摘要:

ISSN 0300-9599 JCFTAR 82 ( I 2 ) 3525-371 9 (1 986) JOURNAL OF THE CHEMICAL SOCIETY Faraday Transactions I Physical Chemistry in Condensed Phases 3525 3535 3553 3561 3569 3587 360 1 361 1 3625 3635 3647 3657 3667 368 1 3697 3709 3717 CONTENTS Fourier Transform Infrared Studies of the Irreversible Oxidation of Cyanide at Platinum Electrodes A. S. Hinman, R. A. Kydd and R. P. Cooney The Dielectric Properties of Zeolites in Variable Temperature and Humidity A. R. Haidar and A. K. Jonscher The Time-domain Response of Humid Zeolites A. K. Jonscher and A. R. Haidar Molecular-orbital Studies of C-H Bond Scission induced by Ionizing Radia- tion T. Tada Zeolites treated with Silicon Tetrachloride Vapour. Part 2.-Sorption Studies M. W. Anderson and J. Klinowski Studies of Propene Oxidation over Mixed Uranium-Antimony Oxides F.J. Farrell, T. G. Nevell and D. J. Hucknall Adsorption and Desorption Kinetics of Oxygen on Tin-Antimony Oxide Cat a1 y st by Quasi -cons tan t Coverage Met hods M-C. Bacchus-Montabonel and CO Adsorption at 77 K on KCl Films. An Infrared Investigation D. Scarano and A. Zecchina The Utilization of Time-resolved Dielectric Loss to probe the Role of the Surface in Heterogeneous Photochemistry C. J. Dobbin, A. R. McIntosh, J. R. Bolton, Z. D. Popovic and J. R. Harbour Polysilicate Equilibria in Concentrated Sodium Silicate Solutions I. L. Svensson, S. Sjoberg and L-0. Ohman Potentials of Ion-exchanged Synthetic Zeolite-Polymer Membranes M. Demertzis and N. P. Evmiridis Adsorption and Reduction of Nitrogen Monoxide by Potassium-doped Carbon T.Okuhara and K. Tanaka Physicochemical Properties and Isomerization Activity of Chlorinated Pt/Al,O, Catalysts A. Melchor, E. Garbowski, M-V. Mathieu and M. Primet Excess Pressures for Aqueous Solutions M. J. Blandamer, J. Burgess and A. W. Hakin Effect of Temperature on the Point of Zero Charge and Surface Dissociation Constants of Aqueous Suspensions of y-Al,O, K. Ch. Akratopulu, L. Vordonis and A. Lycourghiotis Thermal Decomposition of Solid Sodium Bicarbonate M. C. Ball, C. M. Snelling, A. N. Strachan and R. M. Strachan Reviews of Books J-P. Joly I I7ISSN 0300-9599 JCFTAR 82 ( I 2 ) 3525-371 9 (1 986) JOURNAL OF THE CHEMICAL SOCIETY Faraday Transactions I Physical Chemistry in Condensed Phases 3525 3535 3553 3561 3569 3587 360 1 361 1 3625 3635 3647 3657 3667 368 1 3697 3709 3717 CONTENTS Fourier Transform Infrared Studies of the Irreversible Oxidation of Cyanide at Platinum Electrodes A.S. Hinman, R. A. Kydd and R. P. Cooney The Dielectric Properties of Zeolites in Variable Temperature and Humidity A. R. Haidar and A. K. Jonscher The Time-domain Response of Humid Zeolites A. K. Jonscher and A. R. Haidar Molecular-orbital Studies of C-H Bond Scission induced by Ionizing Radia- tion T. Tada Zeolites treated with Silicon Tetrachloride Vapour. Part 2.-Sorption Studies M. W. Anderson and J. Klinowski Studies of Propene Oxidation over Mixed Uranium-Antimony Oxides F. J. Farrell, T. G. Nevell and D. J. Hucknall Adsorption and Desorption Kinetics of Oxygen on Tin-Antimony Oxide Cat a1 y st by Quasi -cons tan t Coverage Met hods M-C.Bacchus-Montabonel and CO Adsorption at 77 K on KCl Films. An Infrared Investigation D. Scarano and A. Zecchina The Utilization of Time-resolved Dielectric Loss to probe the Role of the Surface in Heterogeneous Photochemistry C. J. Dobbin, A. R. McIntosh, J. R. Bolton, Z. D. Popovic and J. R. Harbour Polysilicate Equilibria in Concentrated Sodium Silicate Solutions I. L. Svensson, S. Sjoberg and L-0. Ohman Potentials of Ion-exchanged Synthetic Zeolite-Polymer Membranes M. Demertzis and N. P. Evmiridis Adsorption and Reduction of Nitrogen Monoxide by Potassium-doped Carbon T. Okuhara and K. Tanaka Physicochemical Properties and Isomerization Activity of Chlorinated Pt/Al,O, Catalysts A. Melchor, E. Garbowski, M-V. Mathieu and M. Primet Excess Pressures for Aqueous Solutions M. J. Blandamer, J. Burgess and A. W. Hakin Effect of Temperature on the Point of Zero Charge and Surface Dissociation Constants of Aqueous Suspensions of y-Al,O, K. Ch. Akratopulu, L. Vordonis and A. Lycourghiotis Thermal Decomposition of Solid Sodium Bicarbonate M. C. Ball, C. M. Snelling, A. N. Strachan and R. M. Strachan Reviews of Books J-P. Joly I I7

ISSN:0300-9599    

DOI:10.1039/F198682FX035

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

FA RA DA Y 1RA N SA CTlO N S AND SYMPOSIA From the Royal Society of Chemistry FARADAY TRANSACTIONS II Molecular and Chemical Physics SPECIAL ISSUE - AUGUST 1986 Professor Alan Carrlngton delbred the 1985 Faraday lecture at the Royal lnsmutlon on 10th December, 1985. As a compliment to Professor Carrington, a group of his colleagues and Mends submitted original papers on the general theme of Molecular Dynamics and Spectroscopy. These papers are collected in the present Issue. CONTENTs: The Faraday Lecture: Spectroscopy of Molecular Ions at thelr Dissociation Limits A. Carrington The Spectroscopy, Photophysics and Photochemistry of Clusters of Metrazlne D. H. Levy Spectroscopy of Transient Species produced by Photodissociation or Photoionizatlon in a Supersonlc Free-jet Expansion T.A. Mlller Molecukr-beam Infrared Spectroscopy of the Ar-N,O van der Waals Molecule J. Hodge, G. D. Hayman, T. R. Dykeand B. J. Howard The Estimation of Vlbrational Predissociation Ufetimes M. S. Chlld The Infrared Spectrum of H; and its Isotopomers. A Challenge to Theory and Experlment J. Tennyson and B. T. Sutcliffe The Augmented Secular Equation Method for calculating Spectra of van der Waals Complexes. Application to the Infrared Spechum of Ar-HCI J. M. Hutson Quantum-mechanical Wavepacket Dynamics of the CH Group in Symmetric- top X,CH Compounds using Effecttve Hamiltonians from Hlgh-resolution Spectroscopy R. Marguardt, M. Quack, J. Stohner and f. Sutcllffe Internal Dynamics of Subunits and Bondlng Force Constants In Weakly Bound Dlmers P. Cope, D.J. Mlllerand A. C. fegon Internal Dynamics and HF Bond Lengthenlng In the Hydrogen-bonded Heterodimer CH,CN . . . HF determined from Nuclear Hyper-fine Structure in its Rotational Spectrum P. Cope, D. J. Miller, L. C. Willoughbyand A. C. fegon Pumping and Rshing. Double-resonance Measurements on Molecular Jets U. Veeken, N. Damand J. Reuss nme-resow Fluorescence of Jet-cooled Carbazoles and thelr Weak Complexes A. R. Auty, A. C. Jones and D. Philllps Prediction of the Ct(*P, J/CI(P,J Branching Ratio in the Photodissociation of HCI S. C. GIvertzand 6. C. Ballnt-Kurt/ Hlgh-resolution Laser Photofragment Spectroscopy of CH' P. J. Sam, J. M. Walmsleyand C. J. Whitham Asymmetric Uneshapes associated with Predissociating levels M. N. R. Ashfold, R. N. Dixon, J. D. Prince, B.Tutcherand C. M. Westem A Threshold-photoelectron Ruorescence-Photon Coincidence Study of Radlatlonless TransMons in the B ,iI State of BCN+ €. Castellucci, G. Dulardin and S. leach Cornpetthe Channels in the Interaction of Xe('P,) with CI,, Br, and I,. Atom Transfer, Excitation Transfer, Energy Disposal and Product Alignment K. Johnson, R. Pease, J. P. Simons, P. A. Smith and A. Kvaran Mco Non-RSC Momkn P14.30 ($27.70) RSC Momkn M.00 Payment should accompany ordon for lhlr Ihm. RSC Members should send their orders to: Membershlp Manager, The Royal Society of Chemistry, 30 Russell Square, London WC1 B SDT. Non-RSC Members should send their orders to: The Royal Socleiy of Chemistry, Distribution Centre, Blackhorse Road. Letchwotth, Herts SG6 1 HN. Faraday Discussions No.80 Physical Interactions and Energy Exchange af the Gas-Solid Interface [his publication discusses aspects of current research on the gas-solid Interface: elastic, inelastic and dlssipattve scattering of atoms and molecules from cfystal surfaces; the structure and dynamics of physisorbed species, including overlayers. Emphasis Is placed on the themes of physical interactions and energy exchange rather than on molecular beam technology or the phenomenology of phase tmnsmons in overlayen. The interplay between theory and experlment is stressed as they relate to the nature of atom and molecule-surface interaction potentials including many body effects. Faraday Discussions No. 80 (1986) Softcevor Prico 531 .OO ($60.00) RSC Mombon J66.25 ROYAL SOCIETY OF CHEMISTRY Information Services (xiii)FA RA DA Y 1RA N SA CTlO N S AND SYMPOSIA From the Royal Society of Chemistry FARADAY TRANSACTIONS II Molecular and Chemical Physics SPECIAL ISSUE - AUGUST 1986 Professor Alan Carrlngton delbred the 1985 Faraday lecture at the Royal lnsmutlon on 10th December, 1985.As a compliment to Professor Carrington, a group of his colleagues and Mends submitted original papers on the general theme of Molecular Dynamics and Spectroscopy. These papers are collected in the present Issue. CONTENTs: The Faraday Lecture: Spectroscopy of Molecular Ions at thelr Dissociation Limits A. Carrington The Spectroscopy, Photophysics and Photochemistry of Clusters of Metrazlne D. H. Levy Spectroscopy of Transient Species produced by Photodissociation or Photoionizatlon in a Supersonlc Free-jet Expansion T.A. Mlller Molecukr-beam Infrared Spectroscopy of the Ar-N,O van der Waals Molecule J. Hodge, G. D. Hayman, T. R. Dykeand B. J. Howard The Estimation of Vlbrational Predissociation Ufetimes M. S. Chlld The Infrared Spectrum of H; and its Isotopomers. A Challenge to Theory and Experlment J. Tennyson and B. T. Sutcliffe The Augmented Secular Equation Method for calculating Spectra of van der Waals Complexes. Application to the Infrared Spechum of Ar-HCI J. M. Hutson Quantum-mechanical Wavepacket Dynamics of the CH Group in Symmetric- top X,CH Compounds using Effecttve Hamiltonians from Hlgh-resolution Spectroscopy R. Marguardt, M. Quack, J. Stohner and f. Sutcllffe Internal Dynamics of Subunits and Bondlng Force Constants In Weakly Bound Dlmers P.Cope, D. J. Mlllerand A. C. fegon Internal Dynamics and HF Bond Lengthenlng In the Hydrogen-bonded Heterodimer CH,CN . . . HF determined from Nuclear Hyper-fine Structure in its Rotational Spectrum P. Cope, D. J. Miller, L. C. Willoughbyand A. C. fegon Pumping and Rshing. Double-resonance Measurements on Molecular Jets U. Veeken, N. Damand J. Reuss nme-resow Fluorescence of Jet-cooled Carbazoles and thelr Weak Complexes A. R. Auty, A. C. Jones and D. Philllps Prediction of the Ct(*P, J/CI(P,J Branching Ratio in the Photodissociation of HCI S. C. GIvertzand 6. C. Ballnt-Kurt/ Hlgh-resolution Laser Photofragment Spectroscopy of CH' P. J. Sam, J. M. Walmsleyand C. J. Whitham Asymmetric Uneshapes associated with Predissociating levels M.N. R. Ashfold, R. N. Dixon, J. D. Prince, B. Tutcherand C. M. Westem A Threshold-photoelectron Ruorescence-Photon Coincidence Study of Radlatlonless TransMons in the B ,iI State of BCN+ €. Castellucci, G. Dulardin and S. leach Cornpetthe Channels in the Interaction of Xe('P,) with CI,, Br, and I,. Atom Transfer, Excitation Transfer, Energy Disposal and Product Alignment K. Johnson, R. Pease, J. P. Simons, P. A. Smith and A. Kvaran Mco Non-RSC Momkn P14.30 ($27.70) RSC Momkn M.00 Payment should accompany ordon for lhlr Ihm. RSC Members should send their orders to: Membershlp Manager, The Royal Society of Chemistry, 30 Russell Square, London WC1 B SDT. Non-RSC Members should send their orders to: The Royal Socleiy of Chemistry, Distribution Centre, Blackhorse Road. Letchwotth, Herts SG6 1 HN. Faraday Discussions No. 80 Physical Interactions and Energy Exchange af the Gas-Solid Interface [his publication discusses aspects of current research on the gas-solid Interface: elastic, inelastic and dlssipattve scattering of atoms and molecules from cfystal surfaces; the structure and dynamics of physisorbed species, including overlayers. Emphasis Is placed on the themes of physical interactions and energy exchange rather than on molecular beam technology or the phenomenology of phase tmnsmons in overlayen. The interplay between theory and experlment is stressed as they relate to the nature of atom and molecule-surface interaction potentials including many body effects. Faraday Discussions No. 80 (1986) Softcevor Prico 531 .OO ($60.00) RSC Mombon J66.25 ROYAL SOCIETY OF CHEMISTRY Information Services (xiii)

ISSN:0300-9599    

DOI:10.1039/F198682BX037

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ISSN 0300-9599 JCFTAR 82(10) 3053-3287 3053 3069 308 1 3097 3113 3 125 3141 3149 3163 3175 3185 3 197 3205 3215 3233 3245 J O U R N A L OF THE CHEMICAL SOCIETY Faraday Transactions I Physical Chemistry in Condensed Phases CONTENTS 13C Nuclear Magnetic Resonance Investigations of Carbon Monoxide in Decationated Zeolites of Type Y A. Michael, W. Meiler, D. Michel, H. Pfeifer, D. Hoppach and J. Delmau Thermal Properties, Thermochemistry and Kinetics of the Thermal Dissociation of Hydrochlorides of some Mononitrogen Aromatic Bases J. kubkowski and J. Bladejowski Structural Aspects of Metal-oxide-pillared Sheet Silicates. An Investigation by Magic-angle-spinning Nuclear Magnetic Resonance, Fourier-transform Infrared Spectroscopy and Powder X-Ray Diffractometry D. Tilak, B. Tennakoon, W.Jones and J. M. Thomas A Nuclear Magnetic Resonance Study of the Solvatochromism of a Pyridinium Betaine J. G. Dawber and R. A. Williams Reactions of Hydrocarbons on Alumina-supported Pt-Ir Bimetallic Catalysts. Part 1 .-Exchange of Methane and Cyclopentane with Deuterium, and Hydro- genolysis of Butane, Pentane and Cyclopentane D. Garden, C. Kemball and D. A. Whan Reactions of Hydrocarbons on Alumina-supported Pt-Ir Bimetallic Catalysts. Part 2.-Exchange of Benzene with Deuterium, and Exchange and Hydrogen- olysis of 2,2-Dimethylpropane A. C. Far0 Jr and C. Kemball Effects of Electrostatic Repulsion on the Aggregation of Azo Dyes in Aqueous Solution K. Hamada, S. Take, T. Iijima and S. Amiya Phosphorus and Proton Nuclear Magnetic Resonance Studies of Transition- metal Complexes of Triphosphate and Pyrophosphate in Aqueous Solution 0.Laurie, J. Oakes, J. W. Rockliffe and E. G. Smith Ordered Solution Structure of a Monodispersed Polystyrene Latex as studied by the Reflection Spectrum Method T. Okubo Transmitted Light Spectrum Measurements. A New and Convenient Technique for the Study of the Ordered Structure of a Monodispersed Polystyrene Latex in Solution and Film T. Okubo Ordered Solution Structure of a Monodispersed Polystyrene Latex as studied by the Transmitted Light Spectrum Method T. Okubo Adsorption and Decomposition of Ethylene and Acetylene on Flat Ru(OO1) and Stepped Ru(l,l,lO) Surfaces Solubilisation of Two-component Oil Mixtures by Micellar Surfactant Solu- tions B. J. Carroll Decay of High-valent Manganese Porphyrins in Aqueous Solution and Cata- lysed Formation of Oxygen A. Harriman, P. A. Christensen, G . Porter, K. Morehouse, P. Neta and M-C. Richoux A Non-equilibrium Configuration Theory of Polyelectrolyte Adsorption W. Barford, R. C. Ball and C. M. M. Nex Kinetics of Metal Oxide Dissolution. Oxide Dissolution of Chromium(II1) Oxide by Potassium Permanganate M. G. Segal and W. J. Williams C. Egawa, S. Naito and Kenzi Tamaru 101 F A R 1Con tents 3255 The Thermodynamics of Solvation of Ions. Part 1.-The Heat Capacity of Hydration at 298.15 K M. H. Abraham and Y. Marcus 3275 Thermodynamic Parameters of Electrolyte Solutions in Nitromethane A. F. Danil de Namor and L. Ghousseini 3287 Corrigendum to ' The Interaction of Aromatic Compounds with Poly(viny1- pyrrolidone) in Aqueous Solution. Part 6.-Polymer Precipitation and Viscosity Studies with Phenols and 0-Substituted Phenols' P. Molyneux and S. Vekavakayanondha

