摘要:

J O U R N A L OF THE CHEMICAL SOCIETY. TRANSAC'1'IONS. A. CHASTON CHAPMAW. A. W. CROSSLEY D.Sc. Ph.D. F.R.S. F. G. DONNAN bl.A. Yh.D. F.R.S. HSKNAILD DYER D.Sc. 11. 0. FOILSTEB D.Sc. Ph.D. F.R.S. A. ]!ARDEN P).Sc. Ph.D. F.R.S. T. M. LOWRY D.Sc. F.R.S. J. C. PIIICIP D.Sc. P1i.D. F. L. PYMAN D.Sc. Ph.D. A. Scoi'*r BLA. D Sc. F.R.S. G. SESTBR D.Sc. Ph.L). S. SMILES D Sc. J. F. THORPE D.Sc. Ph.D. F.R.S. &bitor : J. C. CAIN D.Sc. 1'h.D. Sub-@:bitor : A. J. GREENAWAY. &xir;trrnt %jub-&:bitar : CLARENCE SMITH D.Sc. 1916. Vol. CIX. Part I. pp. 1-648. LONDON: GURNEY & JACKSON 33 PATERNOSTER ROW E.C. 1916 PRINTED IN GREAT BRITAIN BY R,ICHARD CLAY & SONS LIMITID, ERUNBWICK Sl'. STAMFORD ST. S.E., AND BUNGAY SUFFOLK CONTENTS.I.-The Preparation of Ethyl Bromide. By ALFRED HOLT . 11.-The Rate of Growth of Bacteria. By ARTHUR SLATOR . 111.-The Study of Binary Mixtures. Part I. The Densities and Viscosities of Mixtures containing Phenol. By ARTHUR BRAMLEY , IV.-Cont,ributionr to the Chemistry of Cholesterol and Coyrosterol. Part 111. The Ozonides of Cholesterol. Part IV. The Action of Bromine on Cholesteryl Benzoate. By CHARLES I)ORI~E and LIONEL ORANGE V.-Studies in Catalysis. Part 111. Prelimiriary Measure-ments of the Infra-red Absorption Spectra of Hydrogen Chloride Potassium Chloride and Methyl Qcetate in Aqueous Solution. By RAPHAEL HEBER CALLOW WILLIAM CUDMORE MCCULLAGH LEWIS and GERALD NODDER 7-1.-Studies in Catalysis. Part IV. Stoicheiometric and Catalytic Effects Due to the Progressive Displacement of one Reactant by Another in the “Acid” Hydrolysis of Methyl Acetate.By ROBERT OWEN GRIFFITR and WILLIAM CUDMORE MCCULLAGH LEWIS . VI1.-“The Propagation of Flame in Mixtures of Hydrogen and Air. The ‘Uniform Movement.”’ By WILLIAM ARTHUR HAWARD and TATSURO OTAQAWA . VII1.-Polymorphism in Halogen-substituted Anilides . By FREDERICK DANIEL CHATTAWAY and GEORGE ROGER CLEMO . . . 1X.-A Synthesis of Flavones. By BROJENDRA NATH GHOSH. X.-Molecular Volumes of the Hyponitrites of the Alkali Metals and Metals of the Alkaline Earths. By PRAFULLA CHANDRA RAY and RAJEXDRALAL DE. . XI.-Nitromercaptides and their Reaction with the Alkyl Iodides. Compounds of the Disulphonium Series. By PRAFULLA CHANDRA RAY XI1.-Studies on the Oxidation of Unsaturated Fatty Oils and Unsaturated Fatty Acic‘s.Part I. The Formation of Acrolein by the Oxidation of Linseed Oil and Linolenic Acid. XII1.-The Colouring Matter of Cotton Flowers Part 111. BY ARTHUR GEORGE PERKIN . XI V.-Non-aromatic Diazonium Salts. Part V. Diazo-deriv-atives of Aminotriazoles. By GILBERT T. MORGAN and JOSEPH REILLP (1851 Exhibition Research Scholar) . . . By ARTHUR HENRY SALWAY . PAGE 1 2 10 46 55 67 83 89 105 122 131 138 145 1.55 PAPERS COMM‘CTNICATED TO THE CHEMICAL SOCIETY. i v CONTENTS. XV.- The Influence of Different Surfaces on the Decomposition of Methane. X V1.-Arriphoteric Metallic Hydroxides. Part 111. Chromium Hydroxide. By JOHN KERFOOT WOOD and VERA KATHLEEN BLACK .XVI1.-The Iiiteraction of Tetrxnitromethane and Potassium Ferrocyanide. By FREDERICK DANIEL CHATTAWAY and JOHN MALTHOUSE HARRISON . XVII1.--Note on the Preparation of Diethylamine. By WILLIAM EDWARD GARNER and DANIEL TPRER . X1X.-Synthesis of Ketoindopyrnnols. By SOSALE GARALAPURY SASTRY and BROJENDRA NATEI GHOSH . XX.-Decompositions of Sodium Diacetarnide and Potassium Acetamide. By JITENDRA NATH RAKSHIT . XXL-The Estimation of Mixtures of Paracetaldehyde and Acetal. By KENNEDY JOSEPH PREVITB ORTON and PHYLLIS VIOLET MCKIE . XXI1.-Derivatives of Glyoxaline-4(or 5)-formaldehyde and Glyoxaline-4(or 5)-carboxylic Acid. A New Synthesis of Histidine. By FRANK LEE PYMAN . XXIIL-The Preparation and Properties of Colloidal Carbon. By PERCY CYRIL LESLEY THORNE .XX1V.-A Product Obtained in the Manufacture of Natural Indigo. By ARTHUR GEORGE PERKIN . XXV.-The Simultaneous Estimation of Carbon and Halogen by the Chromic Acid Method. By PHILIP WILFRED ROBERTSON . XXV1.-The Relation of Position Isomerism to Optical Activity. Part X. The Menthyl Alkyl Esters of Phthalic Acid and its Nitro-derivatives. By JULIUS BEREND COHEN DAVID WOODROFFE and LEONARD AKDERSON . XXVI1.-Compounds of Iron Manganese Lead and the Metals of Group 11. By SPENCER UMMFREVILLE PICKERING . XXVII1.-Organo-derivatives of Bismuth (Supplementary Note). By FREDERICK CHALLENGER . THE RECENT WORK ON X-RAYS AND CRYSTALS AND ITS BEARING ON CHEMISTRY. Lecture Delivered before the Chemical Society on February 3rd 1916. I!y WILLIAM HENRY BRAGG XX1X.-Additive Compounds of s-Trinitrobenzene with IIetero-cyclic Compounds Containing Nitrogen in the Ring.By SOSALE GARALAPURY SASTRY . XXX.-The Periodic Evolution of Carbon BSonoxide. By JOHN STANLEY MORGAN , XXX1.-Chloro- and Bromo-triethylphosphinoacetaldehycle. By By WILFRID ERNEST SLATER . ~ ~ ' l J T A I A M CALDWELL PAGE 160 1G-L. 171 174 175 180 184 186 202 210 215 222 235 250 252 270 374 28 CONTENTS. V PAGE XXXI1.-The Action of Sulphixric Acid on Alloy Steels. By LESLIE AITCHISON . XXXII1.-The Vapour Pressure of Glyceryl Trinitrate (Nitro-glycerin). By ARTHUR R~ARSHALL and GORDON PEACE . XXX1V.-Synthesis of 8-Phenyl-y-benzopyrone and a y-Phen-XXXV. -The Acid-Gelatin Equilibrium. By HENRY RICHARD-ANNUAL GENERAL MEETING , PRESIDENTIAL ADDRESS .OBITUARY NOTICES . XXXV1.-The Study of Binary Mixtures. Part 11. The Densities and Viscosities of lllixtures containing Sub-stituted Phenols. By ARTHUR BRAMLEY . XXXVI1.-The Study of Binary Mixtures. Part 111. Freezing-point Curves. By AETHUR BRAMLEY. . XXXVII1.-The Study of Einary Mixtures. Part IT. Heats of Reaction and Specific Heats. By ARTHUR BRAMLEY . XXX1X.-A Kew Form of Distilling flask together with a Note on Benzyl Benzoate. Ey TEIE EARL OF BERKELEY . XT,.-The Ignition of Mixtures of Methane and Air and Hydrogen and Air by Means of the Impulsive Electric Discharge. By SOSALE GARALAPURY SASTRP . XL1.-An Oxidation Product oE Gallic Acid. By HANS BLEULER and ARTHUR GEORGE PERKIN . XLII. -The Colour of Po1phydror;yanthraqninone Dyes.By DAVID B. MEEK and EDWIN XOY WATSON . XLII1.-A Space Foriiiiila for I3erizene. Part 11. By JOHN NORMAN COLLIE XLIV.-3-Phenanthrol-4-alc~ehy~le. By the late JOHN WALTER SMITH . XLV.-0-Chlorobenzyl Bromide arid its Products of Hydrolysis. XLV1.-Studies in the Phenylsuccinic Acid Series. Part 111. The Optically Active Phenylsucciuic Acids and their Derivatives By HENRY WREX and HOWELL WILLIAMS . XLVI1.-Derivatives of 13romotolylhydrazines. Ey FREDERICK DANIEL CHATTAWAY and GEORGE DUFOUR HODGSON . XLVIIL-3 5-Dibrorno- and 3 5-Dichloro-phenylhydraziae. By FREDERICK DANIEL CHATTAWAY and OSCAR CHARLES anthropyrone. By AMY ROSE WATSON . . . SON PROCTER and JOHN ARTHUR WILSON . By ALFRED GODFREY GORCON h O N A R D . ELLINGTON.258 298 303 307 320 338 369 434 469 496 520 523 529 544 561 568 570 572 582 58 vi CONTENTS. X I,IX.-2-Hydroxy-l-keto-4-methylene-l 4dihydronaphthalene. By HARRY F~TZGIBBON DEAN and MAXIMILIAN NIERENSTEIN L.-The Action G f Water on Cupric Thiocyanate. By JAMES CHARLES PHILIP and ARTHUR BRAMLEY IJ.-Mercury Mercaptide Nitrites and their Reaction with the Alkyl Iodides. Part 11. By PRAFULLA CHANDRA RAY . LI1.-Hydrates of Aluminium Nitrate. By RICHARD SELIGMAN and PERCY WILLIAMS . LII1.-Dyes Derived from Phenanthraquinone. By KSHITISH CHANDRA MUKHERJEE and EDWIN KOY WATSON NEWER STAKDPOINTS IN THE STUDY OF NUTRITIOK. Lecture Delivered before the Chemical Society on Nay l s t h 1916. By FREDERICK GOWLAND HOPKINS .L1V.-The Isomerism of the Oximes. Part VIII. Carbanilino-and Carbethoxy-derivatives of the Oximes and the Mechanism of Isomeric Change in the Oximes and their Derivatives. By OSCAR LISLE BRADY and FREDERICK PERCY DUNN . Part IX. Interaction of Alkalis and Alkali Bromoacetates and Bromopropionatos in Ethyl-alcoholic Solution. By GEORGE SENTER and HENRY WOOD . LV1.-Studies on the Walden Inversion Part 111. The Kinetics and Dissociation Constants of Phenylbromoacetic Acid. By GEORGE SENTER and STANLEY HORWOOD TUCKER . LVI1.-Interaction of Iodine and Thioacetamide in Aqueous and Alcoholic Solutions. By PRAFULLA CHANDRA RAY and MANIK LAL DEY . LVIIL-Conversion of Aliphatic Nitrites into Nitro-compounds. By PARCHANAN NEOGI and TARINICHARAN CHOWDHURI .L1X.-The Constitution of Coal. Ey DAVID TREVOR JONES and RICHARD VERNON WHEELER . . LX.-A Cyclic Theory of the Constitution of Metalairiminee and of Ferro- and Ferri-cyanides. By JOHN ALBERT NEWTON FRIEND . LX1.-The Influence of Iron Pyrites on the Oxidation of Coal. LXI1.-The Coagulation of Colloidal Arsenious Sulphide by Electrolytes and its Relation to the Potential Difference a t the Surface of the Particles. By FRANK POWIS . IAXIII.-Some Xanthone Derivatives and Xanthone Colouring Matters. By SURENDRANATH DHAR . LXIV.-The Esscntial Oil of Cinnccrnoi~tzcm Oliveri (Bail.) or Brisbane Sassafras. By GEORGE WATSON HARGREAVES . LXV.-The Reaction of the Alkyl Kitrites with Pyridine and Quinoline. By CHARLES WESLEY ADDY and ALEXANDER KILLEN MACBETH , , .LV.-Reactivity of the Halogens in Organic Compounds. By THOMAS JAMES DRAKELXY . PAC. E 593 597 603 612 617 629 650 68 1 690 698 70 1 707 715 723 734 7 44 75 1 75 CONTENTS. \ I l l PAGE 1 XVI.-Trime thyl- and Triethyl-sulphonium Nitrites. By CHARLES WESLEY ADDY and ALEXANDER KILLEN MACHETH. LXVI1.-The Relation Between the Chemical Constitution and Colour of Szo-compounds. By ANUKUL CHANDRA SIRCAR . . 757 ANNUAL REPORT OF THE INTERNATIONAL COMMITTEE ON ATOMIC LXVII1.-The Valency of Two Directly Linked Nitrogen LX1X.-The Volatile Oil from the JVood of the Indian Deodar Tree. By OSWALDIGBY ROBERTS . . 791 LXX.-Studies in Catalysis. Part V. Quantitative Ex-pressions for the Velocity Temperature-coefficient and Effect of the Catalyst from the Point of View of the Radiation Hypothesis.Ey WILLIAM CUDMORE MCCULLAGH LEWIS . . 796 LXX1.-Cryptopine and Protopine. By WILLIAM HENRY LXXI1.-An Extension of the Theory of Addition to Con-jugated Unsaturated Systems. Part I. Note on the Constitution of the Salts of 1-Benzylidene-%methyl-1 2 3 4-tetrahydroisoquinoline. By ELLICE ETTIE PEDEN HAMILTON and ROBERT EOBINSON . 1029 LXXIX1.-An Extension of the Theory of Addition t o Con-jugated Unsxtnrated Systems. Part 11. The C-Alkylation of Certain Derivatives of P-Aminocrotonic Acid and the Mechanism of the Alkylation of Xthyl Acetoacetate and Similar Substances. By ROBERT ROBINSON . . 1038 LXX1V.-The Functions of the Higher Valencies. By ARTHUR CLAYTON . . . . 1046 LXXV.-Overvoltage Tables.Part I . Cathodic Overvoltnges. By EDGAR NEWBERY . . 1051 LXXVL-Overvoltage Tables. Part 11. Anodic Overvoltages. By EDGAR NEWBERY . . . . 1066 LXXVI1.-Experiments on the So-called Migration of Atoms and Groups. Part I. The Nitration of p-Iodoanisole and other Iodo-Phenolic Ethers. By GERTRUDE MAUD ROBINSON 1078 Part IV. The Influence of the Solvent on the Sign of the Product in the Conversion of Phenylchloroacetic Acid to Phenylamino-acetic Acid (co?ztinued). By GEORGE SENTER and HARRY DUGALD KEITH DREW . . 1091 LXX1X.-Overvoltage Tables. Part 111. Overvoltage and t h e Periodic Law. By EDGAR NEWBERY . . 1107 LXXX.-The Condensation of Pyrrole-%aldehyde with Ketones. 755 WEIGHTS 1917 . . 777 Atoms. Part I. By BAWA KARTAR SINGH . .780 PERKIN jun. . . 815 LXXVII1.-Experiments on the Walden Inversion. By EVA LUBRZYNSKA . . 111 ... V111 C’ONTENTS PAGE LXXX1.-The Constitution of Carbaniides. Part 111. The Reaction of Urea and of Thiourea with Acetic Anhydride. Note on Potassium Thiourea. LXXXIL-The ‘‘ Cyclic Theory ” of tlie Constitution of Com-plex Inorganic Compounds A Criticism. Ey EUSTACE EBENEZER TURNER . . 1130 LXXXII1.-Molecular Weight Determinations in Bromine by LXXX1V.-The Influence of Solvents etc. on the Rotation of Optically Active Compounds. Part XXI. The Relation-ship of the Rotatory Powers of Ethyl Tartrate isoButyl Tartrate and isoButyl Diacetyltartmte. By THOMAS STEWART PATTERSON . . 1139 LXXXV,-The Influence of Solvents etc. on the Rotation of Optically Active Compounds.Part XXII. Rotation-Dispersion. By THOMAS STEWART PATTERSON . . 1176 LXXXV1.-The Jnfluence of Solvents etc. on the Rotation of Optically Active Compounds. Part XXIII. Anomaloiis Rotation-Dispersion and Dynamic Isomerism. By THOMAS STEWART PATTERSON . . 1304 LXXXVT1.-A Period of Induction in the Dehydration of Some Crystalline Hydrates. By WILLIAN NOI~MAN RAE . 1229 LXXXVIIL-Phthalonic Acid and its Derivatives. By JOSEPH TCHERNIAC . . 1236 LXXX1X.-The Interaction of Aldehydes and Thiocarbamides in the Presence of Acids. By AUGUSTUS EDWARD DIXON and JOHN TAYLOR . 1244 A Comparison of the Activities of Certain Strong Acids. By HARKY MEDFORTII DAWSON and THOMAS J$71LLIAM CRANN . 1263 XC1.-The Quantitative Absorption of Light by Simple In-organic Substances.Part I. The Haloids of the Alkali Metals and Hydrogen. Ey PETER JOSEPH BRANNIGAN and ALEXANDER KILLEN MAUDETH . . 1277 By LEO ALEXANDROVITSCH TSCI-IUGAEV and STANISLAV STANISLAVO-VITSCII KILTINOVIC . . 1386 XCII1.-The Reaction between Me thy1 Iodide and Some Metallic Cyanides. By ERNALD GEORGE JUSTINIAN HARTLEY 1296 SC1V.-Some Reactions Produced by Mercuric Iodide. By ERNALD GEORGE JUSTINIAN HARTLEY . . 1302 XCV.-Evidence Indicating tlie Existence of a New Variety of Fructose. A Reactive Forni oE Methylfructoside. By XCV1.-The Constitution of the Disaccharides. Part I. The By WALTER NORMAN HAWORTH and By EMIL ALPHONSE WERNER 11 20 the Air-current Method. By ROBERT WRIGHT . . 1134 XC.-The Dual Theory of Acid Catalysis. XCII. -Ammoniacal Derivatives of Platinous Nitrite. JAMES COLQUHOUN IRVINE and GEORGE ROUERTSON . . 1305 -Structure of Sucrose. JAMES LAW . 131 CONTENTS. ix PAGE XCV1I.-The Preparation of Cyanamide from Calcium Cyan-XCVII1.-Theory of Vegetable Tanning. By HENRY RICHARDSON XC1X.-The Hydrolysis of Iron Ammonium Alum. By WILLIAM C.-Additive Compounds of Trinitrobenzene. By JOHN JOSEPH C1.-Additive Compounds of s-Trinitrobenzene with Amino-amide. By EMIL ALPHONSE WERNER . . . 1325 PKOCTER and JOHX ARTHUR WILSON . . 1327 NORMAN RAE . . 1331 SUDBOROUGH . .. . 1339 derivatives of Complex Aromatic Hydrocarbons. By SHUNKER TRIMBAK CADRE a i d JOHN JOSEPH SUDBOROUGH . 1349 (311.-The Estimation of Arsenic in Organic Compounds. By ARTHUR JAMES EWINS . . * 1355 CII1.-Overvoltage Tables. Part IV. The Theories of Over-voltage and Passivity. B~~EDGAR NEWBERY . . 135

ISSN:0368-1645    

DOI:10.1039/CT91609FP001

出版商:RSC    

年代:1916

数据来源: RSC

摘要:

2 SLATOR THE RATE OF GROWTH OF BACTERIA. 11.-The Rate of Growth of Bacteria. By ARTHUR SLATOR. CHEMICAL reactions brought about by micro-organisms usually proceed under conditions where development of the organism and changes in the composition of the nutrient medium take place simultaneously. I n a study of the dynamics of such reactions the rate of development of the micro-organism is of fundamental importance. The rate of growth of bacteria has received a considerable amount of attention (Lane-Claypon J . Hygiene 1909 9 239; Penfold and Norris ibid. 1913 12 527; ana others). The usual method of investigation is t o allow small seedings of bactteria t o grow in a suitable1 medium under definite conditions SLATOR THE RATE O F GROWTH OF HACTERIA. 3 and to estirnats the number after given intervals of time by the well-known method of “plating.” It has been shown for many growths of bacteria that if the development takes place unre- strictedly the logarithmic law of increase holds good.If care is taken to prevent any change in the nutrient medium in which the bacteria are growing the bacteria can apparently develop indefi- nitely always multiplying a t the same rate. This can be experi- mentally renlised by making repeatedly sub-cultures in the early stages of growth. If the bacteria are allowed t o develop without such precautions the medium changes in composition and owing either to lack of necessary food or to the production of poisons the rate of growth of the bacteria diminishes and they finally cease t o increase.Another phase of bacterial growth has been observed and thoroughly examined by Penfold and by Ledingham and Penfold ( J . Hygietke 1914 14 215). This is the period of bacterial lag which takes place before the bacteria develop a t the constant maximum rate. Bacterial lag is invariably observed when old cultures are seeded into a fresh nutrient medium. I n an investigation on t h e rate of fermentation by yeast celIs growing in lightly hopped wort it has been shown that rates of growth can be determined in other ways (Slator Bio-Chem. J. 1913, 7 197). DuTing the period of unrestricted growth the constant of growth is given by the’ equation 0.434 k = - ’ log- N’n where the t N number of yeast cells increases from N to N+n in the time t. The value of k can also be obtained by measuring the rate of farmenta- tion over this period of growth.The amount of fermentation is proportional t o the number of yeast cells which have developed, and the rate of fermentation is given by the equation 1 s + s 0.434 k = - 10s __ t s , where S is the amount of sugar fermented when the yeast e l l s grow from a very small seeding t o N and S+s the amount when the growth has reached N + n. The values of 12 determined by the two methods showed good agreement. Attempts have now been made to estimabe the rate of growth of bacteria by methods which do not involve the trouble of ‘‘ plating.” The micro-organism chosen f o r the experiments was an organism producing lactic acid and occurring in malt-wort. I f a sample of wort with its natural infection is kept a t a temperature of 45-50° for a short time these organisms alone develop f o r the temperature is too high to admit of the growth of others.I n the technical production of lactic acid these bacilli are of considerable value, B 4 SLATOR THE RATE O F GROWTH OF RACTERTA. and have been very tlioroughly studied (see ‘‘ Garungsbakterio- logisches Praktikum,” W. Henneberg Berlin 1909 445-462). A pure culture of the organism was isolated and attempts were made t o measure the rate of growth in malt-wort.. Experiments were first mad3 in order t o ascertain if the constant of growth could be determined by measurements of the rate of production of lactic acid. Satisfactory results were iiot obtained for it was found that even small quantities of acid diminish the rate of development of the bacilli and that the growth of the organism becomes retarded before sufficiently accurate titrations of the acid can be obtained.Another method however gave good results. The bacterial content of the medium was determined by the following inetliod. Wort in which the bacilli have developed in sufficient an!ount t o cause turbidity shows on qitation 2 characteristic “ silky ’’ appear- ance. If finely divided asbestos is shaken with water a similar “silky” appearance is noticed. I n each case the “silkiness” is caused by the thread-like form of the’ body causing the turbidity. It is possible t o match the turbidity of an aqueous suspension of asbestos with one of the bacilli suspended in the medium in whicli it is growing.By preparing a number of standard asbest’os tubes the relative coiicentration of bacilli in a liquid can be esti- mated by comparison with the standard tubes. These tubes con- tained asbestos in the ratio of amounts 1 J2 :2 :24-2 :4 with the quantity in the first standard adjusted so that the turbidity was just visible. With this method of determining bacterial amounts rates of growth can easily be measured. As an example the measurement of the rate of growth a t 45O of these bacilli can be given. An active growth of the bacilli was obtained by growing the organism in test-tubes of wort (D 1.040) kept a t 45O and re-inocu- lating into fresh wort as soon as thel growth became visible. It was found convenient t o seal up the test-tubes and as a t the beginning of the experiments i t was thought that air retarded the growth of the organism all the tubes were exhausted before sealing.This precaution was subsequently found not to be necessary. When an active growth was obtained it was matched against one of the standards and then diluted with fresh wort to dilutions varying between 1 / 100th and 1 / 100,000th of the original concentration. The tubes were then sealed up and placed in a thermostat atl 4 5 O . The time required for the bacilli in the different tubes t o develop t o an amount measurable by one of the standards was observed. From tlie time ( t ) and the increase ( N + n / A T ) the value of 12 is easily calculated. The following result was obtained SLATOR THE RATE OF GROWTH OF BACTERIA.5 TABLE I. t N + n hours. N 0.434 k 2-53 119 0.82 3-48 800 0.83 4-00 2,260 0-84 4.73 9,060 0.84 5.02 18,100 0.85 0.84 - log 2 Gene.ration time= - ~ 0 . 3 5 8 hour =I 21.5 minute$. 0.84 The method of measuring bacterial amounts is only moderately accurate but satisfactory values of Ic are obtained if the increase is large. Lower values of 7c which increased during the time of the experiment were obtained when the seeding was not grown several times a t 4 5 O before the measurements were made. The Influence of Temperature OH th3e R a t e of Growth. It was found that consistent values of 7i were obtained at other temperatures if the organism was first grown several times a t the given temperature before the measurements were carried out. If this precaution was not" taken lower and variable values of I% were obtained.Thus the organism grows at a rate 0.184 at 30° if the organism is seeded from a growth developing a t 30° but if seeded from one growing a t 45O values increasing from 0.110 t o 0.150 were obtained. The measurements were carried o u t over the whole rang? of tem- peratures a t which i t is possible for the organism to grow. A t lower temperatures the experiments were made under strictly sterile conditions; a t higher ones this care is not necessary. The results are given in the following table: TABLE 11. Temperature. 58.0' 55.0 52.5 50.0 45.0 !%+-n t N No growth. Slight growth. 5.5 614 6.5 5,220 10.3 317,000 3.1 614 3.9 5,220 4-2 31,500 5.4 270,000 See Table 1.0.434k from repcat 0.434 E . experiments. 0.51 0.57 0.64 Mean 0.64 0-90 0-95 1-07 1.01 Mean 0.98 1.04 Mean 0-84 0.080-0.09 6 SLATOR THE RATE OF GROWTH OF BACTERIA. Temperature. t. 40-0 4.0 5.25 6.0 6-8 35.0 30.0 25.0 6-6 8.5 9-6 11.6 18.0 21.0 25.0 29.0 43.0 49.5 22-0 90.0 20.0 No growth N+n N 305 1,550 5,530 23,400 195 830 4,180 29,800 1,840 7,500 46,80 277 3,310 19,900 126 TABLE I1 (contirzued). 0.434 k from repeat 0.434 k. oxp erirnenb. 0.62 0-61 0.62 0.64 Mean 0.62 0.36 0.345 0.38 0.385 Meall 0.365 0.181 0.185 0.187 Mean 0-184 0.084 0.082 0.087 Mean 0.084 0.02 0-081 It will be seen from these figures and from the accompanying curve that the influence of temperature on the rate of growth is very great.A t 203 no growth takes place; a t 2 2 O slow growth; and a t 25O 0.434k=0.083. The value of k increases steadily up to a temperature of about. 50° when the maximum rate of growth occurs. A t 52.5O the rate is nearly halved although it is interesting t o note that consistent values were obtained. A t 5 5 O the growth is slight and soon stops; a t 58O no growth takes place. The temperature-quotient of the rate of growth varies therefore, between it very large value' a t about 20° t o a value of 1 at about 50°; i t then becomes less than 1 and finally zero. There is little advantage to be gained by giving these quotients, f o r no doubt the reactions which have a contlrolling influence on the rate of growth a t one given temperature are different from what they a m a t another temperature and the figures therefore, cannot be considered true temperature-coefficients.The I n f l u e m e of Lactic Acid on the R a t e of Growth. If the bacilli arO allowed t o grow indefinitely in wort having I) 1.040 t.he production of laotic acid is about looo (lo equals a SLATOR THE RATE OF GROWTH OF BACTERIA. 7 increase of lV/lOOO-acid). The main part of the acid is produced when the development is greatly retarded f o r even 5 * 6 O acid is sufficient to lower the rate to 0.55 and 11.l0 to 0.35. It was thought possible that if thO bacilli were grown several times in an acid medium such a growth would be less influenced by acid than the normal bacilli.The following table however, FIG. 1. 15 20 25 30 35 40 45 50 55 60 65' Temperature. shows that this is not the case for the values of k for normal bacilli and those f o r bacilli grown five times in wort to which 5 O lactic acid had been added are' almost identical. TABLE 111. Temperature 45". Addition of acid. Normal bacilli. " Acclimatised " bacilli. 0.434 k=0.76 A - 0.434 lc=0.82 A I- 0" - n 9.0 11-1 U'SS 0.53 0.35 (decreasing). -0.29 (decreasing) 8 SLATOR THE RATE OF GROWTH OF BACTEKTA. .4 l'itratiotb illetlbod of Measuriibg t h e R a t e of Growth. Although the constant of unrestricted growth of the bacilli cannot be determined by direct measurement of the rate of produc- tion of acid the value can be determined indirectly (see Slator, Bio-Chem.J. 1913 7 200). The method is best explained by an example. Two quantities of wort were seeded with actively growing bacilli in a measured ratio (300:l in this experiment) and kept a t 4 5 O . The rate of production of acid was measured by titration from time to time of 25 C.C. of the wort with N/lO-sodium hydroxide, using sensitive litmus paper as an outside indicator. It was found that the one reaction preceded the other by a definite time averag- ing in this case 2.88 hours. This is the time for the one seeding t o N w 2 - grow to the' other that is ~ - 300 in a time 2.88 hours ; 0.4341~~ N therefore is equal to 0.86. Another experiment gave the value 0.89. This result is in fair agreement with the value 0.84 given in table I. The curves given in Fig.2 show differences of times at three different stages 2.9 2.9, and 2-85 hours averaging 2.88. Details of the titrations are given in table IV. TABLE IV. Seeding 1. Seeding 2. T-n. - Time. Titration. hours. hours. 0 1 C.C. 0 1 c.0. 1.0 1.0 5.2 1.0 3.0 3.0 6.0 2-2 3-5 3.2 6.5 3.0 5.0 4.7 7-2 4-2 5-2 5.0 7.5 4.5 5.8 5.3 8.0 4.5 6-8 6.0 8.5 5.5 Average difference of time read off from the curves=2.88 hours, Another experiment gave 0.89. I@ue?ice of t h e Concentration of the Wort. Ths rates of growth of the bacilli in worts having D varying between 1*0€0 2nd 1.008 were found to be the same<; the rate of growth is therefore independent of the concentration of the wort over a large1 range. There is no doubt that in the case of even the dilute wort there is food diffusing t o the cells sufficient t ST,ATOR THE RATE OF GROWTE OF BACTERIA.9 maintain the maximum rate of growth for otherwise the rate would be dependent on the concentration of the medium. PenfieId and Norris have shown in the case) of B. typhqsus that if tlie peptone solution in which the bacteria are growing is diluted sufficiently the rate of growth is diminished and that below 0.2 per cent. the time of generation is inversely proportional to the concen- tration of the peptone. I n this case it is possible that diffusion of the food to the cell plays a controlling part in determining the rate of growth of the bacteria. Slator and Saiid (T. 1910 97 922) hav0 made a calculation of tlie limiting conditions under which diffusion alone would just supply yeast cells efficiently with sugar to enable them t o exert their maximu?n fermentative activity.It' was found that a concentra- FIG. 2. Time in hours. tion of the order of 1 milligram per litre was sufficient for the purpose if the enzyme were evenly distributed throughout the cell. If the enzyme were Concentrated in a few points in the1 cell more concentrated solutions would bO necessary. Judging from these calculations very dilute solutions would be efficient in supplying a micro-organism of the size of a bacterium with food f o r growth if the chemical reactions determining the rate of growth take place throughout the whole cell. If such reactions, however take place in a small part of the interior of the cell more concentrated solutions would be necessary.Whether or not diffusion plays a controlling part in the case investigated by Penfield and Norris could probably bO determined by measurements of the temperature-coefficient for the influence of B' 10 BRAMLEY THE STUDY OF BINARY MIXTURES. PART. I. tlemperature on diffusion is much less than on growth measure,d under conditions of large food supplies. Summary. It is pointed out that the rate of development of micro-organisms is an important factor in the study of the dynamics of chemical reactions brought about by such organisms. A convenient and accurate method of estimating rates of growth of bacilli is described. The method is applicable to bacilli which give a “silky ” turbidity in the medium in which they grow.Such a turbidity can be matched against a standard one of asbestos suspended in water. As an example the rate of growth of an organism of the B. delbrucki type growing in malt-wort has been measured. Measuiements have be’en carried out over the whole range of temperatures a t which growth is possible. The rate of unrestricted growth of t h e organism cannot be determined directly by measuring the rate of production of lactic acid by the growing bacilli for even small quantities of acid greatly retard the’ growth. An indirect method however gives results in agreement with the one mentioned above. The measure- ments are carried out in the following manner. Two quantities of wort are seeded the seeding in one case being considerably larger than in the &her.The development of acidity in the first case occurs earlier than in the second the difference in time being that required for the one seeding to grow to the other. The rate of growth is calculated from this difference in time and the ratio of the seedings. BURTON-ON-TRENT. [Received November 12th 19151 2 SLATOR THE RATE OF GROWTH OF BACTERIA. 11.-The Rate of Growth of Bacteria. By ARTHUR SLATOR. CHEMICAL reactions brought about by micro-organisms usually proceed under conditions where development of the organism and changes in the composition of the nutrient medium take place simultaneously. I n a study of the dynamics of such reactions the rate of development of the micro-organism is of fundamental importance. The rate of growth of bacteria has received a considerable amount of attention (Lane-Claypon J .Hygiene 1909 9 239; Penfold and Norris ibid. 1913 12 527; ana others). The usual method of investigation is t o allow small seedings of bactteria t o grow in a suitable1 medium under definite conditions SLATOR THE RATE O F GROWTH OF HACTERIA. 3 and to estirnats the number after given intervals of time by the well-known method of “plating.” It has been shown for many growths of bacteria that if the development takes place unre- strictedly the logarithmic law of increase holds good. If care is taken to prevent any change in the nutrient medium in which the bacteria are growing the bacteria can apparently develop indefi- nitely always multiplying a t the same rate. This can be experi- mentally renlised by making repeatedly sub-cultures in the early stages of growth.If the bacteria are allowed t o develop without such precautions the medium changes in composition and owing either to lack of necessary food or to the production of poisons the rate of growth of the bacteria diminishes and they finally cease t o increase. Another phase of bacterial growth has been observed and thoroughly examined by Penfold and by Ledingham and Penfold ( J . Hygietke 1914 14 215). This is the period of bacterial lag which takes place before the bacteria develop a t the constant maximum rate. Bacterial lag is invariably observed when old cultures are seeded into a fresh nutrient medium. I n an investigation on t h e rate of fermentation by yeast celIs growing in lightly hopped wort it has been shown that rates of growth can be determined in other ways (Slator Bio-Chem.J. 1913, 7 197). DuTing the period of unrestricted growth the constant of growth is given by the’ equation 0.434 k = - ’ log- N’n where the t N number of yeast cells increases from N to N+n in the time t. The value of k can also be obtained by measuring the rate of farmenta- tion over this period of growth. The amount of fermentation is proportional t o the number of yeast cells which have developed, and the rate of fermentation is given by the equation 1 s + s 0.434 k = - 10s __ t s , where S is the amount of sugar fermented when the yeast e l l s grow from a very small seeding t o N and S+s the amount when the growth has reached N + n. The values of 12 determined by the two methods showed good agreement.Attempts have now been made to estimabe the rate of growth of bacteria by methods which do not involve the trouble of ‘‘ plating.” The micro-organism chosen f o r the experiments was an organism producing lactic acid and occurring in malt-wort. I f a sample of wort with its natural infection is kept a t a temperature of 45-50° for a short time these organisms alone develop f o r the temperature is too high to admit of the growth of others. I n the technical production of lactic acid these bacilli are of considerable value, B 4 SLATOR THE RATE O F GROWTH OF RACTERTA. and have been very tlioroughly studied (see ‘‘ Garungsbakterio- logisches Praktikum,” W. Henneberg Berlin 1909 445-462). A pure culture of the organism was isolated and attempts were made t o measure the rate of growth in malt-wort..Experiments were first mad3 in order t o ascertain if the constant of growth could be determined by measurements of the rate of production of lactic acid. Satisfactory results were iiot obtained for it was found that even small quantities of acid diminish the rate of development of the bacilli and that the growth of the organism becomes retarded before sufficiently accurate titrations of the acid can be obtained. Another method however gave good results. The bacterial content of the medium was determined by the following inetliod. Wort in which the bacilli have developed in sufficient an!ount t o cause turbidity shows on qitation 2 characteristic “ silky ’’ appear- ance.If finely divided asbestos is shaken with water a similar “silky” appearance is noticed. I n each case the “silkiness” is caused by the thread-like form of the’ body causing the turbidity. It is possible t o match the turbidity of an aqueous suspension of asbestos with one of the bacilli suspended in the medium in whicli it is growing. By preparing a number of standard asbest’os tubes the relative coiicentration of bacilli in a liquid can be esti- mated by comparison with the standard tubes. These tubes con- tained asbestos in the ratio of amounts 1 J2 :2 :24-2 :4 with the quantity in the first standard adjusted so that the turbidity was just visible. With this method of determining bacterial amounts rates of growth can easily be measured. As an example the measurement of the rate of growth a t 45O of these bacilli can be given.An active growth of the bacilli was obtained by growing the organism in test-tubes of wort (D 1.040) kept a t 45O and re-inocu- lating into fresh wort as soon as thel growth became visible. It was found convenient t o seal up the test-tubes and as a t the beginning of the experiments i t was thought that air retarded the growth of the organism all the tubes were exhausted before sealing. This precaution was subsequently found not to be necessary. When an active growth was obtained it was matched against one of the standards and then diluted with fresh wort to dilutions varying between 1 / 100th and 1 / 100,000th of the original concentration. The tubes were then sealed up and placed in a thermostat atl 4 5 O .The time required for the bacilli in the different tubes t o develop t o an amount measurable by one of the standards was observed. From tlie time ( t ) and the increase ( N + n / A T ) the value of 12 is easily calculated. The following result was obtained SLATOR THE RATE OF GROWTH OF BACTERIA. 5 TABLE I. t N + n hours. N 0.434 k 2-53 119 0.82 3-48 800 0.83 4-00 2,260 0-84 4.73 9,060 0.84 5.02 18,100 0.85 0.84 - log 2 Gene.ration time= - ~ 0 . 3 5 8 hour =I 21.5 minute$. 0.84 The method of measuring bacterial amounts is only moderately accurate but satisfactory values of Ic are obtained if the increase is large. Lower values of 7c which increased during the time of the experiment were obtained when the seeding was not grown several times a t 4 5 O before the measurements were made.The Influence of Temperature OH th3e R a t e of Growth. It was found that consistent values of 7i were obtained at other temperatures if the organism was first grown several times a t the given temperature before the measurements were carried out. If this precaution was not" taken lower and variable values of I% were obtained. Thus the organism grows at a rate 0.184 at 30° if the organism is seeded from a growth developing a t 30° but if seeded from one growing a t 45O values increasing from 0.110 t o 0.150 were obtained. The measurements were carried o u t over the whole rang? of tem- peratures a t which i t is possible for the organism to grow. A t lower temperatures the experiments were made under strictly sterile conditions; a t higher ones this care is not necessary.The results are given in the following table: TABLE 11. Temperature. 58.0' 55.0 52.5 50.0 45.0 !%+-n t N No growth. Slight growth. 5.5 614 6.5 5,220 10.3 317,000 3.1 614 3.9 5,220 4-2 31,500 5.4 270,000 See Table 1. 0.434k from repcat 0.434 E . experiments. 0.51 0.57 0.64 Mean 0.64 0-90 0-95 1-07 1.01 Mean 0.98 1.04 Mean 0-84 0.080-0.09 6 SLATOR THE RATE OF GROWTH OF BACTERIA. Temperature. t. 40-0 4.0 5.25 6.0 6-8 35.0 30.0 25.0 6-6 8.5 9-6 11.6 18.0 21.0 25.0 29.0 43.0 49.5 22-0 90.0 20.0 No growth N+n N 305 1,550 5,530 23,400 195 830 4,180 29,800 1,840 7,500 46,80 277 3,310 19,900 126 TABLE I1 (contirzued).0.434 k from repeat 0.434 k. oxp erirnenb. 0.62 0-61 0.62 0.64 Mean 0.62 0.36 0.345 0.38 0.385 Meall 0.365 0.181 0.185 0.187 Mean 0-184 0.084 0.082 0.087 Mean 0.084 0.02 0-081 It will be seen from these figures and from the accompanying curve that the influence of temperature on the rate of growth is very great. A t 203 no growth takes place; a t 2 2 O slow growth; and a t 25O 0.434k=0.083. The value of k increases steadily up to a temperature of about. 50° when the maximum rate of growth occurs. A t 52.5O the rate is nearly halved although it is interesting t o note that consistent values were obtained. A t 5 5 O the growth is slight and soon stops; a t 58O no growth takes place.The temperature-quotient of the rate of growth varies therefore, between it very large value' a t about 20° t o a value of 1 at about 50°; i t then becomes less than 1 and finally zero. There is little advantage to be gained by giving these quotients, f o r no doubt the reactions which have a contlrolling influence on the rate of growth a t one given temperature are different from what they a m a t another temperature and the figures therefore, cannot be considered true temperature-coefficients. The I n f l u e m e of Lactic Acid on the R a t e of Growth. If the bacilli arO allowed t o grow indefinitely in wort having I) 1.040 t.he production of laotic acid is about looo (lo equals a SLATOR THE RATE OF GROWTH OF BACTERIA. 7 increase of lV/lOOO-acid).The main part of the acid is produced when the development is greatly retarded f o r even 5 * 6 O acid is sufficient to lower the rate to 0.55 and 11.l0 to 0.35. It was thought possible that if thO bacilli were grown several times in an acid medium such a growth would be less influenced by acid than the normal bacilli. The following table however, FIG. 1. 15 20 25 30 35 40 45 50 55 60 65' Temperature. shows that this is not the case for the values of k for normal bacilli and those f o r bacilli grown five times in wort to which 5 O lactic acid had been added are' almost identical. TABLE 111. Temperature 45". Addition of acid. Normal bacilli. " Acclimatised " bacilli. 0.434 k=0.76 A - 0.434 lc=0.82 A I- 0" - n 9.0 11-1 U'SS 0.53 0.35 (decreasing).-0.29 (decreasing) 8 SLATOR THE RATE OF GROWTH OF BACTEKTA. .4 l'itratiotb illetlbod of Measuriibg t h e R a t e of Growth. Although the constant of unrestricted growth of the bacilli cannot be determined by direct measurement of the rate of produc- tion of acid the value can be determined indirectly (see Slator, Bio-Chem. J. 1913 7 200). The method is best explained by an example. Two quantities of wort were seeded with actively growing bacilli in a measured ratio (300:l in this experiment) and kept a t 4 5 O . The rate of production of acid was measured by titration from time to time of 25 C.C. of the wort with N/lO-sodium hydroxide, using sensitive litmus paper as an outside indicator. It was found that the one reaction preceded the other by a definite time averag- ing in this case 2.88 hours.This is the time for the one seeding t o N w 2 - grow to the' other that is ~ - 300 in a time 2.88 hours ; 0.4341~~ N therefore is equal to 0.86. Another experiment gave the value 0.89. This result is in fair agreement with the value 0.84 given in table I. The curves given in Fig. 2 show differences of times at three different stages 2.9 2.9, and 2-85 hours averaging 2.88. Details of the titrations are given in table IV. TABLE IV. Seeding 1. Seeding 2. T-n. - Time. Titration. hours. hours. 0 1 C.C. 0 1 c.0. 1.0 1.0 5.2 1.0 3.0 3.0 6.0 2-2 3-5 3.2 6.5 3.0 5.0 4.7 7-2 4-2 5-2 5.0 7.5 4.5 5.8 5.3 8.0 4.5 6-8 6.0 8.5 5.5 Average difference of time read off from the curves=2.88 hours, Another experiment gave 0.89.I@ue?ice of t h e Concentration of the Wort. Ths rates of growth of the bacilli in worts having D varying between 1*0€0 2nd 1.008 were found to be the same<; the rate of growth is therefore independent of the concentration of the wort over a large1 range. There is no doubt that in the case of even the dilute wort there is food diffusing t o the cells sufficient t ST,ATOR THE RATE OF GROWTE OF BACTERIA. 9 maintain the maximum rate of growth for otherwise the rate would be dependent on the concentration of the medium. PenfieId and Norris have shown in the case) of B. typhqsus that if tlie peptone solution in which the bacteria are growing is diluted sufficiently the rate of growth is diminished and that below 0.2 per cent. the time of generation is inversely proportional to the concen- tration of the peptone.I n this case it is possible that diffusion of the food to the cell plays a controlling part in determining the rate of growth of the bacteria. Slator and Saiid (T. 1910 97 922) hav0 made a calculation of tlie limiting conditions under which diffusion alone would just supply yeast cells efficiently with sugar to enable them t o exert their maximu?n fermentative activity. It' was found that a concentra- FIG. 2. Time in hours. tion of the order of 1 milligram per litre was sufficient for the purpose if the enzyme were evenly distributed throughout the cell. If the enzyme were Concentrated in a few points in the1 cell more concentrated solutions would bO necessary. Judging from these calculations very dilute solutions would be efficient in supplying a micro-organism of the size of a bacterium with food f o r growth if the chemical reactions determining the rate of growth take place throughout the whole cell.If such reactions, however take place in a small part of the interior of the cell more concentrated solutions would be necessary. Whether or not diffusion plays a controlling part in the case investigated by Penfield and Norris could probably bO determined by measurements of the temperature-coefficient for the influence of B' 10 BRAMLEY THE STUDY OF BINARY MIXTURES. PART. I. tlemperature on diffusion is much less than on growth measure,d under conditions of large food supplies. Summary. It is pointed out that the rate of development of micro-organisms is an important factor in the study of the dynamics of chemical reactions brought about by such organisms. A convenient and accurate method of estimating rates of growth of bacilli is described. The method is applicable to bacilli which give a “silky ” turbidity in the medium in which they grow. Such a turbidity can be matched against a standard one of asbestos suspended in water. As an example the rate of growth of an organism of the B. delbrucki type growing in malt-wort has been measured. Measuiements have be’en carried out over the whole range of temperatures a t which growth is possible. The rate of unrestricted growth of t h e organism cannot be determined directly by measuring the rate of production of lactic acid by the growing bacilli for even small quantities of acid greatly retard the’ growth. An indirect method however gives results in agreement with the one mentioned above. The measure- ments are carried out in the following manner. Two quantities of wort are seeded the seeding in one case being considerably larger than in the &her. The development of acidity in the first case occurs earlier than in the second the difference in time being that required for the one seeding to grow to the other. The rate of growth is calculated from this difference in time and the ratio of the seedings. BURTON-ON-TRENT. [Received November 12th 19151

ISSN:0368-1645    

DOI:10.1039/CT9160900002

出版商:RSC    

年代:1916

数据来源: RSC

摘要:

