摘要:

Ordinary Members PROFESSOR R. J. DONOVAN 1983 PROFESSOR M. C. R. SYMONS 1983 DR G. J. HILLS 1984 PROFESSOR J. M. THOMAS 1983 PROFESSOR A. J. LEADBETTER 1984 DR J. ULSTRUP 1985 DR I . W. M. SMITH 1985 PROFESSOR G. WILLIAMS 1985 PROFESSOR F. L. SWINTON 1983 DR D. A. YOUNG 1984 Honorarj, Secretarj-: DR G. J. HILLS Honorarj- Treasurer : PROFESSOR P. GRAY The President thanked the retiring members of Council, Vice-presidents Professor Sheppard and Professor Wagner, and Ordinary Members Professor King and Professor Purnell, for their services. 5. Reriew of Futurr Acfirifies A programme of future activities of the Division had been tabled and the President drew attention to the forthcoming General Discussions and Symposia. xiOrdinary Members PROFESSOR R. J. DONOVAN 1983 PROFESSOR M. C. R. SYMONS 1983 DR G. J. HILLS 1984 PROFESSOR J. M. THOMAS 1983 PROFESSOR A. J. LEADBETTER 1984 DR J. ULSTRUP 1985 DR I . W. M. SMITH 1985 PROFESSOR G. WILLIAMS 1985 PROFESSOR F. L. SWINTON 1983 DR D. A. YOUNG 1984 Honorarj, Secretarj-: DR G. J. HILLS Honorarj- Treasurer : PROFESSOR P. GRAY The President thanked the retiring members of Council, Vice-presidents Professor Sheppard and Professor Wagner, and Ordinary Members Professor King and Professor Purnell, for their services. 5. Reriew of Futurr Acfirifies A programme of future activities of the Division had been tabled and the President drew attention to the forthcoming General Discussions and Symposia. xi

ISSN:0300-9599    

DOI:10.1039/F198278FX017

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

3 708 REVIEW OF BOOKS is the absence of any reference to possible new and potentially significant applications for polymer latices. Novel applications may well be found in at least two directions, namely, those which exploit the large polymer-aqueous-phase specific surface area of latices, and those which exploit the electrical dissymmetry which is present at the interface between polymer and aqueous phase in the case of electrostatically stabilised latices. No reference is made in this book to the efforts which have so far been made to exploit for medical purposes the adsorptive and binding potentialities of the large area of polymer-aqueous-phase interface in latices. Nor is there any mention of possible catalytic applications of this large interfacial area. So far, catalytic applictions have been confined to those which rely essentially upon enhancement of the counter-ion concentration in regions of the electrical double layer which are near to the polymer surface.However, it is at least possible that the adsorptive capacity of the interface may also be useful in catalytic applications. Some discussion of possibilities such as these would have been welcome. D. C. BLACKLEY Received 14th April, 1982 Shock Waves in Chemistry. Ed. by ASSA LIFSHITZ. (Marcel Dekker, New York, 1981). Pp. ix + 390. Price SFr 182. After a somewhat hesitant start, the use of shock waves to study chemical and physical processes at high temperatures has become an accepted technique and reliable kinetic data can be obtained in this way. Several books have been written, notably by Bradley and by Gaydon and Hurle, which describe not only the underlying principles and the experimental procedures but also give some account of the early results obtained using shock waves to provide high temperatures for short, well defined times in the reactant gases.Inevitably, these books have become rather dated. This new book, edited by Lifshitz, is rather different. It is a collection of self-contained review articles on various aspects of shock waves. The first (by Khandelwal and Skinner) is concerned with hydrocarbon oxidation, and the next (by Tsang) describes the results obtained using the comparative rate technique which he has pioneered. Both these articles include extensive lists of references and represent useful summaries of the present situation.Boyd and Burns have contributed a chapter on dissociation-recombination reactions, while Kiefer describes the laser-schlieren method which he has done so much to develop. There is another chapter by an acknowledged expert, Just, on atomic resonance absorption spectrometry. Under shock-tube conditions it is very seldom that the concentrations of radicals and other species reach a steady state, and so the classical Bodenstein steady-state approximation cannot be used. Instead, it is necessary to integrate the differential equations describing the time-variation of species concentration, and Gardiner, Walker and Wakefield have provided a useful guide to the computational procedures available in this and other aspects of shock-tube work.In addition to these contributions there is another by Bar-Nun on Chemical Aspects of Shock Waves in Planetary Atmospheres which, although interesting in itself, fits rather uneasily with its companions. As is inevitable in a book of this type the standard and style of the chapters varies and there is some overlapping material; none of this, however. represents a serious drawback. What is more difficult to understand is the audience for whom the book is intended. Each chapter is a useful and interesting review which will appeal to a fairly restricted readership, but, in the opinion of this reviewer, the whole volume lacks coherence. The time-honoured phrase ‘should be on the shelves of every library’ probably applies, though the price, over &50 at the current exchange rate, must cause all university librarians to flinch in these days of U.G.C.cuts. There is still room for the definitive up-to-date book to be written on shock waves in chemistry. J. A. BARNARD Received 5th April, 19823 708 REVIEW OF BOOKS is the absence of any reference to possible new and potentially significant applications for polymer latices. Novel applications may well be found in at least two directions, namely, those which exploit the large polymer-aqueous-phase specific surface area of latices, and those which exploit the electrical dissymmetry which is present at the interface between polymer and aqueous phase in the case of electrostatically stabilised latices. No reference is made in this book to the efforts which have so far been made to exploit for medical purposes the adsorptive and binding potentialities of the large area of polymer-aqueous-phase interface in latices.Nor is there any mention of possible catalytic applications of this large interfacial area. So far, catalytic applictions have been confined to those which rely essentially upon enhancement of the counter-ion concentration in regions of the electrical double layer which are near to the polymer surface. However, it is at least possible that the adsorptive capacity of the interface may also be useful in catalytic applications. Some discussion of possibilities such as these would have been welcome. D. C. BLACKLEY Received 14th April, 1982 Shock Waves in Chemistry. Ed. by ASSA LIFSHITZ. (Marcel Dekker, New York, 1981). Pp. ix + 390.Price SFr 182. After a somewhat hesitant start, the use of shock waves to study chemical and physical processes at high temperatures has become an accepted technique and reliable kinetic data can be obtained in this way. Several books have been written, notably by Bradley and by Gaydon and Hurle, which describe not only the underlying principles and the experimental procedures but also give some account of the early results obtained using shock waves to provide high temperatures for short, well defined times in the reactant gases. Inevitably, these books have become rather dated. This new book, edited by Lifshitz, is rather different. It is a collection of self-contained review articles on various aspects of shock waves. The first (by Khandelwal and Skinner) is concerned with hydrocarbon oxidation, and the next (by Tsang) describes the results obtained using the comparative rate technique which he has pioneered.Both these articles include extensive lists of references and represent useful summaries of the present situation. Boyd and Burns have contributed a chapter on dissociation-recombination reactions, while Kiefer describes the laser-schlieren method which he has done so much to develop. There is another chapter by an acknowledged expert, Just, on atomic resonance absorption spectrometry. Under shock-tube conditions it is very seldom that the concentrations of radicals and other species reach a steady state, and so the classical Bodenstein steady-state approximation cannot be used. Instead, it is necessary to integrate the differential equations describing the time-variation of species concentration, and Gardiner, Walker and Wakefield have provided a useful guide to the computational procedures available in this and other aspects of shock-tube work.In addition to these contributions there is another by Bar-Nun on Chemical Aspects of Shock Waves in Planetary Atmospheres which, although interesting in itself, fits rather uneasily with its companions. As is inevitable in a book of this type the standard and style of the chapters varies and there is some overlapping material; none of this, however. represents a serious drawback. What is more difficult to understand is the audience for whom the book is intended. Each chapter is a useful and interesting review which will appeal to a fairly restricted readership, but, in the opinion of this reviewer, the whole volume lacks coherence. The time-honoured phrase ‘should be on the shelves of every library’ probably applies, though the price, over &50 at the current exchange rate, must cause all university librarians to flinch in these days of U.G.C. cuts. There is still room for the definitive up-to-date book to be written on shock waves in chemistry. J. A. BARNARD Received 5th April, 1982

ISSN:0300-9599    

DOI:10.1039/F198278BX019

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

JOURNAL OF THE CHEMICAL SOCIETY FARADAY TRANSACTIONS, PARTS I A N D I 1 The Journal of The Chemical Society is issued in six sections: Journal of The Chemical Society, Chemical Comniunications Journal of The Chemical Society, Dalton Transactions Journal of The Chemical Society, Faraday Transactions, I Journal of The Chemical Society, Faraday Transactions, I I Journal of The Chemical Society, Perkin Transactions, I Journal of The Chemical Society, Perkin Transactions, I I Thus, five of the sections are directly associated with three of the Divisions of The Royal Society of Chemistry: the sixth is Chemical Communications. This continues to be the medium for the publication of urgent, novel results from all branches of chemistry. Communications should not normally exceed one printed page in length and authors are required to submit three copies of the typescript and two copies of a statement of the reasons and justification for seeking urgent publication of the work.This Section is intended to be essentially a journal for inorganic chemists containing papers on the structure and reactions of inorganic compounds and the application of physical chemistry techniques to, e.g. the study of inorganic and organometallic compounds and problems, including work on the kinetics and mechanisms of inorganic reactions and equilibria, and spectroscopic and crystallographic studies of inorganic compounds. Journal of the Chemical Society, Faraday Transactions, I and I I These are, respectively, physical chemistry and chemical physics journals. P A R T 1 (physical chemistry) includes papers on such topics as radiation chemistry, gas-phase kinetics, electrochemistry (other than preparative), surface and interfacial chemistry, heterogeneous catalysis, physical properties of polymers and their solutions and kinetics of polymerization, etc.P A R T I I (chemical physics) contains theoretical papers, especially those on valence and quantum theory, statistical mechanics, intermolecular forces, relaxation phenom- ena, spectroscopic studies (including i.r., e.s.r., n.m.r., and kinetic spectroscopy, etc.) leading to assignments of quantum states, and fundamental theory, and also studies of impurities in solid systems, etc. Journal of The Chemical Society, Chemical Communications Journal of The Chemical Society, Dalton Transactions Journal of The Chemical Society, Perkin Transactions, I and II These are, respectively, the organic chemistry and the physical organic chemistry sections of the Journal.P A R T I (organic and bio-organic chemistry) is designed to contain papers on all aspects of synthetic, and natural product organic and bio-organic chemistry and to deal with aliphatic, alicyclic, aromatic, carboncyclic and heterocyclic compounds. Papers on organometallic topics are considered for either the Dalton or the Perkin Transactions. 1PART I I (physical organic chemistry) is for papers on reaction kinetics and mechanistic studies of organic systems and the use of physico-chemical, spectroscopic, and crystallographic techniques in the solution of organic problems.Notice to Authors ( 1 ) Although authors need not be members of the Royal Society of Chemistry it is hoped that they will be. (2) Authors must indicate the Part of the Journal they wish their paper to appear in. This preference will be respected unless it is obviously erroneous in terms of the scientific content of the paper. (3) Since all papers will be subjected to refereeing, in parallel, by two independent referees, the original typescript (quarto or A4 size) and two good-quality copies should be provided. (4) All papers should be sent to the Director of Publications, The Royal Society of Chemistry, Burlington House, Piccadilly, London W 1 V OBN. ( 5 ) For details of manuscript preparation, preferred usages, etc. the Instructions to Authors, previously available from the Faraday Society, and now obtainable from The Royal Society of Chemistry, should be consulted. (6) The Society will adopt the following abbreviations for the new journals in all its publications.J . Chem. SOC., Chem. Commun. J . Chem. SOC., Dalton Trans. J . Chem. SOC., Faraday Trans. I J . Chem. SOC., Faraday Trans. 2 J. Chem. SOC., Perkin Trans. 1 J. Chem. SOC., Perkin Trans. 2 * The author to whom correspondence should be addressed is indicated by an asterisk after his name in the heading of each paper. 11F A R A D A Y D I V I S I O N OF THE ROYAL SOCIETY OF CHEMISTRY A S S O C I A Z I O N E I T A L I A N A D I C H l M l C A F l S l C A S O C l i T i DE C H l M l E PHYSIQUE DEUTSCHE BUNSEN GESELLSCHAFT F U R P H Y S I K A L I S C H E CHEMIE F A R A D A Y D I S C U S S I O N NO.7 4 Electron and Proton Transfer University of Southampton, 14-1 6 September 1982 This meeting will be concerned with fundamental aspects of the chemical kinetics of electron and proton transfer reactions in solution and with particular reference to well defined biological systems. Attention will be focused on (i) the theory of charge transfer, (ii) critical experiments designed to test those theories and (iii) their application to the understanding of charge transfer reactions in molecules of biological interest. The meeting will encompass well characterised reactions in solution, redox reactions and elementary biochemical reactions; particular attention will be paid t o isotope effects, to electron and proton tunnelling, to intermolecular and intramolecular transfers and to related questions concerning the organisation of biological systems. Among those who have agreed to take part are R.A. Marcus, R. R. Dogonadze, H. Gerischer, J. Jortner, R. M. Kuznetsov, N. Sutin, R. J. P. Williams, H. L. Friedman, J. M. Saveant, J. F. Holzwarth, F. Willig, J. C. Mialocq, M. Kosower, L. I. Krishtalik, E. F. Caldin, H. H. Limbach, W. J. Albery, M. M. Kreevoy, J. J. Hopfield, P. Rich, H. A. 0. Hill, K. Heremans, C. Gavach and D. 6. Kell. The final programme and application form may be obtained from: Mrs Y. A. Fish, The Royal Society of Chemistry Burlington House, London W1V OBN ... I l lFARADAY D I V I S I O N O F THE ROYAL S O C I E T Y OF CHEMISTRY SYMPOSIUM NO. 1 7 The Hydrophobic Interaction University of Reading, 15-1 6 December 1982 This term refers to interactions between chemically inert residues arising from perturbations in the unique spatial and orientational correlations in liquid water.These effects provide a major contribution to many of the non-covalently bonded structures that form the basis of life processes. Current advances in the statistical mechanics of polar fluids, intermolecular forces, computer simulation, and membrane physics are providing a new basis for the re-examination of various aspects of hydrophobic effects, their origin and their quantitative description. Such theoretical treatments will be confronted with recent experimental work on simple model systems which, it i s hoped, will lead to a better understanding of hydrophobic interactions in more complex processes.The following have provisionally agreed to contribute to the symposium : A. Ben-Naim, H. J. C. Berendsen, D. L. Beveridge, S. D. Christian, L. Cordone, D. Eagland, D. Eisenberg, R. Lumry, P. J. Rossky, M. C. R. Symons, H. Weingartner, M. D. Zeidler The preliminary programme may be obtained from : Mrs Y. A. Fish, The Royal Society of Chemistry Burlington House, London W1 V OBN THE FARADAY DIVISION OF THE ROYAL SOCIETY O F CHEMISTRY GENERAL DISCUSSION NO. 75 I nt ra mo I ecu I a r K i net i cs University of Warwick, 18-20 April 1983 Organising Committee Professor J. P. Simons (Chairman) Dr M. S. Child Professor R . J. Donovan Dr G. Hancock Experimental and theoretical interest in the time-dependent behaviour of isolated molecules, radicals or ions is strong and increasing. The Discussion will be concerned with the kinetics of processes which occur in isolated species following their preparation in states with non-equilibrium energy distributions (e.g.by photon absorption or collisional activation). Topics covered will include: ( a ) theoretical and experimental studies of energy redistribution in isolated species; ( b ) observation and theoretical modelling of the competition between intramolecular energy redistribution and radiative decay or radiationless processes (e.9. internal conversion, fragmentation, isomerisation). Contributions for consideration by the Organising Committee are invited. Titles should be submitted as soon as possible and abstracts of 300 words by 31 May 1982.Full papers for publication in the Discussion Volume will be required by 15 December 1982. Titles and abstracts should be sent to: Professor J. P. Simons, Department of Chemistry, University of Nottingham, University Park, Nottingham NG7 2RD. Dr D. M. Hirst Professor K. R. Jennings Dr R . Walsh 1vTHE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY GENERAL DISCUSSION NO. 76 Concentrated Colloidal Dispersions Loughborough University of Technology, 14-1 6 September 1983 The meeting will discuss the experimental investigation and the theoretical description of the properties of concentrated colloidal dispersions, i.e. those systems in which the particle-particle interactions are strong enough to cause significant deviations from ideal behaviour. Both the structural and dynamic features of concentrated systems as determined by scattering, rheological and other techniques will be considered.It is anticipated that a range of dispersion types will be discussed. These will include both 'model' systems and dispersions of importance to industry provided that the data from the measurements can be interpreted. Contributions for consideration by the organising committee are invited and abstracts of about 300 words should be sent by 31 st August 1982 to: Professor I?. H. OttewiII, School of Chemistry, University of Bristol, Cantock's Close, Bristol BS8 1TS FARADAY DIVISION INFORMAL AND GROUP MEETINGS Gas Kinetics Group Seventh International Symposium on Gas Kinetics To be held at the University of Gottingen, West Germany on 23-27 August 1982 Further information from Dr R.Walsh, Department of Chemistry, University of Reading, Whiteknights, Reading RG6 2AD Colloid and Interface Science Group with the Colloid and Surface Chemistry Group of the SCI Adsorption from Solution To be held at the University of Bristol on 8-1 0 September 1982 Further information from Dr W. D. Cooper, Department of Chemistry, University of Edinburgh, West Mains Road, Edinburgh EH9 3JJ Industrial Physical Chemistry Group Supercritical Fluids: Their Chemistry and Application To be held at Girton College, Cambridge on 13-15 September 1982 Further information from Dr W. R. Ladner, National Coal Board, Coal Research Establishment, Stoke Orchard, Cheltenham GL52 4RZ Neutron Scattering Group and Polymer Physics Group with the Institute of Physics The Neutron and its Applications To be held in Cambridge on 13-1 7 September 1982 Further information from The Meetings Officer, Ins!itute of Physics, 47 Belgrave Square, London SW1 X 8QX Theoretical Chemistry Group Molecular Electron Structure Theory and Potential Energy Surface To be held at the University of Bristol on 15-1 6 September 1982 Further information from Dr G.G. Balint-Kurt;, School of Chemistry, University of Bristol, Cantock's Close, Bristol 6S8 1TS Molecular Beams Group Molecular Beams and Molecular Structure To be held at the University of Bristol on 16-1 7 September 1982 Further information from Dr J. C. Whitehead, Department of Chemistry, University of Manchester, Manchester M13 9PL VFARADAY DIVISION INFORMAL AND GROUP MEETINGS Division Autumn Meeting: Energy and Chemistry To be held at Heriot-Watt University, Edinburgh on 21 -23 September 1982 Further information from Dr J.F. Gibson, The Royal Society of Chemistry, Burlington House, London W1 V OBN Statistical Mechanics and Thermodynamics Group with the British Society of Rheolog y Microstructure and Rheology To be held at Trinity Hall, Cambridge on 21-24 September 1982 Further information from Dr P. Richmond, Unilever Research, Port Sunlight, Wirral, Merseyside L62 3JW High Resolution Spectroscopy Group High Resolution Fourier Transform, Laser Infrared and Electronic Spectroscopy To be held at the University of Newcastle-upon-Tyne on 22-24 September 1982 Further information from Dr P. J. Sarre, Department of Chemistry, University of Nottingham, Nottingham NG7 2RD Polymer Physics Group Polymer Electronics To be held in London on 20 October 1982 Further information from the Meetings Officer, The Institute of Physics, 47 Belgrave Square, London SWlX 8QX Division with Polymer Physics Group and Macrogroup UK Annual Chemical Congress: Copolymers To be held at the University of Lancaster on 11 -1 3 April 1983 Further information from Dr J.F. Gibson, The Royal Society of Chemistry, Burlington House, London W1 V OBN Polymer Physics Group, Macrogroup UK and the Plastics and Rubber Institute Polyethylenes To be held in London on 8-10 June 1983 Further information from The Plastics and Rubber Institute, 11 Hobart Place, London SW1W OHZ V iPublications from The Royal Society of Chemistry SPECIALIST PERIODICAL REPORTS Catalysis VOlm 4 Senior Reporters: C.Kemball and D. A. Dowden This volume reviews the recent literature published up to mid 1980. Brief Contents: The Design and Preparation of Supported Catalysts: Aspects of Characterization and Activity of Supported Metal and Bimetallic Catalysts; Metal Clusters and Cluster Catalysis; Olefin Metathesis; Superbasic Heterogeneous Catalysts; Hydration and Dehydration by Heterogeneous Catalysts; Sulphide Catalysts: Characterization and Reactions including Hydrodesulphurization; Carbon as a Catalyst and Reactions of Carbon. Hardcover 266pp 0 851 86 554 2. Price €29.00 ($62.00). RSC Members f 17.50 Gas Kinetics and Energy Transfer Vol. 4 Senior Reporters: P. G. Ashmore and R. J.Donovan A review of the literature published up to early 1980. Brief Contents: Reactions Studied by Molecular Beam Techniques; Reorientation by Elastic and Rotationally Inelastic Transitions; infrared Multiple Photon Excitation and Dissociation: Reaction Kinetics and Radical Formation; Ultraviolet Multiphoton Excitation: Formation and Kinetic Studies of Electronically Excited Atoms and Free Radicals; Gas Phase Reactions of Hydroxyl Radicals; Gas Phase Chemistry of the Minor Constituents of the Troposphere. Hardcover 252pp 0 85186 786 3. Price €45.00 ($96.00). RSC Members f25.00 Mass Spectrometry VOlm 6 Senior Reporters: R. A. W. Johnstone This volume reviews the literature published between July 1978 and June 1980. Brief Contents: Theory and Energetics of Mass Spectrometry; Structures and Reactions of Gas-phase Organic Ions; Gas-phase Ion Mobilities, Ion - Molecule Reactions, and Interaction Potentials; Interaction of Electromagnetic Radiation with Gas-phase Ions; Aspects of Secondary Ion Emission; Development and Trends in Instrumentation in Mass Spectrometry; Applications of Computers and Microprocessors in Mass Spectrometry; Gas Chromatography- Mass Spectrometry and High- performance Liquid Chromatography- Mass Spectrometry; Reactions of Negative Ions in the Gas Phase; Natural Products; The Use of Mass Spectrometry in Pharmacokinetic and Drug Metabolism Studies; Organometallic, Co-ordination, and Inorganic compounds Investigated by Mass Spectrometry.Hardcover 368pp 0 85186 308 6. Price €39.50 ($88.00). RSC Members €23.00 ORDERING RSC Members should send their orders to: The Royal Society of Chemistry, The Membership Officer, 30 Russell Square, London WC1B 5DT.Non-RSC Members should send their orders to: The Royal Society of Chemistry, Distribution Centre, Blackhorse Road, Letchworth, Herts SG6 1 HN. The Royal Society of Chemistry Burlington House Piccadilly London W1 V OBN viiNOTES I t has always been the policy of the Faraday Transactions that brevity should not be a factor influencing acceptability for publication. In addition however to full papers both sections carry at the end of each issue a section headed “Notes”, which are short self-contained accounts of experimental observations, results, or theory that will not require enlargement into “full” papers. The “Notes” section is not used for preliminary communications. The layout of a “ Note” is the same as that of a paper.Short summaries are required. The procedure for submission, administration, refereeing, editing and publication of “Notes” is the same as for “full” papers. However, “Notes” are published more quickly than papers since their brevity facilitates processing at all stages. The Editors endeavour to meet authors’ wishes as to whether an article is a full paper or a “Note”, but since there is n o sharp dividing line between the one and the other, either in terms of length or character of content, the right is retained to transfer overlong ‘‘ Notes” to the “full papers” section. As a guide a ‘’ Note” should not exceed 1500 words or word-equivalents. NOMENCLATURE AND SYMBOLISM For many years the Society has actively encouraged the use of standard IUPAC nomenclature and symbolism in its publications as an aid to the accurate and unambiguous communication of chemical information between authors and readers.In order to encourage authors to use IUPAC nomenclature rules when drafting papers, attention is drawn to the following publications in which both rules themselves and guidance on their use are given. Physicochemical Quantities and Units. Manual of Symbols and Terminology for Physicochemical Quantities and Units. (Pure and Appl. Chem., Vol. 51, No. I , 1979, pp. 1-41. Also available as a soft-cover booklet from Pergamon Press, Oxford.) Surface Chemistry. ‘ Definitions, Terminology, and Symbols in Colloid and Surface Chemistry - I .’ (Pure and Appl. Chem., Vol. 31, No. 4, 1972, pp. 577-638.) ’ Definitions, Terminology, and Symbols in Colloid and Surface Chemistry - 11. Heterogenous Catalysis.’ (Pure and Appl. Chem., Vol. 46, No. I , 1976, In addition. the terminology and symbols for the following subject areas are available either in the form of soft-cover booklets from Pergamon Press (denoted by *) or have been the subject of articles in Pure and Applied Chemistry in recent years: activities;* chromatography; electrochemistry; electron spectroscopy; equilibria, fluid flow; ion exchange; liquid-liquid distribution; molecular force constants; Mossbauer spectra; nuclear chemistry; pH ; polymers; quantum chemistry; radiation;* Raman spectra; reference materials (recommended reference materials for the realization of physico- chemical properties: general introduction, enthalpy, optical rotation, surface tension, optical refraction, molecular weight, absorbance and wavelength, pressure-volume- temperature relationships, reflectance, potentiometric ion activities, testing distillation columns) ; solution chemistry ; spectrochemical analysis ; surface chemistry ; thermo- dyfiamics, and zeolites. Finally, the rules for the naming of organic and inorganic compounds are dealt with in the following publications from Pergamon Press: ‘Nomenclature of Organic Chemistry, Sections A, B, C, D, E, F, and H’, 1979. ’ Nomenclature of Inorganic Chemistry’, 1971. pp. 71-90.) A complete listing of all IUPAC nomenclature publications appears in the 198 1 Index issues of J . Chem. SOC. ... Vlll