ISSN:0300-9599    

DOI:10.1039/F198682FP127

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

1569 1605 1621 1635 1647 1657 1669 1685 1701 1721 1739 1753 1763 1789 1801 1817 1871 1871 JOURNAL OF THE CHEMICAL SOCIETY Faraday Transactions 11, Issue 10,1986 Molecular and Chemical Physics For the benefit of readers of Faraday Transactions Z, the contents list of Faraday Transactions ZZ, Issue 10, is reproduced below. This issue contains the proceedings of Faraday Symposium 20 on Phase Transitions in Adsorbed Layers, including the Lennard-Jones Lecture by Professor J. S. Rowlinson. General Introduction : Interface Wandering in Adsorbed and Bulk Phases, Pure and Impure M. E. Fisher Motion in Surface Layers on N, on Graphite D. J. Tildesley and R. M. Lynden-Bell Long-period Commensurate Structures near the Incommensurate-Commen- surate Phase Transition in Xenon Layers absorbed on Graphite observed by X-Ray Diffraction Multilayers of Methane adsorbed on Graphite A.Inaba, Y. Koga and J. A. Morrison C. W. Mowforth, T. Rayment and R. K. Thomas Growth Mode of Multilayer Methane on Graphite at Low Temperatures H. K. Kim, Q. M. Zhang and M. H. W. Chan Simulation Studies of the Fluid-Solid Monolayer Transition in Argon absorbed on Graphite at 77.5 K D. Nicholson and N. G. Parsonage Phase Transition, Thermodynamics and Structural Analysis of Ethane Films adsorbed on Graphite. A Low-energy Electron-diffraction Study J-M. Gay, J. Suzanne and R. Wang Wetting at a Fluid-Wall Interface. Computer Simulation and Exact Statistical Sum Rules F. van Swol and J. R. Henderson What Controls the Thickness of Wetting Layers? R. F. Kayser, M. R. Moldover and J.W. Schmidt Interfacial Phase Transitions in Molecular Fluids and Multicomponent Mix- tures M. M. Telo da Gama and J. H. Thurtell Interfacial Phase Transitions of Microemulsions C. Borzi, R. Lipowsky and B. Widom Studies of Phase Transitions in Langmuir Monolayers by Fluorescence Micros- copy B. Moore, C. M. Knobler, D. Broseta and F. Rondelez Capillary Condensation and Adsorption in Cylindrical and Slit-like Pores R. Evans, U. Marini Bettolo Marconi and P. Tarazona Fluid Behaviour in Narrow Pores B. K. Peterson, J. P. R. B. Walton and K. E. Gubbins The Lennard-Jones Lecture: The Statistica J. S. Rowlinson General Discussion Zndex of Names List of Posters Mechanics of Sma t SystemsThe following papers were accepted for publication in J. Chem. SOC., Faraday Trans. I, during July 1986.511258 51 1633 511758 5/22 13 6/01 1 6/ 189 61371 61520 61537 6/616 61737 61745 6/773 6/78 1 61792 61820 61831 61834 61860 61880 61933 61934 Fourier-transform Infrared Studies of the Irreversible Oxidation of Cyanide at Plantinum Electrodes R. P. Cooney, A. S. Hinman and R. A. Kydd Polarographic Evidence for the Interaction of Reduced Nitromidazole Deriva- tives with DNA Bases P. J. Declerck and C. J. de Ranter Electrocatalysis under Temkin Adsorption Conditions A. Saraby-Reintjes Mobilities and Molar Volumes of Multi-charged Cations in N,N-Dimethyl- formamide at 25 "C W. Grzybkowski and M. Pilarczyk Surface Potential Measurements in Pentanol-Sodium Dodecyl Sulphate Mic- elles G. V. Hartland, F. Grieser and L. R. White The Mechanism of Oxygen Reactions at Porous Oxide Electrodes.Part 1 .--Oxygen Evolution at RuO, and Ru,Sn,-,O, Electrodes in Alkaline Solution under Vigorous Electrolysis Conditions M. E. G. Lyons and L. D. Burke Kinetics of Metal Oxide Dissolution. Oxidative Dissolution of Chromium from Mixed Nickel-Iron-Chromium Oxides by Potassium Permanganate A. B. O'Brien, M. G. Segal and W. J. Williams Local Polarity of Solvent Mixtures in the Field of Electronically Excited Molecules and Exciplexes P. Suppan Alluminium Distribution in Y Zeolites Dealuminated with Silicon Tetrachloride Vapour Extraction with H,Na,EDTA and X.P.S. Studies L. Kubelkova, L. D- udikova, Z. Bastl, G. Borbely and H. K. Beyer The Oriental Order of Solutions of a Dye Molecule in Liquid-crystalline Solvents J.W. Emsley, G. R. Luckhurst, G. N. Shilstone and I. Sage Direct Measurement of the Diffusion Coefficients of H Atoms in Six Gases G. Blyth, A. A. Clifford, P. Gray and J. I. Waddicor INDO and CND0/2 Calculations for Substituted Benzene Cations H. Chandran, M. C. R. Symons and A. Hasegawa Cross-relaxation of Adsorbed Hydrogen Atoms S. R. Seyedmonir and R. F. Howe Thermodynamics of Mixtures with a Hexane Isomer. Excess Volumes of 1-Chlorohexane-Hexane Isomer at 298.15 K A. Compostizo, A. C. Colin, R.- G. Rubio and M. D. Pena The Dielectric Properties of Ethyl and Methyl Carbamates in Aqueous Sol- ution J. B. Bateman and M. Thurai Kinetics of Hydrolysis of Phenyl Dichloroacetates in Aqueous Salt Solutions : Cosphere-Cosphere Overlap Effects J. B. F. N. Engberts, M.J. Blandamer, J. Burgess, B. Clark and A. W. Hakin The Effect of Gasification by Air (623 K) or CO, (1098 K) in the Development of Microporosity in Activated Carbons J. Garrido, A. Linares-Solano, J. M. Martin-Martinez, M. Molina-Sabio, F. Rodriguez-Reinoso and R. Torregrosa Polarity of Salt Solutions: A General Empirical Equation M. C. Rezende and L. I. Dal Sasso On the Consolidation of Concentrated Suspension. Part 1 .-The Theory of Sedimentations R. Buscall and L. R. White Light-induced Hydrogen Formation and Photouptake of Oxygen in Colloidal Suspension of a-Fe,O, J. Kiwi and M. Gratzel Effects of the Oxidation State of Cerium on the Reactivity and Sulphur Resistance of Ni"/Zeolite, Ni"/Silica and Ni"/CeO, Catalysts in Butane Hydr- ogenolysis I. Akalay, M.F. Guilleux, J. F. Tempere and D. Delafosse A Fourier-transform Intrared Study of the Role of Zeolite alewis Sites in Methanol Reactions over HZSM-5 Surfaces M. B. Sayed (ii)6/940 Monolayer Adsorption of Non-spherical Molecules on Solid Surfaces. Part 1 .-Adsorption of Hard Dumb-bells on Flat Surfaces L. Lajtar, J. Penar and S. Sokolowski 6/942 Ferrous Colbalt Ferrites. Preparation and Interfacial Behaviour S. Ardizzone, A. Chittofrati and L. Formaro 6/969 An Electron Spin Resonance Study of Single Crystals of X-Irradiated L-Ascorbic Acid at Room Temperature. Experimental Results and Semiempirical Calcul- ations J. T. Masiakowski and A. Lund 6/979 On the Identification of the Structural Transition Temperature in the Solutions of Dialkylsulphoxides S. A.Markarian, K. R. Grigorian and L. K. Simonian 6/ 10 1 7 Properties of Capillary-condensed Benzene W. D. Machin and P. D. Golding 6/ 1071 The Influence of Metal-Support Interactions on the Reactions of Methylcycl- ophentane over Supported Rh Catalysts J. B. F. Anderson, R. Burch and J. A. Cairns 6/01 73 The Effects of Titania and Alumina Overlayers on the Hydrogenation of CO over Rhodium M. E. Levin, M. Salmeron, A. T. Bell and G. A. Somorjai Promotion of Platinum-based Catalysts for Methanol Synthesis from Syngas P. Meriaudeau, K. Albano and C. Naccache 6/1075 Manipulation of the Selectivity of Rhodium by the use of Supports and Promoters G. Van der Lee, A. G. T. M. Bastein, J. Van den Boogert, B. Sch- uller, H. Lue and V. Ponec 6/ 1076 Alkali, Chlorine and other Promoters in the Silver-catalysed Selective Oxidation of Ethylene R.B. Grant, C. A. J. harbach, R. M. Labert and S. A. Tan 6/1077 Promotion in Heterogeneous Catalysis A. Kiennemann, R. Breault, J. P. Hin- dermann and M. Laurin 6/ 1078 The Role of Co in Sulphidised Co-Mo Hydrodesulphurisation Catalysts supported on Carbon and Alumina J. P. R. Vissers, V. H. J. de Beer and R. Prins 6/ 1079 Adsorption and Dissociation of CO, on a Potassium-promoted RhIII Surface F. Solymosi and L. Bugyi 6/ 1080 Adsorption and Reaction of Strained-metal Overlayers D. W. Goodman and C. H. F. Peden 6/ 108 1 Coordination of Carbon Monoxide to Transition-metal Surfaces R. A. Van Santen 6/1085 A Comparison of the Effects of Cu and Au on the Surface Reactivity of Ru (0001) B. Sakakini, A.J. Swift, J. C. Vickerman, C. Harendt and K. Christmann 6/ 1086 The Promotion of Surface-catalysed Reactions by Gaseous Additives: The Role of a Surface Oxygen Transien 6/ 1087 Comparison of Hydrogenation and Hydrogenolysis on Unsupported and Silica-supported Rh-V,O, and Pt-V,O, You-Jyh Lin, D. E. Resasco and G. L. Haller 6/ 1 101 Ultraviolet-Visible Reflectance Studies of Hydrogen Adsorption and CO-H, Interaction at MgO and CaO Surfaces E. Garrone and F. S. Stone 6/ 1 1 13 Rotating-disc Electrodes : ECE and DISPI Processes R. G. Compton, R. G. H- arland, P. R. Unwin and A. M. Waller 6/ 1 121 Origin of Idealized Static Thermodynamic Forces, inducing Solid Particle Motion, Orientation and Related Effects in a Solute Concentration Gradient E. A. Boucher C. T. Au and M.W. Roberts 6/1123 Bond Formation in Momentum Space 6/ 1 136 Study of Adsorption Sites on Thoria by STEM and Fourier-transform Infrared Spectroscopy Adsorption and Desorption of Water and Methanol X. Montagne, J. Lamotte, J. Lynch, E. Freund and J. C. Lavalley D. L. Cooper and N. L. Allan (iii)6/ 1 157 Reaction of Neopentane with Hydrogen over Platinum Metals Z. Karpinski, W. Juszczyk and J. Piekaszek 6/ 1 176 Promotion of Methanol Synthesis and the Water-gas Shift Reactions by Adsorbed Qxygen on Supported Copper Catalysts G. C. Chinchen, M. S. Spencer, K. C. Waugh and D. A. Whan 6/ 12 16 Oxygen Diffusion-concentration in Phospholipidic Model Membranes. An Electron Spin Resonance Saturation Study A. Vachon, C. Lecomte, P.- Braquet, V. Roman, M. Fatome and F.Berleur 6/1344 The Promoting Role of Cr and K Catalysts for High-pressure and High- temperature Methanol and Higher Alcohol Synthesis A. Riva, F. Trifiro, A. Vaccari, G. Busca and L. Mintchev 6/ 1434 Promotion of Nitrogen and Hydrogen Chemisorption and Ammonia Synthesis on Alumina-supported Hexagonal Tungsten Bronze K,WO, S. Stevenson and P. A. Sermon 6/ 1497 Propanaol Adsorption and Decomposition on Alkali-promoted Zinc Oxide D. Chadwick and P. J. R. O’MalleyCumulative Author Index 1986 Abraham, M. H., 3255 Abu-Gharib, E.-E. A., 1471 Abuladze, N. A., 2481 Adams, D. M., 1020 Adams, M., 1979 Aicart, E., 2977 Aida, M., 1619 Aika, K-i., 2269 Al-Hakim, M., 1575 Albery, W. J., 1033 Allen, G. C., 1367 Alwis, U. de, 1265 Amiya, S., 3141 Ammann, D., 1179 Anderson, J.A., 191 1 Anderson, M. W., 569, 1449, Andersson, S. L. T., 1537 Andersson, T., 767 Andrt, O., 2423 Antoniou, A. A., 483 Appleyard, I. P., 2817 Araya, P., 1351, 2473 Atherton, N. M., 3042 Attwood, D., 1903 Augustine, R. L., 3025 Avent, A. G., 1589 Aveyard, R., 1031, 2951, 1752 Balasubramanian, K., 2665 Baldwin, R. R., 89 Balk, R. W., 933 Ball, R. C., 3233 Barford, W., 3233 Barone, G., 2089 Barouch, E., 2801 Bartlett, J. R., 597 Bartlett, P. N., 1033 Battisti, A. De, 2481 Baur, J., 1081 Becker, K. A., 2151 Beezer, A. E., 2863, 2929 Belton, P. S., 451 Benecke, J. I., 1945 Bennett, C. O., 2155 Berei, K., 3003 Berezin, I. V., 319 Bernstein, T., 1879 Berroa de Ponce, H., 2811 Berry, F. J., 1023 Berti, P., 2547 Bhattacharyya, S. N., 2103 Bieth, H., 1935 Binks, B.P., 125, 1031, 1755 Biswas, P. K., 1973 Blackburn, G. M., 2965 285 1 125 Blake, P. G., 723 Blandamer, M. J., 1022, 1471, Blesa, M. A., 2345 Bloemendal, M., 53 Bloor, D., 21 11 Blakejowski, J., 3069 B.Nagy, O., 1789 Boelhouwer, C., 1945, 2704 Bond, G. C., 1985, 3043 Booth, B. L., 2007 Booth, C., 1865 Boucher, E. A., 1589 Bozonnet-Frenot, M-P., 2185 Brereton, I. M., 1999 Brett, C. M. A., 1071 Brigandi, P. W., 1032 Brillas, E., 495, 1781 Brown, 0. R., 3045 Bruckenstein, S., 1105 Buck, R. P., 1169 Bui, V. T., 899 Burch, R., 1985 Burgess, J., 1471, 2989 Busca, G., 3019 Cabani, S., 2547 Cameron, P., 1389 Campos, A., 2781 Canet, D., 2185 Carley, A. F., 723 Carpenter. T. A., 545 Carroll, €3. J., 3205 Casal, B., 1597 Cass, A. E. G., 1033 Castro, V. Di, 723 Castronuovo, G., 2089 Celda, B., 2781 Cenens, J., 281 Cesteros, L.C., 1321 Champion, J. V., 439 Chang, C. D., 1032 Chao, K-J., 2645 Chapman, D., 3048 Chiou, C. T., 243 Chitale, S. M., 663 Christensen, P. A., 3215 Christenson, H. K., 2735 Chung, J. S., 2155 Claesson, P. M., 2735 Clark, B., 1471 Clark, S., 125 Clarke, R. J., 2333 Clewley, J. D., 2589 Clifford, A. A., 2235 Coates, J. H., 2123, 2333 Cochran, S. J., 1721 2989 Cohen de Lara, E., 365 Cohen, H., 2627 Coller, B. A. W., 943 Compostizo, A., 1839 Conti, G., 2547 Contreras Viguria, E., 281 1 Cooney, R. P., 597 Copperthwaite, R. G., 1007 Cortks, J., 2473 Cortes, J., 1351 Corti, H. R., 921 Costas, M., 2977 Covington, A. K., 1209, 3042 Craston, D. H., 1033 Craven, J. R., 1865 Crespo Colin, A., 1839 Crilly, J. F., 439 Crowther, N.J., 2791 Crudden, J., 2195, 2207 Dalas, E., 2897 Danil de Namor, A. F., 3275, 2811, 349 Das, M. N., 1973 Dawber, J. G., 3097, 119 de Haas, M. P., 2933 De Schrijver, F. C., 281 Dean, C. E., 89 Dearden, S. J., 1627 Del Vecchio, P., 2089 Delaney, G. M., 2195,2207 Delannay, F., 2423 Delaval, Y., 365 Delmau, J., 3053 Delmon, B., 2423 Dharmalingam, P., 359 Dias Peiia, M., 1839 Domen, K., 2269 Domtnech, J., 1781 Dore, J. C., 241 1 Duatti, A., 1429 Duce, P. P., 1471 Dupldtre, G., 2825 Eagiand, D., 2008, 2791 Eastland, G., 2729 Eastland, G. W., 2833 Ebeid, E-Z. M., 909 Eden, J., 2945 Edmonds, R. N., 2515 Edwards, P. P., 2515 Egawa, C., 3197 Egdell, R. G., 2003 Ekechukwu, A. D., 1965 El-Daly, S. A., 909 Elbing, The Late E., 943 Elia, V., 2089 Elworthy, P. H., 1903AUTHOR INDEX Espenscheid, M.W., 1051 Espinosa-Jiminez, M., 329 Evans, D. F., 1829 Everett, D. H., 2589, 2605, 2915 Ewen, R. J., 1127 Farnia, G., 1885 Far0 Jr, A. C., 3125 Feakins, D., 2195, 2207, 563 Fegan, S. G., 785, 801 Fernandez-Prini, R., 92 1 Fernandez-Valverde, S. M., 2825 Findenegg, G. H., 2001, 2691 Fink, P., 1879 Fisher, D. T., 119 Flanagan, T. B., 2175, 2589 Fletcher, A. J. P., 2605 Fletcher, P. D. I., 2651, 231 1 Folman, M., 2025 Foulds, N. C., 1259 Fraser, I. M., 607, 2747 Freiser, H., 1217 Freund, P. L., 2277 Fricke, R., 263, 273 Fukuda, H., 1561 Funabiki, T., 35, 707, 1771 Fyles, T. M., 617 Gaboriaud, R., 2301 Gabrys, B., 1923, 1929 Ganghi, N. S., 2367 Garbassi, F., 2043 Garbowski, E., 1893 Garden, D., 3 113 Garrido, J. A., 1781 Gellan, A., 953 Geoffroy, M., 521 Gervasini, A., 1795 Ghatak-Roy, A.R., 1051 Ghoneim, M. M., 909 Ghousseini, L., 349, 3275 Gilbert, R. G., 1979, 2247 Gilhooley, K., 431 Gobolos, S., 2423 Gomez-EstCvez, J. L., 2167 Gonzilez-Caballero, F., 329 Gonzalez-Elipe, A. R., 739 Gonzalez-Fernindez, C. F., 329 Gopalakrishnan, R., 2635 Gormally, J., 157, 2497 Gorton, L., 1245 Gosal, N., 1471 Green, M. J., 1237 Green, N. J. B., 2673 Grieser, F., 1813, 1829 Gritzner, G., 1955 Grzybkowski, W., 1381, 1703, Guardado, P., 1471 Haddad-Fahed, O., 2301 Haggett, 3. G. D., 1033 Haines, A. J., 2817 Haines, G. A. J., 2817 Hakin, A. W., 1471, 2989 Hall, D., 21 11 Halle, B., 401, 415 1745 Hamada, K., 3141 HansCn, O., 77 Harriman, A., 3215 Harris, R. K., 2817 Harrison, M. R., 2515 Havredaki, V. I., 2531 Heatley, F., 255 Heberger, K., 2621 Hedges, W.M., 179 Hellring, S. D., 1032 Hemfrey, J. P., 1589 Hersey, A., 1271 Hewitt, E. A., 869 Hey, M. J., 1805, 2817 Heyrovsky, M., 585 Hidaka, H., 2615 Higgins, J. S., 2004, 1923, 1929, Higson, S., 157 Hill, C. A. S., 1127 Hill, H. A. O., 1237 Hill, T., 349 Hitchman, M. L., 1223 Hobert, H., 1527, 2505 Hobson, D. B., 869 Homer, J., 533 Honeybourne, C. L., 1127 Honeyman, M. R., 89 Hooper, A., 11 17 Hoppach, D., 3053 Houghton, J. D., 1127 Howe, A. hiz., 2411 Howe, R. F., 2887 Hronec, M., 1405 Hsu, W. P., 851 Huang, W-S., 2385 Hubbard, C. D., 1471 Hummel, A., 2933,2945 Humphrey, B. D., 2385 Humphreys, F. J., 1020, 2006 Hunt, D. J., 189 Hunter, W. H., 2863, 2929 Hussian, S. M., 2221 Hutchings, G. J., 1007 Ige, J., 2011 Iijima, T., 3141 Iizuka, T., 1681, 61 Ikeda, H., 61 Ikeda, O., 1561 Indaratna, K., 2755, 2763 Indelli, A., 1429 Inomata, S., 1733 Inoue, M., 2175 Inoue, T., 168 1 Ishii, T., 2615 Ishikawa, T., 2401 Issa, R.M., 909 Iwamoto, M., 1713 Jackson, S. D., 431, 189, 2719 Jaeger, N., 205 Japaridze, J. I., 2481 Japaridze, S. S., 2481 Jayasuriya, D. S., 457,473 Jensen, M., 1351 3047 Johns, A. I., 2235 Johnson, D. C., 1081 Johnson, J., 1081 Johnston, P., 1007 Jones, W., 545, 3081 Jonson, B., 767 Jose, C. I., 663, 681, 691 Kadhum, A. A. H., 2521 Kakuta, N., 1553 Kamat, P. V., 1031 Kaminade, T., 707 Kaner, R. B., 2323 Karaiskakis, G., 2897 Katime, I., 1321, 1333 Katsanos, N. A., 2897 Kavetskaya, 0. I., 319 Kawaguchi, T., 1441 Kawai, S., 527 Kawai, T., 527 Kazusaka, A., 1553 Kelly, H.C., 1271 Kelly, K. P., 3025 Kelly, R. G., 1195 Kemball, C., 3044, 3113, 3125 Kevan, L., 213 Khan, S. U. M., 291 1 Khoo, K. H., 1 Kido, K., 2269 Kinoshita, N., 2269 Kishimoto, S., 2175 Kiss, I., 3003 Kjellander, R., 2735 Kleine, A,, 205 Klinowski, J., 569, 1449, 2851 KodejS, Z., 1853 Komatsu, T., 1713 Komiyama, M., 1713 Kondo, S., 2401 Kondo, Y., 2141 Koreeda, A., 527 Koresh, J. E., 2057 Koster, F., 2691 Kowalak, S., 2151 Kremer, M. L., 2133 Krishnasamy, V., 2665 Kuji, T., 2589 Kurotaki, K., 2843 Kusabayashi, S., 2141 Kuzuya, M., 1441 Lamotte, J., 3019 Lancz, M., 883 Lang, J., 109 Langevin, D., 2001 Larkins, F. P., 1721 Larsson, R., 767 Laurie, O., 3149 Lavalley, J-C., 3019 Lawless, T. A., 1031, 2951 Lawrence, K. G., 563, 2195, Lawrence, M. J., 1903 Leaist, D. G., 247 Lelikvre, J., 2301 Gonard, J., 899 2207Lilley, T.H., 2965 Lim, T.