of, 165.structure of, 164.Anthrone, chloro-, 188.Antibodies, H- and O-types, heat stabilityH- and R-types, 224.production of, 220.purification of, 221.reactions of, with antigens, 222.valency of, 223.alcohol-soluble, 21 8.artificial, 218.bacterial, 215.reactions of, with antibodies, 222.valency of, 223.Wassermann, 218.Anthracenes, reactivity of, 190.Anthraquinone, action of Grignard re-of, 222.Antigens, 215.Antisera, precipitation of, with nucleic acids,220.Araban, 234.Arecaidine, synthesis of, 202.Argon, liquid, structure of, 100.Armepavine, 204.Aromatic compounds, polycyclic, 155, 174.Arylamines, restricted rotation in, 119.Asymmetric synthesis, 126.partial, 127.transformation, 125250 INDEX OPAtaxia, enzootic, 213.Atisines, 206.Auxin, association of, with leaf proteins, 229.Avena, coleoptiles, auxin in, 230.Avidin, 230.biydoAza-alkanes, 200.7-Azaoxindole, 196.Bacilhu3 anthrac&s, antigen from, 215.Bacteria, antigens from, 215.B&&m dysenterice, antigem from, ctrti.f?cid,218, 219.Flexner, antigen from, 216.Shiga, antigen from, 215.Bacterium tyghowm, antigens from, 216, 218.Bebeerame, 205.Bebeerine, structure of, 204.4 : 6-Benzacridans, 196.1 : 2-Benzanthracene, antimony trichloridecomplex, 108.derivatives of, 158.oxidation of, 187.struoture of, 164.meeoBenzanthrone, halogenides of, 167.Benzene, bromo-, chloro-, aad iodo-, reactionof, with sodium, 45.Benzene4 : 4 : 6-tri-p-azopheayhrsonic soid,antibody to, 221.Bemoio wid, eeterifbtion of, by propylene,129.Benzoylbemoic mi&, dehydration of, to an-t hraquinones, 17 7.o-Benzoylbmzoic acids, oyclieation of, toanthraquinone derivativee, 176.2-Benzoylfluorene, 9-&-nitro-, optical activ-ity of, 117.Benzoylformic aaid, (-) menthyl ater,conversion of, into active atxolactinicacid, 127.1 : 12-Benzperylene, 173.from coal tar, 157.preparation of, 184.3 : .l-Bemphenanthrene, carcinogenio deriv-atives of, 183.derivatives of, 158.3 : 4-Benz-l-phenanthroic acid, 184.Benzpyrenes, 156.3 : 4-Benzpyrene, derivatives of, 168.diazo-reaction with, 191.structure of, 180.synthesis of, 182.4 : 5-Benzpyrene, synthesis of, 181, 182.Benzretene, constitution of, 184.Bikhaconitine, 206.Biochemistry, 209.Bios substanoee, 231.Biotin, assay of, 230.preparation of, 230.structure of, 105.Boron fluoride, mtalytic activity of, 128.dl-Brazilic acid, 195.Braeilin, synthesis of, 194.Brucdb bronchkeptica, toxins from, 217.Bush sickness, 212.Butane, 2 : 3-dibromo-, reaotion of, withsilver acetate, 121.iUBJECTS.2-Butmo1, 3-bromo-, bromination of, 122.Butea frondosa, colouring matter of, 192.Butein, 238.A''-Butylene, catalytic polymerisation of, 133.ieoButyrylisobutyric acid, ethyl eater, fromethyl isobutyrate, 137.Cabbage, dehydration of, 210.Cadmium halides, reaotion of, with sodium,Calcium aluminate, structure of, 100.Calgon, inhibition by, of complement, 226.Cameras, X-ray powder, 96.d-Camphor, resolution of, 140.Cannabiscitrh, 193.Carbalkoxylation, 138, 139.Carbon, struoture of, 99.Carbon diaulphide, reaction of, with sodium,Carbonic anhydrase, zinc in, 212.Carbonyl oompowds, reagent for, 140.Carboxypeptidase, panareatic, 226.Carcinogenics, hydrocarbon, 157.Cardiolipin from ox heart, 218.Carotenea, 237.Carotenoids of algae, 229.Carrots, dried, 210.Cassaic mid, 208.Cassaidine, 208.Caaeahe, 208.Cassanic acid, 208.Catalysts, boron fluoride, 128.Catalytic alkylation, 134.39.44.with meroury and trichloroaoetio acidadditions, 131.esterification, 