ISSN:0300-9599    

DOI:10.1039/F198278FP033

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

J . Chem. SOC., Faraday Trans. 1, 1982, 78, 1345-1357 The State and Phase Transitions of an Adsorbate on a Homogeneous Solid Surface Part 1.-A Model of a Two-dimensional Associated van der Waals Gas and of a Two-dimensional Phase Transition BY GEORGY I. BEREZIN AND ANDREJ V. KISELEV* Laboratory of Surface Chemistry, Institute of Physical Chemistry, U.S.S.R. Academy of Sciences, Moscow 1 1707 1, U.S.S.R. Received 16th June, 1980 The state of an adsorbate on a mathematically homogeneous surface of a non-specific adsorbent over a wide range of temperature and surface coverage is considered. A model of a two-dimensional associated van der Waals gas is used. Besides the averaged long-distance intermolecular attraction considered in the usual van der Waals gas model, the new model also takes into account the formation of molecular associates at short distances (in a quasi-chemical approximation).These associates are formed not only as a result of the non-specific intermolecular interactions, but also additional specific intermolecular interactions, especially hydrogen bonds and complexing processes depending on the specific structure of the adsorbate molecule. Expressions have been derived for the thermodynamic characteristics of a two-dimensional associated van der Waals gas at temperatures above and below the two-dimensional critical temperature. A method is proposed for the determination of the critical parameters of the two-dimensional state, and the thermodynamic properties of a two-phase, two-dimensional system are considered.The state of an adsorbate is at present the subject of extensive study. Attempts have mainly been made to construct approximate molecular models for the first adsorbed layer on the homogeneous surface of a solid with the aid of molecular statistics [see, for example, ref. (1) and (2)]. Quantitative results for a non-specific adsorbent on a homogeneous surface (graphitized thermal carbon black, GTCB) have been obtained only for low (zero) surface coverage, 0, in the atom-atom approximation [see ref. (2) and (3)]. So far no advances have been made in obtaining quantitative results in the range of large B for complicated molecules. The use of chemical thermodynamics methods, in combination with the semi-empirical equation of state, is of interest in describing the properties of both the two-dimensional (adsorbed) state and three- dimensional real fluids (gases and liquids).Here the intermolecular interactions at short distances are viewed as association reactions combining all types of intermol- ecular interactions. For example, when molecular associations in an ideal gas are taken into account we obtain an equation of state whose virial coefficients are related to the equilibrium constants of quasi-chemical reactions for the formation of double, triple, etc. associate^.^ An important step in the development of this method for describing the states of fluids lies in accounting for molecular associations in the usual van der Waals gas (VWG) m0de1.~ The exact calculation of the equilibrium constants for the formation and decom- position of associates in fact involves the same difficulties which are encountered in calculating the virial coefficients by means of the molecular statistical method.The problem here therefore lies in finding a convenient and simple approximate method of calculating these equilibrium constants. As a criterion for judging the validity of the 13451346 PHASE TRANSITIONS OF A N ADSORBATE model and calculation procedure chosen we can use the completeness of the semi-empirical description of the thermodynamic properties of a real system. As a first approximation we assumed that the association constants of quasi-chemical reactions of the addition of a single molecule to associates of any multiplicity are the same.6-8 As a consequence of this assumption, we can easily sum the series in the equation of state for an associated van der Waals gass and derive an equation of state in a closed form both for a three-dimensional associated van der Waals gas (3DAVWG)8 and for the corresponding two-dimensional associated gas (2DAVWG).7 A verification of these assumptions can be made by comparing the thermodynamic functions calculated within the framework of the models adopted with the corresponding experimental values.A comparison thus made for ammonias and some other three-dimensional gases has demonstrated the satisfactory agreement between theory and experiment not only for the compressibility coefficient but also for the internal energy and heat capacity of gases over wide ranges of temperature and density. In Part 1 of this communication we shall construct a model for a 2DAVWG and derive the equation of state and the main thermodynamic functions of an adsorbate on a mathematically homogeneous surface.In contrast to our earlier papers, in the present study we have considered mainly the thermodynamic characteristics of a 2DAVWG below the critical temperature of the two-dimensional state. In Parts 2 and 3 of this seriesg we shall make a comparison between the calculations made using this model and the thermodynamic properties of a real system adsorbed on GTCB, particularly the heat capacity, over a wide range of temperature and surface coverage. We shall consider the thermodynamic properties of a two-phase, two-dimensional adsorbed fluid and its corresponding critical parameters.A review of the literature dealing with the two-dimensional state and heat capacity of an adsorbed system, and with two-dimensional phase transitions will be published elsewhere. lo THEORY MODEL OF A TWO-DIMENSIONAL ASSOCIATED VAN DER WAALS GAS (2DAVWG) AND I T S EQUATION OF STATE In the case of adsorbate-adsorbent intermolecular interactions associations of the adsorbed molecules with the adsorbent arise irrespective of the nature of this interaction. Indeed, the minimum in the potential curve describing the intermolecular interaction suggests that the mean lifetime of ‘contacting’ molecules is non-zero, and that it increases with increasing depth of the potential well and with decreasing temperature. A van de Waals gas should also be associated to some degree.This property underlies the van der Waals gas model as a collective system of hard spheres attracting each other. In the framework of a more complete van der Waals model, the formation of associates from other associates of lesser multiplicity or from unitary molecules, in particular, can be described as a quasi-chemical rea~tion.~9 7 + We shall now consider the state of the first adsorbed layer as a mixture of two-dimensional van der Waals gases consisting of the same single molecules associated to varying degrees. At a given temperature T, each of the gases i on the surface of area A contains Ni associates comprising i single molecules. Thus the total number of free single molecules and their associates isG . I. BEREZIN AND A. V. KISELEV 1347 while the total number of single molecules in the mixture (both free, i = 1, and in the (2) associates, i = 2, 3, .. .) is N = iNf. The degree of association is (3) The two-dimensional pressure due to the ith gas solely consisting of the ith (4) where ai and a, are the constants of mean van der Waals attraction for the ith gas and the gas mixture, respectively; A - bz is the free surface area unoccupied by the molecules, where bZ is the area occupied by all the molecules. The total pressure due to two-dimensional gases associated to different degrees is i-1 /? = ( N - Na,)/N = 1 - Na,/N. In the usual van der Waals model, i = 1, N,, = N and p = 0. associates is given by the expression ni = NikT/(A - b,) - (ai)! (a&/A2 n = [NikT/(A-b)]-(ai):(a&/A2. i-1 By virtue of the combinatorial rules for the van der Waals constants of molecules of the same kind, we have5 (ai)+ = iNi(al)i (a& = C.iNi(a,)i = N(a,)i i ' b, = Nb, where a, and b, are the van der Waals constants of single molecules of an unassociated two-dimensional gas. Substituting these expressions into eqn (9, and by virtue of eqn (1) and (2), we obtain Substituting the mean area per adsorbed molecule, w = A / N , and substituting eqn (3) into eqn (7), we obtain an expression of state for a two-dimensional associated van der Waals gas in the form n = [Na,kT/(A-Nb,)]-N2a,/A2. (7) n = [kT(l - / ? ) / ( o - b , ) ] - a , / w 2 . (8) When /I = 0, this equation becomes the usual equation of a two-dimensional van der Waals gas," but the higher the value of /?, the smaller n becomes.The quantity l/w = N / A = a represents the total surface concentration of the asdorbate in a two-dimensional layer, i.e. the total number of single molecules (non-associated and associated) per unit area of surface. In the cases considered below the concentration of the adsorbate in the gas volume over the adsorbent is low, therefore a z r, the Gibbs' value of adsorption. In these cases it is more convenient to introduce a dimensionless quantity, uiz. the surface coverage 8 = a / a , = ab, = b,/w where a , = l / b , is the dense monolayer capacity. In the usual 2DAVWG (two- dimensional van der Waals gas) model, the quantities a, and b,, and consequently N , = a , A, are assumed to be independent of temperature. In terms of total surface coverage 9, eqn (8) can be rewritten in a more convenient form as (8') n = (kT/b,) [( 1 -p)9/( 1 - 9)] - (a,/bf) g2.1348 PHASE TRANSITIONS OF A N ADSORBATE In terms of 6, we can also rewrite eqn (3) as follows: p = 1 -6,,/6 (3’) where OaS is the surface coverage by associates of different multiplicities (monomers, dimers, etc.).The magnitudes of /? and O,, depend on 6. In order to find this dependence and thus eliminate from eqn (V), we shall use a quasi-chemical reaction model for the formation and decomposition of the associates. The formation of the ith associate on the surface we consider here to be the result of successive addition of single molecules6 M, + M, f M, M2 + Mi + M3 Mi + f Mi+1. (9) The equilibrium constants of these reactions can be expressed through the surface coverage of free area by the i-multiple associates, Oil( 1 - O), where Oi is the degree of coverage of the whole surface by the i-multiple associates. Thus, Kas,i = L/[(k)(x)], 1-8 1-0 1-0 Furthermore we assume [as earliers-8 in correspondence with the approximation discussed in ref.(12)] that K,, is the mean association constant for all the reactions proceeding via scheme (9), i.e. Kas,2 x K,,,, x . . , = Kas. (1 1) This approximation is of course quite rough, and this should be borne in mind in assessing the agreement of the theory with experiment. Nonetheless, it is valid for the chain association resulting, in particular, in the formation of hydrogen bonds. This approximation allows us to derive the equations for the thermodymanic characteristics of two-dimensional adsorption in a closed form.Under this assumption, from eqn (10) it follows that 6i = [Kas/( 1 - 6)p-l @. (12) The total surface coverage by all associates is and the total surface coverage by the single molecules (both by the individual molecules and by the molecules contained in associates) is 6 = 61+262+36,+. . . +iOi. (14) Substituting expression (12) into eqn (13) and (14) we obtain K,, O1/( 1 - 6) < 1, therefore the sequence (16) converges (see the Appendix). Hence we haveG. I. BEREZIN A N D A. V. KISELEV 1349 Since sequence (16) is the derivative of sequence (15) with respect to the factor KasO1/(l -8), from eqn (15') it follows that e = e,/[i -Kase/(i -0)y. 81 = eas/[l +Kaseas/(1 -@I* ( 16') (17) From eqn (15') and (16') we obtain Thus for eqn (1 7) [Ka, e/(l- 8)] (8,,/8)2 + e,,/e- 1 = 0.Substituting 8,,/# = 1 -p from eqn (3') into eqn (18) we obtain B = [4Kas8/(1-@1/{1+ V"1+4Kas8/(1 -@]I2- (19) p = K,,e. (20) From eqn (19) it follows that p = 0 when Kas = 0, while for any K,, > 0 and small 8 such that 4Ka,8 < 1, we have As 8 --+ 0, p --+ 0; similarly, as 8 + 1, p --+ 1. The dependence of /? on 8 for different Ka, values is shown in fig. 1 of ref. (7). We now return to the equation of state of a 2DAVWG. Substituting expression (19) into eqn (87, we obtain From this equation it is seen that for a given 8, an increase in Kas results in a decrease in n. We now shall examine the influence of association on the state of a two-dimensional gas in greater detail. For this purpose we express eqn (21) as follows: -- a, 82.nb, 8 2 --- k T - 1-81+~[1+4Ka,8/(1-8)] blkT In the framework of the van der Waals model the ratio a,/b,kT = K2 at constant temperature is constant, and for the sake of convenience in calculation let us take it to be equal to 1. By varying the association constant we can study the influence of association on n. The family of the corresponding two-dimensional equation of state nb,/kT =fl8) for the same temperature and different K,, values is shown in fig. 1, which shows that an increase in the association constant leads to conditions that give rise to two-dimensional condensation. In fig. 1 the curve corresponding to K,, = 4 for a,/b, k T = 1 passes close to the critical two-dimensional temperature of the system vapour-condensate, Tcz.A further increase in Kas at the same temperature gives rise to an undulating section corresponding to a two-dimensional condensation region in the isotherm. ISOTHERM AND HEAT OF ADSORPTION FOR THE 2DAVWG MODEL After finding the value of dn from eqn ( 8 9 , substituting it in the Gibbs equation d Inp = b,dn/kT8 (in which the adsorption r is expressed through 8), and finally integrating, we obtain P = "1 -B>"/K1(1 -a exp " - 8 ) W -@1-(2a,lkTb,)81 (22) or P = (B/KlKas) ~ X P {[(I-P) ~/(l-@I-2ai82/kTbJ (22')1350 PHASE TRANSITIONS OF AN ADSORBATE 1 .o 0.5 k 1 9 -0.5 -1.0 rrooo FIG. 1.-Isotherms of nb,/kT against l/O calculated by eqn (21'). For all curves a,/b,kTwas taken to be equal to 1 ; the values of the association constant, K,, are shown on the curves.bJkTis taken to be constant for all the curves. -4 -3 - 2 - 1 0 log,, P FIG. 2.-Effect of the association constant, K,, on the shape of the adsorption isotherm. The value of K,, is shown on the curves. For K,, = 4 the isotherm is close to the critical curve. For all the curves a,/b,kT is taken to be 1.G. I. BEREZIN AND A. V. KISELEV 1351 where Kl is Henry's ~ o n s t a n t . ~ For small 8, by virtue of eqn (20) we obtain P = (8/Kl). (23) Fig. 2 shows the adsorption isotherms of the 2DAVWG model at the same temperature for various K,, values at constant K, and a,/b,kT = 1. It is seen that association shifts the isotherm toward lower values of the pressure p , and the isotherm becomes S-shaped for higher values of the association constant. The latter corres- ponds (as is also evident from fig. 1 ) to two-dimensional condensation.When K,, = 0 /3 = 0, and eqn (22) becomes the Hill-De Boer equation,ll corres- ponding to the ordinary van der Waals equation for the two-dimensional non- associated state. The internal energy of the 2DAVWG is given by U = U , -pAU,, - (a,/b,) 8 (24) where Ul is the internal energy of isolated adsorbed single molecules and AU,, is the change in internal energy of the adsorbate occurring during the association reactions (9). When 8+0, U--+ U,. The integral change in the internal energy occurring during adsorption of a two-dimensional associated van der Waals gas at low p, i.e. from the bulk of an ideal gas, is AU = U - U g = AU, -pAU,, - (a,/b,) 8 (25) where Ug is the internal energy of an ideal gas, and AUl is the limiting value (as 8 -+ 0) of the change in the internal energy during adsorption of isolated single molecules of an ideal gas.The differentional heat of adsorption' of a two-dimensional associated van der Waals gas is qv = - r U = qst-kT = kT2(a lnp/2lJO-kT = -[a(8AU)/28lT = - m l + p [ l +(1 -p)/(l +p)(l -8)]AUa,+2(a,/b,)8 (26) where qst is the isosteric heat of adsorption. When AU,, = 0, this equation is converted into the equation for the heat of adsorption of a two-dimensional non-associated van der Waals gas, i.e. into the linear dependence of the heat of adsorption on 8." When 8 -+ 0, we find qv --+ qV, = - r U l . Fig. 3 shows how qv varies with 8 for different K,, values and constant a,, b, and T. To remove the influence of Kas on the form of dependence of qv on 8 the AU,, value was taken as constant.From fig. 3(a) it can be seen that an increase in association leads to an increase in the differential heat of adsorption in the initial range of 8. The differential heat of adsorption for Kas > 0 in the formation of a dense monolayer tends to high values [fig. 3(b)] because in the 2DAVWG model a small change in the density of the monolayer at almost full coverage results in repeated association, and consequently leads to a high molar differential heat of adsorption. This is clearly seen for the curves corresponding to low K,,; weak association in the low-8 range leads to an insignificant increase in the differential heat of adsorption which steeply increases with increasing 8.The influence of the adsorption of molecules in the second layer on the form of the dependence of qst on 8 has been considered ear lie^.^1352 PHASE TRANSITIONS OF A N ADSORBATE 0 0.1 e FIG. 3.--8-dependence of the contribution of the intermolecular adsorbate-adsorbate interaction in a two-dimensional associated van der Waals gas to the differential heat of adsorption (qv -qv, ,)/kT for different values of K,, (shown on the curves). In the calculations it was taken that AW,,/kT = 1 and ~ , / b , k T = 1 . log,, 0 FIG. 4.--8-dependence of the contribution made to the heat capacity of the adsorbate by the heat of dissociation for various K,, values (shown on the curves). In the calculations it was assumed that AU,,/kT= 7. HEAT CAPACITY OF A TWO-DIMENSIONAL ASSOCIATED VAN DER WAALS GAS Differentiating expression (24) with respect to 7' for constant 8 (at constant b, this is equivalent to constant r) we obtain an expression for the mean molecular heat capacity of a 2DAVWG: (27) C w = Ci+PACas+ID(1-P)/(1 +P)l (AUa,J2/kTG.I. BEREZIN A N D A. V. KISELEV 1353 where C, is the molar heat capacity of isolated single adsorbed molecules, and ACas is the change in the heat capacity of these molecules due to association [in ref. (7) the term PAC,, was omitted]. The quantity C, is an analogue of the mean molecular heat capacity at constant volume of a three-dimensional system, Cv. As 0 + 0, C, --+ C,. The mean molecular change in the heat capacity occurring during the adsorption of an ideal three-dimensional gas with the formation of a two-dimensional associated van der Waals gas on the adsorbent surface is given by (28) AC= Ca-C& = AC,+pACa~+[p(l-~)/(~ +p)] x(Aua,)'/kT.Fig. 4 shows how the contribution of the third term varies with the degree of association for different values of Kas. For any K,, > 0 this contribution passes through a maximum. The peak in this maximum shifts towards the low-0 range with increasing Kas. TWO- PHASE, T W 0 -DIMENSION A L SYSTEM : GENERAL EXPRESS IONS Below the critical temperature of the monolayer the adsorbate decomposes into two phases. We shall first consider the general expressions for the thermodynamic characteristics of an adsorbate consisting of two-dimensional phases. When N molecules are absorbed on a surface of area A , a part A' of this area is occupied by a two-dimensional liquid, while the other part, A", is covered by a two-dimensional vapour : where N1 and N" are the corresponding numbers of adsorbed molecules, and or, and oif are the areas per molecule for these phases. The superscripts 1 and v indicate that these quantities relate to the corresponding two-dimensional phases, while the subscript s denotes their values at the saturation line (binodal).We now introduce the surface coverages by the coexisting two-dimensional phases, O1 and tF. In the limiting cases where the surface is filled by only one of these coexisting phases, i.e. on the binodals, we have A = A'+AV = Wc;oif+Wc~: (29) or, = bl/Oif and u; = bl/O; (30) where c;ok and co: are the areas (per single molecule) in each of these coexisting phases.Since A = N,b,, O1 = N1/Nm and 0" = N"/Nm, where N , is the total number of single molecules in a dense monolayer, from the expressions given above we obtain the following expressions for the distribution of adsorbate between the two coexisting two-dimensional phases : (31) 0" = o;(or, - O)/(Oif - 0:) 01 = o;(e - s:)/(sg - 03. The mean internal energy of a two-phase, two-dimensional system for a given 0 per single molecule is given by where Uk and G are the mean internal energies of adsorbate in the corresponding phases on the saturation line, while the ratios 01/8 and OV/8 represent the fractions of adsorbate in each of these phases. The internal energy change under two-dimensional evaporation when the phases 1 and v coexist, i.e. the heat of the two-dimensional phase transition, is given by the expression U = (Ul,O1+q@')/6 (32) AUs = G- Uif.(33)1354 PHASE TRANSITIONS OF A N ADSORBATE On differentiating eqn (32) with respect to Tfor constant 8, we obtain an expression for the mean molecular heat capacity of a two-phase, two-dimensional system : C = (C:6'+ ~~v)/6-(AU,,/6)(d8'/dT). (34) Here Ci and are the mean molecular heat capacities of the adsorbate in the corresponding phases. The first term in this equation represents the heat capacity of a two-phase system, while the second term is the contribution of the changes in internal energy occurring when a substance passes from one two-dimensional phase into the other due to temperature variations, in other words the contribution from the heat of the two-dimensional phase transition, AUs, to the heat capacity.A TWO-PHASE SYSTEM I N TERMS OF THE TWO-DIMENSIONAL ASSOCIATED VAN DER WAALS GAS MODEL TWO-DIMENSIONAL CRITICAL STATE On equating the first and second derivatives of the two-dimensional pressure with respect to temperatures given by eqn (21) to zero, we obtain the following expressions for the parameters of the critical state of a two-dimensional associated van der Waals (35) gas : (36) From eqn (36) it follows that the relative critical area oc2/b1, equal to the reciprocal of the critical surface coverage, Bc2, is related to the Kas at c2. When K,, = 0, the ratio ac2/bl is equal to three (8c2 = 0.333), i.e. equal to the value which corresponds to the usual two-dimensional van der Waals gas model.With increasing Kas, the ratio oc2/bl decreases rapidly and approaches the limit equal to two (6,3 = 0.5). From eqn (36) we find that the critical temperature Tc2 increases with increasing Kas. a c 2 = bl/ec2 = b1{2- 3Kas + V"l+ (3&s)21) Tc2 = 2(0c2 - bd2 a1 V' [ 1 + 4Kas b, /(a, 2 - b1)1/ k ( ~ d ~ - 0 1 2 Kas FIG. 5-Dependence on the association constant, Kas, of the ratio of the critical temperature of a two-dimensional associated van der Waals gas, c, (2DAVWG), to the critical temperature of a normal two-dimensional non-associated van der Waals gas, c, (2DVWG).G . I. BEREZIN A N D A. V. KISELEV 1355 The effect of association on the critical temperature is shown in fig. 5 . The ratio of the critical temperature, calculated by eqn (36) with due regard for eqn (35) for different Kas, to the critical temperature of a non-associated van der Waals gas (Ka, = 0) is plotted along the x-axis.From fig. 5 it can be seen that even a small association constant (say, Kas = 0.5) has a strong effect on the critical temperature of a two-dimensional gas (in our case by ca. 50%). TWO-DIMENSIONAL SATURATED VAPOUR PRESSURE According to Maxwell's rule, the phase transition of a gas into a liquid at T c TC2 is determined by the expression n, (av - a') = jIr n d a (37) where n, is the pressure of a two-dimensional saturated vapour. From eqn (8), (21) and (36) we obtain an expression for the pressure of a two-dimensional vapour on the saturation line : where /?f and /?; are, respectively, the degrees of association in a two-dimensional liquid and in a two-dimensional vapour at the saturation line.Using numerical methods, from eqn (8), (19) and (38) we can find 6; and @ for given al/b,, K,, and AUa, at different temperatures. The curve drawn through the values of 0: and 0: thus found is the binodal of a two-dimensional fluid. The binodal peak corresponds to the critical parameters of the system. The binodal of a two-dimensional associated van der Waals gas is shown in fig. 6. For the sake of comparison, the discontinuous curve in this figure shows the binodal for the usual non-associated van der Waals gas, which is shifted along the T-axis to the T,, value for 2DAVWG. Fig. 6 shows how association affects the properties of a two-dimensional liquid and a two-dimensional vapour: association leads to an increase in density for the two-dimensional liquid, but to a decrease in density for the two-dimensional gas existing in equilibrium with the liquid.I 1 0.8 0.9 1 .O TITO, FIG. 6.-Typical temperature dependences of the density of a two-dimensional liquid and a two-dimensional vapour existing in equilibrium (binodal) for the two-dimensional associated van der Waals gas model (2DAVWG, solid curve) and for normal two-dimensional van der Waals model (2DVWG, discontinuous curve). The ratio T/T,, is plotted along the abscissa. (c, of the dashed line is below cz of the solid line.)PHASE TRANSITIONS OF A N ADSORBATE 1356 The quantity is the coefficient (1 1 4 ) (dwf,/dT) = - (1 /Of,) (d@f,/d T ) of thermal expansion of a two-dimensional liquid, and (1 /ax) (dwl/d T ) = - (1 /O,”) (d@/d 7) is the coefficient of thermal expansion of a two-dimensional saturated vapour.They have different signs and tend to infinitely large values near the critical temperature. HEAT OF TWO-DIMENSIONAL EVAPORATION From eqn (33) and (24) it follows that for a two-dimensional associated van der Waals gas we have At T + T,, we find that pf, + 1, & + 0, 6; + 1 and 0,“ + 0. In this case, the heat of two-dimensional evaporation is AUS = @b -m AUas + (a,/b,) (0; - 0,”). (39) AUs = AWas+a,/b,. (39’) However, near the critical temperature + &, 0; + 0: and AUs -+ 0. HEAT CAPACITY OF A TWO-PHASE SYSTEM The heat capacities for the binodal branches, CA and q, differ in magnitude from Cw [eqn (27)]. Under equilibrium conditions for a two-phase system C,l and are defined to be the derivatives of the change in internal energy of a substance adsorbed in each phase, 1 and v, with respect to two variables, viz.the temperature and the corresponding w in this phase. In terms of the thermodynamics of a three-dimensional fluid, these definitions relate to the heat capacity of phases co-existing along the saturation line. Applying these definitions to a two-phase system corresponding to the 2DAVWG model, after differentiating eqn (27) with respect to T we obtain As before, the subscript s in these equations stands for the values of the corresponding parameters along the saturation line, while the superscripts 1 and v denote that these quantities belong to the corresponding two-dimensional phases. Using eqn (34) we can calculate the heat capacity of a two-phase system at the temperature far away from the critical temperature with the help of eqn (39)-(41). However, near the critical temperature we should take into account the contributions made to the heat capacity by fluctuations.CONCLUSIONS In this paper a two-dimensional associated van der Waals gas (2DAVWG) model has been considered for the monomolecular adsorbed layer. This model can be used to study the effect of association on all thermodynamic properties of this layer both above and below the critical temperature of the two-dimensional state. In comparing the theoretical dependences of the thermodynamic parameters on 0 and T with the corresponding experimental dependences due account should be taken of the effectG.I. BEREZIN AND A. V. KISELEV 1357 of formation of a second layer on the experimental parameters. The 2DAVWG model has been applied to describe the experimental data in Parts 2 and 3 of this paperg below and above the critical temperature of the two-dimensional state. The influence of the formation of the second layer has been accounted for with the help of a simple model. APPENDIX For the identity i-1 to be satisfied, where 9 < 1, it is necessary that this series be convergent. We shall now determine the ratio of the ith term of this series to the (i- 1)th term: where 0 < q < 1 and R is the radius of convergence of the series R = (i- l)/i, (R -+ 1 as i --+ a). Consequently, the series will converge, provided KasoI/(l-@ 4 Now from eqn (10) We thus obtain Hence ei G qei-l. K,, = ei(i - e)/(oiPl ol). Kasel/(i -e) = oi/eiP1. This result is the necessary condition for the identity (A 1) to be satisfied. W. A. Steele, The Solid-Gas Interface, ed. E. A, Flood (Marcel Dekker, New York, 1967). N. N. Avgul', A. V. Kiselev and D. P. Poshkus, Adsorbtciya gasov iparov na odnorodnykhpoverkhno- stuyakh (Adsorption of gases and vapours on homogeneous surfaces) (Izd. Khimiya, Moscow, 1975). A. V. Kiselev and D. P. Poshkus, Faraday Symp. Chem. Soc., 1980, 15, 13. E. A. Masson and T. H. Spurling, The Virial Equation of State (International Encyclopedia of Physical Chemistry and Chemical Physics, Topic 10, The Fluid State), ed. J. S. Rowlinson (Pergamon Press, Oxford, 1969), vol. 2. N. P. Vukalovich and I. I. Novikov, Uravnenie sostoyaniya real'nykh gasov (Equation of state of real gases) (Gosenergoisdat, Moscow and Leningrad, 1948). A. V. Kiselev, Dokl. Acad. Nauk SSSR, 1957, 117, 1023. G. I. Berezin, Dokl. Acad. Naut SSSR, 1973, 212, 1134. G. I. Berezin and A. V. Kiselev, J . Chem. Soc., Faraday Trans. I , 1982, 78, in press. ' G. I. Berezin and A. V. Kiselev, J . Colloid Interface Sci., 1974, 46, 203. lo G. I. Berezin and A. V. Kiselev, Adv. Colloid Interface Sci., in preparation. l1 J. H. De-Boer, Dynamic Character of Adsorption (Oxford University Press, London, 1953). l2 H. Kempter and R. Mecke, 2. Phys. Chem. Abr. B, 1940,46, 229. (PAPER 0/928)