-K., 69 Lin, J-C., 2645 Lincoln, S . F., 2123, 2333, 1999 Llars, S . , 767 Llinares, A., 521 Lockhart, J. C., 1161 Loewenschuss, A., 993, 2873 Logan, S . R., 161 Lomen, C. E., 1265 Lorenzelli, V., 3019 Lowe, B. M., 785, 801 Lowe, C. R., 1259 Lukacs, J., 2621 Lundin, S . T., 767 MacCallum, J. R., 607, 2747 MacDiarmid, A. G., 2323, 2385 Mactaggart, J. W., 1805 Mahnke, R., 1413 Malliaris, A., 109 Mandal, P. C., 2103 Manes, M., 243 Marabini, A. M., 2043 Maran, F., 1885 Marchal, J-P., 2185 Marcus, Y., 233, 993, 2873, 3255 Marczewski, M., 1687 Maroto, A. J. G., 2345 Marshall, W. L., 2283 Martin, C. R., 1051 Maruthamuthu, P., 359 Marx, U., 2505 Mastikhin, V. M., 1879 Matheson, R. A., 2755, 2763 Mathieu, M-V., 1893 MatijeviC, E., 2801 Matsuda, T., 1357 McCarthy, S., 943 McQuillan, A.J., 2755, 2763 Mead, J., 125, 1031, 1755 Meiler, W., 3053 Melchor, A., 1893 Miale, J. N., 1032 Miasik, J. J., 1117 Michael, A., 3053 Michel, D., 3053 Midgley, D., 1187 Milburn, P. J., 2965 Miles, R. J., 2929 Minami, Z., 1357 Mishima, S., 1307 Mishra, S . P., 521 Miura, H., 1357 Miyake, Y ., 15 15 Miyamoto, A., 13 Mobbs, R. H., 1865 Mol, J. C., 1945, 2707 Mollett, C . C., 1589 Mollica, V., 2547 Molyneux, P., 291, 635, 3287 Moore 111, R. B., 1051 Morando, P. J., 2345 Morazzoni, F., 1795 Morehouse, K., 3215 AUTHOR INDEX Morgan, H., 143 Mori, K., 13 Moyes, R. B., 189, 2719 Mulla, S . T., 681, 691 Murakami, Y., 13 Murata, M., 2615 Muscetta, M., 2089 Nagano, S . , 1357 Naito, S., 3197 Najbar, M., 1673 Nakajima, T., 1307 Nakamatsu, H., 527 Nakanishi, M., 1441 Nakano, A., 2141 Napper, D.H., 1979,2247 Narayana, M., 213 Neta, P., 3215 Neto, M. M. P. M., 1071 Neuburger, G. G., 1081 Nex, C. M. M., 3233 Nikitas, P., 977 Nitta, S . , 2401 Nock, A., 2817 Noyes, R. M., 2999 tubkowski, J., 3069 Nyasulu, F. W. M., 1223 Oakes, J., 2079, 3149 Oesch, U., 1179 Ogino, Y., 1713 Ohlmann, G., 263, 273 Okazaki, S . , 61 Okubo, T., 3163, 3175, 3185 Okuda, T., 1441 Oldfield, M. J., 2673 Oldham, K. B., 1099 Onai, T., 2615 Onishi, T., 2269 Ooe, M., 35 OpaHo, M., 339 Orchard, S . W., 1007 Oref, I., 1289 O’Reilly, P. J., 2195, 2207 Ortiz, A., 495 Owen, A. E., 1195 Parbhoo, B., 1789 Park, C-N., 2589 Parry, D. E., 3051 Parsegian, V. A., 2801 Parsons, B. J., 1575 Patterson, D., 2977 Pease, W.R., 747, 759 Peeters, G., 963 Peeters, S., 963 Penboss, I. A., 2247 Penner, R. M., 1051 Perry, M. C., 533 Pethig, R., 143 Petropoios, J. H., 2459 Petropoulos, J. H., 2531 Pettersson, A., 2435 Pfeifer, H., 3053 Pham, H. V., 1179 Phillips, G . O., 1575 Piculell, L., 387, 401, 415 Piekarska, A., 513 Piekarski, H., 513 Pilarczyk, M., 1703, 1745 Pilling, M. J., 2673 Pimblott, S . M., 2673 Pinna, F., 1795 Pletcher, D., 179 Polta, J. A., 1081 Polta, T. Z., 1081 Porter, G., 3215 Porter, S . J., 2323 Pouchli, J., 1605 Primet, M., 1893 Puchalska, D., 1381 Puttock, S . J., 2773, 3013, 3033 Quintana, J. R., 1333 Radulovic, S., 1471 Rajaram, R. R., 1985 Ramakrishna Rao, D. N., 2367 Ramdas, S., 545 Rao, D. N. R., 2833 Rashid, S . , 2235 Rebenstorf, B., 767 Richards, W.G., 3047 Richardson, P. J., 869 Rideout, J., 167 Rigby, S., 431 Rizkallah, P. J., 1589 Roberts, M. W., 723 Robinson, B. H., 1271, 2311, Robinson, P. J., 869 Rochester, C. H., 953, 1805, Rockliffe, J. W., 3149 Rodriguez, R. M., 178 1 Rooney, J. J., 2005 Rosenholm, J. B., 77, 2435 Rossi, P. F., 3019 Rouw, A. C . , 53 Rubio, R. G., 1839 Rudham, R., 2817 Ruiz-Hitzky, E., 1597 Russell, D., 2729 Ryder. P. L., 205 Sacchetto, G. A., 1853 Saez, C., 1839 Saleh, J. M., 2221 Salmon, G. A., 161, 2521 Sanchez, F., 1471 Sandona, G., 1885 Sangster, D. F., 1979 Saris, P., 2435 Sarkany, A., 103 Saur, O., 3019 Saville, G., 3041, 3046 Sawada, K., 1733 Scharpf, O., 1923, 1929 Schiller, R. L., 2123 Schlosserova, J., 1405 Schmelzer, J., 1413, 1421 Schmitt, K. D., 1032 Schoonheydt, R.A., 281 Richoux, M-C., 3215 241 1 191 1, 2569, 2773, 3013, 3033 (vii)Schulz-Ekloff, G., 205 Scott, J. M. W., 2989 Scott, R. P., 1389 Segal, M. G., 3245 Segall, R. L., 747, 759 Seloudoux, R., 365 Seni5, M., 2065 Sharma, S., 2497 Shibata, Y., 1357 Shigeto, M., 1515 Shindo, H., 45 Shubin, A. A., 1879 Sidahmed, 1. M., 2577 Siiman, O., 851 Simmons, R. F., 1965 Simon, W., 1179 Sircar, S., 831, 843 Smallridge, M. J., 1589 Smart, R. St C., 747, 759 Smith, D. G., 2569 Smith, E. G., 3149 Smith, I., 869 Smith, J. A. S., 2004 Snowdon, S., 943 Soffer, A., 2057, 2627 Sokoll, R., 1527, 2505 Solymosi, F., 883 Somsen, G., 53, 933 Soria, J., 739 Soria, V., 2781 Soriyan, O., 2011 Spiess, B., 1935 Spiro, M., 2277, 3048 Spotswood, T. M., 1999 Stenius, P., 2735 Stephens, A., 2729 Strazielle, C., 1321 Strukul, G., 1795 Strumolo, D., 1795 Sugiyama, K., 1357 Suppan, P., 509 Sutherland, I.O., 1145 Suzuki, T., 1733 Swallow, A. J., 1575 Symanski, J. S . , 1105 Symons, M. C. R., 167, 2367, 2833, 2729 Szentirmay, M. N., 1051 Tabony, J., 231 1 AUTHOR INDEX Tadros, Th. F., 3045 Take, S., 3141 Tamaru, K., 3197 Tamura, H., 1561 Tamura, K., 1619 Tanaka, T., 35 Tanaka, Y., 2065 Tang, A. P-C., 1081 Taniewska-Osinska, S., 1299 Taniewska-Osinska, S., 5 13 Tardajos, G., 2977 Tatam, R. P., 439 Tawarah, K., 21 11 Tear, S. P., 1022 Tejero, R., 2781 Tennakoon, B., 3081 Tennakoon, D. T. B., 545 Teramoto, M., 1515 Thijs, A., 963 Thomas, J. D. R., 1135 Thomas, J. M., 545, 2851, 3081 Thomson, A. J., 2009 Tiddy, G. J. T., 3043 Tilak, D., 3081 Tobias, H., 2627 Tofield, B.C., 11 17 Toprakcioglu, C., 241 1 Townsend, R. P., 1019 Trasatti, S., 2481 Tunuli, M. S., 2911 Turner, J. C. R., 3052 Turner, P. S., 747, 759 Tyler, J. W., 1367 van de Ven, T. G. M., 457,473 van Lith, D., 2945, 2933 Vansant, E. F., 963 Vadros, L., 3003 Vekavakayanondha, S., 291, Venkatasubramanian, L., 359 Verhaert, I., 963 Veseli, V., 1405 Vidoczy, T., 2621 Vijlder, M. De, 2377 Vink, H., 2353 Viswanathan, B., 2635 Volkov, A. I., 815 Volpe, P. L. O., 2863, 2929 Vonk, D., 1945 635, 3287 Waghorne, W. E., 563, 2195, Walker, R. W., 89 Wallwork, S. C., 1589 Walton, A. J., 1023 Wang, Z-C., 375 Ward, R. J., 2915 Warhurst, P. R., 119 Warman, J. M., 2933, 2945 Warr, G. G., 1813, 1829 Watson, J. T. R., 2235 Watts, P., 1389 Weale, K. E., 1020, 2002 Weiss, E., 2025 Wells, C.F., 2577 Wells, P. B., 189, 2719 Whalley, P. D., 1209 Whan, D. A., 31 13 Whyman, R., 189, 2719 Wiens, B., 247 Williams, G., 3049, 3050 Williams, R. A., 3097 Williams, W. J., 3245 Wilson, G. S., 1265 Wilson, I. R., 943 Wojcik, D., 1381 Woinicka, J., 1299 Wren, B. W., 167 Wright, K. M., 451 Wright, P. G., 2557, 2565 Wu, E. L., 1032 Wu, Q., 2423 Wuthier, U., 1179 Wyn-Jones, E., 21 11 Wysocki, S., 715 Xiaoding, X., 1945, 2707 Yamashita, H., 1771, 707 Yamazaki, A., 1553 Yatsimirsky, A. K., 319 Yeates, S. G., 1865 Yeo, I-H., 1081 Yoshida, N., 2175 Yoshida, S., 35, 707, 1771 Yoshikawa, M., 707, 1771 You-Sing, Y., 2887 Zana, R., 109 Zanderighi, L., 1795 Zund, R., 1179 2207 (viii)NOMENCLATURE A N D SYMBOLISM Units and Symbols. The Symbols Committee of The Royal Society, of which The Royal Society of Chemistry is a participating member, has produced a set of recommendations in a pamphlet ‘Quantities, Units, and Symbols’ (1 975) (copies of this pamphlet and further details can be obtained from the Manager, Journals, The Royal Society of Chemistry, Burlington House, London W1 V OBN).These recommendations are applied by The Royal Society of Chemistry in all its publications. Their basis is the ‘Systbme International d‘Unit6s’ (SI). A more detailed treatment of units and symbols with specific application to chemistry is given in the IUPAC Manual of Symbols and Terminology for Ph ysicochemical Quantities and Units (Pergamon, Oxford, 1979). Nomenclature. 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 the rules themselves and guidance on their use are given: Nomenclature of Organic Chemistry, Sections A, B, C, D, E, F, and H (Pergamon, Oxford, 1979 edn). Nomenclature of Inorganic Chemistry (Butterworths, London, 1971, now publis- hed by Pergamon). Biochemical Nomenclature and Related Documents (The Biochemical Society, London, 1978). A complete listing of all IUPAC nomenclature publications appears in the January issues of J. Chem.SOC., Faraday Transactions. It is recommended that where there are no IU PAC rules for the naming of particular compounds or authors find difficulty in applying the existing rules, they should seek the advice of the Society’s editorial staff.THE FARADAY DIVISION O F THE ROYAL SOCIETY OF CHEMISTRY SYMPOSIUM No. 2 2 Interaction-induced Spectra in Dense Fluids and Disordered Solids University of Cambridge, 10-1 1 December 1986 Organ king Committee : Professor A. D. Buckingham (Chairman) Dr R. M. Lynden-Bell Dr P. A. Madden Professor E. W. J. Mitchell Dr J. Yarwood Dr D. A. Young MrsY.A. Fish Whilst interaction-induced spectra have been studied in the gas phase for many years, their importance in the spectroscopy of condensed matter has been appreciated only relatively recently.At present a considerable number of studies of induced spectra are taking place in what are (nominally) widelyseparated fields of study. It is highlydesirable to bring these communities together so that common issues can be identified and the progress of one field appreciated in another. The final programme and application form may be obtained from: Mrs Y. A. Fish, The Royal Society of Chemistry, Burlington House, London W1V OBN ~~~~~~~ ~~~ ~ THE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY GENERAL DISCUSSION No. 83 Brownian Motion University of Cambridge, 7-9 April 1987 Organising Committee: Dr M. La1 (Chairman) Dr R. Ball Dr E. Dickinson Dr J. S. Higgins Dr P. N. Pusey Dr D. A. Young Mrs Y. A. Fish The Faraday Discussion on Brownian Motion will be introduced by Professor J.M. Deutch of MIT and will include contributions from P. Mazur, P. Meakin, R. Jullien, D. A. Weitz, M. Fixman, P. N. Pusey, R. H. Ottewill, A. Vrij, J. A. McCammon, B. A. Ackerson and V. Degiorgio dealing with hydrodynamics, fractals, Brownian dynamics of aggregation processes and photon correlation spectroscopy. There will be a poster session for which contributions are invited in the form of a brief abstract to be sent by 31 January 1987 to: Dr M . Lal, Unilever Research, Port Sunlight Laboratory, Bebington, Wirral L63 3JW. The preliminary programme may be obtained from: Mrs Y. A. Fish, The Royal Society of Chemistry, Burlington House, London W1V OBNTHE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY GENERAL DISCUSSION N o .84 Dynamics of Elementary Gas-phase Reactions University of Birmingham, 14-16 September 1987 Organising Committee: Professor R. Grice (Chairman) Dr M. S. Child Dr J. N. L. Connor Dr M. J. Pilling Professor I. W. M. Smith Professor J. P. Simons The Discussion will focus on the development of experimental and theoretical approaches to the detailed description of elementary gas-phase reaction dynamics. Studies of reactions at high collision energy, state-to-state kinetics, non-adiabatic processes and thermal energy reactions will be included. Emphasis will be placed on systems exhibiting kinetic and dynamical behaviour which can be related to the structure of the reaction potential- energy surface or surfaces. Further information may be obtained from: Professor R.Grice, Chemistry Department, University of Manchester, Manchester M13 9PL THE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY SYMPOSIUM N o . 2 3 Molecular Vibrations University of Reading, 15-16 December 1987 Organising Committee: Professor I. M. Mills (Chairman) Dr J. E. Baggott Professor A. D. Buckingham Dr M. S. Child Dr N. C. Handy Dr B. J. Howard The Symposium will focus on recent advances in our understanding of the vibrations of polyatomic molecules. The topics to be discussed will include force field determinations by both ab initiu and experimental methods, anharmonic effects in overtone spectroscopy, local modes and anharmonic resonances, intramolecular vibrational relaxation, and the frontier with molecular dynamics and reaction kinetics.Contributions for consideration by the Organising Committee are invited. Titles should be submitted as soon as possible, and should be sent before the end of October 1986 to: Professor 1. M. Mills, Department of Chemistry, University of Reading, Reading RG6 2AD.THE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY Marlow Medal and Prize Applications are invited for the award of the Marlow Medal for 1987 and Prize of f 100. The award will be open to any member of the Faraday Division of the Royal Society of Chemistry, who by the age of 32, had made in the judgement of the Council of the Faraday Division, the most meritorious contribution to physical chemistry or chemical physics. The award will be made on the basis of publications (not necessarily in the Transactions) on any subject normally published in J.Chem. SOC., Faraday Transactions land 11, that carry a date of receipt for publication not later than the candidate’s 32nd birthday. Candidates should be members and under 34 on 1st January 1987, the closing date for applications, which may be made either by the candidate himself or on his behalf by another member of the Society. Copies of the rules of the award and application forms may be obtained from: Mrs Y. A. Fish, The Royal Society of Chemistry, Burlington House, London WIV OBN. (xii)JOURNAL OF CHEMICAL RESEARCH Papers dealing with physical chemistry/chemical physics which have appeared recently in 1, Chern. Research, The Royal Society of Chemistry‘s synopsis + microform journal, include the following: Phenanthrene Hydroconversion over Nickel and Molybdenum Sulphides Supported on Alumina: Effect of the Sulphidation Method Jean-Louis Lemberton and Michael Guisnet (1 986, Issue 6) Deposition of Platinum onto CdS Aqueous Suspensions under Ultraviolet Illumination Javier Domenech, John Curran, Nicole Jaffrezic-Renault and Robert Philippe (1986, Issue 6) A Prototype Model for Artificial Photosynthetic Membranes: Water-swollen Chelate Filter Paper with Adsorbed Tris(2,2’-bipyridine)ruthenium(2+) and Methyl Viologen Yoshimi Kurimura, Noriko Matsuo, Etsuko Kokuta, Yasuyuki Takagi and Yoshiharu Usui (1 986, Issue 7) ln situ Electrochemical Electron Spin Resonance Spectrometry: the Anodic Oxidation of Triphenylmethanol Richard G. Compton, Barry A.Coles and Michael J.Day (1986, Issue 7) Racemization of Peptides. An MNDO Study of the c-(Gly-Gly) Aniom Miguel Pons, Josep M. Bofill and Ernest Giralt (1 986, Issue 7) Hydrophobic Interactions of Gaseous Hydrocarbons derived from Studies of their Solubilities in Mixtures of Water and Ethanol Robert W. Cargill and Donald E. MacPhee (1 986, Issue 8) FARADAY DIVISION INFORMAL AND GROUP MEETINGS Division Half-day Endowed Lecture Symposium - Chemistry at Surfaces: Metals, Oxides and Semiconductors including the Tilden Lecture by Professor J. Pritchard and the Meldola Lecture by Dr J. S. Foord To be held at the Scientific Societies Lecture Theatre, London on 4 November 1986 Further information from Mrs Y. A. Fish, The Royal Society of Chemistry, Burlington House, London WIV OBN Colloid and Interface Science Group with Macrogroup UK Polymer-Polymer Interfaces To be held at the Scientific Societies Lecture Theatre, London on 15 December 1986 Further information from Dr R.Aveyard, Department of Chemistry, The University, Hull HU6 7RX Colloid and Interface Science Group with the Colloid and Surface Group of the SCI Nucleation and Growth in Colloidal Systems To be held at the Society of Chemical Industry, 14 Belgrave Square, London on 16 December 1986 Further information from Dr R. Aveyard, Department of Chemistry, The University, Hull HU6 7RX Neutron Scattering Group Neutron Crystallography To be held at Imperial College, London on 17-19 December 1986 Further information from Dr R. J. Newport, Physics Laboratory, The University, Canterbury, Kent CT2 7NR Electrochemistry Group The Photoelectrochemical Properties of Colloids To be held at the University of Southampton on 7-8 January 1987 Further information from Dr S.P. Tyefield, CEGB Berkeley Laboratories, Berkeley, Gloucestershire GL13 9PB (xiii)Electrochemistry Group with the SC/ and the Institute of Corrosion Science Technology Electrochemical Techniques for Corrosion Scientists To be held at St. Catherine's College, Oxford on 7-8 January 1987 Further information from Dr S. P. Tyefield, CEGB, Berkeley Nuclear Laboratories, Berkeley, Gloucestershire GL13 9PB Division jointly with Perkin Division Half-Day Endowed Lecture Symposium on Radical Ion and Carbon ion Chemistry including the lngold Lecture by F. G. Bordwell To be held at University College, London on 12 March 1987 Further information from Mrs Y.A. Fish, The Royal Society of Chemistry, Burlington House, London WIVOBN Neutron Scattering Group Neutron Scattering and Phase Transitions To be held at the University of Warwick on 30-31 March 1987 Further information from Dr D. McK. Paul, Department of Physics, University of Warwick, Coventry CV4 7AL Electrochemistry Group Spring Informal Meeting To be held at the University of Bristol on 1-3 April 1987 Further information from Dr A. R. Hillman, School of Chemistry, University of Bristol, Cantock's Close, Bristol BS8 ITS Division - Annual Congress The Chemistry and Physics of Intercalation To be held at University College, Swansea on 13-1 6 April 1987 Further information from Professor J. H. Purnell, Department of Chemistry, University College, Singleton Park, Swansea SA2 8PP Polymer Physics Group Fundamental Aspects of Polymer Flammability To be held at Baden Powell House, London on 14-15 April 1987 Further information from Dr G.C. Stevens, CERL, Kelvin Avenue, Leatherhead KT22 7SE Division Full-day Endowed Lecture Symposium on Intramolecular Dynamics and Chemical Reactivity including the Centenary Lecture by S. A. Rice and the Tilden Lecture by M. S. Child To be held at the Scientific Societies Lecture Theatre, London on 6 May 1987 Further information from Mrs Y. A. Fish, The Royal Society of Chemistry, Burlington House, London WIV OBN Polymer Physics Group Electroactive Polymers To be held at the Geological Society, London on 14 May 1987 Further information from Dr G.C. Stevens, CERL, Kelvin Avenue, Leatherhead KT22 7SE Electrochemistry Group with Macro Group UK Polymer Electrolytes To be held at the University of St. Andrews on 18-19 June 1987 Further information from Dr C. A. Vincent or Dr J. R. MacCallum, Department of Chemistry, University of St. Andrews, St. Andrews KY16 9ST Division Xlth International Symposium on Molecular Beams To be held at the University of Edinburgh on 13-17 July 1987 Further information from Dr J. F. Gibson, The Royal Society of Chemistry, Burlington House, London WIV OBNIndustrial Ph ysical Chemistry Group The Physical Chemistry of Small Carbohydrates (as part of the International Symposium on Solute-Solute-Solvent Interactions) To be held at the University of Regensburg, West Germany on 10-14 August 1987 Further information from Dr F.Franks, Pafra Ltd, 150 Science Park, Milton Road, Cambridge CB4 4GG Polymer Ph ysics Group Biennial Meeting To be held at University of Reading on 9-1 1 September 1987 Further information from Dr D. Bassett, Department of Physics, University of Reading, Reading RG7 2AD Neutron Scattering Group Applications of Neutron and X-Ray Optics To be held a t the University of Oxford on 14-15 September 1987 Further information from Dr R. K. Thomas, Physical Chemistry Laboratory, South Parks Road, Oxford OX1 302 Polymer Ph ysics Group New Materials To be held at the University of Warwick on 22-25 September 1987 Further information from Dr M. J. Richardson, Division of Materials Applications, National Physical Laboratory, Queens Road, Teddington, Middlesex TWl1 OLW Division Autumn Meeting Spectroscopy of Gas-phase Molecular Ions and Clusters To be held at the University of Nottingham on 22-24 September 1987 Further information from Professor J. P. Simons, Department of Chemistry, University of Nottingham, Nottingham NG7 2RDIndustrial Ph ysical Chemistry Group The Physical Chemistry of Small Carbohydrates (as part of the International Symposium on Solute-Solute-Solvent Interactions) To be held at the University of Regensburg, West Germany on 10-14 August 1987 Further information from Dr F. Franks, Pafra Ltd, 150 Science Park, Milton Road, Cambridge CB4 4GG Polymer Ph ysics Group Biennial Meeting To be held at University of Reading on 9-1 1 September 1987 Further information from Dr D. Bassett, Department of Physics, University of Reading, Reading RG7 2AD Neutron Scattering Group Applications of Neutron and X-Ray Optics To be held a t the University of Oxford on 14-15 September 1987 Further information from Dr R. K. Thomas, Physical Chemistry Laboratory, South Parks Road, Oxford OX1 302 Polymer Ph ysics Group New Materials To be held at the University of Warwick on 22-25 September 1987 Further information from Dr M. J. Richardson, Division of Materials Applications, National Physical Laboratory, Queens Road, Teddington, Middlesex TWl1 OLW Division Autumn Meeting Spectroscopy of Gas-phase Molecular Ions and Clusters To be held at the University of Nottingham on 22-24 September 1987 Further information from Professor J. P. Simons, Department of Chemistry, University of Nottingham, Nottingham NG7 2RD