129.hydrogenation, stereoohemistry of, 11 7.nitration, with boron fluoride, 136.sulphonation, with boron fluoride, 136.Cathepsin, 228.Cattle, diseases of, deficiency, 212, 213.due to toxic trace elements, 214.Cevadine, 207.Cevanthridine, 207.Cevanthrol, 207.Cevine, 207.Chlorine, reaction of, with methane andsodium, 38.Cholanthrene, preparation of, 183.synthesis of, 178.Choleic acids, 168.Chondrodystrophy, 213.Chromana, polycyclic, 194.Chromatogram, new, 237.Chromatography, use of, in inveetigation ofprotein hydrolysates, 237.Chrysene, derivatives, substitution in, 187.synthesis of, 186.Chymotrypsin, speai&ity of, 228.Cinchona alkaloids, 206.Clayden effect, 64.Clostrirlium wehhii, antigens from, 216.226.a-toxin, opalescence produced in sere by,preparation and purihtion of, 226.e-toxin, 224MDEX OR’ BUBJWXI’S.261Coal tar, hydrocarbons of, 156.Coast disease, 212.Cobalt, as trace element, 211..Complement, 225.Composits, pigment of, 238.Compounds, molecular, 167.Configuration, relative, 124.Configurational relationship, 124.JI-Conhydrine, structure of, 202.Conidendrin, 149.Coniferyl alcohol, 142.Copper, as trace element, 211.compounds in mammalian tiesues, 212.deficiency of, in farm animals, 212, 213.function of, in animals, 213.deficiency of, in farm animals, 212, 213.amount of, fixed by antigen-antibodycomplexes, 226.Coreopsis douglasii, pigment from, 239.Coreopsis grandijkwa, pigment from, 239.Coronene, from coal tar, 167.Cmy&a.!i8 o p h i m r p , ophiocarpine from, 203.Compebacteriurn dipphtheriae, mtigens from,Corynebacterium hojmnnii, antigens from,Cotton-wood.See Populus rnacdoqpli.Coumingidine, 208.Coumingine, 208.Crocoite, structure of, 100.Cro.fuhria grantiana, grantianine from, 203.Crystals, structure analysis of, 96.Crystallography, 95.Cyanoethyl groups, introduction of, intocompounds with active methyleneCyanogen, reaction of, and of its halideswith sodium, 42,Cyanurio trimide, structure of, 106.Cymene peroxide, 190.Dahlia var2bbili.q pigment from, 238.Deamination, 142.Decevinic acid, 207.Dehydroeugenol, 149.Dehydrotetrahydroelliptone, synthesis of,Dmi8 wuhccensi8, constituents of, 194.Denteropyrroles, spectra of, 195.Dextrins, orystalline, from starch, 233.Discridens, 200.1 : 2 : 6 : 6-Dibenzanthracene, antigens con-1 : 2 : 7 : 8-Dibenzanthracene, preparation of,Dibenzphenanthrenes, preparation of, 183,1 : 2 : 3 : 4-Dibenzphe~nthrene~ carcinogenic3 : 4 : 5 : 6.Dibenzphenanthrene, structure of,Diems, eynthesee with, 184.up-Diethyldibenzyl series, stereochemistrysynthesis of, 173.216.216.groups, 140.194.taining, 219.183.184.action of, 159.173.of, 116.carcinogenio aotion of, 168.Diethylmethylamine, di-p-ohloro-, condenn-5 : 10-Dihydroarsanthren, stereoisomeridesDihydrobenzpyrene, 181.Dihydro-#-conhydrinemethine, oxidation of,10 : 10’-Dihydro-9 : 9’-diphenmthrylidene,d-Dihydroheliotridanmethine, synthesis of,Dihydrohexaoene, structure of, 160.Dihydroniquine, 206.9 : 10-Dihydrophenanthrene, spectrum andstructure of, 168.9 : 10-Dimethoxyanthracene, photo-oxidation3 : 7-Dimethoxy-1 : 2 : 5 : 6-dibenzanthra-3 : 3‘-Dimethoxydibenzyl, 6 : 6’-diiodo-, heat-2 : 7-Dimethoxy-9 : 10-dihgdrophenanthrene,1 : CDimethoxy-9 : lO-diphenylanfhrace,9 : 10-Dimethyl-1 : 2-benzanthraoene, 179.4 : 9-Dimethyl-6 : 6-benzthiophanthren, car-1 : 2-DimethylchryseneY carcinogenic action6 : 7=Dirnethylchryeene, 172.NN’-Dimethyldiacriden, fission of, with4 : 