ISSN:0300-9599    

DOI:10.1039/F19827801345

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

J, Chem. SOC., Faraday Trans. I , 1982, 78, 1359-1367 Adsorption and Micellisation of a Surface-active Dye in Aqueous Methanol Solutions BY TOYOKO TMAE, CHISAKO MORI AND SHOICHI IKEDA* Department of Chemistry, Faculty of Science, Nagoya University, Chikusa, Nagoya 464, Japan Received 19th January, 198 1 The surface tension of aqueous methanol solutions of a surface-active dye, p-t-octylphenol yellow amine poly(ethy1ene oxide), has been measured by means of the drop weight method. At a given methanol content, < 20%, the surface tension of methanol+ water solutions of the dye exhibits two break points when plotted against the logarithm of the dye concentration, which indicates two-step micellisation of the dye molecules. The adsorption of the dye on aqueous surfaces is cooperative and multimolecular.An analysis of the surface tension data leads to the following results: the primary micelles are formed when the concentration is higher than ca. mol dm-3, and their aggregation number, at most, 15, suggesting that the micellar structure consists of a stack of molecules; the secondary micelles are formed when the concentration exceeds ca. lo-* mol dm-3, and their formation is induced by the hydrophobic mechanism, as for the common surfactant. In methanol + 0.1 mol dm-3 HC1 solutions the dye is protonated and is adsorbed monomolecu- larly on aqueous surfaces: the micellisation occurs as for the common surfactant. Most dyes associate into reversible aggregates in aqueous so1utions,1-6 with a structure composed of a stack of planar dye molecules. On the other hand, surfactants in aqueous solutions associate into micelles above a certain critical micelle concentration (c.M.c.).' We anticipate that a combination of an amphiphilic structure with a conjugated n-electron system in a single molecule will develop a new feature of micellisation.p-t-Octylphenol yellow amine poly(ethy1ene oxide), whose chemical formula is given by OH x + y =10,20 H3C-C -CH, HCH I I I CH3 H,C-C--CH, is a non-ionic surface-active dye, consisting of a hydrophilic part and a hydrophobic part linked by an azobenzene group. The surface activity and micelle formation of this dye can be observed in the behaviour of the surface tension of its aqueous solutions. In this paper we describe the results of surface tension measurements and discuss the adsorption of the surface-active dye on aqueous surfaces and its micellisation in solutions.Since the dye is sparingly soluble in water, aqueous methanol solutions of 13591360 SURFACE TENSION OF A SURFACE-ACTIVE D Y E low methanol content are used as solvent. To examine the effect of protonation of the dye on the adsorption and micellisation, we also measure the surface tension of acid solutions of the dye, for which the solvents are mixtures of methanol and 0.1 rnol dm-3 HCl. EXPERIMENTAL MATERIALS Two samples of p-t-octylphenol yellow amine poly(ethy1ene oxide) with different degrees of polymerisation of the polyethylene oxide parts were kindly donated by Dr F. Tokiwa of the Kao Soap Co. Ltd. The hydrophobic part of the dye, which is called p-t-octylphenol yellow amine, was prepared by reacting an aqueous solution of p-nitrophenyldiazonium chloride with an alkaline ethanol solution of p-t-octylphenol followed by reaction of the product with an alkaline ethanol solution of Na,S.Both dyes were dark red oils and were used without further purification. The average degree of polymerisation of the polyoxyethylene parts, x+y, of the given dye samples was 10 and 20, respectively. Methanol was a spectrograde reagent from Nakarai Chemical Co. Ltd and water was redistilled from alkaline KMnO, in a glass still. METHODS A stock solution of the surface-active dye was prepared by dissolving the dye in methanol and then adding water to the solution until the desired methanol content was obtained. The composition of the solvent is expressed as the volume fraction or molar concentration of methanol.The methanol + water solution of the dye was prepared by diluting the stock solution with the solvent mixture. The methanol + 0.1 mol dm-3 HCl solution of the dye was prepared in the same way using 0.1 mol dm-3 HC1 in place of water. The surface tension, y (mN m-l), of a solution was measured by the drop weight method and was calculated from the weight, w, of a drop falling from a capillary tip having an outer radius, r, by the Harkins-Brown equation whereg is the gravitational acceleration and Fis the correction factor having an argument V/r3, in which V is the volume of a drop. Two capillary tips were used and their effective radii, r, were determined to be 0.3504 and 0.3984 cm, respectively, referring to the surface tension of water.The temperature was kept at 25 & 0.01 OC. The upper end of the capillary tip was connected to a micrometer syringe and its lower end was kept in a glass bottle dipped in a water thermostat. Each drop was suspended on the tip for 5 min before being allowed to fall, when it was detached by a final screwing of the micrometer. Five drops of each solution were collected in a weighing bottle for weighing. It was confirmed that the equilibrium value of the surface tension was reached within 3 min even in the range around the c.m.c. RESULTS SURFACE TENSION Fig. 1 shows the variation of surface tension, y, with the logarithm of molar concentration, log C, of the dye having x +y = 10 in methanol + water solution.With increasing concentration of the dye, the surface tension changes in roughly four steps. At low concentrations it is almost constant and remains essentially equal to the value for the methanol+water mixture. Above ca. 9 x low6 mol dm-3 it decreases sharply and almost linearly until it reaches a break point. Beyond this break it decreases less sharply but almost linearly and reaches a second break, above which the surface tension remains constant and is almost independent of methanol content.T. IMAE, C. MORI A N D S. IKEDA 1361 80 60 - I E z f + . 4 0 20 I 1 I - 5 - 4 - 3 - 2 log (C/mol dm-? FIG. 1.-Plot of surface tension against the logarithm of the molar concentration of the dye for methanol+water solutions of the dye having x+y = 10. Methanol content (volume %): 0, 1 ; 0, 2; a, 5 ; (3, 10; 0, 20.TABLE ~.-SUFWACE ADSORPTION AND MICELLISATION OF THE DYE HAVING x + y = 10 IN METHANOL WATER SOLUTIONS methanol C,I C;' r:, r g content /lop5 /10-lo A O m Yc (vol. %) mol dm-3 mol dmp3 mol cm-2 mol cm-2 rn molecule-la /mN m-I ~ 1 3.50 36.3 9.8 1.71 5.8 17.0 28 .O 2 2.91 23.4 11.4 1.54 7.4 14.6 27.8 5 1.38 10.0 25.2 1.67 15.1 6.59 28.2 10 1.83 9.8 11.3 1.54 7.3 14.9 28.2 20 3.20 10.3 5.3 1.55 3.4 31.3 28.5 In the region where the surface tension decreases with increasing concentration, the dye is adsorbed on aqueous surfaces, and at the two break points the first and second micellisations occur. The primary micelles are formed beyond the first break and the secondary micelles are formed above the second break, beyond which a constant surface tension is exhibited. The values of the first and second c.m.c., C,I and C;*, are listed in table 1 , together with the value of the constant surface tension, yc.Note that the rate of lowering of surface tension is largest for a methanol content of 5%. Addition of methanol to aqueous solutions of the dye generally reduces the surface tension, but in the presence of (1-2) x mol dmP3 of the dye the surface tension is lowest at a methanol content of 5%. The surface tension curves of methanol+water solutions of the dye having x+y = 20 are similar to those of the dye having x+y = 10 above, as shown in fig. 2. The values of the two c.m.c., C,I and Cil, are listed in table 2. The rate of lowering of the surface tension with an increase in dye concentration is lowest at a methanol content of 2%.1362 SURFACE TENSION OF A SURFACE-ACTIVE DYE The change in surface tension of the methanol + water solutions of the dye can be represented in the form of the Gibbs adsorption isotherm -dy = RT(TLdlnC,+T'dlnC) (2) where R is the gas constant, T the temperature, C, the molar concentration of methanol and r k and r' are the apparent surface excess densities of methanol and the dye, respectively.The apparent surface excess density of the dye is given by It is clear that r' = 0 above the second c.m.c. FIG. 2.-Plot of surface tension against the logarithm of the molar concentration of the dye for the methanol+water solutions of the dye having x+y = 20. Methanol content (volume %): (>, 2; a, 5; a, 20.TABLE 2.-sURFACE ADSORPTION AND MICELLISATION OF THE DYE HAVING X + y = 20 IN METHANOL + WATER SOLUTIONS methanol Ci CO'I r:, r: content /1O-lo /10-lo Yc (vol. %) mol dm-3 mol dmP3 mol cmP2 mol cm-2 rn molecule-' /mN m-' 2 0.95 6.18 14.3 2.44 5.9 11.7 25.9 5 1.23 3.27 9.6 2.21 4.3 17.3 26.4 20 1.33 6.17 5.4 2.12 2.5 31.0 27.6 Fig. 3 shows the surface tension of methanol + 0.1 mol dmP3 HC1 solutions of the dye as a function of molar concentration of the dye having x+y = 20. In this case the surface tension curve has a single break point corresponding to the c.m.c. The rate of lowering of surface tension is larger in 2 and 5% methanol solutions than in the 20% methanol solution. This behaviour is also similar to that observed without HCl. In the methanol+O.l mol dm-3 HCl solutions the surface-active dye should beT.IMAE, C. MORI A N D S. IKEDA 1363 protonated at the azo or amino group, since the pK value of 4-di- methylaminoazobenzene is 3.5 in water8 and the substitution of a hydroxyl group at its 2'-position would raise the pK value. We assume that all the dye molecules are protonated when the dye concentration is lower than the c.m.c. As shown in table 3, the value of the c.m.c., C,, of the dye in the acid solutions is larger than that of the second c.m.c. charged dye. 80 60 I E ZE 1 ?- 40 20 in the neutral solutions. This is in accord with the formation of I I 1 1 - 5 - 4 -3 log (C/mol dm-3) FIG. 3.-Plot of surface tension against the logarithm of the molar concentration of the dye for the methanol+O. 1 mol dm-3 HCI solutions of the dye having x+y = 20.Methanol content (volume %): 0 9 2 ; 0, 5 ; a), 20. TABLE 3.-sURFACE ADSORPTION AND MICELLISATION OF THE DYE HAVING X - k y = 20 IN METHANOL + 0.1 mol dm-3 HCl SOLUTIONS methanol (vol. %) mol dm-3 /10-lo mol cm-2 /A2 molecule-' /mN m-l content CO G H C l A0 Yc ~ 2 8.02 4.80 34.6 29.7 5 10.6 4.72 35.2 29.5 20 34.5 3.39 49.0 30.5 Since we are not concerned with the change in HCl concentration but fix it at 0.1 mol dm-3 HCl in water, we can relate the change in surface tension in terms of eqn (2) and (3). Again it is clear that r' = 0 above the c.m.c. THE GIBBS ADSORPTION ISOTHERM METHANOL+WATER SOLUTIONS OF THE DYE The Gibbs adsorption isotherm for the methanol+water solutions of the dye is examined in a general way. The species present in the solutions are water (W), methanol (M), and monomer (I), primary micelle (m) and secondary micelle (mn) of the dye.Then the Gibbs equation is written1364 SURFACE TENSION OF A SURFACE-ACTIVE DYE where Ti and pi are adsorption density and chemical potential of the species, i, respectively. Two-step micellisation equilibria hold for the dye, D. We assume ideal solutions, so that chemical potentials of solute species are expressed by pi=&+RTlnCi i = M , 1,mandmn (7) where p: is the standard chemical potential of the species, i, at Ci = 1 mol dmP3 in pure water, and Ci is its molar concentration. Introducing the Gibbs convention, TW = 0, and the equilibrium conditions Km = % (5’) Clm Cmn Kmn = - Gl we have the Gibbs equation in terms of C,, C,, m, n, Km and Kmn, where K , and Kmn are the association constants.(6’) At a given methanol content, m, Km, n and Kmn are constant. Then we have - dy = RT(T, + rn Tm + mn rmn) d In C,. C = C, + m Cm + mn Cmn. (8) (9) Here the total dye concentration is given by We apply eqn (8) separately for the three regions divided by the two c.m.c. (0 C<C,I: C , = C and Tm= Tmn=O. Eqn (8) reduces to -dy = RTTldln C. (10) Comparison of eqn (10) with eqn (2) gives r/ = rl. Then the saturated adsorption density of the dye, rb, is given by the apparent surface excess density of the dye at the first c.m.c. and just below it. The values of r b are given in tables 1 and 2. The forms of the surface tension curve, i.e. the constancy of surface tension and its steep decrease above 9 x mol dm-3, indicate abrupt or cooperative adsorption of the dye on the aqueous surface above (approximately) this concentration.(ii) Ci < C < Cil: C,+mCm= C and Tmn=O. dlnC. Eqn (8) leads to When m Cm % C, we have -dy = RT(r,+mrm) C1+ m2Cm r/ = rl+mrm m If the primary micelle is not adsorbed on an aqueous surface, because of its more hydrophilic nature, we may set r’ = 1. (14) r mT. IMAE, C. MORI AND S. IKEDA 1365 The constant slope of the surface tension curves in this region gives the adsorption density, rg, as given in tables 1 and 2. In eqn (14) r’ = r g and rl is equal to rb. Then we can obtain the aggregation number of the primary micelle by using eqn (14). These values are listed in tables 1 and 2. The aggregation number of the primary micelle of the dye is rather small, as compared with that of the surfactant micelle.(iii) C > Cil: C,+mCm+mnCmn = C. Eqn (8) gives d In C. c C, +m2Cm + (mn)2Cmn -dy = RT(Tl+mrm+mnTmn) Using similar assumptions to those used in (ii), we have In this region of dye concentrations, r’ = 0 and rl = rb. Consequently, the values of mn should be large, and the secondary micelle of the dye has a size as large as that of the usual surfactant. METHANOL+O.~ mol dm-3 HCl SOLUTIONS OF THE DYE In methanol + 0.1 mol dmP3 HC1 solutions the monomeric dye is completely ionized, but the dye molecules incorporated in the micelle are deprotonated and non-ionic, as the absorption spectra Then the Gibbs equation is written in terms of neutral species as -dy = r w d ~ w + ~ M ~ P M + r HCidpHCi + r i d p i + r p d ~ p (17) where the suffixes, 1 and p , represent the protonated monomer (DHCl) and the non-ionic micelle (D,) having aggregation number p .The micellisation equilibrium is expressed by p DHCl D, + p HCl. (18) We assume that the solution is ideal so that eqn (7) holds for i = M and p , and the chemical potentials of HCl and 1 are given by pHC1 = p&Cl -kRTln (cA-c,) C A (194 p, = p:+RTlnC,C, (19b) where pGC, and p: are the standard chemical potentials of pure HC1 and pure protonated monomer at CA = 1 mol dm-3 and C, = 1 mol dm-3, respectively, and CA is the molar concentration of added HC1. Then the equilibrium condition is given by where K , is the equilibrium constant. constant. Then we have At a given methanol content and the fixed HCl concentration, p and K , are Here the Gibbs convention, Tw = 0, has been introduced and also we have c = c,+pcp. (21)1366 SURFACE TENSION OF A SURFACE-ACTIVE DYE In the present work we may set CA % C,.(9 C < C, : C, = C and r,=O. Eqn (20) reduces to -dy = RTT,dln C. (22) Comparison of eqn (22) with eqn (2) leads to r’ = r l . The saturated adsorption density of the protonated dye, rgHC1, is given by the apparent surface excess density of the dye at the c.m.c. and just below it. The values of rODHCl are given in table 3. (ii) c > Co: C,+pC, = c . When p C, B C,, we have If the micelle is not absorbed on an aqueous surface, we may set r (25) r’ = 1. As seen in fig. 3, r’ = 0 and rl is equal to TgHC1. Consequently, the aggregation number of the micelle, p , of the dye is very large, as in the case of the usual surfactant.P DISCUSSION ADSORPTION DENSITY AND MOLECULAR AREA OF THE DYE The behaviour of the surface tension of methanol+water solutions of the dye is unique, showing two-step micellisation. As can be seen in tables 1 and 2, however, the adsorption density, rh, of the dye, ca. mol cma2, is several times larger than that of the usual surfactant.1°-12 The molecular area, A , = l/NTb, of the dye, where 1V is Avogadro’s number, ca. 15 A2 or less, is also correspondingly small when compared with the values obtained for polyoxyethylene alkylphenyl ether on aqueous surfaces. lo, l3 The surface pressure-area curves of insoluble monolayers of several 1,3-diglycerides consisting of an azo group and a long-chain fatty acid group gave limiting molecular areas of ca.40 A2.14 Note also that the observed molecular area of the dye depends on the methanol content of the solvent. This observation together with the small molecular area imply that the dye is adsorbed multimolecularly on surfaces of methanol + water solutions. In methanol+O.l mol dmP3 HCl solutions the surface tension behaves as for aqueous solutions of the usual surfactant. The adsorption density, rgHC1, is equal to ca. (4.0+0.8) x lo-’, mol cm-2, and is close to the values for various ionic surfactants.129 15 The corresponding molecular area, A , = 1 / N r G H c l , equal to 40 A2, indicates that the protonated dye is adsorbed monomolecularly on aqueous surfaces. CRITICAL MICELLE CONCENTRATION The first c.m.c.of the dye in methanol + water solutions occurs at ca. mol dm-3 and is lower than the c.m.c. of the usual non-ionic surfactants by approximately an order of magnitude.’,. 11* l3 The primary micelle of the dye in methanol + water solutions consists of, at most, 15 dye molecules. The formation of such a small micelle would not be caused merely by the hydrophobic effect, and the structure of the primaryT. IMAE, C. MORI AND S. IKEDA 1367 micelle would consist of a stack of dye molecules, as proposed for some other dye micelles.2y 5* The second c.m.c. of the dye in methanol+water solutions has values comparable with the c.m.c. of polyoxyethylene derivatives having dodecyl,16-18 octylphenyl or nonylphenyll0, 1 3 9 l9 groups. The secondary micelle has an aggregation number sufficiently high to protect the hydrophobic part of the dye molecules from water.Similarly, the micelles of the dye in methanol + 0.1 mol dm-3 HCl solutions would have a structure similar to those of the usual surfactants. Note that in methanol+water solutions the c.m.c. of the more hydrophilic dye having x +y = 20 are lower than those of the homologue having x +y = 10. Such an apparently opposing phenomenon has been observed with some polyoxyethylene derivatives of n-o~tadecanol.~~~ 2o Consequently, this seems to be characteristic of polyoxyethylene derivatives having a strongly hydrophobic group. THE EFFECT OF ADDED ALCOHOL It is k n o ~ n ~ l - ~ ~ that the addition of a small amount of short-chain alcohol to aqueous surfactant solutions lowers the c.m.c., but further addition of alcohol increases the c.m.c.Consequently, the minimum c.m.c. is observed at a certain content of added alcohol. The present observation of the effect of added methanol on the second c.m.c. in methanol+water solutions is in agreement with values previously found. However, in methanol + 0.1 mol dm-3 HCl solutions the effect of added methanol is to raise the c.m.c., and this would be caused by the effect of deprotonation of the dye upon micellisation. P. Alexander and K. A. Stacey, Proc. R. SOC. London, Ser. A , 1952, 212, 274. H. P. Frank, J. Colloid Sci., 1957, 12, 480. N. Mataga, Bull. Chem. SOC. Jpn, 1957, 30, 375. P. Mukerjee and A. K. Ghosh, J. Am. Chem. Soc., 1970, 92, 6403. D. Pugh, G. H. Giles and D. G . Duff, Trans. Faraday SOC., 1971, 67, 563. B. H. Robinson, A. Loffler and G. Schwarz, J . Chem. SOC., Faraday Trans. 1 , 1973, 69, 56. I. M. Klotz, H. A. Friess, J. Y. Chen Ho and M. Mellody, J. Am. Chem. SOC., 1954, 76, 5136. T. Imae, C. Mori and S. Ikeda, J. Chem. SOC., Faraday Trans. 1, 1982, 78, 1369. ’ C. Tanford, The Hydrophobic Eflect (Wiley, New York, 1973). lo L. Hsiao, H. N. Dunning and P. B. Lorentz, J. Phys. Chem., 1956, 60, 657. l1 J. Stauff and J. Rasper, Kolloid Z., 1957, 151, 148. l 2 S. Ozeki and S. Ikeda, Bull. Chem. SOC. Jpn, 1980, 53, 1832. l3 M. J. Schick, J. Colloid Sci., 1962, 17, 803. l4 J. Heeseman, J. Am. Chem. SOC., 1980, 102, 2167. l5 K. Tajima, Bull. Chem. Soc. Jpn, 1971, 44, 1767. l6 J. M. Corkill, J. F. Goodman and R. H. Ottewill, Trans. Faraday SOC., 1961, 57, 1627. l 7 E. H. Crook, D. B. Fordyce and G. F. Trebbi, J. Phys. Chem., 1963, 67, 1987. l8 H. Lange, Kolloid Z., 1965, 201, 131. l9 S. Ikeda and K. Kakiuchi, J. Colloid Interface Sci., 1967, 23, 134. *O M. J. Schick, S. M. Atlas and F. R. Eirich, J. Phys. Chem., 1962, 66, 1326. 21 A. F. Ward, Proc. R . SOC. London, Ser. A, 1940, 176, 412. 2* W. D. Harkins, R. Mittelman and M. L. Corrin, J. Phys. Colloid. Chem., 1949, 53, 1350. 23 B. D. Flockhardt, J. Colloid Sci., 1957, 12, 557. 24 P. Becher and S. E. Trifiletti, J. Colloid Interface Sci., 1973, 43, 485. 25 D. E. Guveli, J. B. Kayes and S. S. Davis, J . Colloid Interface Sci., 1979, 72, 130. (PAPER 1 /084)