ISSN:0300-9599    

DOI:10.1039/F198682BP129

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. I , 1986,82, 3053-3067 13C Nuclear Magnetic Resonance Investigations of Carbon Monoxide in Decationated Zeolites of Type Y Andreas Michael, Wolfgang Meiler, Dieter Michel* and Harry Pfeifer Sektion Physik, Karl-Marx- Universitat Leipzig, DDR- 701 0 Leipzig, German Democratic Republic Dieter Hoppach Sektion Chemie, Karl-Marx- Universitat Leipzig, DDR-7010 Leipzig, German Democratic Republic Jean Delmau Laboratoire de Spectroscopie, U.E.R. de Physique, Universitk Claude Bernard Lyon 1, 69621 Villeurbanne, France The 13C n.m.r. of carbon monoxide adsorbed on decationated zeolites of type Y is characterized by extremely large resonance shifts to lower magnetic field. The strong dependence of this resonance shift on temperature and on coverage can be explained by means of an exchange model described in the 1iterature.lS This treatment yields the resonance shift of carbon monoxide molecules in the adsorption complex (350 ppm relative to gaseous CQ) and the number of sites in zeolites effective for the complex formation (0.07 active sites per cavity). Investigations of the 13C nuclear magnetic relaxation of CO and CO, adsorbed on zeolite NaY and its decationated forms suggest that the large resonance shifts are due to an interaction with Lewis-type centres in the form of extra-lattice aluminium ions.Thus, investigations of the CQ adsorption by means of the I3C n.m.r. method are suitable for a characterization of Lewis-acid sites in zeolites. In a previous paper3 it was shown that the 13C n.m.r. lines of CO adsorbed on decationated zeolites show very large resonance shifts to lower magnetic fields with respect to the gaseous state.For zeolites NaA, NaX and NaY a similar effect could not be detected. Since the n.m.r. shift has a sensitive dependence on the conditions of pretreatment of the decationated zeolites which produce Lewis-acid sites it was suggested that the shifts are due to interactions of carbon monoxide molecules with this kind of acid sites. Although this hypothesis is supported by the results of quantum-chemical ab initio calculations for an idealized A13+/C0 ~ornplex,~ there remains some doubt concerning the influence of paramagnetic impurities. It is the aim of this paper to present more arguments for the interpretation of the n.m.r. shifts observed and for a detailed description of their dependence on coverage and on temperature.In order to understand better thz role of paramagnetic impurities in the zeolites used, the 13C nuclear magnetic relaxation of CO and CO, adsorbed on NaY and its decationated form has been studied. On this basis an interpretation of the observed resonance shifts has been made possible. Experimental 13C n.m.r. spectra were run at 22.63 MHz in the temperature range 120-450 K. The 13C nuclear magnetic relaxation times and T, were measured at 25.1 MHz and at 22.63 MHz using the inversion-recovery and Hahn’s spin-echo method, respectively. To 3053 101-23054 N.M.R. of CO in Zeolites Fig. 1. 13C n.m.r. spectra of carbon monoxide adsorbed on zeolite 30DeNaY as a function of coverage, N, (in molecules per cavity) at 273 K (a) and for different temperatures (K) with a coverage of 0.4 molecules per cavity (b).increase the signal-to-noise ratio, especially for the n.m.r. relaxation time measurements, the adsorbates CO and CO, were enriched with 13C nuclei to values between 50 and 90%. The use of these substances also allowed measurements of 13C n.m.r. spectra at very low coverages. The NaX and NaY zeolites studied in the present paper were purchased from VEBA . Michael et al. 3055 Chemiekombinat Bitterfeld-Wolfen, GDR, and are characterized by Si/Al ratios of 1.35 and 2.6, respectively. The decationated forms of zeolite Y, abbreviated as DeNaY, were prepared from NH,NaY zeolites with different degrees of ammonium exchange through thermal decomposition under self-steaming conditions (deep-bed activation3).For instance, 30DeNaY denotes such a zeolite where 30% of the sodium ions were exchanged by NH:. Results of 13C N.M.R. Shift Measurements Basic Results The 13C n.m.r. shifts of CO and CO, molecules adsorbed on NaX, NaY and NaA zeolites are not very different from the values in the gaseous state3 (- 1 to - 4 ppm for CO and -0.5 to -2 ppm for CO,, where the negative sign indicates shifts for the adsorbed molecules to higher fields). The same result holds for these adsorbate-adsorbent systems if Na+ is exchanged by other alkali cations or T1+. The resonance shifts depend only weakly on the temperature. Since the values for the shifts are of the same order of magnitude as the corrections due to the specimen susceptibility (ca.0.7 ppm to higher fields) and to van der Waals interactions, they are without further interest in this work. Strong 13C n.m.r. shifts occur when CO molecules are adsorbed on decationated DeNaY zeolites. The values are largest for deep-bed treated specimens, but also for other treatments of decationated forms of zeolite Y they are much greater than for zeolites containing alkali cations. No such effects could be observed for CO,.,? The 13C n.m.r. shifts for CO adsorbed on DeNaY zeolites increase monotonically with decreasing coverage N and with increasing temperature between ca. 200 and 280 K. These variations, which are completely reversible, and the influence of the treatment and of other modifications of the zeolite specimen are important for the interpretation of the results.Dependence on Coverage To derive values for the concentration of the active sites interacting with CO molecules in the decationated zeolites and for the n.m.r. shift of a molecule in such a complex, the dependence of the observed n.m.r. shift on the number of adsorbed molecules has to be analysed. With decreasing coverage (which we were able to measure downward to a value of N x 0.1 molecule per cavity) the observed 13C n.m.r. shift increases strongly, while for large coverages ( N > 3) it approaches the value for CO in the gaseous state (fig. 1). This behaviour is typical for an exchange between physisorbed (P) and chemisorbed (C) species being characterized by different resonance shifts dp and d,, respectively.The task is to extract the difference 6, - B p and the number of complexed molecules by treating the complex formation as a chemical reacti0n.l. 2? Two models will be considered. In Model 1 the decomposition of the complex is only determined by its equilibrium constant Kg. If we denote by P a physisorbed molecule, by A an active site (total number NA) and by C a surface complex (total number Nc), the equilibrium is described by (1) P + A e C K* from which the equation3056 N.M.R. of CO in Zeolites i 0.5 Y 0 200 - 100 - A 1 I I 1 2 3 N Fig. 2. Dependence of the n.m.r. shift on coverage for the system C0/30DeNaY at 303 (0) and 273 K (A) (b) and plot of the experimental quantities y us. x [according to eqn (9)] (a). Table 1. The resonance shift, 6,, number, NA, of active sites per cavity in zeolite 30DeNaY and the equilibrium constant, Kg, calculated from the experimentally determined quantities Am (= 6,/6,) and yoa using Model 1 Kg __ T / K Am Yo 6, (PPm) N A _ _ ~ - 300 0.96k0.02 0.08k0.01 310+20 0.07&0.01 400f300 273 0.85f0.02 0.10&0.01 350+30 0.07f0.01 78f25 a yo = (x = 0) from the plot of the experimental quantities y us.x [cf. eqn (9)]. can be derived. Here N - N , is the number of physisorbed molecules which are not bound in complexes and NA - N , is the number of unoccupied active sites. For a fast exchange and N 2 N , the observed n.m.r. shift is given by iVc N-N, N 6 = --6,+N6p. If we refer the n.m.r. shifts to the physisorbed state (6, = 0) we find Combining eqn (2)-(4) we obtain the formula (3)A .Michael et al. 3057 In Model 2 it is assumed that a decomposition of the complex occurs only when an unoccupied site (denoted as F) for physisorption exists. If we denote the total number of physisorption sites by NPh and the number of unoccupied sites by N, = NPh + N, - N we may derive from the equilibrium the equation KL P + A e C + F l-NPh+KLNA KL-l A- - +--(l -A). KL NA KL NA (7) As can be seen from eqn ( 5 ) and (7), a decrease of the coverage (expressed by the shift value of N) does lead to a maximum value for the n.m.r. shift S(N -+ 0) = 6, which, however, must not be equal to 8,. By means of eqn ( 5 ) and (7), we find for S,/& = Am Am = KgNA (Model 1) 1 +KgNA Am = KL(NA/NPh) (Model 2) + KL(NA/NPh) and hence the final equations I X 1 y = --NA- (Model 1) Am A& X 1 Am KL-1 A& y = - - - KL N A P (Model 2). A plot of the experimental quantity y = N(S/6m)2 (1 -S/Sm)-l us.x = N(6/Sm)(1 -S/S,)-l allows the determination of Am and NA or NA[KL/(KL- l)]. Thus, in general, independently of the model used an unambiguous determination of both the n.m.r. shift 6, for the complex and the number NA of active sites is only possible if K , & 1 holds. Both models were applied to analyse the experimental results for the n.m.r. shifts of the system C0/30DeNaY shown in fig. 2. The following conclusions can be drawn.6 (1) In the temperature range 273-300 K, the condition KL $ 1 holds and thus Model 1 is valid. This conclusion results from the relation [cf. eqn (S)] : KL=----- Am NPh 1-Am NA taking the values for Am from table 1 and considering that the number of active sites is much less than the number of sites for physisorption.(2) According to the values derived for the equilibrium constant Kg and for NA (cf. table l), the CO molecules undergo a strong interaction with a small number of active sites, viz. 0.07 centres per cavity. The relatively large uncertainty for the value of Kg is due to the rather large error (up to 50%) for the term Am/(l -Am). (3) For T < 273 K the condition of fast exchange is no longer fulfilled. Temperature Dependence For T < 180 K the 13C n.m.r. shifts are only small (-4 to -6 ppm) and to higher magnetic field with respect to CO in the gaseous state. The values are typical for physisorption and may be explained in terms of van der Waals interactions [cf.also ref. (3)]. With increasing temperature the resonance lines are shifted to lower magnetic field (positive values of S) and broadened [fig. 1 (b) and 31. Similar to the dependence of the3058 N.M.R. of CO in Zeolites t t I L B 3 4 5 6 Fig. 3. Dependence of (a) the 13C n.m.r. shift and (b) the linewidth, A V , , ~ , of the system CO/30DeNaY on temperature. The symbols and A denote experimental values and theoretical data calculated by the formulae of Swift and Connick, respectively. The coverage is 0.5 molecules per cavity. resonance shift on the coverage, this behaviour may be explained by an exchange process, viz. a fast exchange at room temperature and above and a slow exchange at low temperatures, where the influence of the active centres cannot be detected.Since we know from the analysis in the preceding section that the overwhelming majority of molecules is physisorbed (shift 6,) and that only a very small fraction N , 4 N p is involved in complexes with large resonance shifts 6,, the formula of Swift and Connick7 can be applied to analyse the experimental n.m.r. shift 6 and half linewidth Av: Here z denotes the mean residence time of a molecule in region C and < its transverse relaxation time. Avp is the half linewidth in the physisorbed state. For fast exchange occurring at higher temperatures the conditions z 4 T, and TAU, 4 1 are valid, leading to the simple equations: 1 N , 1 6 = ( N , / N ) S , and Av = Av,+---. n N T ,A . Michael et al. At lower temperatures, i.e. for slow exchange (z % TJ, and the linewidth 1 N , 1 AV = A v ~ + - - - 7 r N z is only determined by Av,.For the intermediate 3059 the resonance shift 6 goes to zero (14) region the formulae predict a monotonic increase of-the n.m.r. shift 6 with increasing temperature and a maximum for the linewidth Av as a function of z/&. To describe the temperature dependence of the 13C n.m.r. shift 6 and the linewidth Av for the system C0/30DeNaY400DB with a coverage of N = 0.5 molecules per cavity (fig. 3) we use the values N,/N = 1/6.25 and 6, = 350 ppm, derived from the analysis of the coverage dependence (table 1). The linewidths Avp for the physisorbed molecules are taken from the resonance lines at low temperatures. For simplification we assume that Avp is constant within the temperature interval considered.From the fit of the linewidths and the resonance shifts us. temperature we obtain an Arrhenius plot for the lifetime z and for the transverse relaxation time with activation energies of 18 and 10.5 kJ mol-1 and pre-exponential factors 1.5 ns and 10 ms, respectively. The agreement between the experimental and theoretical curves (fig. 3) is relatively good at low and high temperatures. Deviations up to 25% occur at medium temperatures. The decrease of the n.m.r. shifts at T > 300 K, which is still more pronounced for samples with a higher temperature of pretreatment, cannot be explained in terms of this exchange model. Adsorption Centres for CO According to the interpretation of the 13C n.m.r. shifts, at room temperature a fast exchange between physically adsorbed CO molecules and those which are strongly adsorbed on a very small number ( N , < 0.07 per large cavity for 30DeNaY 400DB) of adsorption sites occurs.It has been suggested2p3 that these latter are Lewis-acid sites (preferentially extra-lattice A13+ ions) which are formed during thermal treatment of decationated zeolites. In accordance with this suggestion we have found6 a correlation between the number of these sites and the magnitude of the 13C n.m.r. shifts. The shifts increase with rising the treatment temperature for zeolites DeNaY and are appreciably smaller for stabilized decationated zeolites with a strongly reduced aluminium content .8 For those stabilized decationated zeolites where extra-lattice aluminium species were completely rern~ved,~ no significant 13C n.m.r.shifts could be observed for the adsorbed CO (6 = +0.5 ppm relative to gaseous CO for N = 1.8). A similarly small shift was also found for CO adsorbed on a special aluminium-free zeolite Y prepared by means of Beyer’s method.1° Furthermore, we have studied zeolites where Na+ and H+ ions at cationic sites were replaced by A13+ ions (formula for the unit cell: A~,,Na,H,A~,,Si,,,O,,,, x + y = 23, ratio Si/Al = 2.42 for the skeleton). Here similar resonance shifts as for CO adsorbed on heat-treated zeolites DeNaY were measured. Appreciable resonance shifts were also observed for H mordenites and zeolite HNaZSM-5 having A13+ ions at extra-lattice sites. In spite of these measurements, which strongly recommend us to identify the adsorption centres responsible for the large resonance shifts of adsorbed CO, with A13+ ions on extra-lattice sites there are two facts which deserve further consideration.(i) The number of extra-lattice aluminium ions determined by 27Al n.m.r. measurements11 (1.1 ions per large cavity for 80DeNaY 400DB) is at least by a factor of about 5-10 larger than N,. This discrepancy may be explained by the well known experimental fact that the number of strong Lewis-acid sites acting in catalytic reactions is much smaller than3060 N.M.R. of CO in Zeolites 10’: 1 lo-’ rA 1 L- 1 o-2 1 o - ~ . l n 1 m 1 m 1 m l m 1 m 1 3 4 5 6 7 8 9 1 0 3 KIT Fig. 4. 13C nuclear magnetic relaxation times, ( I = 1,2), of (a) carbon monoxide and (b) carbon dioxide adsorbed on zeolites NaX (0, T,; 0, T,) and NaY (m, T,; n, T,) as a function of temperature.The coverage is 2 molecules per cavity.A . Michuel et al. 306 1 1 Id m 1 h- 1 0-: 1 . ;@’ , / / / /’O / ;6 / ,o’ 3 4 5 6 7 8 9 103 KIT Fig. 5. 13C nuclear magnetic relaxation times, (I = I , 2), of carbon monoxide (8, T,, N = 6 ; 0, T,, N = 1.2; 0, G, N = 1.2) and of carbon dioxide (0, T,, N = 2; 0, T,, N = 0.7; 0, q, N = 2) adsorbed on zeolite SODeNaY as a function of temperature. the total number of extra-lattice aluminium ions. (ii) Adsorption shifts of ca. 300 ppm relative to gaseous CO are very large compared with known coordination shifts in diamagnetic systems. Hence, it seems reasonable to consider also a possible influence of a small number of paramagnetic extra-lattice sites (such as Fe3+), which could be formed simultaneously with the A13+ sites owing to the presence of paramagnetic impurities in the zeolites used.13C Nuclear Magnetic Relaxation According to the preceding analysis, the strong 13C n.m.r. shifts of CO molecules adsorbed on decationated zeolites are caused by interactions with a small number of active sites ( N , < 0.07 per cavity). The remaining question is whether Lewis-acid centres or paramagnetic impurities at extra-lattice sites are responsible for this behaviour. In the following investigation 13C nuclear magnetic relaxation time measurements will be used to solve this question. and T, of CO and CO, molecules adsorbed on zeolites NaX and NaY are shown as a function of temperature In fig. 4 the longitudinal and transverse relaxation times3062 N.M.R. of CO in Zeolites Table 2.The minimum of the 13C nuclear magnetic relaxation time, T,, and the ratio of TIT, for different interaction mechanisms of CO interaction r/nm (Wrdmin Tmin/S (TI Qmin - ~~ ~ . _ _ _ _ _ _ _ ~~~ ~ 013c.. .13co 0. I 5 (0.2) 0.616 1.68 (9.4) 1.60 27~13+. . .13co 0.206” 26.96 0.278 4.42 23Na+. . . 13C0 0.261 19.0 3.7 4.42 Fe3+. . . l3COC 0.2 1 4.1 x 1.83 CSAd (A0 = 406 ppm)e 1 0.640 3.67 a Ref. (4). Ref. (1 5). Fe paramagnetic centre. CSA = chemical shift anisotropy. Value for free CO at 4.5 K [ref. (16)]. between 400 and 11 5 K. For lower temperatures (T < 210 K), the longitudinal relaxation time shows a minimum, while the transverse relaxation time increases monotonically with rising temperature.At higher temperatures (T > 170 K for CO and T > 300 K for CO,) both relaxation times q and T, decrease. For CO molecules adsorbed on decationated zeolites the 13C nuclear magnetic relaxation times q and T, increase with decreasing temperature in the total interval of measurements. For adsorbed CO,, maxima occur at intermediate temperatures ( T z 180 K for and T z 150 K for T,) (fig. 5). If we compare the temperatures for which the longitudinal relaxation time exhibits its minimum, we find that the mobility of the adsorbed CO, increases in the order NaX + NaY + DeNaY. An analogous statement can be drawn for the adsorbed CO molecules. Obviously, the mobility in decationated zeolites is so high that the minimum of the longitudinal relaxation time is shifted to lower temperatures, outside the interval of measurements. A decisive question for an analysis of the 13C nuclear magnetic relaxation times is which of the possible interaction mechanisms dominates.In table 2 we summarize theoretical results for the contributions q! of various interaction mechanisms to the total longitudinal relaxation rate 7;-l and for the respective ratios T,,/qi at the minima The discussion includes the magnetic dipolar interaction among the 13C nuclei, between the 13C and 27Al or 23Na nuclei or paramagnetic impurities (Fe3+) and the interaction due to an anisotropy of the chemical shift (CSA). In spite of the small amount of paramagnetic impurities (ca. 550 ppm Fe3+ ions for the zeolites X and Y used) this contribution dominates. The influence of chemical shift anisotropy and of the magnetic dipolar interaction between different nuclei is negligible.For instance, the CSA contribution could only be significant if the anisotropy were larger than 1700 ppm, i.e. ca. one order of magnitude higher than for CO-iron(r1r) porphyrin complexes,12 where the chemical shift anisotropy has been found to be dominant for the 13C nuclear magnetic relaxation of carbon monoxide. For I3C nuclear magnetic relaxation processes due to interactions with paramagnetic impurities, the measured relaxation times q (1 = 1 longitudinal, I = 2 transverse) are proportional to the relative amount of CO molecules, N,m/N, near the paramagnetic of Ti. sites: Here NI and N denote the number of paramagnetic ions and the total number of CO molecules per cavity, respectively, and m is the number of CO molecules in the first coordination sphere.The quantities q(ion) denote the relaxation times of a single 13C nucleus near a paramagnetic ion. It is interesting to note that the values of Tmin are about the same for CO and CO, adsorbed on zeolites NaX and NaY (cmin z 40 ms)A , Michael et al. 3063 Table 3. Results of adsorption isotherm measurementsa CO,/NaX 8.2 298b 2.0 0.3 0.1 CO,/NaY 8.1 29gb 2.0 3.3 1.5 CO/NaX 7.1 1 98b 1.7 0.9 0.9 CO/NaY 8.4 273" 2.0 53.3 31.0 CO/SODeNaY 1.7 20Od 1.2 4.5 1.4 CO/SODeNaY 1.7 273d 1.2 59.4 13.5 a The symbols have the following meaning: free sample volume, V,; temperature, T ; total number, N , of molecules per cavity of the given zeolite; adsorbate pressure, p , in the sample and amount of desorbed molecules, N,,,/N.Ref. (17). Ref. (18). Ref. (19). if the same coverage is chosen. Since the number of impurities is comparable for both zeolites, we can conclude that the time constants q(ion)min are the same for CO and CO, and hence that the distances between the 13C nuclei and the paramagnetic ion should be the same. For decationated zeolites the relaxation time qmin of CO, is much greater than for zeolites NaX and Nay. According to fig. 5 a value of Tmin = 800 ms has been found near the minimum at 125 K. Further characteristic properties to be seen in fig. 5 are the large value for the ratio of q/T, near the minimum of q and the decrease of both relaxation times and T, with increasing temperature.The latter behaviour can be explained either by an increasing desorption of the molecules at higher temperatures or the onset of an exchange process. In the case of desorption, both relaxation times should decrease according to eqn (1 5 ) with decreasing N . To check this influence quantitatively, the number of desorbed molecules was evaluated by the aid of adsorption isotherms (table 3).6 The reduction factors are only small for CO, in zeolites NaX and Nay, which is in qualitative agreement with the relatively small decrease of the relaxation times and T, for these zeolites at higher temperatures [cf. fig. 4(b)]. decreases from 46 ms at 200 K to 14 ms at 273 K, which corresponds to a factor of 3.3. The fraction of carbon monoxide desorbed, however, is only 0.3.The discrepancy becomes even more apparent for CO adsorbed on decationated zeolites, where thr; decrease of the relaxation times by more than one order of magnitude between 200 and 273 K must be compared with the desorption which amounts here to ca. 15%. The large values for the ratio of q/T, and the decrease of both relaxation times and T, at higher temperatures for CO and CO, adsorbed on decationated zeolites can be explained by molecular exchange between physically adsorbed molecules bound to a small number of extra-lattice aluminium (subscript ea) and iron ions (subscript ei) which are created during the stabilization of decationated zeolites under self-steaming conditions. The temperature dependence of the linewidth Av which is proportional to G1 has been already explained on the basis of the Swift-Connick formula [c$ eqn (12)] and hence for the present case we have to write For CO adsorbed on NaY the relaxation time3064 N.M.R.of CO in Zeolites T-’/s-’ Fig. 6. Schematic representation of the influence of different sites on the 13C nuclear magnetic relaxation times, 7 (l = 1,2) as a function of temperature. Details are described in the text. For the longitudinal relaxation time it follows [by placing Amj = 0 and exchanging 2 with 1 in eqn (16)]:13 1 where p. = - * 4 NP The decrease of the transverse relaxation time T, with increasing temperature can be explained if in this temperature interval q is controlled by the residence times z,, and zei of the molecules at A13+ and Fe3+ ions, respectively. This implies that the relaxation times qj and/or the inverse values of chemical shifts for molecules at the respective extra-lattice sites (Amj)-’ must be smaller than the residence times zj.This condition corresponds to the case of slow exchange: at lower temperatures the contribution G& dominates. Its value is given by magnetic dipolar interactions with paramagnetic impurities of the zeolite skeleton. Because of the slow exchange, paramagnetic and aluminium extra-lattice ions are ineffective. The contribution of the different mechanisms to the resulting relaxation times is shown schematically in fig. 6. Owing to the characteristic properties of the relaxation times qei and Tea, the resulting longitudinal and transverse & relaxation times are predominantly determined by interactions with extra-lattice paramagnetic (Fe3+) and extra-lattice diamagnetic (A13+) ions, respectively.As was already mentioned, q1 is proportional to the n.m.r. linewidth Av, so that this interpretation is in agreement with that given above for the observed temperature dependence of Av. It is of special importance that in contrast to the quite different temperature dependences for the resonance shifts of CO and CO, the behaviour of their relaxation times is qualitatively the same as can be seen by comparing fig. 4 6 . Obviously, the typical features of the nuclear magnetic relaxation times of CO, adsorbed on decationated zeolites can only be understood if a specific interaction of CO, molecules with extra-lattice sites also occurs which, however, is not reflected in the 13C n.rn.r.shifts. This behaviour allows the following conclusions.A . Michael et al. 3065 Table 4. Comparison of the changes of the electronic charges, Ap, for CO and CO,, linearly attaching an A13+ ion with respect to the isolated moleculesa and theoretical values of the 13C n.m.r. shift complex APcle Apole APAlle 8 (PPmIb ~- ~- ~ 1 3 + . . . oc - 0.847 0.631 0.216 73.1 ~ 1 3 + . . . co 0.006 - 0.346 0.340 - 80.8 ~ 1 3 + . . . oco -0.265 0.488, - 0.41 6 0.192 - a Electronic charges of isolated molecules: C, 5.844; 0,8.156 for CO. C , 5.192; 0,8.404 for CO,. Positive 6 are to lower field. Table 5. Reaction coordinate, R, stabilization energy, and theoretical resonance shift, 6,b calculated by the CND0/2 methodc complex Rjnm AE/kJ mol-l 6 (ppm) o=c=o H+ o=c=o HS o=c=o 0.301 1 0.3323 0.3742 0.1110 0.1049 0.3776 0.3820 0.1177 0.1071 - 127.1 - 65.9 - 17.4 - 1 184.0 - 906.5 - 18.2 -21.3 - 698.9 - 1016.8 - 35.4 - 25.2 + 0.4 - 74.3 - 34.1 - 0.6 - 1.2 - 2.9 -7.1 - a Stabilization energy AE = Ecomplex - ECO,COp.Theoretical resonance shift 6 = ZC0,C02-~complex; u = + ( 2 ~ ~ ~ - 0 ~ , ) . The interatomic distances (0.1 191 and 0.1230 nm for CO and CO,, respectively) were also optimized by CND0/2. - - Conclusions Influence of Extra-lattice Paramagnetic Sites (i) Extra-lattice paramagnetic sites (e.g. Fe3+ impurities) act as relaxation agents and exert approximately the same influence upon the longitudinal nuclear magnetic relaxation times & of CO and CO, in decationated zeolites. According to Johnston and Grant,14 in this case the longitudinal electron spin relaxation time zlS must be larger than the residence time zei of the molecule at the paramagnetic ion.Since the thermal correlation time z, of qei must be equal to or smaller than zei, the effective correlation time, which is given by (q$ + q1)-l,13 cannot be determined by the electron spin relaxation time. In agreement with this conclusion,3066 exhibits a dependence on temperature which is characteristic for thermally activated molecular motions. (ii) If we assume that in contrast to (i) the interaction with extra-lattice paramagnetic sites would lead to an appreciable n.m.r. shift, then its value should be about the same for CO and CO, molecules. This follows from the fact that the local magnetic fields due to the paramagnetic sites must be approximately the same at the carbon nuclei of CO and CO, molecules in accordance with the similar behaviour of their nuclear magnetic relaxation times.Hence this interpretation is clearly ruled out by the experimental result that for CO, only very small values for the n.m.r. shifts could be observed. N.M.R. of CO in Zeolites Interpretation of the N.M.R. Shifts The preceding discussion suggests explaining the large n.m.r. shifts observed for CO molecules adsorbed on decationated zeolites in terms of a change of the electronic charge at the carbon atom due to an interaction with extra-lattice aluminium ions.* To understand why in the case of CO, molecules only very small values for the 13C n.m.r. shifts could be observed, quantum-chemical calculations for complexes of CO and CO, with A13+,4 Na+15 and H+ ions6 were undertaken.Ab initio calculations of CO and CO, complexes with A13+ ions (table 4) revealed that the change of the electronic charge at the carbon atom is large if the CO molecule is attached to the A13+ ion via its oxygen atom, while for CO, only a smaller change could be found. These statements could be confirmed by means of ab initio calculations for the systems Na+/CO, Li+/C0l5 and by semi-empirical CNDO calculations for CO and CO, on different ions.6 The general trend is that in the case of CO, relatively small influences on the electronic density at the carbon atoms appear. Additional calculations of the 13C n.m.r. shifts revealed that for the systems Na+/CO, and H+/CO, nearly no shifts result, in contrast to pronounced shifts for similar complexes of CO depending on the special arrangement (table 5).As described el~ewhere,~ ab initio calculations yield for A13+/C0 complexes strong 13C n.m.r. shifts to lower magnetic field only if the carbon monoxide molecule is attached to the ion uia its oxygen atom. For the opposite structure, where the carbon atom is attached to the aluminium ion, the 13C n.m.r. shift is to higher magnetic field. This behaviour may explain why at higher temperatures the experimental resonance shift decreases again: owing to the increased thermal energy of the molecules the occupation numbers for both structures become approximately equal and the 13C n.m.r. shifts are averaged to zero. References 1 V. Yu. Borovkov, A. V. Zaiko, V. B. Kazansky and W. K. Hall, J. Catal., 1982, 75, 219. 2 H. Pfeifer, W. Meiler and D. Deininger, NMR of Organic Compounds Adsorbed on Porous Solids, Ann. Reports on NMR Spectroscopy (Academic Press, London, 1983), vol. 15, p. 291. 3 A. Michael, W. Meiler, D. Michel and H. Pfeifer, Chem. Phys. Lett., 1981, 84, 30. 4 Th. Weller, W. Meiler, A. Michael, H-J. Kohler, H. Lischka and R. Holler, Chem. Phys., 1982,72, 155. 5 Th. Bernstein, D. Michel, H. Pfeifer and P. Fink, J. Colloid Interface Sci., 1981, 84, 310. 6 A. Michael, Dissertation A (Karl-Marx-Universitat, Leipzig, 1984). 7 T. J. Swift and R. E. Connick, J. Chem. Phys., 1962, 37, 307. 8 V. BosaEek, D. Brechlerova and M. KEivBnek, Adsorption of Hydrocarbons in Microporous Adsorbents 9 V. BosaCek, V. Patzelova, Z. Tvarbzkova, D. Freude, U. Lohse, W. Schirmer, H. Stach and 10 H. Beyer and I. Belenykaja in Catalysis by Zeolites, ed. B. Imelik et al. (Elsevier, Amsterdam, 1980), 11 D. Freude, T. Frohlich, H. Pfeifer and G. Scheler, Zeolites, 1983, 3, 171. 12 T. Perkins, J. D. Satterlee and J. H. Richards, J. Am. Chem. Soc., 1983, 105, 1350. 13 H. Pfeifer, NMR Basic Principles and Progress (Springer-Verlag, Berlin, 1972), vol. 7, p. 53. IZ (Academy of Science of the GDR, Eberswalde, 1982), vol. 2, p. 26. H. Tharnm, J. Catal., 1980, 61, 435. p. 203.A . Michael et al. 3067 14 E. R. Johnston and D. M. Grant, J. Magn. Reson., 1982,47. 282. 15 Th. Weller, W. Meiler, H. Pfeifer, H. Lischka and R. Holler, Chem. Phys. Lett., 1983, 95, 599. 16 J. W. Gleeson and R. W. Vaughan, J. Chem. Phys., 1983,78, 5384. 17 D. W. Breck, Zeolite Molecular Sieves (Wiley-Interscience, New York, 1974). 18 T. A. Egerton and F. S. Stone, J. Chem. SOC., Faraday Trans. 1 , 1970, 66, 2364. 19 A. Michael, D. Michel and H. Pfeifer, Chem. Phys. Lett., 1986, 123, 117. Paper 511452; Received 21st August, 1985