6-Dimethylphe1mnthe, synthesis of,Dimethylquinolines, 19 9.9 : 10-Dimethyl-1 : 242’ : 3’-thiopheno)an-Dimroth-Van’t Hoff rule, 125.9 : 10-Di-a-naphthylanthracene nzrsodi-Diopside, structure of, 97.Diphenic acids, substituted, optical aotivityDiphenyl, spectrum and structure of, 168.9 I 10-Diphenylanthracene, mesodichloride,9 : 10-Diphenylanthracene-1 : 5-dicarboxylic@-Diphen yl- a- (9-phenanthry1)succinic an-Diphenylselenium dibromide and dichloride,Diphtheria antitoxin, orystalline, 221.Diseases, defioiency, 211, 212, 213.Dispersion, rotatory, equation for, 125.Dithio-p-isoindigos, 197.Di-ptoluidinoanthrquinones, 166.Drude equation, 125.Dyes, sensitising action of, 52.ations of, 199.of, 119.202.hdogenides of, 167.202.of, 189.quhone, 177.ing of, with copper bronze, 183.183,photo-oxidation of, 189.carcinogenic action of, 168, 189.photoeoxidation of, 189.cinogenic aation of, 158.of, 159.preparation of, 183.sulphur, 200.172.thracene, 158.chloride, 163,in, 119.163.syntheeis of, 178.acid, eaterifloation of, 169.hydride, 174.StNUblUe Of, 107.fission of, 22 1.toxin, 224252 INDEX OF SUBJEOTS.Eggs, dried, 209, 210.Egg white, antibiotin factor in, 230.Egonol, 149.Elements, trace, 210.Elliptone, structure of, 194.Emetine, synthesis of, 201.proteolytic, 226.specificity of, 227.Equation, Drude, 125.Equisetrin, 193.Equisetum arvense, equisetrin in, 193.Erechtites hieracifolia, hieracifoline from, 203.Erysodine, 205.Erysonine, 205.Erysopine, 205.Erysovine, 205.Erythraline, 205.Erythramine, 205.Erythratine, 205.Erythrina alkaloids, 205.Erythrophleic acid, 208.Erythrophleine, 208.Erythrophleum alkaloids, 208.Esters, sodium enolates of, 137.Esterification, catalytic, 129.Ethoxyfiuoboric acid, ethyl ester, 12%Ethylene, catalytic polymerisation of, 133,Ethylene dibromide, reaction of, with sodium,Ethylidene dibromide, reaction of, withEthylphenylcarbinol, reaction of, with hy-Fischer reagent, for determination of acids,Fish, dried, 210.Flames, diffusion, reactions with, 40.dilute, reactions in, 36.nozzle, 37.Flavan, derivatives, 191.Flavanones, synthesis of, 191.Flour, protein from petroleum extracts of,Fluoranthrene from natural gas, 157.Fluorene, chemistry of, and its derivatives,synthesis of, with boron fluoride catalyst,130.40.sodium, 40.drogen bromide, 120.alcohols, anhydrides, and water, 139.235.185.from coal tar, 157.reaction of, with acrylonitrile, 140.structure of, 163, 171.Fluorenone, oxidation of, 187.Fluorenone-2 : 3-dicarboxylic anhydride,Mills-Nixon effect in, 163.Fluorides, complex, structure of, 102.Fluorine, as toxic trace element, 211.diseases due to, 214.Fluorosis, 214.in man, 214.Foods, dehydrated, 209.Fourier series, summation of, 97.machine for, 98.Fumaric acid, trans-trans-methyl ester,structure of, 105.Furans, synthesis of, 195.Gas, natural, polycyclic hydrocarbons &om,Glass, crystal andyses of, 98.Globulin, antibodies from, 220.Globulins, association of antibodies with, 222.j-Glucuronosidoglucoses, coupling of, withserum globulin, 219.Glycogen, identification of, with cupricchloride, 234.Grantianine, 203.Graphite, carbon-carbon distances in, 99.Growth substances, 229.Gums, plant, 229.Gutta-percha, structure of, 108.Gymnosperms, differentiation of, from angio-Haemocuprein, 212.Haemocyanin, copper in, 212.Haemoglobin, structure of, 111.Hcemophilus parapertussis, toxins from, 21 7.Haemophilus pertussis, toxin from, 217.polyHalides, inorganic, reaction