ISSN:0300-9599    

DOI:10.1039/F19827801359

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

J. Chem. Soc., Faraday Trans. 1, 1982, 78, 1369-1376 Absorption Spectra and Micellisation of a Surface-active Dye in Aqueous Methanol Solutions TOYOKO IMAE, CHISAKO MORI AND SHOICHI IKEDA* Department of Chemistry, Faculty of Science, Nagoya University, Chikusa, Nagoya 464, Japan Received 19th January, 198 1 The absorption spectra of aqueous methanol solutions of a surface-active dye, p-t-octylphenol yellow amine poly(ethy1ene oxide), have been measured at different dye concentrations, and the absorption bands are assigned to tautomeric forms of the 2’-hydroxy-4-dimethylaminoazobenzene derivative. The K band of the dye appears at 394 nm in methanol+ water solutions when the dye concentration is lower than mol dmP3. Hypochromism occurs and the shoulder band (B band) at 460-480 nm becomes manifest when the primary micelle is formed: the primary micelle is stabilised by the stacking interaction of dye molecules and the increased amount of their o-quinoid hydrazone tautomer.At higher dye concentrations the K band appears at 413 or 406 nm and its intensity increases. This step can be attributed to the second micellisation of dye molecules due to hydrophobic interaction. The methanol+O. 1 mol dm-3 HC1 solutions of the dye exhibit two absorption bands (K’ and Q bands) at 320 and 528 nm, which can be assigned to the ammonium form and the resonance-stabilised azonium tautomer, respectively, of the protonated dye, when the dye concentration is dilute. Above the critical micelle concentration the absorption spectra have the K band at 405 nm, indicating that the dye is deprotonated in the micelle.It has been observed that a surfactant with a chromophoric group undergoes a sudden change in the absorption spectra of its aqueous solutions when it associates into micelles as the concentration increases. Harkins et a1.l were the first to observe this phenomenon for dodecylpyridinium iodide, and Ray and Mukerjee2v measured its absorption spectra in different solvents. Ikeda and Fasman4 found a sudden increase in the molar extinction coefficient of the absorption bands of polyoxyethylene p-t-octylphenyl ether above its critical micelle concentration, and Rehfeld5 observed an abrupt decrease in the absorption intensity of sodium phenylundecanoate upon micellisation. The surface-active dye, p-t-octylphenol yellow amine poly(ethy1ene oxide), is expected to show some spectral changes when micelles are formed in aqueous solutions, because the molecule has an azo group sited between two phenyl groups.From measurements of the surface tension of aqueous solutions of the dye it was found that two-step micellisation occurs for the dye in methanol +water solutions, while the common micelle is formed in methanol + 0.1 mol dm-3 HCl solutions.6 The dye is a derivative of 2’-hydroxy-4-dimethylamino-azobenzene, and the complicated behaviour of the absorption spectra of its aqueous methanol solutions must be understood by referring to its tautomeric forms and resonance. In order to obtain detailed information on the structure of the micelles from the electronic structure of molecules in micelles, we measure the absorption spectra of the dye in methanol + water and methanol + 0.1 mol dmP3 HCl solutions, together with the spectra in various organic solvents, and examine the micellisation of the dye as welQ as its structure.45 1369 FAR 11370 ABSORPTION SPECTRA OF A SURFACE-ACTIVE DYE EXPERIMENTAL p-t-Octylphenol yellow amine poly(ethy1ene oxide) samples were the same as previously used,6 and were kindly donated by Dr F. Tokiwa of the Kao Soap Co. Ltd. They had an average degree of polymerisation of the polyoxyethylene parts of x + y = 10 and 20, respectively. Water, dioxan, 1,2-dichloroethane and 1,2-dibromoethane were redistilled before use, after the necessary purifications. Benzene, chloroform, ethyl acetate and concentrated HCl solution were special-grade reagents.The other organic solvents were spectrograde reagents of the Nakarai Chemical Co. Ltd. The dye was sparingly soluble in water, ethyl ether and cyclohexane. The method of preparation of the aqueous methanol solutions of the dye was the same as previously described.6 The absorption spectra were measured on a Shimadzu UV-200s spectrophotometer and recorded on a U-l25MU recorder. Quartz cells with path lengths of 10, 5, 2 and 1 mm were used. The temperature of the cell chamber was adjusted to 25 f 0.05 O C by circulating water of constant temperature from a Haake thermobath FS. RESULTS The absorption spectra of the surface-active dye having x+y = 10 in 2 and 20% methanol+water solutions are illustrated in fig. 1 and 2, respectively.At concentrations lower than mol dm-3 the absorption spectra have a band at 0' I I I I I 350 400 4 5 0 500 550 w avelengt h/n m FIG. 1.-Absorption spectra of the dye having x+y = 10 in 2% methanol+water solution. Concentration/mol dm-3: -, 4.66 x lop6; ---, 1.86 x lop5; ---, 7.76 x . . . , 6.21 x 394 nm with a shoulder at ca. 450 nm. The spectra of the dye show a feature of the 4-aminoazobenzene derivative in aqueous ethanol solutions, and the main band termed the K band7-10 can be assigned to the lowest n-n* transition of azo d ~ e . ~ ? l l The shoulder band appearing at 460-480 nm cannot be attributed to the n-n* transition of azo dyes since its molar extinction coefficient is too high. In fact, the spectral feature is more similar to that of 2-hydroxyazobenzene12~ l3 and 1 -phenylazo- 2-naphtholl49 l5 in aqueous ethanol solutions, which have a shoulder at wavelengths 50-70 nm longer than the K band.The subsidiary band, originally termed the B band,13 can be assigned to the hydrazone tautomer, as will be described later. With increasing concentration of the dye, the K band shifts to the red and the B band is stronger, but both bands are hypochromic. The concentration dependence of the wavelength of the main bands is shown in fig. 3. The spectra do not change withT. IMAE, C. MORI A N D S. I K E D A 1371 dye concentration in 50% methanol + water solutions. The red shift and hypochromism occur at the range of concentrations around the first c.m.c. and may be attributable to the stacking interaction of the dye molecules.We have previously6 found that the primary micelle has an aggregation number of, at most, 15 and consequently we can imagine that it is composed of a stack of dye molecules. 2 - I I 1 I . . . . . FIG. 2.-Absorption spectra of the dye having x+y = 10 in 20% methanol+water solution. Concentration/mol dm-3: -, 3.82 x ---, 3.05 x lop5; - . - . -, 6.36 x 10-5; ---, 1.27 x 10-4; . . . . -, 5.09 x 10-4. 5 420 x" 400 . 2 - 6 - 5 - 4 - 3 log (C/mol dm-3) FIG. 3.-Wavelength of the main band of the dye having x+y = 10 plotted against concentration in methanol+water solutions. Methanol content (volume %): 0, 1 ; @, 2; 0, 5; 0, 10; (0, 20; $, 50. In the region of the second c.m.c. or exceeding it the K band remains at 413 or 406 nm. The observed spectral shift will be caused by the facts that the molecules in the secondary micelle are free from hydrogen bonding with solvent, as shown below, and they are in a hydrophobic environment.Fig. 4 shows the absorption spectra of the surface-active dye having x+y = 10 in 2% methanol+O.l mol dm-3 HCl solutions. The dye is protonated in these solvents 45-21372 ABSORPTION SPECTRA OF A SURFACE-ACTIVE DYE in which the apparent pH is 1.1, since the pK of 4-dimethylaminoazobenzene is ca. 3.5 in water16 and 2.2 in 50% ethan01.~ The spectra have bands at 320 and 528 nm at concentrations < mol dm-3, and they are similar to those observed for amino- substituted azobenzenes in acid media.99 lo, 1 7 9 l8 The two bands can be assigned as the K' and Q bands, which are attributed to the ammonium form and the resonance- stabilised azonium form, respectively, as will be discussed later.When the dye concentration exceeds l 0-4 mol dm-3 the absorption spectra change drastically and the two bands are replaced by a single band at 405 nm having a shoulder at ca. 480 nm. The main band at 405 nm can be identified with the K band, and in this concentration range the micellisation is manifest from the surface tension I ' I I 1 - 0 E € 1 2 m P 0 1 n W I I 300 400 500 600 wavelength/nm FIG. 4.-Absorption spectra of the dye having x + y = 10 in 2% methanol+O.l mol dm-3 HCl solution. ---, 9.35 x Concentration/mol dm-3: -, 1.87 x ---, 4.67 x . . . . ., 9.35 x C W Vl ? n 2.5 2.0 1.5 ~~ - 5 - 4 - 3 log (C/mol dm-3) FIG. 5.-Molar extinction coefficient of the main bands of the dye having x+y = 20 plotted against concentration in methanol +O.1 mol dm-3 HCI solutions. Top: the Q band; bottom: the K band. Methanol content (volume %): (>, 2; a, 5 ; a, 20.T. IMAE, C. MORI A N D S. I K E D A 1373 measurements.G Thus the surface-active dye must be deprotonated and non-ionic in the micelle, even if the outside media are strongly acidic and the free dye molecules are totally protonated. The dye having x+y = 20 behaves similarly to that having x+y = 10, in both methanol + water and methanol + 0.1 mol dm-3 HC1 solutions. Fig. 5 illustrates the molar extinction coefficients of the two bands at ca. 400 and 530 nm plotted against the dye concentration in methanol + 0.1 mol dmP3 HCl solutions. The plots clearly show the presence of a concentration where the molar extinction coefficient at 530nm suddenly decreases at each methanol content.The break points, (7.1, 9.8 and 25) x mol dm-3 at 2, 5 and 20% methanol contents, can be well compared with the c.m.c. found by the surface tension measurements, (8.02, 10.6 and 34.5) x mol dm-3. TABLE WAVELENGTH OF THE K BAND OF THE DYE IN VARIOUS ORGANIC SOLVENTS carbon tetrachloride benzene toluene c yclohexane ethyl ether acetone ethanol methanol water ethyl acetate dioxan chloroform 1,2-dichloroethane 172-dibromoethane 382.5 387 388 377 388 395 396 395 (394) 39 1 390 385 388 388 - 415 417 41 5 420 398 379 40 1 417 412 - - 20 -21 - 20 ( - 26) -7 1 1 - 16 -31 - 24 Table 1 shows the wavelength of the K band of the dye having x+y = 10 dissolved in various organic solvents.It was found that the wavelength, A, of the K band in five solvents, i.e. ethyl ether, toluene, benzene, carbon tetrachloride and cyclohexane, follows the McRae equation19 - = - + A ( + - ) 1 1 z2 -1 (-L) D-1 C 2 - 1 a aG 2fiD+1 + B D+2 iib4-2 where aG is the wavelength of the K band of the dye in the hypothetical gaseous state, 6, is the refractive index of the solvent at the D line and D is the dielectric constant (relative permittivity) of the solvent. The parameters, aG, A and B, have been evaluated (2) as Table 1 gives the wavelength of the K band calculated for the other solvents using eqn (1) and (2). Although some polar halogenated hydrocarbons must be excluded, the deviation, A1 = Lobs-acal, can be taken as representing the effect of hydrogen bonding between the dye and solvent.In some strongly hydrogen-bonding solvents such as alcohols and acetone, the observed wavelength is 20 nm blue shifted from the calculated one. The calculated wavelength for these solvents, ca. 415 nm, is AG = 309 nm, A = - 28 576 cm-l, B = 4830 cm-l.1374 ABSORPTION SPECTRA OF A SURFACE-ACTIVE DYE approximately equal to the observed wavelength for the secondary micelle. Thus we can imagine that the dye molecules are free from hydrogen bonding in the secondary micelle and are effected only by the polarity of the environment within the micelle. DISCUSSION STRUCTURE OF THE DYE I N NEUTRAL SOLUTIONS Since the stability or presence of the tautomeric forms and resonance species of 2-hydro~yazobenzene~~9 1 3 9 20* 21 and 4-amino- or 4-dirnethylamino-azoben~ene~~~ 18* 22 has been subject to controversy, and since neither 2’-hydroxy-4-aminoazobenzene nor 2’-hydroxy-4-dimethylaminoazobenzene has ever been described, we examine the tautomeric equilibrium and resonance of the surface-active dye in aqueous methanol solutions on the basis of the absorption spectra observed.APPEARANCE OF THE SHOULDER BAND AT 460-480 nm 0spensonl2 and Burawoy and Chamberlain13 showed that 2-hydroxyazobenzene exhibited two absorption bands in aqueous ethanol solutions, one being the strong K band’. l2 and the other the B band at a higher wavelength.l29 l3 For l-phenylazo- 2-naphthol the K band is weaker relative to the B band.l2, 1 3 9 l5 The K band has been assigned to the azo form and the B band to its o-quinoid hydrazone form.l2? 1 4 9 1 5 9 23 In a similar way we may postulate the keto-enol tautomerism of 2’-hydroxy- 4-dimethylaminoazobenzene represented by Q--N;.p-Q--NRR~ / \ == Q N - y o N R R / 0-H O b b e H (I) in which both the azo (I) and o-quinoid hydrazone (11) forms are associated with the intramolecular hydrogen Evidence for the formation of o-quinoid forms of some derivatives of 4- aminoazobenzene can be found in their absorption spectra.Ross and Warwick17 reported that 2’-hydroxy-5’-ni tro-4-di(2-chloroethyl)aminoazobenzene exhibited only the B band in 95% ethanol, which we interpret as the complete shift of the keto-enol equilibrium towards the hydrazone form. They also found that the introduction of a 2’-carboxyl group in 4-dimethylaminoazobenzene or 4-N-methyl-N-(2-chloro- ethy1)aminoazobenzene enhanced the intensity of the B band and they ascribed this effect to the stabilisation of the p-quinoid form.The resonance Raman spectra of aqueous solutions revealed that Tropaeolin 0, i.e. 2,4-dihydroxyazobenzene-4’- sulphonic acid, can assume either or both the quinoid forms (0- and p-) at neutral pH.25 The present surface-active dye has a characteristic shoulder band at 460-480 nm, in addition to the K band, in methanol+water solutions, while the shoulder is not observed in methanol nor in any other organic solvents. As we have referred above to the tautomeric equilibrium of the derivatives of 2’-hydroxy-4-aminoazobenzene, the B band of the dye can be interpreted as being due to the formation of the o-quinoid hydrazone form.Thus we can conclude that the tautomeric equilibrium is displaced towards the azo form in methanol or other organic solvents, while it is considerably shifted towards the hydrazone form in methanol + water solutions. Note also that the hydrazone form is more stabilised in the primary micelles than in the monomer form or in the secondary micelles.T. IMAE, C. MORI AND S. IKEDA 1375 SUPPRESSION OF THE RED SHIFT OF THE K BAND IN WATER Forbes and Milligan18 reported that 4-dimethylaminoazobenzene had the K band at 406 nrn in ethanol and at 446 nm in water, and that Methyl Orange, i.e. 4-dimethylaminoazobenzene-4’-sulphonic acid, had the K band at 463 nm in water, which was shifted to the red by 45 nm from the K band in ethanol. They proposed various modes of solvation of the 4-aminoazobenzene derivatives and attributed the spectral difference to the formation of a hydrated form that could stabilise a polarised resonance species through an equilibrium.Brode et aZ.22 also explained the spectral shift by postulating the formation of a hydrated form of the dye which was hydrogen- bonded with water at the azo nitrogen atom. On the other hand, the surface-active dye has the K band at 394 nm in both methanol and methanol+water solutions. This behaviour of the dye can again be explained by the effect of substitution of a 2’-hydroxyl group. From the scheme by Forbes and Milligan, it is likely that the substitution of a 2’-hydroxyl group prevents the azo-hydrated form of the dye from being formed, because the azo group is protected from water molecules by its intramolecular hydrogen bond.Then some of resonance species, such as the polarised one, which should be more stabilised in more polar solvents, are not formed even in aqueous solutions, and the number of resonance species remains the same in aqueous solutions as in methanol. Thus the location of the K band of the dye is not influenced by the nature of the hydrogen-bonding solvent. MICELLISATION OF THE DYE IN NEUTRAL SOLUTIONS The surface-active dye in methanol + water solutions shows spectral changes in two steps, approximately following the two-step micellisation.6 The first step includes the red shift and hypochromism of the main bands, which can be attributed to the stacking interaction of dye molecules, leading to the formation of the primary micelle.The second step is characterised by a further spectral shift of the K band and a steady increase in intensity of the main bands. These changes can be ascribed to the release of an azo chromophore from hydrogen bonding and its transfer into the more hydrophobic environment. The former can be seen from the results of eqn (1) and (2) in table 1. The latter is supported by the observation that the K band of 4-dimethylaminoazobenzene shifts from 420 nm to 401 or 398 nm on going from 50% ethanolg* lo to cyclohexane18 or iso-octane,26 if the spectral shift is regarded as a blue shift on going from the primary micelle to the secondary micelle. The B band becomes stronger relative to the K band upon the first micellisation, indicating formation of the o-quinoid form, but it becomes weaker at the second micellisation.STRUCTURE AND MICELLISATION OF THE DYE IN ACIDIC SOLUTIONS The two absorption bands exhibited by the dye in acidic solutions can be attributed to the free protonated dye, i.e. the monomeric cation, which can assume tautomeric isomers and resonance species as shown by + + Q . N = J O ! R ~ - i_ Q ~ + ~ N R R / t-) Q - ~ - ; ~ ~ ~ R I 0-H 0-H 0--H (rm (IYa 1 ( IYb 1 While the K’ band at 320 nm is assigned to the ammonium form (111), the Q band at 528 nm is associated with the azonium form (IVa), which is stabilised through resonance with the p-quinoid immonium species (IVb). The formation of p-quinoid1376 ABSORPTION SPECTRA OF A SURFACE-ACTIVE DYE species of 4-dime thylaminoazobenzene derivatives in acidic solutions was demonstrated by means of absorption spectra9? lo, 1 7 9 27 and resonance Raman 28 In acidic solutions it is uncertain whether or not the o-quinoid hydrazone form would arise from the ketwmol tautomerism.mol dm-3 has been interpreted as being due to the micellisation accompanying deprotonation of the dye molecules. Thus the K band appearing at ca. 405 nm means that the dye is non-ionic in the micelles. The pK value of the dye is much lower in the micelles than in the monomeric cations. From the spectral and surface tension behaviour, the micelle of the dye in acidic solutions would have a structure similar to the secondary micelle in neutral solutions. A similar but opposite effect of micellisation on the protonation reaction was also found in the case of dimethyldodecylamine oxide:29 it has a pK value of 4.78 in the free cationic form in water, while it becomes less acidic in the micelle, giving an intrinsic pK value ca.0.85 pH units higher.30 More indirect micellar effects on acid-base equilibria of pH indicators have been the subject of recent investigations to elucidate the polarity and electrostatic potential of m i ~ e l l e s . ~ ~ ~ ~ ~ When Methyl Orange in acidic solutions forms a complex with octadecyltrimethylammonium chloride, in which the surfactant concentration is generally higher than the c. m. c., the deprotonation of Methyl Orange cations occurs, as revealed in the absorption It is likely that the pK value of a N-dimethylamino group is lowered by the micellar The drastic change in spectra of the dye at 33 W.D. Harkins, H. Krizek and M. L. Corrin, J. Colloid Sci., 1951, 6, 576. A. Ray and P. Mukerjee, J. Phys. Chem., 1966, 70, 2138. P. Mukerjee and A. Ray, J. Phys. Chem., 1966, 70, 2144. S. Ikeda and G. D. Fasman, J. Polym. Sci., Part A-I, 1970, 8, 991. S. J. Rehfeld, J, Colloid Interface Sci., 1970, 34, 518. ti T. Imae, C. Mori and S. Ikeda, J. Chem. Soc., Faraday Trans. I , 1982, 78, 1359. A. Burawoy, J. Chem. Sac., 1937, 1865. P. Birnbaum, J. H. Linford and D. W. Style, Trans. Faraday SOC., 1953, 49, 735. G. M. Badger, R. G. Buttery and G. E. Lewis, J. Chem. Soc., 1954, 1888. lo G. E. Lewis, Tetrahedron, 1960, 10, 129. l1 M. B. Robin and W. T. Simpson, J. Chem. Phys., 1962, 36, 580. l2 J. N. Ospenson, Acta Chem. Scand., 1951, 5, 491. l3 A. Burawoy and J. T. Chamberlain, J. Chem. Soc., 1952, 3734. I4 J. N. Ospenson, Acta Chem. Scand., 1950, 4, 1351. l5 A. Burawoy, A. G. Salem and A. R. Thompson, J. Chem. SOC., 1952, 4793. l6 I. M. Klotz, H. A. Fiess, J. Y. Chen Ho and M. Mellody, 1. Am. Chem. SOC., 1954, 76, 5136. l7 W. C. J. Ross and G. P. Warwick, J. Chem. SOC., 1956, 1719. la W. F. Forbes and B. Milligan, Aust. J. Chem., 1962, 15, 841. 2o A. Burawoy and I. Markowitsch, Liebigs Ann., 1933, 503, 180. 21 D. Hadii, J. Chem. SOC., 1956, 2143. 22 W. R. Brode, I. L. Seldin, P. E. Spoerri and G. M. Wyman, J. Am. Chem. SOC., 1955, 77, 2762. 23 R. Kuhn and F. Bar, Liebigs Ann., 1935, 516, 413. 24 S. B. Hendricks, 0. R. Wulf, G. E. Hilbert and U. Liddel, J. Am. Chem. SOC., 1936, 58, 1991. 25 Y. Saito, B-K. Kim, K. Machida and T. Uno, Bull. Chem. SOC. Jpn, 1974, 47, 21 1 1 . 26 W. R. Brode, J. H. Gould and G. M. Wyman, J. Am. Chem. Soc., 1953,753, 1856. 27 A. Hantzsch and A. Burawoy, Chem. Ber., 1930, 63, 1360. 28 K. Machida, B-K. Kim, Y. Saito, K. Igarashi and T. Uno, Bull. Chem. SOC. Jpn, 1974, 47, 78. 28 F. Tokiwa and K. Ohki, J. Phys. Chem., 1966, 70, 3437. 30 H. Maeda, M. Tsunoda and S. Ikeda, J. Phys. Chem., 1974, 78, 1086. 31 M. S. Fernandez and P. Fromberg, J. Phys. Chem., 1977,81, 1755. 32 C. A. Bunton, L. S. Romsted and L. Sepulveda, J. Phys. Chem., 1980, 84, 261 1 . 33 C. F. Hiskey and A. T. Downey, J. Am. Chem. Soc., 1954, 58, 835. E. G. McRae, J. Phys. Chem., 1957, 61, 562. (PAPER 1 /085)