ISSN:0300-9599    

DOI:10.1039/F19868203053

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. I, 1986, 82, 3069-3080 Thermal Properties, Thermochemistry and Kinetics of the Thermal Dissociation of Hydrochlorides of some Mononitrogen Aromatic Bases Jacek Lubkowski and Jerzy BlaBejowski" Institute of Chemistry, University of Gdansk, 80-952 Gdansk, Poland Thermal properties of the hydrochlorides of several nitrogen aromatic bases have been elucidated via thermoanalytical investigations. All the compounds studied undergo decomposition upon heating, leading to their total volatilization. The dissociation process proceeds in two stages having different kinetics. The existence of these stages is explained on the basis of the Jacobs and Russell-Jones model of the kinetics of dissociative sublimation processes. The enthalpies of thermal dissociation of the salts have been evaluated from the non-isothermal thermogravimetric curves, and these values, together with thermochemical data available in the literature, have been used to evaluate the enthalpies of formation and the crystal lattice energies of the compounds.The ' thermochemical' radii for appropriate cations have been determined from the Kapustinskii-Yatsimirskii equation. The influence of the structure of the nitrogen bases on the thermal behaviour of their hydrochlorides is also discussed. Mononitrogen aromatic bases (i.e. pyridine, quinoline, acridine and their homologues and derivatives) (denoted by B) interact in condensed phases with hydrogen chloride forming appropriate monohydrochlorides. Little attention has been devoted to the examination of properties of these compounds, possibly because simple representatives of this series are highly hygroscopic.Neither has much attention been paid to the thermochemistry of these derivati~es.~-~ This paper deals with the latter problem. We undertook investigations on this topic for several reasons. First, considerable the~retical~-~ and experimental8, effort has been made recently to explain the nature of interactions between HCl and nitrogen present in organic molecules. However, these investigations mainly concerned amine-HC1 systems in both the gaseous6' 7 7 9 f lo and condensed phases,8$ 11, l2 less attention having been paid to the case of aromatic molec~les.~~ 11-16 In a previous study of amine-HC1 systems we illustrated the usefulness of thermoanalytical investigations,17 and we believe that in the case of the hydrochlorides of aromatic bases many questions can also be resolved by the application of these methods. Secondly, such studies should provide more information regarding the behaviour of highly unsymmetrical ions.Thirdly, investigations of the thermal dis- sociation of alkylammonium chlorides17 and hexachlorostannatesl* revealed that the non-isothermal kinetics of the process differs significantly from that characteristic of typical chemical reactions. By extending the study we expected to gather more information on this phenomenon. Lastly, a knowledge of the thermochemistry of these compounds is of considerable practical importance.2* 1 3 9 l9 Experimental All reagents were of the best available grades and were used without further purification.The hydrochlorides were prepared by passing a stream of dried HCl through a solution of the appropriate aromatic base in anhydrous ethyl ether.l3? 2o The resulting precipitates 30693070 M - W * 60- E 2 80- 2 .9 ; 100- m - Thermochemistry of Hydrochlorides of some Nitrogen Bases - (A) , I I l 1 I _ I TIK Fig. 1. Thermal analyses of quinolinium chloride (A) and acridinium chloride (B). (a) Differential thermal gravimetry, (b) differential thermal analysis and (c) thermal gravimetry. were filtered and thoroughly washed with ice-cold, dry ether. Most compounds were recrystallized from ethanokthyl ether or isopropyl alcohol-ethyl ether mixtures. The salts were dried in a vacuum desiccator over P205. The compounds were checked for purity by elementary analysis and, in some cases, by mercurometric determination of chloride ion.The purity of the compounds examined was always better than 99%. The thermal analyses were performed on an OD-I03 derivatograph (Monicon, Hungary) with a-A120, as reference, in a dynamic atmosphere of nitrogen. The samples, weighing 45 mg (k 10%) were placed on a platinum plate [see ref. (21), Appendix 1 no. 41. All analyses were carried out at a heating rate of 4.7 k 0.2 K min-l. All the compounds studied, especially the hydrochlorides of pyridine and its derivatives, are highly hygroscopic. Moreover, under certain conditions, compounds of higher HCl contents than that resulting from the formulae B * HC1 can be formed.l9 13, 22-24 To avoid the presence of both H,O and an excess of HCl each sample was heated for 2 h under vacuum (in a vacuum pistol) prior to analysis, just below the temperature of the onset of volatilization.The hydrates of B - HCl and compounds of general formulae B (HCl),, where n > 1, are much less stable than B-HCl and decompose under these conditions.l? 22-24 From the thermogravimetric curves, such as those shown in fig. 1, values of the temperature ( T ) corresponding to certain values of the extent of reaction (a) can be For each compound investigated a set of a us. T data points was determined on the basis of at least three replicate measurements. An example is given in table 1. Results and Discussion Thermal Properties of the Hydrochlorides The results of a typical thermoanalytical measurement are shown in fig.1 . To enable further discussion of the thermal properties of the compounds studied we have compiled in table 2 characteristic parameters resulting from an examination of the thermal analysis curves together with available literature information.J . Lubkowski and J. Blazejowski 307 1 Table 1. a vs. T for the first step of volatilization of quinolinium chloride (A) and acridinium chloride (B) 0.10 0.18 0.26 0.34 0.42 0.50 0.58 0.66 0.74 403.8 414.1 421.1 426.4 430.7 434.6 438.0 440.9 443.9 450.5 462.0 468.7 474.3 478.6 482.0 484.9 487.8 491.0 All the compounds examined undergo total volatilization upon heating. The thermo- gravimetric curves demonstrate, however, that the process exhibits a complex nature. Two distinct stages are always seen in the curves.The first step, in which up to 88% of the sample is volatilized, is followed by a slow process, seen as a ‘tail’, in which the remains of the sample disappears. One possible explanation of this effect might be that the organic fragments of the molecules undergo destruction at high temperatures. To verify this hypothesis we heated the compounds until the first step was complete. After cooling, the samples were subjected to standard chemical analysis. The composition of the remaining substance was always the same as that of an original sample. Therefore this explanation fails. Another possible explanation could come from the mechanism of the volatilization process. We will discuss this issue subsequently. Careful analysis of the thermal analysis curves reveals that only for pyridinium and quinolinium chlorides were additional effects recorded before the onset of the volatilization process.In the case of pyridinium chloride this effect originates from the phase tran~ition,~ while for quinoline hydrochloride it can be ascribed to the fusion of the compound. The peak seen in the thermal analysis curves of quinolinium chloride before the melting point (fig. 1) originates from the loss of water. This effect always appeared when the sample examined was not subjected to heating before the analysis. For the remaining compounds studied we could not detect any endothermic effects before the onset of decomposition. This is a rather unexpected result, since many authors have reported melting points for the majority of these compounds (table 2).However, a comparison of the literature melting points with values of the peak temperatures in differential thermal gravimetry (d. t.g.) and differential thermal analysis (d.t.a.), as well as values of and reveals that these compounds decompose while melting. Moreover, for some hydrochlorides, i.e. 2,4,6-trimethylpyridine, 2-methylquinoline, isoquinoline and acridine, the reported melting points correspond to temperatures at which the volatilization process is far advanced. Since both processes overlap, it is likely that weak thermal effects originating from fusion2 are masked by much stronger effects resulting from the volatilization process. The above discussion clearly shows that our knowledge of the fusion process of hydrochlorides of aromatic bases is incomplete.On the other hand, this study provides basic information for potential users of these compounds. and &.,, values, gradually increase with an increase in the size of the cation of a given aromatic base, However, this tendency is less pronounced than in the case of the boiling points of the appropriate 44 On the other hand, temperture limits for the volatilization of hydrochlorides (AT) show an insignificant dependence on the size of the cation. Peak temperatures in d.t.g. and d.t.a., as well asTable 2. Thermal properties of hydrochlorides of nitrogen aromatic bases 3 2 parameters 8 5 0 peak temperaturea/K describing the onset of the G- d.t.a. W K second stage -& 4 % Q 1 pyridine* 425 427 355;279 414;4 413-415;'' 416.5;" 417;30-32 382.6 424.0 41.4 0.88 431 L 2 2 3-methylpyridine 420 421 338-348 ;35 353 ;36 35637 371.4 416.4 45.0 0.87 421 0 F g 5 3,4-dime th ylp yridine 458 460 - 417.6 457.9 40.3 0.86 465 b 3 10 isoquinoline 456 458 466;48 482;48 493-49Y5 417.9 459.7 41.8 0.82 466 3 c1 3 substance d.t.g. T p Tm G.1 To.74 AT cc T/K _____ 417.6;l33 33 419;2 415.3-419.8;34 41842136 3 4-methylpyridine 438 439 366;3e 432;38 434;37 434-43635 391.8 436.4 44.6 0.84 442 4 2,4-dimethylpyridine 450 451 468-470;39 485-486;40 487-487.5;41 48842 409.4 450.5 41.1 0.84 459 6 3 +dimethylpyridine 447 448 - 407.4 447.0 39.6 0.84 458 7 2,4,6-trimethylpyridine 469 472 498;43 545-54740 429.7 472.4 42.7 0.82 479 8 quinoline 445 448 371 ;" 387-390;35 400;" 406;45 407.6'** 46 403.8 443.9 40.1 0.85 453 9 2-methylquinoline 462 465 493-494;35 49747 425.7 462.3 36.6 0.84 471 b 0 11 acridine 486 489 493.6-495;49 506-50835 450.5 491.0 40.5 0.83 500 2 6 2 ~ _ _ _ a The symbols were taken from ref.(26). is equal to a, and AT = = temperature of the peak, Tm = temperature of melting, T, = temperature at which the fraction reacted G1. * Temperature of phase transition for pyridinium chloride = 355 K [345 K from ref. (4)]. From this work.J . Eubkowski and J . Btatejowski 3073 Nature of the Volatilization Process Numerous investigations have proved that the hydrochlorides of nitrogen aromatic bases behave in the condensed phases a typical ionic substances.*~ 5 7 1 4 7 1 5 9 29 Hence, the transfer of molecules from the condensed to the gaseous phase must involve several stages.Numerous similarities in the thermal behaviour of the compounds studied and of ammonium and alkylammonium saltsl7t 50-52 show that the mechanism of the volatilization process is essentially the same for both groups of compounds. Therefore, we present below a reaction scheme for the process which accounts satisfactorily for all the observed experimental facts : (1) (11) (111) BH+Cl- + B -HC1 c B(ads) + HCl(ads) B(g) + HCl(g). (1) This scheme involves three stages in the thermal dissociation process, namely : (i) proton transfer in the ion couple being at a position of a half crystal site (I) which creates the transition state (11), (ii) dissociation of the molecule in the transition state followed by the formation of a mobile adsorbed phase (111) and (iii) desorption of adsorbed molecules from the surface.The migration of the molecules over the surface followed by their diffusion through the gas phase represents the sublimation process. The mechanism of the process shown in reaction (1) does not consider the behaviour of dissociation fragments in the gas phase. It seems probable that these fragments behave as kinetically free molecules; however, weak interactions between some of them cannot be excluded.69 7 9 9 9 lo, 53 Enthalpy of Volatilization The enthalpy of volatilization (AH,) was evaluated on the basis of the Van? Hoff equation. The dissociation of 1 mol of B-HCl results in the formation of 2 mol of gaseous products. On the basis of our previous c o n s i d e r a t i o n ~ , ~ ~ ~ ~ ~ AH, can thus be derived from the equation (2) In a = -w2RT+constant AHV where R is the gas constant. The main premise of this method is that the system reaches equilibrium at a given temperature, T.This implies that the only energy barrier existing for the process is a thermochemical one, i.e. AHv. Under these conditions the experimental degree of conversion, a, is equal to PIP,, where P is the equilibrium vapour pressure at a given T, and Po is the atmospheric presaure. In following the above assumption the thermal analyses were carried out in such a way as to eliminate, as far as possible, all parameters which could influence the equilibrium conditions. Thus, the analyses were performed at a moderate heating rate and using as small a sample as possible. Furthermore, the compounds were analysed in thin layers placed on a relatively large surface to enable free diffusion of gaseous products and to avoid side processes.Using experimental a us. T dependences we calculated values for AHv, and these are summarized in table 3. An increase of AHv with both the number of aromatic rings and the number of alkyl groups in the molecule is observed. An exception is the mono- methylpyridines, for which AH, values are slightly lower than that for pyridine. On the other hand, the variation of AHv within a given group of isomers depends on specific features of individual compounds. Kinetics of the Thermal Dissociation To examine the kinetics of the volatilization process we invoked the phenomenological theory outlined originally by Jacobs and Rus~ell-Jones.~~ The main premise of this3074 Thermochemistry of Hydrochlorides of some Nitrogen Bases Table 3.Thermochemical and kinetic constants for the volatilization of hydrochlorides of nitrogen aromatic bases from eqn (2)a from eqn (4)b substance E XI x2 /K-k s-4 10-4 K-1 (no. from A 4 table 2) /kJ mol-l constant /kJ mo1-1 1 2 3 4 5 6 7 8 9 10 11 132 115 129 150 158 153 160 149 178 156 184 18.4 16.3 17.5 19.7 20.5 20.3 20.1 19.9 22.8 20.1 22.3 129 107 128 154 159 148 169 149 171 159 187 54 1 317 6 810 7 740 2 280 22 200 4 200 58 700 2 720 46 100 23.9 -4.3 -8.1 - 3.0 3. I -0.56 -4.7 7.3 - 0.63 1.1 - 9.6 - 3.0 a The values of AHv and the constant were evaluated using a standard least-squares procedure. The correlation coefficient was always better than 0.9995.Eqn (4) was rearranged to the form a, =flE, XI, X2, &) and the calculations were continued until ( % , o - % , n ) 2 - T (%,0-a,,n+J2 d where n is the number of iteractions. The values of the function Xi(%* - %, n+1)2, which indicate the ‘goodness of fit’, were always lower than 0.00025. method is that the migration of molecules over the surface of a condensed phase followed by their diffusion through the gas phase is the crucial step in the volatilization process. Despite some criticism of this it describes satisfactorily the kinetics of thermal dissociation of ammonium and several alkylammonium chlorides17 1 54 and hexachlorostannates.Is A detailed discussion of this method, as well as its adaptation to non-isothermal conditions, has been presented e1~ewhere.l~~ 51 The final form of the integral kinetic equation, adequate for linearly increasing temperature conditions, is given by (X2 T - t ) [ 1 - ( 1 -a)+]+4[1-(1 -a)#]- 1 Xl = - ~ T! exp (- E/2RT) (3) 3cD where X I and X2 are the constants and E identifies the activation energy for the process, @ is the heating rate, a, represents the initial radius of particles from which the sublimation process occurs and A denotes the distance between collisions (i.e.the distance which a molecule travels after leaving the condensed phase before a collision occurs). At moderate pressures of a foreign gas (ca. atmospheric pressure) A a,, so that the influence of this parameter on the kinetics of volatilization is negligible. Eqn (3) can then be simplified to the form X2T[1-(1-a)~]+~[1-(1-a)%] =--fiexp(-E/2RT) 1 Xl 3 0 (4) which actually describes the kinetics of the first, fast step of thermal dissociation of the compounds studied.Numerical values of the kinetic constants were found using theJ . Lubkowski and J . Blajejowski 3075 Marquardt least-squares optimization procedure55 and are listed in table 3. This method requires the assumption of certain values of constants in the first approximation. These values were estimated by a procedure described e1sewherel7* la which allows one to calculate both XI and X z at a given E. The variation of E is not, however, completely free. We searched for E values which were possibly close to AHv values. Such a procedure ensured that other minima have not been identified.The derived values of the activation energy are very close to those of the enthalpy of volatilization. Also, dependences of both E and AHv on the size and structure of the base cation are similar. These facts indicate that the process is not required to overcome an activation barrier above that resulting from the enthalpy change. They also indicate that the process is realized near equilibrium. X I and Xz parameters can be considered as mathematical constants without true physical significance. Nevertheless, numerical values of both these constants should be positive. Unfortunately, values of X2 are mostly negative. This discrepancy might result from the neglect of two terms containing A/ao, being formally negative, in eqn (4). As pointed out earlier, all the compounds examined undergo thermal dissociation in two distinct stages.The Jacobs and Russell-Jones model also explains this phenomenon qualitatively. As mentioned earlier, the term A/a, in eqn (3) is small in the first step of volatilization and can be neglected. However, at the end of the process the mean free path for the molecules increases in comparison with the geometric surface area of the drops or crystals, and A approaches a, in magnitude. This causes a gradual decrease in the rate of the process, as both terms containing A/ao are negative. A similar phenomenon to that described above has been observed for the volatilization of ammonium and alkylammonium chloridesl7? 54 and hexachlorostannates,l* as well as several non-ionic compounds (e.g. 2-methylquinoline or acridine).However, this phenomenon does not exhibit a general nature, and does not occur during the thermal decomposition of the majority of chemical substances. Moreover, when the compounds mentioned above, e.g. the alkylammonium chlorides, volatilize in the presence of other solid substances (e.g. PbCl,) this effect may not occur.56 (Note that quaternary alkylammonium chlorides do not exhibit such a although these derivatives may be considered similar to the hydrochlorides of aromatic bases or amines.) In the past several other models have been proposed to describe the kinetics of dissociative sublimation. The reason we applied the Jacobs and Russell-Jones method in this work was that it is, in our opinion, the only one which adequately predicts both the existence of the slow reaction step at the end of the process and the activation-energy values whose magnitudes are physically significant.Using experimental a us. T depen- dences we also calculated Arrhenius parameters from two other equations, namely the Polanyi-Wigner (zero kinetic order) equation and the contracting surface area equation. Both these approaches have often been proposed to describe the kinetics of volatilization proces~es,~~~ 519 5 7 7 58 and they also approximate our experimental data fairly well. Values of the apparent activation energy thus derived are equal to 0.48 and 0.54 of the value of AHv, respectively. Many efforts have been made to explain these low values of E 5 0 7 5 7 9 58 However, it does not seem to be likely that a molecule can reach a high energy level resulting from the enthalpy change for the process if the rate-determining step has such a low activation barrier.If this were the case the process would exhibit a more complex nature. Thermochemical Characteristics Many thermochemical quantities of the compounds studied can be derived from a simple Hess law. The various relations between these quantities are presented in the form of a thermochemical cycle in scheme 1. All magnitudes in the cycle refer to 298 K and atmospheric pressure. AHf denotes the enthalpy of formation of a given substance;3076 Thermochemistry of Hydrochlorides of some Nitrogen Bases Scheme 1. U+2RT is the lattice enthalpy; U represents the lattice energy; A , denotes the proton affinity of the aromatic base and AH, identifies the enthalpy of volatilization.From scheme 1 the following relationships result : A%, c(B * HCl) = AG,g(B) + AG,g(HC1) - AH", AG,g(BH+) = AG,g(B) + AG,g(H+) - A , u" = AG,g(BHf) + AG,,(CI-) - A G , , ( B * HCl) - 2RT. (5) (6) (7) The AHv values derived in this work do not refer to 298 K. Therefore, they should be modified according to the equation PT = A H ~ + A H & + Z A ~ - J AC;dT 298 where AH, is the enthalpy of volatilization derived from eqn (2) and A K n denotes the enthalpy of fusion; the term ZAG arises from any polymorphic transitions which these compounds may undergo between 298 K and the onset of the fusion or volatilization process, while jigs ACp dT accounts for the change in enthalpy resulting from changes in the heat capacities of the reactants.The value of ZAPo for pyridinium chloride was taken from the work of Ripmee~ter.~ On the basis of an empirical relationship relating to the entropy of fusion A q n = AGn/Tm = constant and known values of A G n and T, for pyridinium chloride2 we estimated the enthalpy of fusion for quinolinium chloride to be 9.2 kJ mol-l. We neglect the term A G n for the remaining compounds since they always melt and volatilize simultaneously. The magnitude and sign of the heat-capacity term is more difficult to assess in the absence of C, values for all salts and most gaseous products. During the study of the sublimation process of ammonium chloride we found that ACp can be approximated by the expression 7.0 - 0.13 I T/J mol-1 I C 1 . 1 7 We used this relationship to estimate appropriateJ. Lubkowski and J .Blajejowski 3077 heat capacity changes in the temperature limit 298 K-T,.,. Some idea of the validity of this assumption may be gauged from the fact that the expected increase in heat capacity caused by an increase in the size of an aromatic base should be identical for salts and appropriate bases. The corrected values of A P v are listed in table 4. The values of the enthalpy of formation of appropriate nitrogen bases have been taken from the literature (cf. table 4). Proton affinities listed in table 4 were also taken from the literature and were corrected relative to the proton affinity of NH,, 860 kJ mo1-1.62$ 63 Since the pertinent value of A , for 2,4,6-trimethylpyridine was not available we estimated it from a linear relationship relating proton affinities of methyl-substituted pyridines and free energies of p r ~ t o n a t i o n .~ ~ The necessary data were taken from ref. (66). Using values of AG,g(B) and A , (table 4) together with the values of A P v evaluated in this work (table 4), and taking AG,g(H+) = 1536.2 kJ mol-1,6* A%,,(Cl-) = - 233.1 kH mo1-1,68 and A%,,(HCl) = - 92.3 kJ mo1-1,68 we derived values of the enthalpies of formation of crystalline hydrochlorides, crystal lattice energies and enthalpies of formation of gaseous BH+ ions in the standard state. These values are listed in table 4. To facilitate further discussion of the crystal lattice energy, we invoked an approximate method developed originally by Kapu~tinskii~~ and improved later by Yat~imirskii.~~ According to the authors the crystal lattice energy can be expressed by the equation71 (En) Z K 2, ( 0.0345 U/kJ mol-l = 120.2 1 -___ + 0.087(r, + rA)) Y K + r ~ YK + r~ (9) where (En) is the total number of ions in the simplest formula unit of the molecule, 2, and 2, denote the numerical values of the charges of cation and anion, respectively, and rK and YA are the ‘thermochemical’ ionic radii (in nm).Assuming rcl = 0.172 nm72 we derived values for rK. They are shown in table 4. Final Remarks The experimental method applied in this work for the determination of AHv values is actually a non-equilibrium technique. Furthermore, in using this approach we assume perfect behaviour for all gaseous products. It is thus difficult to estimate the precision of the method directly.This method has also been applied in the evaluation of AHv for NH,Cl and more than 30 alkylammonium chlorides.l79 54 It was thus possible to compare the values derived by this method with those available in the literature. The agreement is satisfactorily considering that values taken from various sources sometimes differ markedly. Generally, this approach leads to lower values of AHv in comparison with those derived from calorimetric measurements, and we believe this to be due to the non-perfect behaviour of the gaseous products. This approach has been successfully used to estimate the enthalpies of the thermal decomposition of several other substances [see ref. (25) and (56) and references cited therein]. The standard enthalpy of formation of pyridinium chloride evaluated in this work is slightly lower in comparison with the literature value, which has been determined by a standard calorimetric method.However, the agreement is good considering that both values have been obtained by the application of totally different techniques. Despite this discrepancy the results of this work undoubtedly support the concept that gaseous reactants predominantly exist as kinetically independent fragments. The enthalpies of formation of hydrochlorides of aromatic bases are much lower than the enthalpies of formation of alkylammonium ch10rides.l~- 54 This mostly results from the fact that all aromatic bases used in this work are thermodynamically unstable. On the other hand, this feature of B - HC1 compounds explains several of their properties, e.g.their relatively low thermal stabilities, their tendency to form hydrates, solvates and compounds of higher HC1 contents than those resulting from the B * HCl stoichiometry, and so on.Y Table 4. Thermal characteristics for mononitrogen aromatic bases, their cations and hydrochlorides at 298 K i3. B * HCl 3 8 3 5 B+ AG, c B /kJ mol-l this u" A%& YK q /nm + * literature /kJ mol-l /kJ mol-l AH", work A%, 298,g 4 substance B /kJ mol-l /kJ mol-1 /kJ mol-1 p yridine 3-met hylpyridine 4-met h ylp yridine 2,4-dimethylpyridine 3,4-dimethylpyridine 3,5-dimethylpyridine 2,4,6-trimethylpyridine quinoline 2-methylquinoline isoquinoline acridine 144.6" 106.4" 102.1U 63.9a 70.0" 72.ga 44.7b 216.7c 201.7" 292.5c - 928.9d,e 938.Y 942.7f 956.lf 953.2f 950.2f 945.6" 945.2e 968.0e 972 - 141 117 132 153 162 156 164 161 182 160 189 - - 88.7 - 103 - 122 - 181 - 184 - 176 -212 - 36.6 - - 50.6 11.2 -98.2' 603 569 580 - 587 599 - 597 583 606 605 61 1 - - - - - - - - - 75 1.9 703.7 695.6 644.0 653.0 658.8 608.9 807.3 792.7 860.7 - z 0.220 5 0.215 g 0.207 q 0.218 5 0.201 ii! 0.202 3 % cs 0.203 0.229 z 0.206 - 2 0.197 as rb Y a Ref. (59). Ref. (60). Ref. (61). Ref. (64). Ref. (65). f Values taken from ref. (66), corrected relative to the proton affinity for pyridine given in table 4. g Ref. (69). E 2J . Lubkowski and J . Biaiejowski 3079 We were surprised with the results regarding crystal lattice energies. Despite the obvious increase in the size of the cation over the series : pyridine, quinoline and acridine on the one hand, and pyridine and methyl-substituted pyridines on the other hand, the observed differences between U values are insignificant. Note also that the variation in U values does not always follow the Kapustinskii-Yatsimirskii equation.The ‘ thermo- chemical’ radii for appropriate cations show also only minor changes. All these facts allow one to conclude that the distances between charged centres in the crystal lattices of all the compounds studied remain principally unchanged, since the lattice energy is affected mostly by Coulombic interactions. Similar regularities have been observed for alkylammonium chlorides and hexachloro~tannates.~~~ 18, 54 A comparison of the values of U from this work with those for alkylammonium chloridesl7~ 54 shows that the latter are always higher.This would indicate that the distances between the charged centres are greater in the crystal lattices of the hydrochlorides of aromatic bases. Further progress in this area would be made by calculating the crystal lattice energies on the basis of the crystal lattice parameters. We are currently working on this problem. We thank Professor J. Szychlinski for valuable discussions and Mrs I. Nikel for experimental assistance. References 1 P. A. Kilty and D. Nicholls, Chem. Ind. (London), 1963, 1123. 2 H. Bloom and V. C. Reinsborough, Aust. J . Chem., 1967, 20, 2583. 3 J. Kuthan, N. V. Koshmina, J. Pdlecek and V. Skala, Collect. Czech. Chem. Commun., 1970,35, 2787. 4 J. A. Ripmeester, Can. J . Chem., 1976, 54, 3453.5 J. Kuthan, S. Boehm and V. Skala, Collect. Czech. Chem. Commun., 1979, 44, 99. 6 Z. Latajka, S. Sakai, K. Morokuma and H. Ratajczak, Chem. Phys. Lett., 1984, 110, 464. 7 A. Brciz, A. Karpfen, H. Lischka and P. Schuster, Chem. Phys., 1984, 89, 337. 8 0. Knop, I. A. Oxton and M. Falk, Can. J . Chem., 1979, 57, 404. 9 C. G. de Kruif, J. Chem. Phys., 1982, 77, 6247. 10 P. Goldfinger and G. Verhaegen, J. Chem. Phys., 1969, SO, 1467. 11 F. Genet, Bull. Soc. Fr., Miner. Crist., 1965, 88, 463. 12 T. L. Gremyachkina, Issled. Obl. Neftekhim., 1976, 25: Chem. Abstr., 1978, 89, 899851.1. 13 M. Goffman and G. W. Harrington, J. Phys. Chem., 1963, 67, 1877. 14 R. Foglizzo and A. Novak, J. Chim. Phys., Phys. Chim. Biol., 1969, 66, 1539. 15 S. E. Odinokov, A. A. Mashkovsky, V.P. Glazunov, A. V. Iogansen and B. V. Rassadin, Spectrochim. 16 J. Szydlowski and M. Zielinski, J. Phys. Chem., 1979, 83, 2122. 17 J. Blazejowski, Thermochim. Acta, 1983, 68, 233. 18 E. Kowalewska and J. Blazejowski, Thermochim. Acta, in press. 19 R. Royer and P. Demerseman, Bull. SOC. Chim. Fr., 1968, 2633. 20 M. D. Taylor and L. R. Grant, J . Chem. Educ., 1955, 32, 39. 21 Atlas of Thermoanalytical Curves, ed. G. Liptay (Akademiai Kiado, Budapest, 1973). 22 F. Ephraim, Chem. Ber., 1914,47, 1828. 23 F. Ephraim and E. Hochuli, Chem. Ber., 1915, 48, 629. 24 V. G. Ovchinnikov, V. I. Kosorotov, I. B. Grebenyuk, L. N. Makovei and R. V. Dzhagatspanyan, Zh. 25 J. Blazejowski, J. Szychlinski and K. Windorpska, Thermochim Acta, 1981, 46, 147. 26 R. C. Mackenzie, J . Therm.Anal., 1981, 21, 173. 27 P. F. Trowbridge and 0. C. Diehl, J. Am. Chem. SOC., 1897, 19, 558. 28 Handbook of Chemistry and Physics, 66th edn. (CRC Press, Florida, 66th edn, 1985/86). 29 D. Cook, Can. J . Chem., 1961, 39, 2009. 30 V. Prey, Chem. Ber., 1941, 74, 1219. 31 D. Klamm, Monatsh. Chem., 1952,83, 1398. 32 D. M. Gruen and R. L. McBeth, J . Inorg. Nucl. Chem., 1959, 9, 290. 33 L. F. Audrieth, A. Long and R. E. Edwards, J . Am. Chem. SOC., 1936, 58, 428. 34 B. E. Tate and P. D. Bartlett, J. Am. Chem. Soc., 1956,78, 5575. 35 S. A. Heininger, J. Org. Chem., 1957, 22, 704. 36 M. Katecka and T. Urbanski, Bull. Acad. Pol. Sci., Ser. Sci. Chim., 1967, 15, 413. 37 E. A. Coulson and J. I. Jones, J . Soc. Chem. Ind., 1946, 65, 169. Acta, Part A , 1976, 32, 1355. Prikl.Khim., 1978, 51, 2312.3080 Thermochemistry of Hydrochlorides of some Nitrogen Bases 38 D. Rostafinska, Roczn. Chem., 1955, 29, 813. 39 F. B. Ahrens and R. Gorkow, Chem. Ber., 1904, 37, 2062. 40 E. M. Gepshtein, Proizvod. Ispol’z. Novykh Koksokhim. Produktov, Vost. Nauch.-Issled. Uglekhim. Inst., Sb. Statei, 1960, 22; Chem. Abstr., 1962, 57, 16546d. 41 E. A. Coulson, J. D. Cox, E. F. G. Herington and J. F. Martin, J. Chem. Soc., 1959, 1934. 42 D. Rostafinska, Roczn. Chem., 1955, 29, 803. 43 W. Shive, E. G. Ballweber and W. W. Ackermann, J. Am. Chem. SOC., 1946, 68, 2144. 44 W. Utermark and W. Schicke, Schmelzpunkttabellen Organischer Verbindtmgen (Friedr. Vieweg and Sons, Braunschweig, 2nd edn, 1963). 45 E. G. V. Percival and W. Wardlaw, J. Chem. Soc., 1929, 1505.46 0. Eckstein, Chem. Ber., 1906, 39, 2135. 47 G. Heller and A. Souvlis, Chem. Ber., 1908, 41, 2692. 48 E. Bergmann and W. Rosenthal, J. Prakt. Chem., 1932, 135, 267. 49 K. Ishikawa, K. Akiba and N. Inamoto, Bull. Chem. Soc. Jpn, 1978, 51, 2684. 50 R. F. Chaiken, D. J. Sibbett, J. E. Sutherland, D. K. Van de Mark and A. Wheeler, J. Chem. Phys., 51 P. W. M. Jacobs and A. Russell-Jones, J. Phys. Chem., 1968,72, 202. 52 C. Guirao and F. A. Williams, J. Phys. Chem., 1969, 73, 4302. 53 L. A. Curtiss, D. J. Frurip, C, Horowitz and M. Blander, Thermal C’onducticity, 1979, 16, 577; Chem. 54 J. Blazejowski and E. Kowalewska, Thermorhim. Acta, submitted for publication. 55 D. W. Marquardt, J. Soc. Ind. Appl. Math., 1963, 11, 341. 56 J. Blazejowski, J. Szychlinski and E. Kowalewska, Thermochim. Acta, 1983. 66, 197. 57 R. D. Schultz and A. 0. Dekker, J. Phys. Chem., 1956,60, 1095. 58 K. Kishore and V. R. Pai-Verneker, J. Chim. Phys., Phys. Chim. Bid., 1977,74, 997. 59 J. D. Cox and G. Pilcher, Thermochemistry of Organic and Organometallic Compounds (Academic Press, New York, 1970). 60 Yu. G. Papulov and P. P. Isaev, Zh. Fiz. Khim., 1977, 51, 1405. 61 S. E. Stein and B. D. Barton, Thermochim. Acta, 1981, 44, 265. 62 H. D. B. Jenkins and D. F. C. Morris, Mol. Phys., 1976, 32, 231. 63 H. D. B. Jenkins and D. F. C. Morris, J. Chem. Soc., Faraday Trans 2, 1984,80, 1 167. 64 D. H. Aue, H. M. Webb and M. T. Bowers, J. Am. Chem. Soc., 1975,97,4137. 65 M. Moet-Ner (Mautner), J. Am. Chem. Soc., 1979, 101, 2399. 66 Gas-phase Ion Chemistry, ed. M. T. Bowers (Academic Press, New York, 1979), vol. 2, chap. 9. 67 D. H. 4ue, H. M. Webb, M. T. Bowers, C. L. Liotta, C. J. Alexander and H. P. Hopkins, J. Am. 68 D. D. Wagman, W. H. Evans, V. B. Parker, R. H. Schumm, I. Halow, S. M. Bailey, K. L. Churney 69 G. 0. Shmyreva and A. M. Mosin, Termodin. Org. Soedin., 1983, 82; Chem. Abstr., 1984, 101, 70 A. F. Kapustinskii, Q. Rev. Chem. Soc., 1956, 10, 283. 71 K. B. Yatsimirskii, Zh. Neorg. Khim., 1961, 6, 518. 72 H. D. B. Jenkins and K. P. Thakur, J. Chem. Educ., 1979, 56, 576. 1962, 37, 23 1 1. Abstr., 1983, 98, 186795~. Chem. SOC., 1976, 98, 854. and R. L. Nuttall, J. Phys. Chem. Ref. Data, 1982, 11, suppl. no. 2. 1991 13w. Paper 511945; Received 5th November, 1985