velocities of,Haptens, reaction of, with antibodies, 222.Heliotridan, 202.Hellebore alkaloids, 207.Helminthic diseases, 213.Hemicelluloses, 233.Hepatocuprein, 212.Heptacene, and its hydrides, 161.Heptacenequinone, tetruhydroxy-, 161.Herschel effect, 51, 64.Heteratisine, 207.Heterocyclic compounds, 191.Hetisine, 207.Hexacene, structure of, 160.Hexahydrobenzoyl chloride, 140.Hexamethylethane, structure of, 104.cycloHexane, reaction of, with benzoyl per-cycEoHexane, 1 : 2-dibromo-, reaction of,trans-1 : 2-dibromo-, formation of, 122.Hibiscus cannabinus, glucoside from, 193.Hieracifoline, 203.Hunnemanine, 203.Hunnemunnia fumricefolia, hunnemaninefrom, 203.Hydrazinium difluoride, structure of, 102.Hydrocarbons, acetylenic, reactions of,catalysed with boron fluoride in presenceof mercuric oxide, 131.aromatic, alkylation of, by olehs, 135.polycyclic, oompounds of, with poly-carcinogenic, 167.from coal tar, 156.ethylenic, polymerisation of, catalysed byboron fluoride, 133.Hydrogen fluoride, cyclisation of acids bymeans of, 181.peroxide, structure of, and of its ureaderivative, 103.sulphide, reaction of, with sodium, 44.157.serum, antigenic properties of, 220.sperms, 143.42.oxide and oxalyl chloride, 140.with silver acetate, 121.nitro-compounds, 167INDEX OF SUBJECTS.263Hydroxyamino-acids, isolation of, in proteinHypaconitine, 206.Hypaphorine, 205.“ Hyperol,” atructure of, 103.Immunochemistry, 215.Indaconitine, 206.Indene, condensation of, with methylene-anthrone, 185.fi-isoIndigo, 197.Indigotin, colour and structure of, 198.Indogenides, colour and resonance in, 198.Indole, alkylation of, 196.and its carboxylic acids, 141.synthesis of, 196.Indoles, acetylation of, 196.Indophenines, 196.Integerrimine, 203.Intervalency angles, 116.Iodine, in animals, 211.Iris pseudacorus, carotenoids of, 238.Iron, in animals, 21 1.Isatidine, 203.Isatinecine, 203./?-Isoprenesulphone, structure of, 110, 195.Jaconecine, 203.Jervine, 207, 208.Karanjin, 193.Keratin, structure of, 111.8-Keto-esters, preparation of, 139.Ketotetrahydrobenzpyrene, 180.Kidneys, proteolytic enzymes of, 228.Kobusine, 207.Kynurenine, structure of, and its derivatives,198.hydrolysates, 236.Laccase, copper in, 212.I;csctdmcillus casei E, use of, in biotin assay,230.Laevans, 234.Lappaconithe, 206.Larch wood, arabogdactan from, 234.Lariciresinol, 149.Lecithovitellin, reaction of, with a-toxin, 225.Lecksucht, 213.Leuconostoc rnesenteroides, antisera from, 216,217.Lichenin, 234.Lignans, 149.Lignin, analyses of, 144.butanol-, extraction of, 154.constitution of, 142.Erdtman’s views on, 149.Freudenberg’s views on, 145.Hibbert’s views on, 150.cuproxam, 143.hydrogenation of, 152, 153.jute, 143.methanol-, 143.physical properties of, 143.glycol-, 143.Linariu vulgaris, glycosides of, 193.W i n , 193.neolinarin, 193.Liquida, containing non-spherical particles,shear of, 19.Longilobine, 203.Lucigenin, 200.Lumineacence, 48.Lupinus luteus, utilisation of alkaloids by,Luteins, 237, 238.Lycopenes, 237.Lycorenine, 205.Lycorine, 205.Lycoris alkaloids, 205.Maclura pornifera, pigments from, 193.Magnolia fuacata, magnoline from, 204.Magnoline, 204.Magnolol, 149.Malaccol, 194.Maleic anhydride, addition of, to anthracenes,190.Malonic acid, substituted esters of, frommonocarboxylic esters, 138.d-Mandelic acid, (- ))-menthy1 ester, dia-stereoisomerides of, 117.Manganese, as trace element, 211, 213.Mhrasmus, enzootic, 212.Meat, dried, 210.Melamine, structure of, 106.