10 BRAMLEY THE STUDY OF BINARY MIXTURES. PART. I. 111.-The Study of Binary 2Mixtum.s. Pa?-t I. The Densities und Viscosities of Mixtures contuin- ing Phenol. By ARTHUR BRAMLEY. THE study of the physical properties of binary mixtures has engaged the attention for a great many years of a large number of chemists and physicists. It rarely happens that any physical property of a mixture of two substances is exactly the same as that calculateld according to a simple mixture rule but often the difference between the actual and calculated values is small compared with the magni BRAMLEY THE STUDY OF BINARY MIXTURES. PBRT I 11 ture of the property itself. These differences are generally ascribed t o constitutional changes brought about by the interaction of the component substances.When the substances brought together react chemically the physical properties of the product frequently differ enormously from those of the original substances; on the other hand when mixing is not accompanied by chemical action the physical properties of the mixture generally approach closely the values calculated. I n a large number of cases where two liquids mix i t is very difficult t o ascertain whether the mixing is accom- panied by chemical action and still more difficult if not impos- sible t o determine the extent of such action. The study of this last class of mixtures really consists in a critical examination of their physical properties and comparing the values obtained by direct measurement with those calculated according t o some formula.It has long been known that the viscosities of mixtures differ from those calculated to a much greater extent than most other physical properties; in contradistinction from other physical properties the constitutional effect is large compared with the additive value. For this reason the study of the viscosity of pure and mixed liquids has attracted a considerable amount of attention but so far no satisfactory explanation of the effect of the interaction of two liquids on the viscosity is forthcoming. The viscosity-concent'ration curves of liquid mixtures have been classified by Dunst'an and Thole into three divisions according as they deviate but slightly froin a straight line exhibit a maximum, o r a minimum. The first class includes mixtures of liquids that are chemically indifferent towards each other the molecules of which are not associated.I n the second class the constituents of each mixture are supposed t o react chemically forming a compound which is the cause of the high viscosity. The third class consists of mixtures of liquids which do not form a compound with one another and a t least one of which contains associated molecules, the low viscosity of the mixture being attributed to the dissocia- tion of complex into simpler molecules as a result of the interaction of the t8wo substances. This method of classification frequently leaves one in doubt about the category t o which some mixtures belong. No mixture has an exactly linear viscosity-composition curve and it is obviously only necessary tot choose two substances which are1 chemically indifferent towards each other and of about the same viscosity to obtain a mixture having a minimum on its viscosity-composition curve yet the deviation from linearity may be very small.Further there are a considerable number of mixtures which show no minimum on their viscosity-composition curves but the latter lie far below the straight B* 12 BRAMLEY THE STUDY O F BINARY MIXTURES. PART T. lines representing the simple mixture rule. Examples of such are mixtures of phenol with benzene chlorobenzene nitrobenzene and acetone which are described later. The cause of the very great deviations in these cases is undoubt'edly the dissociating effect of the second substance on tlie molecular complexes of phenol. Again, there are many mixtures which have viscosities considerably higher than those calculated according t o the simple mixture formula but their viscosity-composition curves show no maximum.These cases are not provided f o r in the above system of classification. Much importance is attached t o the existence of a maximum on the viscosity-composition curve by Dunstan and his collaborators (T. 1904 85 817; 1905 87 11; 1907 91 83; 1908 93 1919; 1309 95 1556); Tsakalotos (Bull. SOC. cJ~im. 1908 [iv] 3 234), and Faust (Zeitsch. pJiysikal. C'izem. 1912 79 97). As regards the cause of a maximum tlie general consensus of opinion seems t o be t h a t i t is the result of the formation of a compound the latter having a higher viscosity than either of the pure' substances. Nearly all attempts t o determine the composition of the complex, however from the position of the maximum have proved fut'ile.This is only t o be expected since in almost every case the position of the maximum point depends on the temperature. Faust (loc. c i t . ) found that with rise in temperature the maximum point on the viscosity-composition curve for mixtures of acetone and chloroform disappears as also does the minimum point on the acetone-carbon disulphide curve. The mixtures of alkylthiocarbimides and secondary bases examined by Kurnakov and Shemtschushni (Zeitsch. pJ~ysi7iul. Chem. 1913 83 481) are exceptional. The viscosity isotherms f o r these mixtures rise t o a definite cusp a t a concentration corresponding with equimolecular proportions ; more- over the composition of the mixture which has the maximum viscosity does not vary with temperature.Other investigators, notably Findlay (Zeitsch. physikal. Chem. 1909 69 203) and Denison (Trans. Faraday SOC. 1912 8 20) attribute more import- ance to the position of maximum deviation of the viscosity curve from a straight line. Denison examined the viscosity curves of a number of mixtures and found t h a t for each mixture the position of maximum deviation was constant at different temperatures and occurred a t a point corresponding with some simple molecular pro- portion; this he took t o indicate the composition of the compound formed I n a large number of cases Denison's test of maximum deviation fails because1 not only does i t draw no distinction between deviations which are positive and negative that is above and below the straight line but as will be shown subsequently the position of maximurn deviation frequently varies considerably with the tern- BRAMLEY THE STUDY OF BINARY MIXTURES.PART I. 13 perature. Findlay (Zoc. cii.) presumably with the object of obtain- ing conditions approximating t o those of corresponding states, measured the viscosities of a number of mixtures a t temperatures just below their boiling points. His results failed t o reveal any simple relatioiiship bettween either the maximum viscosity or the maximum deviation and the nature1 of the compound formed. Denison’s method of analysing the viscosity-composition curves, although yielding definite results in but few cases suggests another system of classification which appears simpler and preferable to that of Dunstan and Thole.Clnss d .-The viscosity-composjtion curves lie below the straight line joining the viscosities of the pure components. Gloss B.-The viscosity-composition curves lie above the straight line joining the viscosities of the pure components. Jn general class A consists of mixtures of substances which do not react chemically with one another t o form a compound the only effect of the interaction of the two pure substances being one of dissociation of complex into simpler molecules. Since there are no linear viscosity isotherms f o r mixtures this would imply that all liquids contain some more or less associated molecules a conclu- sion for which there is a t present no entirely satisfactory criteyion.Class B consists of mixtures of substances which react chemically, forming compounds with each other. Some exceptions arise as for example mixtures of phenol and acetone; this pair of substances yields viscosity isotherms of the type A although the formation of a compound undoubtedly takes place. There are also a number of rYlixtGres giving inflected viscosity isotherms. A discussion of these exceptions will be given subsequently. The work described below was undertaken with the object of determining what factors influence the viscosit’y of liquid mixtures, and whether it is possible t o ascertain from viscosity measurements the composition of a compound that might exist in a liquid mixture of two substances. F o r this purpose it was decided to determilne the densities and viscosities of mixtures of phenol and several bases of different strengths over as wide a range of temperature as possible.I n order t o be able t o compare the viscosity-composition curves of these mixtures with the general type of curve given by mixtures of phenol and indifferent substances mixtures of phenol with benzene chlorobenzene and nitrobenzene were examined. Two classes only are required: EXPERIMENTAL. I n the determination of the densities of the liquids examined, pyknometers of the Sprengel type were employed. ’ F o r tempera- tures up t o 20° a pyknometer of about 8 C.C. capacity with a bul 14 BRAMLEY THE STUDY OF BINARY MIXTULZES. PAlZ'I' I. on one limb was used. Two plain Sprengel tubes of about 10 C.C.capacity were employed f o r the higher temperatures. Duplicate experiments with diff ererit pyknometers showed that the figure in the fourth decimal place was trustworthy but a general considera- tion of the different factors influencing the determinations of density indicate that this was about the limit of accuracy attained. All the densities are referred t o that of water a t 4O. The viscosities were measured by means of viscometers of the Ostwald type. On account of the great differences in the viscosities of the liquids examined three pairs of viscometers were used. One pair was suit'able f o r liquids of low viscosity another pair for liquids of intermediate viscosity and a third pair for liquids of high viscosity. I n this way the periods of flow f o r the different liquids were of about the same order.The hygroscopic nature of the substances used rendered i t necessary to protect them from the moisture in the air; this was done by attaching tubes containing anhydrous calcium chloride t o the viscometers. The periods of flow were measured by means of two tested stopwatches which marked off time in fifths of a second. I n the calculation of absoIute viscosities Thorpe and Rodgers's value for water a t 25O namely, 0.00891 was taken as standard. It is well known that in the determination of viscosities of liquids generally the maintenance a t a steady temperature of the bath in which the viscometer is placed is of great importance. Few experiments however had been made in the early part of this investigation before i t became apparent that on account of the enormous temperature-coefficients of the viscosity of some of the liquids at the lower temperatures the precision with which the temperatures of the baths must be maintained was of a high order.This will be evident from the following example The viscosity of phenol a t 20° is 0.1104; a t loo it is 0.2010; for a temperature interval of loo the) viscosity has almost doubled; hence if an accuracy of 0.1 per cent. is t o be attained the temperature must not vary more than O * 0 l o . The thermo-regulation of a bath so thatu its extreme variations never exceed this amount is a matter of some delicacy especially so when the t'emperature required is below that of the room. For a constant temperature bath a t Oo a 4-litre beaker was filled with a mixture of clean ice broken into small pieces and distilled water.The following arrangement was fonnd to be suitable for temperatures between about 5 O and t h a t of the room. The bath consist'ed of a rectangular copper vessel with two opposite sides of glass and had a capacity of about 30 litres. Cooling was accomplished in the following way A coil of about 10 metres of lead tubing (1.2 cm. in diameter) was embedded i BRAMLEY THE STWDY OF BINARY MIXTURES. PART I. 15 rougkly broken ice in a tank of about 30 litres capaci€y; one end of this coil was connected t o a water-tap and the other to one end of a second coil of similar tubing having a dozen turns of 7.5 cm. diameter. The other end of this coil which was immersed in the constant-temperature bath was connected to a pipe leading t o a sink.A steady stream of water was run from the tap through the first coil where it was cooled to a temperature near zero; i t then passed through the second coil thus cooling the water in the bath; finally i t escaped to the sink. The speed of the water was adjusted so that it would cool the bath to a temperature two or three degrees below that required the final adjustment of the temperature being made by means of an electrical heater actuated by an electrical tliermo-regulator of the form described by Gouy ( J . Physique 1597 6 479). A full description with diagrams of this excellent regulator is given by Barnes (Phil. Trans. 1902 [A], 199 208). For temperatures between that of the room and 50° the same method was used excepting of course the cooling arrange- ment which was dispensed with.Auxiliary heating by means of a small gas flame was necessary for temperatures above 30°. The water in the tank was kept thoroughly mixed by means of a t8urbine stirrer. Trial runs extending over a full day showed that, by ths above method the temperature of the bath could be kept constant within tlwo or three thousandths of a degree. For the more viscous liquids the1 temperature-coefficient of viscosity dimin- ishes rapidly with rise in temperature ; consequently the very high degree of precision in the maintenance of a constant temperature which is required for the lower temperatures is not so necessary for higher temperatures. The baths used for temperatures above 50° were beakers of 4-litres capacity and ordinary toluene thermo- regulators were employed.Above looo the water in the baths was replaced by glycerol and aniline was used instead of toluene in the regulators. The substances used were purified as follows : PhemZ.-Pure phenol was distilled and the1 middle fraction, which had a constant boiling point was used. Henzen e.-Pure commercial benzene was shaken with successive portions of concentrated sulphuric acid until the latter was no longer coloured; the benzene was then washed several times with water and dried first with calcium chloride and afterwards with sodium. Finally i t was distilled through a fractionating column, the portion of constant boiling point being used. ChZorobenzene.-Kahlbaum’s chlorobenzene was kept over anhy- drous calcium chloride for a few days and then distilled.Its boiling point was constant 16 BRAMLEY THE STUDY OF BINARY MIXTUHES. PART I. Nitrobenzene.-Some nitrobenzene of unknown origin was washed with sodium carbonate solution then with water dried o'ver calcium chloride and distilled. 9 piiliiie.-Commercial aniline was fractionated several times; filially a quantity was obtained which had a constant boiling point. Fro. 1. It had a const,ant boiling point. 0 Weight per cent. of phenol. II. , , chlorobenzene. I. Phenol and benzene. III. , , nitrobenzene. DimethyZarziZine.-Kahlbaum's preparation was redistilled and ths middle fraction of constant boiling point used. p- To1 uidin e D i ph e 1) y la ?n i i i e o ti d D i plz e n y 1 m e thy la rn in e .-These substances obtained from Kahlbaum were redistilled under diminished pressure the first and last portions being rejected.&uinoZi.ne.-Kahlbaum's synthetic quinoline was allowed to remain in contact with solid potassium hydroxide for several days, and then distilled ; a portion of constant boiling point was obtained. f'yridiize.-Purs commercial pyridine was kept in contact wit BRAMLEY THE STUDY OF BINARY MIXTURES. PART I. 17 solid potassium hydroxide for several days and then fractionated by distillation. This process was repeated several times and ulti- mately a quantity was obtained which had a range of boiling point less than onetenth of a degree. 1'11 eiietoZe.-Tliis substance obtained from Kahlbaum was redis- tilled.,4 cetone.-Commercially pure acetone was treated with aqueous potassium permanganate solution and allowed t o remain for three or four days more permanganate solution being added as the colour disappeared. The acetone was distilled off on the steam-bath dried over calcium chloride and finally distilled using a fractionating column. From 2500 C.C. of commercial acetone 1500 C.C. were obtained which had a range of boiling point of less than one-tenth of a degree. Mixtures of phenol with ( n ) benzene (6) chlorobenzene and (c) nitrobenzene were examined a t one temperature only namely, 20O. The values obtained are given in t'ables In I h and Ic and Fig. 1 represents graphically the relations between the viscosity and composition of these mixtures.The density-composition curves for these mixtmes are! almost straight lines; they lie very slightly below the linear. TABLE Ia. 171 ixtzires of Phenol and Benzene. of phenol. at 20". at 20". It had a constant boiling point. Weight per cent. Density Viscosity ~0.00 0-8772 0.00629 6-04 0.8880 0-00683 9.84 0.8949 0.00724 20.01 0.9133 0.00865 32-40 0.9370 0.01126 42.09 0.9549 0.0 1401 53.02 0.9766 0.0191 1 63-65 0.9976 0.02642 74.1 1 1.0194 0.0381 1 83.20 1.0383 0.0535 100-00 1.0762 0.1104 TABLE I b . Mixtures of Pheml and Chlorobeitzene. Weigh per cent. Density Viscosity of phenol. at 20". at 20". i 0.00 1.1051 0.00768 4.93 1.1034 0.00825 8-78 1.1018 0.00888 21.73 1.0980 0.01122 30.43 1.0954 0.01374 38.90 1.0930 0-01673 49.90 1.0898 0.02218 58.15 1.0874 0.02748 71.41 1.0836 0.04070 81.45 1.0806 0.05555 1 oo*oo 1.0752 0.110 18 BKAMLEY THE STUDY OF BINARY MIXTURES.PART 1. TABLE Ic. Weight per cent. of phenol. 0.00 4.16 8.84 18.12 27-41 37-96 49.73 58.64 71.03 84.68 100~00 Densi t,y at 20". 1.202 1 1.1957 1.1888 1.1756 1.1635 1.1495 1.1346 1.1233 1.1085 1.0927 1-0752 Viscosity at 20". 0-01931 0.0 1975 0.0204 1 0.02208 0-02460 0.02845 0.0353 0-0419 0.0540 0.0759 0.1104 The densities and viscosities of mixtures of phenol and aniline were determined for a number of temperatures ranging from 20" t o 125O. Tables I1 and IIa contain the results obtained and the viscosity isotherms are shown in Fig.2. The viscosity-composition curves were drawn on a large1 scale for each temperature and the position of maximum deviation from the straight line joining the viscosities of the two components was deterininecl. The compositioii of the mixtures which show the greatest divergence in viscosity from the values calculated according to the simple mixture rule, together with the corresponding temperatures are given in table 116. These quantities are represent'ed graphically in the small curve inset in Fig. 2. The density curves f o r these mixtures are almost straight lines the densities found being slightly higher than those calculated. The viscosities of mixtures of phenol and aniline have been measured for one temperature namely 3 5 O by Thole, lhnstan and Mussel1 (T.1913 103 1114) URAMLEY THE STUDY OF U N A R Y MIXTUREJ. PART I. 19 FIG. 2. Phenol and aniline. 20 40 60 80 Weight per cent. of phenol BRAMLEY THE STUDY OF BINARY MIXTURES. PART I. 21 Weight per cent. of phenol . 0.00 7.94 15.31 23-34 31-28 39-39 47.56 53-81 62.50 69-52 77.02 85.02 92-28 1 00*00 TABLE IIa. M i x t w e s of PJaenol and Aniline. Densities. Viscosities. 20". 1.0219 1.0276 1.0326 1.0380 1.0434 1.0485 1.0532 1.0569 1.061 1 1.0644 1.0675 1-0704 1.0729 1.0760 125" 0.9288 0.9342 0.9390 0.9436 0.9482 0.9527 0.9671 0.9606 0.9648 0.9690 0.9727 0.9762 0-9795 0.9828 , 20". 0.0428 0.0509 0.0610 0.0735 0.0889 0.1059 0.1215 0.1320 0.1418 0.1447 0.1421 0.1331 0.1221 0.1104 , 126".0.00637 0.00666 0.00693 0.00723 0*00740 0.00770 0-00788 0.00799 0.0081 1 0.008 17 0.008 18 0.00813 0.00797 0.00770 TABLE IIB. Jlixtzcres of Plieiiol and Aniline. Weight per cent. of phenol .. 63.0 62.5 62-0 61-25 60.6 50.5 Temperature ........................ 20" 30" 40" 60" 80" 125" Mixtures of phenetole and aniline were examined a t several temFeratures ranging from Oo to 80°. The figures obtained for the densities and viscosities are given in table 111. The density-com- position curves for mixtures of these substances are straight lines, the values found coinciding with those calculated. Fig. 3 contains the viscosity isotherms %lie character of which i t will be seeen is altogether different from that' of the corresponding curves for mixtures of phenol and aniline 22 BRAMLEY THE STUDY OF BINARY MIXTURES.PART 1 FIU. 3. Aniline and phenetole. Weight per cent. o j phenetole Weight per Mixtures Densities. TABLE I11 of I'henetole and Aniline. cent. of / \ / phenetole. 0" 9.9" 20.2" 29.6" 40.0" 60.0" 80.0" 0" 9.9" 0.00 11-74 21.75 32.30 43.68 54.66 65.69 76.54 88.49 100.0 1.0390 1.0303 1.0214 1.0134 1.0045 0.9872 0.9700 0.1005 0.0631 1.0326 1,0239 1.0149 1.0070 0.9980 0.9809 0.9637 0.0799 0.0523 1.0272 1.0184 1.0094 1.0014 0.9922 0.9748 0,9573 0.0655 0.0443 1.0216 1.0128 1.0037 0.9956 0.9864 0.9688 0.9511 0.0548 0.0375 1.0155 1.0066 0.9974 0.9893 0.9801 0.9623 0.9444 0.0451 0.0321 1.0096 1.0007 0.9916 0,9833 0.9740 0.9560 0.9379 0.0379 0.0276 1.0036 0.9946 0.9854 0.9771 0.9677 0,9497 0.9315 0.03205 0.02385 0.9978 0.9888 0.9795 0.9712 0.9616 0.9434 0.9249 0.02720 0.02075 0.9914 0.9823 0.9729 0.9644 0.9547 0.9363 0.9176 0.02265 0.01780 0.9852 0.9760 0.9666 0,9580 0.9483 0.9298 0.91 10 0.01860 0.0153 24 BRAMT,EY THE STUDY OF BlNhElY MIXTURES.PART I. Tables I V and IVu contain the densities and viscosities of mix- tures of phenol and p-toluidine for several temperatures between 40° and 175O. The densities are somewhat higher than the calcu- lated values ; consequently the density curve is slightly concave towards the composition axis. The viscosity curves are shown in Fig. 4. Ta,ble IVb contains tlie composition of mixtures the viscosi- ties of which show a maximum deviation from the calculated values, together with the corresponding temperatures.These were obtained as already described for phenol and aniline and the small curve inset in Fig. 4 depicts tlie relationship between the two. This mixture has been examined a t 30° by Thole Dunstan and Mussel1 (Zoc. c i t . ) who found that the maximum point on the viscosity-composition curve was a t 63 per cent. of phenol,. Accord- ing t’o the experiments described in this paper the position of the maximum point on the viscosity curve corresponds with a composi- tion of 67 per cent. of phenol and since in this case the position of the maximum does not vary appreciably with the temperature, there appears t o be a discrepancy of about 4 per cent. in the com- position of the mixture of maximum viscosity.Independent experi- ments carried out in this laboratory in which Kahlbaum’s chemicals redistilled before use were employed gave the following results for a temperature of 31.5O: Weight per cent. of phenol ......... 50.6 63.3 66.0 68.8 70.7 79.1 86.0 Viscosity ...... ..... 0.0846 04921 0.0931 0.0931 0.0928 0.0867 0.0803 These figures indicate a maximum viscosity a t a point corre- sponding with a composition of about 67.5 per cent. of phenol, which agrees well with the value found by the aut,hor BRAMLEY THE STUDY OF BINARY MIXTURES PART I. 25 Weight per cent. of pheno2 Weight per cent. of phenol. 0.00 9.85 20.67 29-86 38-57 46.25 55.09 62-70 71.11 80.19 89.76 100*00 TABLE IV. Jf iztvres of Phenol and p-Toluidine.Densities. / A 4 39.9" 59.9" 79.8" 99.9" 125" 0.9703 0.9534 0.9365 0.9189 0.8962 0.9808 0.9640 0.9470 0.9295 0.9068 0.9913 0.9744 0-9574 0.9398 0.9172 1.0004 0.9835 0.9665 0.9488 0.9261 1.0087 0.9919 0.9750 0,9376 0.9348 1.0160 0.9991 0.9820 0.9645 0-9418 1,0239 1.0069 0.9898 0.9723 0.9495 1-0305 1.0135 0.9965 0.9795 0.9567 1.0372 1.0201 1.0031 0.9856 0-9628 1.0441 1.0270 1.0099 0.9924 0.9696 1.0512 1.0340 1.0170 0.9995 0.9766 1.0585 1.0414 1.0243 1.0065 0-9833 39.9" 0~02080 0.02632 0.03352 0.04090 0.0482 0.0543 0-06015 0-0629 0.0627 0.05915 0.0538 0.0479 59.9" 0.01398 0-01649 0.01983 0.02283 0.02564 0.02810 0.03015 0.03115 0.03110 0.02990 0.02780 0.0252 BRAMLEY THE STUDY OF BINARY MIXTURES.PART I. 27 Weight per cent. of phenol. 0.00 16.62 23.1 1 34.42 38-24 45.61 56.31 65.69 76-25 79-43 86.34 100~00 TABLE IVa. Mixtures of Yh,enol mid p-Toluzdine. Densities. Viscosities. -\ 7- 150". 0.8734 0.8898 0-8961 0.9069 0-9106 0.9 172 0.9264 0.9341 0.942 1 0.9443 0.9491 0.9572 175". 0.8502 0.8668 0-8732 0.8842 0.8878 0.8943 0.9034 0.9110 0.9188 0.9210 0-9256 0.9337 150". 0.00491 0.00541 0.00560 0.00594 0.00603 0.00619 0.00636 0.00641 0.00635 0-00630 0.00618 0.00592 175". 0.00423 0.00456 0.00468 0.00490 0.00496 0*00507 0.00517 0.00520 0.00515 0*00612 0*00508 0.00492 TABLE IVb. Mixtzires of Phenol and p-Toluidine. Weight per cent.of phenol.. 61.5 60-0 59.0 58.0 57-0 56.0 55.0 Temperature .............. 39.9" 59.9" 79.8" 99.9" 125" 150" 175" Mixtures of phenol and dimethylaniline were examined a t eight temperatares between loo and 1 8 0 O . The figures obtained are given in tables V and Vu. The viscosity-composition curves are given in Fig. 5. On account of the sinuous nature of these curves, the deviations from t h s straight line are both positive and negative, that is some of the viscositia are greater and some less than the values calculated according to the simple mixture rule. Table V b contains the composition of mixtures the viscosities of which show a maximum positive deviation and the corresponding temperatures. The connexion between the two is shown graphically by the curve inset in Fig.5. The densities of these mixtures are but slightly higher than the calculated values. TABLE Vb. Mixtures of Phenol and Dirnethylaniline. Weight per cent. of phenol.. .......... 87.5 84-0 82.0 80.0 78-0 76.0 73.0 71.5 Temperature ........ 10" 20" 29-8" 40.2' 59.9" 80-0" 126.0" 177 28 BRAMLEY THE SlUDY OF BINARY MIXTURES. PART I. FIU. 5. PJzenol and dimethglaniline. Weight per cent. of phenol Weight per cent. of phenol. 0.00 9.07 17-30 33-94 33.08 40.39 48.27 55.75 62.82 69.97 78.14 85-39 92.76 100-00 TABLE V. Jf ixtzires of Phenol arid Dimethylaniline. Densities. 100 0.9647 0.9753 0.9851 0.9932 1.0041 1.0136 1.0236 1.0327 1.0413 1-0500 1.0595 14678 1.0759 1.0835 20" 0.0564 0.9670 0.9768 0-9849 0.9959 1.0053 1-0150 1.0243 1.0330 1.0416 1.0512 1.0595 1.0676 1.0752 126" 0.8679 0.8776 0.8863 0.S939 0.9040 0.9127 0.9216 0.9302 0.9387 0.947 1 0.9564 0.9648 c.9734 0-9Sl5 177" 0.8225 0.8318 0.8403 0.8472 0.3568 0.8647 0.8733 0.8817 0.8895 0.8978 0.9069 0-9152 0.9237 0.0316 10" 0.01654 0.02076 0.02586 0.03145 0.04 185 0.053 15 0.0696 0.0869 0,1094 0-1347 0,1639 0-1850 0.1964 0.201 Weight per cent.of phenol. 0.00 7.93 16.61 24.60 32.71 41-46 49.19 56.14 63-95 70.99 78.83 86.08 93.19 100~00 TABLE V a . Mixtures of Phenol and Dirnethylaniline. Densit,ies. f A \ / 29.8" 40.2" 59.9" 80.0" 29.8" 0.9482 0.9574 0.9677 0.9772 U.9872 0.9981 1.0076 1,0158 1.0252 1.0329 1.043 1 1.0516 1.0594 1.0668 0.9400 0,9492 0.9593 0.9888 0.9788 o .9 m 0.9990 1.0073 1.0166 1.0243 1.0346 1.0431 1-0509 1.0582 0-9234 0.9325 0.9425 0.9520 0.9619 0.9724 0.981 8 0.8899 0.9991 1.0069 1.0171 1.0256 1.0335 1.0414 0.9070 0.9159 0.9258 0.9352 0,9449 0.9553 0.9645 0.9726 0-9817 0.9895 0.9997 1.0082 1.0162 1.0242 0.01 173 0.01351 0.01629 0.0 1036 0.02315 0.0283 0.0333 0.0393 0.04705 0.05325 0.06025 0-0648 0.068G 0.070 BRAMLEY THE STUDY OF BINARY MIXTURES. PART I. 31 Phenol and diphenylamine mixtures were examined at four temperatures between 30° and 8l0 and the results obtained are given in table VI.The figures in brackets for the viscosities of diplienylamine a t 30° and 40° were obtained by extrapolation. Fig. 6 shows graphically the relationship between the viscosity and cornposition of these mixtures. The density curves in this case FIG. 6. Phenyl and diphenylamine. 0 40 60 Weight per cent. oJ phenol. 3 10 are straight lines. A few mixtures of these two compounds have been examined a t 50° by Thole Dunstan and Mussel1 (Zoc. cit.). Mixtares of phenol and diphenylmethylamine were examined a t six temperatures between loo and 80° and the measurements are recorded in tables VII and VIIa. The density curves are almost straight lines the calculat,ecl densities being slightly higher tha Weight per cent. of phenol.0.0 3 7.87 15.18 23.29 30.87 38.60 46-56 53.43 59.65 68.84 76.60 84.59 92-04 100~00 TAELE VI. Slixtures of Phenol and Diphenylamine. Densities. / h 4 30" - 1.0790 1.0780 1.0769 1.0759 1.0740 1.0738 1.0729 1.0721 1.0709 1.0699 1-0688 1.0678 1.0667 40" - 1.0711 1.0700 1.0689 1.0678 1.0668 1.0656 1.0647 1.0639 1.0626 1.0616 1.0605 1-0595 1.0584 61" 1.0543 1.0533 1.0523 1.0511 1.0501 1.0490 1.0478 1.0470 1.0461 1.0448 1.0438 1.0427 1.0416 1.0405 81" 1.0377 1.0386 1.0356 1.0344 1.0333 1.0321 1.0309 1-0300 1.0292 1.0278 1.0268 1-0256 1:0245 1.0233 30" (0.1357) 0.1257 0.1165 0.1072 0.1003 0.0947 0.0898 0-0861 0.0828 0.0791 0.0765 0.0742 0.0 7255 0.070 ERAMLEY STUDY OF THE BINBRY MIXTURES.PART I. 33 the observed values. Fig. 7 shows the connexion between the viscosities and compositions of these mixtures. FIG. 7. Phenol arid diphenylmetlaylamine. 0.20 0.1 8 0.16 0-14 0.12 & 2 0.10 *+ G? -* P O*OE O*OE 0.04 0.02 C TVeight per cent. OJ phenol. TABLE VII. U i x t w e s of Piieiiol and Diphenylmethylamine. phenol. 9.8' 20.1" 9.8" 20.l0 Weight per Densities. Viscosities. cent. of 7- - 0.00 1.0595 1.0515 0.1096 0.0722 4.92 1.0605 1-0523 0.1090 0.0708 9-45 1.0615 1.0532 0-1097 0.0715 17.18 1.0632 1.0548 0.1145 0-0730 27.69 1.0657 1.0572 0-1226 0.0764 36-92 1.0679 1.0593 0.1306 0.0798 48.42 1.0708 1.0621 0-1411 0.0845 56.87 1.0728 1.0641 0.1502 0.0885 67.10 1.0753 1.0665 0.1614 0.0935 78.79 1.0782 1.0693 0.1752 0.0995 89-30 1.0809 1.0720 0.1877 0.1048 100~00 1.0836 1.0750 0.2010 0.1104 VOL.CIX. TABLE VIIa. Mixtures of Phenol and Diphei2ylmethylamine. Weight per cent. of phenol. 0.00 4.98 10.21 20.04 35.34 49.87 62-12 73.58 82-22 90-05 100*00 Densities. 3 0% 1.0438 1.0449 1.0461 1.0483 1.0519 1.0552 1.0581 1.0607 1.0627 1.0645 40' 1.0359 1.0369 1.0379 1.0401 1.0435 1.0467 1.0495 1.0521 1.0542 1.0560 60" 1.0198 1.0207 1.0217 1.0237 1.0269 1.0301 1.0328 1.0354 1.0373 1.0391 Pi00 1.0040 1.0048 1.0058 1.0076 1.0104 1.0136 1.0164 1.0188 1.0206 1.0223 30" 0.0513 0.0510 0.0513 0,0519 0-0542 0.0570 0.0597 0.0626 0.0650 0.0673 / 1.0668 1.0584 1-0414 1.0242 0.070 BRAMLEY STUDY OF THE BINARY MIXTURES.PART I. 35 The densities and viscosities of mixtures of phenol and quinoline were determined f o r eight temperatures ranging from 9 * 8 O to 1 7 5 O , FIG. 8. Phenol and quinoline. Weight per cent. of phenol. and the measurements are recorded in tableg V I I I and VIIIa. Fig. 8 contains the viscosity curves and the small curve inset shows tho connexion between the temperature and composition of the c Weight per cent. of phenol. 0.00 7.54 14.56 22.73 29.76 37.52 45.08 53.20 60.30 68.21 76.88 83-37 92.06 100*00 TABLE vrrr. Mixtures of Pheizol and Quinoline. Densities. I \ 7 9.8" 20.1" 125" 175' 9.so 1.1004 1.1021 1.1037 1.1056 1.1071 1.1078 1.1074 1,1057 1.1030 1.0994 1.0950 1.0916 1.0875 1.0836 1.0925 1,0944 1.0960 1.0977 1-0992 1.0999 1-0994 1.0977 1.0950 1.0914 1-0869 1.0835 1.0791 1.0750 1.0085 1-0103 1.0119 1.0133 1-0140 1.0139 1.0127 1.0099 1.0065 1.0021 0-9969 0.9927 0.9876 0.9828 0.9673 0.9687 0.9698 0.9707 0.9705 0.9696 0.9678 0.9643 0,9602 0.9550 0.9492 0.9445 0.9389 0.9337 0-04805 0.0619 0.0810 0.1170 0.1677 0.2644 0.3750 0.5065 0.5259 0-4740 0.3836 0.3156 0.2420 0.201 Weight per cent.of phenol. 0.00 7.77 14-92 21.96 29.82 37-14 44-62 52.31 59.89 67.92 75.75 83.49 91.79 100*00 TABLE VIIIa.iMixtures of Quinoline and Phenol. Densities. 29:9' 1.0851 1.0870 1.0886 1-0904 1.0917 1.0924 1.0917 1.0901 1.0874 1.0837 1-0795 1.0756 1.0711 1.0668 40' 1,0773 1.0792 1.0808 1.0823 1-0838 1.0843 1.0837 1.0820 1.0793 1.0755 1.0713 1.0672 1.0628 1-0584 60" 1,0615 1-0635 1.0651 1-0666 1.0679 1.0682 1-0675 1.0658 1.0629 1.0590 1.0548 1.0503 1.0458 1.0414 80' 1-0458 1.0478 1.0494 1-0509 1.0518 1.0521 1.0514 1.0496 1.0465 1-0425 1.0382 1.0335 1.0288 1.0242 29.9" 0.02943 0.03645 0.04495 0.05605 0.07425 0-0965 0.1221 0.1436 0.1480 0-1344 0.1177 0.1010 0.0844 0.070 38 BRAMLEY THE STUDY OF BlNARY MIXTURES. PART I.mixture which has a viscosity deviating t o the greatest extent from the calculated value. The data from which this curve is drawn were obtained from the viscosity curves as previously indicated and these are' enumerated in the follo,wing table : TABLE VIIIb. n-lixtures of Phenol and Quinoline. Weight per cent. of phenol.. . . . . . . . . . 58.0 57.0 56.0 55-5 54.0 53.0 50.5 48.0 Temperature . . . . . . . 9.8" 20.1" 29.9" 40.0" 60.0" 80.0° 125" 175" The broken line curve in Fig. 8 is the density curve for 20*1°. Unlike all the other mixtures described in this paper the density- composition curve f o r mixtures of phenol and quinoline? has a very pronounced maximum. Moreover this maximum does not become less marked with rise in temperature; the density curves for the othm temperatures run very approximately parallel to the one shown in the figure.Mixtures of phenol and pyridine were examined a t seven tem- peratures ranging from loo to l l O o and the results are set out in tables IX and IXn. I n Fig. 9 arel given the viscosity curves. The scale of viscosities for the three highest temperatures 60° 80°, and l l O o is considerably extended in order to bring out more clearly the character of the curves. Inset in Fig. 9 is the curve connecting temperature with the composition of the mixture which shows a maximum positive deviation in viscosity. Data for the construction of this curve are given in table 1x6. TABLE IXb. rllixtzires of PJbenol and Pyra'dine. Weight per cent. of phenol . . 93.0 88.0 84.5 82.0 77-0 73.5 70.0 Temperature .. . . . . . . . 10" 20" 30" 40" 60" 80" 110" The densities of mixtures of these two substances show a con- siderable deviation from the calculated values; this is especially so with those mixtures co,ntaining a greater proportion of phenol. The broken line curve shows the connexion between density and com- position for one temperature namely 20°. The density curves for all the other tempera-tures run closely parallel t80 this BRAMLEY THE STUDY OF BINARY MIXTURES. PART 1. 39 FIG. 9. Phenol and pyridine. Weight per cent. of phenol TABLE IX. Mixtures of Phenol and Pyridine. Densities. / b .. / 20" 30' 40" 60" 80" 110" 20" 30" 0.9819 0.9723 0.9627 0,9424 0.9218 0.8900 0.00941 0.00821 0.9916 0.9820 0.9724 0.9524 0.9324 0.9014 0.01 109 0.00940 1.0005 0.9909 0.9814 0.9620 0.9428 0.91 19 0.01321 0.01106 1.0096 1.0002 0.9909 0.9720 0.9532 0.9234 0.01597 0.01339 1.0187 1.0096 1.0005 0.9821 0.9639 0.9347 0.02010 0.01634 1.0258 1.0167 1.0077 0.9897 0.9718 0.9430 0.02411 0.01963 1-0349 1.0260 1.0173 0.9995 0.9821 0.9535 0.03215 0.02515 1.0429 1.0343 1.0257 1.0083 0.9910 0.9626 0.0437 0.03245 1.0514 1.0426 1.0341 1.0168 0.9995 0.9719 0.05945 0.0429 1.0568 1.0482 1.0397 1.0226 1.0053 0.9779 0.0748 0.0517 1.0620 1.0534 1.0449 1.0279 1.0107 0.9833 0.0907 0.0604 1.0668 1.0583 1.0498 1.0327 1.0155 0.9878 0.1004 0.06625 1.0710 1.0625 1-0540 1.0369 1.0198 0.9923 0.1072 0.06915 1.0752 1.0668 1.0584 1.0414 1.0242 0.9967 0.1104 0.0709 Weight per cent.of phenol 0.00 8-30 16.12 24.36 32.58 38.94 47.13 54.96 63-81 70.49 71.94 85.17 92-45 100~0 BRAMLEY THE STUDY OF BJNARY MIXTURES.PART I. 41 Weight per cent. of phenol 0.00 17-26 26.01 35.14 45.48 51.89 TABLE IXQ. Mixtiires of Phenol and Pyriditle. Density a t 10" 0.9916 1.0101 1.0196 1.0294 1-0412 1.0479 Visc:osit.y at' 10" 0.01108 0.01594 0.02050 0.02675 0.0376 0.0505 lTTeight per cent. of Density phenol. a t 10" 58-46 1.0544 66.99 1-0618 76.81 1.0700 82-86 1.0742 91.89 1.0787 100~00 1.0836 Viscosity a t 10" 0.0680 0-0988 0.1416 0.1644 0.1880 0.2010 Tables X and X n contain a record of densities and viscosities of mixtures of phenol and acetone for five t>emperatures ranging from 9'95O t'o 49.8O.The densities of these mixtures are almost thel same as those calculated; the density curves are almost straight lines. The viscosity curves are given in Fig. 10. TABLE X. Mixtures of Phenol cciid Acetone. Weight per cent. of phenol. 0.00 14-19 26.72 38.06 49.43 57-79 65.22 73.74 78.94 85.39 92.85 100*00 Weight per cent. of phenol 0.00 9.57 19.53 27.70 37.42 44.67 53.79 60-24 67.1 9 74.25 80.76 87.98 92.81 100~00 Densities. 9.95" 20.05" 0.8031 0.7912 0.8425 0.8315 0.8768 0.8662 0.9085 0.8983 0.9406 0.9308 0.9642 0.9547 0.9851 0.9757 1.0090 0.9998 1-0237 1.0146 1.0420 1.0334 1,0623 1.0538 1.0836 1.0752 Viscosities, /- 9.95" 20.05" 0.00360 0.00323 0.00486 0.00429 0.00635 0.005CO 0.00868 0.00755 0.01256 0.01055 0.01688 0.01379 0.02358 0.01853 0.03670 0.02750 0.0495 0-0359 0.0748 0.0497 0.1193 0,0730 0*2010 0.1104 TABLE Xa.Mixtures of Phenol and Acetone. 7 29.8" 0.7799 0.8056 0.8336 0-8572 0.8855 0-9063 0-9330 0-9520 0.9724 0.9935 1.01 15 1.0327 1.0466 1-0668 Densities. 40.1" 0.7676 0.7940 0.8223 0.8460 0.8743 0.8952 0.9220 0-9416 0-9625 0.9837 1.0022 1.0237 1.0378 1.0584 7 49.8" 0.7559 0.7830 0.8116 0.8335 0.8636 0.8846 0.9115 0.9316 0.9530 0.9733 0.9933 1.0150 1.0293 1.0503 7 29.8" 0.00295 0.00360 0.00441 0.0052 1 0.00670 0.00808 0.01058 0.013 19 0.01658 0.021 80 0.02910 0-03915 0-04905 0.07 100 Viscosities - 40.1" 0.00270 0.00328 0.00399 0.00470 0.00590 0.00711 0.00904 0~01101 0.01363 0.01741 0.02230 0.02875 0.03465 0.0474 - 49.8" 0.00248 0.00299 0.00360 0.00422 0.00530 0-00628 0-00794 0.00950 0.01150 0.01425 0*01785 0.02245 0.02615 0.0328 C 42 BRBMLEY THE STUDY OF BINARY MIXTURES.PART JI. 0.20- 0.18- 0.16- 0.14- 0.12- s; s u .* CI 8 0.10- 0.08 0.06 0.04 0.02 FIU. 10. Pheizol and acetone. - - - - - __ - - - - ~ - - - - - 0*00,, Weight per cent. of phenol. Discussion of Results. A general survey of the work described in the'foregoing pages (1) When a substance such as phenol which in the liquid state, leads to the following general conclusions BRAMLEY THE STUDY OF BINARY MIXTURES.PART J. 43 consists largely of associated molecules is mixed with another indif- f erent substance such as benzene chlorobenzene or nitrobenzene, the viscosity of the mixture is considerably lower than that calcu- lated according to the simple mixture rule. The viscosity-composi- tion curve shows a very marked curvature convex towards the axis of composition. (2) When the formation of a compound takes place as for example when phenol is mixed with such bases as aniline p-tolu- idine and quinoline the viscosity of the mixture is in general, higher than that calculated by the simple mixture rule. The viscosity-composition curve is concave towards the cornposition axis. That the high viscosity of these mixtures is caused by the acidic nature of phenol is shown clearly by displacing the hydroxyl hydrogen by an alkyl group thus obtaining a neutral substance, phenetole f o r example which with aniline yields mixtures the viscosity curves of which are convex towards the axis of com- position.(3) The position of the maximum point on the viscosity curves generally varies with the temperature and can therefore give little indication as t o the composition of any compound that might be present. Further in every case described in this paper where there is a positive deviation the composition of the mixture the viscosity of which shows a maximum positive deviation depends on the temperature a t which the viscosities are measured; the nature of the relationship between these two quantities is indicated by the small curves inset in the various figures.This is altogether different from what Denison found. On examining these small curves it will be seen that they are all convex towards the composi- tion axis; they lie entirely on on0 side of the ordinate representing molecular proportions and if produced in the direction of higher temperatures they would approach this ordinate asymptotically. These peculiarities are in all probability the result of an unequal thermal dissociation of the molecular complexes of phenol and of the compound formed the opposing influence of these two kinds of molecular complexes being altered in favour of the compound. This is well illustrated by &he mixtures of pyridinei and phenol. A t loo only a very small portion of the viscosity curve is concave towards the axis of composition.This curve resembles very closely the one for phenol and benzene and were it not f o r the slight inflexion near that end of the curve where the mixtures are rich in phenol it might be taken as a typical case of a mixture of two indifferent su bst'ances one of which contained associated molecules. Several unsuccessful attempts were made to measure the viscosity of pllenol a t Oo with the object of ascertaining whether the inflexiar, c+ 44 BRAMLEP THE STIJnY OF RINARY MIXTURES. ]'ART I. of the viscosity curve would disappear at' this temperature. With rise in temperature the part' of the curve which is concave t'owards the composition axis becomes more pronounced and above 40° a well defined maximum appears.The viscosity curve f o r this mixture a t loo has only about 15 per cent. of its length above the straight line joining the viscosities of the pure components whereas a t l l O o there is only about the same proportion below the corre- sponding straight line. A t the lower temperatures thO effect of the associated phenol molecules on the viscosity is predominant whilst a t the higher temperatures the influence of the pyridine-phenol complexes is more assertive. The development of a maximum on the viscosity curve of a mixture with rise in temperature appears to be rather unusual. Faust. found the opposite effect with mixtures of ether and acetone. The difference between the viscosity curves for mixtures of phenol with aniline and dimetliylaniline may be accounted for by the weaker basic nature of the latter.As regards the viscosity curves for mixtures of phenol with diphenylamine and diphenylmethylamine it will be noticed that in neither case do thO curves fall as far below the straight line joining the viscosities of the pure components as is the case f o r mixtures of phenol and benzene. Since there is no reason to suppose &hat either diphenylamine or diphenylmethylamine has a weaker dissociating action on the associated phenol molecules than has benzene the smaller curvature of thO viscosity curves f o r mixtures of phenol and these two very weak bases may be the result of the counteracting influence of compound-f ormation. The viscosity curves f o r mixtures of phenol and acetone unlike those of phenol and pyridine show no change in charactler with rise in temperature.The associated phenol molecules are pre- dominant throughout. It should be borne in mind however that the range of temperature over which this mixture has h e n examined is relatively small and although as has been shown by Schmidlin and Lang (Ber. 1910 43 2812) compound-formation certainly takes place. the amount of compound formed is un- doubtedly much smaller than in the phenol-pyridine mixtures a t the same temperatures. A number of experiments on the influence of pressure on the viscosity of liquids have been made by Warburg and Sachs (Ann. Physiil. 1884 [iii] 22 518) and Cohen (ibid, 1892 [iii] 45 666), who found that increase in pressure generally causes an increase in viscosity.This increase in viscosity may not be the direct result of the increase in pressure but of the decrease in voluine resulting from the latter. Cohen found that the change in viscosity is no BRAMLEY THE STUDY OF BINARY MIXTURES. PART I. 45 proportional to the increase in pressure. Bogojawlenski and Tam- mann (Zeitsch. phtysikal. Chem. 1898 27 457) investigated the relative effect of pressure on the viscosity and specific volume of water a t Oo and found that they are not proportional. That influ- ences besides mere volume changes play a far more important part in determining the viscosity of a liquid is shown in a most striking manner by the results obtained in this investigation. I n every case studied the density-composition curves for the different tem- peratures run very nearly parallel; in no case is this so with the viscosity curves.A reference t o the mixtures of phenol with quinoline and pyridine will illustrate this. The former of these two mixtures has a marked maximum on the density curves. The maximum increase in density or the greatest contraction in volume produced by mixing takes place a t all temperatures when the mixture contains about 47 per cent. of phenol and further the amount of this contraction is very approximately the same f o r every temperature studied. A glance a t the viscosity curves in Fig. 8 will show how different these are. I n the phenol-pyridine mixtures there is a maximum contraction in volume a t 59 t o 60 per cent. of phenol. I n this case also both the composition of the mixture which shows maximum contraction and the extent of this contraction are almost independent of temperature.Nothing could be more different than the viscosity curves for this mixture; not only is there a lack of parallelism but they change in character from one extreme to the other with rise in temperature. If con- traction in volume resulting from whatever cause were the dominant factor in increasing the viscosity a t least some1 approach to a proportional relative increase in viscosity and volumscontrac- tion might be expected. Experiment shows that this is not the case. I wish to express my sincere thanks to Professor Philip f o r placing the resources of his laboratory a t my disposal and for the kind interest he has taken in the work throughout.THE IMPERIAL COLLEGE OF SCIENCE AND TECHNOLOGY, SOUTH KENSINGTON, LONDON S.W. [Received Nocenzber 1 2 t h 1915 10 BRAMLEY THE STUDY OF BINARY MIXTURES. PART. I. 111.-The Study of Binary 2Mixtum.s. Pa?-t I. The Densities und Viscosities of Mixtures contuin- ing Phenol. By ARTHUR BRAMLEY. THE study of the physical properties of binary mixtures has engaged the attention for a great many years of a large number of chemists and physicists. It rarely happens that any physical property of a mixture of two substances is exactly the same as that calculateld according to a simple mixture rule but often the difference between the actual and calculated values is small compared with the magni BRAMLEY THE STUDY OF BINARY MIXTURES. PBRT I 11 ture of the property itself.These differences are generally ascribed t o constitutional changes brought about by the interaction of the component substances. When the substances brought together react chemically the physical properties of the product frequently differ enormously from those of the original substances; on the other hand when mixing is not accompanied by chemical action the physical properties of the mixture generally approach closely the values calculated. I n a large number of cases where two liquids mix i t is very difficult t o ascertain whether the mixing is accom- panied by chemical action and still more difficult if not impos- sible t o determine the extent of such action. The study of this last class of mixtures really consists in a critical examination of their physical properties and comparing the values obtained by direct measurement with those calculated according t o some formula.It has long been known that the viscosities of mixtures differ from those calculated to a much greater extent than most other physical properties; in contradistinction from other physical properties the constitutional effect is large compared with the additive value. For this reason the study of the viscosity of pure and mixed liquids has attracted a considerable amount of attention but so far no satisfactory explanation of the effect of the interaction of two liquids on the viscosity is forthcoming. The viscosity-concent'ration curves of liquid mixtures have been classified by Dunst'an and Thole into three divisions according as they deviate but slightly froin a straight line exhibit a maximum, o r a minimum.The first class includes mixtures of liquids that are chemically indifferent towards each other the molecules of which are not associated. I n the second class the constituents of each mixture are supposed t o react chemically forming a compound which is the cause of the high viscosity. The third class consists of mixtures of liquids which do not form a compound with one another and a t least one of which contains associated molecules, the low viscosity of the mixture being attributed to the dissocia- tion of complex into simpler molecules as a result of the interaction of the t8wo substances. This method of classification frequently leaves one in doubt about the category t o which some mixtures belong.No mixture has an exactly linear viscosity-composition curve and it is obviously only necessary tot choose two substances which are1 chemically indifferent towards each other and of about the same viscosity to obtain a mixture having a minimum on its viscosity-composition curve yet the deviation from linearity may be very small. Further there are a considerable number of mixtures which show no minimum on their viscosity-composition curves but the latter lie far below the straight B* 12 BRAMLEY THE STUDY O F BINARY MIXTURES. PART T. lines representing the simple mixture rule. Examples of such are mixtures of phenol with benzene chlorobenzene nitrobenzene and acetone which are described later. The cause of the very great deviations in these cases is undoubt'edly the dissociating effect of the second substance on tlie molecular complexes of phenol.Again, there are many mixtures which have viscosities considerably higher than those calculated according t o the simple mixture formula but their viscosity-composition curves show no maximum. These cases are not provided f o r in the above system of classification. Much importance is attached t o the existence of a maximum on the viscosity-composition curve by Dunstan and his collaborators (T. 1904 85 817; 1905 87 11; 1907 91 83; 1908 93 1919; 1309 95 1556); Tsakalotos (Bull. SOC. cJ~im. 1908 [iv] 3 234), and Faust (Zeitsch. pJiysikal. C'izem. 1912 79 97). As regards the cause of a maximum tlie general consensus of opinion seems t o be t h a t i t is the result of the formation of a compound the latter having a higher viscosity than either of the pure' substances.Nearly all attempts t o determine the composition of the complex, however from the position of the maximum have proved fut'ile. This is only t o be expected since in almost every case the position of the maximum point depends on the temperature. Faust (loc. c i t . ) found that with rise in temperature the maximum point on the viscosity-composition curve for mixtures of acetone and chloroform disappears as also does the minimum point on the acetone-carbon disulphide curve. The mixtures of alkylthiocarbimides and secondary bases examined by Kurnakov and Shemtschushni (Zeitsch. pJ~ysi7iul. Chem. 1913 83 481) are exceptional.The viscosity isotherms f o r these mixtures rise t o a definite cusp a t a concentration corresponding with equimolecular proportions ; more- over the composition of the mixture which has the maximum viscosity does not vary with temperature. Other investigators, notably Findlay (Zeitsch. physikal. Chem. 1909 69 203) and Denison (Trans. Faraday SOC. 1912 8 20) attribute more import- ance to the position of maximum deviation of the viscosity curve from a straight line. Denison examined the viscosity curves of a number of mixtures and found t h a t for each mixture the position of maximum deviation was constant at different temperatures and occurred a t a point corresponding with some simple molecular pro- portion; this he took t o indicate the composition of the compound formed I n a large number of cases Denison's test of maximum deviation fails because1 not only does i t draw no distinction between deviations which are positive and negative that is above and below the straight line but as will be shown subsequently the position of maximurn deviation frequently varies considerably with the tern- BRAMLEY THE STUDY OF BINARY MIXTURES.PART I. 13 perature. Findlay (Zoc. cii.) presumably with the object of obtain- ing conditions approximating t o those of corresponding states, measured the viscosities of a number of mixtures a t temperatures just below their boiling points. His results failed t o reveal any simple relatioiiship bettween either the maximum viscosity or the maximum deviation and the nature1 of the compound formed.Denison’s method of analysing the viscosity-composition curves, although yielding definite results in but few cases suggests another system of classification which appears simpler and preferable to that of Dunstan and Thole. Clnss d .-The viscosity-composjtion curves lie below the straight line joining the viscosities of the pure components. Gloss B.-The viscosity-composition curves lie above the straight line joining the viscosities of the pure components. Jn general class A consists of mixtures of substances which do not react chemically with one another t o form a compound the only effect of the interaction of the two pure substances being one of dissociation of complex into simpler molecules. Since there are no linear viscosity isotherms f o r mixtures this would imply that all liquids contain some more or less associated molecules a conclu- sion for which there is a t present no entirely satisfactory criteyion.Class B consists of mixtures of substances which react chemically, forming compounds with each other. Some exceptions arise as for example mixtures of phenol and acetone; this pair of substances yields viscosity isotherms of the type A although the formation of a compound undoubtedly takes place. There are also a number of rYlixtGres giving inflected viscosity isotherms. A discussion of these exceptions will be given subsequently. The work described below was undertaken with the object of determining what factors influence the viscosit’y of liquid mixtures, and whether it is possible t o ascertain from viscosity measurements the composition of a compound that might exist in a liquid mixture of two substances.F o r this purpose it was decided to determilne the densities and viscosities of mixtures of phenol and several bases of different strengths over as wide a range of temperature as possible. I n order t o be able t o compare the viscosity-composition curves of these mixtures with the general type of curve given by mixtures of phenol and indifferent substances mixtures of phenol with benzene chlorobenzene and nitrobenzene were examined. Two classes only are required: EXPERIMENTAL. I n the determination of the densities of the liquids examined, pyknometers of the Sprengel type were employed. ’ F o r tempera- tures up t o 20° a pyknometer of about 8 C.C.capacity with a bul 14 BRAMLEY THE STUDY OF BINARY MIXTULZES. PAlZ'I' I. on one limb was used. Two plain Sprengel tubes of about 10 C.C. capacity were employed f o r the higher temperatures. Duplicate experiments with diff ererit pyknometers showed that the figure in the fourth decimal place was trustworthy but a general considera- tion of the different factors influencing the determinations of density indicate that this was about the limit of accuracy attained. All the densities are referred t o that of water a t 4O. The viscosities were measured by means of viscometers of the Ostwald type. On account of the great differences in the viscosities of the liquids examined three pairs of viscometers were used. One pair was suit'able f o r liquids of low viscosity another pair for liquids of intermediate viscosity and a third pair for liquids of high viscosity.I n this way the periods of flow f o r the different liquids were of about the same order. The hygroscopic nature of the substances used rendered i t necessary to protect them from the moisture in the air; this was done by attaching tubes containing anhydrous calcium chloride t o the viscometers. The periods of flow were measured by means of two tested stopwatches which marked off time in fifths of a second. I n the calculation of absoIute viscosities Thorpe and Rodgers's value for water a t 25O namely, 0.00891 was taken as standard. It is well known that in the determination of viscosities of liquids generally the maintenance a t a steady temperature of the bath in which the viscometer is placed is of great importance.Few experiments however had been made in the early part of this investigation before i t became apparent that on account of the enormous temperature-coefficients of the viscosity of some of the liquids at the lower temperatures the precision with which the temperatures of the baths must be maintained was of a high order. This will be evident from the following example The viscosity of phenol a t 20° is 0.1104; a t loo it is 0.2010; for a temperature interval of loo the) viscosity has almost doubled; hence if an accuracy of 0.1 per cent. is t o be attained the temperature must not vary more than O * 0 l o . The thermo-regulation of a bath so thatu its extreme variations never exceed this amount is a matter of some delicacy especially so when the t'emperature required is below that of the room.For a constant temperature bath a t Oo a 4-litre beaker was filled with a mixture of clean ice broken into small pieces and distilled water. The following arrangement was fonnd to be suitable for temperatures between about 5 O and t h a t of the room. The bath consist'ed of a rectangular copper vessel with two opposite sides of glass and had a capacity of about 30 litres. Cooling was accomplished in the following way A coil of about 10 metres of lead tubing (1.2 cm. in diameter) was embedded i BRAMLEY THE STWDY OF BINARY MIXTURES. PART I. 15 rougkly broken ice in a tank of about 30 litres capaci€y; one end of this coil was connected t o a water-tap and the other to one end of a second coil of similar tubing having a dozen turns of 7.5 cm.diameter. The other end of this coil which was immersed in the constant-temperature bath was connected to a pipe leading t o a sink. A steady stream of water was run from the tap through the first coil where it was cooled to a temperature near zero; i t then passed through the second coil thus cooling the water in the bath; finally i t escaped to the sink. The speed of the water was adjusted so that it would cool the bath to a temperature two or three degrees below that required the final adjustment of the temperature being made by means of an electrical heater actuated by an electrical tliermo-regulator of the form described by Gouy ( J . Physique 1597 6 479).A full description with diagrams of this excellent regulator is given by Barnes (Phil. Trans. 1902 [A], 199 208). For temperatures between that of the room and 50° the same method was used excepting of course the cooling arrange- ment which was dispensed with. Auxiliary heating by means of a small gas flame was necessary for temperatures above 30°. The water in the tank was kept thoroughly mixed by means of a t8urbine stirrer. Trial runs extending over a full day showed that, by ths above method the temperature of the bath could be kept constant within tlwo or three thousandths of a degree. For the more viscous liquids the1 temperature-coefficient of viscosity dimin- ishes rapidly with rise in temperature ; consequently the very high degree of precision in the maintenance of a constant temperature which is required for the lower temperatures is not so necessary for higher temperatures.The baths used for temperatures above 50° were beakers of 4-litres capacity and ordinary toluene thermo- regulators were employed. Above looo the water in the baths was replaced by glycerol and aniline was used instead of toluene in the regulators. The substances used were purified as follows : PhemZ.-Pure phenol was distilled and the1 middle fraction, which had a constant boiling point was used. Henzen e.-Pure commercial benzene was shaken with successive portions of concentrated sulphuric acid until the latter was no longer coloured; the benzene was then washed several times with water and dried first with calcium chloride and afterwards with sodium.Finally i t was distilled through a fractionating column, the portion of constant boiling point being used. ChZorobenzene.-Kahlbaum’s chlorobenzene was kept over anhy- drous calcium chloride for a few days and then distilled. Its boiling point was constant 16 BRAMLEY THE STUDY OF BINARY MIXTUHES. PART I. Nitrobenzene.-Some nitrobenzene of unknown origin was washed with sodium carbonate solution then with water dried o'ver calcium chloride and distilled. 9 piiliiie.-Commercial aniline was fractionated several times; filially a quantity was obtained which had a constant boiling point. Fro. 1. It had a const,ant boiling point. 0 Weight per cent. of phenol. II. , , chlorobenzene. I. Phenol and benzene. III. , , nitrobenzene. DimethyZarziZine.-Kahlbaum's preparation was redistilled and ths middle fraction of constant boiling point used.p- To1 uidin e D i ph e 1) y la ?n i i i e o ti d D i plz e n y 1 m e thy la rn in e .-These substances obtained from Kahlbaum were redistilled under diminished pressure the first and last portions being rejected. &uinoZi.ne.-Kahlbaum's synthetic quinoline was allowed to remain in contact with solid potassium hydroxide for several days, and then distilled ; a portion of constant boiling point was obtained. f'yridiize.-Purs commercial pyridine was kept in contact wit BRAMLEY THE STUDY OF BINARY MIXTURES. PART I. 17 solid potassium hydroxide for several days and then fractionated by distillation. This process was repeated several times and ulti- mately a quantity was obtained which had a range of boiling point less than onetenth of a degree.1'11 eiietoZe.-Tliis substance obtained from Kahlbaum was redis- tilled. ,4 cetone.-Commercially pure acetone was treated with aqueous potassium permanganate solution and allowed t o remain for three or four days more permanganate solution being added as the colour disappeared. The acetone was distilled off on the steam-bath dried over calcium chloride and finally distilled using a fractionating column. From 2500 C.C. of commercial acetone 1500 C.C. were obtained which had a range of boiling point of less than one-tenth of a degree. Mixtures of phenol with ( n ) benzene (6) chlorobenzene and (c) nitrobenzene were examined a t one temperature only namely, 20O.The values obtained are given in t'ables In I h and Ic and Fig. 1 represents graphically the relations between the viscosity and composition of these mixtures. The density-composition curves for these mixtmes are! almost straight lines; they lie very slightly below the linear. TABLE Ia. 171 ixtzires of Phenol and Benzene. of phenol. at 20". at 20". It had a constant boiling point. Weight per cent. Density Viscosity ~0.00 0-8772 0.00629 6-04 0.8880 0-00683 9.84 0.8949 0.00724 20.01 0.9133 0.00865 32-40 0.9370 0.01126 42.09 0.9549 0.0 1401 53.02 0.9766 0.0191 1 63-65 0.9976 0.02642 74.1 1 1.0194 0.0381 1 83.20 1.0383 0.0535 100-00 1.0762 0.1104 TABLE I b . Mixtures of Pheml and Chlorobeitzene. Weigh per cent. Density Viscosity of phenol.at 20". at 20". i 0.00 1.1051 0.00768 4.93 1.1034 0.00825 8-78 1.1018 0.00888 21.73 1.0980 0.01122 30.43 1.0954 0.01374 38.90 1.0930 0-01673 49.90 1.0898 0.02218 58.15 1.0874 0.02748 71.41 1.0836 0.04070 81.45 1.0806 0.05555 1 oo*oo 1.0752 0.110 18 BKAMLEY THE STUDY OF BINARY MIXTURES. PART 1. TABLE Ic. Weight per cent. of phenol. 0.00 4.16 8.84 18.12 27-41 37-96 49.73 58.64 71.03 84.68 100~00 Densi t,y at 20". 1.202 1 1.1957 1.1888 1.1756 1.1635 1.1495 1.1346 1.1233 1.1085 1.0927 1-0752 Viscosity at 20". 0-01931 0.0 1975 0.0204 1 0.02208 0-02460 0.02845 0.0353 0-0419 0.0540 0.0759 0.1104 The densities and viscosities of mixtures of phenol and aniline were determined for a number of temperatures ranging from 20" t o 125O.Tables I1 and IIa contain the results obtained and the viscosity isotherms are shown in Fig. 2. The viscosity-composition curves were drawn on a large1 scale for each temperature and the position of maximum deviation from the straight line joining the viscosities of the two components was deterininecl. The compositioii of the mixtures which show the greatest divergence in viscosity from the values calculated according to the simple mixture rule, together with the corresponding temperatures are given in table 116. These quantities are represent'ed graphically in the small curve inset in Fig. 2. The density curves f o r these mixtures are almost straight lines the densities found being slightly higher than those calculated.The viscosities of mixtures of phenol and aniline have been measured for one temperature namely 3 5 O by Thole, lhnstan and Mussel1 (T. 1913 103 1114) URAMLEY THE STUDY OF U N A R Y MIXTUREJ. PART I. 19 FIG. 2. Phenol and aniline. 20 40 60 80 Weight per cent. of phenol BRAMLEY THE STUDY OF BINARY MIXTURES. PART I. 21 Weight per cent. of phenol . 0.00 7.94 15.31 23-34 31-28 39-39 47.56 53-81 62.50 69-52 77.02 85.02 92-28 1 00*00 TABLE IIa. M i x t w e s of PJaenol and Aniline. Densities. Viscosities. 20". 1.0219 1.0276 1.0326 1.0380 1.0434 1.0485 1.0532 1.0569 1.061 1 1.0644 1.0675 1-0704 1.0729 1.0760 125" 0.9288 0.9342 0.9390 0.9436 0.9482 0.9527 0.9671 0.9606 0.9648 0.9690 0.9727 0.9762 0-9795 0.9828 , 20".0.0428 0.0509 0.0610 0.0735 0.0889 0.1059 0.1215 0.1320 0.1418 0.1447 0.1421 0.1331 0.1221 0.1104 , 126". 0.00637 0.00666 0.00693 0.00723 0*00740 0.00770 0-00788 0.00799 0.0081 1 0.008 17 0.008 18 0.00813 0.00797 0.00770 TABLE IIB. Jlixtzcres of Plieiiol and Aniline. Weight per cent. of phenol .. 63.0 62.5 62-0 61-25 60.6 50.5 Temperature ........................ 20" 30" 40" 60" 80" 125" Mixtures of phenetole and aniline were examined a t several temFeratures ranging from Oo to 80°. The figures obtained for the densities and viscosities are given in table 111. The density-com- position curves for mixtures of these substances are straight lines, the values found coinciding with those calculated.Fig. 3 contains the viscosity isotherms %lie character of which i t will be seeen is altogether different from that' of the corresponding curves for mixtures of phenol and aniline 22 BRAMLEY THE STUDY OF BINARY MIXTURES. PART 1 FIU. 3. Aniline and phenetole. Weight per cent. o j phenetole Weight per Mixtures Densities. TABLE I11 of I'henetole and Aniline. cent. of / \ / phenetole. 0" 9.9" 20.2" 29.6" 40.0" 60.0" 80.0" 0" 9.9" 0.00 11-74 21.75 32.30 43.68 54.66 65.69 76.54 88.49 100.0 1.0390 1.0303 1.0214 1.0134 1.0045 0.9872 0.9700 0.1005 0.0631 1.0326 1,0239 1.0149 1.0070 0.9980 0.9809 0.9637 0.0799 0.0523 1.0272 1.0184 1.0094 1.0014 0.9922 0.9748 0,9573 0.0655 0.0443 1.0216 1.0128 1.0037 0.9956 0.9864 0.9688 0.9511 0.0548 0.0375 1.0155 1.0066 0.9974 0.9893 0.9801 0.9623 0.9444 0.0451 0.0321 1.0096 1.0007 0.9916 0,9833 0.9740 0.9560 0.9379 0.0379 0.0276 1.0036 0.9946 0.9854 0.9771 0.9677 0,9497 0.9315 0.03205 0.02385 0.9978 0.9888 0.9795 0.9712 0.9616 0.9434 0.9249 0.02720 0.02075 0.9914 0.9823 0.9729 0.9644 0.9547 0.9363 0.9176 0.02265 0.01780 0.9852 0.9760 0.9666 0,9580 0.9483 0.9298 0.91 10 0.01860 0.0153 24 BRAMT,EY THE STUDY OF BlNhElY MIXTURES.PART I. Tables I V and IVu contain the densities and viscosities of mix- tures of phenol and p-toluidine for several temperatures between 40° and 175O. The densities are somewhat higher than the calcu- lated values ; consequently the density curve is slightly concave towards the composition axis.The viscosity curves are shown in Fig. 4. Ta,ble IVb contains tlie composition of mixtures the viscosi- ties of which show a maximum deviation from the calculated values, together with the corresponding temperatures. These were obtained as already described for phenol and aniline and the small curve inset in Fig. 4 depicts tlie relationship between the two. This mixture has been examined a t 30° by Thole Dunstan and Mussel1 (Zoc. c i t . ) who found that the maximum point on the viscosity-composition curve was a t 63 per cent. of phenol,. Accord- ing t’o the experiments described in this paper the position of the maximum point on the viscosity curve corresponds with a composi- tion of 67 per cent. of phenol and since in this case the position of the maximum does not vary appreciably with the temperature, there appears t o be a discrepancy of about 4 per cent.in the com- position of the mixture of maximum viscosity. Independent experi- ments carried out in this laboratory in which Kahlbaum’s chemicals redistilled before use were employed gave the following results for a temperature of 31.5O: Weight per cent. of phenol ......... 50.6 63.3 66.0 68.8 70.7 79.1 86.0 Viscosity ...... ..... 0.0846 04921 0.0931 0.0931 0.0928 0.0867 0.0803 These figures indicate a maximum viscosity a t a point corre- sponding with a composition of about 67.5 per cent. of phenol, which agrees well with the value found by the aut,hor BRAMLEY THE STUDY OF BINARY MIXTURES PART I. 25 Weight per cent. of pheno2 Weight per cent.of phenol. 0.00 9.85 20.67 29-86 38-57 46.25 55.09 62-70 71.11 80.19 89.76 100*00 TABLE IV. Jf iztvres of Phenol and p-Toluidine. Densities. / A 4 39.9" 59.9" 79.8" 99.9" 125" 0.9703 0.9534 0.9365 0.9189 0.8962 0.9808 0.9640 0.9470 0.9295 0.9068 0.9913 0.9744 0-9574 0.9398 0.9172 1.0004 0.9835 0.9665 0.9488 0.9261 1.0087 0.9919 0.9750 0,9376 0.9348 1.0160 0.9991 0.9820 0.9645 0-9418 1,0239 1.0069 0.9898 0.9723 0.9495 1-0305 1.0135 0.9965 0.9795 0.9567 1.0372 1.0201 1.0031 0.9856 0-9628 1.0441 1.0270 1.0099 0.9924 0.9696 1.0512 1.0340 1.0170 0.9995 0.9766 1.0585 1.0414 1.0243 1.0065 0-9833 39.9" 0~02080 0.02632 0.03352 0.04090 0.0482 0.0543 0-06015 0-0629 0.0627 0.05915 0.0538 0.0479 59.9" 0.01398 0-01649 0.01983 0.02283 0.02564 0.02810 0.03015 0.03115 0.03110 0.02990 0.02780 0.0252 BRAMLEY THE STUDY OF BINARY MIXTURES.PART I. 27 Weight per cent. of phenol. 0.00 16.62 23.1 1 34.42 38-24 45.61 56.31 65.69 76-25 79-43 86.34 100~00 TABLE IVa. Mixtures of Yh,enol mid p-Toluzdine. Densities. Viscosities. -\ 7- 150". 0.8734 0.8898 0-8961 0.9069 0-9106 0.9 172 0.9264 0.9341 0.942 1 0.9443 0.9491 0.9572 175". 0.8502 0.8668 0-8732 0.8842 0.8878 0.8943 0.9034 0.9110 0.9188 0.9210 0-9256 0.9337 150". 0.00491 0.00541 0.00560 0.00594 0.00603 0.00619 0.00636 0.00641 0.00635 0-00630 0.00618 0.00592 175". 0.00423 0.00456 0.00468 0.00490 0.00496 0*00507 0.00517 0.00520 0.00515 0*00612 0*00508 0.00492 TABLE IVb.Mixtzires of Phenol and p-Toluidine. Weight per cent. of phenol.. 61.5 60-0 59.0 58.0 57-0 56.0 55.0 Temperature .............. 39.9" 59.9" 79.8" 99.9" 125" 150" 175" Mixtures of phenol and dimethylaniline were examined a t eight temperatares between loo and 1 8 0 O . The figures obtained are given in tables V and Vu. The viscosity-composition curves are given in Fig. 5. On account of the sinuous nature of these curves, the deviations from t h s straight line are both positive and negative, that is some of the viscositia are greater and some less than the values calculated according to the simple mixture rule. Table V b contains the composition of mixtures the viscosities of which show a maximum positive deviation and the corresponding temperatures.The connexion between the two is shown graphically by the curve inset in Fig. 5. The densities of these mixtures are but slightly higher than the calculated values. TABLE Vb. Mixtures of Phenol and Dirnethylaniline. Weight per cent. of phenol.. .......... 87.5 84-0 82.0 80.0 78-0 76.0 73.0 71.5 Temperature ........ 10" 20" 29-8" 40.2' 59.9" 80-0" 126.0" 177 28 BRAMLEY THE SlUDY OF BINARY MIXTURES. PART I. FIU. 5. PJzenol and dimethglaniline. Weight per cent. of phenol Weight per cent. of phenol. 0.00 9.07 17-30 33-94 33.08 40.39 48.27 55.75 62.82 69.97 78.14 85-39 92.76 100-00 TABLE V. Jf ixtzires of Phenol arid Dimethylaniline. Densities.100 0.9647 0.9753 0.9851 0.9932 1.0041 1.0136 1.0236 1.0327 1.0413 1-0500 1.0595 14678 1.0759 1.0835 20" 0.0564 0.9670 0.9768 0-9849 0.9959 1.0053 1-0150 1.0243 1.0330 1.0416 1.0512 1.0595 1.0676 1.0752 126" 0.8679 0.8776 0.8863 0.S939 0.9040 0.9127 0.9216 0.9302 0.9387 0.947 1 0.9564 0.9648 c.9734 0-9Sl5 177" 0.8225 0.8318 0.8403 0.8472 0.3568 0.8647 0.8733 0.8817 0.8895 0.8978 0.9069 0-9152 0.9237 0.0316 10" 0.01654 0.02076 0.02586 0.03145 0.04 185 0.053 15 0.0696 0.0869 0,1094 0-1347 0,1639 0-1850 0.1964 0.201 Weight per cent. of phenol. 0.00 7.93 16.61 24.60 32.71 41-46 49.19 56.14 63-95 70.99 78.83 86.08 93.19 100~00 TABLE V a .Mixtures of Phenol and Dirnethylaniline. Densit,ies. f A \ / 29.8" 40.2" 59.9" 80.0" 29.8" 0.9482 0.9574 0.9677 0.9772 U.9872 0.9981 1.0076 1,0158 1.0252 1.0329 1.043 1 1.0516 1.0594 1.0668 0.9400 0,9492 0.9593 0.9888 0.9788 o . 9 m 0.9990 1.0073 1.0166 1.0243 1.0346 1.0431 1-0509 1.0582 0-9234 0.9325 0.9425 0.9520 0.9619 0.9724 0.981 8 0.8899 0.9991 1.0069 1.0171 1.0256 1.0335 1.0414 0.9070 0.9159 0.9258 0.9352 0,9449 0.9553 0.9645 0.9726 0-9817 0.9895 0.9997 1.0082 1.0162 1.0242 0.01 173 0.01351 0.01629 0.0 1036 0.02315 0.0283 0.0333 0.0393 0.04705 0.05325 0.06025 0-0648 0.068G 0.070 BRAMLEY THE STUDY OF BINARY MIXTURES.PART I. 31 Phenol and diphenylamine mixtures were examined at four temperatures between 30° and 8l0 and the results obtained are given in table VI. The figures in brackets for the viscosities of diplienylamine a t 30° and 40° were obtained by extrapolation. Fig. 6 shows graphically the relationship between the viscosity and cornposition of these mixtures. The density curves in this case FIG. 6. Phenyl and diphenylamine. 0 40 60 Weight per cent. oJ phenol. 3 10 are straight lines. A few mixtures of these two compounds have been examined a t 50° by Thole Dunstan and Mussel1 (Zoc. cit.). Mixtares of phenol and diphenylmethylamine were examined a t six temperatures between loo and 80° and the measurements are recorded in tables VII and VIIa.The density curves are almost straight lines the calculat,ecl densities being slightly higher tha Weight per cent. of phenol. 0.0 3 7.87 15.18 23.29 30.87 38.60 46-56 53.43 59.65 68.84 76.60 84.59 92-04 100~00 TAELE VI. Slixtures of Phenol and Diphenylamine. Densities. / h 4 30" - 1.0790 1.0780 1.0769 1.0759 1.0740 1.0738 1.0729 1.0721 1.0709 1.0699 1-0688 1.0678 1.0667 40" - 1.0711 1.0700 1.0689 1.0678 1.0668 1.0656 1.0647 1.0639 1.0626 1.0616 1.0605 1-0595 1.0584 61" 1.0543 1.0533 1.0523 1.0511 1.0501 1.0490 1.0478 1.0470 1.0461 1.0448 1.0438 1.0427 1.0416 1.0405 81" 1.0377 1.0386 1.0356 1.0344 1.0333 1.0321 1.0309 1-0300 1.0292 1.0278 1.0268 1-0256 1:0245 1.0233 30" (0.1357) 0.1257 0.1165 0.1072 0.1003 0.0947 0.0898 0-0861 0.0828 0.0791 0.0765 0.0742 0.0 7255 0.070 ERAMLEY STUDY OF THE BINBRY MIXTURES.PART I. 33 the observed values. Fig. 7 shows the connexion between the viscosities and compositions of these mixtures. FIG. 7. Phenol arid diphenylmetlaylamine. 0.20 0.1 8 0.16 0-14 0.12 & 2 0.10 *+ G? -* P O*OE O*OE 0.04 0.02 C TVeight per cent. OJ phenol. TABLE VII. U i x t w e s of Piieiiol and Diphenylmethylamine. phenol. 9.8' 20.1" 9.8" 20.l0 Weight per Densities. Viscosities. cent. of 7- - 0.00 1.0595 1.0515 0.1096 0.0722 4.92 1.0605 1-0523 0.1090 0.0708 9-45 1.0615 1.0532 0-1097 0.0715 17.18 1.0632 1.0548 0.1145 0-0730 27.69 1.0657 1.0572 0-1226 0.0764 36-92 1.0679 1.0593 0.1306 0.0798 48.42 1.0708 1.0621 0-1411 0.0845 56.87 1.0728 1.0641 0.1502 0.0885 67.10 1.0753 1.0665 0.1614 0.0935 78.79 1.0782 1.0693 0.1752 0.0995 89-30 1.0809 1.0720 0.1877 0.1048 100~00 1.0836 1.0750 0.2010 0.1104 VOL.CIX. TABLE VIIa. Mixtures of Phenol and Diphei2ylmethylamine. Weight per cent. of phenol. 0.00 4.98 10.21 20.04 35.34 49.87 62-12 73.58 82-22 90-05 100*00 Densities. 3 0% 1.0438 1.0449 1.0461 1.0483 1.0519 1.0552 1.0581 1.0607 1.0627 1.0645 40' 1.0359 1.0369 1.0379 1.0401 1.0435 1.0467 1.0495 1.0521 1.0542 1.0560 60" 1.0198 1.0207 1.0217 1.0237 1.0269 1.0301 1.0328 1.0354 1.0373 1.0391 Pi00 1.0040 1.0048 1.0058 1.0076 1.0104 1.0136 1.0164 1.0188 1.0206 1.0223 30" 0.0513 0.0510 0.0513 0,0519 0-0542 0.0570 0.0597 0.0626 0.0650 0.0673 / 1.0668 1.0584 1-0414 1.0242 0.070 BRAMLEY STUDY OF THE BINARY MIXTURES.PART I. 35 The densities and viscosities of mixtures of phenol and quinoline were determined f o r eight temperatures ranging from 9 * 8 O to 1 7 5 O , FIG. 8. Phenol and quinoline. Weight per cent. of phenol. and the measurements are recorded in tableg V I I I and VIIIa. Fig. 8 contains the viscosity curves and the small curve inset shows tho connexion between the temperature and composition of the c Weight per cent. of phenol. 0.00 7.54 14.56 22.73 29.76 37.52 45.08 53.20 60.30 68.21 76.88 83-37 92.06 100*00 TABLE vrrr.Mixtures of Pheizol and Quinoline. Densities. I \ 7 9.8" 20.1" 125" 175' 9.so 1.1004 1.1021 1.1037 1.1056 1.1071 1.1078 1.1074 1,1057 1.1030 1.0994 1.0950 1.0916 1.0875 1.0836 1.0925 1,0944 1.0960 1.0977 1-0992 1.0999 1-0994 1.0977 1.0950 1.0914 1-0869 1.0835 1.0791 1.0750 1.0085 1-0103 1.0119 1.0133 1-0140 1.0139 1.0127 1.0099 1.0065 1.0021 0-9969 0.9927 0.9876 0.9828 0.9673 0.9687 0.9698 0.9707 0.9705 0.9696 0.9678 0.9643 0,9602 0.9550 0.9492 0.9445 0.9389 0.9337 0-04805 0.0619 0.0810 0.1170 0.1677 0.2644 0.3750 0.5065 0.5259 0-4740 0.3836 0.3156 0.2420 0.201 Weight per cent.of phenol. 0.00 7.77 14-92 21.96 29.82 37-14 44-62 52.31 59.89 67.92 75.75 83.49 91.79 100*00 TABLE VIIIa. iMixtures of Quinoline and Phenol. Densities. 29:9' 1.0851 1.0870 1.0886 1-0904 1.0917 1.0924 1.0917 1.0901 1.0874 1.0837 1-0795 1.0756 1.0711 1.0668 40' 1,0773 1.0792 1.0808 1.0823 1-0838 1.0843 1.0837 1.0820 1.0793 1.0755 1.0713 1.0672 1.0628 1-0584 60" 1,0615 1-0635 1.0651 1-0666 1.0679 1.0682 1-0675 1.0658 1.0629 1.0590 1.0548 1.0503 1.0458 1.0414 80' 1-0458 1.0478 1.0494 1-0509 1.0518 1.0521 1.0514 1.0496 1.0465 1-0425 1.0382 1.0335 1.0288 1.0242 29.9" 0.02943 0.03645 0.04495 0.05605 0.07425 0-0965 0.1221 0.1436 0.1480 0-1344 0.1177 0.1010 0.0844 0.070 38 BRAMLEY THE STUDY OF BlNARY MIXTURES.PART I. mixture which has a viscosity deviating t o the greatest extent from the calculated value. The data from which this curve is drawn were obtained from the viscosity curves as previously indicated and these are' enumerated in the follo,wing table : TABLE VIIIb. n-lixtures of Phenol and Quinoline. Weight per cent. of phenol.. . . . . . . . . . 58.0 57.0 56.0 55-5 54.0 53.0 50.5 48.0 Temperature . . . . . . . 9.8" 20.1" 29.9" 40.0" 60.0" 80.0° 125" 175" The broken line curve in Fig. 8 is the density curve for 20*1°. Unlike all the other mixtures described in this paper the density- composition curve f o r mixtures of phenol and quinoline? has a very pronounced maximum.Moreover this maximum does not become less marked with rise in temperature; the density curves for the othm temperatures run very approximately parallel to the one shown in the figure. Mixtures of phenol and pyridine were examined a t seven tem- peratures ranging from loo to l l O o and the results are set out in tables IX and IXn. I n Fig. 9 arel given the viscosity curves. The scale of viscosities for the three highest temperatures 60° 80°, and l l O o is considerably extended in order to bring out more clearly the character of the curves. Inset in Fig. 9 is the curve connecting temperature with the composition of the mixture which shows a maximum positive deviation in viscosity. Data for the construction of this curve are given in table 1x6.TABLE IXb. rllixtzires of PJbenol and Pyra'dine. Weight per cent. of phenol . . 93.0 88.0 84.5 82.0 77-0 73.5 70.0 Temperature . . . . . . . . . 10" 20" 30" 40" 60" 80" 110" The densities of mixtures of these two substances show a con- siderable deviation from the calculated values; this is especially so with those mixtures co,ntaining a greater proportion of phenol. The broken line curve shows the connexion between density and com- position for one temperature namely 20°. The density curves for all the other tempera-tures run closely parallel t80 this BRAMLEY THE STUDY OF BINARY MIXTURES. PART 1. 39 FIG. 9. Phenol and pyridine. Weight per cent. of phenol TABLE IX. Mixtures of Phenol and Pyridine. Densities. / b ../ 20" 30' 40" 60" 80" 110" 20" 30" 0.9819 0.9723 0.9627 0,9424 0.9218 0.8900 0.00941 0.00821 0.9916 0.9820 0.9724 0.9524 0.9324 0.9014 0.01 109 0.00940 1.0005 0.9909 0.9814 0.9620 0.9428 0.91 19 0.01321 0.01106 1.0096 1.0002 0.9909 0.9720 0.9532 0.9234 0.01597 0.01339 1.0187 1.0096 1.0005 0.9821 0.9639 0.9347 0.02010 0.01634 1.0258 1.0167 1.0077 0.9897 0.9718 0.9430 0.02411 0.01963 1-0349 1.0260 1.0173 0.9995 0.9821 0.9535 0.03215 0.02515 1.0429 1.0343 1.0257 1.0083 0.9910 0.9626 0.0437 0.03245 1.0514 1.0426 1.0341 1.0168 0.9995 0.9719 0.05945 0.0429 1.0568 1.0482 1.0397 1.0226 1.0053 0.9779 0.0748 0.0517 1.0620 1.0534 1.0449 1.0279 1.0107 0.9833 0.0907 0.0604 1.0668 1.0583 1.0498 1.0327 1.0155 0.9878 0.1004 0.06625 1.0710 1.0625 1-0540 1.0369 1.0198 0.9923 0.1072 0.06915 1.0752 1.0668 1.0584 1.0414 1.0242 0.9967 0.1104 0.0709 Weight per cent.of phenol 0.00 8-30 16.12 24.36 32.58 38.94 47.13 54.96 63-81 70.49 71.94 85.17 92-45 100~0 BRAMLEY THE STUDY OF BJNARY MIXTURES. PART I. 41 Weight per cent. of phenol 0.00 17-26 26.01 35.14 45.48 51.89 TABLE IXQ. Mixtiires of Phenol and Pyriditle. Density a t 10" 0.9916 1.0101 1.0196 1.0294 1-0412 1.0479 Visc:osit.y at' 10" 0.01108 0.01594 0.02050 0.02675 0.0376 0.0505 lTTeight per cent. of Density phenol. a t 10" 58-46 1.0544 66.99 1-0618 76.81 1.0700 82-86 1.0742 91.89 1.0787 100~00 1.0836 Viscosity a t 10" 0.0680 0-0988 0.1416 0.1644 0.1880 0.2010 Tables X and X n contain a record of densities and viscosities of mixtures of phenol and acetone for five t>emperatures ranging from 9'95O t'o 49.8O.The densities of these mixtures are almost thel same as those calculated; the density curves are almost straight lines. The viscosity curves are given in Fig. 10. TABLE X. Mixtures of Phenol cciid Acetone. Weight per cent. of phenol. 0.00 14-19 26.72 38.06 49.43 57-79 65.22 73.74 78.94 85.39 92.85 100*00 Weight per cent. of phenol 0.00 9.57 19.53 27.70 37.42 44.67 53.79 60-24 67.1 9 74.25 80.76 87.98 92.81 100~00 Densities. 9.95" 20.05" 0.8031 0.7912 0.8425 0.8315 0.8768 0.8662 0.9085 0.8983 0.9406 0.9308 0.9642 0.9547 0.9851 0.9757 1.0090 0.9998 1-0237 1.0146 1.0420 1.0334 1,0623 1.0538 1.0836 1.0752 Viscosities, /- 9.95" 20.05" 0.00360 0.00323 0.00486 0.00429 0.00635 0.005CO 0.00868 0.00755 0.01256 0.01055 0.01688 0.01379 0.02358 0.01853 0.03670 0.02750 0.0495 0-0359 0.0748 0.0497 0.1193 0,0730 0*2010 0.1104 TABLE Xa.Mixtures of Phenol and Acetone. 7 29.8" 0.7799 0.8056 0.8336 0-8572 0.8855 0-9063 0-9330 0-9520 0.9724 0.9935 1.01 15 1.0327 1.0466 1-0668 Densities. 40.1" 0.7676 0.7940 0.8223 0.8460 0.8743 0.8952 0.9220 0-9416 0-9625 0.9837 1.0022 1.0237 1.0378 1.0584 7 49.8" 0.7559 0.7830 0.8116 0.8335 0.8636 0.8846 0.9115 0.9316 0.9530 0.9733 0.9933 1.0150 1.0293 1.0503 7 29.8" 0.00295 0.00360 0.00441 0.0052 1 0.00670 0.00808 0.01058 0.013 19 0.01658 0.021 80 0.02910 0-03915 0-04905 0.07 100 Viscosities - 40.1" 0.00270 0.00328 0.00399 0.00470 0.00590 0.00711 0.00904 0~01101 0.01363 0.01741 0.02230 0.02875 0.03465 0.0474 - 49.8" 0.00248 0.00299 0.00360 0.00422 0.00530 0-00628 0-00794 0.00950 0.01150 0.01425 0*01785 0.02245 0.02615 0.0328 C 42 BRBMLEY THE STUDY OF BINARY MIXTURES.PART JI. 0.20- 0.18- 0.16- 0.14- 0.12- s; s u .* CI 8 0.10- 0.08 0.06 0.04 0.02 FIU. 10. Pheizol and acetone. - - - - - __ - - - - ~ - - - - - 0*00,, Weight per cent. of phenol. Discussion of Results. A general survey of the work described in the'foregoing pages (1) When a substance such as phenol which in the liquid state, leads to the following general conclusions BRAMLEY THE STUDY OF BINARY MIXTURES.PART J. 43 consists largely of associated molecules is mixed with another indif- f erent substance such as benzene chlorobenzene or nitrobenzene, the viscosity of the mixture is considerably lower than that calcu- lated according to the simple mixture rule. The viscosity-composi- tion curve shows a very marked curvature convex towards the axis of composition. (2) When the formation of a compound takes place as for example when phenol is mixed with such bases as aniline p-tolu- idine and quinoline the viscosity of the mixture is in general, higher than that calculated by the simple mixture rule.The viscosity-composition curve is concave towards the cornposition axis. That the high viscosity of these mixtures is caused by the acidic nature of phenol is shown clearly by displacing the hydroxyl hydrogen by an alkyl group thus obtaining a neutral substance, phenetole f o r example which with aniline yields mixtures the viscosity curves of which are convex towards the axis of com- position. (3) The position of the maximum point on the viscosity curves generally varies with the temperature and can therefore give little indication as t o the composition of any compound that might be present. Further in every case described in this paper where there is a positive deviation the composition of the mixture the viscosity of which shows a maximum positive deviation depends on the temperature a t which the viscosities are measured; the nature of the relationship between these two quantities is indicated by the small curves inset in the various figures.This is altogether different from what Denison found. On examining these small curves it will be seen that they are all convex towards the composi- tion axis; they lie entirely on on0 side of the ordinate representing molecular proportions and if produced in the direction of higher temperatures they would approach this ordinate asymptotically. These peculiarities are in all probability the result of an unequal thermal dissociation of the molecular complexes of phenol and of the compound formed the opposing influence of these two kinds of molecular complexes being altered in favour of the compound.This is well illustrated by &he mixtures of pyridinei and phenol. A t loo only a very small portion of the viscosity curve is concave towards the axis of composition. This curve resembles very closely the one for phenol and benzene and were it not f o r the slight inflexion near that end of the curve where the mixtures are rich in phenol it might be taken as a typical case of a mixture of two indifferent su bst'ances one of which contained associated molecules. Several unsuccessful attempts were made to measure the viscosity of pllenol a t Oo with the object of ascertaining whether the inflexiar, c+ 44 BRAMLEP THE STIJnY OF RINARY MIXTURES. ]'ART I. of the viscosity curve would disappear at' this temperature. With rise in temperature the part' of the curve which is concave t'owards the composition axis becomes more pronounced and above 40° a well defined maximum appears.The viscosity curve f o r this mixture a t loo has only about 15 per cent. of its length above the straight line joining the viscosities of the pure components whereas a t l l O o there is only about the same proportion below the corre- sponding straight line. A t the lower temperatures thO effect of the associated phenol molecules on the viscosity is predominant whilst a t the higher temperatures the influence of the pyridine-phenol complexes is more assertive. The development of a maximum on the viscosity curve of a mixture with rise in temperature appears to be rather unusual. Faust. found the opposite effect with mixtures of ether and acetone. The difference between the viscosity curves for mixtures of phenol with aniline and dimetliylaniline may be accounted for by the weaker basic nature of the latter. As regards the viscosity curves for mixtures of phenol with diphenylamine and diphenylmethylamine it will be noticed that in neither case do thO curves fall as far below the straight line joining the viscosities of the pure components as is the case f o r mixtures of phenol and benzene. Since there is no reason to suppose &hat either diphenylamine or diphenylmethylamine has a weaker dissociating action on the associated phenol molecules than has benzene the smaller curvature of thO viscosity curves f o r mixtures of phenol and these two very weak bases may be the result of the counteracting influence of compound-f ormation. The viscosity curves f o r mixtures of phenol and acetone unlike those of phenol and pyridine show no change in charactler with rise in temperature. The associated phenol molecules are pre- dominant throughout. It should be borne in mind however that the range of temperature over which this mixture has h e n examined is relatively small and although as has been shown by Schmidlin and Lang (Ber. 1910 43 2812) compound-formation certainly takes place. the amount of compound formed is un- doubtedly much smaller than in the phenol-pyridine mixtures a t the same temperatures. A number of experiments on the influence of pressure on the viscosity of liquids have been made by Warburg and Sachs (Ann. Physiil. 1884 [iii] 22 518) and Cohen (ibid, 1892 [iii] 45 666), who found that increase in pressure generally causes an increase in viscosity. This increase in viscosity may not be the direct result of the increase in pressure but of the decrease in voluine resulting from the latter. Cohen found that the change in viscosity is no BRAMLEY THE STUDY OF BINARY MIXTURES. PART I. 45 proportional to the increase in pressure. Bogojawlenski and Tam- mann (Zeitsch. phtysikal. Chem. 1898 27 457) investigated the relative effect of pressure on the viscosity and specific volume of water a t Oo and found that they are not proportional. That influ- ences besides mere volume changes play a far more important part in determining the viscosity of a liquid is shown in a most striking manner by the results obtained in this investigation. I n every case studied the density-composition curves for the different tem- peratures run very nearly parallel; in no case is this so with the viscosity curves. A reference t o the mixtures of phenol with quinoline and pyridine will illustrate this. The former of these two mixtures has a marked maximum on the density curves. The maximum increase in density or the greatest contraction in volume produced by mixing takes place a t all temperatures when the mixture contains about 47 per cent. of phenol and further the amount of this contraction is very approximately the same f o r every temperature studied. A glance a t the viscosity curves in Fig. 8 will show how different these are. I n the phenol-pyridine mixtures there is a maximum contraction in volume a t 59 t o 60 per cent. of phenol. I n this case also both the composition of the mixture which shows maximum contraction and the extent of this contraction are almost independent of temperature. Nothing could be more different than the viscosity curves for this mixture; not only is there a lack of parallelism but they change in character from one extreme to the other with rise in temperature. If con- traction in volume resulting from whatever cause were the dominant factor in increasing the viscosity a t least some1 approach to a proportional relative increase in viscosity and volumscontrac- tion might be expected. Experiment shows that this is not the case. I wish to express my sincere thanks to Professor Philip f o r placing the resources of his laboratory a t my disposal and for the kind interest he has taken in the work throughout. THE IMPERIAL COLLEGE OF SCIENCE AND TECHNOLOGY, SOUTH KENSINGTON, LONDON S.W. [Received Nocenzber 1 2 t h 1915