ISSN:0300-9599    

DOI:10.1039/F19868203069

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

J. Chem. SOC., Faraduy Trans. I , 1986, 82, 3081-3095 Structural Aspects of Metal-oxide-pillared Sheet Silicates An Investigation by Magic-angle-spinning Nuclear Magnetic Resonance, Fourier-transform Infrared Spectroscopy and Powder X-Ray Diffractometry D. Tilak B. Tennakoon, William Jones* and John M. Thomas* Department of Physical Chemistry, Lens$ield Road, Cambridge CB2 1EP Variable-temperature powder X-ray diffractometry, Fourier-transform infrared spectroscopy and 27Al and 29Si magic-angle spinning nuclear magnetic resonance spectroscopy have been used to elucidate both the nature of alumina pillars introduced into sheet aluminosilicates and to monitor changes in structure of the host matrix following the introduction of the pillars. The pillars are neutral oxide columns which are structurally similar to y-alumina; they are linked through oxygen to aluminium and magnesium atoms within the octahedral layer of the clay sheet.The layer charge of the clay is balanced by protons, released during calcination, residing in the aluminosilicate structure. The use of various cation-exchanged natural and synthetic clay minerals as efficient Brmsted-acid catalysts for a variety of organic conversions continues to attract interest.l+ Such clay minerals, e.g. montmorillonite, have a layered structure with individual sheets separated by an interlayer space within which reside charge-balancing, exchangeable cations which are extensively hydrated. Such clay minerals readily expand, and numerous organic reactions catalysed by clays occur within the interlamellar space, with the actual protons provided by the highly polarised hydration spheres of small interlayer cations.Thus the microenvironment provided by the interlayer space is favourable for the organic reactions to proceed. On calcination of a clay mineral, however, collapse of the layer structure occurs and, as a consequence, the interlayer space is virtually eliminated. Therefore spent clay catalysts cannot be regenerated by calcination in air to burn off carbonaceous deposits, and attention has subsequently focused on the so-called pillared clays, prepa~ed~-~ by replacing the interlayer cations with bulky, polymeric, inorganic cations such as [A11,04(0H)24(H20)12]7+ and [Zr4(OH)14(H20),,]z+. These precursor pillared clays are then calcined at ca.5OO0C, resulting in dehydration and dehydroxylation of the polymeric cations. High-surface-area materials are formed (ca. 250 m2 g-1)39 which are stable to temperatures > 500 "C. There has been considerable work aimed at characterising these materials, including pore-volume and surface-area measurement~,~ neutron diffraction,B magic-angle-spinning n.m.r.,7t as well as 13C n.m.r. of adsorbatesgl lo and i.r. studies on the nature of acid sites.ll The nature of the pillars and their relationship to the clay sheets, however, is still not well understood. In addition, the catalytic properties of the pillared clays show a marked dependence on temperature, and the reasons behind this also remain unclear. We report here detailed investigations of the structure of pillared clays, utilising variable-temperature X-ray powder diffraction, 27Al and 29Si magic-angle-spinning n.m.r.(m.a.s.n.m.r.) spectroscopy and F.t.i.r. spectroscopy. In the first part we examine in detail the structure of the pillaring species and the nature of the pillar which is formed during calcination, using well established preparative procedures. We also consider the accompanying proton diffusion which takes place. In the second part we examine the evidence which suggests that a direct chemical link is formed at 500 "C between the pillar and the octahedrally coordinated aluminium or magnesium present in the original clay sheet. 308 13082 Metal-oxide-pillared Sheet Silicates Experimental N.M.R. Studies M.a.s.n.m.r. spectra were recorded on Bruker WH 400 and MSL 400 spectrometers using either a Bruker or purpose-built m.a.s.probe. Samples were packed in Delrin Andrew-Beams type spinners. 27Al spectra were recorded with a typical pulse width of 5 ,us (ca. 30°), spectral width of 80 kHz and a recycle delay of 0.3 s. For the 29Si spectra a pulse width of 3 ,us (ca. 20"), spectral widths of 25 kHz and recycle delays of 1 s were used. X-Ray Diffractometry X-ray patterns were obtained with a Phillips APD 10 powder diffractometer. High- temperature studies employed an Anton Parr high-temperature attachment.12 Infrared Spectra Infrared spectra were recorded using a Nicolet MX1 F.t.i.r. spectrometer. The samples were prepared as thin self-supporting films and studied in a specially designed evacuable cell which was provided with facilities for heating, in situ, the sample film.Materials The clays used were Gelwhite L, a montmorillonite kindly supply by English China Clays Ltd, and a natural hectorite (designated SHCa-1 and supplied by the Source Clay Minerals Repository, University of Missouri). The Gelwhite was used as supplied but the hectorite was purified by sedimentation followed by treatment with dilute acetic acid to remove calcite impurity. Chlorhydrol, a commercial aluminium polymeric salt, was kindly supplied by Reheis Chemicals, Ireland. Preparation of Pillared Clays Pillared clays were prepared by either of two standard (1) where polymeric aluminium solutions were prepared from chlorhydrol solutions by ageing at 80 "C for 1 h or (2) following hydrolysis of 0.025 mol dm-3 aluminium chloride solutions with 2.25 equivalents of NaOH at 80 "C and ageing for 1 h.The zirconyl polymers were obtained by dissolving zirconyl chloride in water to produce a ca. 0.2 mol dm-3 solution. Following exchange with the polymeric cations the precursor pillared clay was calcined at different temperatures for varying lengths of time. Results and Discussion Characterisation of the Pillared Clays X-Ray Diflraction The pillared clays used in this study were alumina- and zirconia-pillared Gelwhite and alumina-pillared hectorite. Irrespective of the method of preparation all alumina-pillared clays showed (001) layer spacings of ca. 18 A. Considering that the aluminosilicate sheet of the clay has a thickness of 9.6 A, the pillar height is estimated to be ca.8.4 A. This compares well with the dimensions of the polymeric cations used for ~i1laring.l~ The zirconia-pillared clays showed (001) layer spacing of 16.2 A.D. r. B. Tennakoon, W. Jones and J . M . Thomas 2 55 II \ 3083 1 1 I I I 100 50 0 -50 -100 PPm Fig. 1. 27Al m.a.s. spectra of (a) Na-Gelwhite, (b) a precursor-pillared Gelwhite prepared from chlorhydrol, (c) a calcined-alumina-pillared Gelwhite prepared from chlorhydrol and ( d ) a calcined-alumina-pillared Gelwhite prepared from hydrolysed aluminium chloride. I 6 56 I 1 1 I I 100 50 0 - 50 -100 PPm Fig. 2. 27Al m.a.s. spectra of (a) natural hectorite, (b) precursor-pillared hectorite prepared from chlorhydrol, ( c ) calcined-alumina-pillared hectorite prepared from chlorhydrol and (a) calcined- alumina-pillared hectorite prepared from hydrolysed aluminium chloride solution.102 F A R 13084 SSB Metal-oxide-pillared Sheet Silicates OH SSB Fig. 3. Schematic diagram of the proposed structure of the [All,0,(OH),,(H,0),,]7+ cation (see text). 6 I I I I I I I I I 150 100 50 0 -50 -100 -150 PPm Fig. 4. 27Al m.a.s. spectra of the sodium selenate double salt of the ‘Al13’ cation (a) as prepared and (b) after heating to 500 “C for 3 h. Solid-state Nuclear Magnetic Resonance Nature of the Pillaring Species. For alumina-pillared Gelwhite, 27Al m.a.s.n.m.r. proved of little help because of the near-coincidence of the signals from the constituent aluminium and from the aluminium within the pillar. Spectra for the parent clay are shown in fig. 1 along with those for the precursor pillared clay (before calcination) andD.T. B. Tennakoon, W. Jones and J. M. Thomas 6 3085 I I I I I I I I 150 100 50 0 -50 -100 -150 PPm Fig. 5. 27Al m.a.s. spectra of chlorhydrol (a) as supplied and (b) after heating to 500 "C for 3 h. the pillared analogue. All show two signals centred around 0 and 53 ppm which correspond to aluminium in octahedral and tetrahedral oxygen coordination, respectively. The precursor pillared clay [fig. 2(b)] shows an additional absorption at 67 ppm corresponding to the exchanged polymeric ions. However, using a natural hectorite containing little structural aluminium, it is possible to obtain the 27Al spectra of the pillar i t ~ e l f . ~ The relevant spectra are shown in fig. 2. The parent clay, fig.2(a), has little aluminium. The spectra [fig. 2(c) and (d)] corresponding to alumina-pillared hectorite therefore show the aluminium distribution within the pillar. Fig. 2(b) is that of the precursor-pillared hectorite prepared from chlorhydrol; there are two signals, the first at 7 ppm and the second at 67 ppm, corresponding to octahedral and tetrahedral aluminium, respectively. No reliable quantitative information can be gained from the intensities of the aluminium ~igna1s.l~ In order to interpret the above spectra it is necessary to consider the structure of the polymer used for the pillaring. This ion is believed to be the [Al,,04(OH)2,(H20)12]7+ ion descrilxd by Johansson,13 and the proposed structure is shown in fig. 3. It consists of four sets of three-edge-sharing aluminium octahedra which are linked to a centrally placed tetrahedral aluminium.The corner positions are occupied by oxygen, bridging hydroxy groups and bridging water molecules. Thus the ion contains twelve peripheral octahedral and one central tetrahedral aluminium atoms. The 27Al n.m.r. spectrum of this ion in solution shows only one sharp signal, at 59 ppm, corresponding to tetrahedral a1~minium.l~ It is thought that the octahedral signals of the quadrupolar aluminium nuclei are broadened by electric field gradient effects. The more symmetrically placed central tetrahedral atom is relatively free from such effects. The solid-state m.a.s. spectrum of the sodium selenate double salt of the above ion also shows only a tetrahedral signal at 62 ppm, fig.4(a). Solid chlorhydrol, which in solution gives the same polymeric ion, however, shows broad signals for both tetrahedral and octahedral aluminium, as shown in fig. 5(a). On the other hand, when the interlamellar cations in hectorite are replaced by the polymeric ion, the spectrum of the precursor pillared clay shows relatively sharp signals for both tetrahedral and octahedral 102-23086 Me tal-oxide-pillared Sheet Silicates 7 - 9 5 -95.5 I I I I I -70 -80 -90 -100 -110 PPm Fig. 6. 29Si m.a.s. spectra of (a) Na-Gelwhite, (b) precursor-pillared Gelwhite prepared from chlorhydrol, (c) a calcined-alumina-pillared Gelwhite prepared from chlorhydrol and ( d ) from hydrolysed aluminium chloride. SSB I I I I I -70 -80 -90 -100 -110 PPm Fig. 7. 29Si m.a.s.spectra of (a) natural hectorite and (b) calcined-alumina-pillared hectorite prepared from chlorhydrol.D . T. B. Tennakoon, W. Jones and J. M . Thomas 3087 94.5 A I \ -110 I I I -93.5 -60 - 80 -1 00 -1 20 PPm Fig. 8. 29Si m.a.s. spectra of (a) the precursor-pillared Gelwhite prepared from chlorhydrol, (b) after calcination at 500 "C and (c) following subsequent treatment at 300 "C with ammonia. aluminium, fig. 2(b). It appears, therefore, that the broadening of the octahedral signal depends on the immediate environment of the polymeric ion; in solution or in the double salt structure the electric field gradient effects are large and the octahedral signal is lost, whereas in the solid chlorhydrol structure or when the ion is placed within the sheets of a clay mineral the field gradient effects appear to be less significant.Fig. 4(b) and 5(b) show the 27Al m.a.s. spectra of the sodium double salt and chlorhydrol, respectively, after calcination at 500 "C. The spectra are similar to those of pillared hectorite in so far as both tetrahedral and octahedral signals appear distinct. In the case of chlorhydrol the spectrum of the calcined material is merely a sharper version of that of the uncalcined material. Likewise, the spectrum of the pillared hectorite is similar to that of the precursor clay. On calcination, therefore, the basic aluminium distribution is maintained and little reordering of the relative atomic positions occurs, with the structure of the pillar following closely that of the polymeric ion. It has been shown elsewhere that the replacement of the bridging hydroxy groups and water molecules with oxygen atoms converts the structure shown in fig.3 to that of y-alumina.16 During calcination of the precursor material a similar change occurs, such that an oxide column resembling that of the y-alumina structure is left behind. Associated Changes in the Clay Framework. The 29Si m.a.s.n.m.r. spectra of Na-Gelwhite, [All,04(OH),4(H,0)l,]7+ ion-exchanged Gelwhite and alumina-pillared Gelwhite are shown in fig. 6. The parent clay shows two signals, the first at - 93.5 ppm corresponding to tetrahedral silicon linked through oxygen to three other silicons and to either3088 Metal-oxide-pillared Sheet Silicates 500 400 300 250 200 100 25 l . , , r l , , , , l ' , , ' ~ , , , ~ \ 2 4 6 8 28/" Fig.9. (001) X.r.d. traces of a precursor-alumina-pillared Gelwhite calcined at various temperatures and subsequently cooled to 25 "C in air prior to measurement, showing that an irreversible change occurs only at temperatures > 400 "C. aluminium or magnesium in the octahedral 1a~er.l~ The resolution is insufficient to separate out silicon adjacent to tetrahedral aluminium. A second signal appearing at - 110 ppm was assigned to silica impurity associated with the parent clay.ls The 29Si spectrum of the precursor pillared clay is similar to that of the parent clay, but following calcination the 29Si signal at -93.5 ppm shifts to -95.5 ppm with a concomitant broadening of the signal. Similar observations were made for the pillared hectorite, fig.7, where the parent clay absorbed at-95 ppm whilst in the pillared analogue this is shifted to --98.5 ppm. The chemical shifts of the pillared Gelwhite and pillared hectorite compare well with those of the uncharged analogues of the parent clays, pyrophyllite and talc, re~pective1y.l~ We have reported similar changes in 29Si chemical shifts on calcination of Li+, A13+ and NHZ ion-exchanged forms of other clay minerals,ls and have shown that the above changes reflect the neutralisation of the layer charge by small cations diffusing into the aluminosilicate structure. The calcination of the precursor pillared clay therefore results in the neutralisation of the clay layer charge. In the precursor material the layer-charge is balanced by the [Al,,0,(OH)2,(H20)l,]7+ ions, but after calcination this ion is converted into an oxide, with the layer charge then being balanced by the release of an equivalent number of protons.These protons are then able to migrate into the clay structure. This is in agreement with the very low cation-exchange capacities observed for the pillared clays, since the charge-balancing protons locked within the clay sheets would, under normal exchange conditions, be unavailable for replacement. Furthermore, it is known that the treatment, at elevated temperature, of a pillared clay with NH, increases the available cation-exchange capacity to nearly that of the parent clay;19 probably the base is able to abstract the trapped protons and transfer them, as NHZ ions, to the interlamellar space. This is indeed reflected in the 29Si chemical shift (fig. 8) where, following treatment with NH,, the 29Si signal is restored to a value close to that of the parent clay.The presence of NHZ ions in the interlayer was confirmed by an absorption in the i.r. spectrum at 1430 cm-l. It is noteworthy that the pillared clayD. T. B. Tennakoon, W. Jones and J. M. Thomas 3089 I I I I I 3000 2000 1500 1000 500 wavenumber/cm-' Fig. 10. F.t.i.r. spectra of a self-supporting film of a precursor alumina heated in situ in a vacuum of Torr at various temperatures. The lattice hydroxyl stretching mode is at 3630 cm-l, and librational modes are at 913 and 844 cm-l. The mode at 795 cm-l corresponds to silica impurity. Note the marked decrease in the intensity of the -OH stretching mode at 3630 cm-l.which is obtained when [Zr,(OH)l,(H20)l,]2+ ion is used also shows a 2 ppm shift. Proton diffusion into the octahedral framework appears to be a general phenomenon. Transformations Occurring during Calcination of the Precursor-pillared Material Variable- temperature X- Ray Difraction Studies Although the (001) layer spacing of the precursor clay decreased progressively upon heating to 400 "C, it did not fall to the 18 A layer spacing of the pillared clay which results from calcination at 500 "C (fig. 9). Furthermore, heating to 400 "C does not prevent re-expansion upon exposure to moist air. The pillared clay obtained by calcination at 500 "C shows no tendency to expand. Kinetic factors were studied by calcining at 400 "C for periods varying from 12 h to 1 week, exposure to moist air and the determination of the (001) repeat distance.All samples showed expanded layer spacings of 19.5 A, and it is therefore clear that in the temperature range 400-500 "C an3090 Me tal-oxide-pillared Sheet Silicates m L I I I I 1 I 4000 3000 2000 1500 1000 500 0 wavenumberlcm-' Fig. 11. F.t.i.r. spectra of a self-supporting film of NHi-Gelwhite heated in situ in a vacuum of 10-3 Torr to various temperatures. The lattice hydroxy stretching mode is at 3630 cm-l, and librational modes at 913 and 844 cm-l. The modes at 1430 and 795 cm-l correspond to the ammonium ion and silica impurity, respectively. The associated temperatures are: (a) 25, (b) 150, (c) 250, ( d ) 400 and (e) 500 "C. All for 30 min heating time. Spectrum (f) is for 3 h at 500 "C.irreversible contraction in the layer spacing occurs, during which the pillars are irreversibly held within the aluminosilicate sheets. Infrared Studies Calcination was monitored using F. t.i.r. spectroscopy, with thin self-supporting films of the precursor pillared clays calcined in situ under a vacuum of lop3 Torr. Fig. 10 shows the transmission infrared spectra obtained at different calcination temperatures. Commencing at ca. 250 "C, there is a progressive reduction in intensity of the stretching modes of the structural hydroxy groups at 3630 cm-l and the Al---OH---A1 and A1---OH---Mg librational modes at 913 and 844 cm-l, respectively. DuringD , T. B. Tennakoon, W. Jones and J . M . Thomas 309 1 100 50 0 -50 -100 PPm Fig. 12. 27Al m.a.s.n.m.r.spectra of Na-Gelwhite (a) before and (b) after heating to 500 "C for 1 h, (c) to 550 "C for 200 h, ( d ) to 700 "C for 1 h and (e) to 900 "C for 1 h. Note the loss in intensity of the peak at ca. 2 ppm above 500 "C. calcination the intensities of the librational modes decrease at a much faster rate, similar to that seen for straight ion-exchanged clays18 and previously shown to be due to the migration of small interlayer cations into the clay structure, thereby interacting with the librational modes of the lattice hydroxy groups. The parallel behaviour shown by the pillared clays provides additional evidence for the proton migration mechanism proposed above from n.m.r. investigations. At 500 "C a drastic reduction in the stretching mode of the hydroxy groups at 3630 cm-l is observed: this drastic reduction has no counterpart in a normal ion-exchanged clay, such as NH,+ ion-exchanged Gelwhite (fig.11). Thus, although the changes in the intensity of the librational modes are common to the pillared and ion-exchanged clays, the drastic reduction in the intensity of the stretching mode is unique to the pillared forms. At ca. 500 "C a substantial reduction in the number of structural hydroxy groups therefore occurs for the pillared clays which is not directly related to proton diffusion. Along with the X-ray results presented above, this may be rationalised on the basis of condensation of those terminal hydroxy groups which are present on the polymeric ions, with the lattice hydroxy groups on the clay sheets.The oxide pillars then become directly linked via oxygen to the aluminium and magnesium atoms in the octahedral layer. Such a situation would explain the rigid layer structure and the resistance to expansion shown by pillared clays calcined at 500 "C. This idea is further supported by the following evidence. It has been previously reported,18 on the basis of infrared spectroscopy and 27A1 and 29Si m.a.s.n.m.r., that calcination of different cation-exchanged clays at 700 "C results in nearly complete dehydroxylation of the structure. As shown in fig. 12, on heating Na+-Gelwhite to3092 Me tal-oxide-pillared Sheet Silicates n -111 PPM -50 -100 PPm Fig. 13. 29Si m.a.s.n.m.r. spectra of Na+-Gelwhite (a) before and (b) after heating to 500 "C for 1 h, (c) 550 "C for 200 h, ( d ) 700 "C for 1 h and (e) 900 "C for 1 h.700 "C, the signal at 2 ppm due to octahedral aluminium disappears, but without a corresponding increase in the intensity of the tetrahedral signal, because five-coordinated aluminium, formed on dehydroxylation, is n.m.r.-silent.209 21 On the other hand, the 29Si signal at -93.5 ppm of the parent clay is replaced by a new signal at - 100 ppm, fig. 13, which corresponds to silicon in the tetrahedral sheet linked via oxygen to three other silicon atoms and a trigonal bipyramid-shaped pentacoordinated aluminium or magnesium in the previously octahedral, central layer of the clay sheet.'* When, however, a zirconia-pillared Gelwhite, where the only aluminium present is that of the clay structure, is heated at 700 "C, the 27Al spectrum shows a distinct signal for octahedral aluminium, as shown in fig.14. The i.r. spectrum of the same sample does not show any absorption attributable to lattice hydroxy groups, suggesting that under the above calcination conditions, although the structural hydroxy groups are lost, a substantial proportion of the aluminium in the clay sheets remains octahedrally co- ordinated. This is possible only if the octahedral coordination around the aluminium is maintained by new linkages, via oxygen, in place of the hydroxy groups in the parent clay. Such a situation may be rationalised if it is postulated that the pillars are linked to the octahedral cations from both above and below the sheet.D. T. B. Tennakoon, W. Jones and J . M . Thomas 3093 56 3 lirz I I I I I 100 50 0 - - PPm Fig.14. 27Al m.a.s.n.m.r. spectra of zirconia-pillared Gelwhite calcined for 1 h at (a) 500 and (b) 700 "C. L I I I I I I -70 -80 -90 -100 -110 PPm Fig. 15. 2gSi m.a.s.n.m.r. spectra of alumina-pillared Gelwhite calcined at (a) 500 "C for 1 h and (b) 700 "C for 10 h. 29Si m.a.s.n.m.r. spectra of pillared clays calcined at 700 "C confirm the above possibility. Thus, as shown in fig. 15, the 29Si signal at -95.5 ppm present in the pillared clay persists in the same material after calcination at 700 "C for 10 h. It is clear that a substantial proportion of the silicon remains linked to octahedral aluminium. The spectrum shows, however, a signal at - 100 ppm, indicating that part of the octahedral aluminium and magnesium is converted, through dehydroxylation, to a fivefold coord- inated aluminium.3094 Metal- oxide -p illar ed Sheet Silicates The number of interlayer cations in a clay is determined by the extent of isomorphous substitution and the charge on the cation.In the case of the precursor pillared clays the interlayer species are the multivalent, polymeric cations. The precursor material therefore has many more lattice hydroxy groups than there are polymeric cations. Thus when one or more of the terminal hydroxy groups on the polymeric ion condense with lattice hydroxyls, a substantial number of hydroxy groups would remain in the clay sheets. On heating to 700 "C these remaining hydroxy groups would be lost, to produce an equivalent number of pentacoordinated sites in the clay sheet.This situation accounts for the 2sSi signal at - 100 ppm. The aluminium or magnesium atoms already linked to the pillar would remain octahedral at 700 "C, provided that both bridging hydroxy groups originally linked to the octahedral atom had condensed with the polymeric cation. This would require that the octahedral atom is linked to two oxide pillars, one on either side of the sheet. Conclusions The 27A1 m.a.s.n.m.r. studies show that the oxide pillar has an aluminium distribution similar to that of the calcined polymeric aluminium salts. Comparing the spectra of calcined and uncalcined chlorhydrol, it is seen that the aluminium distribution changes little on calcination. Considering the fact that one can convert the polymeric ion structure into that of alumina merely by replacing the bridging hydroxy groups and water molecules with oxygen atoms, we conclude that the oxide pillar is essentially alumina.The i.r. and n.m.r. studies also show that, on calcination of the precursor pillared clay the polymeric ion is converted to a neutral oxide with the release of protons as charge-balancing cations, and that these protons diffuse into the clay structure under the above conditions. The low cation-exchange capacity and the low catalytic activity at moderate temperatures, exhibited by the pillared clays clearly reflect the inaccessibility of these protons from the interlamellar space. Further, these protons may occupy at least two different sites within the aluminosilicate structure. In the freshly prepared material the protons appear to be directly associated with the negatively charged isomorphously substituted centre.The studies on the calcination process and the further calcination of the pillared clay at 700 "C together show that the alumina and zirconia pillars are linked to the aluminium and magnesium atoms in the octahedral layers of the clay. We appreciate helpful discussions with Drs R. Schlogl, T. Rayment, T. A. Carpenter and P. A. Diddams and the collaboration with Prof. J. H. Purnell and J. A. Ballantine, University College of Swansea. Financial support by the S.E.R.C. and B.P. (Sunbury) is gratefully acknowledged. References 1 J. M. Thomas, in Intercalation Chemistry, ed. M. S . Whittingham and A. J. Jacobson (Academic Press, 2 T. J. Pinnavaia, Science, 1983, 220, 365.3 D. E. W. Vaughan, R. J. Lussier and J. S. Magee, German Patent 2.825,769 (1979). 4 D. E. W. Vaughan, R. J. Lussier and J. S. Magee, US. Patent 4,176,090 (1979). 5 S . Yamanaka and G. W. Brindley, Clays Clay Miner., 1979,27, 119. 6 T. J. Pinnavaia, V. Rainey, Ming-Shin Tzou and J. W. White, J . Mol. Catal., 1984, 27, 213. 7 D. Plee, F. Borg, L. Gatineau and J. J. Fripiat, J . Am. Chem. Soc., 1985, 107, 2262. 8 P. A. Diddams, J. M. Thomas, W. Jones, J. A. Ballantine and J. H. Purnell, J . Chem. SOC., Chem. 9 D. T. B. Tennakoon, W. Jones, J. M. Thomas, L. J. Williamson, J. A. Ballantine and J. H. Purnell: New York, 1982), pp. 56-92. Commun., 1984, 1340. Proc. Indian Acad. Sci., Sect. A , , 1983, 92, 27. 10 M. Matsumoto, S. Shimoda, H. Takahashi and Y. Sato, Bull. Chem. Soc. Jpn, 1984, 57.D. T. B. Tennakoon, W. Jones and J . M . Thomas 3095 11 M. L. Occelli and J. E. Lester, Ind. Eng. Chem., Prod. Res. Dev., 1985, 24, 27. 12 D. T. B. Tennakoon, R. Scholgl, T. Rayment, J . Klinowski, W. Jones and J. M. Thomas, Clay Miner., 1983, 18, 357. 13 G. Johansson, Acta Chem. Scand., 1960, 14, 771. 14 K. F. M. G. J. Scholle, A. P. M. Kentgens, P. Frenken and W. S. Veeman, J . Phys. Chem., 1983,88. 5. 15 J. W. Akitt and A. Farthing, J . Chem. SOC., Dalton Trans., 1981, 1617. 16 S. Ramdas, personal communication. 17 M. Magi, E. Lippmaa, A. Samoson, G. Engelhardt and A-T. Grimer, J . Phys. Chem., 1984,88, 1518. 18 D. T. B. Tennakoon, T. A. Carpenter, W. Jones and J. M. Thomas, J . Chem. SOC., Faraday Trans. 1, 19 D. E. W. Vaughan, personal communication. 20 N. C. M. Alma, G. R. Hays, A. V. Samoson and E. T. Lippmaa, Anal. Chem., 1984,56, 729. 21 R. L. Frost and P. F. Barron, J . Phys. Chem., 1984,88, 6206. 1986, 82, 545. Paper 511982; Received 1 1 th Nouemher, I985