ISSN:0300-9599    

DOI:10.1039/F19827801369

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

J . Chem. SOC., Faraday Trans. I, 1981,78, 1377-1388 Viscosity B Coefficients of Some Alkyltrimethylammonium Bromides and the Effect of Added Alkan-1-01s D O ~ A N E. G U V E L ~ ~ Department Physicochimie des Solutions, Universite Pierre et Marie Curie, 11 Rue Curie, 75231 Paris, France Received 20th February, 198 1 The viscosity B coefficients for aqueous solutions of alkyltrimethylammonium bromides and decyltri- methylammonium bromide containing various concentrations of aliphatic alcohols have been determined at 25 OC. Viscosity B coefficients for the alkyltrimethylammonium bromides were positive and increased as the alkyl chain length increased; the B coefficient increase per methylene group was 0.084. The addition of aliphatic alcohols decreased the viscosity B coefficient of decyltrimethylammonium bromide.As the alcohol concentration was increased the measured partial molal volume of decyltrimethylammonium bromide first decreased and then increased, and this was correlated with the viscosity B coefficient. The addition of alcohol is thought to effect the solvent structure and solute-solvent interactions by changing the composition and dielectric constant of the environment. The viscosity B coefficient is considered to be a measure of long-range forces, solvent structural effects and size and shape factors.' Feakins et aL2 developed a relationship between the relative viscosity, viscosity B coefficient and partial molal volume of the solute, applying transition-state theory. They examined the effect of methanol on viscosity B coefficients of electrolytes.The solute-solvent interactions in aqueous alkyltrimethylammonium bromides and alkysulphate solutions have been examined by Tanaka et al.3 in terms of the Jones-Dole equation. Additional valuable information about solute-solvent interactions can be obtained from partial molal volume studies. Kaneshina et al.4 reported that partial molal volume (E) values below the critical micelle concentration of the sodium alkyl sulphates decreased at low added methanol concentrations and then increased. This result was supported by the observation of Vikingstad and K~ammen,~ who studied the effect of alkan-1-01s on the partial molal volume of sodium decanoate, and the increase in was related5 to the hydrophobic hydration of the surfactant monomer.In earlier studies6* the effects of alkan-1-01s on the hydrodynamic and volumetric properties of micellar systems of alkyltrimethylammonium bromides in aqueous solution were examined. In this work the effect of alkan-1-01s on water structure and the surfactant-solvent interaction in the singly dispersed state are discussed in terms of the viscosity B coefficient and the partial molal volume of the surfactant monomer. EXPERIMENTAL MATERIALS The preparations and purification of the alkyltrimethylammonium bromides and the purification of the alkan-1-01s (methanol, ethanol, propanol and butanol) have been described elsewhere.6 Water was doubly distilled from an all-glass still of specific conductance < 1 x R-' cm-l. t Present address: School of Pharmacy, University of Bradford, Bradford BD7 IDP.13771378 VISCOSITY B COEFFICIENTS DENSITY AND VISCOSITY MEASUREMENTS The densities of the alkyltrimethylammonium bromides C,TAB (x = 10, 12, 14 and 16) and decyltrimethylammonium bromide (C,,TAB) solutions containing various concentrations of alkan-1-01s at 25 f 0.02 'C were made using a Paar density meter model DMA 02. The viscosity measurements were carried out at 25 f 0.02 OC using two U-tube capillary viscometers. The viscometers were calibrated with pure water and 20% sucrose solutions according to British Standard 188.* The viscosities were calculated from average flow times, t , and densities, d, based on the equation. q = d(Ct-:) where the characteristic constants C and B of the viscometers were 3.5 x lop3, 3.8 x lop3 and 36.52, 0.3, respectively. The accuracies of the density and viscosity measurements were (2f 1) x g C M - ~ and O.Ol%, respectively, and the density of water was taken as 0.9971 g cm-3.9 RESULTS AND DISCUSSION VISCOSITY STUDIES Jones and Dole, studying dilute electrolyte solutions, ShowedlO that the viscosity of electrolytes can be related to the following relationship q = qo(l + A d C t BC) (2) where ylo is the viscosity of solvent, C is the concentration, and A and B are the characteristic constants of the solute.Since pre-micellar aggregation was not observed in the solutions of ionic surfactant an association of monomers below the critical micelle concentration (c.m.c.) is unlikely to occur. Therefore the viscosity data of aqueous surfactant solutions and decyltrimethylammonium bromide solution containing various concentration of alkan- 1-01s below the c.m.c. were examined by the Jones-Dole equation.Eqn (2) can be rearranged as A plot of (qr- l)/z/C against d C yields the constants A and B. Eqn (2) is generally used to provide information about the solute-solvent interaction. The viscosity of unsolvated spherical colloidal suspensions can be represented by (4) Einstein's equation" qr = 1+2.50 where CD is the volume fraction of the solute; by combining with eqn (2), eqn (4) becomes 2.5 CD = A d C + BC. ( 5 ) Since the A2/C term can be neglected in comparison with BC, and 0 = CV, where Vis the partial molal volume, then eqn ( 5 ) takes the form 0.0025 V = B. (6) In the ideal case the B coefficient is a linear function of the solute partial molal volume ( V ) with a slope equal to 0.0025.12 The B coefficient can be interpreted1 as consisting of two terms B = Bsize + Bsolv (7)D.E. G U V E L ~ 1379 where Bsize is the effect of solute and Bsolv is the contribution arising from solute-solvent interaction. Thus Bsolv12 Bsolv = B-0.0025 V. (8) In order to check the accuracy and precision of the viscosity technique for the investigation of solute-solvent interactions in dilute solutions alone and with the added alcohols, the B values of a series of alkyltrimethylammonium bromides (C,TAB) below the critical micelle concentration in aqueous solutions were determined by a least-squares fit of a plot of (v, - l)/z/C against d C . The viscosity A and B constants obtained for C,TAB and the standard errors for the estimation of viscosity constants, including correlation c’: iefficients, are given in table 1.The A coefficients for C,TAB, which represent the contribution from interionic electrostatic forces,13 were small. Fig. 1 illustrates the dependence of (qr- l ) / d C on z/C, which is linear up to the c.m.c. of C,TAB, and then increases non-linearly as the concentration increases. The linear increase in (q, - l ) / d C below the c.m.c. is a result of the structuring of water molecules around the hydrocarbon chain of the surfactant molecule and the solvation of the hydrophilic group.6 However, the change in the slope of the plots (at c.m.c.) is related to the release of ordered water molecules around the surfactant monomers, and the apparent increase in (qr- l)/z/C above the c.m.c. is considered to be due to the presence of associated monomers, electrostatic interactions and hydration.6 Fig.2 shows that there is a linear increase in the B coefficient with increasing hydrophobicity of the surfactant molecule. This can be attributed to the hydrogen-bonded cluster size produced through hydrophobic interactions.6 The B values obtained for C,,TAB and C,,TAB, 0.77 and 0.95, respectively, compare favourably with those found by Tanaka et al.,3 (0.76 and 0.91 for C,,TAB and C , ,T A B, respectively ). The relationship between the B coefficient and the alkyl chain length for C,TAB can be expressed by the following linear equation (9) B = a + B(CH,)CH where a and B(CH,) are constants. The value obtained for B(CH,) (0.084) is in good agreement with those reported for alkyltrimethylammonium bromides14 (0.080), amino acids15 (0.084) and alkyl sulphates16 (0.079), whilst being slightly higher than the value (0.076) per methylene group given by Tanaka et al.3 The ionic B values (Bion), obtained by Kaminsky’s17 procedure using his data for Brian ( - 0.042), are given in table 1.The present data show that the ionic Bcoefficient follows similar trends as does the B coefficient, and that the B coefficient increases linearly with increasing for C,TAB was linear with a slope (0.0051) higher than the value (0.0025) given by eqn (6). This result is related to the non-ideality which occurs because of the hydration of monomers and electroviscous effects. The value of Bsize indicates an increase with increasing alkyl chain length.However, in contrast to Bsize, the values of Bsolv were reasonably constant for the various C,TAB used. In fact, the major contribution to the B coefficient arises from Bsize and is related to the solute-size structuring effect on water structure through hydrophobic interactions. (fig. 2). Thus a least-squares fit of a plot of B against PARTIAL MOLAL VOLUME STUDIES The partial molal volume of a solute in solution provides important information on the nature of solute-solvent interaction^.^ The apparent partial molal volume ofTABLE l.-V~scos~ry B COEFFICIENTS FOR THE ALKYLTRIMETHYLAMMONIUM BROMIDES IN WATER AT 25 OC ~~~~ - standard standard C error error E7 below ~ r - 1 110-3 mol tiona tiona coeffic- below / mol for the for the correla- /cm3 dm+ c.m.c.6 estima- estima- tion m o P C,TAB c.m.c.4 C dm-3 A (4 B (B) ient Bsize Bsolv c.m.c. Bion B; B(CH,) 60 50 40 30 20 10 16 14 10 4 2 1 3 2 1 0.8 0.6 0.7 0.6 0.5 0.4 0.1 I 0.202 0.185 0.170 0.148 0.123 0.091 0.121 0.1 12 0.078 0.059 0.047 0.106 0.084 0.079 0.072 0.068 0.077 0.075 0.07 0.059 65 16.8 3.7 0.8 0.015 0.017 0.04 0.046 1.86 5.41 7.36 1.54 0.77 0.95 1.13 1.27 0.99 0.061 0.191 0.07 1 0.999 0.991 0.959 0.995 1.32 -0.55 258.2 1.42 -0.47 278.1 1.62 -0.49 318.5 1.80 -0.53 352 5 m cl - 2 3 0.812 - 4 4 tu cl s 2 crl 0.992 -0.042 0.084 2 m z 4 m 1.172 - - - 1.312 - a All errors assumed in (qr - l ) / d C .FIG. 1 .-Plots of (q, - D. E. GUVELi 1381 I c.rn.c I 1 I 1 0.1 0.2 0.3 0.4 dC A, Cia; 0, Clz; 0, Cia; x 9 Cl6.l)/dC against d C for the alkyltrimethyl ammonium bromides in water at 25 O C : FIG. 2.-Plots of the viscosity B coefficients against the partial molal volume and alkyl chain length for the alkyltrimethylammonium bromides : 0, carbon number; V, partial molal volume.1382 VISCOSITY B COEFFICIENTS decyltrimethylammonium bromide in aqueous solution with added alkan- 1-01s was calculated from18 (10) where no, n and n, are the numbers of moles of water, surfactant and additive, and M,, M and M, are the corresponding molecular weights. To obtain the partial molal volume (c) of the monomer below the c.m.c., the calculated a value was plotted against concentration and extrapolated to zero. - M n,M,+nM+n,M, a V = - - (dp, T,non2 d d2 EFFECT OF ALIPHATIC ALCOHOLS ON PARTIAL MOLAL VOLUME In previous work7 the effect of alkan-1-01s at higher concentrations on the partial molal volume (E) of dodecyltrimethylammonium bromide was examined.The value for C,,TAB showed an increase with increasing alcohol concentration, and for higher added alcohol concentrations (> 1 mol dm-3) there was no evidence of a minimum in E. In order to examine the effect of alkan-1-01s on solute-solvent 260 7 258 E 6 256 - 0 rn 13” 2 54 1 0.1 OI.3 0’5 0.7 0.9 1.1 alcohol concentration/mol dmd3 FIG. 3.-Effect of concentration of added alcohols on the partial molal volume of decyltrimethylammonium bromide below the c.m.c. at 25 O C : 0, methanol; A, ethanol; 0, propanol; +, butanol. I interactions in the singly dispersed state of C,,TAB, a series of aliphatic alcohols (methanol, ethanol, propanol and butanol) at different concentrations (0.2- 1 mol dm-3) was added to aqueous solutions of C,,TAB.Fig. 3 shows that the added alcohols have a considerable effect on for C,,TAB. Dealing first with methanol, at low alcohol concentrations decreased slightly showing a minimum in E, then increased as the alcohol concentration was raised. Ethanol was less effective with regard to changes in E, and the minimum became less pronounced as the chain length of the alcohoIs increased. This effect has also been observed by Kaneshina et aZ.,4 who have reported that the values of sodium alkyl sulphates decreased at low alcohol concentrations and then increased. This result was also confirmed by the observationD.E. G U V E L ~ 1383 for sodium dodecanate by Vikingstad and K~ammen.~ It is known19 that at low alcohol concentrations the order of solvent structure increases and that the transfer of a hydrocarbon from a non-polar environment to water results in a negative volume change, which reflects the ordering of water molecules around the hydrocarbon chains.20 In this work at low concentrations, the changes in are similar to those in v f o r hydrophobic in alcoholic solutions. The minimum in E indicates the maximum structuring effect of the surfactant monomer on the water-alcohol structure. The minimum and increase in E in the ternary systems are related to solute-solvent interactions. It has been reported5 that the addition of alcohol reduces the electrostriction of water molecules at the ionic head group of the surfactant molecule, which increases the value of V of this group. However, the increase in E at higher alcohol concentrations is due to the effect of alcohols on the water structure, both in terms of a decrease in the hydrophobic hydration of the hydrophobic part of the surfactant monomer, as well as a decrease in the hydrophilic hydration of the ionic head group.' The concentration giving rise to the minimum in E appears to be dependent on the alkyl chain length of the alcohol; the longer the chain length, the greater the effect on because of the hydrophobic effect associated with the hydrophobic part of the alcohol molecule.EFFECT OF ALIPHATIC ALCOHOLS O N THE VISCOSITY B COEFFICIENT The addition of aliphatic alcohols to decyltrimethylammonium bromide affects the viscosity of the surfactant solution by changing the solvent composition. The increase in viscosity is due to the structuring of water molecules around the hydrocarbon chain of the surfactant monomer and solvation of the hydrophilic groups.6 Thus, when various concentrations of methanol were added the relationship between (q, - l)/z/C and z/C differed in the low and high alcohol concentration regions.Fig. 4 shows that (vr - l)/z/C falls linearly as the alcohol concentration increases, followed by a steady increase above the c.m.c. The other alcohols showed similar effects. The viscosity A and B coefficients of decyltrimethylammonium bromide containing various concen- trations of aliphatic alcohols below the c.m.c.in aqueous solution are shown in table 2. They were determined by a least-squares fit of a plot of (qr- l)/z/C against z/C, and include the standard errors for the estimation of viscosity constants and correlation coefficients. It is apparent that the coefficient B changes with increasing alcohol concentration. Changes in B for strong electrolytes with added alcohol have been observed by several 2 1 p 22 Table 2 shows that the addition of alcohol to an aqueous solution of C,,TAB does not contribute in the expected way to an increase in the B coefficient. Thus, at a given concentration of added alcohol (for example 0.2 mol dmP3 methanol), the viscosity B coefficient of C,,TAB decreased while the A coefficient showed an increase, table 2. The decrease in the B coefficient of C,,TAB with equal concentrations of alkan-1-01s was linear in the order butanol > propanol > ethanol > methanol (fig.5). Changes in the B coefficients of strong electrolytes with added ethanol were also observed by Padova,21 who reported that the B coefficient decreased with increasing ethanol concentration until the alcohol concentration was ca. 40% and that the effect of ethanol on B was related to the negative interaction coefficients and the decrease in entropy of the mixed solvents. Feakins et aZ.22 studied the effect of added methanol on the B coefficient of the electrolytes and they observed that small amounts of methanol enhanced the structure of the mixture compared with that of water, and an increase in the structure-breaking effect caused a decrease in the B values of electrolytes.The concentration causing a decrease in the B coefficient for C,,TAB appeared to be dependent on the alkyl chain length of the alcohol; the longer the chain length, the greater the lowering effect on1384 VISCOSITY B COEFFICIENTS B. This can be attributed to the effect of lowering the dielectric constant, which increases as the alcohol chain length increases. Although many studies have appeared5? 6 y ‘ 9 which have concluded that alkan-1-01s at low concentrations in micellar solutions and in water alone increase the ordering of water molecules, the observed results for the B coefficient of C,,TAB in alcoholic solutions do not show similar trends. There are a number of complexities to consider concerning the behaviour of alkan- 1-01s in an aqueous surfactant solutions.The competing effects involved that lead to a negative viscosity B coefficient have yet to be resolved. Cox and W ~ l f e n d e n ~ ~ ‘t “a A’ A. / H , I 0.1 0.2 0.3 0.4 dc FIG. 4.-Plot of (qr - l)/v’C against d C for solutions of decyltrimethylammonium bromide containing various concentrations of methanol at 25 OC: 0, 0.2; V, 0.4; 0, 0.7; A, 1 mol dm-3. concluded that the negative viscosity B coefficient was due to depolymerization of the water structure. Gurney24 reported that if a solute causes a local loosening of the water structure, then the viscosity B coefficient is negative. The existence of a negative B coefficient was also related12 to the incompatible ordering effect of the ion on solvent structure.It is apparent that in the solvent mixture the observed B values do not depend on the solvent composition in a simple way. However, Tominaga et c 1 1 . ~ ~ demonstrated the occurrence of decreased counter-ion binding to the surfactant monomer with increasing alcohol concentration. This finding was in agreement with the observation of Larsen and Tipley,26 who reported an increase in free counter-ion concentration in alcoholic surfactant solution. It is expected that the addition of alcohol gives rise to an increase in the concentration of free bromide ion in terms ofD. E. G U V E L ~ -1 - -3- B -5- - 7- 1385 TABLE 2.-EFTECT OF ALIPHATIC ALCOHOLS ON THE B COEFFICIENT OF DECYLTRIMETHYLAMMONIUM BROMIDE AT 25 *c standard standard error for error for alcohol the the B con.es tima tiona estimationa correlation per mole /moldm-3 A (4 - B (B) coefficient /cm3 rno1-I of alcohol 0.2 0.4 0.7 1 .o 0.2 0.4 0.7 1 .o 0.2 0.4 0.7 1 .o 0.2 0.4 0.7 1 .o 0.36 0.64 0.82 1.08 0.51 0.88 1.02 2.05 0.76 1.16 1.99 2.71 0.84 1.35 2.35 3.49 0.021 0.023 0.078 0.075 0.024 0.03 0.058 0.147 0.04 0.016 0.042 0.1 10 0.029 0.074 0.120 0.042 CH,OH 0.58 0.121 1.48 0.137 1.79 0.426 2.22 0.416 C,H,OH 0.82 0.052 1.7 0.156 3.0 0.302 4.47 0.724 C,H,OH 1.75 0.239 2.26 0.08 4.23 0.206 5.70 0.544 C,H,OH 1.7 0.151 2.45 0.365 4.6 0.593 7.05 0.209 0.893 0.987 0.918 0.968 0.995 0.987 0.990 0.974 0.965 0.998 0.997 0.990 0.992 0.978 0.983 0.999 255.2 - 1.88 254.59 254.88 256.00 255.60 -4.54 254.71 255.06 256.20 256.10 -5.7 255.00 255.36 256.97 256.59 -6.82 256.22 256.52 257.92 a All errors assumed in (qr- l)/z/C.'I i i i i alcohol chain length FIG. 5.-B coefficient of decyltrimethylammonium bromide containing equal concentration of alkan- 1-01s as a function of the hydrocarbon chain length of the alcohol molecule at 25 O C : 0, 0.2; 0, 0.4; V, 0.7; A, 1 mol dm-3.1386 VISCOSITY B COEFFICIENTS -1 - -3 - - 5 - B -7- the changing solvent composition in singly dispersed C,,TAB solution. This was also observed by Larsen and Tipley,26 who noted that the concentration of free counter-ions in aqueous solution of C,,TAB followed the changes in the solvent structure. Since the bromide ion is a structure-breaker, an increase in the free ion concentration will decrease the ordering of water molecules. In addition, as the concentration of alcohol is increased, there is dehydration of the monomer6 and a lowering of the dielectric constant of the solvent mixture.As the dielectric constant decreases, a decrease in charge density2' is expected together with a decrease in the surface ionic field. This might increase the structure-breaking effect of the surfactant monomer, which is in accord with the suggestion of Feakins and Lawrence.28 L. I I I I 254 255 256 257 258 - V,jcm3 mol-' FIG. 6.-B coefficient of decyltrimethylammonium bromide containing equal concentrations of alkan- 1-01s as a function of at 25 OC: 0, 0.2; n, 0.4; 7, 0.7; A, 1 mol dm-3. The structural attractive forces arise when there is compatibility between structural influences and repulsive forces, and the establishment of structural order near the non-polar solute is favoured.This produces a hydrogen-bonded water network around the surfactant monomer which leads to a decrease in E. However, when there is incompatibility between structural effects, and the disrupted water molecules exceed the structured molecules in the total volume of the solution, then the structure-breaking effect will be reflected in the solute-solvent interactions. Thus, changes in the B values of C,,TAB with added alcohols were more pronounced and became more negative as the alcohol concentration increased. Fig. 6 shows the relationship between the B coefficient of C,,TAB containing the same amount of alkan-1-01s and E at 25 O C . It is apparent that the viscosity B coefficient decreases non-linearly, reaching a minimum value with increasing E.Dealing first with low alcohol concentrations there is little effect on B until the concentration reaches 0.4 mol dm-3. However, the tendency of B to decrease is reduced when the chain length of the alcohol is increased. This is also reflected in a change in E with the added alkan-1-01s (fig. 3). The hydrophobic hydration results in a structured medium where the solute B coefficient is positive and the slope dB/dD. E. G U V E L ~ 1387 - 2. - 4- B - 6 - deviates from the Einstein relation, being more positive for the larger hydrophobic salts.12 This is in accord with the observation for C,TAB in aqueous solutions, whereas the dependence of B on E, fig. 6, shows that the Einstein relation is no longer valid for C,,TAB with added alcohols, owing to the complicated effects of the alcohols on solute-solvent interactions.Fig. 7 indicates that the change in B coefficient per mole alcohol chain length FIG. 7.-B coefficient of decyltrimethylammonium bromide per mole of alcohol as a function of the hydrocarbon chain length of the alcohol molecule at 25 O C : 0, alcohol chain length. of alcohol, derived from a least-squares fit of a plot of B coefficient against alcohol concentration, decreases with increasing alcohol chain length (table 2). However, the effect as the homologous series is ascended is non-linear, which is related to the fact that the hydrophobic effect associated with the hydrophobic part of the alcohol molecule becomes greater as the length of the hydrocarbon chain of the alcohol increases.I thank Prof. R. Schaal and Dr D. Eagland for their comments on this work. R. H. Stokes and R. Mills, Viscosity of Electrolytes and Related Properties (Pergamon Press, Oxford, 1965). D. Feakins, D. J. Freemantle and K. G. Lawrence, J. Chem. SOC., Faraday Trans. 1, 1974, 70, 795. M. Tanaka, S. Kaneshina, W. Nishimoto and H. Takabatake, Bull. Chem. SOC. Jpn, 1973, 46, 364. S. Kaneshina, M. Manabe, G. Sugihara and M. Tanaka, Bull. Chem. SOC. Jpn, 1976,49, 876. E. Vikingstad and 0. Kvammen, J. Colloid Interface Sci., 1980, 74, 16. D. E. Guveli, J. B. Kayes and S. S. Davis, J. Colloid Interface Sci., 1979, 72, 130. 'I D. E. Guveli, J. B. Kayes and S. S. Davis, J. Colloid Interface Sci., 1981, 82, 307. * British Standard, BS 188 (1957). @ G. S. Kell, J. Chem. Eng. Data, 1967, 12, 66. lo G. Jones and M. Dole, J. Am. Chem. SOC., 1929,51, 2950. l2 J. E. Desnoyers and G. Perron, J. Solution Chem., 1972, 1, 199. l 3 H. Falkenhagen and E. L. Vernon, Phys. Z., 1932, 33, 140. l4 J. E. Desnoyers, M. Are1 and P. A. Leduc, Can. J. Chem., 1969, 47, 547. l5 W. Devine and B. M. Lowe, J. Chem. Soc. A , 1971, 21 13. A. Einstein, Ann. Phys, 1906, 19, 289, 191 1, 34, 591.1388 VISCOSITY B COEFFICIENTS K. Tamaki, Y. Ohara and Y. Isomura, Bull. Chem. SOC. Jpn, 1973, 46, 1551. '' M. Kaminsky, Discuss. Faraday Soc., 1957, 24, 171. E. Hutchinson and C. S. Mosher, J . Colloid Sci., 1956, 11, 352. l9 F. Franks and D. J. G. Ives, Quart. Rev., 1966, 20, I . 2o G. Nemethy and H. A. Scheraga, J . Chem. Phys., 1962, 36, 3401. 21 J. Padova, J . Chem. Phys., 1963, 38, 2635. 22 D. Feakins, B. C. Smith and L. Thakur, J . Chem. SOC. A , 1966, 714. 23 W. M. Cox and J. H. Wolfenden, Proc. R. SOC. London, Ser. A , 1934, 145, 147. 24 R. W. Gurney, Ionic Processes in Solution (McGraw Hill, New York, 1953). 25 T. Tominaga, B. T. Stem and D. F. Evans, Bull. Chem. Soc. Jpn, 1980, 53, 795. 26 J. W. Larsen and L. B. Tepley, J . Colloid Interface Sci., 1974, 49, 113. 27 K. Shirahama and T. Kashiwabara, J. Colloid Interface Sci., 1971, 36, 65. D. Feakins and K. G. Lawrence, J . Chem. SOC. ,4, 1966, 212. (PAPER 1 /297)