ISSN:0368-1645    

DOI:10.1039/CT9160900010

出版商:RSC    

年代:1916

数据来源: RSC

摘要:

J O U R N A L OF THE CHEMICAL SOCIETY. TRANSACTIONS. A. CHASTON CHAPMAN. A. W. CROSSLEY D.Sc. Ph.D. F.R.S. F. G. DONNAN M.A. Ph.D. F.R.S. BERNARD DYER D.Sc. M. 0. FORSTER D.Sc. Ph.D. F.R.S. A. HARDEN D.Sc. Ph.D. F.R.S. T. M. LOWRY D.Sc. F.R.S. J. C. PHILIP D.Sc. Ph.D. F. L. PYMAN D.Sc. Ph.D. A. SCOTT M.A. D.Sc. F.R.S. G. SENTER D.Sc. Ph.D. S. SMILES D.Sc. J. F. THORPE D.Sc. Pli.D. F.R.S. &,bitor : J. C. CAIN D.Sc. Ph.D. Sub- @hitar : A. J. GREENAWAY. %riei#tatrt Sixrb-6;bitar : CLAREKCE SMITH D. Sc. 1916 Vol. CIX. Part II. pp. 649-end. LONDON: GURNEY & JACKSON 33 PATERNOSTER ROW E.C. 1916 PRINTED IN GREAT E~UTAIN BY RICn ARD CLAY & SONS LIWTED, BRUNSWICK SS. STAMFORD ST. \.C., ASD BUNGAY SUFFOLIC

ISSN:0368-1645    

DOI:10.1039/CT91609FP011

出版商:RSC    

年代:1916

数据来源: RSC

摘要:

46 DORkE AND ORANGE CONTRIBUTIONS TO THE CHEMISTRY 1V.-Contributions to the Chemistry of Cholesterol and Cop-osterol. Part 111. The Ozonides o j Cholesterol. Part IV. The Action of Bromine on Cholesteryl Benzoate. By CHARLES DORBE and LIONEL ORANGE. RECENT researches on the constitution of cholesterol may be sum- inarised in the following formula : CH CH, /-\ I n addition to the double linking situated in the end vinyl group, the possible presence of another which must be placed in the complex CllHl, has been the subject of considerable discussion. The evidence rests almost entirely on the1 nature of the action of ozone on cholesterol but in different hands this reagent has given very different results. I n 1908 one of us firs€ showed that whilst cholesterol formed an ozonide C27H4G0*03 the corresponding ketone gave a compound C2,H,,0*0,.This number was in keeping with the views then held f o r the number of oxygen atoms which should combine witlh a doubly unsaturated ketone and the statement was made that “itl may be that cholestenone contains a second ethylene linking which is not rendered apparent by the usual reagents ” (T.. 1908 93 1330). Langheld (Bey. 1908 41 1023) mentioned that cholesterol appears to combine’ with a t leash two ozone complexes. Diels, however found that the crude ozonide after treatment with alcohol lost oxygen and that a definite compound C27H4@,04, resulted for which he suggested a peroxide formula (Ber. 1908, 41 2596). Molinari and Fenaroli obtained in ethereal solution, products containing between 3 and 4 atoms of combined oxygen per molecule but later using ozonised air of 1 per cent.ozone concentration they obtained ozonidea C27H4G0,0G stable a t 60°, from cholesterol and phytosterol. They stated that these results confirmed the existence of a second double linking in chblesterol, and further claimed that thO increase in weight of an unsaturated compound after treatmentl with ozone gave an ‘ozone number’ which was a measure of the number of ethylene linkings presen OF CHOLESTEROL AND COPROSTEROL. PARTS 111. AND IV. 47 ( I ~ w . 1908 41 2785). This method was extended by one of us to a series of derivatives of cholesterol (T. 1909 95 638). Under prolonged action of the ozone stream cholesterol and cholestenone were found to combine with two molecular proportions of ozone.The saturated dihydrocholesterols P-cholestanol and coprosterol, still gave ozonides C,7H4&),03 but the ketones corresponding with these showed an addition of 0,. This inconsistency has recently been criticised by von Furth and Felsenreich (Biochem. Zeitsch., 1915 69 416) but it serves to show that the action of ozone is not as simple as was a t first assume'd. The behaviour of the dibasic acid C&&,O (Diels BET. 1906 36 3177) was also examined. It is insoluble in ordinary solvents but the ozonides are readily soluble. A t the solution point (six hours) an ozonide C2,H4,04,03, formed by saturation of the characteristic double linking was obtained ; after one hundred hours a compound C,,H,,O,,O, was produced.Harries (Ber. 1912 45 936; Annalen 1912 390 235) has brought evidence t o show that the 16 per cent. ozone used by him contains two constituents normal ozone 0, and oxozone 0, in the proportion 20 0,. I f the crude ozone is passed through dilute sodium hydroxide solution the oxozone is destroyed. This 'washed' ozone gives normal ozonides by the addition of 0 for each unsaturated linking. The ' crude' ozone gives not only the ozonide X,O, and tlhe oxozonide' X,O, but also polymerides of these (XO,) and (X04)2. Further treatment with crude ozone leaves the normal ozonide unchanged but the bisozonide is con- verted into the bisoxozonide. These results may account for the varying compositions assigned to the cholesterol ozonides but they do not seem to have been taken into account by von Furth and Felsenreich who working with 8 per cent.ozone and using the ozonsnumber method with chloroform as a solvent found that the molecule of cholesterol absorbs 1.5-4 molecules of ozone! dihydrocholesterol 1.8-4.3 and cholestane 1-5-4.8 molecules. They conclude that only one double linking exists in the cholesterol molecule. The ozonide results are however, of little value. We have found that chloroform is strongly attacked by ozone and that i t is impossible to obtain a product free from chlorine when it is employed. Harries has found (Zoc. cit.) that cholesterol with washed ozone in hexane solution gives an ozonide, C27H460,03 but in carbon tetrachloride solution an ozonide, c 2 7 H 4 6 0 0 6 is formed. The method of working namely evapora- tion of the solvent in the 8 per cent.ozone current and analysis of the crude residue obtained does not seem very satisfact,ory. 1% has moreover lately been recognised that for the study of con 48 DORJkE AND ORANGE CONTRIBUTIONS TO THE CHEMISTRY stitution a weaker ozone of 1 to 2 per cent. concentration is desirable. We have therefore made some preparations and analyses of the ozonides of cholesterol and of the acid C27n4404 in the hops of explaining some of these points which are of importance both from the point of view of the constitution of cholesterol and of the nature1 of ozone action. EXPERIMENTAL. The ozonised oxygen used contained 2 per cent. of ozone. Passage through sulphuric acid did not alter the concentration but after 'washing' through 5 per cent.sodium hydroxide it became 0.8 per cent. The cholesterol was ozonised in glacial acetic acid suspension except in one or two cases. The acid C27H4404 was suspended in acetone,. Acetic acid appears to be quite unaffected by ozone and acetone is only slightly attacked. I n order to form the ozonide C,,H4,O*O, 1 gram of cholesterol requires 0.12 gram of ozone. Under t8Ke conditions of experiment this quantity was passed through per hour. All the products obtained had very similar properties and were worked up in the same way. The acetic x i d solution was diluted with its own volume of light petroleum and poured into1 twice the volume of water. The ozonide floated as a clot on the1 aqueous layer and was easily removed. The aqueous layer was extracted with ether the ethereal solution shaken with dilute alkali and the ether evaporated.The total ozonide was dissolved in a little ether and precipitated with light petroleum this process being several times repea$.ed. The viscous clot obtained foamed in a high vacuum and dried to a britttle solid which was easily powdered. It was kept for fourteen days in a vacuum before analysis. The acetone solutions were allowed t o evaporate and the ozonide treated as before. The following tables give the theoretical composition of the various ozonides : TABLE I. Theoretical Composition of the Ozoiiides of Cholesterol, [H = 1.008]. C,vH,,O f0.q +O -to +Oc +O to,. C ...... ......... 83-85 74.59 71.94 69.52 67.17 65.01 62.98 H ............... 12.00 10.68 10.30 9.95 9.62 9-31 9.0 O F CHOLESTEROL AND COPROSTEROL.PARTS 111. AND 1V. 49 TABLE 11. Yheoretical Composition of the Ozorzides of t h e Acid C27H44O4. C&f&4O* 3 - 0 3 + 0 4 t-06 +o + 0 7 C ............... 74.95 67.45 65-32 63.28 61-36 59.55 67-85 13 ............... 10.26 9.23 8.87 8.44 8.19 8.09 7-85 The first series of experimelnts table 111 was made t o discover whether cholesterol formed a higher ozonide than C,,H4,0 + O,, and if so what was the maximum number of oxygen atoms which could combine. TABLE 111. The Extre,me Action of Ozone on Cholesterol. Total time Time in Rotation of treatment hours per Ozone of Analysis with ozone 1 gram passed ozonide. Analysis indicates 13xpt. in hours. treated. through [a]:* C/H addition of 1.16 3 H,SO A" 66.66 66.52 O6 9-57 9.55 2. 18 0 H2S04 - 64.97 65.56 0 7 9.21 9.05 3. 21 10 - 21" 65.01 65.27 0, 8.93 8.94 NaOH 67.35 67.28 0, 4. 60 24 3- 31" 9-75 9.83 %SO4 Glacial acetic acid was used except in e'xperiment 1 which was in acetone! solution. Rotations were measured in ethereal solution. It being evident that a maximum of 6 or 7 atoms of oxygen could combine atltempta were made to asce'rtain the composition of the ozonide formed in the first stsage of the action that is during the saturation of the double1 link in the terminal vinyl grouping. The results are given in table IV 50 DORkE AND ORANGE CONTRIRUTIONS TO THE CHEMISTRY TABLE IV. The Action of Ozoite O?L Cholesterol in Gluciul Acetic A c i d S m p e n - sion t h e experinzeizts being stopped as soon as tlze cholesterol had passed i n t o solutioiz as ozonide.Total Timein Rotation time of hours per Ozone Expt. treatment 1 gram passed in hours. treated. through 5. 8 2+ H,SO, 6. 5 24 - - 7. 17 3 NaOH 8. 3 1; + H7W4 %SO4 NaOH 9. 10 4 + of Analysis ozonide. Analysis indicates [.I1 C/H. addition of 35" 73.03 72.33 03-4 10.49 10.41 35" 71.11 70.98 0, 10-35 10.10 23"* 71-13 71.21 0, 10.63 10.46 74.89 75.01 0,- + - 10.27 10.86 73.38 73.24 O,+ 45' 10-71 10.64 * Chloroform solution. I n so far as the solution point can be taken as indicating the end of the first stage these results point t o an addition of O4 with crude ozone and with less cert'ainty of 0 with normal ozone. The product obtained in experiment 8 was fully saturated towards bromine.In the following experiments 10-12 made with the acid C27H4404 the solution point was far sharper. I n experiment 11 the action was stopped short of complete solution and the excess of acid removed by filtration. Experiments 13 and 14 show the effect of prolonged treatment with crude and washed ozone respectively. TABLE V. Action of Ozone on t h e Acid C27H4404 m. p . 290O. Total Time in Rotation time of hoursper Ozone of Analysis Expt. treatment 1 gram passed ozonide. Analysis indicates in hours. treated. through [a]:* C/H. addition of 10. 6 2 H,SO 30" 64-45 63-85 04-5 11. 13 4 H,SO 28" 65-03 64.71 0, 8-96 8-83 9.07 9.0 OF CHOLESTEROL AND COPROSTEROL. PARTS 111. AND IV. 51 TABLE v (continued).,4ctioi~ of Ozoize on t h e Licid C27n4404 m. p . 290'. Total Time in Rotation time of house per Ozone of Analysis Expt. treatment 1 grain passed ozonide. Analysis indicates in hours. treated. through [a]:8 CIH. addition of NaOH 12. 8 4 + - 64.09 64.20 04-5 H2SO4 8.91 9.49 13. 25 8 H,S04 13" 57-78 57-66 0s 7.55 7.40 NaOH 59.21 59.74 0, H?SOg - 8.00 8-15 14. 33 20 + Statement ixnd Discussion of Results. (1) It will be seen that cholesterol combines with a maximum of seven atoms of oxygen under the action of ozone. With 'washed' or normal ozone the number is exactly six. The acid C2,Hi404 forms compounds [ + O,] and [+ O,] on treatment with washed and crude ozone respectively. (2) Expariments made to determine the first stage in the action point to the formation of an ozonide [+O,] with cholesterol and washed ozone but t o an addition of 0 in other cases.(3) All the products obtained were brittle solids which melted between 8 5 O and 95O and decomposed above looo. Like other saturated cholesterol derivatives they were dextrorotatory the values f o r those made with washed ozone1 being higher than those made with crude ozone. No information was obtained by a study of their decomposition products resinous acids and neutral sub- stances being formed. The formation of an ozonide C27H460,06 as an end-product with ozone must we think bel taken as evidence that a second double linking has entered into combinatlon. The comparatively slow rate a t which the reaction takes place leads to the conclusion that this linking does not exist pre-formed in the cholesterol molecule, but is developed under the action of ozone.Harries has suggested that water is eliminate'd from the grouping -CH(OH)*CH,* with the formation of a cholesterylene but this explanation scarcely holds in the case of the ketones and the acid C27H4404 which also form higher ozonides. The explanation previously given by one of us that the secondary unsaturation is due to the opening up of a bridged ring under the influence of ozone seems more probable, and has been adopted by von Furth and Felsenreich. Theanalogie 52 DOREE AND ORANGE CONTRIBUTIONS TO THE CHEMISTRY of cholesterol with the terpenes render the presence of such a ring in the complex CllHl by no means unlikely. The ozonide C27H460,0 must be formed by the addition of 0 to an ozonide C2,H4,0,0, and experiments 8 and 9 point to this composition a t the first stage of the action.The presence of oxozone when crude ozone is used tends to the production of com- pounds [+04J and in each case such an ozonide is indicated by analysis. Prolonged treatment increases the proportion of oxozone combining and explains the formation of ozonides [ + O ] and We attempted to obtain some further evidence by the preparation of a cholesterol thio-ozonide. Erdmann (Annah 1908 362 133) put forward the view that the reactive dark-coloured form of sulphur obtained by cooling the molten substance to 160° contains molecules S, and is the analogue of ozone. He found that it reacted with unsaturated organic compounds forming thio-ozonides especially with those containing a terminal vinyl group.Linalyl acetate f o r example, gave a derivative C,,H,,O,,S, whatever the proportion of sulphur employed ; linalool instead of the expected CloHlsO,Ss gave a compound C,,H,,O,S, with the evolution of hydrogen sulphide. c + 081. The following experiments were carried out with cholesterol : ( a ) Ten grams were mixed with 3.5 grams of sulphur (S requ'ires 2.5) and heated a t 160-165O f o r eight hours. Odour of hydrogen sulphide was not observed. The dark red mass was dissolved in carbon disulphide and the solvent allowed t o evaporak. The uncombined sulphur crystal- lised out and was separated by washing with cold benzene. Sulphur recovered 1.1 grams; combined 2.4 grams. The compound was insoluble in alcohol but moderately soluble in light petroleum and very readily so in ether o r benzene.It was purified by repeated precipitation with alcohol from benzene solution and obtained as a dark red brittle solid melting indefi- nitely a t 1 1 1 O : 0.1264 gave 0.1957 BaS04. S=21*2. C,,H4,OS3 requires S = 20.0 per cent. ( b ) Several preparations were made in which the proportion of sulphur and the method of purification were varied. The sulphur- content in each case was between 19 and 21 per cent. The sub- stances proved extremely difficult to burn and no satisfactory analyses could be obtained the values found f o r carbon being 1-2 per cent. tool low. F o r example 10 grams of cholesterol were heated with 8 grams of sulphur and the compound was purified b OF CHOLESTEROL AND COPROSTEKOI,.PARTS 111. AND IV. 53 fractional precipitation by alcohol from ethereal solution and cxtrsction of ths product with much boiling alcohol : 0.1654 gave 0.3947 CO and 0.1291 H20. 0.2234 , 0.3254 BaSO,. S=19.65. C,7H4,0S requires C = 67.2 ; H = 9.5 ; S = 20.0 per cent. These analyses point t o the formation of a compound having the composition of a thio-ozonide of cholesterol and if Erdmann’s assumptions are correct they indicate that only one unsaturated linking capable of combining with sulphur is present in cholesterol. C-65.1; H=8*6. Part IV.-TJhe Action of Bromine on Cholesteryl Benzoate. Obermuller (Zeitsch. pkysiol. Chem. 1891 15 42) observed that cliolesteryl benzoate when treated with bromine in carbon disul- phide solution gave a monobromo-substituted derivative, C,7H4,0Br*C7H,0 (m.p. 138O) instead of the expected cholesteryl benzoab dibromide. As the benzoate in this respect appears t o differ from the other esters of cholesterol we have repeated these experiments and extended them t-o the nitrobenzoyl esters. An ethereal solution of cholesteryl benzoate was found to be indifferent towards bromine when treated with a solution of the reagent in glacial acetic acid. No absorption of bromine appeared to bake place when the benzoate was dissolved in chloroform o r carbon disulphide and treated a t Oo with a solution of bromine in the same solvent but in a few minutes the colour of the bromine disappeared and tliereaf ter it was instantly absorbed until the reaction was complete.Hydrogen bromide was evolved in small amount during and after treatlment. Approximately 2 atoms of bromine were required per molecule of the benzoate. Two experi- ments showed (a) 3-35 ( b ) 3.45 grams of bromine per 10 grams of benzoate whilst the absorption of 2 atoms of bromine requires 3.26 grams. The crystalline product was very readily soluble in ether, chloroform or benzene sparingly so in light petroleum or acetone, and insoluble in alcohol. The microscope showed i t to be it mixture which was resolved as follows After crystallisation from a mixture of alcohol and chloroform the mixture was crystallised from acetone. The first crops after washing with warm acetone con- sisted of soft needles melting a t 1 3 8 O (product A).The acetone filtrates on keeping deposited heavy flat hexagonal tablets whicb were hard and gritty (product B). Shaking with light petroleum and repetition enabled them to be separated from the lighter nillases of product A 54 CONTRIBUTIONS TO THE CHEMISTRY OF CHOLESTEROL ETC. Prcduct A was osbtaiiied in fine needles which when magnified, were seen to consist of narrow six-sided prisms. These melted a t 140-142O and in other respects corresponded with the description of the monobromocholesteryl benzoate given by Obermuller. Product B was purified by solution in acetone. On slow evapora- tion crystals more than 1 cm. in diameter could be obtained which melted a t 168-169O. The yield was not more than 20 per cent. of the t-otal product : 0.1253 gave 0.2847 CO and 0.0857 H,O.C=62.0; H=7*7. 0.2819 required 8-76 C.C. N / 10-AgNO,. Br = 24.85. On treatment with zinc dust and acetic acid it was quantitatively reduced t o cholesteryl benzoate. It would appear therefore to bO the hitherto unknown cholesteryl benzoate dibromide. Cholesteryl m-nitrobenzoate CsH4904N was prepared by heatc ing together m-nitrobenzoyl chloride with cholesterol a t 165O for five minutes. The product was almost insoluble in cold more readily in boiling alcohol ; sparingly soluble in acetone and readily so in benzene o r chloroform. After crystallisation from a mixture of alcohol and benzene the m-nitrobenzoate melted a t 137O to a turbid liquid which became clear a t 170O. On cooling a charac- teristic play of colours violet green red and orange was observed.Cholesteryl p-nitrobenzoate was also prepared and closely resembled the m-compound. The crude product was extracted with alcohol and the residue crystallised from acetone and obtained in glistening plates which melted a t 185O (turbid liquid) decomposing a t 250O. A play of colour violet green and red was here again observed. On the addition of bromine t o a solution of these esters in carbon disulphide solutiion an induction period was again noticed after which absorption rapidly took place. Two atoms of bromine per molecule of the nitrobenzoate were required f o r complete reaction. Measurement of the hydrogen bromide evolved gave the following results : One gram-molecule (535 grams) of cholesteryl m-nitrobenzoate required 159 grams (2 atoms) of bromine of which 52 grams were evolved as hydrogen bromide.These numbers indicate a substitu- tion of 52 and an addition of 55 grams of bromine leaving some portion of the1 ester unaltered. Examination of the product con- firmed this. A crystalline monobromo-derivative was separated in a yield of 20 per cent. of the theoretical. Tl10 residue consisted of a resinous mass probably the impure dibromide which could not be crystallised. Br o rn oc h oles t eryl m-ni t iro b e tz x oa t e was isolated from the crude CsH,,O,Br requires C = 62.7 ; H = 7.8 ; Br = 24.6 per cent STUDIES IN CATALYSIS. PART 111. 55 preparation by extraction with acetone. ethyl acetate i t formed leaf-life crystals melting at 149O: 0.7624 gave 0.2423 AgBr. Br=13*5.C,H4,0,NBr requires Br= 13.0 per cent. From these results i t would appear that it is the negative char- acter of the benzoyl groupings that hinders the reactivity of the unsaturated linking. The acid C2,H4,04 in which this linking is also present is quite indifferent towards bromine although active towards ozon0. Summary. On recrystallisation from (1) Evidence obtained from the action of ozone indicates that only one double linking exists in cholesterol. (2) The secondary unsaturation shown by the existence of ozonides C,,H,,O-O, is probably due to the opening up of a bridged sing. (3) With sulphur cholesterol yields a compound which approxi- mates to the composition of a ' monothio-ozonide,' C,,H,,O*S,. (4) The cholesteryl benzoates unlike the other esters form mono- brorno-substitution derivatives with bromine as well as a dibromide.The expenses of Part IIL. of this series were covered by a grant from the Government Grant Committee of the Royal Society and those of Part IV. by a grant from the Research Fund of the Chemical Society for both of which we desire to express our thanks. CHEMISTRY DEPARTMENT, BOROUGH POLYTECHNIC INSTITUTE S.E. [Received December 9th 1 915. 46 DORkE AND ORANGE CONTRIBUTIONS TO THE CHEMISTRY 1V.-Contributions to the Chemistry of Cholesterol and Cop-osterol. Part 111. The Ozonides o j Cholesterol. Part IV. The Action of Bromine on Cholesteryl Benzoate. By CHARLES DORBE and LIONEL ORANGE. RECENT researches on the constitution of cholesterol may be sum- inarised in the following formula : CH CH, /-\ I n addition to the double linking situated in the end vinyl group, the possible presence of another which must be placed in the complex CllHl, has been the subject of considerable discussion.The evidence rests almost entirely on the1 nature of the action of ozone on cholesterol but in different hands this reagent has given very different results. I n 1908 one of us firs€ showed that whilst cholesterol formed an ozonide C27H4G0*03 the corresponding ketone gave a compound C2,H,,0*0,. This number was in keeping with the views then held f o r the number of oxygen atoms which should combine witlh a doubly unsaturated ketone and the statement was made that “itl may be that cholestenone contains a second ethylene linking which is not rendered apparent by the usual reagents ” (T..1908 93 1330). Langheld (Bey. 1908 41 1023) mentioned that cholesterol appears to combine’ with a t leash two ozone complexes. Diels, however found that the crude ozonide after treatment with alcohol lost oxygen and that a definite compound C27H4@,04, resulted for which he suggested a peroxide formula (Ber. 1908, 41 2596). Molinari and Fenaroli obtained in ethereal solution, products containing between 3 and 4 atoms of combined oxygen per molecule but later using ozonised air of 1 per cent. ozone concentration they obtained ozonidea C27H4G0,0G stable a t 60°, from cholesterol and phytosterol. They stated that these results confirmed the existence of a second double linking in chblesterol, and further claimed that thO increase in weight of an unsaturated compound after treatmentl with ozone gave an ‘ozone number’ which was a measure of the number of ethylene linkings presen OF CHOLESTEROL AND COPROSTEROL.PARTS 111. AND IV. 47 ( I ~ w . 1908 41 2785). This method was extended by one of us to a series of derivatives of cholesterol (T. 1909 95 638). Under prolonged action of the ozone stream cholesterol and cholestenone were found to combine with two molecular proportions of ozone. The saturated dihydrocholesterols P-cholestanol and coprosterol, still gave ozonides C,7H4&),03 but the ketones corresponding with these showed an addition of 0,. This inconsistency has recently been criticised by von Furth and Felsenreich (Biochem. Zeitsch., 1915 69 416) but it serves to show that the action of ozone is not as simple as was a t first assume'd.The behaviour of the dibasic acid C&&,O (Diels BET. 1906 36 3177) was also examined. It is insoluble in ordinary solvents but the ozonides are readily soluble. A t the solution point (six hours) an ozonide C2,H4,04,03, formed by saturation of the characteristic double linking was obtained ; after one hundred hours a compound C,,H,,O,,O, was produced. Harries (Ber. 1912 45 936; Annalen 1912 390 235) has brought evidence t o show that the 16 per cent. ozone used by him contains two constituents normal ozone 0, and oxozone 0, in the proportion 20 0,. I f the crude ozone is passed through dilute sodium hydroxide solution the oxozone is destroyed. This 'washed' ozone gives normal ozonides by the addition of 0 for each unsaturated linking.The ' crude' ozone gives not only the ozonide X,O, and tlhe oxozonide' X,O, but also polymerides of these (XO,) and (X04)2. Further treatment with crude ozone leaves the normal ozonide unchanged but the bisozonide is con- verted into the bisoxozonide. These results may account for the varying compositions assigned to the cholesterol ozonides but they do not seem to have been taken into account by von Furth and Felsenreich who working with 8 per cent. ozone and using the ozonsnumber method with chloroform as a solvent found that the molecule of cholesterol absorbs 1.5-4 molecules of ozone! dihydrocholesterol 1.8-4.3 and cholestane 1-5-4.8 molecules. They conclude that only one double linking exists in the cholesterol molecule.The ozonide results are however, of little value. We have found that chloroform is strongly attacked by ozone and that i t is impossible to obtain a product free from chlorine when it is employed. Harries has found (Zoc. cit.) that cholesterol with washed ozone in hexane solution gives an ozonide, C27H460,03 but in carbon tetrachloride solution an ozonide, c 2 7 H 4 6 0 0 6 is formed. The method of working namely evapora- tion of the solvent in the 8 per cent. ozone current and analysis of the crude residue obtained does not seem very satisfact,ory. 1% has moreover lately been recognised that for the study of con 48 DORJkE AND ORANGE CONTRIBUTIONS TO THE CHEMISTRY stitution a weaker ozone of 1 to 2 per cent. concentration is desirable.We have therefore made some preparations and analyses of the ozonides of cholesterol and of the acid C27n4404 in the hops of explaining some of these points which are of importance both from the point of view of the constitution of cholesterol and of the nature1 of ozone action. EXPERIMENTAL. The ozonised oxygen used contained 2 per cent. of ozone. Passage through sulphuric acid did not alter the concentration but after 'washing' through 5 per cent. sodium hydroxide it became 0.8 per cent. The cholesterol was ozonised in glacial acetic acid suspension except in one or two cases. The acid C27H4404 was suspended in acetone,. Acetic acid appears to be quite unaffected by ozone and acetone is only slightly attacked. I n order to form the ozonide C,,H4,O*O, 1 gram of cholesterol requires 0.12 gram of ozone.Under t8Ke conditions of experiment this quantity was passed through per hour. All the products obtained had very similar properties and were worked up in the same way. The acetic x i d solution was diluted with its own volume of light petroleum and poured into1 twice the volume of water. The ozonide floated as a clot on the1 aqueous layer and was easily removed. The aqueous layer was extracted with ether the ethereal solution shaken with dilute alkali and the ether evaporated. The total ozonide was dissolved in a little ether and precipitated with light petroleum this process being several times repea$.ed. The viscous clot obtained foamed in a high vacuum and dried to a britttle solid which was easily powdered.It was kept for fourteen days in a vacuum before analysis. The acetone solutions were allowed t o evaporate and the ozonide treated as before. The following tables give the theoretical composition of the various ozonides : TABLE I. Theoretical Composition of the Ozoiiides of Cholesterol, [H = 1.008]. C,vH,,O f0.q +O -to +Oc +O to,. C ...... ......... 83-85 74.59 71.94 69.52 67.17 65.01 62.98 H ............... 12.00 10.68 10.30 9.95 9.62 9-31 9.0 O F CHOLESTEROL AND COPROSTEROL. PARTS 111. AND 1V. 49 TABLE 11. Yheoretical Composition of the Ozorzides of t h e Acid C27H44O4. C&f&4O* 3 - 0 3 + 0 4 t-06 +o + 0 7 C ............... 74.95 67.45 65-32 63.28 61-36 59.55 67-85 13 ............... 10.26 9.23 8.87 8.44 8.19 8.09 7-85 The first series of experimelnts table 111 was made t o discover whether cholesterol formed a higher ozonide than C,,H4,0 + O,, and if so what was the maximum number of oxygen atoms which could combine.TABLE 111. The Extre,me Action of Ozone on Cholesterol. Total time Time in Rotation of treatment hours per Ozone of Analysis with ozone 1 gram passed ozonide. Analysis indicates 13xpt. in hours. treated. through [a]:* C/H addition of 1. 16 3 H,SO A" 66.66 66.52 O6 9-57 9.55 2. 18 0 H2S04 - 64.97 65.56 0 7 9.21 9.05 3. 21 10 - 21" 65.01 65.27 0, 8.93 8.94 NaOH 67.35 67.28 0, 4. 60 24 3- 31" 9-75 9.83 %SO4 Glacial acetic acid was used except in e'xperiment 1 which was in acetone! solution. Rotations were measured in ethereal solution. It being evident that a maximum of 6 or 7 atoms of oxygen could combine atltempta were made to asce'rtain the composition of the ozonide formed in the first stsage of the action that is during the saturation of the double1 link in the terminal vinyl grouping.The results are given in table IV 50 DORkE AND ORANGE CONTRIRUTIONS TO THE CHEMISTRY TABLE IV. The Action of Ozoite O?L Cholesterol in Gluciul Acetic A c i d S m p e n - sion t h e experinzeizts being stopped as soon as tlze cholesterol had passed i n t o solutioiz as ozonide. Total Timein Rotation time of hours per Ozone Expt. treatment 1 gram passed in hours. treated. through 5. 8 2+ H,SO, 6. 5 24 - - 7. 17 3 NaOH 8. 3 1; + H7W4 %SO4 NaOH 9. 10 4 + of Analysis ozonide. Analysis indicates [.I1 C/H.addition of 35" 73.03 72.33 03-4 10.49 10.41 35" 71.11 70.98 0, 10-35 10.10 23"* 71-13 71.21 0, 10.63 10.46 74.89 75.01 0,- + - 10.27 10.86 73.38 73.24 O,+ 45' 10-71 10.64 * Chloroform solution. I n so far as the solution point can be taken as indicating the end of the first stage these results point t o an addition of O4 with crude ozone and with less cert'ainty of 0 with normal ozone. The product obtained in experiment 8 was fully saturated towards bromine. In the following experiments 10-12 made with the acid C27H4404 the solution point was far sharper. I n experiment 11 the action was stopped short of complete solution and the excess of acid removed by filtration. Experiments 13 and 14 show the effect of prolonged treatment with crude and washed ozone respectively.TABLE V. Action of Ozone on t h e Acid C27H4404 m. p . 290O. Total Time in Rotation time of hoursper Ozone of Analysis Expt. treatment 1 gram passed ozonide. Analysis indicates in hours. treated. through [a]:* C/H. addition of 10. 6 2 H,SO 30" 64-45 63-85 04-5 11. 13 4 H,SO 28" 65-03 64.71 0, 8-96 8-83 9.07 9.0 OF CHOLESTEROL AND COPROSTEROL. PARTS 111. AND IV. 51 TABLE v (continued). ,4ctioi~ of Ozoize on t h e Licid C27n4404 m. p . 290'. Total Time in Rotation time of house per Ozone of Analysis Expt. treatment 1 grain passed ozonide. Analysis indicates in hours. treated. through [a]:8 CIH. addition of NaOH 12. 8 4 + - 64.09 64.20 04-5 H2SO4 8.91 9.49 13. 25 8 H,S04 13" 57-78 57-66 0s 7.55 7.40 NaOH 59.21 59.74 0, H?SOg - 8.00 8-15 14.33 20 + Statement ixnd Discussion of Results. (1) It will be seen that cholesterol combines with a maximum of seven atoms of oxygen under the action of ozone. With 'washed' or normal ozone the number is exactly six. The acid C2,Hi404 forms compounds [ + O,] and [+ O,] on treatment with washed and crude ozone respectively. (2) Expariments made to determine the first stage in the action point to the formation of an ozonide [+O,] with cholesterol and washed ozone but t o an addition of 0 in other cases. (3) All the products obtained were brittle solids which melted between 8 5 O and 95O and decomposed above looo. Like other saturated cholesterol derivatives they were dextrorotatory the values f o r those made with washed ozone1 being higher than those made with crude ozone.No information was obtained by a study of their decomposition products resinous acids and neutral sub- stances being formed. The formation of an ozonide C27H460,06 as an end-product with ozone must we think bel taken as evidence that a second double linking has entered into combinatlon. The comparatively slow rate a t which the reaction takes place leads to the conclusion that this linking does not exist pre-formed in the cholesterol molecule, but is developed under the action of ozone. Harries has suggested that water is eliminate'd from the grouping -CH(OH)*CH,* with the formation of a cholesterylene but this explanation scarcely holds in the case of the ketones and the acid C27H4404 which also form higher ozonides.The explanation previously given by one of us that the secondary unsaturation is due to the opening up of a bridged ring under the influence of ozone seems more probable, and has been adopted by von Furth and Felsenreich. Theanalogie 52 DOREE AND ORANGE CONTRIBUTIONS TO THE CHEMISTRY of cholesterol with the terpenes render the presence of such a ring in the complex CllHl by no means unlikely. The ozonide C27H460,0 must be formed by the addition of 0 to an ozonide C2,H4,0,0, and experiments 8 and 9 point to this composition a t the first stage of the action. The presence of oxozone when crude ozone is used tends to the production of com- pounds [+04J and in each case such an ozonide is indicated by analysis. Prolonged treatment increases the proportion of oxozone combining and explains the formation of ozonides [ + O ] and We attempted to obtain some further evidence by the preparation of a cholesterol thio-ozonide.Erdmann (Annah 1908 362 133) put forward the view that the reactive dark-coloured form of sulphur obtained by cooling the molten substance to 160° contains molecules S, and is the analogue of ozone. He found that it reacted with unsaturated organic compounds forming thio-ozonides especially with those containing a terminal vinyl group. Linalyl acetate f o r example, gave a derivative C,,H,,O,,S, whatever the proportion of sulphur employed ; linalool instead of the expected CloHlsO,Ss gave a compound C,,H,,O,S, with the evolution of hydrogen sulphide. c + 081. The following experiments were carried out with cholesterol : ( a ) Ten grams were mixed with 3.5 grams of sulphur (S requ'ires 2.5) and heated a t 160-165O f o r eight hours.Odour of hydrogen sulphide was not observed. The dark red mass was dissolved in carbon disulphide and the solvent allowed t o evaporak. The uncombined sulphur crystal- lised out and was separated by washing with cold benzene. Sulphur recovered 1.1 grams; combined 2.4 grams. The compound was insoluble in alcohol but moderately soluble in light petroleum and very readily so in ether o r benzene. It was purified by repeated precipitation with alcohol from benzene solution and obtained as a dark red brittle solid melting indefi- nitely a t 1 1 1 O : 0.1264 gave 0.1957 BaS04. S=21*2. C,,H4,OS3 requires S = 20.0 per cent.( b ) Several preparations were made in which the proportion of sulphur and the method of purification were varied. The sulphur- content in each case was between 19 and 21 per cent. The sub- stances proved extremely difficult to burn and no satisfactory analyses could be obtained the values found f o r carbon being 1-2 per cent. tool low. F o r example 10 grams of cholesterol were heated with 8 grams of sulphur and the compound was purified b OF CHOLESTEROL AND COPROSTEKOI,. PARTS 111. AND IV. 53 fractional precipitation by alcohol from ethereal solution and cxtrsction of ths product with much boiling alcohol : 0.1654 gave 0.3947 CO and 0.1291 H20. 0.2234 , 0.3254 BaSO,. S=19.65. C,7H4,0S requires C = 67.2 ; H = 9.5 ; S = 20.0 per cent.These analyses point t o the formation of a compound having the composition of a thio-ozonide of cholesterol and if Erdmann’s assumptions are correct they indicate that only one unsaturated linking capable of combining with sulphur is present in cholesterol. C-65.1; H=8*6. Part IV.-TJhe Action of Bromine on Cholesteryl Benzoate. Obermuller (Zeitsch. pkysiol. Chem. 1891 15 42) observed that cliolesteryl benzoate when treated with bromine in carbon disul- phide solution gave a monobromo-substituted derivative, C,7H4,0Br*C7H,0 (m. p. 138O) instead of the expected cholesteryl benzoab dibromide. As the benzoate in this respect appears t o differ from the other esters of cholesterol we have repeated these experiments and extended them t-o the nitrobenzoyl esters.An ethereal solution of cholesteryl benzoate was found to be indifferent towards bromine when treated with a solution of the reagent in glacial acetic acid. No absorption of bromine appeared to bake place when the benzoate was dissolved in chloroform o r carbon disulphide and treated a t Oo with a solution of bromine in the same solvent but in a few minutes the colour of the bromine disappeared and tliereaf ter it was instantly absorbed until the reaction was complete. Hydrogen bromide was evolved in small amount during and after treatlment. Approximately 2 atoms of bromine were required per molecule of the benzoate. Two experi- ments showed (a) 3-35 ( b ) 3.45 grams of bromine per 10 grams of benzoate whilst the absorption of 2 atoms of bromine requires 3.26 grams.The crystalline product was very readily soluble in ether, chloroform or benzene sparingly so in light petroleum or acetone, and insoluble in alcohol. The microscope showed i t to be it mixture which was resolved as follows After crystallisation from a mixture of alcohol and chloroform the mixture was crystallised from acetone. The first crops after washing with warm acetone con- sisted of soft needles melting a t 1 3 8 O (product A). The acetone filtrates on keeping deposited heavy flat hexagonal tablets whicb were hard and gritty (product B). Shaking with light petroleum and repetition enabled them to be separated from the lighter nillases of product A 54 CONTRIBUTIONS TO THE CHEMISTRY OF CHOLESTEROL ETC. Prcduct A was osbtaiiied in fine needles which when magnified, were seen to consist of narrow six-sided prisms.These melted a t 140-142O and in other respects corresponded with the description of the monobromocholesteryl benzoate given by Obermuller. Product B was purified by solution in acetone. On slow evapora- tion crystals more than 1 cm. in diameter could be obtained which melted a t 168-169O. The yield was not more than 20 per cent. of the t-otal product : 0.1253 gave 0.2847 CO and 0.0857 H,O. C=62.0; H=7*7. 0.2819 required 8-76 C.C. N / 10-AgNO,. Br = 24.85. On treatment with zinc dust and acetic acid it was quantitatively reduced t o cholesteryl benzoate. It would appear therefore to bO the hitherto unknown cholesteryl benzoate dibromide. Cholesteryl m-nitrobenzoate CsH4904N was prepared by heatc ing together m-nitrobenzoyl chloride with cholesterol a t 165O for five minutes.The product was almost insoluble in cold more readily in boiling alcohol ; sparingly soluble in acetone and readily so in benzene o r chloroform. After crystallisation from a mixture of alcohol and benzene the m-nitrobenzoate melted a t 137O to a turbid liquid which became clear a t 170O. On cooling a charac- teristic play of colours violet green red and orange was observed. Cholesteryl p-nitrobenzoate was also prepared and closely resembled the m-compound. The crude product was extracted with alcohol and the residue crystallised from acetone and obtained in glistening plates which melted a t 185O (turbid liquid) decomposing a t 250O. A play of colour violet green and red was here again observed.On the addition of bromine t o a solution of these esters in carbon disulphide solutiion an induction period was again noticed after which absorption rapidly took place. Two atoms of bromine per molecule of the nitrobenzoate were required f o r complete reaction. Measurement of the hydrogen bromide evolved gave the following results : One gram-molecule (535 grams) of cholesteryl m-nitrobenzoate required 159 grams (2 atoms) of bromine of which 52 grams were evolved as hydrogen bromide. These numbers indicate a substitu- tion of 52 and an addition of 55 grams of bromine leaving some portion of the1 ester unaltered. Examination of the product con- firmed this. A crystalline monobromo-derivative was separated in a yield of 20 per cent.of the theoretical. Tl10 residue consisted of a resinous mass probably the impure dibromide which could not be crystallised. Br o rn oc h oles t eryl m-ni t iro b e tz x oa t e was isolated from the crude CsH,,O,Br requires C = 62.7 ; H = 7.8 ; Br = 24.6 per cent STUDIES IN CATALYSIS. PART 111. 55 preparation by extraction with acetone. ethyl acetate i t formed leaf-life crystals melting at 149O: 0.7624 gave 0.2423 AgBr. Br=13*5. C,H4,0,NBr requires Br= 13.0 per cent. From these results i t would appear that it is the negative char- acter of the benzoyl groupings that hinders the reactivity of the unsaturated linking. The acid C2,H4,04 in which this linking is also present is quite indifferent towards bromine although active towards ozon0. Summary. On recrystallisation from (1) Evidence obtained from the action of ozone indicates that only one double linking exists in cholesterol. (2) The secondary unsaturation shown by the existence of ozonides C,,H,,O-O, is probably due to the opening up of a bridged sing. (3) With sulphur cholesterol yields a compound which approxi- mates to the composition of a ' monothio-ozonide,' C,,H,,O*S,. (4) The cholesteryl benzoates unlike the other esters form mono- brorno-substitution derivatives with bromine as well as a dibromide. The expenses of Part IIL. of this series were covered by a grant from the Government Grant Committee of the Royal Society and those of Part IV. by a grant from the Research Fund of the Chemical Society for both of which we desire to express our thanks. CHEMISTRY DEPARTMENT, BOROUGH POLYTECHNIC INSTITUTE S.E. [Received December 9th 1 915.