ISSN:0300-9599    

DOI:10.1039/F19868203081

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. I , 1986,82, 3097-3 112 A Nuclear Magnetic Resonance Study of the Solvatochromism of a Pyridinium Betaine J. Graham Dawber* and Richard A. Williams Department of Chemistry and Biology, North Stafordshire Polytechnic, Stoke-on-Trent ST4 2DE The lH and 13C nuclear magnetic resonance spectra of the solvatochromic indicator 2,4,6-triphenylpyridinium-N-3,5-diphenylphenol betaine have been measured in CDCl,, CD,OD, [2H,]DMS0 and [2H,]acetone solvents and 1 : 1 molar mixtures of CDC1,-DMSO, CDC1,-CD,OD, CDC1,- acetone, DMSO-acetone, DMSO-D,O and CD,OH-D,O. An attempt has been made to assign the resonance signals and to measure the changes in resonance positions with change of solvent. The major changes in resonance positions that accompany change of solvent are for those parts in the molecule close to the centres of charge. The effects of the differences in solvation at these centres, which are likely to be associated with the extreme solvatochromic properties of the betaine, are also transmitted to other parts of the molecule.The important effect of solvents upon chemical reactions has long been known and there have been several attempts to correlate empirical and experimental parameters of solvents with their polarity and solvation properties. These include1 the so-called Y values, X values, R values, II* values, 2 values and ET values. Of these, the 2 values and the ET values1 depend upon the solvatochromic behaviour of dyes, the term solvatochromism referring to the shift of an electronic absorption band when varying the polarity of the medium.Such changes result from physical intermolecular solute- solvent interaction forces, which usually involve alterations in the electronic ground state or may also involve the excited state of the absorbing species. The u.v.-visible solvatochromism of such dyes, while providing useful solvent polarity parameters, gives no information regarding the localised changes within the solute concerned when the solvent polarity is changed. The proton-decoupled 13C n.m.r. spectrum, on the other hand, can, in principle, give such information since each different carbon atom gives rise to a single line frequency. Dimroth, Reichardt et aL2 discovered a pyridinium N-phenol betaine (I) (2,4,6- trip hen ylp yridinium-N- 3,5 -dip hen ylphenol bet aine), whose solva toc hr omism was very large (A,,, = 810 nm in diphenyl ether to 450 nm in water) and which was used to establish the solvent polarity ET sca1e.l It was decided, therefore, to undertake a lH and a 13C n.m.r.study of this interesting compound to give possible information concerning the localised intramolecular changes which occur when the solvent in which it is dissolved is changed, but also to study solvent mixtures with a view to detecting possible preferential solvation by one of the solvent components. A similar study has recently been carried out with a merocyanine dye which has also been used as a solvatochromic indicator .3 30973098 Solvatochromism of a Pyridinium Betaine H H H H H H H 0- Experimental 2,4,6-Triphenylpyridnium-N-3,5-diphenylphenol betaine (I) This compound was synthesised by the condensation of 2,4,6-triphenylpyrilium perchlorate4y with 2,6-diphenyl-4-aminopheno16 (11) to give the perchlorate salt which is the precursor of the betaine (I); m.pt.273-274 "C, (30% yield). Treatment of the perchlorate salt with sodium methoxide in methanol as described by Dimroth et aL2 gives, after recrystallisation from methanol-water (1 : 3), the required betaine (I), m.pt. 269-273 "C, which is in fact a dihydrate. The N.M.R. Spectra The n.m.r. spectra were recorded on a Jeol FX90Q Fourier-transform spectrometer (proton resonance at 89.55 MHz, carbon resonance at 22.49 MHz). The lH spectra were obtained using a 90" tip-angle and a spectral width of 900 Hz. The noise-decoupled 13C spectra were obtained with a tip-angle of 30" with a pulse repetition time of 1 s and ca.50 000 scans. Quantitative decoupled 13C spectra without nuclear Overhauser enhance- ment (NNE) for (I) in CDCl, were obtained using a tip-angle of 30" and a pulse repetition time of 30 s with gated decoupling during acquisition. Undecoupled 13C spectra were obtained with nuclear Overhauser enhancement (NOE) using gated decoupling with the proton decoupler switched off during acquisition and a pulse repetition time of 3 s. For all the 13C spectra the responses were acquired into 8K and zero-filled to 16K data points, and an experimental broadening of 0.9 Hz was applied prior to Fourier transformation. The lH and 13C chemical shifts were all measured relative to TMS as internal standard at S = 0.0 ppm.Spectra were recorded in CDCl,, CD,OD (also CD,OH), [2H6]DMS0 and [2H6]acetone solvents and 1 : 1 molar mixtures of CDC1,-DMSO, CDC1,-CD,OD, CDC1,-acetone, DMSO-acetone, CD,OD-D,O, and DMSO-D,O, and also with added HCl and NaOD to the latter mixture. TheJ. G. Dawber and R. A . Williams 3099 i j k I /I 4 a b c I 8.0 7.5 7.0 6, (PPm) Fig. 1. (a) lH n.m.r. spectrum of (I) in [2H6] acetone. (b) lH n.m.r. spectrum of (I) in CDC1,.3100 Solvatochromism of a Pyridinium Betaine Table 1. 'H chemical shifts of pyridinium betaine (I) in various solvents 'H chemical shift, 6 solvent abc g ij k n 9 r S ~~ ~~~ CDC1, [ 2H6]ace t o ne CD,OD 7.18 6.72 7.36 7.19 7.01 7.46 (split) (split) 7.19 6.78 7.43 [ 2H ;I DMSO 7.23 [2H6]DMSO-[2H6]acetOne 7.24 (split) CDC1,-CD,OH 7.20 [2H,]DMSO-CDCl, 7.20 CDC1,-[2H6]ace tone 7.17 (split) [2H6]DMSO-D20 7.2 1 (split) CD30H-D20 7.24 6.98 6.99 6.70 6.91 6.80 (split) 6.83 6.82 7.48 7.47 split) 7.47 7.47 7.41 7.46 split) 7.50 8.01 7.83 7.54 7.26 8.48 8.24 7.67 7.33 8.34 8.07 7.60 7.26 8.56 8.30 7.63 7.29 8.55 8.32 7.63 7.35 8.26 7.98 7.65 7.25 8.45 8.18 7.60 7.28 8.27 8.06 7.58 ? 8.40 8.21 7.64 7.31 8.39 8.09 7.69 ? ~ ~~~ -~ deuterated solvents (Aldrich) were used as received and had the following isotopic purities: CDCI, (99.8% D), CD,OD(99.5% D), [2H,JDMS0 (99 +% D) and [2H6]acetone (99.5 % D).The CDC1, was checked for acidity by extraction with water and measurement of the pH of the aqueous extract. 1 cm3 of CDCl, was shaken with 50 cm3 of distilled water, the pH of which changed from 6.4 to 5.8, corresponding to an acidity in the CDCl, of ca.2.5 x mol drn-,; this is insignificant in comparison with the concentration of betaine used, which is three orders of magnitude greater than this. The lH and 13C spectra were measured on the same filtered solution, which normally contained ca. 20 mg of betaine in 1.5 cm3 solvent. Results and Discussion The greatest resolution of the lH resonances was obtained for a solution of (I) in [2H6]acetone (and also in the [2H6]DMSO-[2H6]aCetOne mixture and the CDC1,- [2H6]acetone mixture) and this is shown in fig. 1 (a). The least splitting was observed for CDCl, solvent, shown in fig. 1 (b). From the integration traces of the lH spectra and the numbers of similar protons in different parts of the molecule a tentative assignment of the signals was made.These lH assignments and their chemical shifts in the various solvents are given in table 1. The proton spectra in the various solvents show qualitative similarities and differences. For example, the splittings of the bands due to Habc and Hijk depend quite markedly on solvent, and yet for these protons the chemical shifts do not change greatly. The overall changes in proton chemical shift for the rest of the molecule lie in the range 0.1-0.7 ppm. Several factors can contribute to solvent-induced chemical-shift changes.' These include contributions from the bulk solvent susceptibility, anisotropy, reaction field, hydrogen bonding, specific interactions and van der Waals forces.' In the case of electronic spectra, effects accompanying changes in solvent may also have contributions from differences between electron density distributions and dipole moments in the ground and excited states.However, in the case of n.m.r. spectra the majority of the molecules are in their electronic ground state and the lifetimes of the excited states are very short on the n.m.r. timescale. In any case, the n.m.r. spectra measured on the Jeol FX90Q instrument are carried out with the sample in light-free environment. Conse-J. G. Dawber and R. A . Williams 3101 quently the changes observed in the n.m.r. chemical shift produced by the solvent are likely to be caused by the normal solvent effects due to shape and polarity of the solute and solvent rather than excited-state contributions.Proton-proton couplings are fairly insensitive to solvent but the lH chemical shifts are solvent dependent. However, Benson and Murrel* concluded that no simple relation can be derived between the lH chemical shifts and the n-electron densities in aromatic molecules. It has been showng that the 13C and lH n.m.r. chemical shifts of the simpler CHCl, molecule in a range of solvents show that the changes in 6, range from 0 to 4.2 ppm going from cyclohexane to hexamethylphosphorimide, for which the corres- ponding changes in 6, are approximately half those for 6,. The values of AS, for CHCl, going from cyclohexane to methanol and acetone solvents are 1.35 and 1.76 ppm downfield, respectively, with the values being about half of these. Similarly, the addition of benzene to non-polar solutions of the more complex molecules of flavones produces characteristic proton chemical shifts.1° However, if the chemical-shift differences between non-equivalent protons are reduced in a particular solvent then the splittings will be affected.The splitting of the lH resonances of (I) is most apparent in acetone solvent alone or any mixture containing acetone. Thus the chemical-shift differences between adjacent protons are decreased in solvents other than acetone and the coupling constants are insufficient to split the peaks. Poly-halogen compounds, such as CCl, and CHCl,, are known to form weak charge-transfer complexes with aromatic compounds,ll and this must be a distinct possibility, with the pendant phenyl rings attached to the main betaine structure thereby influencing the chemical shifts of Habc and Hijk. In addition, the extent of splittings in the bands from H, and H, are also slightly solvent dependent.The proton resonances which are most affected by solvent are those from H, and H,, which might be expected in view of their positions in the molecule. Also affected, rather surprisingly, is the resonance from H,. The greatest solvent effect was upon the H, resonance, which is no doubt due to these protons being nearest to the cationic centre of the betaine molecule and being favourable to solvation by polar solvent molecules. The H, protons are deshielded by solvation in the order DMSO > acetone > CD,OH > CDCl,. In the case of binary solvent mixtures it is possible in principle to judge the presence of any preferential solvation by comparing the resonances of the solute in the solvent mixtures with the corresponding values in the pure solvents.If a particular line is nearer to its position in one of the pure solvents when the solute is dissolved in a solvent mixture then this could indicate preferential solvation at that particular site. From the H,, H, and H, chemical shifts it can be seen that there is little or no preferential solvation of these protons in the CDC1,-CD,OD mixture and the CDC1,-acetone mixture. It must be borne in mind, however, that the lH spectra will probably only reflect the extent of changes in hydrogen bonding, and to a lesser extent any changes in van der Waals complexing or n-complexing, which are more likely to be observed in the 13C spectra.In the case of the DMSO-CDCl, and DMSO-acetone mixtures the evidence suggests that there is some preferential solvation of the betaine in the region of H, by DMSO. The 13C chemical shifts of (I) in the various solvent systems are listed in table 2. There are a total of 19 distinguishable carbon atoms in the structure of (I), and in CDCl, solvent there are 19 lines in the noise-decoupled 13C n.m.r. spectrum (fig. 2). These lines are numbered 1 to 19 starting from the low-field end. The NNE decoupled 13C n.m.r. spectrum allowed the number of carbon atoms associated with each signal to be established and showed whether or not they were quaternary or proton-bearing. The carbon atoms in the betaine are labelled a to s as shown in structure (I).The 13C n.m.r. spectrum of the precursor of the betaine, namely 3,Sdiphenyl- 4-hydroxyaniline (11) was measured in CDCl, and the resonances designated by means of standard tables of assignments;I2-l4 these, along with other relevant assignments used, are given in structures (11)-(VIII). Using this information and that from the undecoupledTable 2. 13C chemical shifts, 6 (ppm) of the pyridinium betaine (I) in various solvents relative to TMSa peak no. from low CDC13- CDC13- CDCl3- [2H,]DMSO- [2H6]DMSO- field CDC1, [2H,]DMS0 [2H,]acetone CD30D [2H,]DMS0 CD30D [2H,]acetone [2H,]a~etone D2O 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 161.3 156.9 155.7 138.7 134.2 133.7 132.0 130.26 130.19 I 129.81 129.41 129.11 128.78 128.51 128.10 127.80 126.40 126.00 124.23 161.0 156.5 154.0 139.7 134.0 133.5 132.0 129.47 128.96 128.73 128.60 128.49 (sh) (sh) 127.21 128+oo 1 125.50 125.07 124.00 162.9 158.0 156.I 141.0 135.1 135.0 133.0 131.1 (sh) 130.61 130.02 129.89 l g 6 } 129.17 128.25 126.61 126.55 125.20 164.4 158.7 157.2 141.6 135.7 135.5 133.2 132.9 (sh) 131a0 (sh) 1 130.89 I 130.45 130.2 I 129.50 129.39 128.48 127.98 126.82 125.9 1 161.1 156.5 154.6 139.5 133.9 133.7 132.0 129.61 (sh) 129.53 139.20 128.80 128.69 128.36 128.20 127.30 125.71 125.40 124.10 (sh) 164.2 157.7 156.5 140.2 134.2 134.0 132.9 132.3 130.76 130.35 129.78 129.52 129.10 128.91 128.40 127.91 126.51 126.09 123.87 164.0 157.3 155.6 140.4 134.7 134.6 132.4 130.50 130.17 130.07 129.86 129.40 128.90 128.80 128.52 127.70 126.21 125.97 123.26 1 (sh) 163.1 157.1 154.7 140.8 134.9 134.3 132.4 i I 130.04 129.62 129.22 128.90 128.56 127.60 1 125.79 J 123.59 ~- a (sh).shoulder: a brace indicates the merging together of two or more lines. 164.6 156.6 153.9 140.6 134.1 133.6 132.0 129.61 129.41 128.84 128.56 128.29 128.05 127.15 125.13 121.76 CD30H- D2O _____. 158.4 157.3 151.3 148.6 140.5 135.1 134.9 133.42 132.59 131.11 130.90 130.57 130.18 129.50 129.16 128.80 127.53 125.34 - -~ ~J . G. Dawber and R. A . Williams 3103 I , , , I 12 I , I , , , , $ I I , , , , I , , , , , ( I ( , I , , , ) , , , , I 13C spectrum of the betaine in CDCl, (obtained with Overhauser enhancement) (fig. 3) the assignment of the 19 lines was attempted making use of the carbon-hydrogen coupling constants (lJCH, 2JCH, 3JcH).The results of this and relevant comments are given in table 3. A few of the assignments may not be completely unequivocal, but for the few doubtful cases intuitive suggestions are made in table 3. In table 4 are presented the assignments of the 13C resonances of the betaine in CDCl, along with the changes (Ad,,,,,) produced when the betaine is dissolved in [2H,]DMS0, [*H],acetone and CD,OD. Also presented are the data for 50 : 50 mol % of CD30H-D20 and 50: 50 mol % of [2H6]DMSO-D20. In general, all of the 6, values of (I) were found to be influenced by solvent, the effects for the pure solvents being approximately in the order CD,OD > acetone > CDCl, > DMSO. The solvent-induced 13C chemical-shift changes, Adc, fall in the range &3 ppm, and this is comparable with values found for simpler molecules, e.g.CHC1,.g Also, by comparison the values of Adc for benzene in different solvents are between -0.1 and 0.3 ppm, while those for the polar solute of ani- line range from - 2.7 to + 2.8 going from CCl, to DMSO s01vent.l~ Unlike proton chemi- cal shifts,s the changes in 13C chemical shift correlate with charge-density calculations for the various carbon atoms. l5 Solvent-induced chemical-shift changes can cause over- lapping of 13C resonance signals, which in one solvent may be separate, but in another solvent they are partially overlapped, for example the aromatic compound 3-bromo- biphenyl in cyclohexane and DMSO.lG Similar effects were observed for the betaine (I), as shown in table 2. Thus the chemical-shift changes, Adc, for (I) produced by different solvents appear to be similar in magnitude to those for simpler molecular systems, and hence the observed effects are likely to correspond to normal solvent effects arising from polarity changes.Table 4 shows that the greatest solvent-induced chemical-shift changes are for those carbon atoms which we have designated as being nearest the centres of charge in the molecule, and the least effkcts are for those carbon atoms which are3104 Solvatochromism of a Pyridinium Betaine 127.6 129.2 127.7 136.0 123.8 149.9 OH 118.2 132.6 148.6 128.5 0 141.9 0- 121.2 132.6 123.3 (VII) 118.1 cvrrr)J. G. Dawber and R . A . Williams 1 8 1 8 134 132 125 4 h 3105 ! 4 . , . , , , , , , , , , , , ~ , , l ' " ' ~ " ' ' 1 " ' ~ ' ' 160 140 120 6, (PPm) Fig.3. Undecoupled 13C n.m.r. spectrum of (I) in CDCI,. furthest away from the charge centres; this adds confidence to our assignments for the 13C n.m.r. signals. The carbon atoms which were influenced most by solvent were Cf, C,, C,, C,, C,, C, and C,, which correspond to the quaternary carbon atoms nearest to the centres of charge within the betaine. The C, and C, carbon atoms are less affected, probably owing to their being hydrogen-bearing. The downfield shift (higher 6) accompanying increased polarity of the solvent is greater for the C, and C , carbons, near the 0- centre, than for the C, and C, atoms adjacent to the N+ centre. A It is of interest to compare our results for the betaine (I) with those of another solvatochromic substance, the merocyanine dye (IX), the I3C n.m.r.spectra of which were measured in CDCl,, DMSO and CD,0H,3 and for which values of AdCDCl3 were calculated for comparison with our results. In the case of the change from CDCl, to DMSO, the largest negative AdCDCl3 change is for C,, and this corresponds to the same carbon atom in (I), i.e. Ch, while the largest positive AdCDCl3 change is for C,, at the negative 0- centre. In the case of (I) the largest positive Ad was for the adjacent carbon atom, i.e. C,. In the case of the change from CDCl, to CD,OH for (IX), all the solvent-induced chemical-shift changes were positive as for (I). The largest positive changes were for carbon atoms C,, C,, C, and C,, and these correspond to C1, C, and C, in (I). Thus the two systems are similar.The 13C chemical shifts of (IX) have been3106 Solvatochromism of a Pyridinium Betaine Table 3. 13C assignments of betaine (I) line no. of 6 (ppm) assign- H- comments on undecoupled spectrum no. C atoms in CDC1, ment bearing and assignment 1 1 2 2 3 1 4 2 5 1 6 2 7 1 8 2 9 2 11 4 12 4 13 4 14 2 161.3 156.9 155.7 138.7 134.2 133.7 132.0 130.26 130.19 129.40 129.11 128.78 128.51 f m h e P 1 S d 9 k C j n Likely to be C, since this will be the most deshielded C atom, [cf. structures (VII) and (VIII)]. Undecoupled gives "JCH with protons on C, (J z 8 Hz) Undecoupled gives singlet which is slightly broadened which could be due to unre- solved 2JcH splitting. Should be C, since 2JcH could occur with H,. Likely to be C , rather than C, since orlho to N+ at lowerfield (higher 6) than ortho to 0- [cf.structures (VI) and (VIII)] Undecoupled splits into narrow triplet with J z 4 Hz which is probably 'JCH coupling, i.e. C , split by the two H, atoms Undecoupled gives broadened singlet which could be C , since ortho to -C-O- at higherfield (lower 6) than ortho to N+ Probably C, since this will be shielded relative to C,. In principle there should be "JCH triplet splitting with H,, but the undecoupled spectrum shows shoulders on this peak Undecoupled broadened which may be a narrow triplet, i.e. unresolved 2JcH split- ting. Assigned as C, rather than C, since by comparison C, is at lower field (higher 6) than C, Undecoupled gives 2 triplets, lJCH = 162 Hz and "JCH = 8 Hz. Thus, this must be a proton-bearing carbon with possi- bility of 3JCH coupling, i.e.C, since only one C atom involved; 2JcH not apparent Undecoupled slightly broadened : it may be a narrow triplet (cf. line 6) i.e. unresolved 'JCF. Assigned as C, rather than C, since C, is at higher field (lower 6) than C , by comparison Undecoupled gives two peaks (lJCH = 162 Hz with some messy splitting from overlapping. Likely to be C,, and C,. Assign peak 9 as C, and 10 as C since ortho slightly higher field (lower j) than meta [cf. structure (VIII)] Undecoupled gives quarter, 'JCH and ortho 2JcH [cf. phenoxide anion (VIII)]; may be C, or C,: designate as C, As above (line 11) Probably Cj; see line 16 Undecoupled gives doublet ('JCH = 162 Hz) may be C, or C,; assign as C, since meta of X lowerfield (higher 6) than meta of VI.Coupling with H, and H, probably weak due to possible non- planarity of adjacent phenyl ringsJ . G. Dawber and R. A . Williams Table 3. (cont.) 3107 line no. of 6 (ppm) assign- H- comments on undecoupled spectrum no. C atoms in CDC1, ment bearing and assignment 17 18 19 15 2 128.10 g yes Undecoupled gives doublet (lJCH = 162 Hz); see line 14 16 4 127.80 b yes Undecoupled gives 2 large peaks (lJCH = 162 Hz) with slight broadening due to 2JCH coupling; probably cb or cj : assign as cb, [cf. (VIII) and (VI)] 2 126.40 a yes Undecoupled gives 2 peaks with triplet structure: probably C, or Ci 2 126.00 1 yes As for line 17 1 124.23 0 no Undecoupled gves very narrow triplet, i.e. 2JCH only. Possible for C, in pyridinium ring. Any 3JcH would involve another ring which is probably-not coplanar _ _ ~ ~ _ _ shown to give a linear correlation with the electron densities in the ground state.Although there will be a change in the permanent dipole accompanying an electronic transition, the calculation of the ground-state dipole moment of (IX) shows significant changes when the solvent polarity is changed. Such changes in the ground-state dipole moment induced by the surrounding solvent cage are thought to be particularly important in the case of solvatochromic l7 Thus this is likely to be the case not only for the merocyanine (IX) but also for the betaine (I). The 6, values for (I) in [2H,]DMS0 seem to be out of line with the relative polarity of this solvent as judged by its dielectric constant, E . This is illustrated in fig.4 for carbon atoms nearest the centres of charge, Cf, C,, and C,, where the points for DMSO at E = 50 are very different from the rest. This may be due to a difference in the type of bonding in the solvation of the betaine with DMSO. The relationship between E and ET is not linear, and for DMSO the value of ET seems low for its value of 6.l The changes in the 13C resonances for each carbon atom produced by the addition of D20 to DMSO and CQOH (50: 50 mol%) are plotted in fig. 5. Large changes do occur for the carbon atoms near the centres of charge. The resonances for C, move in the same direction for (I) in both solvents on addition of D20, but move in opposition for Cf. For Ch and C,, i.e. atoms near to the N+ centre, the resonances are affected by the addition of D20 to the CD,OH solvent, but not significantly in the case of DMSO, which may mean that there is strong interaction of the DMSO near the N+ centre of the betaine molecule.For reasons which are not clear, the C, resonance (for DMSO solvent) and C, (for CD30H solvent) are considerably affected by the addition of D20 to the organic solvent. Since D20 in high concentrations (50 mol%) does produce large changes for some carbon atoms of (I) it is perhaps pertinent to question the influence of small amounts of water in the supposedly pure organic solvents. Water is known to affect the 6, values of dipolar solutes by hydrogen bonding. For example, the 6, values for quinoline and isoquinolinel8 [slightly related to (I)] when dissolved in acetone change by up to ca.2 ppm on addition of water, but 50% by volume of water is required to do this. For a few percent (e.g. 2.5%) of water the changes are of the order of 0.2 ppm. Although solvatochromic indicators such as (I) have been suggested as a means of determining the water content of organic solvents, the estimation is based upon an empirical equation obtained for certain mixtures, and presumably with the indicator in the concentration3108 Solvatochromism of a Pyridinium Betaine 1 152 . \ ‘- X 10 20 30 40 50 E Fig. 4. The variation of 13C chemical shift (6,) with solvent polarity (dielectric constant E ) for atoms f, m and h: 0, CDC1,; A, [2H,]acetone; 0, CD,OD; x , [‘%,]DMSO. Fig. 5. Changes in chemical shifts (Ad) for (I) in [2H,]DMS0 ( x ) and CD,OD (0) produced by addition of D,O (1 : 1).range 10-3-10-4 mol dm-3 for visible spectroph~tometry.~~ The samples of pure deuterated solvents used in this n.m.r. study were freshly opened samples, and although they were not dried any further we are confident that by comparison with other n.m.r. work18 any traces of water played only a small role compared with the bulk solvent effect in the observed solvent-induced chemical-shift changes. We also measured the effect upon the 13C spectrum of the betaine dissolved in 50 : 50J. G. Dawber and R. A . Williams 3109 Table 4. Changes in 13C chemical shifts in various media change, Ad, from CDC1, line position 100% 100% 100% atom no. in CDC1, (PPm) (PPm (PPm) carbon line (PPm) [2H,]DMS0 [2H,]acetone CD,OD a b d e f g h i j k 1 m n P 9 r C 0 S 17 16 12 8 4 1 15 3 18 13 11 6 2 14 19 5 9 10 7 126.40 127.80 129.11 130.26 138.7 161.3 128.10 155.70 126.00 128.78 129.40 133.70 156.90 128.51 124.23 134.20 130.19 129.8 1 132.0 - 0.90 -0.59 -0.38 - 0.79 1 .oo -0.30 -0.10 - 1.70 - 0.93 -0.18 - 0.45 -- 0.20 - 0.40 - 0.02 - 0.23 - 0.20 - 0.72 - 0.34 0.00 0.2 1 0.45 0.78 0.84 2.30 1.60 0.15 0.40 0.55 0.48 0.6 1 1.30 1.10 0.66 0.97 0.90 0.42 0.8 1 1 .oo 1.58 0.68 1.10 2.64 2.90 3.10 1.29 1.50 0.82 0.72 1.04 1.80 1.80 0.99 1.68 1 S O 0.8 1 1.08 1.20 carbon line atom no. change from change from excess chemical shift, d,, in CD,OH in DMSO in 50 : 50 m o l x mixtures 50: 50 mixture 50: 50 mixture a 17 b 16 c 12 d 8 e 4 f 1 g 15 h 3 i 18 j 13 k 11 1 6 m 2 n 14 o 19 P 5 q 9 r 10 S 7 with D,O Ad (PPm> -0.45 0.48 0.36 0.52 7.00 - 6.00 - 0.23 - 5.90 - 0.22 0.68 0.45 - 0.40 - 1.40 0.00 - 0.57 4.80 1.59 0.22 1.70 with D,O Ad (PPO CDC1,- DMSO (PPm) - 0.37 - 0.06 -0.36 0.14 0.90 3.60 0.05 -0.10 0.06 - 0.04 -0.12 0.10 0.10 - 0.20 - 2.24 0.10 - 0.06 - 0.06 0.00 - 0.24 -0.21 -0.12 - 0.27 0.30 0.15 - 0.05 - 0.25 -0.14 0.00 0.02 0.10 -0.20 -0.14 - 0.02 -0.20 -0.30 -1.11 0.00 CDC1,- CD,OD (PPm) - 0.68 - 0.23 -0.14 0.72 0.05 1.35 0.05 -0.35 - 0.32 - 0.04 -0.15 - 0.60 -0.10 -0.10 - 1.20 -0.75 0.17 0.00 0.30 CDCI,- ace tone (PPm) DMSO- acetone (PPm) -0.30 -0.33 -0.10 -0.18 0.35 1.90 0.35 - 0.30 -0.31 - 0.09 0.15 0.25 -0.15 - 0.04 - 1.46 - 0.05 - 0.23 -0.14 -0.10 0.23 -0.13 - 0.09 - 0.25 0.45 1.15 0.44 -0.35 - 0.02 0.32 0.13 0.05 -0.15 0.07 - 1.01 0.35 0.00 0.00 -0.1031 10 Solvatochromism of a Pyridiniurn Betaine @ CDC13 [ 2H6]DMS0 Fig.6. Preferential solvation profile in the CDCl,-[2H,]DMS0 mixture. @ CDC13 CD30D Fig. 7. Preferential solvation profile in the CDC1,-CD,OD mixture. [2H,]DMSO-D,0 caused by the addition of H+ ion and OD- ion (by small additions of concentrated HCl and concentrated NaOD). In the case of added acid there were, in general, downfield shifts of 1-2 ppm. However, atom C , experiences a large downfield Adc of 9.7 ppm, and Cf and C, experienced upfield changes of shifts of - 1.2 and - 0.5 ppm. Addition of NaOD to the betaine produced large upfield changes for C, and C , (- 10.1 and - 7.8 ppm) which might be expected in view of their close proximity to the N+ centre of the betaine. Thus the effects of acid and base are considerable when appreciable amounts are added.The infinitesimally small amount of acidity present in the CDCl, (see Experimental section) is unlikely to have affected the 13C chemical shifts in this solvent.J. G. Dawber and R. A . Williams CDClj ['H6 ]acetone 3111 Fig. 8. Preferential solvation profile in the CDC1,-[2H,]acetone mixture. @ [ *H6 lacetone @ [ 2 ~ , ~ ~ ~ ~ ~ Fig. 9. Preferential solvation profile in the [2H,]acetone-[2H,]DMS0 mixture. For the 50: 50 mol % mixtures of organic solvents it is possible to test for preferential solvation by calculating an excess chemical shift, 6,, defined as: 6 E = 6,-0.5(6,+6,) (1) where 6, and 6, are the resonances of a given carbon atom in the pure solvents and 6, is the corresponding resonance in the 50: 50 solvent mixture.The values of A are given in table 4 where, for the systems involving CDCI,, a negative value for dE corresponds to a given line in the solvent mixture being closer to its value in pure CDCI,, (i.e. preferential solvation by CDCI,), whereas a positive value of 6, corresponds to31 12 Solvatochromism of a Pyridinium Retaine preferential solvation at that position of the molecule by the other solvent component. For the DMSO-acetone solvent mixture a negative value of B E [from eqn (l)] corresponds to preferential solvation at a given position by the DMSO. From these data graphs similar to fig. 5 were plotted, namely, 6, against carbon atom position, and from such graphs were constructed the solvation profiles of the betaine in the solvent mixtures; these are shown in fig.6-9. Clearly this approach is a simplification, but it does illustrate the possibility of preferential solvation, For the organic solvent mixtures involving CDC1, (fig. 4-8) it appears that preferential solvation by the other more polar component occurs within the vicinity of the CO- centre, whereas the N+ centre seems to prefer solvation by the CDCl,. For the DMSO-acetone mixture (fig. 9) the CO- centre appears to favour acetone, whereas the N+ centre prefers solvation by DMSO in the solvent mixture. The latter finding is similar to the result for the DMSO-D,O solvent mixture. Conclusion From the lH and 13C n.m.r. studies of (I) in various solvent systems it can be seen that, owing to solvation phenomena, the polarity of the medium influences the positions of the n.m.r.signals. The sites in the molecule most influenced by the solvent are those centres nearest to the positive and negative charges within the molecule, and this is seen in the n.m.r. chemical shifts. The resulting differences in electron density near the ionic sites, produced by solvation differences, are transmitted to more distant sites of the molecule by the extensively delocalised n-electron system of this remarkable solvato- chromic compound. By means of an excess chemical shift, B E , it is possible to obtain preferential solvation profiles in mixed solvent systems, although this approach is likely to be a simplified concept. The helpful comments of a referee and also Dr J. W. Akitt, University of Leeds, are gratefully acknowledged.References 1 K. Dimroth and C. Reichardt, Palette No. ZI (Sandoz AG, Basel, Switzerland); C . Reichardt, Angew. Chem.. Int. Ed. Engl. 1965, 4, 29; C . Reichardt, Solvent Efsects in Organic Chemistry (Verlag Chemie, Weinheim, 1979). 2 K. Dimroth, C. Reichardt, T. Siepmann and F. Bohlmann, Liebigs Ann. Chem., 1963, 661, 1. 3 A. Botrel, A. Le Benze, P. Jacques and H. Strub, J . Chem. Soc., Faraday Trans. 2, 1984, 80, 1235. 4 K. Dimroth, G. Arnoldy, S. von Eichen and G. Schiffler, Liebigs Ann. Chem., 1957, 604, 221. 5 K. Dimroth, Angew. Chem., 1960,72, 331. 6 H. B. Bull, C. A. Soch and G . Oenslager, J . Am. Chem. Sac., 1900, 24, 1. 7 J. W. Akitt, N.M.R. and Chemistry (Chapman and Hall, London, 2nd edn, 1983). 8 H. G. Benson and J. N. Murrel, J. Chem. Soc., Faraday Trans. 2, 1972, 68, 137. 9 R. L. Lichter and J. D. Roberts, J . Chem. Phys., 1970, 74, 912. 10 R. G. Wilson, J. H. Bowie and D. H. Williams, Tetrahedron, 1968, 24, 1407. 1 1 See for example J. G. Dawber, J . Chem. Soc., Faraday Trans. I , 1984,80,2133; 1978,74, 1702; 1709; 1979, 75, 370, and references therein. I2 H. Gunther, N.M.R. Spectroscopy (Wiley, Chichester, 1980). 13 F. W. Wehrli and T. Wirthlin, Interpretation of Carbon-I3 N.M.R. Spectra (Wiley, Chichester, 1983). 14 J. D. Memory and N. K. Wilson, N.M.R. of Aromatic Compounds (Wiley, Chichester, 1982). 15 G. L. Nelson, G. C . Levy and J. D. Cargioli, J. Am. Chem. SOC., 1972, 94, 3090; G. C. Levy, R. L. Lichter and G. L. Nelson, Carbon-I3 Nuclear Magnetic Resonance Spectroscopy (Wiley, New York, 2nd edn, 1980). 16 G. C. Levy, J. D. Cargioli and F. A. L. Anet, J . Am. Chem. SOC., 1973,95, 1527. 17 R. Radeglia, W. Sieffert, G . Hohlweicher, C. Jutz and H. L. Springer, Tetrahedron Lert., 1966,41,5053, R. Radegli, G. Engelhardt, E. Lipmaa, T. Pelk, K. D. Nolte and S . Dahne, Org. M a p . Reson., 1972, 4, 571 ; S. Dahne and K. D. Nolte, J . Chem. Soc., Chem. Commun., 1972, 1056. 18 E. Breitmaier and K. H. Spohn, Tetrahedron, 1973, 29, 1145. 19 H. Langhals, Angew. Chem., Int. Ed. Engl., 1982, 21, 724. Paper 512030; Receiced 18th Nocember, 1985