(-)-Menthol, reaction of, with phosphoruspentachloride, 120.Z-Menthyl N-aminocarbamate, as reagentfor carbonyl compounds, 140.Mercury halides, reaction of, with sodium, 39.Mesaconitine, 206.Methane, reaction of, with chlorine andMethane, bromo- and chloro-derivatives,Methoxyanthrone, 188.Methoxyfluoboric acid, and its mercuric salt,methyl ester, 129.10-Methyl-1 : 2-benzanthracene, oxidation of,Methyl bromide, reaction of, with sodium, 45.Methyl a/3-&ibromo-/3-phenylethyl ketone,optically active, formation of, 127.2-Methyl-3-butanone, 2-hydroxy-, 132.20-Methylcholanthrene, carcinogenic actionof, 158.238.sodium, 38.reaction velocities of, 42.128.191.diazo-reaction with, 191.synthesis of, 178.indene, 185.structure of, 169.183.ment of, 124.in various solvents, 117.ation of, 171.Methyleneanthrone, condensation of, withMethylenecydobutane, structure of, 104.4 : 6-Methylene-9 : 10-dihydrophenanthrene,4 : 5-Methylenedioxychrysene, preparation of,y-Methyl-a-ethylallyl alcohol, rearrange-(+)-y-Methyl-n-heptane, rotatory power of,y-8-Methyl-2-naphthylbutyric acid, cyclis264 INDEX OF SUBJECTS.1 -Methylphenanthrene, preparation of, from9 -Met h ylphenant hrene, synthesis of, 1 7 8.2-Methylpyrrolizidine, synthesis of, 201.Methylquinolines, 199.Microscopes, X-ray, 97.Milk, dried, 209.Molecular compounds, 167.rearrangements, 123.weight, viscosity and, 20.Molybdenosis, 214.Molybdenum, as toxic trace element, 211.diseases due to, 214.Muconic acid, trans-trans-methyl ester, struc-Mushrooms, copper in, 212.Myosin, structure of, 11 1.Nagler reaction, 226.Nakuruitis, 213.Napelline, 207.Naphthacene from coal tar, 156.Naphthacenequinone, structure of, 164.Naphthalene, catalytic alkyhtion of, 135.Nickel nitrate ammoniate, structure of, 102.Niobium disilicide, structure of, 101,Niquidine, 206.Niquine, 206.Nitration, catalytic, with boron fluoride, 136.Nitriles, carbalkoxylation of, 139.synthesis of, with boron fluoride catalyst,polyNitro-compounds, compounds of, withNitrogen organic compounda, cyclic, 195.oxides, reaction of, with sodium, 44.Nobiletin, synthesis of, 192.Nucleic acids, precipitation of, with anti-bacterial sera, 220.Nutrition, 209.Oat seeds, constituents of, 234.Octacenequinone, tetrahydroxy-, 161.Octahydropyrrocolines, synthesis of, 202.Octameth ylant hraquinone, 184.Oleanolic aaid, esterification of, 140.Ophioaarpine, 203.O p t i d activity, due to molecular dis-due t o symmetrically-placed deuterium andinduced, 125.retene, 183.ture of, 105.130.axomatic hydrocarbons, 167.symmetry, 110.hydrogen, 119.Orange, Osage.See dlaclura pomijexa.Organic chemiatry, 113.Osajin, 193.Otosenine, 203.Oxidaaee, copper in, 212.Oxindoles, preparation of, 196.Oxygen, reaction of, with sodium, 43.Oxyhexafiuoniobate ions, structure of, 102.Palm od, lipoid pigmentB from, 237.Pantothenic acid, analogues of, growth effectsring compounds of, 191.structure of, 100.of, 231.Pantothenic acid, assay of, 231.Papaver armmiacum, armepavine from, 204.Papaya latex, enzymes of, 228.Pectic substances, 229.Pectolinarin, 193.Pelletierine acetal, synthesis of, 202.Penicillamine, 237.Pentacene, reaction of, with sulphur intrichlorobenzene, 161.structure of, 160.Pentacenequinone, addition of Grignardreagents to, 165.cydoPenf;adiene, reaction of, with acrjlo-nitrile, 140.Pent amet h ylanisole, 1 35.cycZoPentene oxide, reaction of, with sodiumethyl