ISSN:0300-9599    

DOI:10.1039/F19827801377

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. 1, 1982, 78, 1389-1400 Influence of the Lithium Content on the State, Dispersion and Reducibility of Nickel Oxide Supported on y-Alumina BY MARWAN HOUALLA,~ JACQUES LEMAITRE* AND BERNARDELMON Groupe de Physico-chimie Minerale et de Catalyse, Universite Catholique de Louvain, Place Croix du Sud 1, B- 1348 Louvain-la-Neuve, Belgium Received 9th March, 1981 The influence of lithium on the dispersion, active phase-carrier interaction and reducibility of nickel oxide supported on alumina carriers has been investigated using X-ray diffraction analysis (X.r.d.), ultraviolet diffuse reflectance spectroscopy (d.r.s.), X-ray photoelectron spectroscopy (X.P.S.) and temperature- programmed reduction (t.p.r.). On Li-free alumina, nickel is present in the form of a NiAl,O, compound and dispersed NiO.Increasing the Li content leads to a decrease in the formation of NiA1,0,, due, presumably, to the concurrent formation of Li-containing spinel phases, and a concurrent increase in the reducibility of the nickel-containing phase. The addition of alkali metal ions to alumina-based catalysts is widely used to control unwanted side reactions due to the carrier. This effect is mainly due to modification of the acidity of the surface.’ However, the action of an additive is not restricted to alteration of the surface properties of the solid carrier and other consequences of fundamental importance may take place. They pertain, as shown by Lo Jacono et a1.,2 to the physico-chemical properties of the active phase deposited on the modified aluminas (e.g.modification of the dispersion of the supported material and formation of a chemical association between the active phase and the support273). Many investigations have already been devoted to the study of the influence of additives, but most of the published literature has emphasized the first aspect, e.g. the variations of the acid-base properties of the modified carrier in connection with its catalytic activity. Little attention has been given to the influence of modifiers on the nature of the deposited phase and its dispersion over the carrier. The present paper is part of a systematic study we have recently undertaken to shed more light on the role of additives on the properties of the supported oxides. Recent communications have reported preliminary results concerning the influence of alkali metal ions (Li+, Na+, K+) on the nature, dispersion and distribution of deposited nickel 1 oxide on a y-Al,O, s ~ p p o r t .~ - ~ The salient features of more detailed studies7? in the case of nickel and cobalt oxides supported on a series of Na-modified aluminas are that increasing the Na content of the alumina brings about segregation of the deposited phase to the outer parts of the catalyst particle (inhomogeneous repartition). Concomitantly, an enhancement of the dispersion of the ‘ Co30, ’ and, to a lesser extent, ‘ NiO ’ phases takes place. A tentative interpretation of these results has been p r o p ~ s e d . ~ ~ ~ It rests on the influence of ‘free’ Na-containing species present in Na-rich carriers on the pH of the solution introduced in the pores of the modified carrier during the impregnation by Co- or Ni-containing salts (nitrates).This modification of the pH brings about, through hydrolysis and t Present address : Department of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania 15260, U.S.A. 13891390 EFFECT OF Li ON NiO/y-Al,O, precipitation of the metal ions, a better anchoring of the Co or Ni on the surface of the modified support; but this also causes irregular repartition of the species with the depth in the pores. The purpose of our present work is to investigate the influence of the nature of the alkali metal ions and hopefully to establish general patterns for the action of alkali metal additives. In the present paper we report a study, using X-ray diffraction (X.r.d.), ultraviolet diffuse reflectance spectroscopy (d.r.s.), X-ray photoelectron spectroscopy (X.P.S.) and temperature-programmed reduction measurements (t.p.r.), of the influence of Li content in an alumina carrier on the nature of the deposited nickel ‘ oxide’ phases, the dispersion of these phases and their interaction with the support.EXPERIMENTAL MATERIALS The series of Li-modified Al,O, was prepared by pore volume (‘ incipient wetness’) impregnation of a powdered y-Al,O, (Houdry 0415, surface area: 140 m2 g-l, pore volume: 0.45 cm3 g-*) with solutions containing various concentrations of LiNO,. The samples were then dried at 110 OC and calcined for 12 h at 600 OC. The lithium-modified alumina samples are designated as AlLix, where x indicates the approximate Li content, expressed as wt% Li,O.The exact Li content of the samples is given in table 1. TABLE I.-LITHIuM CONTENT OF THE AlLix SAMPLES AlLix AlLi 0.4 AlLi 0.9 AlLi 1.7 AlLi 2.5 AlLi 3.4 wt% Li20 0.43 0.86 1.71 2.54 3.36 The series of supported nickel oxides was obtained from the AlLix series by pore volume impregnation with a solution of nickel nitrate. The impregnated samples were then dried for 2 h at 110 OC and calcined for 6 h at 500 OC. The nickel content is approximately equal in all the samples and corresponds to 10 wt% NiO. The samples are designated as NiAlLix. Pure NiO and NiAl,O, samples, prepared as model compounds for d.r.s. measurements, were obtained, respectively, by calcination of nickel nitrate at 950 O C and by the ‘hot kerosene drying’ methods using an aqueous solution of nickel and aluminium sulphates, followed by calcination for 16 h at 850 OC.PH Y SI CO-C HE M IC A L CHAR A C TER IZ A TI ON SURFACE AREA MEASUREMENTS The surface areas of AlLix samples were determined by the B.E.T. method using nitrogen. The apparatus used allowed the simultaneous evacuation of all samples of the series AlLix and swift measurement (overall time for 6 samples: 5 h). This minimizes the possible errors. Although the absolute error, without use of a proper standard, might be of the order of 10 m2 g-l, the relative error (from sample to sample in the same series) is & 5 m2 g-l. X-R A Y D I FFR ACT 10 N X-ray diffraction patterns were obtained using a Philips Norelco diffractometer with Cu K, radiation (Ni filtered). NiO crystallite sizes were estimated from the X-ray line broadening (X.r.1.b.) of the (200) line, using Scherrer’s formula.U.V. DIFFUSE REFLECTANCE MEASUREMENTS Diffuse reflectance spectra were obtained with a Beckman Acta IV spectrophotometer, using, as a reference specimen, the pure y-Al,O, carrier pretreated under the same conditions as the Ni-containing samples. In the case of the model compound NiO, a MgO pellet was used as a standard. The measurements were made in the wavelength range 800-350 nm.M. HOUALLA, J. LEMAITRE A N D B. DELMON 1391 X-RAY PHOTO E LE c TRO N s PE c T R 0s co PY General. In this work, we used X.P.S. exclusively for signal intensity measurements. In principle, the position of the peaks associated with the supported species, namely Ni, could have provided valuable information concerning the compounds in which Ni is bound (NiO or NiAl,O,).In fact, the shifts, taking into account a previous study on catalysts on Na-modified support^,^ are probably relatively small, 0.5 eV, and reliable measurements, in the present case, would need additional experiments. Techniques and operating procedures. X.P.S. measurements were carried out with a Vacuum Generators ESCA 2 spectrometer equipped with an aluminium anode (hv = 1486.6 eV) operated at 50 mA and 10 kV. The samples were finely ground and the powder was dusted on double-sided adhesive tape. A Tracor Northern NS 560 signal averager was used in order to improve the signal-to-noise ratio. The residual pressure inside the spectrometer was 1 0-8 Torr.* The signal intensity corresponding to each element is represented by the area under the corresponding peaks. The background produced by the inelastically scattered electrons is assumed to vary linearly with energy. The order of recording of the various lines was kept strictly the same for all samples, namely C( Is), A1(2s), Ni(2pt), A1(2s), C( Is), O( 1s). The intensity of the Ni(2pt) level includes the associated satellite peak. Basis for quantitative measurements. The X.P.S. relative intensity measurements reported here concern the ratio of the intensities, I , and I,, of two peaks associated, respectively, with the supported phase [Ni(2p,)] and the carrier [A1(2s)]. Several models have been proposed to correlate the measured X.P.S.intensity ratio to the composition of the catalyst and dispersion of the species. They take into account the surface area, the shape of the carrier particles and the dispersion of the supported species (monolayer on the carrier, or cry~tallites).'~-~~ Following the models proposed by Defosse et al. and generalized by Kerkhof and Moulijn13 to include the case where crystallites form, we have, assuming a sheet-like structure of the carrier particles, three different situations. (a) Monolayer formation m D, 0, 1 1 + exp (- 2/0, A, S,) I t - s D, us p,A,S, 1 -exp(-2/aSl,S,) - - ~ ~ _ _ _ _ - where m/s is the atomic ratio of the characteristic elements representative of the supported phase m and the carrier s, D,, D, are the detector efficiencies for m and s, which', vary as l/v, v being the kinetic energy of the corresponding electrons, 0, are the photoelectron cross-sections of the Ni(2pJ and Al(2s) levels taken from Scofield,15 Am are the escape depths of the Ni(2p9 and Al(2s) electrons, estimated, respectively, to be 1.051s,17 and 2.5 nm,18 p,.is the density of the carrier (3.8 g ~ m - ~ ) and S, is the specific surface area of the support.It can be readily demonstrated that, in the case of alumina and in the range of surface areas corresponding to our AlLix samples, the factor is very close to unity, so that eqn (1) may be simplified to the following expression: [1 + ~ X P (-2/~sAmSs)l/ 11 - ~ X P (-2/psLS,)I As all the samples studied here have the same nickel content, the intensity ratio I:/It dependent only on the surface area of the sample, so that eqn (1 a) may be written as follows: S,/I$Zt = k .(b) Crystallites where c is the dimension of the edge of the equivalent cubic crystallites of the deposited phase and I z / I t is the intensity ratio which would be predicted in the case of a monolayer. * 1 Torr x 1.33 x lo2 Pa.1392 EFFECT OF Li ON NiO/y-Al,O, Taking into account eqn (1 b), we may write: Eqn (2a) shows, similarly to eqn (1 b), that the factor SsIrn/Is remains unchanged when there is no variation in the particle size of the deposited species. An increase or a decrease in the dispersion of the supported phase leads, respectively, to an increase or a decrease in Ss Irn/Is. When dealing with large particle size (c + Am), a further simplification of eqn (2a) is possible.We may write: In this latter case, Ss Irn/Is varies linearly with the dispersion (1 /c) of the supported phase. (c) Ideal statistical mixing on the atomic scale (solid solution) Limitations. X.P.S. intensity measurements suffer two limitations, of different origin : the first is inherent to the nature of the element analysed. Some elements (e.g. lithium) which have a very low photoionisation cross-section, can hardly be detected within reasonable analysis time. The second limitation is related to the validity of the hypotheses used for predicting the X.P.S. relative intensity ratio Im/Is for various configurations of the supported phase over the carrier. For example, depending on the adopted value for the mean free path of electrons in the supported phase or the carrier, the calculated intensity ratio Irn/Is may suggest different configurations of the supported phase.Similarly, the variation in the repartition of the deposited phase (relative amount of species deposited in the inner parts of the elementary catalyst particle and at the mouth of the pores or on the external part of the grain) may cause misleading conclusions about the dispersion of the supported materiaL5 REDUCTION STUDIES TEMPER ATURE-PROGR A M MED REDUCTION Apparatus. The system, designed according to Robertson et aI.,19 is made entirely of stainless steel and is grease free. The reactor is made of quartz. The system and the reactor were improved according to the suggestions of Cvetanovic and Amenomiya.20 Operating procedure.A sample weighing ca. 50 mg is used. Before reduction, the sample is heat-treated in situ for 1 h at 400 "C under pure oxygen flux (20 cm3 min-I). After cooling, the reactor is purged with a reducing mixture (5% hydrogen in argon) at ambient temperature. The reactor is then heated at a linear rate (10 O C min-I) from ambient temperature to 900 O C , while maintaining a constant flow of the reducing gas (35 cm3 min-I). Downstream of the reactor, water produced by the reduction processes is removed from the gas by passing it through a cold trap, consisting of a U-shaped tube immersed in a frozen heptane bath (-91 "C). The composition of the dried gas is monitored using a thermal conductivity detector. Basis for quantitative interpretation. Owing to the fact that t.p.r.experiments were carried out at the same heating rate and that the ordinate value indicates an instantaneous rate of hydrogen consumption, it follows that the area under the peaks is proportional to the amount of hydrogen consumed. As all the samples examined here contain the same NiO loading, which is in principle the only reducible phase, and H, adsorption is negligible, one can expect practically the same value for the total area under the peaks for all the NiAlLix samples.M. HOUALLA, J. LEMAITRE AND B. DELMON 1393 RESULTS LI THIU M-MODI FIED s A M PLES : AlLix SURFACE AREAS The variation of the specific surface area, S,, as a function of the lithium content of the alumina is shown in fig. 1. Taking account of the method of measurement, these variations are significantly larger than the maximum possible relative errors (vide supra).2 0 0 0 2 4 100 Li,O (wt %) FIG. 1.-Variation of the specific surface area S, of the modified aluminas AlLix as a function of Li,O content . After a slight increase in the surface area for 0 < x < 1 wt% Li,O, there is a quite significant decrease with further lithium addition (this decrease is larger than that corresponding to a decrease in the amount of Al,O, in each gram of modified carrier). Similar variations were reported by Levy and Bauer in the course of their study of the effect of foreign ions on the stability of activated alumina.21 Small changes in the values obtained through the B.E.T. method may reflect either real variations of the surface area, or relatively large changes of pore geometry.We did not measure the pore size distribution. However, in view of the relatively small effects observed, the simplest interpretation is that the effect is mainly related to a change in surface area. X-RAY D I FFR A c TION The X-ray diffraction spectra of representative lithium-modified alumina samples (x = 0, 1.7, 3.4) are shown in fig. 2. Analysis of the line positions and intensities of the original alumina indicates good agreement with those reported for y-Al,O,. X.r.d. spectra of lithium-containing samples are closely related to those of the unmodified carrier. However, as the lithium content increases, sharper lines are obtained, indicating a slightly better crystallinity. ‘NICKEL OXIDE’ SUPPORTED ON Li-MODIFIED ALUMINAS: NiAlLix X-RAY DIFFRACTION X-ray diffraction spectra of nickel-containing samples (fig.3) show, in addition to the diffraction pattern of the corresponding aluminas, lines which may be attributed to NiO. The intensity of these lines steadily increases as the lithium content of the carrier is increased. The average NiO particle size of NiAlLix samples determined by the X.r.1.b. method varies between 5.3k0.3 nm for x = 0 and 6.0k0.4 nm for x = 3.4. The presence of a NiAl,O, phase in NiAlLix samples cannot be easily established1394 EFFECT OF Li ON NiO/y-Al,O, I 0 60 50 40 291" FIG. 2.-X-ray diffraction patterns of representative AlLix samples. (A) AlLi 0; (B) AlLi 1.7; (C) AlLi 3.4. I 1 1 0 60 50 40 3c 291" FIG. 3.-X-ray diffraction patterns of typical NiAlLix samples.(A) NiAlLi 0; (B) NiAlLi 3.4.M. HOUALLA, J. LEMAITRE A N D B. DELMON 1395 using X.r.d. measurements.2 This is due, essentially, to the low amount of nickel involved and the closely related pseudo-spinel structures of y-A1203 and NiA1204. In addition, the latter might be amorphous instead of crystalline. In only a few instances2 has the incipient formation of NiA120, been proved by a small variation in the intensity of the y-A1203 diffraction lines. A similar attempt in our case is impeded by the relatively high degree of crystallinity of the alumina carrier. DIFFUSE REFLECTANCE SPECTRA The d.r.s. spectra of the NiAlLix series are shown in fig. 4. All the NiAlLix samples exhibit one adsorption band centred at 420 nm and one at 620 nm. Absorption in these wavelength regions corresponds2 to the most intense bands for octahedral and tetrahedral Ni2+, respectively.The band at 620 nm is relatively weak, but it is well above the level of the base line fluctuations for the samples with low Li content (upon calcination at higher temperatures, this band develops and displays the fine structure characteristic of Ni2+ in a NiA1204 spinel environment). In addition to the bands at 420 and 620 nm, there is probably a broad and weak band at ca. 720 nm, which would correspond to octahedral Ni2+, as in NiO. wavelength/nm FIG. 4.-Diffuse reflectance spectra of NiAlLix. X . P . S . INTENSITY MEASUREMENTS Thevariation of X.p.s. intensity ratio INi(2p)/IA1(2,) against theoveralllithium content is plotted in fig. 5. We observe that as the lithium content of the carrier increases, the X.P.S. intensity ratio INi(2p)/IA1(28) first slightly increases and then steadily decreases. The same trend is observed for the factor &1Ni(&p)/1A1(28), which must reflect the dispersion of the supported phase [fig.5(b)], as shown above [eqn (1 b) and (2a)l.1396 EFFECT OF Li ON NiO/y-Al,O, 0 0.6 I I ( A ) I ' O O i 0 2 4 0 2 4 FIG. 5.+A) Variation of the X.P.S. intensity ratio ZNi(2p)/IA,(2s), as a function of Li,O content. (B) Variatior of the product SsZNj(2p)/ZA1(2s), as a function of Li,O content. Li,O (wt %) Li,O (wt %) 300 500 700 900 T/"C FIG. 6.-Temperature programmed reduction spectra of NiAlLix samples (arbitrary units).M. HOUALLA, J . LEMAITRE A N D B. DELMON 1397 TE M PER ATURE-PROGR AM MED REDUCTION The t.p.r.patterns of the NiAlLix samples are presented in fig. 6. One may distinguish two main reduction waves and T,. Their maxima lie at 600-650 O C (q) and 750-800 OC (Q. Increasing the lithium content of the support causes the progressive disappearance of the high-temperature band or shoulder T, and a concomitant emergence of a large, possibly complex, band at ca. 675 OC. On the other hand, the overall area of the reduction peaks is not altered by the presence of lithium in the carrier, as postulated in the Experimental section. DISCUSSION THE ROLE OF Li I N THE NiAlLix SERIES EFFECT OF Li ON THE CHEMICAL STATE OF THE Ni2+ SPECIES Depending on the degree of interaction between the active phase and the support, the nickel-containing species may be present as NiO or NiAl,O, (surface or bulk).The distinction between these compounds is in principle. readily provided by d.r.s. measurements. Ni2+ in NiO is exclusively in octahedral symmetry and characterized by an absorption band at 720 nm (fig. 4, spectrum of NiO), whereas Ni2+ in NiAl,O, is in both octahedral and tetrahedral environments and is characterized by a complex band at ca. 620 nm (fig. 4, spectrum of NiAl,O,). The d.r.s. spectra of the samples with low Li content have a band at 620 nm, thus indicating the presence of NiAl,O,. In spite of the weakness of this band, the comparison of the spectra of the various samples suggests that it decreases when the Li content increases, i.e. less NiAl,O, is present. The interpretation of the spectra at ca.720 nm is speculative, because of the extreme weakness of the observed effects. If the effect mentioned above, i.e. an enhancement of this band for higher Li content, really exists, it would be in agreement with the enhancement of the X.r.d. lines due to NiO, thus suggesting that the addition of Li promotes the formation of NiO. EFFECT OF Li ON THE DISPERSION OF THE NICKEL SPECIES We shall attempt to gain insight into the dispersion of the nickel-containing species from the X.P.S. data. More precisely, we shall compare these data with the theoretical results corresponding to the models presented in the Experimental section. Monolayer dispersion of nickel on AlLix. The model corresponding to a monolayer dispersion of nickel on AlLix also includes the case where Ni2+ species would be present as a surface nickel aluminate having an average thickness much lower than the escape depth for Ni(2pg) electrons.The predicted intensity ratio INi(2p$)/IA1(2,3) [eqn ( I a)] for various lithium contents, shown in fig. 7, is from five to nine times higher than the corresponding experimental values. Statistical distribution of nickel in alumina. This configuration would correspond to the situation when Ni is homogeneously distributed on the atomic scale throughout the carrier (e.g. formation of homogeneous bulk nickel aluminate or solid solution of Ni in A1,0,). The calculated intensity ratio [eqn (3)] INi@pf)/IA1(Ps) remains about three times higher than the observed values (fig. 7). NiO particles over AlLix. Assuming that all the nickel-containing phase is present as NiO, eqn (2a) enables us to determine the size of the NiO particles eventually present.The calculated values range from 5.5 nm for x = 0 to 8.5 nm for x = 3.4 wt% Li,O. They are in reasonable agreement with those deduced from X.r.1.b. analysis (e.g. 5.3-6.0 nm).1398 EFFECT OF Li O N NiO/y-Al,O, 0 2 4 Li,O (wt %) FIG. 7.-Predicted X.P.S. intensity ratios ZNi(2p)/ZA1(28) for various configurations of the nickel-containing phase over the Li-modified aluminas. (----) Monolayer model; (-) statistical distribution model; (-@-) experimental observations. ASSESSMENT OF THE STATE AND DISPERSION OF THE NICKEL-CONTAINING PHASE, AS DETERMINED FROM D.R.S., X.P.S. INTENSITY MEASUREMENTS AND T.P.R. RESULTS The X.P.S.intensity results discussed above suggests that neither surface nor bulk NiAl,O, formation can solely account for the observed X.P.S. intensity ratios INi(2p)/IA1(2s). The values of the latter must be explained by the presence of Ni in NiO aggregates; however, as d.r.s. data suggest that part of the Ni2+ species is present as nickel aluminate, we propose that the actual situation is intermediate, i.e. that the nickel-containing phases comprise both NiO and NiAl,O,. In line with this assumption, the decrease in the dispersion of Ni2+ species with increasing lithium content (x > 1) illustrated by the steadily declining values of the factor &INi(2p)/lA1(2~), reflects in the fraction of nickel species present as NiO-like species, at the expense of NiA120,, and an eventual growth of the NiO particle size. The slight increase of the factor SslNi(zp)/IA~(zS) for the initial lithium addition (0 < x < l), if it is not an artifact, may be tentatively explained by increased formation of surface nickel aluminate at the expense of the bulk compound (fig.7). The variatiom of the reactivity of NiAlLix samples as a function of Li content of the carrier corroborate the above results. Indeed, the decrease of the hardly reducible fraction (peak T,) of the nickel-containing phase (fig. 6) may simply be assigned, in agreement with the assumed role of the additive, to the progressively more difficult formation of a bulk-like nickel aluminate. This has been confirmed by the fact that a t.p.r. diagram of a bulk nickel aluminate exhibits a reduction wave which coincides with peak &.The first reduction wave (peak q) would then be attributable to the reduction of NiO and, eventually, some surface nickel aluminate. TENTATIVE INTERPRETATION OF THE MECHANISM FOR THE ACTION OF LITHIUM The action of modifiers on the state and dispersion of a deposited phase may essentially correspond to two types of mechanisms: (i) via an alteration of the pH ofM. HOUALLA, J. LEMAITRE A N D B. DELMON 1399 the impregnating medium, during the deposition of the active phase, thus promoting or inhibiting the adsorption, hydrolysis or precipitation of the cations; (ii) via variations of the physico-chemical properties of the carrier through solid-state reactions of the additive with the support, thus influencing the reactivity of the latter toward the deposited phase.As reported in the introduction, mechanism (i) has been proposed to be responsible for the action of sodium in the case of cobalt and nickel oxides supported on modified aluminas. It is believed that ‘ free’ Na-containing species present in Na-rich carriers would cause an alteration of the pH of the solution filling the pores and thus modify the nickel deposition process. The effect of these Na-containing species is to bring about an enhancement of the dispersion of the deposited phase. Our results clearly indicate that lithium, when present in large quantities on alumina, has the opposite effect, i.e. a substantial decrease in the dispersion of the deposited nickel phase (fig. 5). We are thus led to assume that lithium exerts its influence via mechanism (ii), i.e. by modifying the solid-state reactions of the additive with the support.This may be substantiated by the known high reactivity of lithium toward alumina, due to the easier diffusion of the small lithium cations into the alumina lattice, as compared with larger Na ions. The consequence would be, in the case of Li, the absence of ‘free’ alkali-metal-containing species, which are required for mechanism (i) to be operative. Indeed, a mixed spinel LiAl,O, formation has been reported to take place as a consequence of calcination at 600 OC of lithium nitrate21 or carbonate,, impregnated aluminas. At lower temperatures, the formation of a-LiAlO, has been reported.22 We have no direct evidence, from our X.r.d.measurements, establishing the presence of the mixed spinel LiAl,O, phase in AlLix specimens. This is not surprising, since y-Al,O, and LiAl,O, both have spinel structures and nearly identical lattice parameters. However, other indications, though not very conclusive, suggest the incorporation of lithium in the y-Al,O, lattice. In particular, the initial increase in the surface areas of AlLix samples (for x < 1) suggests a stabilization of the pseudo-spinel y-Al,O, induced, as proposed by Kordes,,, by filling the y-Al,O, vacancies to form eventually a mixed lithium aluminate type compound. In line with that, the subsequent enhanced crystallization of AlLix specimens with increasing lithium content may be ascribed, as suggested by Levy and Bauer,22 to the growth of crystallites of the mixed lithium aluminate phase.In conclusion, we may conceive that, after calcination of lithium nitrate impregnated aluminas, lithium species are essentially present as a solid solution of Li ions in alumina or a mixed lithium aluminate. Nickel impregnation on lithium-free or -deficient aluminas leads, after drying and calcination, to the diffusion of part of the Ni2+ ions into the y-Al,O, framework to form nickel aluminate NiAl,O,, whereas the remaining part is present as dispersed NiO. On lithium-rich aluminas, the increasing thickness of the surface lithium aluminate hinders the diffusion of Ni2+ species into the y-Al,O, lattice during calcination, thus inhibiting the formation of nickel aluminate and concomitantly increasing the fraction of nickel present as NiO.Note, however, that alternative explanations of the effect of lithium in line with mechanism (ii) cannot be excluded. For instance, one may conceive that the effect of lithium addition is simply to alter the pore distribution, in a way which would preclude deep penetration of the nickel solution into the inner porosity of lithium-modified aluminas. This would, in turn, minimize the interaction of the nickel-containing phase with the carrier, increase the extent of NiO formation and favour the obtention of larger NiO particles. Such a mechanism could equally take place in the case of larger alkali-metal cations, but only as a secondary phenomenon, as compared with theI400 EFFECT OF Li ON NiO/y-A1203 predominant effect due to the presence in that case of weakly interacting alkali- metal-containing species [mechanism (i)].The conclusion that lithium exerts its influence via mechanism (ii) is consistent with other observations. Lo Jacono et aZ.24 and Cimino et al.25 studied the influence of small additions to y-A1203 of cations (for example Ga3+) having a preference for tetrahedral sites. Their work concerned a system related to ours, namely Co/Al,O,. These authors proposed that these cations, by increasing the anion polarization towards the tetrahedral sites of the alumina lattice, enhanced the diffusion rate of Co2+ into the spinel, thus decreasing the formation of segregated c0304. This process would favour the formation of a surface spinel CoAl,04 with a higher fraction of cobalt in tetrahedral sites, as compared with gallium-free alumina.We thus speculate that addition to A1203 of cations having a preference for octahedral sites will bring about the opposite phenomenon (e.g. segregation of c0304 and the existence of a lesser fraction of Co2+ in tetrahedral sites). On the basis of these assumptions and taking into account that lithium reportedly23 occupies octahedral sites of the alumina pseudo-spinel lattice, one can explain, in the case of NiAlLix samples, the increased formation of NiO at high lithium content and the concomitant decrease of the hardly reducible fraction of the nickel-containing phase (presumably Ni2+ in tetrahedral coordination). We thank the ‘Services de la Programmation de la Politique Scientifique’ (Actions Concertees Interuniversitaires ‘Catalyse’) for a research fellowship (M.H.) and for equipment and functioning expenses. K. Tanabe, Solid Acid and Bases (Academic Press, New York, 1970). M. Lo Jacono, M. Schiavello and A. Cimino, J. Phys. Chem., 1971, 75, 1044. H. Lafitau, E. Nee1 and J. C. Clement, Preparation of Catalysts I, ed. B. Delmon, P. A. Jacobs and G. Poncelet (Elsevier, Amsterdam, 1976), p. 393. M. Houalla and B. Delmon, C.R. Acad. Sci., Sir. C, 1979, 289, 77. F. Delannay, M. Houalla, D. Pirotte and B. Delmon, Surf. Interface Anal., 1979, 1, 172. * M. Houalla and B. Delmon, C.R. Acad Sci., Sir. C, 1980, 290, 301. M. Houalla, F. Delannay and B. Delmon, J. Phys. Chem., in press. A. Lycourghiotis, C. Defosse, F. Delannay, J. Lemaitre and B. Delmon, J. Chem. Soc., Faraday Trans. I , 1980, 76, 1677. P. Reynen and H. Bastius, Powder Metall. Int., 1976, 8, 91. P. B. Wells and F. C. Tompkins (The Chemical Society, London, 1976), vol. 2, p. 61 1. lo P. J. Angevine, J. L. Vartuli and W. N. Delgass, Proc. VIth Int. Cong. Catal., ed. G. C. Bond, l1 C. Defosse, P. Canesson, P. G. Rouxhet and B. Delmon, J. Catal., 1978, 51, 269. l2 S. C. Fung, J. Catal., 1979, 58, 454. l3 F. P. J. M. Kerkhof and J. A. Moulijn, J. Phys. Chem., 1979, 83, 1612. l4 J. C. Helmer and N. H. Weichert, Appl. Phys. Lett., 1968, 13, 266. l5 J. H. Scofield, J. Electron Spectrosc. Relat. Phenom., 1976, 8, 129. l6 D. R. Penn, J. Electron Spectrosc. Relat. Phenom., 1976, 9, 29. l7 C. C. Chang, SurJ Sci., 1975, 48, 9. l9 S. D. Robertson, D. D. McNicol, J. M. de Baas and K. S. Kloet, J. Catal., 1975, 37, 424. 2o R. J. Cvetanovich and Y. Amenomiya, Adu. Catal., 1967, 17, 103. 22 W. H. J. Stork and G. T. Pott, J. Phys. Chem., 1974, 78, 2496. 24 M. Lo Jacono, M. Schiavello, V. H. J. de Beer and G. Minelli, J. Phys. Chem.. 1977, 81, 1583. 25 A. Cimino, M. Lo Jacono and M. Schiavello, J. Phys. Chem., 1975, 79, 243. A. Bame, Chem. Phys. Lett., 1973, 19, 109. R. M. Levy and D. J. Bauer, J. Catal., 1967, 9, 76. E. Kordes, 2. Kristallogr., 1935, 9, 193. (PAPER 1 /391)