ISSN:0368-1645    

DOI:10.1039/CT9160900046

出版商:RSC    

年代:1916

数据来源: RSC

摘要:

STUDIES IN CATALYSIS. PART 111. 55 V.-Studies in Catalysis. Part 111. Preliminary Measurements of t h In fra-red Absorptson Spectra o j Hydrogen Chloride Potassium Chloride aqad Methyl Acetate in Aqueous Solution. By RAPHAEL HEBER CALI.OW WILLIAM CUDMORE MCCULLAGH LEWIS and GERALD XODDER. IN Parts I. and 11. of this series (Lamble and Lewis T. 1914 105, 2330; 1915 107 233) it has been shown on the assumption t h a t homogeneous catalysis (such as the effect of acid on the inversion of sucrose and on the hydrolysis of methyl acetate) is due t o the infra-red radiation emitted in quanta by the catalyst and absorbe 56 c z r m w LEWIS AND KODDER-: by the reacting substance that the radiation from the hydrogen ion, and therefore flie absorption bands possessed by it should lie in the short infra-red region.Thus from measurements of the tern- perature-coefficient of tlie rate of inversion of sucrose by hydrogen chloride it was calculat'ed that the characteristic frequency * of the hydrogen ion (neglecting any modification in numerical values which may be attributed to the effect of the solvent) should corre- spond with a wave-length of the order 1 . 2 3 ~ ~ and similarly from the rate of hydrolysis of methyl acetate it was inferred that the position of the hydrogen-ion band should be 1 . 9 ~ . The mean of these two values is 1.56 p . From the results obtained by Coblentz (" Investigations of Infra-red Spectra," 7 parts Pub. Carnegie Inst. 1905-1908) i t is evident that radiation of this order of wave- length could be absorbed by sucrose and by methyl acetIab although the most marked bands of these substances occur a t longer wave- lengths.These observations however refer t o the substances in the pure state. N-o data are available as regards the position of the dissolved methfl acetate or hydrogen-ion bands since neither ester nor strong acid has been examined in aqueous solution. It is of obvious importance f o r thO radiation view of catalysis that such data should be obtained and f o r this purpose the present investigation was carried out. According to Coblentz (Zoc. cit.) the band a t 3 p in the water- transmission spectrum is due to the hydroxyl group and the heavier the group the longer is the wavelength of the absorption band associated with it for example the SO group gives rise to bands a t 4*55p 8 .7 ~ ~ and 9 . 1 ~ . From this it would follow that if the hydrogen ion gave rise t o any bands in the infra-red region one or more of these should lie between 0 . 8 ~ and 3p provided the solvent itself did not' mask or fundamentally modify the vibration of the solute. As a matter of fact we have found that this modi- fication takes place. The bands in the water spectrum in this region occur a t 0,77 p l*O,u 1-25 p 1*50,u 2.0 p and 3.0 p of which those in heavy type are the most prominent. I n principle the present measurements consist in comparing the transmjssion of a layer of water with a layer of solution having the same thick- ness with respect to water and isolating the effect of the solute by * It is not implied that the hydrogen ion possesses only one band in the infra-red.From analogy to other substances it is probable that there are several bands and further it is impossible at the present stage to attribute the catalytic activity to a single type of vibration. The term " characteristic frequency '' might be called the " equivalent frequency," that is the single frequency which would give rise to a single quantum of sufficient size t o account quanta.titively for the effects observed. Whether this is due actually t o a single quantum or to more than one cannot a t present be stated STUDIES IN CATALYSIS. PART In. 57 taking the water transmission value as unity at a series of wave- lengths. E x P E R I MEN T AL. The infra-red spectrometer employed was of the IIilger pattern with a range 0-5-1Opu.The source of radiation was a Nernst filament lamp (0.6 ampere and 190 volts) regulation of current being effected by means of a variable resistance. The beam of radiation is rendered parallel and focussed by means of metallic mirrors dispersion being obtained by means of a. rock-salt prism, to which is attached a plane mirror t o give automatically minimum deviation for all wave-lengths after a single setting. The telescope of the instrument is fitted for a thermopile to be used in conjunc- tion with a galvanometer. Many attempts were made with different thermopiles used in conjunction with a Broca galvanometer to obtain readings but owing t o the lack of steadiness of the galvano- ineter zero due to variable magnetic effects the thermopile-galvano- meter system had to be :ibandoned and in its place a radio- tnicromet'er was employed which was found t o give consistent read- ings.This instrument which was constructed in the laboratory, was of the type reconimended by H. C. Jones (2. -4mer. Chent. SOC. 1915 37 776). The suspended system consisted of a loop of pure silver wire supported on a glass frame and connected t o a thermo-couple of the same kind as that employed by Jones. The Compensating junction was found to nullify to a very large extent the disturbing influence of local variations in temperature due to the warming of the slit on exposure1 t o the radiation. The radia- tion ent'ered through a fluorite window and was received on a thin plate of pure silver blackened and attached to the couple.The deflexioii o f the suspended system was observed on an illuminated scale by means of a reading telescope. I n all cases it was found most advant.ageous t o use the instrument ballistically that is t o take the reading of the deflexion produced and completed in seven or eight seconds after raising the water-cooled shutter placed in front of the Nernst la-mp. The water and hydrogen chIoride solutions were examined in quartz cells two such cells being mounted on a horizontal carrier (placed directly in front of the collimator slit) so that the cells could be rapidly inkrchanged f o r comparative measurements. By means of stops i t was possible t o bring the cells to exactly the same position throughout a series of measurements in order that the amount of absorption due t o the quartz walls would reniain con- stant.The two cells most. frequently employed were made by cenleiitiiig quartz plates (each 1.5 mm. thick) on to the sides of 58 CALLOW LEWIS AND NODDER thin U-shaped piece of glass with marine glue. The internal thick- ness of each cell was approximately 0.9 mm. Sensitivity of tJze Rndiomicrometer.-Sensitivity is defined as the deflexion in cm. per sq. mm. of exposed surface for a standard candle and scale each a t a distance of 1 metre. The candle and scale were set up a t this distance from the instrument and the deflexion observed. Seven concordant readings gave as a mean value 6.0 cm. The width of the slit was 0.5 mm. and the area of exposed vane 1.5 mmq2. Hence the1 sensitivity is 4.0.I n t h i s measurement tke instrument was exhausted t o a pressure of 2-3 mm. Jones (loc. c i t . ) obtained a sensitivity of 5 with his instrument under the ordinary pressure. Our instrument in its present form is therefore less sensitive but has the advantage of possessing a very short period. CJbange in Sensitivity with Time.-Before taking any set of read- ings the deflexion with the spectrometer drum set a t 2 p was in all cases noted the width of the slit of the radiomicrometer being permanently adjusted to 0.5 mm. and an empty quartz cell being placed in Dosition in front of the collimator. The1 deflexion remained the s3,me for several days the maximum variation f o r widely separateld intervals of time being from 161 to 155 mm. divisions c n the scale.All values were reduced where necessary to the 161 mm. deflexion as standard. Drum Calzbration.-Tkle position of the sodium-yellow and potassium-red lines indicated that the1 drum was correct f o r wave- lengths shorter than 0 . 8 ~ approximately. It was also found that the quartz band occurred a t 2.9p as Coblentz had observed and that the water bands a t 3 . 0 ~ and 4 . 7 ~ were correctly registered. Instead of oKtaining the water band a t 1*5p however the drum reading was 1 0 4 5 ~ . This relatively small error in calibration has been allowed i o r in the data quoted later. Percentage Error in the Deflexion Readings.-The region in- vestigated ext'ended approximately from 0 . 9 ~ to 2.1 p. A t each point a t least six independent readings were taken and the experi- mental error thus obtained.For deflexions greater than 200 mm. the error is negligible. F o r smaller deflexions the average peroent? age error is as follows: Deflexion in mm. ...... 150 100 50 25 10 5 Per cent. error ......... 5 1 . 0 j 1 . 2 5 k 2 - 0 f4.0 &6 &lO Comparison of t?be Transmission of the Quartz Cells. Over the range investigat,ed no measurable difference could be detected. When the cells were filled with water differences were observed, Readings were taken through each cell when empty STUDIES IN CATALYSIS. PART JII. 59 due to the difference in thickness of the water layers. A table of values was constructed from which observations made on one cell could be seduced t o that of the other. This correction will be referred to as the “ correction for the dimensions” of the cell.Infra-red Absorption of Hydrogen Chloride in Aqueous Solution. The acid was investigated a t three concentrations namely N , 2LT and 3 N . The solution was placed in one of the cells ( A ) , water in the other ( B ) arid readings taken alternately through each. I n order that the deflexions may be comparable i t is neces- sary to correct f o r the displacement of water by the acid in the SCiL tions. This was effected by determining the absorption-coeffi- cieqt k of water. One of the cells was made considerably thinner tl*aii the other the difference being 0.55 mm. The law of absorp- tim is expressed thus : I = Ioe - kd, where Z is the intensity of the incident radiation Z the intensity of the emergent beam and d the thickness of the water-layer traversed.Knowing the thickness of the two cells exposed to the same intensity I, it is possible t o calculate k f o r water a t various wavelengths. I n this way a table of values was prepared the unit of length being the millimetre. Returning to the comparison of cells A and B when the solution in A was compared with the water in B i t was possible from a knowledge of the density of the solution to determine the effective thickness of water actually traversed; knowing k it was then possible t o alter the values of the deflexion observed through cell B t o the values which would have been given if the beam had traversed in B the same thickness of water as exists in cell A . This was the general method of correction adopted and will be referred t o as the “correction for density.” A single illustration of the method may be given in the case of AT-hydrochloric acid.The density of this solution is 1.017. The concentrat,ion of hydrogen chloride is 0.0365 gram per c.c., and hence the concentration of water (in cell A containing the solute) is 0.9805 gram per C.C. The concentration of water in cell 11 is taken as unity. Hence, Effective thickness of water in A - 0 9805 Thickness of water in B 1 ’ --__-- - ____ The actual thickness of the water-layer in R is 0.92 1nm. Hence the effective thickness of water in A is 0.90 mm. The intensity Z, of the emergent beam from cell B thus requires to be increased to the value i t would have (1’) if the cell were 0.90 mm.thick. It is evident that I’=Zek(0*9~--0*9(’) =ZeO*ozk. Witll the aid of the know 60 CALLOW LEWIS AND NODDER: Deflexion Deflexion through through value of I; one can thus calculate I' from the observed value I. On dividing the defle'xion through the acid solution by that through the water (corrected) a t the same wave-length the absorption due to the acid alone can be determined. The following table (I) con- tains in full the results obtained and the extent of the corrections in the case of ,!!-hydrochloric acid. The remaining tables do not contain the corrections the final corrected results only being given. Deflexion Deflexion through through TABLE I. 1-55 96.0 114.4 0.84 1.65 85-4 99.1 0.86 1.74 64.5 79.3 0.81 1.84 43.2 57.0 0.76 1-93 32.2 45-8 0.70 2.03 23.3 38.8 0.60 2-13 21.0 32.3 0.65 N-Hydrochloric 4 cid.1-55 79.5 116.8 0.68 1.65 68.1 103.0 0.66 1.74 41.4 73.7 0.56 1.84 25.2 60.0 0.42 1-93 16.0 40.6 0-40 2.03 11.0 37.4 0-30 2.13 10.0 31.9 0-31 Waye-leng th in p. 1.0 1.12 1.23 1-34 1.45 1.50 1.55 1.65 1.74 1.84 1.93 2-03 2-13 Deflexion through acid solution (cell A ) . 40.0 mm. 67.0 82-0 66.0 54.5 99.0 109.0 96.8 66.6 42.0 34.9 30-3 23.0 Deflexion observed through water (cell B). 39.3 mm. 63-5 78.8 62.0 78.5 95.5 105.5 95-5 68-0 45.1 34.6 29.9 23.6 Deflexion through B corrected for dimensions. 38.5 mm. 65.4 80.4 65.1 84.8 104.1 117.1 102.2 75.5 50.1 40.1 35.9 30.0 Deflexioll Deflesion through B inA corrected corrected for deflexion density.in B. 38.6mm. 1.04 65-6 1.02 80-8 1.01 65-7 1.00 85.8 0.98 106.4 0.93 119.6 0.91 104.0 0.93 76.9 0.86 51.1 0.82 41-3 0.86 37.2 0.81 31.4 0.73 The last colunin gives the fractional absorption exertled by the solutel. TABLE 11. 2N-Hydr o c hloric A c id. TABLE 111. 3N-Hydrochloric A cid STUDIES IN CXTALTSIS. PART 111. 61 considerable absorption in the short infra-red region. As far as this goes it supports qualitatively the conclusion already drawn as regards the effect of the acid in increasing the radiation density in this region. A t the shorter wave end of the region examined the solution is actually more transparent than water. This pheno- menon has been observed by others and must be attributed to a change in the nature and therefore in the absorbing power of a certain fraction of ths water molecules brought about by the presence of the dissolved snbstance.This may mean either that the FIG. 1. Tlrave-length in p, solute has become hydrated the water of hydration being physically different from water in the uncombined state or t'he presence of the solute may have altered the position of equilibrium obtaining between the various polymerised forms of water and have thus altered the absorption power. The most important conclusion, however which can be drawn from the data obtained has t o do with the positions and relative intensities of the bands of the acid. There appear t o be a t least three bands in this region namely a weak one a t 1.12 p a stronger band a t 1.55 p and one still stronge 62 CALLOW LEWIS AND NODDER : in the neighbourhood of 2.1 p.* Thel resemblance of these bands in respect of position and intensity to the bands exhibited by water itself namely 1*Op 1.5 p and 2*0p is so remarkable that it is fairly evident that the solvent has exerted a very marked effect on the type of vibration possessed by the solute.I n view of this apparent effect it is necessary to examine the behaviour of other substances dissolved in water. F o r the sake of comparison methyl acetate was therefore investigated in aqueous solution as being a substance totally unlike hydrogen chloride in respect of chemical constitution and yetl catalysable by the acid a d further potassium chloride was examined as being similar to hydrogen chloride yet lacking its catalysing power.Infra-red Absorption of Methyl Acetate in Aqueous Solution. The concentration of methyl acetate employed was 10 C.C. in 50 C.C. of solution or approximately 2.6N. This concentration was chosen as i t corresponded with one of the cases investigated by Mr. Griffith and described in the succeeding paper (Part IV). The results are given in the following table: Wave-length in p. 1.0 1.12 1.23 1.34 1-45 1-50 1.55 1-65 1.74 1.84 1.93 2-03 2-13 TABLE IV. 2*6N-iliethyZ Acetate. Deflexion through Deflexion through ester solution water (corr.1 Deflexion through A . (cell A). (cell B). DeAexion through B. 43.2 mm. 42.1 mm.1.03 71.3 73.2 0.96 97.3 96.6 1.00 82.5 79.3 1.04 96.6 92.1 1.05 119.0 127.1 0.93 133.5 144.4 0.92 132.8 137.9 0.96 101.7 105-2 0.97 68.4 71.4 0.95 52.3 61.0 0.86 40.8 53.2 0.77 34.6 61.6 0.67 These data are shown by one of the curves in Fig. 2. It will be observed t$at the ester exerts considerably less absorption than hydrogen chloride a t the same concentration. Again however, bands appear a t 1.12 p 1.55 p and another beyond 2.1 p. Coblelntz (Zoc. c d . ) has examined the pure subst,ance from the wave-length 1.7 p approximately a t which point the fractional absorption was * Gaseous hydrogen chloride has according to Angstrom and Palmer an emission band at 3.98 p S'L'UDIES IN CATALYSIS. PART 111. 63 Wave- through water in length solution (corrected) DTflexion in p.(cell A ) . (cell B). in B. 1.0 38.0 mm. 34-9 mm. 1.09 1-12 62.0 61.0 1.01 1.23 75.0 72.7 1.03 1.34 68.0 62-8 1.08 1-45 99.0 97.2 1.02 1.50 121.0 114.8 1.06 1.55 124-0 126.9 0.98 1.65 108.0 104.5 1.03 1.74 74.0 74.7 0.99 1.84 56-0 53.3 1.05 1-93 46.6 45.8 1-01 2.03 41.0 39.8 1.03 2.13 32.0 35.9 0.89 , 0.81 a feeble band manifesting itself a t 2.2-2.4 p and a vexy marked band a t 3 . 3 ~ . ' through water in A solution (corrected) Deflexion (cell A ) . (cell B). in B. 39 mm. 37.0mm. 1.05 64 63.0 1.01 81 80.0 1.01 67 64.0 1-05 92 85-5 1-07 111 113.0 0.98 z 22 127-0 0.96 109 108.0 1.00 77 79.5 0.97 54 54.0 1.00 43 47.4 0.91 37 42-0 0.88 30 36.5 0.82 Infra-red Absorption of Yotussizim Chloride in Aqueous Solution.The so'lutions examined werel respectively N Z J and 3lv. The results are given in the following tables : TABLE VII. 3 N-1'0 t tzssiti rn Chloride. Wave-length in ,u. 1.0 1.12 1.23 1.34 1.45 1-50 1.55 1-65 1.74 1-84 1-93 2.03 2.13 Deflexion through solution (cell A ) . 37 mni. 61 74 75 110 131 138 121 85 G I 51 45 35 Deflexion through water (corrected) (cell B). 34-3 mm. 61.8 73.4 63.0 91.8 109.0 124.0 99.5 7 1.0 51.0 44.5 39.6 33.0 Deflexion in A Deflexion in B 1.08 0.99 1.08 1.20 1-20 1.20 1.11 1.20 1.20 1.20 1.15 1.15 1-08 These data are given in Fig. 2. It is evident that potassium chloride exerts very little absorption compaxed with hydrogen chloride'.It can theref ore only slightly affect the radiation density of the system and its transparency would even lead one t o expect that its behaviour should be the reverse of hydrogen chloride. Sinc 64 CALLOW LEWIS AND NODDER: the latter is a strong catalyst one would anticipate therefore on the radiation view that potassium chloride would exert no positive catalytic effect and might even be a feeble negative catalyst. So far as absence of effect is concerned a t the ordinary temperature experiment bears out this conclusion. It will be observed that the 3N-solution of potassium chloride is more transparent than those more dilute but that there is a reversal of the relative positions of the N - and 2N-solutions. Whether this has any considerable significance cannot be determined a t present for in those cases in which the transmission of the solution lies in the region of 100 per FIa.2. Wave-length in p. cent. a small experimental error would unduly displace the curve. The contrast in behaviour of potassium and hydrogea cliLoridss is, however unmistlakable. As in the case of the other solutes the curves for potassium chloride show apparent bands a t 1-12 p 1.55 p, and increasing absorption beyond 2.1 p. There is also evidence of an extra band a t 1 . 9 3 ~ . Discussion of Results. Owing to the influence of the solvent on the mode of vibration of the solute it is not possible a t this stage t'o att,ribute any of th STUDIES IN CATALYSIS. PART 111. 65 bands observed in the case’ of hydrochloric acid over the region investigated to hydrogen ion as distinct from the acid as a whole.What has been shown however is that hydrogen chloride possesses certain well-marked bands in the shortrwave region and further, in contrast with the behaviour of potassium chloride the extent of absorption increases with the concentration. So far as position of bands is concerned hydrogen chloride is capable of functioning as a catalyst in the sense of the radiation hypothesis already advanced although i t cannot be stated that the catalytic effect may be attributed t o one band only. Having drawn attention t o the contrast between the behaviour of hydrogen chloride and potassium chloride i t is at the same time1 important t o point out the striking similarity in their curves in respect of the position of the bands.These occur a t the same places (within the limit of experimental error) namely a t slightly longer wave-lengths than the bands of water itself. The solvent has produced forced vibration of similar type in both these solutes and in methyl acetate as well. Coblentz (Zoc. cit.) had previously investigated a few solutions in the infra-red down to 5p the solvent being carbon tetrachloride and the solutes being respectively diphenyl, naphthalene and azobenzene. Coblentz concludes that these sub- stances absorb infra-red radiation and further “ that the selective absorption of a solid in solution and that of the solvent are identi- cal.” Accosding t o the results obtained in aqueous solution it would appear that Coblentz’s conclusion is not accurately true in all cases and this is further emphasised by the fact that ‘ I solvent” and “ solute ” are ultimately interchangeable terms.From the present data and also from Coblentz’s curves it is appaxent that the intsnsity of a given band is the determining factor. Thus carbon tetrachloride has two bands a t 3 . 0 ~ and 4.5 p respectively of which the 3 p band is slightly the stronger. Diphenyl itself possesses one marked band at 3 . 2 5 ~ ~ whilst the solution of diphenyl exhibits two bands one a t 3 . 2 5 ~ and the other at 4 . 4 5 ~ approximately. It appears therefore that the strong diphenyl band (at 3 . 2 5 ~ ) has masked the solvent band which should have manifested itself a t 3.0p whilst the solvent band a t 4 .5 ~ persists in the solution (with slight displacement) since the solute alone has no1 band at this point. It is immaterial whether one of the substances is the solvent, except in so far as the concentration of the solvent is usually large, and intensifies its own bands. The results with naphthalene are the same the’ strong naphthalene band a t 3 . 2 5 ~ being the only one that appears in thatl region even when naphthalene is in the dis- solved state whilst a second band appears at- 4-45,~ due to the solvent. I n the case of water which is characterised by intense VOTA. CIX. 66 STUDIES IN CA'I'ALYSIS. PART 111. absorption i t may be inferred that solutions containing this solvent will exhibit the solvent bands (slightly displaced) and this too, when the absorption of the solvent has already been allowed for.This point of view brings Coblentz's observations into line with our own. We have finally to consider what mechanism underlies this masking effect produced by strong bands. The simplest view is to ascribe the phenomenon to general solvation the solvated (hydrated) compound produced possessing the most intense bands of both its constituents and exhibiting these a t slightly displaced positions the displacement being due t o the change in the electro- magnetic friction term denoted by " g " on the Lorentz notation. As a particular case therefore ,it would be inferred that when methyl acetate is dissolved in water a t least one additive compound is produced although not necessarily in large quantity. It is signi- ficant that the bands of the supposed additive compound occur just on the long-wave side of the bands exhibited by the solvent alone.On t-he basis of the electromagnetic theory of radiat'ion it is known that the radiation density is abnormally high for the wave-length region just below'a band. This means that the additive compound exists in a medium for which the radiation density is high just a t the wave-lengths which the complex can most readily absorb. On the radiation view of chemical reactivity put forward in the previous communications it' may be anticipated that this compound in the presence of excess of water is chemically activated and therefore tends to decompose into its two constituents ester and water. Although water is necessary for the production of the com- pound in a stoicheiometric sense its presence in excess is harmful in a cat'alytic-radiation sense.I n other words water a d s as a negative catalystb and the amount of compound formed will not be in strict accordance with the principle of mass-action. This con- clusion from the radiation hypothesis-which would appear to be a general one-bears very directly on the results obtained in the case of the hydrolysis of methyl acetate described in the succeed- ing paper (Part IV). Summary. (1) Employing a radiomicrometer of the H. C. Jones type the infra-red absorption spectra of hydrogen chloride potassium chloride and methyl acetate in aqueous solution have been examined over the range 1 p-2.1 p. (2) These solutes exhibit remarkably similar behaviour in respect of the position of the absorption bands which occur a t 1*12p, 1.55~1 and 2 .1 ~ (approx.) and thus correspond with the bands o STUDIES IN CATALYSIS. PAKT 1V. 67 water itself with a slight displacement in each case in the direction of longer wave-length. (3) It is suggested that the behaviour observed is clue to the formation of additive compounds between the solute and the solvent. (4) It is shown that hydrogen chloride possesses much greater absorptive power in the short infra-red region than does potassium chloride. This is in agreement with the radiation theory of homo- geneous catalysis according t o which a positive catalyst is one which increases the radiation density a t certain parts of the short infra-red region and it is well known that the existence of an absorption band increases the radiation density abnormally f o r wave-lengths below the band and diminishes i t f o r wave-lengbhs above the band.MUSPRATT LABORATORY OF PHYSICAL AND ELECTRO -CHEMISTRY, THE UNIVERSITY OF LIVERPOOL [Received November 12th 19153 STUDIES IN CATALYSIS. PART 111. 55 V.-Studies in Catalysis. Part 111. Preliminary Measurements of t h In fra-red Absorptson Spectra o j Hydrogen Chloride Potassium Chloride aqad Methyl Acetate in Aqueous Solution. By RAPHAEL HEBER CALI.OW WILLIAM CUDMORE MCCULLAGH LEWIS and GERALD XODDER. IN Parts I. and 11. of this series (Lamble and Lewis T. 1914 105, 2330; 1915 107 233) it has been shown on the assumption t h a t homogeneous catalysis (such as the effect of acid on the inversion of sucrose and on the hydrolysis of methyl acetate) is due t o the infra-red radiation emitted in quanta by the catalyst and absorbe 56 c z r m w LEWIS AND KODDER-: by the reacting substance that the radiation from the hydrogen ion, and therefore flie absorption bands possessed by it should lie in the short infra-red region.Thus from measurements of the tern- perature-coefficient of tlie rate of inversion of sucrose by hydrogen chloride it was calculat'ed that the characteristic frequency * of the hydrogen ion (neglecting any modification in numerical values which may be attributed to the effect of the solvent) should corre- spond with a wave-length of the order 1 . 2 3 ~ ~ and similarly from the rate of hydrolysis of methyl acetate it was inferred that the position of the hydrogen-ion band should be 1 .9 ~ . The mean of these two values is 1.56 p . From the results obtained by Coblentz (" Investigations of Infra-red Spectra," 7 parts Pub. Carnegie Inst. 1905-1908) i t is evident that radiation of this order of wave- length could be absorbed by sucrose and by methyl acetIab although the most marked bands of these substances occur a t longer wave- lengths. These observations however refer t o the substances in the pure state. N-o data are available as regards the position of the dissolved methfl acetate or hydrogen-ion bands since neither ester nor strong acid has been examined in aqueous solution. It is of obvious importance f o r thO radiation view of catalysis that such data should be obtained and f o r this purpose the present investigation was carried out.According to Coblentz (Zoc. cit.) the band a t 3 p in the water- transmission spectrum is due to the hydroxyl group and the heavier the group the longer is the wavelength of the absorption band associated with it for example the SO group gives rise to bands a t 4*55p 8 . 7 ~ ~ and 9 . 1 ~ . From this it would follow that if the hydrogen ion gave rise t o any bands in the infra-red region one or more of these should lie between 0 . 8 ~ and 3p provided the solvent itself did not' mask or fundamentally modify the vibration of the solute. As a matter of fact we have found that this modi- fication takes place. The bands in the water spectrum in this region occur a t 0,77 p l*O,u 1-25 p 1*50,u 2.0 p and 3.0 p of which those in heavy type are the most prominent.I n principle the present measurements consist in comparing the transmjssion of a layer of water with a layer of solution having the same thick- ness with respect to water and isolating the effect of the solute by * It is not implied that the hydrogen ion possesses only one band in the infra-red. From analogy to other substances it is probable that there are several bands and further it is impossible at the present stage to attribute the catalytic activity to a single type of vibration. The term " characteristic frequency '' might be called the " equivalent frequency," that is the single frequency which would give rise to a single quantum of sufficient size t o account quanta.titively for the effects observed.Whether this is due actually t o a single quantum or to more than one cannot a t present be stated STUDIES IN CATALYSIS. PART In. 57 taking the water transmission value as unity at a series of wave- lengths. E x P E R I MEN T AL. The infra-red spectrometer employed was of the IIilger pattern with a range 0-5-1Opu. The source of radiation was a Nernst filament lamp (0.6 ampere and 190 volts) regulation of current being effected by means of a variable resistance. The beam of radiation is rendered parallel and focussed by means of metallic mirrors dispersion being obtained by means of a. rock-salt prism, to which is attached a plane mirror t o give automatically minimum deviation for all wave-lengths after a single setting. The telescope of the instrument is fitted for a thermopile to be used in conjunc- tion with a galvanometer.Many attempts were made with different thermopiles used in conjunction with a Broca galvanometer to obtain readings but owing t o the lack of steadiness of the galvano- ineter zero due to variable magnetic effects the thermopile-galvano- meter system had to be :ibandoned and in its place a radio- tnicromet'er was employed which was found t o give consistent read- ings. This instrument which was constructed in the laboratory, was of the type reconimended by H. C. Jones (2. -4mer. Chent. SOC. 1915 37 776). The suspended system consisted of a loop of pure silver wire supported on a glass frame and connected t o a thermo-couple of the same kind as that employed by Jones.The Compensating junction was found to nullify to a very large extent the disturbing influence of local variations in temperature due to the warming of the slit on exposure1 t o the radiation. The radia- tion ent'ered through a fluorite window and was received on a thin plate of pure silver blackened and attached to the couple. The deflexioii o f the suspended system was observed on an illuminated scale by means of a reading telescope. I n all cases it was found most advant.ageous t o use the instrument ballistically that is t o take the reading of the deflexion produced and completed in seven or eight seconds after raising the water-cooled shutter placed in front of the Nernst la-mp. The water and hydrogen chIoride solutions were examined in quartz cells two such cells being mounted on a horizontal carrier (placed directly in front of the collimator slit) so that the cells could be rapidly inkrchanged f o r comparative measurements.By means of stops i t was possible t o bring the cells to exactly the same position throughout a series of measurements in order that the amount of absorption due t o the quartz walls would reniain con- stant. The two cells most. frequently employed were made by cenleiitiiig quartz plates (each 1.5 mm. thick) on to the sides of 58 CALLOW LEWIS AND NODDER thin U-shaped piece of glass with marine glue. The internal thick- ness of each cell was approximately 0.9 mm. Sensitivity of tJze Rndiomicrometer.-Sensitivity is defined as the deflexion in cm. per sq. mm. of exposed surface for a standard candle and scale each a t a distance of 1 metre.The candle and scale were set up a t this distance from the instrument and the deflexion observed. Seven concordant readings gave as a mean value 6.0 cm. The width of the slit was 0.5 mm. and the area of exposed vane 1.5 mmq2. Hence the1 sensitivity is 4.0. I n t h i s measurement tke instrument was exhausted t o a pressure of 2-3 mm. Jones (loc. c i t . ) obtained a sensitivity of 5 with his instrument under the ordinary pressure. Our instrument in its present form is therefore less sensitive but has the advantage of possessing a very short period. CJbange in Sensitivity with Time.-Before taking any set of read- ings the deflexion with the spectrometer drum set a t 2 p was in all cases noted the width of the slit of the radiomicrometer being permanently adjusted to 0.5 mm.and an empty quartz cell being placed in Dosition in front of the collimator. The1 deflexion remained the s3,me for several days the maximum variation f o r widely separateld intervals of time being from 161 to 155 mm. divisions c n the scale. All values were reduced where necessary to the 161 mm. deflexion as standard. Drum Calzbration.-Tkle position of the sodium-yellow and potassium-red lines indicated that the1 drum was correct f o r wave- lengths shorter than 0 . 8 ~ approximately. It was also found that the quartz band occurred a t 2.9p as Coblentz had observed and that the water bands a t 3 . 0 ~ and 4 . 7 ~ were correctly registered. Instead of oKtaining the water band a t 1*5p however the drum reading was 1 0 4 5 ~ .This relatively small error in calibration has been allowed i o r in the data quoted later. Percentage Error in the Deflexion Readings.-The region in- vestigated ext'ended approximately from 0 . 9 ~ to 2.1 p. A t each point a t least six independent readings were taken and the experi- mental error thus obtained. For deflexions greater than 200 mm. the error is negligible. F o r smaller deflexions the average peroent? age error is as follows: Deflexion in mm. ...... 150 100 50 25 10 5 Per cent. error ......... 5 1 . 0 j 1 . 2 5 k 2 - 0 f4.0 &6 &lO Comparison of t?be Transmission of the Quartz Cells. Over the range investigat,ed no measurable difference could be detected. When the cells were filled with water differences were observed, Readings were taken through each cell when empty STUDIES IN CATALYSIS.PART JII. 59 due to the difference in thickness of the water layers. A table of values was constructed from which observations made on one cell could be seduced t o that of the other. This correction will be referred to as the “ correction for the dimensions” of the cell. Infra-red Absorption of Hydrogen Chloride in Aqueous Solution. The acid was investigated a t three concentrations namely N , 2LT and 3 N . The solution was placed in one of the cells ( A ) , water in the other ( B ) arid readings taken alternately through each. I n order that the deflexions may be comparable i t is neces- sary to correct f o r the displacement of water by the acid in the SCiL tions.This was effected by determining the absorption-coeffi- cieqt k of water. One of the cells was made considerably thinner tl*aii the other the difference being 0.55 mm. The law of absorp- tim is expressed thus : I = Ioe - kd, where Z is the intensity of the incident radiation Z the intensity of the emergent beam and d the thickness of the water-layer traversed. Knowing the thickness of the two cells exposed to the same intensity I, it is possible t o calculate k f o r water a t various wavelengths. I n this way a table of values was prepared the unit of length being the millimetre. Returning to the comparison of cells A and B when the solution in A was compared with the water in B i t was possible from a knowledge of the density of the solution to determine the effective thickness of water actually traversed; knowing k it was then possible t o alter the values of the deflexion observed through cell B t o the values which would have been given if the beam had traversed in B the same thickness of water as exists in cell A .This was the general method of correction adopted and will be referred t o as the “correction for density.” A single illustration of the method may be given in the case of AT-hydrochloric acid. The density of this solution is 1.017. The concentrat,ion of hydrogen chloride is 0.0365 gram per c.c., and hence the concentration of water (in cell A containing the solute) is 0.9805 gram per C.C. The concentration of water in cell 11 is taken as unity. Hence, Effective thickness of water in A - 0 9805 Thickness of water in B 1 ’ --__-- - ____ The actual thickness of the water-layer in R is 0.92 1nm.Hence the effective thickness of water in A is 0.90 mm. The intensity Z, of the emergent beam from cell B thus requires to be increased to the value i t would have (1’) if the cell were 0.90 mm. thick. It is evident that I’=Zek(0*9~--0*9(’) =ZeO*ozk. Witll the aid of the know 60 CALLOW LEWIS AND NODDER: Deflexion Deflexion through through value of I; one can thus calculate I' from the observed value I. On dividing the defle'xion through the acid solution by that through the water (corrected) a t the same wave-length the absorption due to the acid alone can be determined. The following table (I) con- tains in full the results obtained and the extent of the corrections in the case of ,!!-hydrochloric acid.The remaining tables do not contain the corrections the final corrected results only being given. Deflexion Deflexion through through TABLE I. 1-55 96.0 114.4 0.84 1.65 85-4 99.1 0.86 1.74 64.5 79.3 0.81 1.84 43.2 57.0 0.76 1-93 32.2 45-8 0.70 2.03 23.3 38.8 0.60 2-13 21.0 32.3 0.65 N-Hydrochloric 4 cid. 1-55 79.5 116.8 0.68 1.65 68.1 103.0 0.66 1.74 41.4 73.7 0.56 1.84 25.2 60.0 0.42 1-93 16.0 40.6 0-40 2.03 11.0 37.4 0-30 2.13 10.0 31.9 0-31 Waye-leng th in p. 1.0 1.12 1.23 1-34 1.45 1.50 1.55 1.65 1.74 1.84 1.93 2-03 2-13 Deflexion through acid solution (cell A ) . 40.0 mm. 67.0 82-0 66.0 54.5 99.0 109.0 96.8 66.6 42.0 34.9 30-3 23.0 Deflexion observed through water (cell B).39.3 mm. 63-5 78.8 62.0 78.5 95.5 105.5 95-5 68-0 45.1 34.6 29.9 23.6 Deflexion through B corrected for dimensions. 38.5 mm. 65.4 80.4 65.1 84.8 104.1 117.1 102.2 75.5 50.1 40.1 35.9 30.0 Deflexioll Deflesion through B inA corrected corrected for deflexion density. in B. 38.6mm. 1.04 65-6 1.02 80-8 1.01 65-7 1.00 85.8 0.98 106.4 0.93 119.6 0.91 104.0 0.93 76.9 0.86 51.1 0.82 41-3 0.86 37.2 0.81 31.4 0.73 The last colunin gives the fractional absorption exertled by the solutel. TABLE 11. 2N-Hydr o c hloric A c id. TABLE 111. 3N-Hydrochloric A cid STUDIES IN CXTALTSIS. PART 111. 61 considerable absorption in the short infra-red region.As far as this goes it supports qualitatively the conclusion already drawn as regards the effect of the acid in increasing the radiation density in this region. A t the shorter wave end of the region examined the solution is actually more transparent than water. This pheno- menon has been observed by others and must be attributed to a change in the nature and therefore in the absorbing power of a certain fraction of ths water molecules brought about by the presence of the dissolved snbstance. This may mean either that the FIG. 1. Tlrave-length in p, solute has become hydrated the water of hydration being physically different from water in the uncombined state or t'he presence of the solute may have altered the position of equilibrium obtaining between the various polymerised forms of water and have thus altered the absorption power.The most important conclusion, however which can be drawn from the data obtained has t o do with the positions and relative intensities of the bands of the acid. There appear t o be a t least three bands in this region namely a weak one a t 1.12 p a stronger band a t 1.55 p and one still stronge 62 CALLOW LEWIS AND NODDER : in the neighbourhood of 2.1 p.* Thel resemblance of these bands in respect of position and intensity to the bands exhibited by water itself namely 1*Op 1.5 p and 2*0p is so remarkable that it is fairly evident that the solvent has exerted a very marked effect on the type of vibration possessed by the solute. I n view of this apparent effect it is necessary to examine the behaviour of other substances dissolved in water.F o r the sake of comparison methyl acetate was therefore investigated in aqueous solution as being a substance totally unlike hydrogen chloride in respect of chemical constitution and yetl catalysable by the acid a d further potassium chloride was examined as being similar to hydrogen chloride yet lacking its catalysing power. Infra-red Absorption of Methyl Acetate in Aqueous Solution. The concentration of methyl acetate employed was 10 C.C. in 50 C.C. of solution or approximately 2.6N. This concentration was chosen as i t corresponded with one of the cases investigated by Mr. Griffith and described in the succeeding paper (Part IV). The results are given in the following table: Wave-length in p.1.0 1.12 1.23 1.34 1-45 1-50 1.55 1-65 1.74 1.84 1.93 2-03 2-13 TABLE IV. 2*6N-iliethyZ Acetate. Deflexion through Deflexion through ester solution water (corr.1 Deflexion through A . (cell A). (cell B). DeAexion through B. 43.2 mm. 42.1 mm. 1.03 71.3 73.2 0.96 97.3 96.6 1.00 82.5 79.3 1.04 96.6 92.1 1.05 119.0 127.1 0.93 133.5 144.4 0.92 132.8 137.9 0.96 101.7 105-2 0.97 68.4 71.4 0.95 52.3 61.0 0.86 40.8 53.2 0.77 34.6 61.6 0.67 These data are shown by one of the curves in Fig. 2. It will be observed t$at the ester exerts considerably less absorption than hydrogen chloride a t the same concentration. Again however, bands appear a t 1.12 p 1.55 p and another beyond 2.1 p.Coblelntz (Zoc. c d . ) has examined the pure subst,ance from the wave-length 1.7 p approximately a t which point the fractional absorption was * Gaseous hydrogen chloride has according to Angstrom and Palmer an emission band at 3.98 p S'L'UDIES IN CATALYSIS. PART 111. 63 Wave- through water in length solution (corrected) DTflexion in p. (cell A ) . (cell B). in B. 1.0 38.0 mm. 34-9 mm. 1.09 1-12 62.0 61.0 1.01 1.23 75.0 72.7 1.03 1.34 68.0 62-8 1.08 1-45 99.0 97.2 1.02 1.50 121.0 114.8 1.06 1.55 124-0 126.9 0.98 1.65 108.0 104.5 1.03 1.74 74.0 74.7 0.99 1.84 56-0 53.3 1.05 1-93 46.6 45.8 1-01 2.03 41.0 39.8 1.03 2.13 32.0 35.9 0.89 , 0.81 a feeble band manifesting itself a t 2.2-2.4 p and a vexy marked band a t 3 . 3 ~ . ' through water in A solution (corrected) Deflexion (cell A ) .(cell B). in B. 39 mm. 37.0mm. 1.05 64 63.0 1.01 81 80.0 1.01 67 64.0 1-05 92 85-5 1-07 111 113.0 0.98 z 22 127-0 0.96 109 108.0 1.00 77 79.5 0.97 54 54.0 1.00 43 47.4 0.91 37 42-0 0.88 30 36.5 0.82 Infra-red Absorption of Yotussizim Chloride in Aqueous Solution. The so'lutions examined werel respectively N Z J and 3lv. The results are given in the following tables : TABLE VII. 3 N-1'0 t tzssiti rn Chloride. Wave-length in ,u. 1.0 1.12 1.23 1.34 1.45 1-50 1.55 1-65 1.74 1-84 1-93 2.03 2.13 Deflexion through solution (cell A ) . 37 mni. 61 74 75 110 131 138 121 85 G I 51 45 35 Deflexion through water (corrected) (cell B). 34-3 mm.61.8 73.4 63.0 91.8 109.0 124.0 99.5 7 1.0 51.0 44.5 39.6 33.0 Deflexion in A Deflexion in B 1.08 0.99 1.08 1.20 1-20 1.20 1.11 1.20 1.20 1.20 1.15 1.15 1-08 These data are given in Fig. 2. It is evident that potassium chloride exerts very little absorption compaxed with hydrogen chloride'. It can theref ore only slightly affect the radiation density of the system and its transparency would even lead one t o expect that its behaviour should be the reverse of hydrogen chloride. Sinc 64 CALLOW LEWIS AND NODDER: the latter is a strong catalyst one would anticipate therefore on the radiation view that potassium chloride would exert no positive catalytic effect and might even be a feeble negative catalyst. So far as absence of effect is concerned a t the ordinary temperature experiment bears out this conclusion.It will be observed that the 3N-solution of potassium chloride is more transparent than those more dilute but that there is a reversal of the relative positions of the N - and 2N-solutions. Whether this has any considerable significance cannot be determined a t present for in those cases in which the transmission of the solution lies in the region of 100 per FIa. 2. Wave-length in p. cent. a small experimental error would unduly displace the curve. The contrast in behaviour of potassium and hydrogea cliLoridss is, however unmistlakable. As in the case of the other solutes the curves for potassium chloride show apparent bands a t 1-12 p 1.55 p, and increasing absorption beyond 2.1 p.There is also evidence of an extra band a t 1 . 9 3 ~ . Discussion of Results. Owing to the influence of the solvent on the mode of vibration of the solute it is not possible a t this stage t'o att,ribute any of th STUDIES IN CATALYSIS. PART 111. 65 bands observed in the case’ of hydrochloric acid over the region investigated to hydrogen ion as distinct from the acid as a whole. What has been shown however is that hydrogen chloride possesses certain well-marked bands in the shortrwave region and further, in contrast with the behaviour of potassium chloride the extent of absorption increases with the concentration. So far as position of bands is concerned hydrogen chloride is capable of functioning as a catalyst in the sense of the radiation hypothesis already advanced although i t cannot be stated that the catalytic effect may be attributed t o one band only.Having drawn attention t o the contrast between the behaviour of hydrogen chloride and potassium chloride i t is at the same time1 important t o point out the striking similarity in their curves in respect of the position of the bands. These occur a t the same places (within the limit of experimental error) namely a t slightly longer wave-lengths than the bands of water itself. The solvent has produced forced vibration of similar type in both these solutes and in methyl acetate as well. Coblentz (Zoc. cit.) had previously investigated a few solutions in the infra-red down to 5p the solvent being carbon tetrachloride and the solutes being respectively diphenyl, naphthalene and azobenzene.Coblentz concludes that these sub- stances absorb infra-red radiation and further “ that the selective absorption of a solid in solution and that of the solvent are identi- cal.” Accosding t o the results obtained in aqueous solution it would appear that Coblentz’s conclusion is not accurately true in all cases and this is further emphasised by the fact that ‘ I solvent” and “ solute ” are ultimately interchangeable terms. From the present data and also from Coblentz’s curves it is appaxent that the intsnsity of a given band is the determining factor. Thus carbon tetrachloride has two bands a t 3 . 0 ~ and 4.5 p respectively of which the 3 p band is slightly the stronger. Diphenyl itself possesses one marked band at 3 .2 5 ~ ~ whilst the solution of diphenyl exhibits two bands one a t 3 . 2 5 ~ and the other at 4 . 4 5 ~ approximately. It appears therefore that the strong diphenyl band (at 3 . 2 5 ~ ) has masked the solvent band which should have manifested itself a t 3.0p whilst the solvent band a t 4 . 5 ~ persists in the solution (with slight displacement) since the solute alone has no1 band at this point. It is immaterial whether one of the substances is the solvent, except in so far as the concentration of the solvent is usually large, and intensifies its own bands. The results with naphthalene are the same the’ strong naphthalene band a t 3 . 2 5 ~ being the only one that appears in thatl region even when naphthalene is in the dis- solved state whilst a second band appears at- 4-45,~ due to the solvent.I n the case of water which is characterised by intense VOTA. CIX. 66 STUDIES IN CA'I'ALYSIS. PART 111. absorption i t may be inferred that solutions containing this solvent will exhibit the solvent bands (slightly displaced) and this too, when the absorption of the solvent has already been allowed for. This point of view brings Coblentz's observations into line with our own. We have finally to consider what mechanism underlies this masking effect produced by strong bands. The simplest view is to ascribe the phenomenon to general solvation the solvated (hydrated) compound produced possessing the most intense bands of both its constituents and exhibiting these a t slightly displaced positions the displacement being due t o the change in the electro- magnetic friction term denoted by " g " on the Lorentz notation.As a particular case therefore ,it would be inferred that when methyl acetate is dissolved in water a t least one additive compound is produced although not necessarily in large quantity. It is signi- ficant that the bands of the supposed additive compound occur just on the long-wave side of the bands exhibited by the solvent alone. On t-he basis of the electromagnetic theory of radiat'ion it is known that the radiation density is abnormally high for the wave-length region just below'a band. This means that the additive compound exists in a medium for which the radiation density is high just a t the wave-lengths which the complex can most readily absorb.On the radiation view of chemical reactivity put forward in the previous communications it' may be anticipated that this compound in the presence of excess of water is chemically activated and therefore tends to decompose into its two constituents ester and water. Although water is necessary for the production of the com- pound in a stoicheiometric sense its presence in excess is harmful in a cat'alytic-radiation sense. I n other words water a d s as a negative catalystb and the amount of compound formed will not be in strict accordance with the principle of mass-action. This con- clusion from the radiation hypothesis-which would appear to be a general one-bears very directly on the results obtained in the case of the hydrolysis of methyl acetate described in the succeed- ing paper (Part IV). Summary. (1) Employing a radiomicrometer of the H. C. Jones type the infra-red absorption spectra of hydrogen chloride potassium chloride and methyl acetate in aqueous solution have been examined over the range 1 p-2.1 p. (2) These solutes exhibit remarkably similar behaviour in respect of the position of the absorption bands which occur a t 1*12p, 1.55~1 and 2 . 1 ~ (approx.) and thus correspond with the bands o STUDIES IN CATALYSIS. PAKT 1V. 67 water itself with a slight displacement in each case in the direction of longer wave-length. (3) It is suggested that the behaviour observed is clue to the formation of additive compounds between the solute and the solvent. (4) It is shown that hydrogen chloride possesses much greater absorptive power in the short infra-red region than does potassium chloride. This is in agreement with the radiation theory of homo- geneous catalysis according t o which a positive catalyst is one which increases the radiation density a t certain parts of the short infra-red region and it is well known that the existence of an absorption band increases the radiation density abnormally f o r wave-lengths below the band and diminishes i t f o r wave-lengbhs above the band. MUSPRATT LABORATORY OF PHYSICAL AND ELECTRO -CHEMISTRY, THE UNIVERSITY OF LIVERPOOL [Received November 12th 19153

ISSN:0368-1645    

DOI:10.1039/CT9160900055

出版商:RSC    

年代:1916

数据来源: RSC

摘要:

STUDIES IN CATALYSIS. PAKT 1V. 67 VT.-Studies in Catalysis. Pui-t I V. Xtoicheiomctric and Catalytic Efects Due to the Progressive Displacement of oyie Reactant by Another in the ‘‘ Acid” Hydrolysis o f Methyl Acetate. By ROBERT OWEN GRIFFITH and WILLIAM CUDMORE MCCULLAGH LEWIS. IT is a remarkable fact although it does not appear t o have attracted as much attention as i t deserves that the velocity- constant k for the reaction CH3*C0,*CH3 + H,O + CH3*OH + CH,=CO,H, as obtained from the equation k = Llog - varies with the initial concentration of the methyl acetate. Ostwald who first investi- gated the reaction obtained the following results ( J . p. Chem., 1883 [ii] 28 449). At the temperature of 25*2O with the same concentration of hydrochloric acid in each case (23/3) the con- centrations of ester employed in the experiments were 2 c.c.1 c.c., 0.5 c.c. and 0-3 C.C. respectively in 15 C.C. of solution. C.C. of methyl acetate ,t=24*19. >> 9 ) 7 , k=21.96. >7 0.5 ,> 97 , li=20*71. 7 0.3 ,> 7 , k=19-91. a t a-x With 68 GRIFFITH AND LEWIS: Ostwald attributes this decrease in the velocity-constants to the fact that the end condition (in each case attained after two days) varies with the initial Concentration in such a way that in the solution intially containing least ester the decomposition progresses to the furthest extent and thus the velocity-constants differ because of the existence of an equilibrium point. He illustrates this in the following way. If we take in each case the final titra- tion value in C.C.of alkali and subtract the value due to the hydro- chloric acid we obtain the value due t o the acetic acid formed by the hydrolysis of the ester. Dividing these figures by the initial amount of ester in C.C. in each experiment' one would expect the same value if the total amount or some fraction of the' ester were decomposed in each case. The table given below shows however, that this is not so. TABLE I. C.C. of barium hydroxide solution required for 1 C.C. of ester. End point in C.C. of barium hydroxide solution. Initial ester in C.C. 2.0 25.96 12.98 1.0 14-34 14.34 0.5 7.66 15.32 0.3 4.66 15-53 Ostwald states that the magnitudes given in the last column of table I. and the corresponding velocity-constants are inversely proportional.A fairly constant result is obtained by multiplying the figures in the last column by the corresponding velocity- constants. From this he deduces the empirical relation which should ho'ld for all different initial concentrations of ester namely, a log - = k? t. a-x p in which p is the initial concentration of ester and a x and t have their usual significance. This relation holds independently of the nature of the acid. In criticising this result the first point to be noted is that the constants are not truly comparable as they stand for the concen- tration of the water differs in each experiment' and the reaction- velocity should be proportional to the concentration of the water. I n the experiments with the largest amounts of methyl acetate there is less water than in the; others.This obvious correction, which was applied by Rosanoff t o the case of the inversion of sucrose (J. Amer. Chem. SOC. 1913 35 248> has the effect of making the constants diverge still more from each other. This correction which we will call the " stoicheiometric " correction for water can be very approximately applied tol Ostwald's results b STUDIES IN CATALYSIS. PART IV. 69 assuming that the methyl acet'ate displaces its own volume of water from the solution. I n this way we obtain the following. TABLE 11. Constants corrected for variat,ion in water content. Ostwald's unimolecular consCants. 24-19 1.888 21-96 1-590 20.71 1.447 19.99 1.378 The corrected constants are obtained by dividing those of Ostwald by the weight of water present in 15 c.c.the total volume of solution in Ostwald's experiments. As the figures are purely relative the above method serves f o r comparison. From the tables given above it is seen that2 the divergence is now greater than before so that Ostwald's empirical relation loses much of its significance. I n order therefore t o attempt to determine satis- factorily whether the change in the velocity-constant is duel t o the variation in the end point that is the equilibrium point with different initial concentrations of the ester the present research was undertaken. The A cc.zcmfe ?7clocity Equation. The reaction is obviously bimolecular and reversible although in dilute aqueous solution a unimolecular constant is necessarily obtained.Denoting the initial concentration of the methyl acetate in mols. per litre by b the number of mols. decomposed a t time t by x and the concentration of the water in mols. per litre by w thevelocity- constant for the decomposition of the ester by k and the velocity- constant' for the formation of ester by k, we have the differential equation dx - k,(b - X)(PU - z) - k&X? at - Writing k! = K the equilibrium constant where k , [ester] x [wat,er] [alcohol] x [acetic acid]' K= __ ___.__~- ~ - _ _ _ and also setting y=iu+ b and 3 = J(w+ b V + 4(R- l)wb integra- tion gives reme'mbering that when t=O x=O, k = Kk 70 GRIFFITH AND LEWIS As is t o be expected neglect of the change in the concentration of water during an experiment makes very little difference in the value of k obtained.Assuming that 2 ~ is a constant we have @ = k p ( b - x) - k,2? d t This on integration gives a value f o r h. which is very near t.0 the value obtained as above (generally within 1 per cent.) altliough the variatJon in the concentration of water from beginning to end of a reaction is in the case of the most concentrated solution of ester employed more than 4 per cent. of t,he initial concentration of water. The more complete equation given above has however, been employed throughout. E X P E R I M E N T A L . Methyl acetate was redistilled and the portion boiling between 55'4O and 56'0° was used; it contained very little free acetic acid, the specific conductivity of an approximately normal solution of the ester in water being 40 x 10-6 mho.The concentration of the catalyst hydrochloric acid was N / 2 in the first set of measure- ments N in the second and N / l O in the third. The concentra- tions of ester were respectively about 8 5 3 1 and 0.5 C.C. in 50 C.C. of total solution. The exact concentrations in mols. per litre are given in the table of results. For the measurements of velocity the usual method due to Ostwald was employed. The reactions were carried out in Jena-glass flasks which had been thoroughly steamed out and fitted with well-paraffined corks the flasks being kept in a thermostat maintained a t a constant tempera- ture. From time t o time a convenient volume of the reaction mixture was withdrawn by a pipette run into 40 C.C. of conductivity water and titrated against standard baryta.For each experi- ment two dsterminations were made and the means of the two titres a t corresponding times were taken t o calculate the velocity- constants. The end points were taken aft.er allowing the reaction t o proceed for a time sufficient for equilibrium t o be attained. Before starting each experiment the required amount of stock acid solution was run into the flask together with that amount of water (previously determined) which with the ester gave a total volume of 50 C.C. This was then placed in the thermostat for thirty minutes and the required amount of ester added at a noted time. Thus for each experiment the concentration of each con- stituent was known STUDIES IN CATALYSIS. PART IV. 71 Determination of t h e Equilibrium Constant of t h e Reaction.To determine K the equilibrium-constant a knowledge is r e quired of the initial conceiitration of ester the final concentration of the acetic acid and that of the water. The final concentration of the acetic acid which is identical with that of the methyl alcohol is arrived a t from the end-point titra obtained after ten times the time of half decomposition of the ester has elapsed by which time the system is not appreciably removed from its con- dition of equilibrium. The initial concentration of the ester was not determined by weighing into a certain volume but by treating the exact amount taken with an excess of baryta solution and then titrating with hydrochloric acid. The ester was treated with an excess of standard baryta solution f o r eight hours precautions being taken to prevent entry of carbon dioxide from the air and then the excess of baryta was determined.The difference between the initial concentration of ester and the amount decomposed after the solution had reached its position of equilibrium gives the amount of ester present a t equilibrium those of the methyl alcohol and acetic acid being of course equal t o that of the decomposed methyl acetate. The equilibrium-constant. was determined only in the case of the two more concentrated solutions of methyl acetate as experi- mental errors are magnified largely in calculating K for more dilute solutions owing to the very large excess of one of the reactants-water. There is no reason however to believe that there is any variation in the equilibrium-constant with change in the concentration of the ester.The following values were obtained for the equilibrium-constant ester x water methyl alcohol x acetic acid' I<= ~ - _ _ _ _ (1) In the preseiice of N/2-hydrochloric acid 4.30 4.52 4.66, 4.80; (2) in the presence of N / 10-hydrochloric acid 4-52 ; and (3) in the presence of N-hydrochloric acid 4.34 4.83. These are constant within experimental error and the mean value 4-6 was adopted. It may be noted however that the absolute value of K is not of great importance for the object of this research as a fairly large variation in K makes but a small change in the value of the velocity-constant k (constant for ester decomposition) although, of course from formula (1) the value of k is directly proportional to K .The equilibrium-constant K has been determined for the pure substances with no added catalyst by Berthelot and Pean d 72 GRIFFITH AND LEWIS: St. Gilles (Ann. cJhim. PJL~s. 1863 [iii] 68 225) who found that, starting with equivalent amounts of acetic acid and methyl alcohol, 67.5 per cent. of each combined. This gives K=4.31. Further, Menschutkin (Annalen 1879 195 334) obtained 69.5 per cent. of combination which gives K = 5.18. Now Lap- worth and Jones (T, 1911 99 1427) have shown that for ethyl acetate the presence of concentrated hydrochloric acid affects the value of K which increases with decreasing values of the ratio H,O /HC1. However they found that provided the concentration of the hydrochloric acid is below normal the equilibrium-constant is very nearly independent of the hydrochloric acid present.Assuming that in this respect methyl acetate acts similarly t o ethyl acetate it is likely that no very sensible change in the value of K is t o be expected in these experiments. The equilibrium- constant for this reaction in the presence of hydrochloric acid has also been ,determined by Worley (PYOC. Roy. SOC. 1912 [ A ] 87, SSZ) who deduces by extrapolation the value K=6.6 for the constant< with no added catalyst' present a value decidedly higher than those obtained above. The value 4.6 obtained above lies between the two. Velocity Mecrsiirements. Below are given the experimental results. The following is the notation used: t=time in minutes from the start.x=number of gram-molecules of ester decomposed a t time t . 1 CL Ji = unimolecular constant - log -. t a-x k = velocity-constant for ester ,decomposition \ k = velocity-constant for ester formation from equation (1) Tables 111. and IV. serve as an indication of the general character of the velocity-constants obtained in these experiments, whilst tables V. t o VII. summarise the results of the experiments with hydrochloric acid as catalyst STUDIES IN CATALYSIS. PART IV. 73 TABLE 111. Hydrolysis of methyl acetate with 0*5lV-hydrochloric acid a t Initial concentration of methyl acetate = 0.6857 mols. per litre. 24-3O k 0.03. 9 ,9 , water =51.74 , 9 9 9 t. 60 90 120 160 280 340 460 m X. 0.1192 0- 1747 0.2183 0.2746 0.4044 0.4513 0.5141 0.6482 Mean .........k. 0.00338 0.00348 0.00343 0.00344 0.00349 0.00351 0.00343 - 0.003465 k,. - kl. 0.0000616 0.0000633 - 0.0000620 Mean 0.0000622 0.000284 0.0000628 - 0.0000629 - 0.000061 3 - - - - 0.0000623 0-000284 TABLE IV. Hydrolysis of methyl acetate with 0-13-hydrochloric acid a t Initial concentration of ester = 0.7013N. 25*00° f 0.03. 9 9 9 , water = 52.19.T. t. 200 2 80 445 620 1515 1705 1g30 2. 0.08455 0.1171 0.1727 0.2311 0.4299 0.4588 0.4877 0.6624 k. 0.000683 0.000695 0.000680 0.000692 0*00069 1 0.000692 0.000698 - Mean . . . .. 0.000690 4. - k - 0.00001 237 0*00001255 - 0.0000 1225 Mean 0.00001247 0.0000565 0.0000 1235 - 0.00001236 - 0.00001254 - - - 0*00001236 0.0000565 TABLE V.Catalyst N-Hydrochloric acid a t 25*0°. Initial con- centration No. of C.C. in mols. of ester of est,er in 50 C.C. per litre. of solution. 1-741 8 approx. 1.151 5 ,, 0.7093 3 ,, Mean con- centration of water in mols. per litre=w*. k. klw. kl. k*. 45.38 0.00903 0.0001991 0.0001652 0.000755 48.76 0.00818 0.0001679 0-0001482 0-000677 50.94 0-00764 0.0001467 0.0001392 0.000636 * The mean concentration of water is the concentration midway through an experiment that is when half the methyl acetate has been decomposed. D 74 GRIFFITH AND LEWIS: TABLE VI. Catalyst 0.5A -Hydrochloric acid a t 24*3O.* Initial con- centration No. of C.C. in mols. of ester of ester in 50 C.C.per litre. of solution. 1.744 8 approx. 1.155 5 ,, 0.6857 3 ,) 0-2420 1 ,, 0.1270 0-5 ,, Mean con- centration of water in mols. per litre=w.t k. k/w. k - k2. 46.10 0.00382 0.0000829 0.0000687 0.000313 49.14 0.00363 0.0000739 0.0000648 0.000295 5 1.42 0.00346 0.0000674 0*0000623 0.000284 53.73 0.00310 0*0000578 0.0000569 0.000256 54.15 0.00300 0.0000553 0.0000551 0.000251 * The thermometer used in this series of measurements registered 25.0", t The mean concentration of water is the Concentration midway through but on calibration was found to be in error. an experiment that is when half the methyl acetate has been decomposed. TABLE VII. Catalyst O*lN-Hydrochloric acid a t 25*0°. Initial concen- tration in mols. of est.er per litre.1.156 0.7013 0.2425 0.1304 Mean concen- No. of C.C. tration of ester of water in 50 C.C. 5 approx. 49.69 0.000726 0.00001461 0.00001302 0.0000595 3 ? 9 51.86 0.000690 0*00001330 0.00001236 0*0000565 1 Y Y 54.25 0.000660 0*00001217 0*00001177 0.0000538 0.5 1 j 54-86 0.000606 0*00001104 0*00001093 0.0000499 in mols. per of solution. litre= w. k. kjw. El. Ic,. Throughout each single experiment the values of k k, and k , are constant within the experimental error that is there is no trend observable even in the most concentrated solutions of methyl acetate. From the summarising tables given above it will be seen that the values of the unimolecular constants k fall steadily with decreasing concentration of ester as Ostwald has already shown ; the values of k / w fall more rapidly still.Also the values of k,-the true velocity-constant-still show the same decided fall although not nearly so marked as that of k / w . The effect of the correction term k2x2 in the differential equation, 5!!? = k,(zo - z)( b - z) - k2z2, d t may be seen in the above tables by comparison of corresponding values of k / w and k,. It is seen that these two values in the case of the most dilute solution of methyl acet,ate employed (0.1N) are identical within experimental error but as the concentration of ester is increased k / w increases inore rapidly than k,. Assuming the law of mass-action to be valid throughout the constants given in the column headed k should be of equal value STUDIES IN CATALYSTS PART IV.75 as they include the correction for the reverse reaction the forma- tion of ester. Ostwald's statement therefore that the variation in the velocity-constants with change in the concentration of ester is due to there being an equilibrium point is insufficient and we must therefore consider other explanations to account for this variation. A first suggestion is that the concentration of the catalyst is not the same in each experiment; although the total concentration of hydrochloric acid is the same in each series the degree of dissociation of the acid might change with the amount of ester present. Presumably the addition of the ester would decrease the dissociation of the acid thus lowering the hydrogen ion concentration; a t first sight i t appears if this view be correct, that the velocity-constants should decrease as the concentration of the methyl acetate is increased.However from the recent work of Snethlage (Zeitsch. physikd. C'henz. 1913 85 253) Acree (Amel.. C'hem J. 1912 48 352) and others the results of measurements of velocity of reaction can be interpreted so as t o make it appear possible that the undissociated molecule of hydrochloric acid has a catalytic activity about twice as great as that of the hydrogen ion. On $he assumption then, that the presence of methyl acetate decreases the percentage ionisation of the catalysing acid hydrochloric it is t o be expected that the velocity-constants should be greater the more ester is present. To test this therefore i t is necessary t o know the dis- sociation of the acid in each solution used and for this purpose measurements of the conductivities of the acid in mixtures of water and methyl acetate were carried out.Determination of the Degree of Dissociation of Hydrochloric Acid in Mixtures of Water and Methyl Ace!ate. The electrical conductivities of several of the solutions used in the experiments on the velocity of the reaction were determined, and also those of solutions of N/2000-hydrochloric acid in the presence of varying amounts of methyl acetate. Assuming com- plete dissociation of the hydrochloric acid at the latter dilution, the degrees of dissociation for the other concentrations were calcu- lated in the usual way. The ordinary Kohlrausch method and apparatus was used the bridge being calibrated by Strouhal and Barus' method whilst the cell constant was determined by the use of iV/50-potassium chloride which has a specific conductivity of 0.002728 a t 24.3O and of 0*002768 a t 25O.For the more dilute acid solutions the cell employed was of the Arrhenius type whilst for the more concentrated solutions the conductivity vessel was a U-tube with closely fitting ebonite caps which held the electrodes D* 76 GIUFFITH AND LEWIS : firmly in position. Tables VIII. and IX. give the results of the measurements ; interpolated values in these tables are bracketed. The electrical conductivities of the solutions were determined immediately after making up and allowing them to attain the temperature of the thermostat' but nO very great change was observed in the conductivities of the solution with time the acetic acid gradually formed being very little dissociated in the presence of concentrated hydrochloric acid whilst the hydrolysis of the ester is so slow with iV/2000-hydrochloric acid that' no difficulty presented itself.To obtain the specific conductivities of the more dilute acid solutions that of the water (1.2 x 10-6 mho) was sub- trated from the observed conductivities. The small conductivity of the methyl acetate was however not subtracted it being no doubt due chiefly t o acetic acid which even in the presence of N / 2000-hydrochloric acid would be' largely undissociated and thus have very little influence on the final conductivity. No correc- tions for viscosity are introduced in the case of the results for N-hydrochloric acid as the applicability of such a correction is doubtful in this case.TABLE VIII. Equivalent Conductivities and Ilegrees of Dissociatioiz (a) for N- and N/ IO-Hydroc?iloric Acid iri Jfixtures of TVater c t d Methyl Acetate at 2 5 O . Concentration of methyl Equivalent conductivity a. acetate in / \ - mols. per litre. N-HCl. N/10-HCl. N/2000-HCl. N-HCI. N/lO-HCl 1.741 247.4 - 336.8 0.734 - 1.151 274.8 - 370.2 0.742 (0.853) 0.704 296.8 344.5 396.9 0.748 0.868 0.2425 - 367.1 409.0 - 0-897 0.1300 - 374.4 412.5 - 0.908 A TABLE IX. Epivaleiat Conductivities and Degrees of Dissociation (a) for N/2-Hydrochloric Acid in Mixtures of Water amd Methyl Acetate at 24.3'. Concentration of Equivalent conductivity methyl acetate in I \ a.mols. per litre. NIB-HCl. N/ZOOO-HCl. for NI2-HCI. 1.744 262.6 333.0 0.789 1-155 - - (0.796) 0.6857 317-6 392.4 0.810 0.2420 - - (0.837) 0.1270 315.0 407.8 0.84 STUDIES 1N CATALYSIS. PART IV. 77 The abave tables show that the degree of dissociation of HC1 obtained in this way falls as the concentration of methyl acetate is increased. Trelocit?i-coizstaiits Corrected for t h e Chaiige in the Degree of Dissociation of the Catalyst. I n order t o correct the velocity-constants for the change in total catalyst the most probable value of the ratio of the catalytic activity of the undissociated molecule of hydrochloric acid to that of the hydrogen ion must be ascertained. This “constant,” which is denoted by kM/kH is undoubtedly dependent on the reaction which is catalysed by the hydrochloric acid and following Snethlage (loc.cit.) and H. S. Taylor (Zeitscli. Elektrochem. 1914 20 201) we take kM/kH=2 although this is probably only a first approxi- mation. The correction is applied t o the velocity-const’ants in the following way. If ?a be the concentration of the acid and a its degree of dissociation then the catalytic power of the acid is given by the expression ?zakH+n(l -a)k,=nakH+2n(1 -a)kH=nkH(2 -a). As k. is a constant its value is immaterial and comparative corrected results are obtained by dividing the velocity-constants k l given in tables V. VI. and VII. by the corresponding values of n(2-a). This has been done and the results are given in Tables X.-XII. the new velocity-constants being given in the column headed (‘ Corrected velocity-constants.” TABLE X.Velocity-co?istants with N-Hydrochloric Acid at 25O Corrected for th,e Change it^ Catalytic Activity of the Acid. Concentration of Corrected velocity- methyl acetate. a for N-HC1. k * constants. 1-741N 0.734 0*0001652 0*0001305 1.151N 0.742 0.0001 482 0.0001179 0.7093N 0.748 0.0001 392 0*0001112 TABLE XI. ‘C’elocity-co,ista?tts with O*5N-Hydrochloric Acid at 24.3O Corrected for the Change in Catalytic Activity of the Acid. Concentration of Corrected velocity- methyl acetate. a for 0-5 N-HC1. k * constants. 1-744N 0.780 0-0000686 0.0001 134 1.155N 0.796 0.0000648 0.0001076 0-6857N 0*810 0-0000623 0.000 1047 0-242OiV 0.83’7 0.0000569 0.000097 8 0.1270N 0.846 0.0000551 0.000095 78 GRIFFlTH AND LEWlS : TABLE XII.TTelocity-coiistaizts with Q*IN-Iiydrochloric Acid at 25*Q0 Corrected for the Change in Catalgtic Activity of the Acid. Concentration of Corrected velocity- methyl acetate. n for 0.1 N-HC1. k * constants. 1.156N 0.853 O*OOOO 1302 0.0001 135 0.7013N 0.768 0.00001 236 0~0001091 0.2 42 5 N 0.897 0*00001177 0.0001067 0.1304N 0.908 0*00001093 0~0001001 The values in the column headed " Corrected velocity-constants " in the three above tables exhibit the same rise on increasing the concentration of the ester although not so markedly as the k, values themselves ; that is the introduction of the correction which allows for the actual change in the degree of dissociation of the acid as the concentration of ester increases and also allows for the catalytic activity of the undissociated niolecule of hydrochloric acid is insufficient t o explain the phenomenon quantitatively.This assumes that the eonductivity method gives correct values for a and that the value of k,/k is 2. It is important t o notice that in order to make all the corrected constants of table XI. identical it would be necessary t o assume k,/lc to be of the order 20-a value which on other grounds is inadmissible. The abnormality still remains. I n order however to make certain that the phenomenon is not connected with the fact that in the case of hydrochloric acid the value of k is greater than that of k (that is the undissociated molecule more catalytically active than the hydrogen ion) it was thought advisable to repeat the experiments using as catalyst another acid f o r which kM/lcH is known t o be less than unity.Trichloroacetic acid was eventually chosen it having the advantage of being a very strong acid with a dissociation-constant of 1-21 at 25O and H. S. Taylor (ilfedd. K . Tretemkapsakad. Nobelinst., 1913 2 No. 37) has shown conclusively that the catalytic activity of its undissociated molecule is about a third that of the hydrogen ion. Experinierits using Tricldoroacetic Acid as Catalyst. The trichloroacetic acid first obtained melted fairly sharply a t 55O (corr.) and in order to purify it from any possible liquid impurities i t was allowed to drain for two days on a porous plate in a vacuum. The.resulting acid melted a t 56O but a suitable solvent froni which to recrystallise it was not found.That there was no serious amount of impurity present was shown by a deter STUDIES IN CATALYSIS. PART IV. 79 mination of its equivalent. Two experiments on the velocity of hydrolysis of methyl aceiate were carried out a t 25* using O.5N-trichloroacetic acid as catalyst the concentration of the ester being 1 and 5 C.C. respectively in 50 C.C. of solution. The former gave a unimolecular constant of 0.00250 whilst the latter gave a constant which rose slightly from 0*00260 to 0.00279. With regard to the latter experiment a distinct odour of the methyl ester of trichloroacetic acid was observed after the reaction was completed and the formation of this ester is presumably the cause of the rising “constant.” In any case the second constant for the hydrolysis of 5 C.C.of ester is unmistakably higher than the former the hydrolysis of 1 C.C. of ester so that the same effect is observed as in the presence of hydrochloric acid as the catalyst. However the presence of trichloroacetic acid complicates the system as the acid combines with some methyl alcohol and there- fore no attempt could be made to determine the equilibrium- constant of the reaction in the presence of trichloroacetic acid and thus to calculate the true velocity-constant of the hydrolysis k,. From analogy how- ever to the case in which hydrochloric acid is the catalyst if the unimolecular constants are increased as the ester content increases, so also are kl and Ic,. An experiment was also attempted with 0*5N-trichloroacetic acid as catalyst and 8 C.C.of methyl acetate in 50 C.C. of solution but in this case two liquid layers resulted, the heavier being chiefly trichloroacetic acid and methyl acetate. The above two experiments however are sufficient to show that the same effect is observed using trichloroacetic acid as catalyst as with hydrochloric acid that is the velocity of hydrolysis increases as the initial concentration of the ester is increased. It was shown for the case of hydrochloric acid that correcting for the reverse reaction the formation of ester did not account for the change and it is very probable that the same is the case with trichloroacetic acid as the catalyst. the velocity-constant with increase in initial concentration of the ester is therefore not causally connected with a particular cata- lysing acid.CH,*CO,*CH + H,O -” CH,*OH + CH3*C0,H The increase in the value of - Experiments with More Concentrated Solutions of Methyl Aceta,te. All the experiments described up t o this point gave well-agreeing velocity-constants but it was found that further increase in the concentration of the ester caused a rising velocity (‘ constant ” throughout a single determination. The figures given in table XIII. illustrate this. The concentration of the ester was abou 80 GRTFFITH AND LEWIS: 13 C.C. in 50 C.C. of solution that of the catalyst (hydrochloric acid) 0.51\T and the temperature 24'3O. TABLE XIII. Unimolecular Time from start. constant k. 39.0 minutes 0.003926 74.5 , 0.003900 120.3 , 0.003914 200.3 ) 0.004 14 6 Unimolecular Time from start.constant k. 240.3 minutes 0.004176 281.3 ,) 0.004249 333.6 3 y 0.004273 402.1 , 0*004491 The unimolecular constant L steadily rises. With a concentra- tion of ester of 10 C.C. in 50 C.C. of solution there was also a rise, which although definite was not so pronounced as in the case of 13 C.C. I n the former case the "constant" rose from 0.00387 at the commencement of th.e reaction to 0.00414 a t the end whilst with 8 C.C. of ester on the other hand 110 deviation from constancy in I; could be observed This particular phenomenon was not pursued further and is only introduced here because of its bear- ing on the results given above concerning the rise in the velocity- constants when the initial concentration of ester is increased.Discussion of R es id t s. The problem raised by the results obtained in the present work is the mode of expressing the active mass of a constituent which acts stoicheiometrically and also catalytically (as solvent). I n the first place i t may be pointed out that the substitution of osmotic pressure f o r concentration as a more correct measure of active mass a suggestion made by Arrhenius (Zeitsch. physikal. Chem., 1899 28 317) in the analogous case of inversion of sucrose does not correspond to the facts a conclusion already reached by Rosanoff (Zoc. cit.). I n the second place it may be suggested that the dependence of the velocity-constant on the initial conditions is due to the formation of intermediate compounds which function as the reactants and resultants respectively.The simplest assump- tion to make is that the ester and water first produce an additive compound which is catalytically decomposed into methyl alcohol and acetic acid. Applying the principle of mass action to the instantaneous formation of complex namely, ester + water (ester water), i t is evident that a greater quantity of complex will be formed on increasing the ester concentration a t the expense of the water up to a certain composition of the mixture. This suggestion may be applied in the following way. If the concentration of ester b STUDIES IN CATALYSIS PART IV. 81 initially a gram-molecules per litre that of the water w and that of the complex e gram-molecules the equilibrium will be given by the relation Kl(a - e)(w - e ) = e, where K / is the equilibrium-constant.Since in the experiments already described a is small compared with w e must be still smaller so that we can rewrite the above relation thus: If the complex be the reactive substance we have (when the reverse reaction is negligible) for reaction K'zua 1 + K'zu' $= ktrue x e = ktroe x ~ where ktrre stands for the velocity-constant of a t t,he beginning the velocity of decomposition of the complex. takes the form On the other hand the equat,ion usually employed 5 = /cobs x a x zu. at where I; obs is the velocity-constant actually measured. It follows therefore that This relation may be tested by,calculating by its aid values of the equilibrium-constant K' by taking any two values of koobs, corresponding with two values of w.This has been done employ- ing the series of '' corrected velocity-constants " given in table XI. Six negative and four positive values of K' were obtained. The equation was also applied to the analogous case of the inversion of sucrose the data being tIhose of Ostwald Spohr and Rosanoff (compare Rosanoff Zoc. cit.). In all castes a negative value for K / is the r-esult. The negative value has of course no physical significance but its frequency of occurrence suggests that it is not due to experimental error. It will be seen however on inspection of the data that the cause of the negative sign is the fact that the (observed) velocity- constant rises too rapidly for a given diminution in initial water- content that is the values of e are greater (as the water-content diminishes) than would be anticipated on the basis of massaction alone.If we accept the explanation that the reactive substance is the additive compound then i t follows that the solvent is exert- ing a catalytic effect a t the same time in the sense that diminu- tion in the water favours the production of the complex. Thi 82 STUDIES IN CATALYSIS. PART IV. is in agreement with what one would expect in view of the strongly dissociating power of water. As the initial water-content is diminished the medium becomes richer in methyl acetate and therefore as a whole must act as a feebler dissociating agent on the additive complex (ester water) which is therefore produced in greater quantity than would be the case if the physical action of the solvent remained constant.This explains not only the apparent negative catalytic effect of water in this case but like- wise its negative catalytic effect on the rate of esterification in- vestigated by Lapworth and Fitzgerald (T. 1908 93 2174), although in the latter case the water plays no stoicheiometric ” part. I n view of the evidence obtained from spectroscopic measure- ments in the infra-red region described in Part 111. of this series, in favour of the existence of an ester-water complex the sugges- tion of a simultaneous stoicheiometric and anticatalytic r61e on the part of the water present as an explanation of the variation of velocity-constant with initial composition of the system receives considerable support.The conclusion is also reached that the phenomenon must be a genera1 one since solvation appears to be general and all liquids possess dissociating power to some degree. This also is borne out by a number of cases quoted by van’t Hoff ( ‘ I Studies in Chemical Dynamics,” p. 28) who concludes that “it may therefore also happen that one of the liquid bodies which is taking part in the reaction is itself an unfavourable medium.” Negative catalysis by one of the constituents of the solvent is there- fore not exceptional even when the constituent participates stoicheiometrically a t the same time. Naturally dissociating power itself requires a physical mechanism. It is suggestive that Kruger (Zeitscli. EZe?;ti*ochem. 1911 17 453) attributes it to the presence of infra-red radiation.Summary. (I) Measurements of the velocity of hydrolysis of methyl acetate with varying initial concentrations of the ester have been carried out using as catalysts O-lN- 0.5N- and N-hydrochloric acid and also 0-5N-trichloroacetic acid. It is found that the unimolecular const,ants k and the constants L, which are corrected for the reverse reaction were constant in each experiment but their value depended on the initial composition of the system the value rising with the concentration of the ester. (2) This increase is not accounted for by correcting the constants for the change in catalysing power of the acid due t o the change in the degree of dissociation as more ester is added PROPAGATION OF FLAME IN HYDROGEN AND AIR.83 (3) The equilibrium-constant of the reaction between water and methyl acetate has been redetermined and is found to be 4.6 in the presence of hydrochloric acid not exceeding N in strength. (4) The variation in velocity-constant with initial conditions can be explained if the reacting substance is considered to be an ester-water additive complex the degree of dissociation of which decreases as the water-content decreases that is the water acts stoicheiometrically and at- the same time as a negative catalyst in virtue of its high dissociating power. MUSPRATT LABORATORY OP PHYSICAL AND ELECTRO-CHEMISTRY. UNIVERSITY OF LIVERPOOL. [Received November 12th 1916. STUDIES IN CATALYSIS. PAKT 1V. 67 VT.-Studies in Catalysis. Pui-t I V. Xtoicheiomctric and Catalytic Efects Due to the Progressive Displacement of oyie Reactant by Another in the ‘‘ Acid” Hydrolysis o f Methyl Acetate.By ROBERT OWEN GRIFFITH and WILLIAM CUDMORE MCCULLAGH LEWIS. IT is a remarkable fact although it does not appear t o have attracted as much attention as i t deserves that the velocity- constant k for the reaction CH3*C0,*CH3 + H,O + CH3*OH + CH,=CO,H, as obtained from the equation k = Llog - varies with the initial concentration of the methyl acetate. Ostwald who first investi- gated the reaction obtained the following results ( J . p. Chem., 1883 [ii] 28 449). At the temperature of 25*2O with the same concentration of hydrochloric acid in each case (23/3) the con- centrations of ester employed in the experiments were 2 c.c. 1 c.c., 0.5 c.c.and 0-3 C.C. respectively in 15 C.C. of solution. C.C. of methyl acetate ,t=24*19. >> 9 ) 7 , k=21.96. >7 0.5 ,> 97 , li=20*71. 7 0.3 ,> 7 , k=19-91. a t a-x With 68 GRIFFITH AND LEWIS: Ostwald attributes this decrease in the velocity-constants to the fact that the end condition (in each case attained after two days) varies with the initial Concentration in such a way that in the solution intially containing least ester the decomposition progresses to the furthest extent and thus the velocity-constants differ because of the existence of an equilibrium point. He illustrates this in the following way. If we take in each case the final titra- tion value in C.C. of alkali and subtract the value due to the hydro- chloric acid we obtain the value due t o the acetic acid formed by the hydrolysis of the ester.Dividing these figures by the initial amount of ester in C.C. in each experiment' one would expect the same value if the total amount or some fraction of the' ester were decomposed in each case. The table given below shows however, that this is not so. TABLE I. C.C. of barium hydroxide solution required for 1 C.C. of ester. End point in C.C. of barium hydroxide solution. Initial ester in C.C. 2.0 25.96 12.98 1.0 14-34 14.34 0.5 7.66 15.32 0.3 4.66 15-53 Ostwald states that the magnitudes given in the last column of table I. and the corresponding velocity-constants are inversely proportional. A fairly constant result is obtained by multiplying the figures in the last column by the corresponding velocity- constants.From this he deduces the empirical relation which should ho'ld for all different initial concentrations of ester namely, a log - = k? t. a-x p in which p is the initial concentration of ester and a x and t have their usual significance. This relation holds independently of the nature of the acid. In criticising this result the first point to be noted is that the constants are not truly comparable as they stand for the concen- tration of the water differs in each experiment' and the reaction- velocity should be proportional to the concentration of the water. I n the experiments with the largest amounts of methyl acetate there is less water than in the; others. This obvious correction, which was applied by Rosanoff t o the case of the inversion of sucrose (J.Amer. Chem. SOC. 1913 35 248> has the effect of making the constants diverge still more from each other. This correction which we will call the " stoicheiometric " correction for water can be very approximately applied tol Ostwald's results b STUDIES IN CATALYSIS. PART IV. 69 assuming that the methyl acet'ate displaces its own volume of water from the solution. I n this way we obtain the following. TABLE 11. Constants corrected for variat,ion in water content. Ostwald's unimolecular consCants. 24-19 1.888 21-96 1-590 20.71 1.447 19.99 1.378 The corrected constants are obtained by dividing those of Ostwald by the weight of water present in 15 c.c. the total volume of solution in Ostwald's experiments.As the figures are purely relative the above method serves f o r comparison. From the tables given above it is seen that2 the divergence is now greater than before so that Ostwald's empirical relation loses much of its significance. I n order therefore t o attempt to determine satis- factorily whether the change in the velocity-constant is duel t o the variation in the end point that is the equilibrium point with different initial concentrations of the ester the present research was undertaken. The A cc.zcmfe ?7clocity Equation. The reaction is obviously bimolecular and reversible although in dilute aqueous solution a unimolecular constant is necessarily obtained. Denoting the initial concentration of the methyl acetate in mols. per litre by b the number of mols.decomposed a t time t by x and the concentration of the water in mols. per litre by w thevelocity- constant for the decomposition of the ester by k and the velocity- constant' for the formation of ester by k, we have the differential equation dx - k,(b - X)(PU - z) - k&X? at - Writing k! = K the equilibrium constant where k , [ester] x [wat,er] [alcohol] x [acetic acid]' K= __ ___.__~- ~ - _ _ _ and also setting y=iu+ b and 3 = J(w+ b V + 4(R- l)wb integra- tion gives reme'mbering that when t=O x=O, k = Kk 70 GRIFFITH AND LEWIS As is t o be expected neglect of the change in the concentration of water during an experiment makes very little difference in the value of k obtained. Assuming that 2 ~ is a constant we have @ = k p ( b - x) - k,2? d t This on integration gives a value f o r h.which is very near t.0 the value obtained as above (generally within 1 per cent.) altliough the variatJon in the concentration of water from beginning to end of a reaction is in the case of the most concentrated solution of ester employed more than 4 per cent. of t,he initial concentration of water. The more complete equation given above has however, been employed throughout. E X P E R I M E N T A L . Methyl acetate was redistilled and the portion boiling between 55'4O and 56'0° was used; it contained very little free acetic acid, the specific conductivity of an approximately normal solution of the ester in water being 40 x 10-6 mho. The concentration of the catalyst hydrochloric acid was N / 2 in the first set of measure- ments N in the second and N / l O in the third.The concentra- tions of ester were respectively about 8 5 3 1 and 0.5 C.C. in 50 C.C. of total solution. The exact concentrations in mols. per litre are given in the table of results. For the measurements of velocity the usual method due to Ostwald was employed. The reactions were carried out in Jena-glass flasks which had been thoroughly steamed out and fitted with well-paraffined corks the flasks being kept in a thermostat maintained a t a constant tempera- ture. From time t o time a convenient volume of the reaction mixture was withdrawn by a pipette run into 40 C.C. of conductivity water and titrated against standard baryta. For each experi- ment two dsterminations were made and the means of the two titres a t corresponding times were taken t o calculate the velocity- constants.The end points were taken aft.er allowing the reaction t o proceed for a time sufficient for equilibrium t o be attained. Before starting each experiment the required amount of stock acid solution was run into the flask together with that amount of water (previously determined) which with the ester gave a total volume of 50 C.C. This was then placed in the thermostat for thirty minutes and the required amount of ester added at a noted time. Thus for each experiment the concentration of each con- stituent was known STUDIES IN CATALYSIS. PART IV. 71 Determination of t h e Equilibrium Constant of t h e Reaction. To determine K the equilibrium-constant a knowledge is r e quired of the initial conceiitration of ester the final concentration of the acetic acid and that of the water.The final concentration of the acetic acid which is identical with that of the methyl alcohol is arrived a t from the end-point titra obtained after ten times the time of half decomposition of the ester has elapsed by which time the system is not appreciably removed from its con- dition of equilibrium. The initial concentration of the ester was not determined by weighing into a certain volume but by treating the exact amount taken with an excess of baryta solution and then titrating with hydrochloric acid. The ester was treated with an excess of standard baryta solution f o r eight hours precautions being taken to prevent entry of carbon dioxide from the air and then the excess of baryta was determined.The difference between the initial concentration of ester and the amount decomposed after the solution had reached its position of equilibrium gives the amount of ester present a t equilibrium those of the methyl alcohol and acetic acid being of course equal t o that of the decomposed methyl acetate. The equilibrium-constant. was determined only in the case of the two more concentrated solutions of methyl acetate as experi- mental errors are magnified largely in calculating K for more dilute solutions owing to the very large excess of one of the reactants-water. There is no reason however to believe that there is any variation in the equilibrium-constant with change in the concentration of the ester.The following values were obtained for the equilibrium-constant ester x water methyl alcohol x acetic acid' I<= ~ - _ _ _ _ (1) In the preseiice of N/2-hydrochloric acid 4.30 4.52 4.66, 4.80; (2) in the presence of N / 10-hydrochloric acid 4-52 ; and (3) in the presence of N-hydrochloric acid 4.34 4.83. These are constant within experimental error and the mean value 4-6 was adopted. It may be noted however that the absolute value of K is not of great importance for the object of this research as a fairly large variation in K makes but a small change in the value of the velocity-constant k (constant for ester decomposition) although, of course from formula (1) the value of k is directly proportional to K . The equilibrium-constant K has been determined for the pure substances with no added catalyst by Berthelot and Pean d 72 GRIFFITH AND LEWIS: St.Gilles (Ann. cJhim. PJL~s. 1863 [iii] 68 225) who found that, starting with equivalent amounts of acetic acid and methyl alcohol, 67.5 per cent. of each combined. This gives K=4.31. Further, Menschutkin (Annalen 1879 195 334) obtained 69.5 per cent. of combination which gives K = 5.18. Now Lap- worth and Jones (T, 1911 99 1427) have shown that for ethyl acetate the presence of concentrated hydrochloric acid affects the value of K which increases with decreasing values of the ratio H,O /HC1. However they found that provided the concentration of the hydrochloric acid is below normal the equilibrium-constant is very nearly independent of the hydrochloric acid present.Assuming that in this respect methyl acetate acts similarly t o ethyl acetate it is likely that no very sensible change in the value of K is t o be expected in these experiments. The equilibrium- constant for this reaction in the presence of hydrochloric acid has also been ,determined by Worley (PYOC. Roy. SOC. 1912 [ A ] 87, SSZ) who deduces by extrapolation the value K=6.6 for the constant< with no added catalyst' present a value decidedly higher than those obtained above. The value 4.6 obtained above lies between the two. Velocity Mecrsiirements. Below are given the experimental results. The following is the notation used: t=time in minutes from the start. x=number of gram-molecules of ester decomposed a t time t . 1 CL Ji = unimolecular constant - log -.t a-x k = velocity-constant for ester ,decomposition \ k = velocity-constant for ester formation from equation (1) Tables 111. and IV. serve as an indication of the general character of the velocity-constants obtained in these experiments, whilst tables V. t o VII. summarise the results of the experiments with hydrochloric acid as catalyst STUDIES IN CATALYSIS. PART IV. 73 TABLE 111. Hydrolysis of methyl acetate with 0*5lV-hydrochloric acid a t Initial concentration of methyl acetate = 0.6857 mols. per litre. 24-3O k 0.03. 9 ,9 , water =51.74 , 9 9 9 t. 60 90 120 160 280 340 460 m X. 0.1192 0- 1747 0.2183 0.2746 0.4044 0.4513 0.5141 0.6482 Mean ......... k. 0.00338 0.00348 0.00343 0.00344 0.00349 0.00351 0.00343 - 0.003465 k,.- kl. 0.0000616 0.0000633 - 0.0000620 Mean 0.0000622 0.000284 0.0000628 - 0.0000629 - 0.000061 3 - - - - 0.0000623 0-000284 TABLE IV. Hydrolysis of methyl acetate with 0-13-hydrochloric acid a t Initial concentration of ester = 0.7013N. 25*00° f 0.03. 9 9 9 , water = 52.19.T. t. 200 2 80 445 620 1515 1705 1g30 2. 0.08455 0.1171 0.1727 0.2311 0.4299 0.4588 0.4877 0.6624 k. 0.000683 0.000695 0.000680 0.000692 0*00069 1 0.000692 0.000698 - Mean . . . .. 0.000690 4. - k - 0.00001 237 0*00001255 - 0.0000 1225 Mean 0.00001247 0.0000565 0.0000 1235 - 0.00001236 - 0.00001254 - - - 0*00001236 0.0000565 TABLE V. Catalyst N-Hydrochloric acid a t 25*0°.Initial con- centration No. of C.C. in mols. of ester of est,er in 50 C.C. per litre. of solution. 1-741 8 approx. 1.151 5 ,, 0.7093 3 ,, Mean con- centration of water in mols. per litre=w*. k. klw. kl. k*. 45.38 0.00903 0.0001991 0.0001652 0.000755 48.76 0.00818 0.0001679 0-0001482 0-000677 50.94 0-00764 0.0001467 0.0001392 0.000636 * The mean concentration of water is the concentration midway through an experiment that is when half the methyl acetate has been decomposed. D 74 GRIFFITH AND LEWIS: TABLE VI. Catalyst 0.5A -Hydrochloric acid a t 24*3O.* Initial con- centration No. of C.C. in mols. of ester of ester in 50 C.C. per litre. of solution. 1.744 8 approx. 1.155 5 ,, 0.6857 3 ,) 0-2420 1 ,, 0.1270 0-5 ,, Mean con- centration of water in mols.per litre=w.t k. k/w. k - k2. 46.10 0.00382 0.0000829 0.0000687 0.000313 49.14 0.00363 0.0000739 0.0000648 0.000295 5 1.42 0.00346 0.0000674 0*0000623 0.000284 53.73 0.00310 0*0000578 0.0000569 0.000256 54.15 0.00300 0.0000553 0.0000551 0.000251 * The thermometer used in this series of measurements registered 25.0", t The mean concentration of water is the Concentration midway through but on calibration was found to be in error. an experiment that is when half the methyl acetate has been decomposed. TABLE VII. Catalyst O*lN-Hydrochloric acid a t 25*0°. Initial concen- tration in mols. of est.er per litre. 1.156 0.7013 0.2425 0.1304 Mean concen- No. of C.C. tration of ester of water in 50 C.C.5 approx. 49.69 0.000726 0.00001461 0.00001302 0.0000595 3 ? 9 51.86 0.000690 0*00001330 0.00001236 0*0000565 1 Y Y 54.25 0.000660 0*00001217 0*00001177 0.0000538 0.5 1 j 54-86 0.000606 0*00001104 0*00001093 0.0000499 in mols. per of solution. litre= w. k. kjw. El. Ic,. Throughout each single experiment the values of k k, and k , are constant within the experimental error that is there is no trend observable even in the most concentrated solutions of methyl acetate. From the summarising tables given above it will be seen that the values of the unimolecular constants k fall steadily with decreasing concentration of ester as Ostwald has already shown ; the values of k / w fall more rapidly still. Also the values of k,-the true velocity-constant-still show the same decided fall although not nearly so marked as that of k / w .The effect of the correction term k2x2 in the differential equation, 5!!? = k,(zo - z)( b - z) - k2z2, d t may be seen in the above tables by comparison of corresponding values of k / w and k,. It is seen that these two values in the case of the most dilute solution of methyl acet,ate employed (0.1N) are identical within experimental error but as the concentration of ester is increased k / w increases inore rapidly than k,. Assuming the law of mass-action to be valid throughout the constants given in the column headed k should be of equal value STUDIES IN CATALYSTS PART IV. 75 as they include the correction for the reverse reaction the forma- tion of ester.Ostwald's statement therefore that the variation in the velocity-constants with change in the concentration of ester is due to there being an equilibrium point is insufficient and we must therefore consider other explanations to account for this variation. A first suggestion is that the concentration of the catalyst is not the same in each experiment; although the total concentration of hydrochloric acid is the same in each series the degree of dissociation of the acid might change with the amount of ester present. Presumably the addition of the ester would decrease the dissociation of the acid thus lowering the hydrogen ion concentration; a t first sight i t appears if this view be correct, that the velocity-constants should decrease as the concentration of the methyl acetate is increased.However from the recent work of Snethlage (Zeitsch. physikd. C'henz. 1913 85 253) Acree (Amel.. C'hem J. 1912 48 352) and others the results of measurements of velocity of reaction can be interpreted so as t o make it appear possible that the undissociated molecule of hydrochloric acid has a catalytic activity about twice as great as that of the hydrogen ion. On $he assumption then, that the presence of methyl acetate decreases the percentage ionisation of the catalysing acid hydrochloric it is t o be expected that the velocity-constants should be greater the more ester is present. To test this therefore i t is necessary t o know the dis- sociation of the acid in each solution used and for this purpose measurements of the conductivities of the acid in mixtures of water and methyl acetate were carried out.Determination of the Degree of Dissociation of Hydrochloric Acid in Mixtures of Water and Methyl Ace!ate. The electrical conductivities of several of the solutions used in the experiments on the velocity of the reaction were determined, and also those of solutions of N/2000-hydrochloric acid in the presence of varying amounts of methyl acetate. Assuming com- plete dissociation of the hydrochloric acid at the latter dilution, the degrees of dissociation for the other concentrations were calcu- lated in the usual way. The ordinary Kohlrausch method and apparatus was used the bridge being calibrated by Strouhal and Barus' method whilst the cell constant was determined by the use of iV/50-potassium chloride which has a specific conductivity of 0.002728 a t 24.3O and of 0*002768 a t 25O.For the more dilute acid solutions the cell employed was of the Arrhenius type whilst for the more concentrated solutions the conductivity vessel was a U-tube with closely fitting ebonite caps which held the electrodes D* 76 GIUFFITH AND LEWIS : firmly in position. Tables VIII. and IX. give the results of the measurements ; interpolated values in these tables are bracketed. The electrical conductivities of the solutions were determined immediately after making up and allowing them to attain the temperature of the thermostat' but nO very great change was observed in the conductivities of the solution with time the acetic acid gradually formed being very little dissociated in the presence of concentrated hydrochloric acid whilst the hydrolysis of the ester is so slow with iV/2000-hydrochloric acid that' no difficulty presented itself.To obtain the specific conductivities of the more dilute acid solutions that of the water (1.2 x 10-6 mho) was sub- trated from the observed conductivities. The small conductivity of the methyl acetate was however not subtracted it being no doubt due chiefly t o acetic acid which even in the presence of N / 2000-hydrochloric acid would be' largely undissociated and thus have very little influence on the final conductivity. No correc- tions for viscosity are introduced in the case of the results for N-hydrochloric acid as the applicability of such a correction is doubtful in this case.TABLE VIII. Equivalent Conductivities and Ilegrees of Dissociatioiz (a) for N- and N/ IO-Hydroc?iloric Acid iri Jfixtures of TVater c t d Methyl Acetate at 2 5 O . Concentration of methyl Equivalent conductivity a. acetate in / \ - mols. per litre. N-HCl. N/10-HCl. N/2000-HCl. N-HCI. N/lO-HCl 1.741 247.4 - 336.8 0.734 - 1.151 274.8 - 370.2 0.742 (0.853) 0.704 296.8 344.5 396.9 0.748 0.868 0.2425 - 367.1 409.0 - 0-897 0.1300 - 374.4 412.5 - 0.908 A TABLE IX. Epivaleiat Conductivities and Degrees of Dissociation (a) for N/2-Hydrochloric Acid in Mixtures of Water amd Methyl Acetate at 24.3'. Concentration of Equivalent conductivity methyl acetate in I \ a. mols. per litre. NIB-HCl. N/ZOOO-HCl. for NI2-HCI. 1.744 262.6 333.0 0.789 1-155 - - (0.796) 0.6857 317-6 392.4 0.810 0.2420 - - (0.837) 0.1270 315.0 407.8 0.84 STUDIES 1N CATALYSIS.PART IV. 77 The abave tables show that the degree of dissociation of HC1 obtained in this way falls as the concentration of methyl acetate is increased. Trelocit?i-coizstaiits Corrected for t h e Chaiige in the Degree of Dissociation of the Catalyst. I n order t o correct the velocity-constants for the change in total catalyst the most probable value of the ratio of the catalytic activity of the undissociated molecule of hydrochloric acid to that of the hydrogen ion must be ascertained. This “constant,” which is denoted by kM/kH is undoubtedly dependent on the reaction which is catalysed by the hydrochloric acid and following Snethlage (loc.cit.) and H. S. Taylor (Zeitscli. Elektrochem. 1914 20 201) we take kM/kH=2 although this is probably only a first approxi- mation. The correction is applied t o the velocity-const’ants in the following way. If ?a be the concentration of the acid and a its degree of dissociation then the catalytic power of the acid is given by the expression ?zakH+n(l -a)k,=nakH+2n(1 -a)kH=nkH(2 -a). As k. is a constant its value is immaterial and comparative corrected results are obtained by dividing the velocity-constants k l given in tables V. VI. and VII. by the corresponding values of n(2-a). This has been done and the results are given in Tables X.-XII. the new velocity-constants being given in the column headed (‘ Corrected velocity-constants.” TABLE X. Velocity-co?istants with N-Hydrochloric Acid at 25O Corrected for th,e Change it^ Catalytic Activity of the Acid.Concentration of Corrected velocity- methyl acetate. a for N-HC1. k * constants. 1-741N 0.734 0*0001652 0*0001305 1.151N 0.742 0.0001 482 0.0001179 0.7093N 0.748 0.0001 392 0*0001112 TABLE XI. ‘C’elocity-co,ista?tts with O*5N-Hydrochloric Acid at 24.3O Corrected for the Change in Catalytic Activity of the Acid. Concentration of Corrected velocity- methyl acetate. a for 0-5 N-HC1. k * constants. 1-744N 0.780 0-0000686 0.0001 134 1.155N 0.796 0.0000648 0.0001076 0-6857N 0*810 0-0000623 0.000 1047 0-242OiV 0.83’7 0.0000569 0.000097 8 0.1270N 0.846 0.0000551 0.000095 78 GRIFFlTH AND LEWlS : TABLE XII. TTelocity-coiistaizts with Q*IN-Iiydrochloric Acid at 25*Q0 Corrected for the Change in Catalgtic Activity of the Acid.Concentration of Corrected velocity- methyl acetate. n for 0.1 N-HC1. k * constants. 1.156N 0.853 O*OOOO 1302 0.0001 135 0.7013N 0.768 0.00001 236 0~0001091 0.2 42 5 N 0.897 0*00001177 0.0001067 0.1304N 0.908 0*00001093 0~0001001 The values in the column headed " Corrected velocity-constants " in the three above tables exhibit the same rise on increasing the concentration of the ester although not so markedly as the k, values themselves ; that is the introduction of the correction which allows for the actual change in the degree of dissociation of the acid as the concentration of ester increases and also allows for the catalytic activity of the undissociated niolecule of hydrochloric acid is insufficient t o explain the phenomenon quantitatively.This assumes that the eonductivity method gives correct values for a and that the value of k,/k is 2. It is important t o notice that in order to make all the corrected constants of table XI. identical it would be necessary t o assume k,/lc to be of the order 20-a value which on other grounds is inadmissible. The abnormality still remains. I n order however to make certain that the phenomenon is not connected with the fact that in the case of hydrochloric acid the value of k is greater than that of k (that is the undissociated molecule more catalytically active than the hydrogen ion) it was thought advisable to repeat the experiments using as catalyst another acid f o r which kM/lcH is known t o be less than unity.Trichloroacetic acid was eventually chosen it having the advantage of being a very strong acid with a dissociation-constant of 1-21 at 25O and H. S. Taylor (ilfedd. K . Tretemkapsakad. Nobelinst., 1913 2 No. 37) has shown conclusively that the catalytic activity of its undissociated molecule is about a third that of the hydrogen ion. Experinierits using Tricldoroacetic Acid as Catalyst. The trichloroacetic acid first obtained melted fairly sharply a t 55O (corr.) and in order to purify it from any possible liquid impurities i t was allowed to drain for two days on a porous plate in a vacuum. The.resulting acid melted a t 56O but a suitable solvent froni which to recrystallise it was not found. That there was no serious amount of impurity present was shown by a deter STUDIES IN CATALYSIS.PART IV. 79 mination of its equivalent. Two experiments on the velocity of hydrolysis of methyl aceiate were carried out a t 25* using O.5N-trichloroacetic acid as catalyst the concentration of the ester being 1 and 5 C.C. respectively in 50 C.C. of solution. The former gave a unimolecular constant of 0.00250 whilst the latter gave a constant which rose slightly from 0*00260 to 0.00279. With regard to the latter experiment a distinct odour of the methyl ester of trichloroacetic acid was observed after the reaction was completed and the formation of this ester is presumably the cause of the rising “constant.” In any case the second constant for the hydrolysis of 5 C.C. of ester is unmistakably higher than the former the hydrolysis of 1 C.C.of ester so that the same effect is observed as in the presence of hydrochloric acid as the catalyst. However the presence of trichloroacetic acid complicates the system as the acid combines with some methyl alcohol and there- fore no attempt could be made to determine the equilibrium- constant of the reaction in the presence of trichloroacetic acid and thus to calculate the true velocity-constant of the hydrolysis k,. From analogy how- ever to the case in which hydrochloric acid is the catalyst if the unimolecular constants are increased as the ester content increases, so also are kl and Ic,. An experiment was also attempted with 0*5N-trichloroacetic acid as catalyst and 8 C.C. of methyl acetate in 50 C.C. of solution but in this case two liquid layers resulted, the heavier being chiefly trichloroacetic acid and methyl acetate.The above two experiments however are sufficient to show that the same effect is observed using trichloroacetic acid as catalyst as with hydrochloric acid that is the velocity of hydrolysis increases as the initial concentration of the ester is increased. It was shown for the case of hydrochloric acid that correcting for the reverse reaction the formation of ester did not account for the change and it is very probable that the same is the case with trichloroacetic acid as the catalyst. the velocity-constant with increase in initial concentration of the ester is therefore not causally connected with a particular cata- lysing acid. CH,*CO,*CH + H,O -” CH,*OH + CH3*C0,H The increase in the value of - Experiments with More Concentrated Solutions of Methyl Aceta,te.All the experiments described up t o this point gave well-agreeing velocity-constants but it was found that further increase in the concentration of the ester caused a rising velocity (‘ constant ” throughout a single determination. The figures given in table XIII. illustrate this. The concentration of the ester was abou 80 GRTFFITH AND LEWIS: 13 C.C. in 50 C.C. of solution that of the catalyst (hydrochloric acid) 0.51\T and the temperature 24'3O. TABLE XIII. Unimolecular Time from start. constant k. 39.0 minutes 0.003926 74.5 , 0.003900 120.3 , 0.003914 200.3 ) 0.004 14 6 Unimolecular Time from start. constant k. 240.3 minutes 0.004176 281.3 ,) 0.004249 333.6 3 y 0.004273 402.1 , 0*004491 The unimolecular constant L steadily rises.With a concentra- tion of ester of 10 C.C. in 50 C.C. of solution there was also a rise, which although definite was not so pronounced as in the case of 13 C.C. I n the former case the "constant" rose from 0.00387 at the commencement of th.e reaction to 0.00414 a t the end whilst with 8 C.C. of ester on the other hand 110 deviation from constancy in I; could be observed This particular phenomenon was not pursued further and is only introduced here because of its bear- ing on the results given above concerning the rise in the velocity- constants when the initial concentration of ester is increased. Discussion of R es id t s. The problem raised by the results obtained in the present work is the mode of expressing the active mass of a constituent which acts stoicheiometrically and also catalytically (as solvent).I n the first place i t may be pointed out that the substitution of osmotic pressure f o r concentration as a more correct measure of active mass a suggestion made by Arrhenius (Zeitsch. physikal. Chem., 1899 28 317) in the analogous case of inversion of sucrose does not correspond to the facts a conclusion already reached by Rosanoff (Zoc. cit.). I n the second place it may be suggested that the dependence of the velocity-constant on the initial conditions is due to the formation of intermediate compounds which function as the reactants and resultants respectively. The simplest assump- tion to make is that the ester and water first produce an additive compound which is catalytically decomposed into methyl alcohol and acetic acid.Applying the principle of mass action to the instantaneous formation of complex namely, ester + water (ester water), i t is evident that a greater quantity of complex will be formed on increasing the ester concentration a t the expense of the water up to a certain composition of the mixture. This suggestion may be applied in the following way. If the concentration of ester b STUDIES IN CATALYSIS PART IV. 81 initially a gram-molecules per litre that of the water w and that of the complex e gram-molecules the equilibrium will be given by the relation Kl(a - e)(w - e ) = e, where K / is the equilibrium-constant. Since in the experiments already described a is small compared with w e must be still smaller so that we can rewrite the above relation thus: If the complex be the reactive substance we have (when the reverse reaction is negligible) for reaction K'zua 1 + K'zu' $= ktrue x e = ktroe x ~ where ktrre stands for the velocity-constant of a t t,he beginning the velocity of decomposition of the complex.takes the form On the other hand the equat,ion usually employed 5 = /cobs x a x zu. at where I; obs is the velocity-constant actually measured. It follows therefore that This relation may be tested by,calculating by its aid values of the equilibrium-constant K' by taking any two values of koobs, corresponding with two values of w. This has been done employ- ing the series of '' corrected velocity-constants " given in table XI.Six negative and four positive values of K' were obtained. The equation was also applied to the analogous case of the inversion of sucrose the data being tIhose of Ostwald Spohr and Rosanoff (compare Rosanoff Zoc. cit.). In all castes a negative value for K / is the r-esult. The negative value has of course no physical significance but its frequency of occurrence suggests that it is not due to experimental error. It will be seen however on inspection of the data that the cause of the negative sign is the fact that the (observed) velocity- constant rises too rapidly for a given diminution in initial water- content that is the values of e are greater (as the water-content diminishes) than would be anticipated on the basis of massaction alone.If we accept the explanation that the reactive substance is the additive compound then i t follows that the solvent is exert- ing a catalytic effect a t the same time in the sense that diminu- tion in the water favours the production of the complex. Thi 82 STUDIES IN CATALYSIS. PART IV. is in agreement with what one would expect in view of the strongly dissociating power of water. As the initial water-content is diminished the medium becomes richer in methyl acetate and therefore as a whole must act as a feebler dissociating agent on the additive complex (ester water) which is therefore produced in greater quantity than would be the case if the physical action of the solvent remained constant. This explains not only the apparent negative catalytic effect of water in this case but like- wise its negative catalytic effect on the rate of esterification in- vestigated by Lapworth and Fitzgerald (T.1908 93 2174), although in the latter case the water plays no stoicheiometric ” part. I n view of the evidence obtained from spectroscopic measure- ments in the infra-red region described in Part 111. of this series, in favour of the existence of an ester-water complex the sugges- tion of a simultaneous stoicheiometric and anticatalytic r61e on the part of the water present as an explanation of the variation of velocity-constant with initial composition of the system receives considerable support. The conclusion is also reached that the phenomenon must be a genera1 one since solvation appears to be general and all liquids possess dissociating power to some degree. This also is borne out by a number of cases quoted by van’t Hoff ( ‘ I Studies in Chemical Dynamics,” p. 28) who concludes that “it may therefore also happen that one of the liquid bodies which is taking part in the reaction is itself an unfavourable medium.” Negative catalysis by one of the constituents of the solvent is there- fore not exceptional even when the constituent participates stoicheiometrically a t the same time. Naturally dissociating power itself requires a physical mechanism. It is suggestive that Kruger (Zeitscli. EZe?;ti*ochem. 1911 17 453) attributes it to the presence of infra-red radiation. Summary. (I) Measurements of the velocity of hydrolysis of methyl acetate with varying initial concentrations of the ester have been carried out using as catalysts O-lN- 0.5N- and N-hydrochloric acid and also 0-5N-trichloroacetic acid. It is found that the unimolecular const,ants k and the constants L, which are corrected for the reverse reaction were constant in each experiment but their value depended on the initial composition of the system the value rising with the concentration of the ester. (2) This increase is not accounted for by correcting the constants for the change in catalysing power of the acid due t o the change in the degree of dissociation as more ester is added PROPAGATION OF FLAME IN HYDROGEN AND AIR. 83 (3) The equilibrium-constant of the reaction between water and methyl acetate has been redetermined and is found to be 4.6 in the presence of hydrochloric acid not exceeding N in strength. (4) The variation in velocity-constant with initial conditions can be explained if the reacting substance is considered to be an ester-water additive complex the degree of dissociation of which decreases as the water-content decreases that is the water acts stoicheiometrically and at- the same time as a negative catalyst in virtue of its high dissociating power. MUSPRATT LABORATORY OP PHYSICAL AND ELECTRO-CHEMISTRY. UNIVERSITY OF LIVERPOOL. [Received November 12th 1916.