malonate, 123.2 : 3-~ycZoPentenophenanthrene, 175.Pepsin, chemical composition of, 226.crystalline swine, fractionation of, 226.specificity of, 227.Peptidases, intestinal, activation of, 226.Perinaphthenone, basic properties of, 168.halogenides, 167.Perosis, 213.Peroxides, transamdar, formation of, 188.Persicaria hydropiper, persicnrin in, 193.Persicarin, 193.Perylene, from coal tar, 156.Phalloidine, 198.Phenanthrene, derivatives, synthesis of, 1’7.5.y-2-Phenanthrylbutyric acid, cyclisation of,y-3-Phenanthrylbutyric acid, cyclisation of‘,fl-3-Phenanthrylpropionyl chIoride, cyclis-a-(9-Phenanthryl)stilbenes, stereoisomeric,Phenols, catalytic alkylation of, 135.9-Phenylanthracene, structure of, 169.10-Phenyl-1 : 2 : 3 : 4-dibenzphenanthrene,p-Phenylenebisiminocamphor, adsorption of,2-Phenylindole, reduction of, with copperPhenylmethylcarbinol, reaction of, withPhenylpyridylbutadienes, 199.Phosphates, acid, hydrogen bridges in,Phosphorus pentachloride, model of, 101.Phosphorybse, potato, 232.Photographic development, 68.image formation8 in, 49.halogenides, 167.synthesis of, 181.substitution in, 186.synthesis of, 174, 183.181.171.ation of, 171.174.C-methylation of, 141.174.on lactose, 117.chromite, 196.hydrogen bromide, 120.100.effects, 60.emulsions, gelatin-silver halide, latent-impurities in, 55.sensitivity specks in, 67.hypersensitisation, 61MDEX OF SUBJECTS.255Photographic intermittency effect, 63.Iatent image, formation of, 41).materials, reciprocity failure of, 61.sensitivity, effect of temperature on, 62.size of, 55.Photometers, integrating, 96.Photo-oxidation, 189.Photo-oxides, 188.Phthalio acids, condensation of, to anthra-quinones, 177.Phthalocyanines, structure of, 99.Phthalyl chloride as reagent in preparation ofacid chlorides, 141.Picene, preparation of, 183, 184.Picenequinone, structure of, 166.Picylene ketone, structure of, 166.Pigments, 237.Pine disease, 212.Pinoresinol, 149.Piperidines, 199.Plants, etiolated, substances utilieed by, 238.growth substances in, 229.hemicelluloses of, 233.trace elements for, 21 1.Plant pigments, 237.Plant products, 229.Platinum oxide, structure of, 100.Platynecine, 203.Pneumococci, polysaccharidee, combinationof, with antibodies, 224.Polyisobutylene, viscosity of, 20.Polychloroprene, structure of, 108.Polymers, high, solubility of, 27.osmotic pressure of, 9.viscosity of, 19.Polysaccharides, 229.Polystyrenes, mol.wts. of, in chloroforlu, 23.solubility of, 27.a-Polythienyls, 195.Pomiferin, 193.Populus d o u g d i , hemicelluloses from, 234.l’otsssium, reactions of, with halogens andhalides, 37.Potassium fluoroiodate, structure of, 107.Potatoes, polyphenoloxidaseof, copperin, 21 2.Precipitin reactions, theories of, 223.Propargyl halides, structure of, 105.2’-a-isoPropyldiphenyl-2-carboxylic aoid, 2’-a-hydroxy-, brucine salt, 125.Propylene, catalytic polymerisation of, 133.esterification by, of benzoicj acid, 129,130.reaction of, with salicylic acid, 130.Proteins, 235.hydrolysis of, isolation of products from,236.iodinated, structure of, 11 I.stereochemistry of, 115.type 111, polysaccharide from, 217.antibody to, 221.dihydrogen arsenate, 100.Proteinases, attack of peptides by, 227.Protew morgunii, use of, in pantothenic acidProtoligain, 142.P~eudornoms pyocyarnecc, reaction of, withassay, 231.lecitho-vitellin, 225.Pterophine, 203.Purapurine, 208.Pyrans, 199.Pyrene, antimony trichloride complex, 1613.chemistry