ISSN:0300-9599    

DOI:10.1039/F19827801389

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

J . Chem. SOC., Faraday Trans. 1, 1982, 78, 1401-1404 Coloured and Colourless Charge-transfer Complexes of Small and Polymeric Quaternary Pyridinium Bromides BY ERNEST A. BOUCHER* AND CHRISTOPHER C . MOLLETT School of Molecular Sciences, University of Sussex, Brighton BNl 9QJ Received 18th March, 1981 The products of quaternization reactions involving alkyl bromides are, under certain conditions, coloured, e.g. 4-methyl-N-n-propylpyridinium bromide is red and poly(4-vinyl-N-n-butylpyridinium bromide) is green when prepared in benzene, say, at 333 K. Neither are coloured, however, when heated in sulpholane. These and other products have been shown to be coloured due to charge-transfer bands; the possible involvement of impurities or side reactions has been eliminated: important in establishing the reliability of kinetic measurements by the authors. Comparison has been made with several iodides under varying conditions: these are invariably colourless in the solid and pale yellow in ethanol.They lack the low-energy charge-transfer bands (colours), but show charge-transfer bands in the U.V. Coleman and FUOSS~ (who were interested in polymer quaternization) found colour changes from yellowish-green to dark green with 4-methylpyridine + n-butyl bromide in nitrobenzene (they concluded that the presence of the 4-methyl group was necessary for colour production). Kronick and FUOSS~ found that some reactions in propylene carbonate gave orange solutions deepening to brown. On the other hand, Coleman and FUOSS~ also found orange solutions with 4-methylpyridine + n-butyl bromide in nitromethane and in NN’-dimethylformamide (DMF), which they attributed to side reactions (side reactions are known,3 but it is not yet clear whether they contribute to the colour).Quite separately Kosower and c o ~ o r k e r s ~ - ~ have extensively studied quaternary pyridinium iodides, which in solution are slightly yellow. [Ref. (1) and (4) are in the same journal volume, but are not cross-referenced.] Hantzsch,s incidentally, was aware of coloured quaternary pyridinium salts. Ray9 and Ray and Makurjee’O investigated the electronic spectra of N-ethylpyridinium bromide in water and N-alkylpyridinium iodides in chloroform. Fosterll provides a suitable background to charge-transfer complexes, and Griffiths and Pugh12 have critically reviewed Kosower’s work.When studies13-15 of the kinetics and mechanism of the quaternization of poly(4- vinylpyridine) with alkyl and arylalkyl bromides commenced, coloured products were thought to arise for one of three possible reasons: (i) impurities in the reactants; (ii) side reactions with solvent or (iii) the products being charge-transfer complexes. Apart from the inhererit interest in the coloured products, it was important to investigate them because the occurrence of (i) or (ii) would undermine the kinetic studies. To determine the nature of the polymeric products, some small quaternary pyridine halides were investigated in the U.V. (bromides and iodides) and visible (bromides) regions. EXPERIMENTAL The preparation of quaternary pyridinium halides in several solvents, usually in the range 320-350 K, from purified chemicals under an atmosphere of nitrogen has been adequately 46 1401 F A R 11402 PYRIDINIUM CHARGE-TRANSFER COMPLEXES described in connection with kinetic studies.3* 13-15 Solutions and solid samples were handled and kept under water-free nitrogen. Electronic spectra were recorded on a Unicam SP800 spectrophotometer at 298 K.Other techniques mentioned were carried out according to standard procedures. A reflectance spectrum of compressed powdered 4-methyl-N-n-propylpyridinium bromide was run under dried nitrogen in a Beckmann mark IV Acta spectrometer. RESULTS AND DISCUSSION Table 1 gives the main features found for five N-pyridinium bromides. The coloured products were actually obtained at 333 K from benzene.Coloured species could be obtained at 333 K (arbitrarily chosen) by mixing the pure liquid reagents alone or in the solvents benzene (2.28), toluene (2.38), nitromethane (35.9), DMF (36.7), and propylene carbonate (65.6). Values of the relative permittivity at 298 K are given in TABLE 1 .-ULTRAVIOLET AND VISIBLE ABSORPTION PROPERTIES OF PYRIDINIUM BROMIDES Rrnaxlnm Emax compound (colour) ref. solvent (main peak) AmaX/nm cmaX 4-methyl-N-n-propyl- chloroform 256 7900 455 95.5 4-methyl-N-n-butyl- chloroform 256 8050 455 95.4 pyridinium bromide” ethanol 256 7800 446 85.6 (red) water 256 7800 44 1 83.8 pyridinium bromide water 256 8100 440 84.3 (red) 4-me th yl-N- benzyl- chloroform 256 6900 490 153 pyridinium bromide water 256 6900 46 1 136 (‘ma~ve’)~ 4-et hyl-N-n-butyl chloroform 228 6700 620 40.2 pyridinium bromide ethanol 229 6400 616 38.7 (green) poly(4-vinyl-N-n-butyl- ethanol 237 68 60 660 38.3 pyridinium bromide)c (green) a Colourless when prepared in water or benzene at 298 K (and ethanol or sulpholane at 333 K); similar properties to these in table for product prepared in toluene or propylene carbonate at 333 K.Similar properties when prepared in propylene carbonate, but no colour when prepared in nitromethane. All solutions were of concentration 0.5 mol dm-3. Similar colour when no solvent present during preparation. the brackets. Colours could not be produced in ethanol (24.3), but a coloured solid retained its colour when dissolved in this solvent. Note also that the products precipitate out of benzene as they are formed. Also, in contrast to ethanol, colours are produced in propylene carbonate, whereas the trend seems to be for the production of the coloured species in solvents of low relative permittivity.No colours have been found in sulpholane (44) under any circumstances. Table 1 indicates the need for a substituent in the 4-position (compared to the colourless N-ethylpyridinium bromide), and the absorption frequency in the visible shows some sensitivity to the nature of this substituent. With a 4-methyl substituentE. A. BOUCHER AND C. C. MOLLETT 1403 the product is red or ‘mauve’, and with a 4-ethyl substituent and the polymer species the products have an even lower transition energy, being green; i.e. A,,, (visible) increases in the order 4-methyl c 4-ethyl c poly(4-vinyl) for the N-n-butyl com- pounds.The colour is not so sensitive to the N substituent, but Amax is slightly greater for N-benzyl compared with N-n-butyl. Regarding the solvents in which the spectra were run, table 1 shows some differences between chloroform and water; ethanol was similar to water. The following are representative observations. When heated above its melting point (395.0 K) red 4-methyl-N-n-propylpyridinium bromide did not change colour, and the colourless sample remained colourless above its observed melting point (395.5 K). However, the colourless sample when dissolved in acetone and heated to 333 K for 4 h turned red; the red colour was not affeLced by 5 fractional recrystallizations from acetone.The red salts retained the colour (3-4 years to date). The green colour of 4-ethyl-N-n-butylpyridinium bromide disappeared over several days at room tem- perature, but could be regenerated by re-heating the sample in 4-ethylpyridine (acting as solvent). The green colour of poly(4-vinyl-N-n-butylpyridinium bromide) persisted for several months. The possible involvement of free halide was eliminated when it was found that the HSO; ion did not affect the yellow iodides or the coloured bromides. Proton n.m.r. and i.r. spectra, although of limited sensitivity, showed no distinction between coloured and colourless varieties and did not reveal any impurities. No free radicals could be detected in the solutions by e.s.r. There is no reason to believe that trapped solvent is present in the solid state (excellent agreement of individual and mixed melting points and no change under vacuum).Coloured solids did not lose any colour when heated under vacuum near their melting points. Vapour above warmed coloured solid did not show evidence of solvent in g.1.c. Colourless solids subjected to a nominal pressure of lo4 kg in a KBr press did not produce any colour. Unfortunately crystal structures have not been determined for any of the solids, and the possibility of small changes in position or separation distances cannot be ruled out as distinguishing coloured from colourless solids. As far as can be judged the solutions and solids are not air-sensitive. The A, values in table 1 for 4-methyl-N-n-propylpyridinium bromide range from 441 to 455 nm depending on the solvent.It was not easy to obtain reflectance spectra of the solid, but that for this compound showed a broad maximum with a definite shift in Amax to ca. 490 nm. This is taken as evidence that the absorption is modified by the presence of solvent, but that the same charge-transfer species exists in the solid and in solution. In connection with a related study (unpublished) of the quaternization of poly(4- vinylpyridine) with iodides, the u.v.-visible spectra were examined of the products in ethanol and water from 4-methylpyridine + ethyl iodide and pyridine +ethyl iodide. These salts are colourless in the solid but appear pale yellow in ethanol solution. No colours other than pale yellow could be produced under conditions where bromides would be coloured.Generally the absorption spectra were very similar to those found for charge-transfer iodides by Kosower et al.4-7 This study, apart from establishing that the coloured pyridinium bromides, including the polymer derivative, are genuine charge-transfer complexes, brings together what have been two separate areas of investigation. There is now no need to avoid solvents in which coloured products are produced for quantitative quaternization studies, unless side reactions also occur (e.g. with DMF present). There is no evidence favouring the participation of impurities in the colour formation or of anomalous species derived from the reactants. 46-21404 PYRIDINIUM CHARGE-TRANSFER COMPLEXES C.C.M. is grateful to the States of Guernsey Education Department for a studentship. We are very grateful to a referee for helpful comments. B. D. Coleman and R. M. Fuoss, J . Am. Chem. SOC., 1955, 77, 5472. P. L. Kronick and R. M. Fuoss, J. Am. Chem. SOC., 1955, 77, 61 14. E. A. Boucher, E. Khosravi-Babadi and C. C. Mollett, J. Chem. SOC., Faraday Trans. 1, 1979, 75, 1728. E. M. Kosower, J. Am. Chem. Soc., 1955, 77, 3883. E. M. Kosower and P. E. Klinedinst J . Am. Chem. SOC., 1956, 78, 3493. E. M. Kosower, Prog. Phys. Org. Chem., 1965, 3, 81. E. M. Kosower and J. C. Burbach, J . Am. Chem. SOC., 1956, 78, 5838. * A. Hantzsch, Berichte, 191 1, 44, 1783; 1919, 52, 1544. A. Ray, J. Am. Chem. SOC., 1971, 93, 7146. lo A. Ray and P. Makurjee, J. Phys. Chem., 1966, 70, 2138. R. Foster, Organic Charge- Transfer Complexes (Academic Press, London, 1969). l 2 T. R. Griffiths and D. C. Pugh, Coord. Chem. Rev., 1979, 29, 129. l3 E. A. Boucher and C. C. Mollett, J. Polym. Sci., Polym. Phys. Ed., 1977, 15, 283. l4 E. A. Boucher, J. A. Groves, C. C. Mollett and P. W. Fletcher, J. Chem. SOC., Faraday Trans. 1,1977, l5 E. A. Boucher and C. C. Mollett, J . Chem. SOC., Faraday Trans. I , 1982, 78, 75. 73, 1629. (PAPER 1 /444)

ISSN:0300-9599    

DOI:10.1039/F19827801401

出版商:RSC    

年代:1982

数据来源: RSC

摘要:

J . Chem. SOC., Furaduy Trans. I , 1981, 78, 1405-1422 Activity Coefficients for the System HC1+ CaC1, + H,O at Various Temperatures Applications of Pitzer’s Equations BY RABINDRA N. ROY,* JAMES J. GIBBONS, LISA K. OVENS, GREG A. BLISS AND JAMES J. HARTLEY Department of Chemistry, Drury College, Springfield, Missouri 65802, U.S.A. Received 24th March, 1981 Activity coefficients of hydrochloric acid in mixed solutions with calcium chloride have been measured at total ionic strengths of 0.1,0.25,0.5, 1 .O, 1.5 and 2.0 mol kg-’ at nine different temperatures from 278.15 to 318.15 K at intervals of 5 K, and at total ionic strengths of 2.5, 3.0, 3.5, 4.0, 4.5 and 5.0 mol kg-’ from 278.15 to 318.15 K at intervals of 10 K, using e.m.f. cells of the type Pt, H,(g, 1 atm) lHCl(m~), CaCl,(mg)J AgCl I Ag.The results have been interpreted in terms of Pitzer’s equations, from which the calculations of the activity coefficients of calcium chloride in the mixtures have been made. The values of and II/MNX (including higher-order electrostatic effects) and OM, and yMNX (excluding higher-order effects) were determined based on Pitzer’s treatment. The trace activity coefficients of HCl, y x (when y,, the ionic strength fraction of CaCl, in the mixture, equals l), are also reported. Within the experimental error, Harned’s rule is obeyed for hydrochloric acid at all experimental temperatures and over the entire range of ionic strengths studied. In the case of calcium chloride, Harned’s rule is a good description only for Z = 0.1,0.25 and 0.5 mol kg-’, over the range of temperatures examined.In recent years there has been considerable interest in describing the thermodynamic behaviour of aqueous electrolyte mixtures. In fields such as geothermal energy, geochemistry, desalination and chemical oceanography, it is essential to have thermodynamic information on aqueous mixed electrolyte solutions as functions of the temperature. The solute activity coefficients are of primary importance to a complete understanding of the chemistry of these acid-salt solutions. The e.m.f. method is quite suitable for the determination of the activity coefficients of the acid in aqueous mixed acid-salt solutions, since it directly gives the activity coefficient of one of the solutes in which the electrodes are reversible. Precise e.m.f.measurements have been recently made on other systems such as HCI + MgC1, + H20,1-4 HCl + CaC1, + H,O and HC1+ SrCI, + H20,6-9 with the latter two systems to a more limited extent. As a continuation of the past work along these lines by various investigators10-21 including the authors’ laboratory, we have now studied the system HC1+ CaC1, + H,O using e.m.f. measurements on cells without liquid junction of the type: Pt, H,(g, 1 atm) I HCW,), CaCl,(m,) I AgCl I Ag (A) at constant total ionic strength, I (where I = mA + 3m,) equal to 0.1, 0.25, 0.5, 1 .O, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5 and 5.0 mol kg-l from 278.15 to 318.15 K. The main objectives of this study were as follows: (a) to supplement the data of Khoo et aL5 at I = 0.25, 1.5 and 2.5 mol kg-l by providing e.m.f. values at these ionic strengths; (b) to extend the investigation to higher ionic strengths, such as I = 3.5,4.01412 ACTIVITY COEFFI c I EN TS FOR HCl + CaCl, + H 2 0 4.5 and 5.0 mol kg-l; (c) to extend the temperature range of the study from 278.15 to 318.15 K at 5 K intervals (up to I = 2.0 mol kg-l) and 10 K intervals (up to I = 5.0 mol kg-l); ( d ) to take into account the higher-order electrostatic effects for unsymmetrical mixing; (e) to test the applicability of Harned's rule to hydrochloric acid, as well as to calcium chloride in the mixture.EXPERIMENTAL E.m.f. measurements were made using a Leeds and Northrup K-5 potentiometer. The cell design, the preparation of the electrodes and the other experimental procedures were the same as those described el~ewhere.~? l5 Analytical reagent-grade hydrochloric acid was distilled twice to the azeotropic composition with retention of the middle fraction.It was then standardized gravimetrically by weighing as silver chloride. ACS certified reagent-grade calcium chloride was recrystallized from water and analysed gravimetrically for chloride. Solutions for all the runs were made by direct weighings of the appropriate stock solutions, and vacuum corrections were applied to all weighings. Solution compositions were well within 0.02% of the nominal values of the ionic strength, I. All e.m.f. readings were corrected to a hydrogen pressure of 1 atm (101.325 kPa) and are precise to within 0.05 mV from I = 0.1 to 2.0 mol kg-l, and to within 0.08 mV from I = 3.5 to 5.0 mol kg-'.There was no evidence of irreversible behaviour due to the possible dissolution of silver chloride from the silver, silver chloride electrodes at higher ionic strengths, even at I = 5.0 mol kg-l. The detailed explanations as to how the poisoning effect (or irreversible behaviour) of the hydrogen electrodes was avoided, have been described previously.21 The temperature fluctuations of the bath were within 0.02 K. RESULTS AND CALCULATIONS Table 1 summarizes the experimental results of the corrected e.m.f. for cell (A) at twelve different values of constant total ionic strengths from I = 0.1 to 5.0 mol kg-l, and at various concentrations of calcium chloride, represented by the ionic strength fraction of the salt as yR, where yB = 3m,/I. An analogous expression for the ionic strength fraction of the acid, y,, is yA = mA/I.The negative logarithm of the mean ionic activity coefficients of hydrochloric acid, listed in the third column of table 1, have been fitted to the equation where k = (RTln 10)/F and Eo is the standard e.m.f. of cell (A), given e1~ewhere.l~ HARNED'S RULE The Harned expressionz2 describes the variation of yA as a function of solution composition in a mixed electrolyte system at constant total ionic strength. It is convenient to express this relation as (2) where y l is the activity coefficient of hydrochloric acid alone at the same total ionic strength I of the mixture, and Q A and R A are called the Harned coefficients, which are constants independent of the composition at a given total ionic strength but functions of both temperature and the particular ionic strength.When R A = 0, eqn (2) is linear and the electrolyte is said to obey Harned's rule.zz An analogous expression for the salt (component B) in the mixture can be written log Y A = log YOA-QAYB-RAYL1414 ACTIVITY COEFFICIENTS FOR HCI+CaCl,+H,OThe values of Q A , RA and log 71, together with the standard deviation of log in eqn (2), are listed in table 2. Also, the values of log yB and yA (required to evaluate QB, RB and log y g , which are themselves presented in table 4), are given in table 3 for twelve different ionic strengths from I = 0.1 to 5.0 mol kg-l at 298.15 K. The equations and methods of obtaining the values of -log yB will now be discussed. Various recent approaches as outlined by Scatchard and C O W O T ~ ~ T S , ~ ~ ~ 24 Reilly et aLZ5R.N. ROY, J. J. GIBBONS, L. K. OVENS, G. A. BLISS, J. J. HARTLEY 1415 Rasaiah and Friedman26 and P i t ~ e r ~ ~ are usually applied. As in our earlier studies,4* 1 5 9 21$ 28 we will once again adopt the recent ion-component treatment of Pitzer2’ in order to obtain a quantitative measure of the binary interactions and triple-ion interactions of the types H+-Ca2+ and H+-Cl--Ca2+, respectively. TABLE 4.-cOEFFICIENTS FOR THE HARNED EQUATION [EQN (3)] FOR HC1+ CaC1, + H,O AT 298.15 K I/mol kg-’ -log y i QBa -R, 0. I 0.25 0.5 1 .o 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.210 71 0.010 79 (1) 0.210 56 0.009 79 (0) 0.272 49 0.027 13 (3) 0.271 92 0.023 91 (0) 0.316 67 0.054 83 (3) 0.314 58 0.046 99 (0) 0.344 14 0.092 47 (1 1, 0) 0.347 70 0.137 96 (15,O) 0.338 69 0.183 58 (19,O) 0.321 70 0.229 47 (21, 0) 0.298 97 0.275 65 (22, 0) 0.271 82 0.321 93 (25,O) 0.241 20 0.368 57 (27, 0) 0.207 72 0.415 40 (27,O) 0.171 86 0.462 45 (30,O) Ob Ob Ob -0.000 94 -0.002 82 -0.005 97 -0.01 1 52 -0.015 93 -0.019 49 -0.022 39 -0.024 81 -0.027 00 -0.028 87 - 0.030 61 -0.032 23 a Parenthesized entries are lo5 0 (log yB) without and with inclusion of R,, respectively. Set equal to zero.PITZER’S EQUATIONS Pitzer’s equations7? 2 7 9 29 for the activity coefficients of HCI (component A) in mixtures with the salt CaC1, (component B) are given by In YHCl = f ’+ (mH +mCl)(BHCl + mCl CHCI) + mCa(BCaCI + mC]CCa,] + ‘@HCa + E@HCa) + mH mCl(B’HCl + CHC1) + mCamC1(&!aC1 + !j vHCaC1) +m,mca(E@’,ca+i W H c a c J (4) ( 5 ) where in which A4 has the value of 0.392 1 1 for water at 298.15 K and b = 1.2 kgi mol-4.The effects associated with the mixing of ions of the same sign are largely incorporated into the mixing parameter, 0, which is divided into two parts by Pitzer7 as shown below: f)’ = - A4[Ii/( 1 + bli) + 2/b In (1 + blr)] Oij = Wij+ EOij ( 6 ) where i and j are the charges of the ions. For the HC1-CaC1,-H,O system, eqn (6) becomes (7) @HCa = ‘@HCa+ EOHCa* Similarly, we have1416 ACTIVITY COEFFICIENTS FOR HCl+CaCl,+H,O where EOHCa and Ee;fCa are the contributions from higher-order electrostatic effects of unsymmetrical mixing with the omission of short-range forces, and 'f?HCa and s8'Hca account primarily for the effects of short-range forces, as well as to some extent the effects due to the use of molalities instead of concentrations and certain other approximations in the Debye-Huckel term.The term 'Ohca arises due to the variation of WHCa with ionic strength I (specifically, because of the dependence of the second virial coefficient Aij on I ) . In accordance with the suggestions of Pitzer and Kim,30 we have made a satisfactory approximation by assuming to be independent of I (hence, V/HCaCl in eqn (4) is the parameter representing the effect of short-range forces between these three ions in a mixed electrolyte, 9hca represents a possible ionic strength dependence of OHCa and is usually neglected, since it is expected to be very small, the symbols E 8 ~ ~ a and E8'Hca contribute the higher-order electrostatic effects listed in earlier paper^,*^^^ and the single-electrolyte parameters B, B and C4 are easily obtainable from the following equations: was set equal to zero).and where a is assumed to be 2.0 kg' mol-4 and the other symbols have their usual physical significance as defined by Pitzer. The quantities $, p1 and C4 for use in eqn (9)-(12) are given by Pitzer and M a y ~ r g a , ~ at 298.15 K. These are as follows: $HCl = 0.1775 $CaC1, = 0.3159 pbcl = 0.2945 /3Facl, = 1.614 C&cl = 0.00080 CcaCI, = -0.00034. The values of B, B' and C4 for HCl and CaCl, are obtained using the above tabulations and are presented in table 5 . A fixed value of Cd normally applies up to quite high ionic strengths of an electrolyte. The expression for the activity coefficients of CaCl, in mixtures with HCl is given by log YB = log 7; + 0 .1 4 76 y,I[4 BHC1-Q BCaC1-k 2 OHCa BkCl-iBLaCl+2 eHCa+?! vHCaCl+t c!!%Cl)l +o.14476y~12[2B~,l-QB~aCl-Q ~ H C a C l + ~ c & C ~ - ~ ( 2 ) i c t i a ~ ~ - 2 e ~ ~ a 1 (13) where the equation relevant to the single-electrolyte solution is In 8; = - 2 A&{[I'/( 1 + 1.2 la)] + (2/ 1.2) In (1 + 1.2 I$)> +'j [t Bc,clI+ i 1 ~ ( 6 BLacl+ 3 (2)' C$aCJI (14) where the symbols have the same meanings as those in Pitzer's papers.R. N. ROY, J. J. GIBBONS, L. K. OVENS, G. A. BLISS, J. J. HARTLEY 1417 TABLE 5.-sINGLE ELECTROLYTE PARAMETERS FOR PITZER'S EQUATIONS AT 298.15 K HC1 (C4 = 0.000 80) CaCl, (Cd = -0.000 34) I/mol kg-2 B B B B' E8' 0.1 0.25 0.5 1 .o 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0.372 90 0.333 14 0.299 15 0.264 97 0.246 43 0.234 46 0.226 02 0.219 72 0.214 84 0.210 94 0.207 76 0.205 11 -0.389 37 -0.189 19 -0.100 10 - 0.047 61 -0.029 00 -0.019 78 -0.014 42 -0.011 00 -0.008 67 - 0.007 01 -0.005 78 -0.004 85 1.386 79 1.168 87 0.982 59 0.795 25 0.693 67 0.628 10 0.581 83 0.547 31 0.520 54 0.499 17 0.481 72 0.467 21 -2.133 93 - 1.036 85 - 0.548 59 -0.260 92 -0.158 95 -0.108 40 -0.079 04 -0.060 30 -0.047 53 -0.038 43 -0.031 70 -0.026 57 1.804 1 0.485 21 0.185 22 0.069 40 0.038 42 0.025 37 0.018 23 0.013 85 0.011 08 0.009 08 0.007 60 0.007 07 HIGHER-ORDER ELECTROSTATIC EFFECTS The most important characteristic features of the two terms EOHCa and EB'HCa from eqn (13) and (14) are that they are dependent on (a) ionic strength, (b) temperature and ( c ) charges of the ions i and j (where i = H and j = Ca in the present case).It has been shown in recent papers4?l5 that these special terms are important for 1-2 mixtures even though they are zero when the ions i and j are of the same charge. These significant effects for unsymmetrical mixing were first proposed by Friedman31 and the equations relating E 6 ~ ~ a and E8'Hca with the corresponding terms of the cluster-integral theory3' were developed by Pitzer :7 and E8HCa = - [E6HCa/zl + (2CazH/812) [xHCas(xHCa) -ixHHs(xHH) - t x c a c a J ' ( x ~ a ~ a ) l (16) where, for example, XHCa = 6zHzcaA41k The values of E&Ca, EB&a and the J terms over the constant total ionic stength range at 298.15 K are given elsewhere,15 with the values of E8'Hca listed again at the end of table 5 .The terms BHCa and vHCaCI (without the inclusion of "BIfCa and "elHca) and the corresponding terms and vHCaCl (with E6HCa and E8HCa) were evaluated by using Pitzer's equation' YHCl/%a = 6HCa+8(mH+mCl) YHCaCl- (17) Here Aln YHCl is the difference between the experimental value of In YHCl (table 1) and that calculated from eqn (4) with the inclusion of all pure electrolyte terms and assuming and t,vHCaCl are equal to zero; and, alternatively, with or without Differences between log y,(expt.) and log y,(calc.) from Harned's expression obtained when RA = 0 and when RA # 0 are shown graphically under the headings A1 and A2 in fig. 1. A plot shown in fig. 2 of the quantity on the left-hand side against the coefficient of WHCaCl from eqn (1 7) gives a straight line with the intercept S 6 ~ ~ a EBHCa and E6HCa.1418 40,- ACTIVITY COEFFICIENTS FOR HCl+CaCl,+H,O 6 0 4 A %.% 0 b t a 0 0 FIG. 1.-Plot of A, and A2 against y, (ionic strength fraction of the salt) at various ionic strengths of HCl+CaCl,+H,O at 298.15 K [A, = log y,(expt.)-log yA(calc.), using least-squares fit to eqn (2) with RA = 0; A, = log y,(expt.) - log y,(calc.), using least-squares fit to eqn (2) with RA # 01. Z = 0, e, 0.1 ; D , b, 1.0; A, A, 2.0; a, 4 , 3.0; 0, ., 4.0; 0, 4, 5.0 (open symbols, A,; closed symbols, A,). and the slope vHCaCI. The points on the graph are more weighted for larger values of the term f(mH +mc,). The values of BHCa, YHCaCl and st+?, YHCaCl at 298.15 K for this work, as well as other similar systems for quantitative comparisons, are summarized in table 6.Using these values of WHCa and YHCaCl thus obtained at 298.15 K, the activity coefficients of calcium chloride were computed from eqn (10) and are shown in table 3. TRACE ACTIVITY COEFFICIENTS The trace activity coefficient, yg, can be written as where y i is the activity coefficient of HC1 in its pure aqueous solution (in the limit yA = 1) with no other added electrolyte and yg is the corresponding trace activity coefficient of HCI in a salt solution when yB = 1. The values of 72, obtained from eqn (18) using the values of QA (linear) from table 2, are shown in table 7 at five temperatures and twelve different ionic strengths. For comparison, our value at 298.15 K for I = 1.0 mol kg-l equals 0.720, which is in satisfactory agreement with yg (0.715) obtained by Khoo et aL5 It is also of interest to compare the value of Q A (0.1 188 at 298.15 K, from table 2) for the present work at I = 2.0 mol kg-l with that (0.0916) found for HCl + MgC1, + H,04 at the same ionic strength.These data clearly suggest that the presence of CaCI, lowers the activity coefficient of HCI more than does MgCI,. DISCUSSION Comparison of the A1 and A2 values in fig. I and the values in parentheses of a(1og yA) in table 2 show that, within the experimental error, Harned's rule is obeyed by hydrochloric acid over the range of ionic strengths and temperatures examined. The data in table 3 were substituted into eqn (3), and these results are listed in table 4, indicating that Harned's rule holds for calcium chloride only at I = 0.1, 0.25 and 0.5 mol kg-l, but additional quadratic terms in eqn (3) are definitely warranted aboveR. N.ROY, J . J. GIBBONS, L. K . OVENS, G. A. BLISS, J. J. HARTLEY 1419 -0.150 0.075 0 -0.075 -0.225 lo 0 0 0 400 0 0 0 cbo 0 0 0 . o 0 1 I I I I 0 1 2 3 4 qm,+ + rnc.-) FIG. 2.-Plot of Aln yA/rnB against !&nH++m,,-) from eqn (17) for HCl+CaCl,+H,O at 298.15 K. (Triangles represent values calculated with EB and EB' terms, whereas the circles refer to values resulting from the exclusion of these terms.) 0.5 mol kg-l, suggesting thereby the presence of appreciable ternary interactions for these ionic strengths. It is of interest to compare our values of Q A (equal to 0.1574) at 298.15 K for I = 3.0 mol kg-l (from table 2) with that given in parenthesis (QA = 0.1618,5 also in table 2).The agreement is satisfactory, even at higher ionic strengths. Within experimental error, the values of QA decrease with an increase in temperature at all experimental ionic strengths. This trend is in accord with that obtained by White et al. for the HCl+ MgCl, + H 2 0 This observation led to the ~uggestion~~ that an analogue of Harned's rule exists. That is, LA, the difference in relative partial molal enthalpy, is a linear function of y,. It has been shown in fig. 2 and table 6 that the inclusion of higher-order electrostatic terms E8 and EB', represented by triangles in fig. 2, is important in order to obtain satisfactory values of 8 and v/ for 1-2 type electrolytes. Note that not all the points determined are shown in fig.2, resulting in the observation that some of the points on the lower plot do not have corresponding values on the upper one. This is due to a selection of data points at random for the final representative graph, to avoid cluttering the figure with all the available results. It was suggested by Pitzer, who reviewed a paper by Robinson et a1.33 that there should be a single value of but different values of t,v for binary mixtures of electrolytes containing different anions (for example, HCl+ CaCl, + H,O and HBr + CaBr, + H,O). As can be seen from table 6, the validity of this suggestion and1420 A C T I V I T Y COEFFICIENTS FOR HCI+CaCl,+H,O TABLE 6.-PITZER'S MIXING PARAMETERS FOR THE SYTEMS HCI + NC1R. N. ROY, J. J. GIBBONS, L.K. OVENS, G. A. BLISS, J. J. HARTLEY 1421 the success of the Pitzer treatment is once more clearly demonstrated. For instance, the value at 298.15 K of = 0.0612 and yHCaCl = 0.008 in the present study can be compared with WHCa = 0.0600 and vHCaBr = 0.0163 obtained by Khoo et al.34 for the HBr +CaBr,+ H,O system. Similarly, it can be also seen from table 6 that other values of '8 are in good agreement: 'O,,,(HBr + MgBr, + H,O) = 0.069, cy = - 0.005, '8HMg(HCI + MgC1, + H,O) = 0.062, y = 0.001 ; '8HBa(HBr + BaBr, + H,O, I = 2.0 mol kg-l) = 0.072, cy = 0.0057, 'OHB,(HCI+ BaCI,+H,O, I = 4.0 mol kg-l) = 0.074, cy = 0.01 37; and finally, s8HCa(HCl + CaC1, + H,O, I = 5.0 mol kg-l) = 0.0612, cy = 0.0008, SOHca(HBr+CaBrz+H,O, I = 2.0 mol kg-l) = 0.0600, cy = 0.0163.I shown above denotes maximum ionic strength studied. It is also interesting to point out the value obtained by Khoo et aL5 for OHCa and YHCaCl (equal to -0.0628 and 0.0467, respectively). The reason that these values do not agree with our results mentioned above is due to the fact that the additional electrostatic terms were not included in their determination. In a later paper by Khoo et a1.l these terms were included and they reported the values of '8HCa and VHCaCl at 298.15 K for HC1+ CaCl, + H,O (equal to 0.0739 and 0.0030, respectively). Even these data are not identical with our results (table 6), for the following reasons: (1) Khoo et al. used a non-linear regression for the data in eqn (4), whereas this work used a linear plot of eqn (17), as recommended by Pitzer; (2) Khoo et al.examined ionic strengths only up to 3.0 mol kg-l, with this study covering up to 5.0 mol kg-l, including seven more ionic strengths. More work is under progress for other systems, such as HCI+SrCl,+H,O and KC1 + KHCO, + H,O at I = 0.01 to 5.0 mol kg-l, to investigate Friedman's31 theories at low ionic strengths, as well as Pitzer's equations at higher values of I. Acknowledgment is made to the donors of the Petroleum Research Fund, admin- istered by the American Chemical Society, for partial support of this research, under PRF no. 10775-B5-C. This research was also supported in part by the National Institutes of Health, grant no. NIH 1 ROl GM 26809-02 BMT. We also thank F. R. Burns and M. Visnaw for some of the experimental work and preliminary data, and Mr L.N. Roy for providing technical assistance. K. H. Khoo, T. K. Lim and C. Y. Chan, J . Chem. SOC., Faraday Trans. I , 1978, 74, 2037. H. S. Harned and R. Gary, J . Am. Chem. SOC., 1955, 77, 1994. K. H. Khoo, C. Y. Chan and T. K. Lim, J . Solution Chem., 1977, 6, 855. R. N. Roy, J. J. Gibbons, D. P. Bliss Jr, R. G. Casebolt and B. K. Baker Jr, J . Solution Chem., 1980, 9, 911. K. H. Khoo, C. Y. Chan and T. K. Lim, J . Solution Chem., 1977, 6, 651. H. S. Harned and T. R. Paxton, J. Phys. Chem., 1953, 57, 531. K. S. Pitzer, J . Solution Chem., 1975, 4, 249. H. S. Harned and R. Gary, J . Am. Chem. SOC., 1955, 77, 4695. C. J. Downes, J . Phys. Chem., 1970, 74, 2153. l o R. N. Roy, J. J. Gibbons et al., unpublished results. l 1 H. S. Harned and R.Gary, J. Am. Chem. SOC., 1954, 76, 5924. l 3 K. H. Khoo, C. Y. Chan and T. K. Lim, J . Chem. Soc., Faraday Trans. I , 1978, 74, 837. l4 C. J. Downes, J . Chem. SOC., Faraday Trans. 1, 1972, 68, 1964. l 5 R. N. Roy, J . J. Gibbons, J. K. Trower and G. A. Lee, J. Solution Chem., 1980, 6, 535. l6 K. H. Khoo, T. K. Lim and C. Y. Chan, J . Solution Chem., 1977, 6, 291. l 7 R. N. Roy, J. J. Gibbons et al., unpublished results. R. N. Roy, J. J. Gibbons et al., unpublished results. R. N. Roy, J. J. Gibbons et al., unpublished results. H. S. Harned and C. G. Geary, J . Am. Chem. SOC., 1937, 59, 2032. 2o K. H. Khoo, T. K. Lim and C. Y. Chan, J . Solution Chem., 1979, 8, 277. 21 R. N. Roy and E. E. Swensson, J . Solution Chem., 1975, 5, 431.1422 A C TI V I TY COE FF I C I E N TS FOR HCl + CaCI, + H,O 22 H. S. Harned and B. B. Owen, The Physical Chemistry of Electroiytic Solutions (Reinhold, New York, 23 G. Scatchard, R. M. Rush and J. S. Johnson, J. Phys. Chem., 1970, 74, 3786. 24 R. M. Rush and J. S. Johnson, J. Phys. Chem., 1968, 72, 767. 25 P. J. Reilly, R. H. Wood and R. A. Robinson, J. Phys. Chem., 1971, 75, 1305. 26 J. C. Rasaiah and H. L. Friedman, J. Chem. Phys., 1968, 48, 2742; 1969, 50, 3965. 27 K. S. Pitzer, J. Phys. Chem., 1973, 77, 268. 28 R. N. Roy, J. J. Gibbons, C. Krueger and T. White, J. Chem. Soc., Faraday Trans. 1, 1976,72,2197. 2p K. S. Pitzer and G. Mayorga, J. Phys. Chem., 1973, 77, 2300. 30 K. S. Pitzer and J. J. Kim, J. Am. Chem. Soc., 1974, 96, 5701. 31 H. L. Friedman, Ionic Solution Theory (Interscience, New York, 1962). 32 D. R. White Jr, R. A. Robinson and R. G. Bates, J. Solution Chem., 1980, 9, 457. 33 R. A. Robinson, R. N. Roy and R. G. Bates, J. Solution Chem., 1974, 3, 837. 34 K. H. Khoo, T. K. Lim and C. Y. Chan, J. Chem. SOC., Faraday Trans. I , 1979, 75, 1067. 3rd edn, 1958), chap. 14. (PAPER 1 /479)

ISSN:0300-9599    

DOI:10.1039/F19827801405

出版商:RSC    

年代:1982

数据来源: RSC