ISSN:0368-1645    

DOI:10.1039/CT9160900067

出版商:RSC    

年代:1916

数据来源: RSC

摘要:

PROPAGATION OF FLAME IN HYDROGEN AND AIR. 83 VII.-“ The Propugation of Flame in Mixtures of By WILLIAM ARTHUR HAWARD and TATSUROTAGAWA. THE work described in this paper was carried out at the Home Office Experimental Station under the direction of Dr. R. V. Wheeler and is a revision of Mallard and Le Chatelier’s work on the same subject (Ann. des ICfines 1883 [viii] 4 314). Such a revision seemed necessary in the light of Wheeler’s recent work on the propagation of flame in mixtures of methanel and air (T. 1914, 105 2606) which shows considerable divergence from Mallard and Le Chatelier’s results (Zoc. cit. p. 324). Mixtures of hydrogen and air differ in their beliaviour from mixtures of methane and air when ignited in horizontal tubes, inasmuch as in the former case the flame travels more rapidly and in some mixtures of hydrogen and air the explosion wave may be developed after the flame has travelled a short distance (about 2 metres) from the open end of the tube.The phenomena that precede the development of the explosion wave o r which occur when the explosion wave is not set up are however similar t o those that obtain with all mixtures of methane and air. Thus when the mixture is ignited a t the open end of a horizontal tube which is closed a t the other end a “ uniform movement ’’ of the flame is set up which is succeeded by a “vibratory movement.” The present paper deals solely with the uniform movement. The unifolrm movement of flame defined by Le Chatelier (“Leqons sur le Carbone,” Paris 1908 p. 273) as resulting from the normal transference of heat from layer t o layer of the mixture Hydrogen and Air.The ‘ Uniform Movement.’ ’ 84 HAWARD AND OTAGAWA “THE PROPAGATION OF by radiation and conduction is according to Wheeler (see Bone’s Presidential Address t o Section B Brit’ish Association 1915) a limited phenomenon obtainable only in tubes within a certain range of diameter large enough to prevent appreciable cooling by the walls but narrow enough t o suppress the influence of convec- tion currents. The diameter of the tube requisite t o ensure that cooling by the walls shall not affect the speed of flame is according t o Mallard and Le Chat’elier “d’autant plus grand que la vitesse de propaga- tion de la flamme est plus faiblel” (Zoc. c i t . p. 320). According t o the following series of experiments madel by them with a mixture of hydrogen and air containing 30 pelr cent.of the combustible gas, in which the speed of propagation of flame is fairly rapid a diameter of 3 mm. should suffice for this purpose. Diameter of tube. Speed of uniform movement. 10 mm. 350 cm. per see. 6 9 9 325 9 ) 9 9 3 9 350 3 ) 9 ) We have made three series of experiments with a number of mixtures of hydrogen and air ranging from the lower t o the upper limit of inflammability using tubes of 9 11.5 and 25 mm. in diameter respectively. We are1 unable t o confirm the said conclu- siQn that the speed of the flame in a given mixture is independent of the diameter of the tube above 3 mm. On the contrary over a great part of the range of infla-mmability of such mixtures the speeds are some 25 t o 30 cm.per second faster in tubes 11.5 mm. than in tubes 9 mm. in diameter; whilst in tubes 25 mm. in diameter the speeds .ue between 75 and 100 cm. per second faster than in a tube of 9 mm. diameter over two-thirds of the range1 of inflammability. I n their series of determinations of the speed of flame in different mixtures of hydrogen and air Mallard and Le Chatelier used a tube 10 mm. in diameter the length of which was varied and they measured the velocity over the’ first metre. They thereby included in most of their measurements (as they themselves subsequently realised) part of the ‘‘ vibratory movement,” since th0 uniform movement does not in the majority of mixtures of hydrogen and air extend over so great a distance as 1 metre.Their general conclusions regarding the variation in the speed of flame with the hydrogen content of the mixture were as follows: (1) The maximum speed is not obtained with the mixture con- taining hydrogen and oxygen in combining proportions (29.5 per cent. of hydrogen) but with a mixture containing 40 per cent. of hydrogen FLAME IN MIXTURES OF EIYDROGEN AND AIR.” 85 (2) On either side of tlie maximum the speeds vary proportion- ately with the hydrogen contents of thel mixtures. W e can in partl confirm the first of these statements. The maximum speed of propagation of flame according to our deter- minations is obtained with mixtures over the range 38-45 per cent. of hydrogen. The explanation advanced by Le Chat’elier (“Leqons sur le Carbone,” p.279) is that the rate of transmission of heatl from thel burnt. to the unburnt layer of mixture (and, 3u-o t u i ~ t o the rate of propagation of flame) depends not only on the temperature of combustion of the mixture but on its thermal conductivity. The thermal conductivity of hydrogen is six times that of air and it may well be that with mixtures containing more than one-third of their volume of hydrogen the enhanced conductivity of the1 mixture more than compensates for its lower heating value. I n the case of mixtures of methane and air the range of inflammability of which lies between 5.4 and 14.3 per cent. of methane Wheeler (Zoc. cit. p. 2608) has shown that no such effect is noticeable. The thermal conductivity of methane is 6-47 x compared with air 5-22? x 10-5 and with hydrogen, Our experiments do not however bear out Mallard and Le Chatelier’s second statemeInt.As already stated tlie maximum speed is obtained with mixtures containing between 38 and 45 per cent. of hydrogen and not with one particular mixture. Le Chatelier in further comment on this subject (“ Le Carbone,” p. 279) states that “ la courbe repr6sentant les vitesses en fonction de la proportion de gaz combustible est compos6e de deux droites, se coup?nt ii angles vifs pour la proportion de 40 p. 100 hydrogBne, avec unO vitesse maximum del 4 met. 37 cm. par seconde.” How- ever in the figure (Fig. 36 p. 278) illustrating this statement he adopts 485 instead of 437 cm. per sec. as the velocity for a mixture containing 40 per cent.of hydrogen.* I n Fig. 1 are plotted all the determinations recorded by Mallard and Le Chatelier and the curve (two straight lines) deduced by them therefrom; as well as f o r purposes of comparison a curve derived from one series of our experiments with a tube 9 mm. in diameter. I n dotted line is shown the upper portion of their curve as i t appears when the figure given in the text and in the table of experiments (437 cm. per sec.) is taken for the velocity of the mixture1 containing 40 per cent. of hydrogen. The entire 31.9 x 10-5. . * It may be noted that the same figure 437 em. per sec. is given by Mallard and Le Chatelier in the account of their researches in the Ann. des N i n e s , (loc. cit.) whereas here also the curve is drawn taking 485 em.per see. as the maximum speed 86 HAWARD AND OTAGAWA (' THE PROPAGATION OF curve now conforms more nearly with that indicated by our ow11 experiments. It will be seen that the central portion of our curve is parabolic whilst the extremities show a tendency to become horizontal. Although the latter tendency is not so marked as with mixtures of methane and air it is sufficient to render unjustifiable Mallard and Le Chatelier's prolongation of either limb of the curve to the zero velocity ordinate t o determine the limit's of inflammability, and accordingly their figures of 6 per cent. and 80 per cent. of hydrogen in air as the' lower and upper limits of inflammability FIU. 1. I I I I I 0 10 20 30 40 50 60 70 Hydrogen in air per cent.respectively must be regarded as erroneous. We have made no attempt to determine accurately the limits for horizontal propaga- tion of flame in mixtures of hydrogen and air but are satisfied that in a tubre 9 mm. in diameter flame does not travel in mixtures a t rormal temperature and pressure containing 11.8 per cent. o r less or 63.5 per cent. or more of hydrogen. An examination of the curves derived from our experiments, shown in Fig. 2 indicates that whereas an increase in the diameter of the tube in which the flame is travelling enhances the speed for many mixtures mixtures rich in hydrogen are not appreciably affect,ed. Thus with mixtures containing up to 35 per cent. of hydrogen the speeds in an 11.5 mm. tube are faster than in FLAME IN MIXTURES OF HYDROGEN AND AIR.” 87 9 mm.tube; whilst with a 25 mm. tube’ the speeds in mixtures containing up to 54 per cent. of hydrogen are considerably faster than those obtained in either a 9 mm. tube or in an 11.5 mm. tube; but the speeds in the richer mixtures are almost identical in the three tubes. These results seem to indicate that the influence of cooling by the walls of the tubes on the speeds of the flames becomes nearly negligible when the thermal conductivity of the1 gaseous mixture is high. EXPERIMENTAL. The method of experiment and the electrical means of recording the speeds were similar to those employed by Wheeler f o r mixtures of methane and air (Zoc. cit. p. 2609). Glass tubes of three difTereiit diameters were used €or separate Fro. 2. 0 10 20 30 40 50 60 70 Hydrogen in air per cent.series of experiments each about 1.5 metres long closed a t one end by a tap and connected a t the other t o a gas-holder by rubber tubing which could readily be removed when i t was desired to ignite the mixture leaving the end open. The speeds were determined between two points 40 cm. apart, the first point being 10 cm. from the point of ignition which was the open end of the tube. Ignition was effected by a lighted taper held to the open end as soon as the rubber tubing connecting with the gas-holder had been removed. The hydrogen used was obtained (from the British Oxygen Com- pany) compressed in cylinders and was of 99.5 per cent. purity. The mixtures were made in iron gas-holders of 70 litres capacity 88 PROPAGATION OF FLAME IN HYDROGEN AND AIR.over water rendered slightly alkaline by potassium hydroxide. The mixture t o be experimented with which was analysed before use, was allowed t o flow slowly through the explosion tube to displace air six times the volume of the tube being used. The initial temperature of the mixture was t h a t of the room (15-20°) and it was saturated with water-vapour atl t h a t temperature. All t h e determinations of speeds made are given in the tables t h a t follow. TABLE I. Internal diamet'er of tabe 9 mm. Hydrogen in mixture. Per cent. 11.80 17.30 21.60 2 1-95 23.65 24.70 25.15 27-25 27-80 30-45 33.85 33.90 36-55 37-15 40.10 42-70 Hydrogen in mixture. Per cent. 17.30 22-70 25.15 27-25 27-80 33-90 37.15 40.10 43-10 Speed of uniform movement.Cm. per sec. No propagation. 124,126 166 175,176 173 194 197,203 206 231 221,223 237 260,264 273,278 312,321 321 370,373 391 360,368,400 388,394 394 384,412,412 415,424 424 394 395 416 Hydrogen in mixture. Per cent. 43-10 46.80 48-80 50.40 50-55 51.15 51.55 55-25 57.00 57.15 58.95 59.05 59.45 62.00 I 63-50 TABLE 11. Internal diamet,er of tube 11.5 mm. Speed of uniform movement. Cm. per see. 148 154 212 218 245 265,267 280 287 299 294 383 390 415 400 42 1 411 435 Speed of uniform movement. Cm. per sec. 413 420 433 393 394 412 379 387 352 365 369 353 354 354 356 346 353 299 328 333 275 282 259 277 223 224 208 216 172 150 157 162 No propagation.Hydrogen in Speed of uniform mixture. movement. Per cent. Cm. per see. 46.80 398 419 48.80 370 410 51.15 348 350 51-55 341 352 382 55.25 303 309 319 57.15 261 280 58.95 175 180 59.45 173 17 1 POLYMORPHISM IN HALOGEE-SUBSTITUTED ANILIDES. 89 TABLE 111. Internal diameter of tube' 25 mrn. rdrogen in Speed of uniform mixture. movement . 'er cent. Cm. per see. 20.15 255 259 279 26.10 354,354 29.70 403,408 408 31.50 435,437,453 36.30 480 480 503 40.50 463 483 500 44.55 437 460 478 Hydrogen in mixture. Per cent. 47-00 47.70 49.15 54.80 60.25 61.60 Speed of uniform movement. Cm. per see. 430 430 411 433 433 380 383 398 309 317 320 153 162 162 139 149 150 We wish t o sxpress our t,hanka t o Mr.T. Mason of the Home Office Experimental Station who prepared and analysed the gas mixtures and also t o Dr. R. V. Wheeler and Professor W. A. Bone f o r helpful criticism of our results. DEPARTMENT OF CHEMICAL TECHNOLOGY, IMPERIAL COLLEGE OF SCIENCE DEPARTMENT OF MINING, TOHOKU IMPERIAL UNIVERSITY, LONDON. [Received November 20th 1915.1 AND TECHNOLOGY JAPAN PROPAGATION OF FLAME IN HYDROGEN AND AIR. 83 VII.-“ The Propugation of Flame in Mixtures of By WILLIAM ARTHUR HAWARD and TATSUROTAGAWA. THE work described in this paper was carried out at the Home Office Experimental Station under the direction of Dr. R. V. Wheeler and is a revision of Mallard and Le Chatelier’s work on the same subject (Ann.des ICfines 1883 [viii] 4 314). Such a revision seemed necessary in the light of Wheeler’s recent work on the propagation of flame in mixtures of methanel and air (T. 1914, 105 2606) which shows considerable divergence from Mallard and Le Chatelier’s results (Zoc. cit. p. 324). Mixtures of hydrogen and air differ in their beliaviour from mixtures of methane and air when ignited in horizontal tubes, inasmuch as in the former case the flame travels more rapidly and in some mixtures of hydrogen and air the explosion wave may be developed after the flame has travelled a short distance (about 2 metres) from the open end of the tube. The phenomena that precede the development of the explosion wave o r which occur when the explosion wave is not set up are however similar t o those that obtain with all mixtures of methane and air.Thus when the mixture is ignited a t the open end of a horizontal tube which is closed a t the other end a “ uniform movement ’’ of the flame is set up which is succeeded by a “vibratory movement.” The present paper deals solely with the uniform movement. The unifolrm movement of flame defined by Le Chatelier (“Leqons sur le Carbone,” Paris 1908 p. 273) as resulting from the normal transference of heat from layer t o layer of the mixture Hydrogen and Air. The ‘ Uniform Movement.’ ’ 84 HAWARD AND OTAGAWA “THE PROPAGATION OF by radiation and conduction is according to Wheeler (see Bone’s Presidential Address t o Section B Brit’ish Association 1915) a limited phenomenon obtainable only in tubes within a certain range of diameter large enough to prevent appreciable cooling by the walls but narrow enough t o suppress the influence of convec- tion currents.The diameter of the tube requisite t o ensure that cooling by the walls shall not affect the speed of flame is according t o Mallard and Le Chat’elier “d’autant plus grand que la vitesse de propaga- tion de la flamme est plus faiblel” (Zoc. c i t . p. 320). According t o the following series of experiments madel by them with a mixture of hydrogen and air containing 30 pelr cent. of the combustible gas, in which the speed of propagation of flame is fairly rapid a diameter of 3 mm. should suffice for this purpose. Diameter of tube. Speed of uniform movement. 10 mm. 350 cm. per see. 6 9 9 325 9 ) 9 9 3 9 350 3 ) 9 ) We have made three series of experiments with a number of mixtures of hydrogen and air ranging from the lower t o the upper limit of inflammability using tubes of 9 11.5 and 25 mm.in diameter respectively. We are1 unable t o confirm the said conclu- siQn that the speed of the flame in a given mixture is independent of the diameter of the tube above 3 mm. On the contrary over a great part of the range of infla-mmability of such mixtures the speeds are some 25 t o 30 cm. per second faster in tubes 11.5 mm. than in tubes 9 mm. in diameter; whilst in tubes 25 mm. in diameter the speeds .ue between 75 and 100 cm. per second faster than in a tube of 9 mm. diameter over two-thirds of the range1 of inflammability. I n their series of determinations of the speed of flame in different mixtures of hydrogen and air Mallard and Le Chatelier used a tube 10 mm.in diameter the length of which was varied and they measured the velocity over the’ first metre. They thereby included in most of their measurements (as they themselves subsequently realised) part of the ‘‘ vibratory movement,” since th0 uniform movement does not in the majority of mixtures of hydrogen and air extend over so great a distance as 1 metre. Their general conclusions regarding the variation in the speed of flame with the hydrogen content of the mixture were as follows: (1) The maximum speed is not obtained with the mixture con- taining hydrogen and oxygen in combining proportions (29.5 per cent. of hydrogen) but with a mixture containing 40 per cent.of hydrogen FLAME IN MIXTURES OF EIYDROGEN AND AIR.” 85 (2) On either side of tlie maximum the speeds vary proportion- ately with the hydrogen contents of thel mixtures. W e can in partl confirm the first of these statements. The maximum speed of propagation of flame according to our deter- minations is obtained with mixtures over the range 38-45 per cent. of hydrogen. The explanation advanced by Le Chat’elier (“Leqons sur le Carbone,” p. 279) is that the rate of transmission of heatl from thel burnt. to the unburnt layer of mixture (and, 3u-o t u i ~ t o the rate of propagation of flame) depends not only on the temperature of combustion of the mixture but on its thermal conductivity. The thermal conductivity of hydrogen is six times that of air and it may well be that with mixtures containing more than one-third of their volume of hydrogen the enhanced conductivity of the1 mixture more than compensates for its lower heating value.I n the case of mixtures of methane and air the range of inflammability of which lies between 5.4 and 14.3 per cent. of methane Wheeler (Zoc. cit. p. 2608) has shown that no such effect is noticeable. The thermal conductivity of methane is 6-47 x compared with air 5-22? x 10-5 and with hydrogen, Our experiments do not however bear out Mallard and Le Chatelier’s second statemeInt. As already stated tlie maximum speed is obtained with mixtures containing between 38 and 45 per cent. of hydrogen and not with one particular mixture. Le Chatelier in further comment on this subject (“ Le Carbone,” p.279) states that “ la courbe repr6sentant les vitesses en fonction de la proportion de gaz combustible est compos6e de deux droites, se coup?nt ii angles vifs pour la proportion de 40 p. 100 hydrogBne, avec unO vitesse maximum del 4 met. 37 cm. par seconde.” How- ever in the figure (Fig. 36 p. 278) illustrating this statement he adopts 485 instead of 437 cm. per sec. as the velocity for a mixture containing 40 per cent. of hydrogen.* I n Fig. 1 are plotted all the determinations recorded by Mallard and Le Chatelier and the curve (two straight lines) deduced by them therefrom; as well as f o r purposes of comparison a curve derived from one series of our experiments with a tube 9 mm. in diameter. I n dotted line is shown the upper portion of their curve as i t appears when the figure given in the text and in the table of experiments (437 cm.per sec.) is taken for the velocity of the mixture1 containing 40 per cent. of hydrogen. The entire 31.9 x 10-5. . * It may be noted that the same figure 437 em. per sec. is given by Mallard and Le Chatelier in the account of their researches in the Ann. des N i n e s , (loc. cit.) whereas here also the curve is drawn taking 485 em. per see. as the maximum speed 86 HAWARD AND OTAGAWA (' THE PROPAGATION OF curve now conforms more nearly with that indicated by our ow11 experiments. It will be seen that the central portion of our curve is parabolic whilst the extremities show a tendency to become horizontal. Although the latter tendency is not so marked as with mixtures of methane and air it is sufficient to render unjustifiable Mallard and Le Chatelier's prolongation of either limb of the curve to the zero velocity ordinate t o determine the limit's of inflammability, and accordingly their figures of 6 per cent.and 80 per cent. of hydrogen in air as the' lower and upper limits of inflammability FIU. 1. I I I I I 0 10 20 30 40 50 60 70 Hydrogen in air per cent. respectively must be regarded as erroneous. We have made no attempt to determine accurately the limits for horizontal propaga- tion of flame in mixtures of hydrogen and air but are satisfied that in a tubre 9 mm. in diameter flame does not travel in mixtures a t rormal temperature and pressure containing 11.8 per cent. o r less or 63.5 per cent.or more of hydrogen. An examination of the curves derived from our experiments, shown in Fig. 2 indicates that whereas an increase in the diameter of the tube in which the flame is travelling enhances the speed for many mixtures mixtures rich in hydrogen are not appreciably affect,ed. Thus with mixtures containing up to 35 per cent. of hydrogen the speeds in an 11.5 mm. tube are faster than in FLAME IN MIXTURES OF HYDROGEN AND AIR.” 87 9 mm. tube; whilst with a 25 mm. tube’ the speeds in mixtures containing up to 54 per cent. of hydrogen are considerably faster than those obtained in either a 9 mm. tube or in an 11.5 mm. tube; but the speeds in the richer mixtures are almost identical in the three tubes. These results seem to indicate that the influence of cooling by the walls of the tubes on the speeds of the flames becomes nearly negligible when the thermal conductivity of the1 gaseous mixture is high.EXPERIMENTAL. The method of experiment and the electrical means of recording the speeds were similar to those employed by Wheeler f o r mixtures of methane and air (Zoc. cit. p. 2609). Glass tubes of three difTereiit diameters were used €or separate Fro. 2. 0 10 20 30 40 50 60 70 Hydrogen in air per cent. series of experiments each about 1.5 metres long closed a t one end by a tap and connected a t the other t o a gas-holder by rubber tubing which could readily be removed when i t was desired to ignite the mixture leaving the end open. The speeds were determined between two points 40 cm.apart, the first point being 10 cm. from the point of ignition which was the open end of the tube. Ignition was effected by a lighted taper held to the open end as soon as the rubber tubing connecting with the gas-holder had been removed. The hydrogen used was obtained (from the British Oxygen Com- pany) compressed in cylinders and was of 99.5 per cent. purity. The mixtures were made in iron gas-holders of 70 litres capacity 88 PROPAGATION OF FLAME IN HYDROGEN AND AIR. over water rendered slightly alkaline by potassium hydroxide. The mixture t o be experimented with which was analysed before use, was allowed t o flow slowly through the explosion tube to displace air six times the volume of the tube being used. The initial temperature of the mixture was t h a t of the room (15-20°) and it was saturated with water-vapour atl t h a t temperature.All t h e determinations of speeds made are given in the tables t h a t follow. TABLE I. Internal diamet'er of tabe 9 mm. Hydrogen in mixture. Per cent. 11.80 17.30 21.60 2 1-95 23.65 24.70 25.15 27-25 27-80 30-45 33.85 33.90 36-55 37-15 40.10 42-70 Hydrogen in mixture. Per cent. 17.30 22-70 25.15 27-25 27-80 33-90 37.15 40.10 43-10 Speed of uniform movement. Cm. per sec. No propagation. 124,126 166 175,176 173 194 197,203 206 231 221,223 237 260,264 273,278 312,321 321 370,373 391 360,368,400 388,394 394 384,412,412 415,424 424 394 395 416 Hydrogen in mixture. Per cent. 43-10 46.80 48-80 50.40 50-55 51.15 51.55 55-25 57.00 57.15 58.95 59.05 59.45 62.00 I 63-50 TABLE 11.Internal diamet,er of tube 11.5 mm. Speed of uniform movement. Cm. per see. 148 154 212 218 245 265,267 280 287 299 294 383 390 415 400 42 1 411 435 Speed of uniform movement. Cm. per sec. 413 420 433 393 394 412 379 387 352 365 369 353 354 354 356 346 353 299 328 333 275 282 259 277 223 224 208 216 172 150 157 162 No propagation. Hydrogen in Speed of uniform mixture. movement. Per cent. Cm. per see. 46.80 398 419 48.80 370 410 51.15 348 350 51-55 341 352 382 55.25 303 309 319 57.15 261 280 58.95 175 180 59.45 173 17 1 POLYMORPHISM IN HALOGEE-SUBSTITUTED ANILIDES. 89 TABLE 111. Internal diameter of tube' 25 mrn. rdrogen in Speed of uniform mixture. movement . 'er cent. Cm. per see. 20.15 255 259 279 26.10 354,354 29.70 403,408 408 31.50 435,437,453 36.30 480 480 503 40.50 463 483 500 44.55 437 460 478 Hydrogen in mixture. Per cent. 47-00 47.70 49.15 54.80 60.25 61.60 Speed of uniform movement. Cm. per see. 430 430 411 433 433 380 383 398 309 317 320 153 162 162 139 149 150 We wish t o sxpress our t,hanka t o Mr. T. Mason of the Home Office Experimental Station who prepared and analysed the gas mixtures and also t o Dr. R. V. Wheeler and Professor W. A. Bone f o r helpful criticism of our results. DEPARTMENT OF CHEMICAL TECHNOLOGY, IMPERIAL COLLEGE OF SCIENCE DEPARTMENT OF MINING, TOHOKU IMPERIAL UNIVERSITY, LONDON. [Received November 20th 1915.1 AND TECHNOLOGY JAPAN

ISSN:0368-1645    

DOI:10.1039/CT9160900083

出版商:RSC    

年代:1916

数据来源: RSC

摘要:

POLYMORPHlSM IN HALOGEN-SUBSTITUTED ANILIDES. 89 VIII.-Polyrnoyphism Halogen-substitu ted A nilides. By FREDERICK DANIEL CHATTAWAY and GEORGE ROGER CLEMO. ALTHOUGH ability to crystallise in more than one form appears t o be a general property of the anilides few of the modifications, unstable in ordinary circumstances have been described. They are formed only under favourable conditions as a first stage in the process of crystallisation and then more or less rapidly redissolve and disappear ; their transformation theref ore unless accompanied by some colour change may easily pass unobserved. 1)escriptions of the phenomena attending the crystallisation of aceto-p-bromo- and aceto-2 4-dibromo-anilide (Chattaway and Lam- bert T. 1915 107 1766) and propiono-piodoanilide (Chattaway and Constable T.1914 105 126) have recently been published; the present paper gives an account of the similar behaviour of other derivatives of 2 4-dihalogen-substitukd anilines. Eoth aceto-2-chloro-4-bronio- and 4-chloro-2-bromo-aililide car1 exist in two modifications. When a saturated solution of either in glacial acetic acid is allowed to cool undisturbed an unstable needle-shaped modification a t first separates the long interlaced crystals com 90 CHATTAWAY AND CLEM0 : pletely filling the liquid. I n the crystalline mass compact crystals of a stable form soon make their appearance; the needle- shaped crystals dissolve and the compact crystals grow until the conversion is complete. If only a few crystals of the stable form appear they grow sIowly if the crystalline mass of the unstable form is left undishrbed; but if it is stirred the transformation is complete in a few seconds small crystals of the compact form falling as a crystalline precipitate.The transformation of thO unstable into the stable form of aceto-2-chloro-4-bromoanilide is much more sapid than the corre- sponding transformation of aceto-4-chloro-2-bromoanilide. The pnitrobenzoyl derivatives of the 2 4-dihalogen-substituted anilines also all crystallise in two forms. On cooling hot saturated alcoholic solutions they separate in slender needles or hairs which convert the liquid into a felt-like mass. After a time crystals of the' stable compact forms make their appearance the original crystals disappear and the new crystals grow until conversion is complete.Of the pnitrobenzanilides pnitrobenzo-2-chloro-4-bromo- anilide changes most rapidly ; if the liquid containing the hair-like form in suspension is shaken conversion is complete in a few minutes. In a series of comparative experiments in which similar amounts of solution were used and the felted masses of hair-like crystals were allowed t o transform undisturbed about thFee days were required in the cases of p-nitrobenzo-4-chloro-2-bromoanilide and of pnitrobenzo-2 4-dibromoanilid0 whilst in the case of pnitrobenzo-2 4-dichloroanilide several weeks elapsed before all the unstable form had disappeared. EXPERIMENTAL, I'repamtiot~ vf 2 4-BibromoadirLe. Tlie preparation of 2 4-dibromoaniline has always offered coil- siderable difficulty.The action of bromine on aniline dissolved in glacial acetic acid is so energetic that in practice it cannot be checked a t intermediate stages the substituted product even if a small quantity of bromine is used being mainly 2 4 6-tribromo- aniline o r if excess of bromine is used the stable perbromide of this compound C,H,Br3*NH,,HBr3. The action of bromine on acet- anilide is less vigorous and if a gram-molecule of bromine is added t o a gram-molecule of acetanilide dissolved in acetic acid in the presence of a gram-molecule of sodium acet'ate aceto-pbromoanilide is quantitatively and exclusively produced. A second atom of bromine can only be made t o enter the acetanilide with much greater difficulty some hours of action a t a high temperature bein POLYMORPHlSM IN HALOGEN-SUBSTITUTED ANILIDES.91 required. During the necessary heating if any water is present hydrolysis takes place so easily even in the presence of excess of sodium acetate that further action results in the production of 2 4 6-tribromoaniline7 just as if aniline itself had been employed. This however can be avoided and aceto-2 4-dibromoanilide can be quantitatively obtained i f precautions are taken to avoid the presence of hydrobromic acid and of water. The best procedure is as follows A gram-molecule of aceto-pbromoanilide * and a gram- molecule of fused and finely powdered sodium acetate are suspended in sufficient glacial acetic acid to make a thick paste. A gram- molecule of bromine dissolved in eight t o ten times its volume of glacial acetic acid is added slowly and the mixture heated for five t o six hours on a water-bath until the colour of bromine has almost disappeared.On diluting the' cooled product with water, aceto-2 4-dibromoanilide (m. p. 1 4 6 O ) separates and should be immediately filtered off and crystallised from alcohol. The aniline is obtained by dissolving the anilide in boiling alcohol mixed with about one-eighth of its bulk of concentrated hydrochloric acid, boiling the liquid under a reflux condenser for eight t o nine hours, distilling off the alcohol in a current of steam and adding a slight excess of sodium hydroxide to the cooled residue. The aniline separates as a white crystalline mass (m. p. 78-79O) the yield being about 90 per cent. of the quantity theoretically obtainable from the aceto-pbromoanilide used.The aniline may be obtained free from colour by distillation in a current of steam but this is somewhat tedious if large quantities are required and it is quite unnecessary if the base is to be used for subsequent preparations. Preparation of 2-Chloro-4-bromoaizililze and 4-Chloro-2-bromo- a d i n e . Tlie preparation of 2-cliloro-4-bromoaniline offers no difficulty as the corresponding aceto-2-chloro-4-bromoanilide is easily prepared either by the action of chlorine on aceto-pbromoanilide or of bromine on aceto-o-cliloroanilide. The former method is more convenient as pbromeacetanilide is so easily procured. The process is best carried out by passing the requisite amount of chlorine into a cooled sus- pension in glacial acetic acid of one gram-molecule of aceto-p-bromo- anilide and one gram-molecule of anhydrous sodium acetate.The action proceeds easily a t the ordinary temperature and after pre- cipitation by dilution and crystallisation from alcohol aceto-2- chloro-4-bromoanilide (m. p. 1 5 1 O ) is obtained in good yield. This * Acetaiiilide may be used employing double the quantity of sodium acetate and of bromine 98 CHATTAWAY AND CLEM0 : can then easily be converted into 2-chloro-4-bromoaniline (m. p. 7 3 O ) by ilie method previously described. I n order to effe'ct the entrance of a bromine1 atom into the ortho- position in aceto-pchloroanilide several hours' heating on a water- bat.h are required. The method which has been found most con- venient is similar to' that described for the preparation of aceto-2 4- dibromoanilide.The presence of hydrobromic acid and of water must be similarly avoided An excellent' yield of pure aceto-4- chloro-2-bromoanilide (m. p. 1 3 7 O ) is obtained in this way from aceto-pchloroanilide and from this 4-chloro-2-bromoaniline (m. p. 6 9 O ) can he obt'ained as above described. Polgmorplzic Forms of the A cetochlorobromoanilides. When a nearly saturated solution of aceto-2-chloro-4-bromoanilide in boiling glacial acetic acid is allowed to cool slowly no separation of crystals takes place as a rule until the liquid has reached a temperature not far removed from the ordinary; tufts of fine, needle-shaped crystals of an unstable modification then appear, often floating in the liquids and grow steadily; very soon compact crystals of the stable form make their appearance generally on the surface of the liquid and having reached a moderate size fall in a shower t o the bottom the tufts of the needle-shaped modification dissolving and disappearing.The transformation is usually so sapid that unless the cooling liquid is observed with a lens the tufts of the unstable crystals may easily escape notice. When a small quantity of such a hot saturated solution in glacial acetic acid is cooled rapidly in ice the unstable modification sepa- rates a t a few points in small tufts of very slender prisms which grow rapidly until a felted mass of interlaced fine crystals is formed. These if left undistarbed may remain for several days before crystalline nuclei of t'he stable compact rhombic form make their appearance and grow slowly but if a few crystals of the stable form are added and the mass is stirred transformation takes place very rapidly and in a few seconds the needledaped modification dissolves and disappears and is replaced by small crystals of the compact form which sink to the bottom of the liquid as a crystalline precipitate.An alcoholic solution of aceto-2-chloro-4-bromoanilide behaves similarly when cooled. Aceto-4-chloro-2-bromoanilide also crystallises in two polymorphic forms. Transformation of the unstable into the stable form is not so rapid as in the case of the isomeric compound. When a nearly saturated solution in hot glacial acetic acid is boiled for some time to destroy all crystalline nuclei and then allowecl to cool slowly the unstable modification separates in cluster POLYMORPHISM IN HALOGEN-SUBSTITUTED ANILIDES.93 of slender needles which completely fill the liquid; these if left undisturbed may remain without change for days but sooner or later crystals of the compact stable modification make their appear- ance and grow slowly to a considerable size the unstable form dissolving and disappearing after some hours. If a few crystals of the stable modification are introduced into the felted mass of needle-shaped crystals and the whole is vigorously stirred transformation is complete in a few minutes small crystals of the compact form being the product. A number of derivatives of the 2 4-dihalogen-substituted anilines, not hitherto described have been prepared in order t o ascertain which can easily be obtained in two' forms.The butyranilides were prepared by heating butyric anhydride with equivalent amounts of the various anilines. n-Bu tyro-2-chloro-4-brorn oanilide, C,H,ClBr~NH*CO.CH,*CH,*CH, is very readily soluble in alcohol but less readily so in light pet(ro1eum. It crystallises in long colourless needle,. melting at 1 loo ; extinction 21° : 0.2234 gave 0-2682 AgCl+ AgBr. n-Bu tyro-4-chZoro-2-bromoanilide is very readily soluble in alcohol, but less readily so in a mixture of light pekroleum and benzene. It crystallises in colourless needles melting a t 1 1 1 . 5 O ; extinction 31° : C1= 12.85 ; Br = 28.97. CloHllONCIBr requires C1= 12.82 ; Br= 28.90 per cent.0.2275 gave 0.2730 AgCl + AgBr. e l = 12.84; Br = 28.96. C,oH,lON@lBr requires C1= 12.82 ; B r = 28.90 per cent. The phenylacet'anilides were prepared by the interaction of equi- valent quantities of phenylacetyl chloride and the aniline dissolved in ether in the presence of an equivalent weight of pyridine. Yl~etzylaceto-2-c~loro-4-bronioan~lide, C6H3C1Br*NH*CO*CH,*C,H,, crystallises from boiling alcohol in which it is readily soluble in colourless needles melting a t 150° ; extinction variable : 0.2187 gave 0.2233 AgCl + AgBr. C14Hl10NC1Br requires C1= 10.92 ; B r = 24-63 per cent. Phenylaceto-4-chlo~o-2-bromoanilide crystallises from boiling alcohol in which it is readily soluble in long colourlms needles melting a t 1 4 8 O : C1= 10.93 ; Br = 24-64.0.1536 gave 0.1568 AgCl+ AgBr. Phenylaceto - 2 4 - dibromoanilide, C1= 10.92 ; Br= 24.64. C14Hl,0NC1Br requires C1= 10.92 ; Br= 24.63 per cent. C,H3Br,*NH*CO*CH,*CH~, crystallises from alcohol in long colourless needles melting a t 160° 94 CHATTAWAY AND CLEM0 : 0.1520 gave 0.1548 AgBr. C,,H,,0NBr2 requires Br =43-32 per cent. The nitrobenzo-2 4-dihalogen-substituted anilides were prepared by adding the nitrobenzoyl chlorides to equivalent amounts of the anilines dissolved in ether in the presence of pyridine or of a concentrated solution of sodium carbonate. Br = 43.34. o-Nit ro b enxo-2-chloro-4-bromoanilide, C,H,ClBr*NH*CO*C,H,-NO,, is fairly readily soluble in boiling alcohol from which i t crystallises in colourless needles melting a t 165O; extinction 25O : 0.1788 gave 0.1663 AgCl+ AgBr.C1= 9.96 ; Br = 22.44. o-Ni t ro b en co-2 4-dib ro moanilide, C,,Hs0,N2ClBr requires C1= 9.97 ; Br = 22.48 per cent. @,H,Br,~NH~CO*C,H,-NO , is readily soluble in boiling glacial acetic acid but less readily so in boiling alcohol than the compound just described. It crystal- lises from alcohol in pale yellow slightly oblique plates melting a t 178O; extinction 24O : 0.2060 gave 0.1937 AgBr. Br=40.01. C,,H,O,N,Br require's Br = 39.96 per cent. o-Nitrob enzo-4-chloro-2-bromoanilicFe is fairly readily soluble in boiling alcohol and crystallises in colourless slightly oblique plates melting a t 166O; extinction 18O : 0.2153 gave 0.2005 AgCl + AgBr. o-;jTitro b enzo-2 4-dici~loroal?ii1ide, C1= 9.97 ; Br = 22.47.Cl3H,O3N2C1Br requires c1= 9-97 ; Br = 22.48 per cent. ~6H&1,*NH*CO*C6H,* NO,, is readily soluble in boiling alcohol from which it crystallises in colourless oblique platep melting a t 153.5O : 0.2299 gave 0.2116 AgCl. C1=22*77. C,,H,O,N,~ requires c1= 22.80 per cent. m-Nitrob er~zo-2-chloro-4-brol?zoanilide is sparingly soluble in boil- ing alcohol but readily so in boiling glacial acetic acid from which it crystallises in long colourless needles melting at 191O : 0.1971 gave 0.1830 AgCl+ AgBr. C1,H,03N,ClBr requires C1= 9-97 ; Br = 22.48 per cent. m-Nitrobemo-2 4-dibromoanilide is moderately readily soluble in both alcohol and glacial acetic acid and crystallises in colourless needles melting a t 165O ; extinction variable : C1= 9.94 ; Br = 22.41.0.1990 gave 0.1860 AgBr. m-Nitrobenzo-4-chEoro-2-bromoanilide is readily soluble in both Br = 39-77. C,,Hs0,N2Br2 requires Br = 39.96 per cent POLYMORPHISM I N HALOGEN-SUBSTITUTED ANILIDES. 95 alcohol and glacial acetic acid. It crystallises from acetic acid in colourless flattened plates melting a t 167.5* which in collvergent light show a positive biaxial figure the axial angle being extremely small : 0.2149 gave 0.2002 AgCl+ AgBr. C1= 9.97 ; Br = 22.48. C,,H,03N,C1Br requires C1= 9-97' ; Br = 22.48 per cent. m-Nitrob enzo-2 4-dichloroanilide crystallises from glacial acetic acid in colourless flattened plates melting a t 183O; straight extinction. Normal to plate emerges the acute bisectrix of a posi- tive biaxial figure'; axial angle medium : 0.2126 gave 0.1957 AgC1.C1= 22.77. C,,H,0,N2Cl requires C1= 22.80 per cent. p-iVitrob etzzo-2-chloro-4-bri,moanilide is moderately readily soluble in boiling glacial acetic acid but sparingly so in boiling alcohol. On rapidly cooling a concentrated boiling alcoholic solution it separah in tufts of very fine almost colourless hair-like1 crystals, which grow together to a felt-like mass. Among these interlaced, hair-like crystals six-sided pale yellow plates with straight extinc- tion almost a t once make their appearance and rapidly grow a t the e'xpense of the hair-like form which dissolves and disap- pears. I f the pulpy mass of fine crystals is shaken the conversion is complete in a few minutes the compact form subsiding as a sandy crystalline powder melting a t 199O : 0.2265 gave 0.2109 AgCl + AgBr.p-Nitrobenzo-2 4-dibromoanilide is somewhat readily soluble in boiling glacial acetic acid but lew readily so in boiling alcohol. On rapidly cooling a boiling saturated alcoholic solution it separates in very fine hair-like crystals which on account of their small size appear almost colourless. On allowing these to remain in the solvent a t the ordinary temperature small pale yellow six- sided compact plates with straight extinction seen under the micro- scope to be stout very much flattened prisms with domed ends, make their appearance and grow a t the expense of the hair-like form which gradually dissolves and disappears. On allowing the felted mass of the hair-like form t o remain undisturbed a t the ordinary hmperature in a small flask the conversion was nearly complete in about twelve hours and all crystals of the first form had disappeared a t the end of three days.Both forms melt a t 194O : C1= 9.96 ; Br = 22.47. C,,H,O,N,ClBr requires C1= 9-97 ; Br = 22.48 per cent. 0.2230 gave 0.2092 AgBr. Br = 39-92. p-Nitrob enzo-4-chloro-2-bronloanilide is readily soluble in boiling C,,H,03N,Br2 requires Br = 39.96 per cent 96 CHATTAWAY AND CLEM0 : glacial acetic acid but less readily so in boiling alcohol. On cooling a boiling saturated alcoholic solution i t separates in very fine almost colourless needles with straight extinction which interlace and finally completely fill the liquid. After some time pale yellow stout, six-sided rhombic plates (82.) with diagonal extinction make their appearance and slowly grow a t the expense of the first form cutting their way down through the pulp-like crystalline mass in a very characteristic manner.The conversion is complete after about three days. Both forms melt a t 1 7 4 O : 0.2329 gave 0.21 66 AgCl + AgBr. p-&-itrob enzo-2 4dichZoronnilide is readily soluble in boiling glacial acetic acid but less readily so in boiling alcohol. On cooling a boiling saturated alcoholic solution it separates in very fine almost colourless needles. After some hours small four-sided pale yellow plat,es make their appearance and slowly grow cutting their way down through the interlaced crystals of the first form which dissolve and disappear in the neighbourhood of the growing crystals.These compact plates when seen under the microscope appear as faces of stout six-sided prisms with domed ends. This transformation is slow; thus in one experiment in which a small flask half filled with a pulp of the needle-shaped form was allowed to remain undis- turbed a t the ordinary temperature three weeks elapsed before all the unstable form had disappeared. The length of time taken in transforming naturally depends on the amount t o be transformed, as well as on the temperature and on the nature of the solvent. Both forms melt a t 174O the unstable form transforming before the melting point is reached: 0.1 725 gave 0.1574 AgCl. C1= 22.56. C,,H,O,N,Cl requires C1= 22.80 per cent. The 2 4-disubstituted phthalanils are easily obtained by heating together equivalent quantities of the anilines and phthalic anhy- dride for several hours to about 180°.>N*C6H,c1Br is f air1 y readily soluble in boiling alcohol and readily so in boiling glacial acetic acid from which it crystallises in colourless prisms melting a t 165O; straight extinction : C1= 9-95 ; Br = 22.44. CI3H,O3N,C1Br requires C1= 9.97 ; Br = 22.48 per cent. CO co Y h t ha 20-2-chl oro-4- b rom o a d C,H < 0.2120 gave 0.2094 AgCl + AgBr. C,,H,O,NClBr requires C1= 10.54 ; Br= 23.75 per cent. Ph thalo4-chloro-2-bromoanil is readily soluble in boiling glacial acetic acid and in boiling alcohol fxom either of which i t crystal- lises in colourless prisms with straight extinction melting a t 140O: C1= 10.57 ; Br = 23.84 POLYMORPHISM IN HALOGEX-SUBSTITUTED ANILIDES.97 0.2299 gave 0.2267 AgCl + AgBr. C1= 10.56 ; Br = 23.79. C,,H702NC1Br requires C1= 1 0.54 ; Br = 23.75 per cent. Phthalo-2 4-dichloroard C?6H4<CO>N*C6H3c12 co is fairly readily soluble in-boiling alcohol and readily so in boiling glacial acetic acid. It crystallises from either in colourless prisms melting a t 155O ; straight extinction : 0.2342 gave 0.2302 AgCl. C,,H702NC12 requires C1= 24-28 pOr cent. Phthalo-2 4-dibromoanil is fairly readily soluble in boiling alcohol and readily so in boiling glacial acetic acid from which i t crystallises in colourless prisms melting a t 153*5O ; straight extinc- tion : C1= 24.31. 0.2198 gave 0.2171 AgBr. Br=42'03. C,,H70,NBr2 requires Br = 41.97 per cent. The methyl and ethyl dihalogen-substituted carbanilates are easily prepared by dissolving equivalent quantities of the corre- sponding aniline and of pyridine in dry ether and adding very slowly as the action is violent the equivalent amount of methyl or ethyl chloroformate.The ether is then distilled off and the solid product well washed with dilute hydrochloric acid to remove the pyridine and crystallised from alcohol. Jf ethyl 2-c hloro-4-bromocclr bandat e C6$13~Br*NH*C0,Me cxys- tallises from alcohol in which i t is very readily soluble in colourless needles melting a t 76.5O; extinction 4 2 O : 02035 gave 0.2524 AgCl + AgBs. C1= 13-28 ; Br = 29.93. C,H70,NC1Br requires C1= 13-40 ; Br = 30.22 per cent. Ethyl 2-c hloro-4- bromoclxr banila t e C,H3C1Br*NH-C02Et is readily soluble in alcohol from which it crystallises in colourless needles melting a t 96O; extinction 38O : 0.2012 gave 0.2380 AgCl + AgBr.C1= 12.66 ; Br= 28-55. CgHg02NC1Br requires C1= 12.73 ; Br = 28-70 per cent. Methyl 2 4-dibromocarbanilate C,H,Br,*NH*CO,Me has been obtained by Hentschel (J. pr. C'hem. [ii] 1886 34 423) by direct bromination and by Fromm and Heyder (Ber. 1909 42 3801) by the action of bromine on phenylthiocarbimide in the presence of aqueous methyl alcohol. It is easily prepared by the general method from 2 4-dibromoaniline and crystallises from alcohol in which it is readily soluble in colourless needles melting a t 97O. Hentschel gives the melting point as 96'5O. (Found Br=51.80. Calc. Br=51*74 per cent.) Ethyl 2 4-dibromocarbanilate C6H3Br2*NH*C0,Et was prepared by Fromm and Heyder (Zoc.cit.) by the action of bromine on phenyl- VOL. CIX. 98 CHATTAWAY AND CLEM0 : thiocarbimide in the presence of aqueous ethyl alcohol. It is easily prepared by the above method. It crystallises from alcohol in which it is readily soluble in colourless needles melting a t lolo, with extinction 40°. (Found Br = 49.58. Calc. Br = 49.49 per cent.) illethyl 4-chloro-2-bromocarbanilat e is very readily soluble in alcohol from which it crystallises in colourless needles melting a t 87.5O : 0.1146 gave 0.1440 Ag@l+ AgBr. C1= 13-45 ; Br= 30.32. Ethyl 4-chloro-2-bromocarbanilate crystallises from alcohol in 0.1510 gave 0.1760 AgCl+ AgBr. Cl= 12.48 ; Br=28-13. CgHgO,NBrC1 requires C1= 12.73 ; Br = 28.70 per cent.Methyl 2 4-dichloro’carbaniZate C,H,C?l,*NH*C’O,Me is ex- tremely readily soluble in alcohol from which it crystallises in colourless needles melting a t 70’5O : C8H702NB&1 requires C1= 13-40 ; Br = 30.22 per cent. which it dissolves very readily in colourless needles melting a t 90° : 0.1113 gave 0.1450 AgCl. CI=32*22. C8H70,NCl requires c1= 32.23 per cent. Ethyl 2 4-dichlorocarbaniZate C,H,C~,-~H*CO,Et crystallises from alcohol in which i t is readily soluble in colourless needles melting atl 89O: 0.1121 gave 0.1443 AgCl. C1=30.64. CgHgO,NC1 requires C1= 30.30 per cent. Lik0 other substituted anilines 2-chloro-4-bromoaniline and 4-chloro-2-bromoaniline react readily with carbamide a t a somewhat elevated temperature ammonia is liberated and the monob and s-di-dihalogen-substituted carbamides are produced.ThO latter compounds are formed in largest amount even when a considerable excess of carbamide is used and me easily isolated owing to their very sparing solubility in ordinary organic solvents. A yield of the s-tetra-substituted carbamide amounting to about 80 per cent. of the theoretical is obtained when a gram-molecule of the aniline is heated with 4 gram-molecules of carbamide to 180° for forty hours. 9-2 2r-Dic hloro-4 4f -dib r omodipheny lcarbamide, C6H3ClBr*T”H CO*NR-C,H,ClBr, is very sparingly soluble in boiling glacial aoetic acid. It is much more readily soluble in hot nitrobenzene from which it separates in colourless very fine hair-like needles with oblique extinction, melting at 279O: 0-1420 gave 0.2133 AgCl+ AgBr.C1= 16-08 ; Br = 36.25. C,3H,0N,CI,Br requires C1= 16.16 ; Br = 36.42 per cent POLYMORPHISM IN HALOGEN-SUBSTITUTED ANILIDES. 99 s - 4 4/-Bichloro-2 2'-dibromodiphenylcar6am~e resembles the isomeric compound very closely. It crystallises from hot nitro- benzene in colourlel;Es fine needles melting a t 274O; extinction 40' : 0.1988 gave 0.2990 AgCl + AgBr. C1= 16-10 ; Br = 36.29. C13HsON2C12Br2 requires C1= 16-16 ; Br = 36.42 per cent. All thO dihalogen-substituted anilines react readily with diethyl oxalate a t a suitably high temperature; using excess of ester the corresponding oxanilide is the chief compound; using excess of aniline the oxanilide predominates. They are easily separated on account of the very sparing solubility of the tetra-substituted uxan- ilides.The 2 4-&halogen derivative are k t prepared by heating 1 gram-molecule of the aniline with 5 gram-molecules of diethyl oxalate to about 220' f o r an hour. On cooling and adding a little alcohol both the oxanilate and the tetra-substituted oxanilide separate out. A further quantity of the two compounds can be obtained by distilling off the alcohol from the filtrate and reheating the residue. The two compounds are readily separated by boiling with alcohol, when the oxanilate is dissolved. The yield of the tetrahalogen- substituted oxanilide is always very small. The disubstituted obxanilic acids are be& prepared by suspending the corresponding ethyl disubstituted oxanilates in about twenty times their weight of water adding rather mom than the equivalent of sodium carbonate and passing steam until the este.rs have dis- appeared.The sodium salts of the acids crystallise out on cooling. On adding an excess of hydrochloric acid to hot aqueous solu- tions of the salts and cooling the acids separate. They crystal- liw from water with one molecule of water of crystallisation. When the monohydrated acids are dehydrated and recrystallised from toluene or benzene the anhydrous acids are obtained. The disubstituted oxanilamides separate as sparingly soluble, crystalline solids when dry gaseous ammonia is passed into warm alcoholic solutions of the corresponding ethyl oxanilahs. Ethyl 2-cltEoro-4-bromo-oxanitate C,H3ClBr*NH*CO*C0,Et is readily soluble in boiling alcohol from which it crystallises in colourless needles melting a t 124O : 0.1700 gave 0.1828 AgCl+ AgBr.C1= 11.51 ; Br=25*95. ClOH,O3NC1Br requires C1= 11.57 ; Br = 26.08 per cent. 2-Chloro-4-bromo-oxa:aniEic acid CGH3CIBr*NH*CO*C0,H crystal- lises in colourless needles melting and decomposing a t 131O: 0.1115 gave 0-1332 AgCl+ AgBr. C1= 12-79 ; Br = 28.83. CsH,O3NC1Br requires C1= 12.73 ; Br= 28.70 per cent. E 100 CHATTAWAY AND CLEM0 : s-2 2/-Dichloro-4 4/-dibromo-oxanilide, C6H3C1Br *NH*CO*CO*NH*C,H,ClBr, is spasingly soluble in boiling alcohol o r glacial acetic acid but dis- solves readily in hot nitrobenzene from which it crystallises in small colourless needlep melting a t 285O : 0.0727 gave 0.1026 AgCl+ AgBr. C1= 15-11 ; Br = 34.06.C,,H80,N2C12Br2 needs C1= 15.19 ; Br = 34.23 per cent. 2 - Chloro - 4 - bromo-oxardamide C6H3C1Br*NH*CO*CO*NH2 is modesately readily soluble in boiling alcohol and readily so in hot nitrobenzene. It crystallises in colourless needles melting a t 243O : 0.0905 gave 0-1079 AgCl + AgBr. C1= 12-76 ; Br= 28-77. Ethyl 4-chloro-2-bromo-oxanilate crystallises from alcohol in 0.1932 gave 0.2087 AgC1+ AgBr. C1= 11.56 ; Br = 26-07, 4-Chloro-2-bromo-oxa~~ilic acid crystallises from boiling benzene in When crystallised from water it separates in colourless needles 0.1.177 gave 0.1320 AgClSAgBr. C1=12*00; Br=27*06. s-4 4f-Dichloro-2 2r-dibrorno-oxanilide is very sparingly soluble in boiling glacial acetic acid. It crystallises from hot nitrobenzene in which it is fairly readily soluble in small colourless needles melt- ing a t 295O: (&H602N,C1Br requires c1= 12.78 ; Br = 28-80 per cent.which it is readily soluble in colourless needles melting a t 121O: C,,H,03NC1Br requires C1= 11.57 ; Br = 26.08 per cent. colo8urless needles melting with decomposition a t 126-127O : containing one molecule of water of crystallisation : C8H,O3NC1Br,I~,O requires C1= 11.96 ; Br = 26.96 per cent. 0.1531 gave 0.2159 AgCl+ AgBr. C1= 15.10; Br = 34-03. 4-Chloro-2-bromo-oxanilamide crystallises from alcohol in colour- 0.1658 gave 0.1963 AgCl+AgBr. C1=12*68; Br=28*57. C8H,02N2ClBr requires C1= 12-78 ; Br = 28.80 per cent. Ethyl 2 4-dibromo-oxanilate C6H,Brz=NH*CO*C0,Et is readily soluble in hot alcohol from which it separates in colourless curved, hair-like crystals melting a t 130° : C1,H,02N,C12Br2 requires C1= 15.19 ; B r = 34.23 per cent.less needles melting a t 236O : 0.1750 gave 0.1870 AgBr. Br=45.47. C,,H,O,NBr requires Br = 45.51 per cent. 2 4-Bibromo-oxanilic acid C,H,Br,*NR*CO*CO,H crystallises from boiling benzene in colourless needles melting and decomposing a t 138O into carbon dioxide and 2 4-dibromoformanilide. It crystal- lises from boiling water in which it is readily soluble in slender POLYMORPHISM IN HALOGEN-SUBSTITUTED AEILIDES. 101 colourlm pxisms containing one molecule of water of crystallisa- tion which is slowly lost a t 100': 0.1711 gave 0.1898 AgBr. Br=47.20. 2 4 2' 4'-Tetrabronzo-oxanilide, C,H,03NBr2,H,0 requires Br = 46.88 per cent. C,H3Br2*NH*CO*CO*NH*C6H3Br2, is very sparingly soluble in boiling glacial acetic acid 150 C.C.dis- solving only about 0.1 gram. It is moderately readily soluble! in boiling nitrobenzene from which it crystallises in colourless needles melting a t 298O ; straight extinction : 0.2955 gave 0.3984 AgBr. Br = 57-37. 2 4-Dibromo-oxaniZamide C,H3Br,*NH*CO*CO*NH3[, crystallises 0.2604 gave 0.3018 AgBr. Br = 49.32. C8~,02N,Br requires Br = 49.66 per cent. It has been shown (Chattaway and Mason T. 1910 97 339) that diethyl malonate a t its boiling point reacts with the halogen- substituted anilines halogen derivatives of malonanilide and of ethyl malonanilate being produced. 'These can be separated owing to the sparing solubility of the substituted malonanilides and the ready solubility of the ethyl malonanilates in all ordinary solvents.The halogen-substituted malonanilic acids are also very easily obtained by hydrolysing the substituted malonanilic esters by suspending them in a dilute solution of sodium carbonate and passing steam through the liquid until the ester disappears. After concentrating th0 solutions and adding a slight excess of hydro- chloric acid the acids separate in a crystalline state. When heated, the malonanilic acids decompose quantitatively into carbon dioxide and the corresponding substituted acetanilide. The malonanilamides are obtained by passing ammonia into alcoholic solutions of the corresponding ethyl malonanilates. cI4II8o2N2Br4 repires Br = 57.52 per cent. from alcohol in colourless neledles melting a t 250° : 2 2/-DichZoro-4 4'-dibromomalonanilide, C,H3ClBr-NH-CO-CH2*CO*NH*C,H3ClBr, is sparingly soluble in boiling alcohol and moderately readily so in boiling glacial acetic acid from which i t crystallises in colourless needles with straight extinction melting at 214O : C,,H,,,0,N2C12Br requires C1= 14.74 ; Br = 33-24 per cent.C,IX3C1Br*NH*CO*CH,*C02Et, is very readily soluble in hot alcohol from which i t crystalliscs in 0.1964 gave 0.2700 A@+ AgBr. C1= 14-72 ; Br= 33.18. E thy1 2-chloro-4-b romomalonanila te 102 CHATTAWAY AND CLEM0 : colourless prismatic plates with straight extinction melting a t 81.5O: 0.1990 gave 0.2061 AgCl+ AgBr. C1= 11.09 ; Br = 24-99. C,,H,,03NClBr requires Cl = 11-06 ; Br = 24.94 per cent. 2-CitZoro-4-bromomaZonaiz~Zic acid C,H3ClBr-NH*CO*CH2*C0,H, crystallises well from hot water or alcohol in both of which it is readily soluble in long slender colourless prisms.When quickly heated it melta a t 165O and evolves carbon dioxide producing aceto-2-chloroo-4-bromoanilide : 0.1598 gave 0.1799 AgCl+ AgBr. C1= 12.05 ; Br = 27.17. 2-Chloro-4-bromomalonanilamide, C,H70,NClBr requires C1= 12.12 ; Br = 27.33 per cent, C,H3C1Br*NH*CO*CH2*CO*NH2, is readily soluble in alcohol from which it crystallises in long, slender colourless needlei. melting a t 149O : 0.0872 gave 0.0994 AgC1+ AgBr. C1=12*20 ; Br= 27-51, C,H,O,N,ClBr requires C1= 12.16 ; Br = 27.42 per aent. 4 41-Dichloro-2 2'-dibromomalonanilide crystallises from boiling glacial acetic acid in which it is moderately readily soluble in colourless prisms melting a t 221° with straight extinction in con- vergent light an acute bisectrix of a wide angle positive biaxial figure almost normal axial plane parallel to elongation of plate : 0-2342 gave 0.3206 AgCl + AgBr.C,,H,o0,N,Cl,Br2 requires C1= 14.74 ; Br = 33.24 per cent. Ethyl 4-chloro-2-bromomaZonanilate is very readily soluble in alcohol from which it separates in colourless prismatic plates melt- ing a t 83.5O: C1= 14.65 ; Br = 33.03. 0.1675 gave 0.1740 AgCl+AgBr. C1=11*12; Br=25*07. 4-ChZoro-2-bromomalonaniZic acid crystallises from hot water in which it is readily soluble in long colourless prisms. It melts when quickly heated a t 161° giving off carbon dioxide and producing aceto-4-chloro-2-bro~moanilide : C,,H,,03NCIBr requires C1= 11 *06 ; Br = 24.93 per cent.0.1143 gave 0.1290 AgCl + AgBr. Cl= 12.08 ; Br = 27.23. C,H703NClBr requires C1= 12.12 ; Br= 27.33 per cent. 4-Chloro-2-bromomalonanilaml.'de crystallises from alcohol in which it is very readily soluble in long slender colourlw needles melting a t 159O. 2 4-Dibromomalonanilamide C,H3Br2*NH*CO*CH2.CO*NH2 is very readily soluble in alcohol and crystalliws in colourless slender needles melting a t 164O : 0.1374 gave 0.1545 AgBr. Br=47:85. C,H,02N2Br2 require@ Br= 47.58 per cent POLYMORPHISM IN HALOGEN-SUBSTITUTED ANILIDES. 103 The dihalogen-substituted sulphonanilides are readily obtained by dissolving the aniline mixed with an equivalent weight of pyridine in dry ether and adding an ethereal solution of the equivalent amount of the sulphonyl chloride.The ether is expelled on the water-bath and the product washed with dilute hydrochloric acid, and crystallised from alcohol or acetic acid. They are stable well-crystallised compounds which react with hypochlorous and hypobromous acids to yield stable N-chlorides and bromides. Benzenesulphon-2-chloro-4-6 romoadide, C6H3C1Br*NH*S0,*C6H , crystallises from hot alcohol in which it is very readily soluble in striated colourless flat prisms with straight extinction melting a t 122O: It was analysed by conversion into its N-chloride and titration of the iodine liberated when a weighed amount of the compound dissolved in chloroform is added to a solution of potassium iodide acidified with acetic acid. Benzenesulphon-2-chloro-4-bromophenylchloroam~de, C,H3C1Br*NC1*S0,*C6H~, crystallises from a mixture of chloroform and light petroleum in colourless stout prims.When rapidly heated it melta a t 111-112O : 0.6272 liberated I=30*6 C.C. N/10-I. Cl as NC1=9*14. C,,H,O,NCl,BrS requires C1 as NC1= 9.30 per cent. Benaenesulphon-4-chloro-2-bromoanilide is very readily soluble in glacial acetic acid o r alcohol from which it crystallises in colourless, flat prisms terminated by domes with straight extinction melting at 128O: Benzenesulpho~~4-chloro-2-bromophenylchloroamide crystallises in colourless stout prisms. When rapidly heated i t melts a t 123O : 0.5133 liberated I=26*6 C.C. N/10-I. C1 as NC1=9-18. p-Toluenesulphon-2-chloro-4-bromoan~lia?e, CI2H,O2NCl2BrS requires C1 as NC1= 9.30 per cent.C6H,CIBr0NH*SO,*C6H4Me, is readily soluble in alcohol and crystallises in colourless plates melting a t 121O. p-Toluenesulphon-2-chloro-4-bronwphenylc hloroamide, C6H3C1Br*NC1*S0,* C6H4Me, is very readily soluble in chloroform from which it separates as thO solvent evaporates in colourless stout prismatic crystals melting a t 78O: 0.1572 liberated I=7*8 C.C. N/10-I. C1 as NCl=8.79. C,3H,o0,NC12BrS requires C1 as NCI = 8-97 per cent 104 POLYMORPHISM IN HALOGEN-SUBSTITUTED ANILIDES. p-Tolueneszdphon-2 4-dibromoanilide, C6H3Br,*NH*S0,=C,HH4Me, crystallises from alcohol in which it is readily soluble in colourless plates with straight extinctions melting a t 134O : p-Toluenesulphon-2 4-dibro moph enylchloroamide, C,H3Br2=NC1*S0,*C6H,Me, crystallises from chloroform in which i t is very readily sohble in colourless stout prisms melting a t 78O: 0.1484 liberated I= 6.95 C.C.NI10-I. C13Hl,02NC12Br2S requires C1 as NC1= 8.07 per cent. p-Toluenesulphon-4-chEoro-2-bronioanilide crystallises from alcohol, in which it is very readily soluble in colourless plates melting a t 12 6'5O. p-Tolueneszdphon - 4 - chloro - 2 - bromophenylchloroamide is very readily soluble in chloroform and separates as the solvent evapor- ates in colourless stout prisms melting a t 77O. C1 as NC1=8*30. p-Nitro tolzi ene-o-sulphon-2-chloro-4-bromoanilide, C,H3ClBr*NH*S02-C,H3Me*N02, is moderately readily soluble in alcohol and readily so in boiling glacial acetic acid from which it crystallises in very pale yellow, obliquely terminated prisms melting a t 164.5O ; extinction 25O.p-Nitro t oluene-o-sulphon-2-chloro-4-b romophenylchloroamide, c6H3C1Br*NC1*S0,*C6H3Me~No~, crystallises in colourless prisms which when rapidly heated melt a t 123-124': 0.5084 liberated I=22*6 C.C. #/lo-I. Cl3H9O4N2Cl,BrS requires C1 as NC1= 8.05 per cent'. p-il7itrotol.z~e~ze-o-sulphon-4-chloro-2-bromoanili~e is moderately readily soluble in alcohol o r glacial acetic acid and crystallises in very pale yellow flattened prisms with 60° and 43O terminal faces, and straight extinction melting at 165O. p~~~itrotoluene-o-sulphon-4-chloro-2-bromophenylchloroamide crys- tallises in colourless prisms melting a t 1 2 2 O when quickly heated : 0.4230 liberated I=18*8 C.C. N/lO-I.p-nitro toluene-o-sulphon-2 4-dibromoanilide, C1 as NC1=7*86. C1 as NC1=7*88. C,3H,0,N2C12BrS requires C1 as NC1= 8-05 per cent. C6H3Br,~~H*S02=C,H3Me~N0,, is moderately readily soluble in alcohol and readily so in glacial acetic acid. It crystallises in very faintly yellow flat prisms with straight extinction melting a t 173O. p-ATitrotoluene-o-s~clphon-2 4-dibromop?~enylchloroamide, C6H3Br2-NCl* SO,*C,H,Me*NO, GHOSH A SYNTHESIS OF FLAVONES. 105 crystallises in colourless prisms melting a t 124-125O when quickly heated : 0.2839 liberated I=11*4 C.C. N/10-I. All the dihalogen-substituted anilines yield diazonium salts, which couple readily in alkaline solution with &naphthol. Several of these compounds are described as they are useful for purposes of identification.C6H,C'IBr -N2*ClOH6*OH, is moderately readily soluble in glacial acetic acid from which it crystallises in plates melting a t 210° which have a beetle-green colour by re'flected light and a red colour in transmitted light': 0.1946 gave 0.1769 AgCl + AgBr. C,6-H,oON,ClBr requires C1= 9.80 ; Br = 22.10 per cent. 4-Ghloro-2-bronzo benzeneazo-P-naphthol crystallises from glacial acetic acid in brilliant xed plates melting a t 193O; they give straight extinction and are sharply pleochroic-orange-red : C1 as NC1=7.12. C13H,0,N2C1Br2S requires C1 as NC1= 7-32 per cent. 2-Chloro-4- b ro ?no b e n ze nea z o-/3-naph t hol, C1= 9.73 ; Br = 21-93. 0.1516 gave 0.1383 AgCl + AgBr. 2 4-Dibromobenzeneazo-P-~ff~?~thol has already been described by Hantzsch and Schmiedel (Ber.1897 30 78). It is sparingly soluble in acetic acid from which it crystlallises in flattened brick- red prisms melting a t 203O. They give straight extinction and in convergent light a biaxial figure is visible. (Found Br= 39.19. C,,H,oON2Br2 requires Br = 39.38 per cent.) C1= 9-76 ; Br = 22.01. Cl,HloON,CIBs re'quires C1= 9.80 ; Br = 22-10 per cent. The authors wish to express their thanks to Mr. T. V. Barker, who has kindly described the crystals. UNIVERSITY CHEMICAL LABORATORY, OXFORD. [Receiued December 14th 19151 POLYMORPHlSM IN HALOGEN-SUBSTITUTED ANILIDES. 89 VIII.-Polyrnoyphism Halogen-substitu ted A nilides. By FREDERICK DANIEL CHATTAWAY and GEORGE ROGER CLEMO. ALTHOUGH ability to crystallise in more than one form appears t o be a general property of the anilides few of the modifications, unstable in ordinary circumstances have been described.They are formed only under favourable conditions as a first stage in the process of crystallisation and then more or less rapidly redissolve and disappear ; their transformation theref ore unless accompanied by some colour change may easily pass unobserved. 1)escriptions of the phenomena attending the crystallisation of aceto-p-bromo- and aceto-2 4-dibromo-anilide (Chattaway and Lam- bert T. 1915 107 1766) and propiono-piodoanilide (Chattaway and Constable T. 1914 105 126) have recently been published; the present paper gives an account of the similar behaviour of other derivatives of 2 4-dihalogen-substitukd anilines.Eoth aceto-2-chloro-4-bronio- and 4-chloro-2-bromo-aililide car1 exist in two modifications. When a saturated solution of either in glacial acetic acid is allowed to cool undisturbed an unstable needle-shaped modification a t first separates the long interlaced crystals com 90 CHATTAWAY AND CLEM0 : pletely filling the liquid. I n the crystalline mass compact crystals of a stable form soon make their appearance; the needle- shaped crystals dissolve and the compact crystals grow until the conversion is complete. If only a few crystals of the stable form appear they grow sIowly if the crystalline mass of the unstable form is left undishrbed; but if it is stirred the transformation is complete in a few seconds small crystals of the compact form falling as a crystalline precipitate.The transformation of thO unstable into the stable form of aceto-2-chloro-4-bromoanilide is much more sapid than the corre- sponding transformation of aceto-4-chloro-2-bromoanilide. The pnitrobenzoyl derivatives of the 2 4-dihalogen-substituted anilines also all crystallise in two forms. On cooling hot saturated alcoholic solutions they separate in slender needles or hairs which convert the liquid into a felt-like mass. After a time crystals of the' stable compact forms make their appearance the original crystals disappear and the new crystals grow until conversion is complete. Of the pnitrobenzanilides pnitrobenzo-2-chloro-4-bromo- anilide changes most rapidly ; if the liquid containing the hair-like form in suspension is shaken conversion is complete in a few minutes.In a series of comparative experiments in which similar amounts of solution were used and the felted masses of hair-like crystals were allowed t o transform undisturbed about thFee days were required in the cases of p-nitrobenzo-4-chloro-2-bromoanilide and of pnitrobenzo-2 4-dibromoanilid0 whilst in the case of pnitrobenzo-2 4-dichloroanilide several weeks elapsed before all the unstable form had disappeared. EXPERIMENTAL, I'repamtiot~ vf 2 4-BibromoadirLe. Tlie preparation of 2 4-dibromoaniline has always offered coil- siderable difficulty. The action of bromine on aniline dissolved in glacial acetic acid is so energetic that in practice it cannot be checked a t intermediate stages the substituted product even if a small quantity of bromine is used being mainly 2 4 6-tribromo- aniline o r if excess of bromine is used the stable perbromide of this compound C,H,Br3*NH,,HBr3.The action of bromine on acet- anilide is less vigorous and if a gram-molecule of bromine is added t o a gram-molecule of acetanilide dissolved in acetic acid in the presence of a gram-molecule of sodium acet'ate aceto-pbromoanilide is quantitatively and exclusively produced. A second atom of bromine can only be made t o enter the acetanilide with much greater difficulty some hours of action a t a high temperature bein POLYMORPHlSM IN HALOGEN-SUBSTITUTED ANILIDES. 91 required. During the necessary heating if any water is present hydrolysis takes place so easily even in the presence of excess of sodium acetate that further action results in the production of 2 4 6-tribromoaniline7 just as if aniline itself had been employed.This however can be avoided and aceto-2 4-dibromoanilide can be quantitatively obtained i f precautions are taken to avoid the presence of hydrobromic acid and of water. The best procedure is as follows A gram-molecule of aceto-pbromoanilide * and a gram- molecule of fused and finely powdered sodium acetate are suspended in sufficient glacial acetic acid to make a thick paste. A gram- molecule of bromine dissolved in eight t o ten times its volume of glacial acetic acid is added slowly and the mixture heated for five t o six hours on a water-bath until the colour of bromine has almost disappeared. On diluting the' cooled product with water, aceto-2 4-dibromoanilide (m.p. 1 4 6 O ) separates and should be immediately filtered off and crystallised from alcohol. The aniline is obtained by dissolving the anilide in boiling alcohol mixed with about one-eighth of its bulk of concentrated hydrochloric acid, boiling the liquid under a reflux condenser for eight t o nine hours, distilling off the alcohol in a current of steam and adding a slight excess of sodium hydroxide to the cooled residue. The aniline separates as a white crystalline mass (m. p. 78-79O) the yield being about 90 per cent. of the quantity theoretically obtainable from the aceto-pbromoanilide used. The aniline may be obtained free from colour by distillation in a current of steam but this is somewhat tedious if large quantities are required and it is quite unnecessary if the base is to be used for subsequent preparations.Preparation of 2-Chloro-4-bromoaizililze and 4-Chloro-2-bromo- a d i n e . Tlie preparation of 2-cliloro-4-bromoaniline offers no difficulty as the corresponding aceto-2-chloro-4-bromoanilide is easily prepared either by the action of chlorine on aceto-pbromoanilide or of bromine on aceto-o-cliloroanilide. The former method is more convenient as pbromeacetanilide is so easily procured. The process is best carried out by passing the requisite amount of chlorine into a cooled sus- pension in glacial acetic acid of one gram-molecule of aceto-p-bromo- anilide and one gram-molecule of anhydrous sodium acetate. The action proceeds easily a t the ordinary temperature and after pre- cipitation by dilution and crystallisation from alcohol aceto-2- chloro-4-bromoanilide (m.p. 1 5 1 O ) is obtained in good yield. This * Acetaiiilide may be used employing double the quantity of sodium acetate and of bromine 98 CHATTAWAY AND CLEM0 : can then easily be converted into 2-chloro-4-bromoaniline (m. p. 7 3 O ) by ilie method previously described. I n order to effe'ct the entrance of a bromine1 atom into the ortho- position in aceto-pchloroanilide several hours' heating on a water- bat.h are required. The method which has been found most con- venient is similar to' that described for the preparation of aceto-2 4- dibromoanilide. The presence of hydrobromic acid and of water must be similarly avoided An excellent' yield of pure aceto-4- chloro-2-bromoanilide (m.p. 1 3 7 O ) is obtained in this way from aceto-pchloroanilide and from this 4-chloro-2-bromoaniline (m. p. 6 9 O ) can he obt'ained as above described. Polgmorplzic Forms of the A cetochlorobromoanilides. When a nearly saturated solution of aceto-2-chloro-4-bromoanilide in boiling glacial acetic acid is allowed to cool slowly no separation of crystals takes place as a rule until the liquid has reached a temperature not far removed from the ordinary; tufts of fine, needle-shaped crystals of an unstable modification then appear, often floating in the liquids and grow steadily; very soon compact crystals of the stable form make their appearance generally on the surface of the liquid and having reached a moderate size fall in a shower t o the bottom the tufts of the needle-shaped modification dissolving and disappearing.The transformation is usually so sapid that unless the cooling liquid is observed with a lens the tufts of the unstable crystals may easily escape notice. When a small quantity of such a hot saturated solution in glacial acetic acid is cooled rapidly in ice the unstable modification sepa- rates a t a few points in small tufts of very slender prisms which grow rapidly until a felted mass of interlaced fine crystals is formed. These if left undistarbed may remain for several days before crystalline nuclei of t'he stable compact rhombic form make their appearance and grow slowly but if a few crystals of the stable form are added and the mass is stirred transformation takes place very rapidly and in a few seconds the needledaped modification dissolves and disappears and is replaced by small crystals of the compact form which sink to the bottom of the liquid as a crystalline precipitate.An alcoholic solution of aceto-2-chloro-4-bromoanilide behaves similarly when cooled. Aceto-4-chloro-2-bromoanilide also crystallises in two polymorphic forms. Transformation of the unstable into the stable form is not so rapid as in the case of the isomeric compound. When a nearly saturated solution in hot glacial acetic acid is boiled for some time to destroy all crystalline nuclei and then allowecl to cool slowly the unstable modification separates in cluster POLYMORPHISM IN HALOGEN-SUBSTITUTED ANILIDES. 93 of slender needles which completely fill the liquid; these if left undisturbed may remain without change for days but sooner or later crystals of the compact stable modification make their appear- ance and grow slowly to a considerable size the unstable form dissolving and disappearing after some hours.If a few crystals of the stable modification are introduced into the felted mass of needle-shaped crystals and the whole is vigorously stirred transformation is complete in a few minutes small crystals of the compact form being the product. A number of derivatives of the 2 4-dihalogen-substituted anilines, not hitherto described have been prepared in order t o ascertain which can easily be obtained in two' forms. The butyranilides were prepared by heating butyric anhydride with equivalent amounts of the various anilines.n-Bu tyro-2-chloro-4-brorn oanilide, C,H,ClBr~NH*CO.CH,*CH,*CH, is very readily soluble in alcohol but less readily so in light pet(ro1eum. It crystallises in long colourless needle,. melting at 1 loo ; extinction 21° : 0.2234 gave 0-2682 AgCl+ AgBr. n-Bu tyro-4-chZoro-2-bromoanilide is very readily soluble in alcohol, but less readily so in a mixture of light pekroleum and benzene. It crystallises in colourless needles melting a t 1 1 1 . 5 O ; extinction 31° : C1= 12.85 ; Br = 28.97. CloHllONCIBr requires C1= 12.82 ; Br= 28.90 per cent. 0.2275 gave 0.2730 AgCl + AgBr. e l = 12.84; Br = 28.96. C,oH,lON@lBr requires C1= 12.82 ; B r = 28.90 per cent. The phenylacet'anilides were prepared by the interaction of equi- valent quantities of phenylacetyl chloride and the aniline dissolved in ether in the presence of an equivalent weight of pyridine.Yl~etzylaceto-2-c~loro-4-bronioan~lide, C6H3C1Br*NH*CO*CH,*C,H,, crystallises from boiling alcohol in which it is readily soluble in colourless needles melting a t 150° ; extinction variable : 0.2187 gave 0.2233 AgCl + AgBr. C14Hl10NC1Br requires C1= 10.92 ; B r = 24-63 per cent. Phenylaceto-4-chlo~o-2-bromoanilide crystallises from boiling alcohol in which it is readily soluble in long colourlms needles melting a t 1 4 8 O : C1= 10.93 ; Br = 24-64. 0.1536 gave 0.1568 AgCl+ AgBr. Phenylaceto - 2 4 - dibromoanilide, C1= 10.92 ; Br= 24.64. C14Hl,0NC1Br requires C1= 10.92 ; Br= 24.63 per cent. C,H3Br,*NH*CO*CH,*CH~, crystallises from alcohol in long colourless needles melting a t 160° 94 CHATTAWAY AND CLEM0 : 0.1520 gave 0.1548 AgBr.C,,H,,0NBr2 requires Br =43-32 per cent. The nitrobenzo-2 4-dihalogen-substituted anilides were prepared by adding the nitrobenzoyl chlorides to equivalent amounts of the anilines dissolved in ether in the presence of pyridine or of a concentrated solution of sodium carbonate. Br = 43.34. o-Nit ro b enxo-2-chloro-4-bromoanilide, C,H,ClBr*NH*CO*C,H,-NO,, is fairly readily soluble in boiling alcohol from which i t crystallises in colourless needles melting a t 165O; extinction 25O : 0.1788 gave 0.1663 AgCl+ AgBr. C1= 9.96 ; Br = 22.44. o-Ni t ro b en co-2 4-dib ro moanilide, C,,Hs0,N2ClBr requires C1= 9.97 ; Br = 22.48 per cent. @,H,Br,~NH~CO*C,H,-NO , is readily soluble in boiling glacial acetic acid but less readily so in boiling alcohol than the compound just described.It crystal- lises from alcohol in pale yellow slightly oblique plates melting a t 178O; extinction 24O : 0.2060 gave 0.1937 AgBr. Br=40.01. C,,H,O,N,Br require's Br = 39.96 per cent. o-Nitrob enzo-4-chloro-2-bromoanilicFe is fairly readily soluble in boiling alcohol and crystallises in colourless slightly oblique plates melting a t 166O; extinction 18O : 0.2153 gave 0.2005 AgCl + AgBr. o-;jTitro b enzo-2 4-dici~loroal?ii1ide, C1= 9.97 ; Br = 22.47. Cl3H,O3N2C1Br requires c1= 9-97 ; Br = 22.48 per cent. ~6H&1,*NH*CO*C6H,* NO,, is readily soluble in boiling alcohol from which it crystallises in colourless oblique platep melting a t 153.5O : 0.2299 gave 0.2116 AgCl.C1=22*77. C,,H,O,N,~ requires c1= 22.80 per cent. m-Nitrob er~zo-2-chloro-4-brol?zoanilide is sparingly soluble in boil- ing alcohol but readily so in boiling glacial acetic acid from which it crystallises in long colourless needles melting at 191O : 0.1971 gave 0.1830 AgCl+ AgBr. C1,H,03N,ClBr requires C1= 9-97 ; Br = 22.48 per cent. m-Nitrobemo-2 4-dibromoanilide is moderately readily soluble in both alcohol and glacial acetic acid and crystallises in colourless needles melting a t 165O ; extinction variable : C1= 9.94 ; Br = 22.41. 0.1990 gave 0.1860 AgBr. m-Nitrobenzo-4-chEoro-2-bromoanilide is readily soluble in both Br = 39-77. C,,Hs0,N2Br2 requires Br = 39.96 per cent POLYMORPHISM I N HALOGEN-SUBSTITUTED ANILIDES.95 alcohol and glacial acetic acid. It crystallises from acetic acid in colourless flattened plates melting a t 167.5* which in collvergent light show a positive biaxial figure the axial angle being extremely small : 0.2149 gave 0.2002 AgCl+ AgBr. C1= 9.97 ; Br = 22.48. C,,H,03N,C1Br requires C1= 9-97' ; Br = 22.48 per cent. m-Nitrob enzo-2 4-dichloroanilide crystallises from glacial acetic acid in colourless flattened plates melting a t 183O; straight extinction. Normal to plate emerges the acute bisectrix of a posi- tive biaxial figure'; axial angle medium : 0.2126 gave 0.1957 AgC1. C1= 22.77. C,,H,0,N2Cl requires C1= 22.80 per cent. p-iVitrob etzzo-2-chloro-4-bri,moanilide is moderately readily soluble in boiling glacial acetic acid but sparingly so in boiling alcohol.On rapidly cooling a concentrated boiling alcoholic solution it separah in tufts of very fine almost colourless hair-like1 crystals, which grow together to a felt-like mass. Among these interlaced, hair-like crystals six-sided pale yellow plates with straight extinc- tion almost a t once make their appearance and rapidly grow a t the e'xpense of the hair-like form which dissolves and disap- pears. I f the pulpy mass of fine crystals is shaken the conversion is complete in a few minutes the compact form subsiding as a sandy crystalline powder melting a t 199O : 0.2265 gave 0.2109 AgCl + AgBr. p-Nitrobenzo-2 4-dibromoanilide is somewhat readily soluble in boiling glacial acetic acid but lew readily so in boiling alcohol.On rapidly cooling a boiling saturated alcoholic solution it separates in very fine hair-like crystals which on account of their small size appear almost colourless. On allowing these to remain in the solvent a t the ordinary temperature small pale yellow six- sided compact plates with straight extinction seen under the micro- scope to be stout very much flattened prisms with domed ends, make their appearance and grow a t the expense of the hair-like form which gradually dissolves and disappears. On allowing the felted mass of the hair-like form t o remain undisturbed a t the ordinary hmperature in a small flask the conversion was nearly complete in about twelve hours and all crystals of the first form had disappeared a t the end of three days. Both forms melt a t 194O : C1= 9.96 ; Br = 22.47.C,,H,O,N,ClBr requires C1= 9-97 ; Br = 22.48 per cent. 0.2230 gave 0.2092 AgBr. Br = 39-92. p-Nitrob enzo-4-chloro-2-bronloanilide is readily soluble in boiling C,,H,03N,Br2 requires Br = 39.96 per cent 96 CHATTAWAY AND CLEM0 : glacial acetic acid but less readily so in boiling alcohol. On cooling a boiling saturated alcoholic solution i t separates in very fine almost colourless needles with straight extinction which interlace and finally completely fill the liquid. After some time pale yellow stout, six-sided rhombic plates (82.) with diagonal extinction make their appearance and slowly grow a t the expense of the first form cutting their way down through the pulp-like crystalline mass in a very characteristic manner.The conversion is complete after about three days. Both forms melt a t 1 7 4 O : 0.2329 gave 0.21 66 AgCl + AgBr. p-&-itrob enzo-2 4dichZoronnilide is readily soluble in boiling glacial acetic acid but less readily so in boiling alcohol. On cooling a boiling saturated alcoholic solution it separates in very fine almost colourless needles. After some hours small four-sided pale yellow plat,es make their appearance and slowly grow cutting their way down through the interlaced crystals of the first form which dissolve and disappear in the neighbourhood of the growing crystals. These compact plates when seen under the microscope appear as faces of stout six-sided prisms with domed ends. This transformation is slow; thus in one experiment in which a small flask half filled with a pulp of the needle-shaped form was allowed to remain undis- turbed a t the ordinary temperature three weeks elapsed before all the unstable form had disappeared.The length of time taken in transforming naturally depends on the amount t o be transformed, as well as on the temperature and on the nature of the solvent. Both forms melt a t 174O the unstable form transforming before the melting point is reached: 0.1 725 gave 0.1574 AgCl. C1= 22.56. C,,H,O,N,Cl requires C1= 22.80 per cent. The 2 4-disubstituted phthalanils are easily obtained by heating together equivalent quantities of the anilines and phthalic anhy- dride for several hours to about 180°. >N*C6H,c1Br is f air1 y readily soluble in boiling alcohol and readily so in boiling glacial acetic acid from which it crystallises in colourless prisms melting a t 165O; straight extinction : C1= 9-95 ; Br = 22.44.CI3H,O3N,C1Br requires C1= 9.97 ; Br = 22.48 per cent. CO co Y h t ha 20-2-chl oro-4- b rom o a d C,H < 0.2120 gave 0.2094 AgCl + AgBr. C,,H,O,NClBr requires C1= 10.54 ; Br= 23.75 per cent. Ph thalo4-chloro-2-bromoanil is readily soluble in boiling glacial acetic acid and in boiling alcohol fxom either of which i t crystal- lises in colourless prisms with straight extinction melting a t 140O: C1= 10.57 ; Br = 23.84 POLYMORPHISM IN HALOGEX-SUBSTITUTED ANILIDES. 97 0.2299 gave 0.2267 AgCl + AgBr. C1= 10.56 ; Br = 23.79. C,,H702NC1Br requires C1= 1 0.54 ; Br = 23.75 per cent. Phthalo-2 4-dichloroard C?6H4<CO>N*C6H3c12 co is fairly readily soluble in-boiling alcohol and readily so in boiling glacial acetic acid.It crystallises from either in colourless prisms melting a t 155O ; straight extinction : 0.2342 gave 0.2302 AgCl. C,,H702NC12 requires C1= 24-28 pOr cent. Phthalo-2 4-dibromoanil is fairly readily soluble in boiling alcohol and readily so in boiling glacial acetic acid from which i t crystallises in colourless prisms melting a t 153*5O ; straight extinc- tion : C1= 24.31. 0.2198 gave 0.2171 AgBr. Br=42'03. C,,H70,NBr2 requires Br = 41.97 per cent. The methyl and ethyl dihalogen-substituted carbanilates are easily prepared by dissolving equivalent quantities of the corre- sponding aniline and of pyridine in dry ether and adding very slowly as the action is violent the equivalent amount of methyl or ethyl chloroformate.The ether is then distilled off and the solid product well washed with dilute hydrochloric acid to remove the pyridine and crystallised from alcohol. Jf ethyl 2-c hloro-4-bromocclr bandat e C6$13~Br*NH*C0,Me cxys- tallises from alcohol in which i t is very readily soluble in colourless needles melting a t 76.5O; extinction 4 2 O : 02035 gave 0.2524 AgCl + AgBs. C1= 13-28 ; Br = 29.93. C,H70,NC1Br requires C1= 13-40 ; Br = 30.22 per cent. Ethyl 2-c hloro-4- bromoclxr banila t e C,H3C1Br*NH-C02Et is readily soluble in alcohol from which it crystallises in colourless needles melting a t 96O; extinction 38O : 0.2012 gave 0.2380 AgCl + AgBr. C1= 12.66 ; Br= 28-55. CgHg02NC1Br requires C1= 12.73 ; Br = 28-70 per cent.Methyl 2 4-dibromocarbanilate C,H,Br,*NH*CO,Me has been obtained by Hentschel (J. pr. C'hem. [ii] 1886 34 423) by direct bromination and by Fromm and Heyder (Ber. 1909 42 3801) by the action of bromine on phenylthiocarbimide in the presence of aqueous methyl alcohol. It is easily prepared by the general method from 2 4-dibromoaniline and crystallises from alcohol in which it is readily soluble in colourless needles melting a t 97O. Hentschel gives the melting point as 96'5O. (Found Br=51.80. Calc. Br=51*74 per cent.) Ethyl 2 4-dibromocarbanilate C6H3Br2*NH*C0,Et was prepared by Fromm and Heyder (Zoc. cit.) by the action of bromine on phenyl- VOL. CIX. 98 CHATTAWAY AND CLEM0 : thiocarbimide in the presence of aqueous ethyl alcohol.It is easily prepared by the above method. It crystallises from alcohol in which it is readily soluble in colourless needles melting a t lolo, with extinction 40°. (Found Br = 49.58. Calc. Br = 49.49 per cent.) illethyl 4-chloro-2-bromocarbanilat e is very readily soluble in alcohol from which it crystallises in colourless needles melting a t 87.5O : 0.1146 gave 0.1440 Ag@l+ AgBr. C1= 13-45 ; Br= 30.32. Ethyl 4-chloro-2-bromocarbanilate crystallises from alcohol in 0.1510 gave 0.1760 AgCl+ AgBr. Cl= 12.48 ; Br=28-13. CgHgO,NBrC1 requires C1= 12.73 ; Br = 28.70 per cent. Methyl 2 4-dichloro’carbaniZate C,H,C?l,*NH*C’O,Me is ex- tremely readily soluble in alcohol from which it crystallises in colourless needles melting a t 70’5O : C8H702NB&1 requires C1= 13-40 ; Br = 30.22 per cent.which it dissolves very readily in colourless needles melting a t 90° : 0.1113 gave 0.1450 AgCl. CI=32*22. C8H70,NCl requires c1= 32.23 per cent. Ethyl 2 4-dichlorocarbaniZate C,H,C~,-~H*CO,Et crystallises from alcohol in which i t is readily soluble in colourless needles melting atl 89O: 0.1121 gave 0.1443 AgCl. C1=30.64. CgHgO,NC1 requires C1= 30.30 per cent. Lik0 other substituted anilines 2-chloro-4-bromoaniline and 4-chloro-2-bromoaniline react readily with carbamide a t a somewhat elevated temperature ammonia is liberated and the monob and s-di-dihalogen-substituted carbamides are produced. ThO latter compounds are formed in largest amount even when a considerable excess of carbamide is used and me easily isolated owing to their very sparing solubility in ordinary organic solvents.A yield of the s-tetra-substituted carbamide amounting to about 80 per cent. of the theoretical is obtained when a gram-molecule of the aniline is heated with 4 gram-molecules of carbamide to 180° for forty hours. 9-2 2r-Dic hloro-4 4f -dib r omodipheny lcarbamide, C6H3ClBr*T”H CO*NR-C,H,ClBr, is very sparingly soluble in boiling glacial aoetic acid. It is much more readily soluble in hot nitrobenzene from which it separates in colourless very fine hair-like needles with oblique extinction, melting at 279O: 0-1420 gave 0.2133 AgCl+ AgBr. C1= 16-08 ; Br = 36.25. C,3H,0N,CI,Br requires C1= 16.16 ; Br = 36.42 per cent POLYMORPHISM IN HALOGEN-SUBSTITUTED ANILIDES. 99 s - 4 4/-Bichloro-2 2'-dibromodiphenylcar6am~e resembles the isomeric compound very closely.It crystallises from hot nitro- benzene in colourlel;Es fine needles melting a t 274O; extinction 40' : 0.1988 gave 0.2990 AgCl + AgBr. C1= 16-10 ; Br = 36.29. C13HsON2C12Br2 requires C1= 16-16 ; Br = 36.42 per cent. All thO dihalogen-substituted anilines react readily with diethyl oxalate a t a suitably high temperature; using excess of ester the corresponding oxanilide is the chief compound; using excess of aniline the oxanilide predominates. They are easily separated on account of the very sparing solubility of the tetra-substituted uxan- ilides. The 2 4-&halogen derivative are k t prepared by heating 1 gram-molecule of the aniline with 5 gram-molecules of diethyl oxalate to about 220' f o r an hour.On cooling and adding a little alcohol both the oxanilate and the tetra-substituted oxanilide separate out. A further quantity of the two compounds can be obtained by distilling off the alcohol from the filtrate and reheating the residue. The two compounds are readily separated by boiling with alcohol, when the oxanilate is dissolved. The yield of the tetrahalogen- substituted oxanilide is always very small. The disubstituted obxanilic acids are be& prepared by suspending the corresponding ethyl disubstituted oxanilates in about twenty times their weight of water adding rather mom than the equivalent of sodium carbonate and passing steam until the este.rs have dis- appeared. The sodium salts of the acids crystallise out on cooling.On adding an excess of hydrochloric acid to hot aqueous solu- tions of the salts and cooling the acids separate. They crystal- liw from water with one molecule of water of crystallisation. When the monohydrated acids are dehydrated and recrystallised from toluene or benzene the anhydrous acids are obtained. The disubstituted oxanilamides separate as sparingly soluble, crystalline solids when dry gaseous ammonia is passed into warm alcoholic solutions of the corresponding ethyl oxanilahs. Ethyl 2-cltEoro-4-bromo-oxanitate C,H3ClBr*NH*CO*C0,Et is readily soluble in boiling alcohol from which it crystallises in colourless needles melting a t 124O : 0.1700 gave 0.1828 AgCl+ AgBr. C1= 11.51 ; Br=25*95. ClOH,O3NC1Br requires C1= 11.57 ; Br = 26.08 per cent.2-Chloro-4-bromo-oxa:aniEic acid CGH3CIBr*NH*CO*C0,H crystal- lises in colourless needles melting and decomposing a t 131O: 0.1115 gave 0-1332 AgCl+ AgBr. C1= 12-79 ; Br = 28.83. CsH,O3NC1Br requires C1= 12.73 ; Br= 28.70 per cent. E 100 CHATTAWAY AND CLEM0 : s-2 2/-Dichloro-4 4/-dibromo-oxanilide, C6H3C1Br *NH*CO*CO*NH*C,H,ClBr, is spasingly soluble in boiling alcohol o r glacial acetic acid but dis- solves readily in hot nitrobenzene from which it crystallises in small colourless needlep melting a t 285O : 0.0727 gave 0.1026 AgCl+ AgBr. C1= 15-11 ; Br = 34.06. C,,H80,N2C12Br2 needs C1= 15.19 ; Br = 34.23 per cent. 2 - Chloro - 4 - bromo-oxardamide C6H3C1Br*NH*CO*CO*NH2 is modesately readily soluble in boiling alcohol and readily so in hot nitrobenzene.It crystallises in colourless needles melting a t 243O : 0.0905 gave 0-1079 AgCl + AgBr. C1= 12-76 ; Br= 28-77. Ethyl 4-chloro-2-bromo-oxanilate crystallises from alcohol in 0.1932 gave 0.2087 AgC1+ AgBr. C1= 11.56 ; Br = 26-07, 4-Chloro-2-bromo-oxa~~ilic acid crystallises from boiling benzene in When crystallised from water it separates in colourless needles 0.1.177 gave 0.1320 AgClSAgBr. C1=12*00; Br=27*06. s-4 4f-Dichloro-2 2r-dibrorno-oxanilide is very sparingly soluble in boiling glacial acetic acid. It crystallises from hot nitrobenzene in which it is fairly readily soluble in small colourless needles melt- ing a t 295O: (&H602N,C1Br requires c1= 12.78 ; Br = 28-80 per cent. which it is readily soluble in colourless needles melting a t 121O: C,,H,03NC1Br requires C1= 11.57 ; Br = 26.08 per cent.colo8urless needles melting with decomposition a t 126-127O : containing one molecule of water of crystallisation : C8H,O3NC1Br,I~,O requires C1= 11.96 ; Br = 26.96 per cent. 0.1531 gave 0.2159 AgCl+ AgBr. C1= 15.10; Br = 34-03. 4-Chloro-2-bromo-oxanilamide crystallises from alcohol in colour- 0.1658 gave 0.1963 AgCl+AgBr. C1=12*68; Br=28*57. C8H,02N2ClBr requires C1= 12-78 ; Br = 28.80 per cent. Ethyl 2 4-dibromo-oxanilate C6H,Brz=NH*CO*C0,Et is readily soluble in hot alcohol from which it separates in colourless curved, hair-like crystals melting a t 130° : C1,H,02N,C12Br2 requires C1= 15.19 ; B r = 34.23 per cent. less needles melting a t 236O : 0.1750 gave 0.1870 AgBr. Br=45.47. C,,H,O,NBr requires Br = 45.51 per cent.2 4-Bibromo-oxanilic acid C,H,Br,*NR*CO*CO,H crystallises from boiling benzene in colourless needles melting and decomposing a t 138O into carbon dioxide and 2 4-dibromoformanilide. It crystal- lises from boiling water in which it is readily soluble in slender POLYMORPHISM IN HALOGEN-SUBSTITUTED AEILIDES. 101 colourlm pxisms containing one molecule of water of crystallisa- tion which is slowly lost a t 100': 0.1711 gave 0.1898 AgBr. Br=47.20. 2 4 2' 4'-Tetrabronzo-oxanilide, C,H,03NBr2,H,0 requires Br = 46.88 per cent. C,H3Br2*NH*CO*CO*NH*C6H3Br2, is very sparingly soluble in boiling glacial acetic acid 150 C.C. dis- solving only about 0.1 gram. It is moderately readily soluble! in boiling nitrobenzene from which it crystallises in colourless needles melting a t 298O ; straight extinction : 0.2955 gave 0.3984 AgBr.Br = 57-37. 2 4-Dibromo-oxaniZamide C,H3Br,*NH*CO*CO*NH3[, crystallises 0.2604 gave 0.3018 AgBr. Br = 49.32. C8~,02N,Br requires Br = 49.66 per cent. It has been shown (Chattaway and Mason T. 1910 97 339) that diethyl malonate a t its boiling point reacts with the halogen- substituted anilines halogen derivatives of malonanilide and of ethyl malonanilate being produced. 'These can be separated owing to the sparing solubility of the substituted malonanilides and the ready solubility of the ethyl malonanilates in all ordinary solvents. The halogen-substituted malonanilic acids are also very easily obtained by hydrolysing the substituted malonanilic esters by suspending them in a dilute solution of sodium carbonate and passing steam through the liquid until the ester disappears.After concentrating th0 solutions and adding a slight excess of hydro- chloric acid the acids separate in a crystalline state. When heated, the malonanilic acids decompose quantitatively into carbon dioxide and the corresponding substituted acetanilide. The malonanilamides are obtained by passing ammonia into alcoholic solutions of the corresponding ethyl malonanilates. cI4II8o2N2Br4 repires Br = 57.52 per cent. from alcohol in colourless neledles melting a t 250° : 2 2/-DichZoro-4 4'-dibromomalonanilide, C,H3ClBr-NH-CO-CH2*CO*NH*C,H3ClBr, is sparingly soluble in boiling alcohol and moderately readily so in boiling glacial acetic acid from which i t crystallises in colourless needles with straight extinction melting at 214O : C,,H,,,0,N2C12Br requires C1= 14.74 ; Br = 33-24 per cent.C,IX3C1Br*NH*CO*CH,*C02Et, is very readily soluble in hot alcohol from which i t crystalliscs in 0.1964 gave 0.2700 A@+ AgBr. C1= 14-72 ; Br= 33.18. E thy1 2-chloro-4-b romomalonanila te 102 CHATTAWAY AND CLEM0 : colourless prismatic plates with straight extinction melting a t 81.5O: 0.1990 gave 0.2061 AgCl+ AgBr. C1= 11.09 ; Br = 24-99. C,,H,,03NClBr requires Cl = 11-06 ; Br = 24.94 per cent. 2-CitZoro-4-bromomaZonaiz~Zic acid C,H3ClBr-NH*CO*CH2*C0,H, crystallises well from hot water or alcohol in both of which it is readily soluble in long slender colourless prisms. When quickly heated it melta a t 165O and evolves carbon dioxide producing aceto-2-chloroo-4-bromoanilide : 0.1598 gave 0.1799 AgCl+ AgBr.C1= 12.05 ; Br = 27.17. 2-Chloro-4-bromomalonanilamide, C,H70,NClBr requires C1= 12.12 ; Br = 27.33 per cent, C,H3C1Br*NH*CO*CH2*CO*NH2, is readily soluble in alcohol from which it crystallises in long, slender colourless needlei. melting a t 149O : 0.0872 gave 0.0994 AgC1+ AgBr. C1=12*20 ; Br= 27-51, C,H,O,N,ClBr requires C1= 12.16 ; Br = 27.42 per aent. 4 41-Dichloro-2 2'-dibromomalonanilide crystallises from boiling glacial acetic acid in which it is moderately readily soluble in colourless prisms melting a t 221° with straight extinction in con- vergent light an acute bisectrix of a wide angle positive biaxial figure almost normal axial plane parallel to elongation of plate : 0-2342 gave 0.3206 AgCl + AgBr.C,,H,o0,N,Cl,Br2 requires C1= 14.74 ; Br = 33.24 per cent. Ethyl 4-chloro-2-bromomaZonanilate is very readily soluble in alcohol from which it separates in colourless prismatic plates melt- ing a t 83.5O: C1= 14.65 ; Br = 33.03. 0.1675 gave 0.1740 AgCl+AgBr. C1=11*12; Br=25*07. 4-ChZoro-2-bromomalonaniZic acid crystallises from hot water in which it is readily soluble in long colourless prisms. It melts when quickly heated a t 161° giving off carbon dioxide and producing aceto-4-chloro-2-bro~moanilide : C,,H,,03NCIBr requires C1= 11 *06 ; Br = 24.93 per cent. 0.1143 gave 0.1290 AgCl + AgBr. Cl= 12.08 ; Br = 27.23. C,H703NClBr requires C1= 12.12 ; Br= 27.33 per cent. 4-Chloro-2-bromomalonanilaml.'de crystallises from alcohol in which it is very readily soluble in long slender colourlw needles melting a t 159O.2 4-Dibromomalonanilamide C,H3Br2*NH*CO*CH2.CO*NH2 is very readily soluble in alcohol and crystalliws in colourless slender needles melting a t 164O : 0.1374 gave 0.1545 AgBr. Br=47:85. C,H,02N2Br2 require@ Br= 47.58 per cent POLYMORPHISM IN HALOGEN-SUBSTITUTED ANILIDES. 103 The dihalogen-substituted sulphonanilides are readily obtained by dissolving the aniline mixed with an equivalent weight of pyridine in dry ether and adding an ethereal solution of the equivalent amount of the sulphonyl chloride. The ether is expelled on the water-bath and the product washed with dilute hydrochloric acid, and crystallised from alcohol or acetic acid.They are stable well-crystallised compounds which react with hypochlorous and hypobromous acids to yield stable N-chlorides and bromides. Benzenesulphon-2-chloro-4-6 romoadide, C6H3C1Br*NH*S0,*C6H , crystallises from hot alcohol in which it is very readily soluble in striated colourless flat prisms with straight extinction melting a t 122O: It was analysed by conversion into its N-chloride and titration of the iodine liberated when a weighed amount of the compound dissolved in chloroform is added to a solution of potassium iodide acidified with acetic acid. Benzenesulphon-2-chloro-4-bromophenylchloroam~de, C,H3C1Br*NC1*S0,*C6H~, crystallises from a mixture of chloroform and light petroleum in colourless stout prims. When rapidly heated it melta a t 111-112O : 0.6272 liberated I=30*6 C.C.N/10-I. Cl as NC1=9*14. C,,H,O,NCl,BrS requires C1 as NC1= 9.30 per cent. Benaenesulphon-4-chloro-2-bromoanilide is very readily soluble in glacial acetic acid o r alcohol from which it crystallises in colourless, flat prisms terminated by domes with straight extinction melting at 128O: Benzenesulpho~~4-chloro-2-bromophenylchloroamide crystallises in colourless stout prisms. When rapidly heated i t melts a t 123O : 0.5133 liberated I=26*6 C.C. N/10-I. C1 as NC1=9-18. p-Toluenesulphon-2-chloro-4-bromoan~lia?e, CI2H,O2NCl2BrS requires C1 as NC1= 9.30 per cent. C6H,CIBr0NH*SO,*C6H4Me, is readily soluble in alcohol and crystallises in colourless plates melting a t 121O. p-Toluenesulphon-2-chloro-4-bronwphenylc hloroamide, C6H3C1Br*NC1*S0,* C6H4Me, is very readily soluble in chloroform from which it separates as thO solvent evaporates in colourless stout prismatic crystals melting a t 78O: 0.1572 liberated I=7*8 C.C.N/10-I. C1 as NCl=8.79. C,3H,o0,NC12BrS requires C1 as NCI = 8-97 per cent 104 POLYMORPHISM IN HALOGEN-SUBSTITUTED ANILIDES. p-Tolueneszdphon-2 4-dibromoanilide, C6H3Br,*NH*S0,=C,HH4Me, crystallises from alcohol in which it is readily soluble in colourless plates with straight extinctions melting a t 134O : p-Toluenesulphon-2 4-dibro moph enylchloroamide, C,H3Br2=NC1*S0,*C6H,Me, crystallises from chloroform in which i t is very readily sohble in colourless stout prisms melting a t 78O: 0.1484 liberated I= 6.95 C.C. NI10-I. C13Hl,02NC12Br2S requires C1 as NC1= 8.07 per cent.p-Toluenesulphon-4-chEoro-2-bronioanilide crystallises from alcohol, in which it is very readily soluble in colourless plates melting a t 12 6'5O. p-Tolueneszdphon - 4 - chloro - 2 - bromophenylchloroamide is very readily soluble in chloroform and separates as the solvent evapor- ates in colourless stout prisms melting a t 77O. C1 as NC1=8*30. p-Nitro tolzi ene-o-sulphon-2-chloro-4-bromoanilide, C,H3ClBr*NH*S02-C,H3Me*N02, is moderately readily soluble in alcohol and readily so in boiling glacial acetic acid from which it crystallises in very pale yellow, obliquely terminated prisms melting a t 164.5O ; extinction 25O. p-Nitro t oluene-o-sulphon-2-chloro-4-b romophenylchloroamide, c6H3C1Br*NC1*S0,*C6H3Me~No~, crystallises in colourless prisms which when rapidly heated melt a t 123-124': 0.5084 liberated I=22*6 C.C.#/lo-I. Cl3H9O4N2Cl,BrS requires C1 as NC1= 8.05 per cent'. p-il7itrotol.z~e~ze-o-sulphon-4-chloro-2-bromoanili~e is moderately readily soluble in alcohol o r glacial acetic acid and crystallises in very pale yellow flattened prisms with 60° and 43O terminal faces, and straight extinction melting at 165O. p~~~itrotoluene-o-sulphon-4-chloro-2-bromophenylchloroamide crys- tallises in colourless prisms melting a t 1 2 2 O when quickly heated : 0.4230 liberated I=18*8 C.C. N/lO-I. p-nitro toluene-o-sulphon-2 4-dibromoanilide, C1 as NC1=7*86. C1 as NC1=7*88. C,3H,0,N2C12BrS requires C1 as NC1= 8-05 per cent. C6H3Br,~~H*S02=C,H3Me~N0,, is moderately readily soluble in alcohol and readily so in glacial acetic acid. It crystallises in very faintly yellow flat prisms with straight extinction melting a t 173O. p-ATitrotoluene-o-s~clphon-2 4-dibromop?~enylchloroamide, C6H3Br2-NCl* SO,*C,H,Me*NO, GHOSH A SYNTHESIS OF FLAVONES. 105 crystallises in colourless prisms melting a t 124-125O when quickly heated : 0.2839 liberated I=11*4 C.C. N/10-I. All the dihalogen-substituted anilines yield diazonium salts, which couple readily in alkaline solution with &naphthol. Several of these compounds are described as they are useful for purposes of identification. C6H,C'IBr -N2*ClOH6*OH, is moderately readily soluble in glacial acetic acid from which it crystallises in plates melting a t 210° which have a beetle-green colour by re'flected light and a red colour in transmitted light': 0.1946 gave 0.1769 AgCl + AgBr. C,6-H,oON,ClBr requires C1= 9.80 ; Br = 22.10 per cent. 4-Ghloro-2-bronzo benzeneazo-P-naphthol crystallises from glacial acetic acid in brilliant xed plates melting a t 193O; they give straight extinction and are sharply pleochroic-orange-red : C1 as NC1=7.12. C13H,0,N2C1Br2S requires C1 as NC1= 7-32 per cent. 2-Chloro-4- b ro ?no b e n ze nea z o-/3-naph t hol, C1= 9.73 ; Br = 21-93. 0.1516 gave 0.1383 AgCl + AgBr. 2 4-Dibromobenzeneazo-P-~ff~?~thol has already been described by Hantzsch and Schmiedel (Ber. 1897 30 78). It is sparingly soluble in acetic acid from which it crystlallises in flattened brick- red prisms melting a t 203O. They give straight extinction and in convergent light a biaxial figure is visible. (Found Br= 39.19. C,,H,oON2Br2 requires Br = 39.38 per cent.) C1= 9-76 ; Br = 22.01. Cl,HloON,CIBs re'quires C1= 9.80 ; Br = 22-10 per cent. The authors wish to express their thanks to Mr. T. V. Barker, who has kindly described the crystals. UNIVERSITY CHEMICAL LABORATORY, OXFORD. [Receiued December 14th 19151

ISSN:0368-1645    

DOI:10.1039/CT9160900089

出版商:RSC    

年代:1916

数据来源: RSC

摘要:

GHOSH A SYNTHESIS OF FLAVONES. 105 IX .-A Synthesis of E’lauones. By BROJENDRA NATH GHOSH. IT has been shown (Jacobson and Ghosh T. 1915 107 1051) that the condensation of ethyl a-phenylacetoacetate with phenols leads t o the production of substituted y-benzopyrones o r flavones. It is shown in the present paper that if instead of the ester th 106 GHOSH A SYNTHESIS OF FLAVONES. nitriles are employed an imide is produced from which the flavone is obtained by boiling with dilute sulphuric aicd: 0 0 NK where X may be either H CH, or C,H,. lysing the imine formed the reaction proceeded as follows: By condensing benzoylacetonitrile and phenols and then hydro-0 0 All the imines dissolve in dilute alcohol giving yellow solutions which show a green fluorescence ; in concentrated sulphuric acid, also yellow solutions are obtained showing in some cases a green fluorescence.On boiling with acetic anhydride and sodium acetate the acetyl derivatives of the corresponding flavones were obtained the imino-group being evidently hydrolysed during the opera tion. That the nitrile is more reactive than the corresponding ester is observed in the case of the condensation of acetylphenylacetonitrile. It has been shown (Zoc. cit.) that ethyl a-phenylacetoacetate and m-cresol do not condense in the presence of concentrated sulphuric acid but acetylphenylacetonitrile and m-cresol under the same con-ditions give the corresponding imine from which the flavone is obtained by boiling with dilute sulphuric acid. The condensation of the nitrile with phenol proceeds best when there is a methyl or preferably a hydroxyl group as substituent in the phenol but with any other substituent the reaction does not take place.With the unsubstituted phenol the yield is very small. As regards the position of the substituent the condensation is favoured when the substituent group occupies the meta-position with respect to the phenolic hydroxyl group but there is always an inhibiting influence when both the meta-positions are occupied. Thus resorcinol and pyrogallol give with acetylphenylacetonitrile a very good yield of the corresponding condensation product but under the same conditions phloroglucinol and the nitrile d o no GEIOSH A SYNTHESIS OF FLAVONES. 107 condense a t all. When the substituent group occupies a position other than the meta the reaction does not take place under the conditions employed.Thus with either o-cresol or quinol and acetylphenylacetonitrile the reaction does not take place in the presence of concentrated sulphuric acid. Four condensing agents were employed in these experiments, namely concentrated sulphuric acid hydrogen chloride phosphoryl chloride and anhydrous zinc chloride. I n the case of acetylphenyl-acetonitrile and phenols concentrated sulphuric acid was the only condensing agent employed with satisfactory results. In the case of f ormylphenylacetonitrile and phenols the product obtained by using concentrated sulphuric acid formed a finely divided, amorphous precipitate which was very difficult to manipulate but by using phosphoryl chloride and dry hydrogen chloride the reactions proceeded satisfactorily.It has been observed (Zoc. cit .) that ethyl a-phenylformylacetate does not condense with phenols in the presence of cold concentrated sulphuric acid probably owing to the existence of the ketonic form a t the ordinary temperature, but formylphenylacetonitrile gives a violet coloration with ferric chloride which shows that a t least a portion of it exists in the enolic form and the hope was entertained that it might react with the aid of sulphuric acid or a similar reagent a t the ordinary temperature. As a matter of fact the condensation takes place a t the ordinarj. temperature with hydrogen chloride and even with phosphoryl chloride. I n the case of benzoylphenylacetonitrile and phenols when con-centrated sulphuric acid was used as the condensing agent the nitrile simply combined with a molecular proportion of water, giving the corresponding amide.With anhydrous zinc chloride the condensation of the nitrile and phenol did n o t take place. The condensation was however brought about by dry 'hydrogen chloride. I n the condensation of benzoylacetonitrile and phenols concen-trated sulphuric acid was the only condensing agent employed In the case of resorcinol a satisfactory yield of the imino-compound was obtained but in the case of a-naphthol or pyrogallol the yield of the corresponding imine was small and with phloroglucinol did not react. Some of the substances here described are coloured and wme are colourless but they all give coloured salts.These facts can be explained on the assumption that those substances which are coloured have a quinonoid structure. The constitutions of the salts may be represented by one of the following formulae 108 GHOSH A SYNTHESIS OF FLAVONES. H HSO, \/ I1 HSO, \/ 0 0 0 0 and those of the free bases by 0 0 0 (VI- 1 (VII. ) Forms I 11 and I11 should all yield a base of the constitution shown in VI. Forms I 11 and VI may a t once be rejected because they do not contain any quinonoid arrangement. Any one of the formuke represented by 111 IV o r V may represent the structure of the salts and in the free state V I I ought t o be the proper representation of the compounds because they all contain the quinonoid arrangement. I n order to test this view m-cresol was condensed with acetylphenylacetonitrile and the free base obtained from this condensation was found to be quite colourless.In this case the base must have the structure No quinonoid arrangement is possible here and the absence of colour is in accordance with the view that the coloured compound CtHOSH A SYNTHESIS OF FLAVONES. 109 have the quinonoid structure. This substance dissolves in concen-trated sulphuric acid with a yellow colour and gives a yellow sulphate and therefore the constitution of this salt must be repre-sented by 111. Thus it becomes evident that formula I11 ought to be the proper representation of the salts whilst when they are converted into their bases rearrangement into the quinonoid struc-ture (VII) takes place when a hydroxyl group is present but they retain their structure (VI) when no such arrangement is possible.E X P E R I M E N T A L . Condensations with Acetylphenylacetonitrile. 4-lmino-7-hydroxy-3-phenyl-2-meth yl-y-b e rzzopyran Sulpha te, HTO, Concentrated sulphuric acid (15 c.c.) was slow,y added to resorcind (4 grams) and acetylphenylacetonitrile (5 grams) (Reckh, Rer. 1898 31 3161). When kept overnight the crystalline mass obtained from the cooled solution was poured on ice. After three hours the clear solution yielded a voluminous crystalline precipi-tate which was crystallised from very dilute sulphuric acid. The yield was almost quantitative : 0.1302 gave 0.2630 CO and 0.0538 H,O. 0.1715 , 6.4 C.C. N2 a t 21° and 768 mm. N=4.23. C,,H,,02N,H2S04 requires C = 55-01 ; H =4*50 ; N = 4.00 per cent.The salt forms short yellow needles melting to a dark red liquid a t 246O; it dissolves in dilute sodium hydroxide solution and in concentrated sulphuric acid to yellow solutions the latter show-ing a green fluorescence which is also exhibited in dilute alcoholic solution. The base prepared by treating an alcoholic solution of the sulphate with an excess of aqueous sodium acetate crystallised from dilute alcohol in brown prisms melting and decomposing a t 221O : 0.1214 gave 0.3300 CO and 0.0628 H,O. C'=74*15; H=5.55. 0.0980 , 4.8 C.C. N2 a t 15O and 752 mm. N=5*67. C,,H,,0,N,~H20 requires C = 73.85 ; H = 5.38 ; N = 5-38 per cent. The p-crate separated in orange yellow needles melting a t 227O: C=55*10; H=4.59.Alcoholic ferric chloride develops a violet coloration 110 CtHMH A SYNTHESIS OF FLAVONES. 0.0760 gave 8.6 C.C. N a t 17O and 764 mm. The perchlorate separated in yellow needles which on heating 0.1232 required 3.5 C.C. N/lO-NaOH. The acetyl derivative crystallised from very dilute alcohol in It dissolved in con-N=11*66. C16H130,N,~6H307N3 requires N = 11 * 66 per cent. deco,mposed with explosive violence : HC10 = 28.4. CIGH,,O,N,HC1O requires HClO = 28-5 per cent. very faintly yellow needles melting a t 135O. centrated sulphuric acid with a yellow colour : 0.1138 gave 0.2992 CO and 0.0526 H20. 0.1150 , 4.3 C.C. N a t 17O and 764 mm. N=4-71. The b enzoyl derivative crystallised from alcohol in colourless 0.1104 gave 0.3070 CO and 0.0467 H20.0*1860 , 6 C.C. N2 a t 16O and 770 mm. N=$*81. C=71*70; H=5*13. C,,H,,O,N,~H,O requires C = 71.52 ; H = 5.30 ; N = 4-70 per cent. needles melting at. 178-179O : @=75.84; H=4*71, C2,H1703N,~H20 requires C= 75.80; H =4*94; N=3:84 per cent. Reaction with Dilute Sulphuric Acid. I n order to prove the constitution 2 grams of the above sulphate were boiled under reflux with about 30 C.C. of 10 per cent. sulphuric acid for two hours. The clear solution yielded an oil which solidified to a pale yellow crystalline mass. This was collected while hot washed with hot water and crystallised from alcohol, when it separated in colourless needles melting a t 224-225O (the melting point of 7-hydroxy-3-phenyl-2-methyl-y-benzopyrone is 226O; Jacobson and Ghosh loc. cit.).The substance developed a bluish-violet fluorescence in sulphuric acid whilst the imine gave a green fluorescence with sulphuric acid. The dilute sulphuric acid after the precipitate had been removed, was boiled with excess of sodium hydroxide when ammonia was evolved showing that the imine had been hydrolysed. 4-Zmino-7 8-dihydroxy-3-p?~ei~,yl-2-met hyl-y-benzomran, 011 0 NH This was prepared from pyrogallol (4 grams) acetylphenylaceto-nitrile (5 grams) and concentrated sulphuric acid (15 c.c.). The alcoholic solution of the yellow substance obtained was treated wit GHOSH A SYKTHESIS OF FLAVONES. 111 an excees of aqueous sodium acetate and the base was crystallised from dilute alcohol : 0.1243 gave 0.3178 CO and 0.0556 H,O. C=69*73; H=4*97. 0.0936 , 4 C.C.N a t 19O and 764 mm. N=4*93. Cl,H1303N,~H,0 requires cT= 69.56 ; H = 5-07 ; N = 5-07 per cent. The compound forms dark brown prisms which shrink a t 125O and melt and decompose a t 142O; it gives a green colour with alcoholic ferric chloride reduces ammoniacal silver nitrate on warm-ing and gives a brown sodium salt with sodium hydroxide solution. The acetyl derivative crystallises from alcohol in pale yellow needles melting a t 194O : O.I-271 gave 0.3176 CO and 0.0540 H,O. C=68.18; H=4.72. 0.1016 , 3.6 C.C. N2 a t 15O and 766 mm. N=4.18. C,,Hl70,N requires C = 68.37 ; H = 4.84 ; N = 3.98 per cent. Reaction with Dilute Sulphuric Acid. Two grams of the above base were boiled with 25 C.C. of 10 per cent. sulphuric acid for four hours. The substance which separated crystallised from acetic acid in yellow prisms melting a t 268O; the acetyl derivative melted a t 212O (the melting point of 7 8-dihydr-oxy-3-phenyl-2-methyl-y-benzopyrone is 268O and that of the acetyl derivative 2 loo).4-74in o-3-ph eny l-2-m e t h y 1-1 4-a-naph t hapyran Sz6lph.a t e, This was prepared from a-naphthol (3 grams) acetylphenylaceto-nitrile (5 grams) and concentrated sulphuric acid (10 c.c.). On pouring the mixture on ice the product separated as a viscous mass, which solidified under water. It crystallised from acetic acid con-taining dilute sulphuric acid in bright orange needles melting and decomposing a t 174O : 0.1164 gave 0.2668 CO and 0.0460 H,O. The base prepared in the usual way crystallised from dilute C= 62.51 ; H=4.40.C20Hl,0N,H2S0 requires C = 62.66 ; H;= 4.43 per cent. alcohol in pale brown prisms melting a t 162O 112 GHOSH A SYNTHESIS OF FLAVONES. 0.1172 gave 0.3503 CO and 0.0585 H,O. 0.0966 , 4.1 C.C. N a t 19O and 764 mm. N=4*90. C=81*52; H=5*55. C,,Hl,0N,~H20 requires C = 81.63 ; H = 5.44 ; N = 4.76 per cent. Reaction with Dilute Sulphuric Acid. One gram of the naphthapyran sulphate was boiled for three hours with 10 C.C. of 10 per cent. sulphuric acid. The product crystallised from alcohol in lemon-yellow needles melting a t 209O (the melting point of 3-phenyl-2-methyl-l 4-a-naphthapyrone is 209O; Jacobson and Ghosh loc. cit.). 4-Imisz 0-3-phe?zyl-2 7-dimet hyl- y-b e n zopyran Sulphat e, 70.4 0 This was prepared from 4 grams of m-cresol 5 grams of acetyl-phenylacet,onitrile and 15 C.C.of concentrated sulphuric acid. The semi-solid mass obtained on pouring the mixture on ice was dis-solved in acetic acid and precipitated with dilute sulphuric acid, when it separated in yellow prisms melting and decomposing at 122O. The yield was small: 0.1152 gave 0.2484 CO and 0.0489 H,O. C=58-80; H=4*71. The base crystallised from dilute alcohol in short colourless I n concentrated sulphuric acid it dissolved CliH1,ON,H,SO4 requires C= 58-78 ; H =4*89 per cent. needles melting a t 89O. with a yellow colour but without any fluorescence: 0.1054 gave 0.3168 CO and 0.0574 H,O. 0.1017 ,? 5 C.C. N a t 20° and 764 mm. N=5*65. C=81*97; H=6.05. C,.,H,,ON requires C = 81-92 ; H = 6-02 ; N = 5.62 per cent. 3-Phemyl-2 7-dimet hyl-2-b en zopyrone, 0 Two grams of the above sulphate were boiled with 20 C.C.of 10 per cent. sulphuric acid for four hours. After cooling the solid substance was crystallised from alcohol from which i t separated in colourless needles melting a t 155O GHOSH A SYNTHESIS OF FLAVONES. 113 0.1206 gave 0.3610 CO and 0.0631 H,O. The compound is readily soluble in alcohol benzene or acetic acid sparingly so in light petroleum and insoluble in water. I n concentrated sulphuric acid i t dissolves t o a colourless solution but develops a bright blue fluorescence. C=81*63; H=5.81. CI7H,,O requires C = 81.60 ; H = 5-60 per cent. C o ?L de ?a sa t io n s w i t IL F o r m y l p h e ?a y la c e t o n i t r i I e. PreparatiotL of Fol.mylphenylaEetonitrile.Ethyl formate (40 grams) was added in small portions a t a time to a well-cooled solution of phenylacetonitrile (58.5 grams) and sodium (11.5 grams) in alcohol when a gelatinous precipitate separated which after keeping a t the ordinary temperature for one hour was boiled on a water-bath for two hours. Alcohol was then removeld as far as possible by distillation. The residue was dissolved in water and after the removal of unchanged phenyl-acetonitrile with ether was acidified when formylphenylacetonitrile separated as a colourless crystmallins mass. Ths yield of the pure product was about 90 per cent. of the theoretical (compare Walther and Schickler J . p. C'hem. 1897 [ii] 55 308). 4-1mit2 0-7-11 ydroxy-3-ph cn yl- y-b e n zopyra n Hydrochloride, ? 0 I *H, Phosphoryl chloride (20 c.c.) was added to resorcinol (3 grams) and f ormylplienylacetonitrile (4 grams) when the reaction started slowly a t the ordinary temperature.The mixture which reacted vigorously on warming a t 50° was cooled and treated with pow-dered ice. The yellow granular mass crystallised from acetic acid containing dilute hydrochloric acid in short dull yellow needles melting and decomposing a t 215O after shrinking a t 205O. The yield was almost quantit-ative : 0*1100 gave 02484 CO and 0.0466 H,O. C = 61.58; H = 4.70. 0.1338 .. 5.6 C.C. N a t ZOO and 768 mm. N=4*64. C,,H,,02N,HCI,H,0 requires C = 61.85 ; H = 4.81 ; N==4*81 per cent. The salt gives a violet colour with alcoholic ferric chloride and VOL. CIX. 114 GHOSH A SYNTHESIS OF FLAVONES.dissolves in aqueous sodium hydroxide giving a yellowish-brown solution which evolves ammonia on warming. The base csystallised from dilute alcohol in bright yellow needles melting a t 226O: 0*1020 gave 02631 CO and 0.0478 H20. C = 70.32 ; H = 5-20. 0'1254 , 6.2 C.C. N a t 20° and 764 mm. N=5.68. The me tyl derivative crystallises from alcohol in colourless needles melting and decomposing a t 186O: 0.1058 gave 0.2840 CO and 0-0438 H20. C=73*20; H=4*60. The benzoyl derivative crystallises from light petroleum in colour-0-1060 gave 0.2862 CO and 0.0458 H,O. U-1740 , 6 C.C. N a t 16O and 770 mm. N=4*07. The perchlorate separates in bright yellow needles : 0.1222 required 3.6 C.C. N / 10-NaOH. Cl,Hl102N,H20 requires C = 70.54 ; H = 5.09 ; N = 5.49 per cent.C17H1303N requires C = 73.11 ; H = 4.66 per cent. less needles melting a t 136O: C=73*63; H=4.80. C2,H150,N,H,0 requires C = 73.73 ; H = 4-93 ; N = 3-90 per cent. HC10 = 29.4. Cl5Hl1O,N,HC1O4 require HClO = 29.6 per cent. Reaction with Dilute Sulphuric Acid. Two grams of the above hydrochloride were boiled with 20 C.C. of 10 per cent. sulphuric acid for three hours. The product crystal-lised from dilute alcohol in pale yellow needles melting at 130° (the melting point of 7-hydroxy-3-phenyl-y-benzopyrone is 13 1 O, Jacobson an6 Ghosh loc. cit.). 4-Imino-7 8-d& ydroxy-3-phenyl- y-b enzopyran, OH 0 Four grams of pyrogallol 4 grams of f ormylphenylacetonitrile, and 20 C.C. of phosphoryl chloride w0re heated a t 50-60° for half an hour.The alcoholic solution of the yellow product was treated with an excess of aqueous sodium acetate and the precipi-tate crystallisd from dilute acetic acid from which it separated in yellow needles melting a t 220° after shrinking a t 208O: 0.1009 gave 0.2462 CO and 0.0459 H20. 0.3000 , 13.4 C.C. N at 1 8 O and 764 mm. N=5*15. C = 66.54 ; H = 5.06. C1,H,,O,N,H20 re8quirea C = 66-42 ; H =4*79 ; N = 5-16 per cent GHOSH A SYNTHESIS OF FLAVONES. 1.15 The compound gives a green colour with alcoholic ferric chloride The acetyl derivative crystallised from dilute alcohol in almost 0*1001 gave 09354 CO and 0.0418 R,O. C = 64*i3 ; H=4.63. 0.2448 , 8 C.C. N at 15O and 764 mm. N=4*19. C,,H,,O,N,H,O requires C = 64-22 ; H = 4.78 ; N= 3.94 per cent. and reduces ammoniacal silver nitrate solution on warming.colourless needles melting a t 171O : 7 8-Dihydros:y-3-phenyl-y-benzopyrone, OH 0 Three grams of 4-imino-7 8-dihydroxy-3-phenyl-y-benzopyran were boiled under reflux with 30 C.C. of 10 per cent. sulphuric acid for three hours. 'The product crystallised from dilute alcohol in pale yellow needles melting a t 215O: 0.1194 gave 0.2729 CO and 0.0489 H,O. C=62.29; H=4.55. The compound is soluble in aqueous sodium hydroxide giving a brown solution. In concentrated sulphuric acid it gives a yellow solution and develops a green fluorescence. It reduces ammoniacal silver nitrate solution on warming. The acetyl derivative crystallised from alcohol in colourless needles melting a t 184O. It was also prepared by boiling the imino-compound with acetic anhydride and fused sodium acetate : Cl,H,40,,H,0 requires C = 64.04 ; H = 4.49 per cent.C,,H,,04,2H,0 requires C = 62.07 ; H = 4.82 per cent. 0.1049 gave 0.2510 CO and 0.0432 H,O. C=64.38; H=4*57. 4-Imino-3-ph eny 1-1 4-a-naph t hapyran Hydrochloride, /\ This was prepared by passing a current of dry hydrogen chloride i n b a well-cooled solution of 3 grams of a-naphthol and 3 grams of formylphenylacetonitrile in 6 C.C. of glacial acetic acid for about an hour. The product after two days was precipitated with dilute hydrochloric acid and was crystallised from acetic aci 116 GHOSH A SYNTHESIS OF FLAl'ONES. containing hydrochloric acid from which it separated in yellow needles melting and decomposing a t 135O : 0.1184 gave 0.3044 CO and 0.0533 H,O.C=70.11; H=5.01. The base crystallised from alcohol in dull yellow prisms which I n dilute alcoholic solution it gave a ClQH,,0N,HC1,H20 requires C = 70.15 ; H = 4-92 per cent. decomposed a t 115-130°. green fluorescence : 0-1272 gave 0.3760 CO and 0.0626 H20. C = 78.68; H= 5-47. 0-1680 , 7.4 N a t 20° and 764 mm. N=5.03. C,,H,,ON,H,O requires C = 78.89 ; H = 5.19 ; N = 4.84 per cent. 3-P h e nyl- 1 4-a-naph t ha pyr o n e, /\ i I 0 This substance prepared in the usual way from the imino-com-pound crystallised from alcohol in pale yellow needles melting a t 169-170O. I n concentrated sulphuric acid i t dissolved with a yellow colour and developed a dark green fluorescence: 0.1094 gave 0.3264 CO and 0.0470 H,O. C=81*36; H=4*77.C,,H,,O,,~H,O requires C = 81 * 14 ; H = 4.63 per cent. 4-lmirzo-7-hydroxy-3-phenyl-5-m e thyl- y-b e II zopyru n , 0 The product obtained by warming oscinol (2 grams) formyl-phenylacetonitrile (2 grams) and phosphoryl chloride (20 c.c.) and then treating the alcoholic solution with an excess of aqueous sodium acetate crystallises from alcohol in bright yellow needles which shrink a t 227O melt a t 235O and decompose a t 241O. It gives a yellowish-green colour with alcoholic ferric chloride : 0.1006 gave 0.2630 CO and 0.0486 H,O. 0.1308 , 6-2 C.C. N a t 20° and 760 mm. N=5*43. C,,H,,O,N,H,O requires C = 71.64 ; H = 5-22 ; N = 5.22 per cent. C= 71.30 ; H = 5.36 GHOSH A SYNTHESIS OF FLAVONES. 117 This substance crystallised from alcohol in pale yellow needles melting a t 224O after shrinking a t 218O.I n concentrated sul-phuric acid i t dissolved with a yellow colour and developed a dark green fluorescence. With ferric chloride i t gave no coloration : 0.1402 gave 0.3418 CO and 0.0696 H,O. C=66*48; H=5.51. C,,H,,0,,2H20 requires C = 66.66 ; H = 5.55 per cent. Co ?zd e ?z sa t i o n s with B e I L z o y I p h e n y l a c e t o nit r il e. Preparation of BeiLzoyl~henylacetonitrile. This was prepared in the same way as formylphenylacetonitrile from 11-5 grams of sodium dissolved in alcohol 58 grams of phenyl-acetonitrile and 75 grams of ethyl benzoate. The addition of ethyl benzoate must be gradual and the mixture must be shaken vigor-ously after each addition. The yieId was about 45 per cent. of the theoretical.Walt’her and Schickler (Zoc. c i t . ) by usii?g sodium methoxide free from alcohol and water obtained a yield of 10 per cent. of the theoretical. I N 13, This was prepared by passing a slow current of dry hydrogen cl1lorid0 into a well-cooled solution of 3 grams of resorcinol and 5 grams of benzoylphenylacetonitrile in 5 C.C. of glacial acetic acid f o r five t o six hours. The pink-coloured needles which separated were collected after three days washed with ether and kept on porous earthenware in a vacuum over potassium hydroxide. They shrink a t 28d0 and melt a t 286O: 0.1272 gave 0.3206 CO and 0.0583 H,O. C=68.73; H=5.09. C2,H,,0,N,HCl,H20 requires C = 68.66 ; H = 4-95 per cent 118 (XHOSH A SYNTHESIS OF FLAVONES. The compound gives a yellowish-brown coloration with alcoholic The base crystallised from alcohol in yellow needles melting 0.1206 gave 0.3546 CO and Om05O0 H,O.0.2164 )) 8.5 C.C. N a t 20° and 760 mm. N=4.49. The acet yl derivative crystallises from alcohol in colourless 0.1064 gave 0.2951 CO and 0.0458 H,O. 0.2280 ,? 8 C.C. N2 a t 20° and 758 mm. N=4*00. ferric chloride. at' 290O: C=80*19; H=4*60. C,,H,,O,N requires C = 80.50 ; H =4.60 ; N = 4.47 per cent. needles melting at 215O: C=75*62; H=4*78. C,,H,,O,N,~H,O requires C = 75.82 ; H'= 4-94 ; N = 3.83 per cent. 7-Hydroxy-2 3-diphenyl-y-b enzopyrone, 0 This was prepared by boiling 3 grams of 4-imino-7-hydroxy-2 3-diphenyl-y-benzopyran hydrochloride with 30 C.C. of 10 per cent. sulphuric acid for two hours. The product separated from alcohol in colourless needles melting a t 288O: 0.1112 gave 0.3262 CO and 0.0458 H,O.The acetyl derivative crystallises from acetic acid in colourless 0.1142 gave 0-3096 CO and 0.0490 H,O. C=80.00; H=4.57. C,,H,,O requires C?= 80.25 ; H = 4.45 per cent. leaflets melting a t 222O: C=73-93; H=4.76. C23H,,0,,H,0 requires C = 73.80 ; H = 4.81 per cent. Decomposition with Potassium Hydroxide. I n order to prove the constitution 3 grams of 7-hydroxy-2:3-diphenyl-y-benzopyrone were boiled with a solution of 15 grams of potassium hydroxide in 30 C.C. of water for four hours. The solu-tion after dilution was extracted with ether which on evaporation gave deoxybenzoin. The alkaline solution from which the ketone had been removed was acidified with dilute sulphuric acid and the precipitate collected.The acid solution on extraction with ether furnished resorcinol. The precipitate obtained after acidification of the alkaline solu-tion was dissolved in aqueous sodium hydronide and a current of carbon dioxide was passed through in order to liberate any un-changed original material. The clear solution was acidified an GHOSH A SYNTHESIS OF FLAVONES. 119 extracted with ether which on evaporation gave a mixture of a solid and an oil the latter being eliminated by treatment witlh ammonia in which it was insoluble. The ammoniacal solution on acidification gave a solid which crystallised from hot water in colourless needles melting a t 204O (the melting point of B-resorcylic acid is 204-206O). 4-Imino-7 8-hydroxy-2 3-d~pherayl-y-benzo~rai~, OH 0 This was prepared from 3 grams of pyrogallol and 5 grams of benzoylphenylacetonitrile.After five days water was added and the precipibate dissolved in alcohol and treated with aqueous sodium acetate. It then crystallised from dilute alcohol in yellow needles melting a t 179-180° : 0.1200 gave 0.3285 CO and 0.0510 H,O. C = 74.65 ; H =4.72. 0.1564 , 5.8 C.C. N a t 1 7 O and 764 mm. N=4.30. C,,H,,O,N,&H,O requires @= 74.55 ; H= 4.73 ; N = 4-14 per cent. The compound gave a green colour with alcoholic ferric chloride, and dissolves on warming in aqueous sodium hydroxide t o a yellow solution with the evolution of ammonia. 7 8-Dihydroxy-2 3-diphenyl-y-benzopyronze, OH 0 Thia substance prepared in the usual way from the imino-com-pound crystallised from alcohol in colourless needles melting a t 185O : 0.1245 gave 0.3470 CO and 0.0478 H,O.The compound is soluble in alcohol acetic acid or benzene but insoluble in light petroleum or water. It gives a green colour with alcoholic ferric chloride and reduces ammoniacal silver nitrate solution on warming. It does not fluoresce with concentrated sulphuric acid. C = 76.01 ; H=4.26. C2,HI4O6 requires C = 76-36 ; H = 4-24 per cent 120 GHOSH A SYNTHESIS OF FLAVOKES. -1 ttempted Condensation of Resorcitiol ajid Benzoylphenyl-ace t onit ril e. A mixture of benzoylphenylacetonitrile (3 grams) resorcinol (2 grams) and concentrated sulphuric acid (10 c.c.) after being kept overnight was poured on ice. The product crystallised from alcohol in colourless needles melting a t 177-178O (the melting point of benzoylphenylacetamide is 177-1 78O ; Waltlier and Schickler Zoc.c z t ) . Calc. C = 75.31 ; H = 5.43 per cent.) (Found C= 75.33 ; H = 5.43. Co n d e n s a t i o ?a s with B e n z o y l n c e t o rL i t r i I e. 4-Imino-7- h y dr o xy-2-ph enyl- y -b e?h z opyra<n, Concentrated sulphuric acid (15 c.c.) was slowly added t o benzoyl-acetonitrile (3.5 grams) and resorcinol (3.5 grams) and the cooled mixture afker being kept overnight was poured on ice. The pro-duct obtained after treating an alcoholic solution of the separated solid with aqueous sodium acetate crystallised from alcohol in yellowish-brown prisms decomposing a t 185-235O : 0.1104 gave 0.2954 CO and 0-0510 H,O.0.1194 , 6.1 C.C. N at 1 7 O and 768 mm. N=5.99. The compound is soluble in alcohol o r acetic acid sparingly so in benzene and insoluble in light petroleum. It gives a violet colora-tion with alcoholic ferric chloride and in dilute alcoholic solution shows a green fluorescence. The p’crate separated in bright orange needles melting a t 238O : 0.1138 gave 12 C.C. N a t 17O and 760 mm. C=72*97; H=5.13. C,,H,,0,N,4H20 requires C’= 73-17 ; H = 4.88 ; N = 5.99 per cent. N=12.33. C,,H,,O2N,C,H3O7N3 requires N = 12.01 per cent. 7-ligdroxy-2-phem~l-y-b enzopyrone, 0 Three grams of the above imino-pyran were boiled under reflux with an excess of 5 per cent,. sodium hydroxide solution until n GHOSH A SYNTHESIS OF FLAVOHES. 131 more ammonia was evolved.The solution was acidified with dilute sulphuric acid and the product crystallised from alcohol in colour-less needles melting a t 243O (Emilewicz and Kostanecki Bey. 1898, 31 703 give 240O). Decomposition zvith Potassium Hydroxide. I n order to prove the constitution 4 grams of 4-imino-7-hydroxy-2-phenyl-y-benzopyran were boiled with a solution of 18 grams of potassium hydroxide iii 35 C.C. of water for three hours. The solu-tion was extracted with ether and on evaporation of the ethereal solution an oil was obtained which was identified as acetophenone. The alkaline solution was treated in the same way as described in the case of the decomposition of 7-hydroxy-2 S-diphenyl-y-benzo-pyrone. The products obtained from the solution were resorcinol and P-resorcylic acid.4Jmino-2-phenyl-1 4-a-naphthapyran Sulphate, H, This was prepared from benzoylacetonitrile (4 grams) a-naphthol (4 grams) and concentrated sulphuric acid (20 c.c.). The product, separated by partly neutralising the clear acid solution obtained by pouring the mixture on ice crystallised from alcohol containing dilute sulphuric acid in prisms which decomposed a t 185-215O. The yield was small : 0.1100 gave 0.2486 CO and 0.0386 H,O. 0.2460 , 8 C.C. N2 a t 16O and 762 mm. N=3*80. The base crystallised from alcohol in pale brown prisms decom-0.1093 gave 0.3152 GO and 0.0505 H 2 0 . 0.1047 , 4.8 C.C. N a t 20° and 764 mm. N=5.27. The picrate formed orange needles melting and decomposing a t 0.1230 gave 12.2 C.C. N2 a t 1 7 O and 764 mm. VOL. CIX.G C=61-63; H=3*89. C,,H,,0N,H2S0 requires C= 61-78 ; H = 4-06 ; N = 3-80 per cent. posing a t 138-148O: C=78*65; H=5-12. C19H130N requires @= 78.89 ; H = 4-80 ; N = 5.1 7 per cent. 274O : N=ll-50. C,,H,,0N,C,H307N requires N = 11.20 per cent 122 RAY AND DE MOLECULAR VOLUMES OF THE Treatment with Dilute Sodium Hydroxide. Three grams of the above imino-compound were boiled with 30 C.C. of 5 per eent. sodium hydroxide solution until the evolution of ammonia ceased. The solution was then acidified when a solid mixed with some resinous material separated. This product after repeated crystallisations from dilute alcohol separated in colour-less needles melting a t 153O (the melting point of 2-phenyl-1 4-a-naphthapyrone is 154-156O; Kostanecki Ber. 1898 31 707).The yield was very small. 4-Imino-7 ; 8-dihydroxy-2-phenyl-y-b enzopyran., OH 0 This was prepared from 4 grams of benzoylacetonitrile 4 grains of pyrogallol and 20 C.C. of concentrated sulphuric acid. The alcoholic solution of the viscous product obtained on pouring the acid mixture on ice was treated with an excess of aqueous sodium acetate ; the substance which separated crystallised from alcohol in brown prisms decomposing a t 145-165O : 0-1064 gave 0-2596 CO and 0.0428 H,O. C,,H,,O,N,H,O requires C = 66.42 ; H = 4.79 per cent. The compound gives a red solution with aqueous sodium hydr-oxide develops a green colour with alcoholic ferric chloride and reduces ammoniacal silver nitrate solution on warming. C= 66-54 ; H=4.57. ORQANIC LABORATOBY, UNIVERSITY COLLEGE, LONDON.[Received November 30th 1915. GHOSH A SYNTHESIS OF FLAVONES. 105 IX .-A Synthesis of E’lauones. By BROJENDRA NATH GHOSH. IT has been shown (Jacobson and Ghosh T. 1915 107 1051) that the condensation of ethyl a-phenylacetoacetate with phenols leads t o the production of substituted y-benzopyrones o r flavones. It is shown in the present paper that if instead of the ester th 106 GHOSH A SYNTHESIS OF FLAVONES. nitriles are employed an imide is produced from which the flavone is obtained by boiling with dilute sulphuric aicd: 0 0 NK where X may be either H CH, or C,H,. lysing the imine formed the reaction proceeded as follows: By condensing benzoylacetonitrile and phenols and then hydro-0 0 All the imines dissolve in dilute alcohol giving yellow solutions which show a green fluorescence ; in concentrated sulphuric acid, also yellow solutions are obtained showing in some cases a green fluorescence.On boiling with acetic anhydride and sodium acetate the acetyl derivatives of the corresponding flavones were obtained the imino-group being evidently hydrolysed during the opera tion. That the nitrile is more reactive than the corresponding ester is observed in the case of the condensation of acetylphenylacetonitrile. It has been shown (Zoc. cit.) that ethyl a-phenylacetoacetate and m-cresol do not condense in the presence of concentrated sulphuric acid but acetylphenylacetonitrile and m-cresol under the same con-ditions give the corresponding imine from which the flavone is obtained by boiling with dilute sulphuric acid.The condensation of the nitrile with phenol proceeds best when there is a methyl or preferably a hydroxyl group as substituent in the phenol but with any other substituent the reaction does not take place. With the unsubstituted phenol the yield is very small. As regards the position of the substituent the condensation is favoured when the substituent group occupies the meta-position with respect to the phenolic hydroxyl group but there is always an inhibiting influence when both the meta-positions are occupied. Thus resorcinol and pyrogallol give with acetylphenylacetonitrile a very good yield of the corresponding condensation product but under the same conditions phloroglucinol and the nitrile d o no GEIOSH A SYNTHESIS OF FLAVONES.107 condense a t all. When the substituent group occupies a position other than the meta the reaction does not take place under the conditions employed. Thus with either o-cresol or quinol and acetylphenylacetonitrile the reaction does not take place in the presence of concentrated sulphuric acid. Four condensing agents were employed in these experiments, namely concentrated sulphuric acid hydrogen chloride phosphoryl chloride and anhydrous zinc chloride. I n the case of acetylphenyl-acetonitrile and phenols concentrated sulphuric acid was the only condensing agent employed with satisfactory results. In the case of f ormylphenylacetonitrile and phenols the product obtained by using concentrated sulphuric acid formed a finely divided, amorphous precipitate which was very difficult to manipulate but by using phosphoryl chloride and dry hydrogen chloride the reactions proceeded satisfactorily.It has been observed (Zoc. cit .) that ethyl a-phenylformylacetate does not condense with phenols in the presence of cold concentrated sulphuric acid probably owing to the existence of the ketonic form a t the ordinary temperature, but formylphenylacetonitrile gives a violet coloration with ferric chloride which shows that a t least a portion of it exists in the enolic form and the hope was entertained that it might react with the aid of sulphuric acid or a similar reagent a t the ordinary temperature. As a matter of fact the condensation takes place a t the ordinarj. temperature with hydrogen chloride and even with phosphoryl chloride.I n the case of benzoylphenylacetonitrile and phenols when con-centrated sulphuric acid was used as the condensing agent the nitrile simply combined with a molecular proportion of water, giving the corresponding amide. With anhydrous zinc chloride the condensation of the nitrile and phenol did n o t take place. The condensation was however brought about by dry 'hydrogen chloride. I n the condensation of benzoylacetonitrile and phenols concen-trated sulphuric acid was the only condensing agent employed In the case of resorcinol a satisfactory yield of the imino-compound was obtained but in the case of a-naphthol or pyrogallol the yield of the corresponding imine was small and with phloroglucinol did not react. Some of the substances here described are coloured and wme are colourless but they all give coloured salts.These facts can be explained on the assumption that those substances which are coloured have a quinonoid structure. The constitutions of the salts may be represented by one of the following formulae 108 GHOSH A SYNTHESIS OF FLAVONES. H HSO, \/ I1 HSO, \/ 0 0 0 0 and those of the free bases by 0 0 0 (VI- 1 (VII. ) Forms I 11 and I11 should all yield a base of the constitution shown in VI. Forms I 11 and VI may a t once be rejected because they do not contain any quinonoid arrangement. Any one of the formuke represented by 111 IV o r V may represent the structure of the salts and in the free state V I I ought t o be the proper representation of the compounds because they all contain the quinonoid arrangement.I n order to test this view m-cresol was condensed with acetylphenylacetonitrile and the free base obtained from this condensation was found to be quite colourless. In this case the base must have the structure No quinonoid arrangement is possible here and the absence of colour is in accordance with the view that the coloured compound CtHOSH A SYNTHESIS OF FLAVONES. 109 have the quinonoid structure. This substance dissolves in concen-trated sulphuric acid with a yellow colour and gives a yellow sulphate and therefore the constitution of this salt must be repre-sented by 111. Thus it becomes evident that formula I11 ought to be the proper representation of the salts whilst when they are converted into their bases rearrangement into the quinonoid struc-ture (VII) takes place when a hydroxyl group is present but they retain their structure (VI) when no such arrangement is possible.E X P E R I M E N T A L . Condensations with Acetylphenylacetonitrile. 4-lmino-7-hydroxy-3-phenyl-2-meth yl-y-b e rzzopyran Sulpha te, HTO, Concentrated sulphuric acid (15 c.c.) was slow,y added to resorcind (4 grams) and acetylphenylacetonitrile (5 grams) (Reckh, Rer. 1898 31 3161). When kept overnight the crystalline mass obtained from the cooled solution was poured on ice. After three hours the clear solution yielded a voluminous crystalline precipi-tate which was crystallised from very dilute sulphuric acid. The yield was almost quantitative : 0.1302 gave 0.2630 CO and 0.0538 H,O.0.1715 , 6.4 C.C. N2 a t 21° and 768 mm. N=4.23. C,,H,,02N,H2S04 requires C = 55-01 ; H =4*50 ; N = 4.00 per cent. The salt forms short yellow needles melting to a dark red liquid a t 246O; it dissolves in dilute sodium hydroxide solution and in concentrated sulphuric acid to yellow solutions the latter show-ing a green fluorescence which is also exhibited in dilute alcoholic solution. The base prepared by treating an alcoholic solution of the sulphate with an excess of aqueous sodium acetate crystallised from dilute alcohol in brown prisms melting and decomposing a t 221O : 0.1214 gave 0.3300 CO and 0.0628 H,O. C'=74*15; H=5.55. 0.0980 , 4.8 C.C. N2 a t 15O and 752 mm. N=5*67. C,,H,,0,N,~H20 requires C = 73.85 ; H = 5.38 ; N = 5-38 per cent.The p-crate separated in orange yellow needles melting a t 227O: C=55*10; H=4.59. Alcoholic ferric chloride develops a violet coloration 110 CtHMH A SYNTHESIS OF FLAVONES. 0.0760 gave 8.6 C.C. N a t 17O and 764 mm. The perchlorate separated in yellow needles which on heating 0.1232 required 3.5 C.C. N/lO-NaOH. The acetyl derivative crystallised from very dilute alcohol in It dissolved in con-N=11*66. C16H130,N,~6H307N3 requires N = 11 * 66 per cent. deco,mposed with explosive violence : HC10 = 28.4. CIGH,,O,N,HC1O requires HClO = 28-5 per cent. very faintly yellow needles melting a t 135O. centrated sulphuric acid with a yellow colour : 0.1138 gave 0.2992 CO and 0.0526 H20. 0.1150 , 4.3 C.C. N a t 17O and 764 mm. N=4-71. The b enzoyl derivative crystallised from alcohol in colourless 0.1104 gave 0.3070 CO and 0.0467 H20.0*1860 , 6 C.C. N2 a t 16O and 770 mm. N=$*81. C=71*70; H=5*13. C,,H,,O,N,~H,O requires C = 71.52 ; H = 5.30 ; N = 4-70 per cent. needles melting at. 178-179O : @=75.84; H=4*71, C2,H1703N,~H20 requires C= 75.80; H =4*94; N=3:84 per cent. Reaction with Dilute Sulphuric Acid. I n order to prove the constitution 2 grams of the above sulphate were boiled under reflux with about 30 C.C. of 10 per cent. sulphuric acid for two hours. The clear solution yielded an oil which solidified to a pale yellow crystalline mass. This was collected while hot washed with hot water and crystallised from alcohol, when it separated in colourless needles melting a t 224-225O (the melting point of 7-hydroxy-3-phenyl-2-methyl-y-benzopyrone is 226O; Jacobson and Ghosh loc.cit.). The substance developed a bluish-violet fluorescence in sulphuric acid whilst the imine gave a green fluorescence with sulphuric acid. The dilute sulphuric acid after the precipitate had been removed, was boiled with excess of sodium hydroxide when ammonia was evolved showing that the imine had been hydrolysed. 4-Zmino-7 8-dihydroxy-3-p?~ei~,yl-2-met hyl-y-benzomran, 011 0 NH This was prepared from pyrogallol (4 grams) acetylphenylaceto-nitrile (5 grams) and concentrated sulphuric acid (15 c.c.). The alcoholic solution of the yellow substance obtained was treated wit GHOSH A SYKTHESIS OF FLAVONES. 111 an excees of aqueous sodium acetate and the base was crystallised from dilute alcohol : 0.1243 gave 0.3178 CO and 0.0556 H,O.C=69*73; H=4*97. 0.0936 , 4 C.C. N a t 19O and 764 mm. N=4*93. Cl,H1303N,~H,0 requires cT= 69.56 ; H = 5-07 ; N = 5-07 per cent. The compound forms dark brown prisms which shrink a t 125O and melt and decompose a t 142O; it gives a green colour with alcoholic ferric chloride reduces ammoniacal silver nitrate on warm-ing and gives a brown sodium salt with sodium hydroxide solution. The acetyl derivative crystallises from alcohol in pale yellow needles melting a t 194O : O.I-271 gave 0.3176 CO and 0.0540 H,O. C=68.18; H=4.72. 0.1016 , 3.6 C.C. N2 a t 15O and 766 mm. N=4.18. C,,Hl70,N requires C = 68.37 ; H = 4.84 ; N = 3.98 per cent. Reaction with Dilute Sulphuric Acid. Two grams of the above base were boiled with 25 C.C.of 10 per cent. sulphuric acid for four hours. The substance which separated crystallised from acetic acid in yellow prisms melting a t 268O; the acetyl derivative melted a t 212O (the melting point of 7 8-dihydr-oxy-3-phenyl-2-methyl-y-benzopyrone is 268O and that of the acetyl derivative 2 loo). 4-74in o-3-ph eny l-2-m e t h y 1-1 4-a-naph t hapyran Sz6lph.a t e, This was prepared from a-naphthol (3 grams) acetylphenylaceto-nitrile (5 grams) and concentrated sulphuric acid (10 c.c.). On pouring the mixture on ice the product separated as a viscous mass, which solidified under water. It crystallised from acetic acid con-taining dilute sulphuric acid in bright orange needles melting and decomposing a t 174O : 0.1164 gave 0.2668 CO and 0.0460 H,O.The base prepared in the usual way crystallised from dilute C= 62.51 ; H=4.40. C20Hl,0N,H2S0 requires C = 62.66 ; H;= 4.43 per cent. alcohol in pale brown prisms melting a t 162O 112 GHOSH A SYNTHESIS OF FLAVONES. 0.1172 gave 0.3503 CO and 0.0585 H,O. 0.0966 , 4.1 C.C. N a t 19O and 764 mm. N=4*90. C=81*52; H=5*55. C,,Hl,0N,~H20 requires C = 81.63 ; H = 5.44 ; N = 4.76 per cent. Reaction with Dilute Sulphuric Acid. One gram of the naphthapyran sulphate was boiled for three hours with 10 C.C. of 10 per cent. sulphuric acid. The product crystallised from alcohol in lemon-yellow needles melting a t 209O (the melting point of 3-phenyl-2-methyl-l 4-a-naphthapyrone is 209O; Jacobson and Ghosh loc. cit.). 4-Imisz 0-3-phe?zyl-2 7-dimet hyl- y-b e n zopyran Sulphat e, 70.4 0 This was prepared from 4 grams of m-cresol 5 grams of acetyl-phenylacet,onitrile and 15 C.C.of concentrated sulphuric acid. The semi-solid mass obtained on pouring the mixture on ice was dis-solved in acetic acid and precipitated with dilute sulphuric acid, when it separated in yellow prisms melting and decomposing at 122O. The yield was small: 0.1152 gave 0.2484 CO and 0.0489 H,O. C=58-80; H=4*71. The base crystallised from dilute alcohol in short colourless I n concentrated sulphuric acid it dissolved CliH1,ON,H,SO4 requires C= 58-78 ; H =4*89 per cent. needles melting a t 89O. with a yellow colour but without any fluorescence: 0.1054 gave 0.3168 CO and 0.0574 H,O. 0.1017 ,? 5 C.C. N a t 20° and 764 mm. N=5*65. C=81*97; H=6.05.C,.,H,,ON requires C = 81-92 ; H = 6-02 ; N = 5.62 per cent. 3-Phemyl-2 7-dimet hyl-2-b en zopyrone, 0 Two grams of the above sulphate were boiled with 20 C.C. of 10 per cent. sulphuric acid for four hours. After cooling the solid substance was crystallised from alcohol from which i t separated in colourless needles melting a t 155O GHOSH A SYNTHESIS OF FLAVONES. 113 0.1206 gave 0.3610 CO and 0.0631 H,O. The compound is readily soluble in alcohol benzene or acetic acid sparingly so in light petroleum and insoluble in water. I n concentrated sulphuric acid i t dissolves t o a colourless solution but develops a bright blue fluorescence. C=81*63; H=5.81. CI7H,,O requires C = 81.60 ; H = 5-60 per cent. C o ?L de ?a sa t io n s w i t IL F o r m y l p h e ?a y la c e t o n i t r i I e.PreparatiotL of Fol.mylphenylaEetonitrile. Ethyl formate (40 grams) was added in small portions a t a time to a well-cooled solution of phenylacetonitrile (58.5 grams) and sodium (11.5 grams) in alcohol when a gelatinous precipitate separated which after keeping a t the ordinary temperature for one hour was boiled on a water-bath for two hours. Alcohol was then removeld as far as possible by distillation. The residue was dissolved in water and after the removal of unchanged phenyl-acetonitrile with ether was acidified when formylphenylacetonitrile separated as a colourless crystmallins mass. Ths yield of the pure product was about 90 per cent. of the theoretical (compare Walther and Schickler J . p. C'hem. 1897 [ii] 55 308).4-1mit2 0-7-11 ydroxy-3-ph cn yl- y-b e n zopyra n Hydrochloride, ? 0 I *H, Phosphoryl chloride (20 c.c.) was added to resorcinol (3 grams) and f ormylplienylacetonitrile (4 grams) when the reaction started slowly a t the ordinary temperature. The mixture which reacted vigorously on warming a t 50° was cooled and treated with pow-dered ice. The yellow granular mass crystallised from acetic acid containing dilute hydrochloric acid in short dull yellow needles melting and decomposing a t 215O after shrinking a t 205O. The yield was almost quantit-ative : 0*1100 gave 02484 CO and 0.0466 H,O. C = 61.58; H = 4.70. 0.1338 .. 5.6 C.C. N a t ZOO and 768 mm. N=4*64. C,,H,,02N,HCI,H,0 requires C = 61.85 ; H = 4.81 ; N==4*81 per cent. The salt gives a violet colour with alcoholic ferric chloride and VOL.CIX. 114 GHOSH A SYNTHESIS OF FLAVONES. dissolves in aqueous sodium hydroxide giving a yellowish-brown solution which evolves ammonia on warming. The base csystallised from dilute alcohol in bright yellow needles melting a t 226O: 0*1020 gave 02631 CO and 0.0478 H20. C = 70.32 ; H = 5-20. 0'1254 , 6.2 C.C. N a t 20° and 764 mm. N=5.68. The me tyl derivative crystallises from alcohol in colourless needles melting and decomposing a t 186O: 0.1058 gave 0.2840 CO and 0-0438 H20. C=73*20; H=4*60. The benzoyl derivative crystallises from light petroleum in colour-0-1060 gave 0.2862 CO and 0.0458 H,O. U-1740 , 6 C.C. N a t 16O and 770 mm. N=4*07. The perchlorate separates in bright yellow needles : 0.1222 required 3.6 C.C.N / 10-NaOH. Cl,Hl102N,H20 requires C = 70.54 ; H = 5.09 ; N = 5.49 per cent. C17H1303N requires C = 73.11 ; H = 4.66 per cent. less needles melting a t 136O: C=73*63; H=4.80. C2,H150,N,H,0 requires C = 73.73 ; H = 4-93 ; N = 3-90 per cent. HC10 = 29.4. Cl5Hl1O,N,HC1O4 require HClO = 29.6 per cent. Reaction with Dilute Sulphuric Acid. Two grams of the above hydrochloride were boiled with 20 C.C. of 10 per cent. sulphuric acid for three hours. The product crystal-lised from dilute alcohol in pale yellow needles melting at 130° (the melting point of 7-hydroxy-3-phenyl-y-benzopyrone is 13 1 O, Jacobson an6 Ghosh loc. cit.). 4-Imino-7 8-d& ydroxy-3-phenyl- y-b enzopyran, OH 0 Four grams of pyrogallol 4 grams of f ormylphenylacetonitrile, and 20 C.C.of phosphoryl chloride w0re heated a t 50-60° for half an hour. The alcoholic solution of the yellow product was treated with an excess of aqueous sodium acetate and the precipi-tate crystallisd from dilute acetic acid from which it separated in yellow needles melting a t 220° after shrinking a t 208O: 0.1009 gave 0.2462 CO and 0.0459 H20. 0.3000 , 13.4 C.C. N at 1 8 O and 764 mm. N=5*15. C = 66.54 ; H = 5.06. C1,H,,O,N,H20 re8quirea C = 66-42 ; H =4*79 ; N = 5-16 per cent GHOSH A SYNTHESIS OF FLAVONES. 1.15 The compound gives a green colour with alcoholic ferric chloride The acetyl derivative crystallised from dilute alcohol in almost 0*1001 gave 09354 CO and 0.0418 R,O. C = 64*i3 ; H=4.63. 0.2448 , 8 C.C. N at 15O and 764 mm. N=4*19.C,,H,,O,N,H,O requires C = 64-22 ; H = 4.78 ; N= 3.94 per cent. and reduces ammoniacal silver nitrate solution on warming. colourless needles melting a t 171O : 7 8-Dihydros:y-3-phenyl-y-benzopyrone, OH 0 Three grams of 4-imino-7 8-dihydroxy-3-phenyl-y-benzopyran were boiled under reflux with 30 C.C. of 10 per cent. sulphuric acid for three hours. 'The product crystallised from dilute alcohol in pale yellow needles melting a t 215O: 0.1194 gave 0.2729 CO and 0.0489 H,O. C=62.29; H=4.55. The compound is soluble in aqueous sodium hydroxide giving a brown solution. In concentrated sulphuric acid it gives a yellow solution and develops a green fluorescence. It reduces ammoniacal silver nitrate solution on warming. The acetyl derivative crystallised from alcohol in colourless needles melting a t 184O.It was also prepared by boiling the imino-compound with acetic anhydride and fused sodium acetate : Cl,H,40,,H,0 requires C = 64.04 ; H = 4.49 per cent. C,,H,,04,2H,0 requires C = 62.07 ; H = 4.82 per cent. 0.1049 gave 0.2510 CO and 0.0432 H,O. C=64.38; H=4*57. 4-Imino-3-ph eny 1-1 4-a-naph t hapyran Hydrochloride, /\ This was prepared by passing a current of dry hydrogen chloride i n b a well-cooled solution of 3 grams of a-naphthol and 3 grams of formylphenylacetonitrile in 6 C.C. of glacial acetic acid for about an hour. The product after two days was precipitated with dilute hydrochloric acid and was crystallised from acetic aci 116 GHOSH A SYNTHESIS OF FLAl'ONES. containing hydrochloric acid from which it separated in yellow needles melting and decomposing a t 135O : 0.1184 gave 0.3044 CO and 0.0533 H,O.C=70.11; H=5.01. The base crystallised from alcohol in dull yellow prisms which I n dilute alcoholic solution it gave a ClQH,,0N,HC1,H20 requires C = 70.15 ; H = 4-92 per cent. decomposed a t 115-130°. green fluorescence : 0-1272 gave 0.3760 CO and 0.0626 H20. C = 78.68; H= 5-47. 0-1680 , 7.4 N a t 20° and 764 mm. N=5.03. C,,H,,ON,H,O requires C = 78.89 ; H = 5.19 ; N = 4.84 per cent. 3-P h e nyl- 1 4-a-naph t ha pyr o n e, /\ i I 0 This substance prepared in the usual way from the imino-com-pound crystallised from alcohol in pale yellow needles melting a t 169-170O. I n concentrated sulphuric acid i t dissolved with a yellow colour and developed a dark green fluorescence: 0.1094 gave 0.3264 CO and 0.0470 H,O.C=81*36; H=4*77. C,,H,,O,,~H,O requires C = 81 * 14 ; H = 4.63 per cent. 4-lmirzo-7-hydroxy-3-phenyl-5-m e thyl- y-b e II zopyru n , 0 The product obtained by warming oscinol (2 grams) formyl-phenylacetonitrile (2 grams) and phosphoryl chloride (20 c.c.) and then treating the alcoholic solution with an excess of aqueous sodium acetate crystallises from alcohol in bright yellow needles which shrink a t 227O melt a t 235O and decompose a t 241O. It gives a yellowish-green colour with alcoholic ferric chloride : 0.1006 gave 0.2630 CO and 0.0486 H,O. 0.1308 , 6-2 C.C. N a t 20° and 760 mm. N=5*43. C,,H,,O,N,H,O requires C = 71.64 ; H = 5-22 ; N = 5.22 per cent. C= 71.30 ; H = 5.36 GHOSH A SYNTHESIS OF FLAVONES.117 This substance crystallised from alcohol in pale yellow needles melting a t 224O after shrinking a t 218O. I n concentrated sul-phuric acid i t dissolved with a yellow colour and developed a dark green fluorescence. With ferric chloride i t gave no coloration : 0.1402 gave 0.3418 CO and 0.0696 H,O. C=66*48; H=5.51. C,,H,,0,,2H20 requires C = 66.66 ; H = 5.55 per cent. Co ?zd e ?z sa t i o n s with B e I L z o y I p h e n y l a c e t o nit r il e. Preparation of BeiLzoyl~henylacetonitrile. This was prepared in the same way as formylphenylacetonitrile from 11-5 grams of sodium dissolved in alcohol 58 grams of phenyl-acetonitrile and 75 grams of ethyl benzoate. The addition of ethyl benzoate must be gradual and the mixture must be shaken vigor-ously after each addition.The yieId was about 45 per cent. of the theoretical. Walt’her and Schickler (Zoc. c i t . ) by usii?g sodium methoxide free from alcohol and water obtained a yield of 10 per cent. of the theoretical. I N 13, This was prepared by passing a slow current of dry hydrogen cl1lorid0 into a well-cooled solution of 3 grams of resorcinol and 5 grams of benzoylphenylacetonitrile in 5 C.C. of glacial acetic acid f o r five t o six hours. The pink-coloured needles which separated were collected after three days washed with ether and kept on porous earthenware in a vacuum over potassium hydroxide. They shrink a t 28d0 and melt a t 286O: 0.1272 gave 0.3206 CO and 0.0583 H,O. C=68.73; H=5.09. C2,H,,0,N,HCl,H20 requires C = 68.66 ; H = 4-95 per cent 118 (XHOSH A SYNTHESIS OF FLAVONES.The compound gives a yellowish-brown coloration with alcoholic The base crystallised from alcohol in yellow needles melting 0.1206 gave 0.3546 CO and Om05O0 H,O. 0.2164 )) 8.5 C.C. N a t 20° and 760 mm. N=4.49. The acet yl derivative crystallises from alcohol in colourless 0.1064 gave 0.2951 CO and 0.0458 H,O. 0.2280 ,? 8 C.C. N2 a t 20° and 758 mm. N=4*00. ferric chloride. at' 290O: C=80*19; H=4*60. C,,H,,O,N requires C = 80.50 ; H =4.60 ; N = 4.47 per cent. needles melting at 215O: C=75*62; H=4*78. C,,H,,O,N,~H,O requires C = 75.82 ; H'= 4-94 ; N = 3.83 per cent. 7-Hydroxy-2 3-diphenyl-y-b enzopyrone, 0 This was prepared by boiling 3 grams of 4-imino-7-hydroxy-2 3-diphenyl-y-benzopyran hydrochloride with 30 C.C.of 10 per cent. sulphuric acid for two hours. The product separated from alcohol in colourless needles melting a t 288O: 0.1112 gave 0.3262 CO and 0.0458 H,O. The acetyl derivative crystallises from acetic acid in colourless 0.1142 gave 0-3096 CO and 0.0490 H,O. C=80.00; H=4.57. C,,H,,O requires C?= 80.25 ; H = 4.45 per cent. leaflets melting a t 222O: C=73-93; H=4.76. C23H,,0,,H,0 requires C = 73.80 ; H = 4.81 per cent. Decomposition with Potassium Hydroxide. I n order to prove the constitution 3 grams of 7-hydroxy-2:3-diphenyl-y-benzopyrone were boiled with a solution of 15 grams of potassium hydroxide in 30 C.C. of water for four hours. The solu-tion after dilution was extracted with ether which on evaporation gave deoxybenzoin.The alkaline solution from which the ketone had been removed was acidified with dilute sulphuric acid and the precipitate collected. The acid solution on extraction with ether furnished resorcinol. The precipitate obtained after acidification of the alkaline solu-tion was dissolved in aqueous sodium hydronide and a current of carbon dioxide was passed through in order to liberate any un-changed original material. The clear solution was acidified an GHOSH A SYNTHESIS OF FLAVONES. 119 extracted with ether which on evaporation gave a mixture of a solid and an oil the latter being eliminated by treatment witlh ammonia in which it was insoluble. The ammoniacal solution on acidification gave a solid which crystallised from hot water in colourless needles melting a t 204O (the melting point of B-resorcylic acid is 204-206O).4-Imino-7 8-hydroxy-2 3-d~pherayl-y-benzo~rai~, OH 0 This was prepared from 3 grams of pyrogallol and 5 grams of benzoylphenylacetonitrile. After five days water was added and the precipibate dissolved in alcohol and treated with aqueous sodium acetate. It then crystallised from dilute alcohol in yellow needles melting a t 179-180° : 0.1200 gave 0.3285 CO and 0.0510 H,O. C = 74.65 ; H =4.72. 0.1564 , 5.8 C.C. N a t 1 7 O and 764 mm. N=4.30. C,,H,,O,N,&H,O requires @= 74.55 ; H= 4.73 ; N = 4-14 per cent. The compound gave a green colour with alcoholic ferric chloride, and dissolves on warming in aqueous sodium hydroxide t o a yellow solution with the evolution of ammonia. 7 8-Dihydroxy-2 3-diphenyl-y-benzopyronze, OH 0 Thia substance prepared in the usual way from the imino-com-pound crystallised from alcohol in colourless needles melting a t 185O : 0.1245 gave 0.3470 CO and 0.0478 H,O.The compound is soluble in alcohol acetic acid or benzene but insoluble in light petroleum or water. It gives a green colour with alcoholic ferric chloride and reduces ammoniacal silver nitrate solution on warming. It does not fluoresce with concentrated sulphuric acid. C = 76.01 ; H=4.26. C2,HI4O6 requires C = 76-36 ; H = 4-24 per cent 120 GHOSH A SYNTHESIS OF FLAVOKES. -1 ttempted Condensation of Resorcitiol ajid Benzoylphenyl-ace t onit ril e. A mixture of benzoylphenylacetonitrile (3 grams) resorcinol (2 grams) and concentrated sulphuric acid (10 c.c.) after being kept overnight was poured on ice.The product crystallised from alcohol in colourless needles melting a t 177-178O (the melting point of benzoylphenylacetamide is 177-1 78O ; Waltlier and Schickler Zoc. c z t ) . Calc. C = 75.31 ; H = 5.43 per cent.) (Found C= 75.33 ; H = 5.43. Co n d e n s a t i o ?a s with B e n z o y l n c e t o rL i t r i I e. 4-Imino-7- h y dr o xy-2-ph enyl- y -b e?h z opyra<n, Concentrated sulphuric acid (15 c.c.) was slowly added t o benzoyl-acetonitrile (3.5 grams) and resorcinol (3.5 grams) and the cooled mixture afker being kept overnight was poured on ice. The pro-duct obtained after treating an alcoholic solution of the separated solid with aqueous sodium acetate crystallised from alcohol in yellowish-brown prisms decomposing a t 185-235O : 0.1104 gave 0.2954 CO and 0-0510 H,O.0.1194 , 6.1 C.C. N at 1 7 O and 768 mm. N=5.99. The compound is soluble in alcohol o r acetic acid sparingly so in benzene and insoluble in light petroleum. It gives a violet colora-tion with alcoholic ferric chloride and in dilute alcoholic solution shows a green fluorescence. The p’crate separated in bright orange needles melting a t 238O : 0.1138 gave 12 C.C. N a t 17O and 760 mm. C=72*97; H=5.13. C,,H,,0,N,4H20 requires C’= 73-17 ; H = 4.88 ; N = 5.99 per cent. N=12.33. C,,H,,O2N,C,H3O7N3 requires N = 12.01 per cent. 7-ligdroxy-2-phem~l-y-b enzopyrone, 0 Three grams of the above imino-pyran were boiled under reflux with an excess of 5 per cent,. sodium hydroxide solution until n GHOSH A SYNTHESIS OF FLAVOHES.131 more ammonia was evolved. The solution was acidified with dilute sulphuric acid and the product crystallised from alcohol in colour-less needles melting a t 243O (Emilewicz and Kostanecki Bey. 1898, 31 703 give 240O). Decomposition zvith Potassium Hydroxide. I n order to prove the constitution 4 grams of 4-imino-7-hydroxy-2-phenyl-y-benzopyran were boiled with a solution of 18 grams of potassium hydroxide iii 35 C.C. of water for three hours. The solu-tion was extracted with ether and on evaporation of the ethereal solution an oil was obtained which was identified as acetophenone. The alkaline solution was treated in the same way as described in the case of the decomposition of 7-hydroxy-2 S-diphenyl-y-benzo-pyrone.The products obtained from the solution were resorcinol and P-resorcylic acid. 4Jmino-2-phenyl-1 4-a-naphthapyran Sulphate, H, This was prepared from benzoylacetonitrile (4 grams) a-naphthol (4 grams) and concentrated sulphuric acid (20 c.c.). The product, separated by partly neutralising the clear acid solution obtained by pouring the mixture on ice crystallised from alcohol containing dilute sulphuric acid in prisms which decomposed a t 185-215O. The yield was small : 0.1100 gave 0.2486 CO and 0.0386 H,O. 0.2460 , 8 C.C. N2 a t 16O and 762 mm. N=3*80. The base crystallised from alcohol in pale brown prisms decom-0.1093 gave 0.3152 GO and 0.0505 H 2 0 . 0.1047 , 4.8 C.C. N a t 20° and 764 mm. N=5.27. The picrate formed orange needles melting and decomposing a t 0.1230 gave 12.2 C.C. N2 a t 1 7 O and 764 mm. VOL. CIX. G C=61-63; H=3*89. C,,H,,0N,H2S0 requires C= 61-78 ; H = 4-06 ; N = 3-80 per cent. posing a t 138-148O: C=78*65; H=5-12. C19H130N requires @= 78.89 ; H = 4-80 ; N = 5.1 7 per cent. 274O : N=ll-50. C,,H,,0N,C,H307N requires N = 11.20 per cent 122 RAY AND DE MOLECULAR VOLUMES OF THE Treatment with Dilute Sodium Hydroxide. Three grams of the above imino-compound were boiled with 30 C.C. of 5 per eent. sodium hydroxide solution until the evolution of ammonia ceased. The solution was then acidified when a solid mixed with some resinous material separated. This product after repeated crystallisations from dilute alcohol separated in colour-less needles melting a t 153O (the melting point of 2-phenyl-1 4-a-naphthapyrone is 154-156O; Kostanecki Ber. 1898 31 707). The yield was very small. 4-Imino-7 ; 8-dihydroxy-2-phenyl-y-b enzopyran., OH 0 This was prepared from 4 grams of benzoylacetonitrile 4 grains of pyrogallol and 20 C.C. of concentrated sulphuric acid. The alcoholic solution of the viscous product obtained on pouring the acid mixture on ice was treated with an excess of aqueous sodium acetate ; the substance which separated crystallised from alcohol in brown prisms decomposing a t 145-165O : 0-1064 gave 0-2596 CO and 0.0428 H,O. C,,H,,O,N,H,O requires C = 66.42 ; H = 4.79 per cent. The compound gives a red solution with aqueous sodium hydr-oxide develops a green colour with alcoholic ferric chloride and reduces ammoniacal silver nitrate solution on warming. C= 66-54 ; H=4.57. ORQANIC LABORATOBY, UNIVERSITY COLLEGE, LONDON. [Received November 30th 1915.

ISSN:0368-1645    

DOI:10.1039/CT9160900105

出版商:RSC    

年代:1916

数据来源: RSC