of, ‘186.chemistry and structure of, 163.from coal tar, 157.from natural gas, 157.oxidation of, 187.Pyrene, l-hydroxy-, 172.1 : 2-Pyrenequinone9 172.Pyrenoylpropionic mid, reduction of, 180.Pyridine alkaloids, 202.Pyrrole, pigment derivatives, 197.Pyrroles, reduction of, cattalysts for, 196.P yrrole-bluee, 197.Pyrrole group, 196.Pyrrolidine, preparation of, 195.Quenohing, 48.Quercetagitrin, 193.epiQuinidine, 206.epiQuinine, 206.Quinolines, Friedel-Crafts reaction with, 199.2 : 3-disubstituted, preparation of, 199.isoQuinoline alkaloids, 203.Radicals, formation of, in reactions ofReactions, fast, measurement of rate of, 36.ionogenic, 46.Retronecine, 202.hydrogenation of, 203.Rhenium disilicide, struoture of, 101.Riboflavin, assay of, 230.Rosmarinecine, 203.Rosmarinine, 203.Rotatory power, theories of, 116.Rubber, equilibrium distribution of, 28.gel and sol forms of, 31.mol. wt.of, 8.molecules, form of, in solution, 20.solubility and fractionation of, 27.solutions, entropy of dilution of, 18.sodium with organic halides, 46.osmotic pressure of, 8.in benzene, 15.in toluene, 16.physical chemistry of, 7.viscosity of, 19.structure of, 8, 108.swelling of, 33.vulcanised, swelling of, 34.Rubicene, synthesis of, 179.Rubidium hwfluogermanate, structure of,Rubrene, structure of, 169.Rubrenes, photo-oxides of, 188.Sabadilla alkaloids, 207.Salicylic acid, reaction of, with propylene,Salvia ofccinalis, seedlings, etiolation of, 238.Scarlatina toxin, 224.Selanthren, structure of, 108.Selenium as toxic trace element, 21 1 ,Selenophens, synthesis of, 195.103.130256 INDEX OF SUBJEUTS.Selenopht hens, 195.Senecic acid, 203.Benecio alkaloids, 202.Sheep, diseases of, deficiency, 212, 213.Silk fibroin, structure of, 220.Silver arsenate, structure of, 100.bromide, photolysis of, 50.halides, crystals, discoloration of, byelectrical and optical properties of, 61,ionic processes in, 54.photoconductivity of, 63.spectra of, absorption, 51.due to toxic trace elements, 214.light, 51.sulphide, in photographic emulsions, 58.reactions of, with cadmium, mercury, andzinc halides, 39.with carbon and hydrogen sulphides,and with nitrogen and sulphur oxides,44.with chlorine and methane, 38.with cyanogen and with its bromide andchloride, 42.with ethylene dibromide, ethylidenedibromide, and trimethylene bromide,40.with halogens and halides, 37.with halogenobenzenes, 45.with hydrogen halides, 42.with methyl bromide, 44.with organic halides, 41.with oxygen, 43.with stannic halides, 39.condensations with, 136.Sodium, fluorescence of, quenching of, 49.Sodium triphenylmethyl, metal enolateSolanum alkaloids, 208.Solanum auriculatum, alkaloids from, 208.SoEanum avicubre, alkaloid from, 208.Solarisation, 64.Solasodine, 208.Solasonine, 208.Solauricidine, 208.Solauricine, 208.Solvents, effect of, on rotatory power, 116.for viscosity measurements, 26.Sophora japonica, glucosides of, 193.Sophorabioside, 193.Sophoraflavonoloside, 193.Sophoricoside, 193.Spleen, proteolytic enzymes of, 228.Squalene, mol.wt, and viscosity of, 24.Staphylococci, enterotoxin and toxin, 224.Staphylococcus aureus, polysaccharides andStarch, action on, of Bacillus macerans, 233.proteins from, 217.maize, structure of, 232.natural and synthetic, 232.structure of, 232.Stearic acids, trihydroxy-, diastereoisomeric,Stereochemistry, 115, 168.Stereoisomerism, 117.Sterols, of algE, 229.cis-Stilbene, conversion of, into trans-118.form, 134.Stilbene, trinitro-, bromination of, 126.Streptococci, haemolytic, polysaccharide of,Streptolyein-0, 224.Strychnine, antigens containing, 219.Substitution in polycyclic aromatic com-Sugars, substitution reactions in, 123.2-(fi-Sulphanilamidoet hyl)pyridine, 19 9.Sulphonamides, antigens containing, 219.Sulphonation, catalytic, with boron fluoride,Sulphur oxides, reaction of, with sodium,Swayback, 213.Swelling, theory of, 33.Tqetes erectu, glucoside from, 193.Talatisamine, 207.Talatisidines, 207.Talatisine, 207.Tantalum disilicide, structure of, 101.Tellurium tetrachloride, structure of, 107.Tetanus, antiserum to, purification of, 221.Tetracarboline-y-carboxylic acids, Adam-kiewicz-Hopkins reaction for, 198.Tetrdeuterothiophen, 195.Tetrahydrobenzpyrene, 181.( +)-Tetrahydrofurfury1 alcohol, 126.y-5 : 0 : 7 : 8-Tetramethyl-2-naphthylbutyricacid, cyclisstion of, 171.1 : 2 : 3 : 4-Tetramethylphenanthrene, car-cinogenic action of, 159.Thioindoxyls, condensation of, with thio-naphthenquinones, 19 7.Thionaphthenquinones, condensation of, withthioindoxyls, 197.Thiophens, synthesis of, 195.Thiophen series, 195.Thiopyrans, 199.Thorium disilicide, structure of, 101.Tin halides, reaction of, with sodium, 39.Tobacco viruses, mosaic, combination of,with its antiserum, 224.molecular structure of, 20.3 - p - Toluenesulphonyl-4-acetyl-6-triphenyl-methyl-2-methyl-a-methylaltroside, re-action of, with alkali, 123.Toluidine-blue, structure of, 165.Toluidine-green, structure of, 165.Tomato viruses, bushy stunt, antigenic pro-Toxins, 224.Trace dements, 210.essential, 21 1.toxic, 211.2 : 4 : 6-Triethylbenzoic acid, esterificationof, 140.Triethylenediaminochromitrichloride, adaorp-tion of, on quartz, 117.Trilobine, 205.6 : 6 : 9-Trimethyl-3-n-amyl-7 : 8 : 9 : 10-tetrahydro-6-dibenzopyran, l-hydroxy-,optically active forms of, 118.2 : 4 : 6-Trimethylbenzoic acid, esterificationof, 140.217.pounds, 186.136.44.perties of, 220INDEX OF SUBJECTS.2572 : 5 : 5-Trimethyl-2-butyl-1 : 3-dioxold-one,Trimethylene glycol, polyesters, structure of,2 : 4 : 6-Triisopropylphenyl isopropyl ether,Trypsin, specificity of, 227.Tryptophan, derivatives, 198.Tryptophan, hydroxy-, structure of, 198.Triptycene, structure of, 170.Tropinonecarboxylic acid, esters, prepar-Truxillamic acids, molecular rearrangementsTungsten, as trace element, 211.Turacin, copper in, 212.Tyrosine, iodination of, 227.Ulva lactw;a, rhamnosan from, 235.Vaccinia virus, antigens of, 218.Vanadium disilicide, structure of, 101.Van’t Hoff’s law, theory of, 12.Vegetables, dried, 209.Velocity of reaction, flame methods, 36, 40.Veratramine, 208.132.108.135.ation of, 139.of, 123.life-period method, 44.Vibrio cholerce, antigens of, 218.Vinylacetylene, methyl alcohol addition to,1 -Vinylanthraquinone, 184.Violaxanthin, 238.Viscosity, intrinsic, 19.mol. wt. and, 20.Vitamin-B, as growth substance, 231.Vitamin-B, as growth substance, 231.Walden inversion, 120.Wheat, Indian, mucilage of, 234.Xanthurenic acid, 198.Yeast, polysaccharide from, 234.Yeast cellulose, 236.Zea, coleoptiles, auxin in, 229.Zeu mays, glycogen in, identification of, 234.Zeaxanthin, 238.Zein, hydrolysis of, 236.Zinc, as trace element, 211.132.solvents for measurement of, 26.compounds in mammalian tissues, 212.deficiency of, in mammals, 214.halides, reaction of, with sodium, 39

ISSN:0365-6217    

DOI:10.1039/AR9423900249

出版商:RSC    

年代:1942

数据来源: RSC