ISSN:0300-9599    

DOI:10.1039/F19868203097

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

J . Chem. Soc., Faraday Trans. I , 1986, 82, 3113-3123 Reactions of Hydrocarbons on Alumina-supported Pt-Ir Bimetallic Catalysts Part 1 .-Exchange of Methane and Cyclopentane with Deuterium, and Hydrogenolysis of Butane, Pentane and Cyclopentane David Garden, Charles Kernball" and David A. Whan-f- Department of Chemistry, University of Edinburgh, West Muins Road, Edinburgh EH9 3JJ, Scotland Exchange reactions with deuterium of cyclopentane (303 K) and of methane (473 K) and the hydrogenolysis of n-butane, n-pentane and cyclopentane have been studied on a series of alumina-supported Pt-Ir catalysts. Metal composition had little effect on the activity of the catalysts for the exchange reactions but influenced the character of the reactions in a regular manner. Rates of hydrogenolysis increased with iridium content, and the change was > 100 for n-butane but only 10 for cyclopentane.The main reaction with cyclopentane was ring-opening, but the extent of further hydrogenolysis increased with iridium content in a manner that suggested the possible requirement of ensembles of three iridium atoms. Interest in supported platinum-iridium catalysts has developed since they were shown1 to have advantages over conventional alumina-supported platinum catalysts for hydro- carbon re-forming, mainly associated with greater and longer-lasting activity. These catalysts are also important from a fundamental point of view since they belong to the class of Group VIII-Group VIII bimetallic catalysts which have perhaps received less attention in the catalytic literature than Group VIII-Group IB combinations.The manner in which catalytic activity and selectivity for different hydrocarbon reactions varies with composition of the bimetallic catalysts can provide a useful guide to the relative importance of electronic factors or ensemble size in determining the properties of the catalysts. The literature on supported iridium catalysts is not as substantial as that on the corresponding platinum catalysts, but it is well known that iridium is more active for hydrogenolysis than platinum. Sinfelt2 found iridium to be lo5 times more active for the hydrogenolysis of ethane at 478 K, and rather lower but still substantial differences between the rates of hydrogenolysis of n-butane and cyclopentane have been rep~rted.~ Comparatively few investigations have been made on the catalytic properties of platinum-iridium bimetallics.The work of the Exxon group has been brought together in a recent book,* R a s ~ e r ~ . ~ has studied re-forming reactions with n-hexane and n-heptane, Bernard et aZ.7 have shown that alumina-supported platinum-iridium is effective for the selective hydrogenolysis of alkanes to ethane and Leclercq et a1.8 have studied the hydrogenation of benzene, the hydrogenolysis of cyclopentane and the dehydrogenation of 1,1,3-trirnethylcyclohexane on catalysts of various compositions supported on a-alumina. The objective of our work was to study exchange reactions of a number of hydrocarbons with deuterium and also hydrogenolysis and isomerisation of hydrocarbons at higher temperatures over catalysts with a range of compositions, in order to define Cleveland TS23 1LB.t Present address: Agricultural Division, Imperial Chemical Industries plc, P.O. Box 1, Billingham, 311331 14 Reactions of Hydrocarbons on Pt-Ir Catalysts as clearly as possible how activities and mechanisms varied across the series from platinum to iridium. It was hoped that useful information would be gained from the variations of selectivity for the higher-temperature reactions and also of the character of the exchange reactions, i.e. the nature and the extent of multiple exchange, with metal composition. One of the major questions of interest about supported platinum-iridium catalysts is the extent to which the component metals constitute a single phase (alloy or bimetallic cluster) or whether some form of phase separation occurs either during preparation or subsequent treatment.The literature on this question has been reviewed in a paperg which describes the preparation and characterisation of the alumina-supported catalysts used in the present work. It was found that the application of low metal loadings to an alumina support gave well dispersed platinum-iridium catalysts with good interaction between the metals. There was evidence that an initial calcination at ca. 773 K during catalyst preparation was beneficial in helping to form catalysts with high metal dispersions, but calcination at this temperature after the metals had been reduced gave rise to phase separation in agreement with the results of other workers.1°-12 In this part, the results for a number of exchange and hydrogenolysis reactions each carried out at one selected temperature are presented.Other reactions studied over a range of temperatures will be reported subsequently. Experiment a1 The preparation of the series of Pt-Ir catalysts containing 0.5 wt% of metal supported on y-Al,O, and their characterisation by hydrogen adsorption, temperature-programmed reduction and temperature-programmed desorption of hydrogen have been described. Samples of catalyst for use (50 mg) were evacuated, raised to 773 K over 3 h in flowing hydrogen, maintained at that temperature for 16 h and then cooled to 573 K and evacuated. The hydrocarbons methane and n-butane were obtained from the Matheson Company, n-pentane from D.S.I.R.and cyclopentane from B.D.H. ; all were 99.9% purity. Deuterium (99 % ) from Matheson was purified by diffusion through a palladium- silver thimble. The hydrogenolysis experiments were carried out in a static system connected by three-way stopcocks and a sampling valve to a Perkin-Elmer F33 gas chromatograph fitted with a flame ionisation detector. The reaction vessel (1 83 cm3) was made of silica glass and contained a quartz sinter to hold the catalyst. Connections to the reaction vessel enabled flowing hydrogen to be used for reducing the catalyst. Gas chromato- graphic analyses were carried out using a 4 m column of bis-2-methoxyethyladipate (1 3.5 % ) and di-2-ethylhexylsebacate (6.5 % ) on Chromosorb P. The exchange experiments were also carried out in a static system in one of a pair of reaction vessels (260 cm3) each linked to an AEI MSlO mass spectrometer by a capillary leak.Some problems were encountered from the water background in the mass spectrometer during the experiments on the exchange of methane with deuterium. Peaks of significant size for m/e = 16 to 18 were observed owing to the water background, and during the course of running exchange experiments, the residual water in the spectrometer exchanged with deuterium and peaks corresponding to m/e = 19 and 20 gradually increased in size. Two special precautions were taken to minimise this trouble. First, the mass spectrometer was baked out overnight with a pressure of 8 kPa of hydrogen in the duplicate reaction vessel, which therefore provided a continuous bleed of light hydrogen into the spectrometer.This helped to reduce the water background in the spectrometer and ensured that such background was mainly due to light water rather than deuterated water. Secondly, a pressure of light hydrogen was maintained in the duplicate reaction vessel while the experiments on the exchange of methane with deuterium were being carried out in the first reaction vessel. This eliminated the shift in the water backgroundD. Garden, C. Kemball and D. A . Whan 31 15 Table 1. Turnover numbersa and values of Mb for the exchange reactions cyclopentane methane at at 303 K 473 K catalyst M RC238 RC239 RC240 RC24 1 RC242 RC243 RC244 0 19 39 46 54 73 100 6.5 4.0 1.4 4.0 3.1 1.2 3.9 3.2 1.1 4.4 2.9 1.3 4.3 2.9 0.69 4.5 2.7 1.1 4.7 1.9 0.78 1.1 1.1 1.2 1.2 1.3 1.3 1.4 a Turnover numbers are given in molecule s-l site-l.M represents the mean number of deuterium atoms entering each reacting hydro- carbon molecule under initial conditions. from peaks corresponding to m / e = 16 to 18 towards m / e = 19 and 20; such a shift was always observed when the reaction mixture was bled into the mass spectrometer with the duplicate vessel evacuated. The supply of light hydrogen from the duplicate vessel to the mass spectrometer simultaneously with the bleed of methane and deuterium from the reaction vessel maintained the background peaks effectively constant during the course of the exchange experiment. Thus it was possible to correct the observed peaks in the range from m/e = 16 to 20 for the contribution from the water background and obtain reasonably accurate analyses of the isotopic methanes throughout the course of the exchange reaction.Analyses were made using 20 eV electrons to ionise the molecules and the peak heights were corrected for naturally occurring heavy isotopes and for fragmentation in order to obtain the amount of the various isotopic hydrocarbons. Standard methods13 were used to calculate the initial rates of exchange of the hydro- carbons and the mean number of deuterium atoms (M) entering each molecule react- ing under initial conditions. All rates were expressed as turnover numbers assuming that the number of metal sites on the catalysts was equivalent to the number of H atoms adsorbed. Results Exchange Reactions Exchange reactions of cyclopentane with deuterium at 303 K and of methane with deuterium at 473 K were carried out over the whole series of platinum-iridium catalysts. Reaction mixtures consisted of a 10: 1 ratio of deuterium to alkane; the pressure of alkane was 800 Pa, corresponding to a charge of 4.6 x 1019 molecules of hydrocarbon in the reaction vessel.The reactions exhibited the usual first-order-kinetic behaviour, and the initial rates of reaction as turnover numbers and the values of M are given in table 1. The exchange of methane was mainly stepwise over platinum and the relative percentages of the initial products D, to D, were 87, 9, 3 and 1 %, respectively. There was a small increase in the extent of multiple exchange with iridium content, and over the 100% iridium catalyst the amounts of the initial products were 79, 14, 4 and 3%.More significant trends were found in the percentages of the various isotopic 103 F A R 131 16 Reactions of Hydrocarbons on Pt-Ir Catalysts Table 2. Initial product distributions for the exchange of cyclopentane at 303 K -~ percentages catalyst D, D, D, D, D, D, D, D, D, D,, RC238 25 8 8 14 27 4 3 3 4 4 RC239 38 11 9 11 18 4 2 2 2 3 RC240 35 11 7 10 25 3 2 2 2 3 RC241 45 13 8 7 16 3 2 2 2 2 RC242 44 15 7 7 15 3 2 2 2 3 RC243 49 16 7 6 14 2 2 1 1 2 RC244 51 25 8 6 6 1 1 1 0.5 0.5 ~~ ______ ~ ~- Table 3. The hydrogenolysis of n-butane at 473 K turnover productsb number,a catalyst NB/ s, S , s, s; MB R4C RC238 1.9 0.40 1.38 0.28 0.002 I .06 0.005 RC239 71 0.17 1.37 0.12 0.002 1.02 0.10 RC240 57 0.22 1.67 0.13 0.002 1.02 0.1 1 RC24 1 93 0.25 1.65 0.14 0.002 1.04 0.25 RC242 140 0.27 1.64 0.15 0.002 1.06 0.55 RC243 230 0.30 1.60 0.16 0.004 1.06 0.63 RC244 280 0.50 1.53 0.15 0.004 1.18 1.8 ~~ -~~~ - ~~~~ ~~ a NB is the number of n-butane molecules reacting s-' site-'.S,, S , and S, give, respectively, the number of methane, ethane and propane molecules formed for each molecule of butane reacted; Si refers to the formation of 2-methylpropane. R, is the ratio of the rate of methane formation to the rate of methane exchange on the same catalyst at 473 K. cyclopentanes with the composition of the catalyst and the full initial product distributions are shown in table 2. Multiple exchange was extensive with cyclopentane particularly over platinum-rich catalysts.Hydrogenolysis Reactions The hydrogenolysis experiments were carried out at 473 K using a 10: 1 ratio of hydrogen to hydrocarbon; the pressure of alkane was 800 Pa, corresponding to a charge of 3.0 x 1019 molecules in the reaction vessel. The course of the reaction was followed by the decrease in the concentration of the reactant expressed as a percentage of the total carbon present in the vessel. Initial rates of reaction were estimated either from direct plots of this quantity against time, for reactions which approximated to zero order; or from the logarithm of the quantity, where first-order behaviour was observed. Selectivities of the reactions were determined by measuring the parameters Si, the number of molecules of the hydrocarbon with i carbon atoms produced initially from each molecule of reactant destroyed.From these parameters it was then possible to calculate MB = si- 1 (1) i where MB indicates the depth of hydrogenolysis. A value of MB of unity shows a stepwise hydrogenolysis with the cleavage of only one C-C bond for each molecule ofD. Garden, C. Kemball and D. A . Whan 31 17 Table 4. The hydrogenolysis of n-pentane at 473 K turnover products number, catalyst Np/10-4 S, s 2 s 3 s4 s; s; MI3 R.5" RC238 5.9 0.29 0.83 0.78 0.18 - - 1.08 0.012 RC240 69 0.17 0.99 0.88 0.05 - - 1.09 0.1 1 RC241 140 0.09 1.10 0.85 0.03 - 0.008 1.07 0.10 RC243 270 0.30 0.90 0.84 0.07 - 0.017 1.09 0.74 RC244 370 0.49 1.04 0.62 0.09 0.005 0.018 1.23 2.3 a R, is the ratio of the rate of methane formation to the rate of methane exchange on the same catalyst at 473 K.Table 5. The hydrogenolysis of cyclopentane at 473 K products catalyst N ~ ~ / 10-4 RC238 RC239 RC240 RC24 1 RC242 RC243 RC244 8.9 40 51 55 67 83 93 0.020 0.038 0.070 0.065 0.099 0.133 0.412 - 0.047 0.098 0.073 0.134 0.152 0.360 s3 - 0.061 0.078 0.085 0.104 0.109 0.206 - 0.009 0.021 0.027 0.017 0.041 s; s.5 - 0.996 0.937 - 0.893 - 0.891 - 0.843 0.837 0.002 0.615 - - ~~ 1.016 1.883 1.148 1.135 1.207 1.228 1.634 reactant destroyed, e.g. the hydrogenolysis of an n-butane molecule to form two molecules of ethane or one of methane and one of propane. Higher values of MB indicate the degree of multiple hydrogenolysis and correspond to the number of C-C bonds broken for each molecule of reactant consumed. Details for the hydrogenolysis of n-butane which showed an approximate first-order reaction are given in table 3.The parameter Si refers to isomerisation to 2-methylpropane which occurred on all catalysts but only to a limited extent. Little more than stepwise hydrogenolysis occurred on most of the catalysts except for 100% iridium. All catalysts showed preferential cleavage of the central bond, particularly those of intermediate composition, which gave relatively more ethane than the monometallic catalysts. A further parameter included in table 3 is R,, the ratio of the rate of formation of methane from the hydrogenolysis of n-butane to the rate of methane exchange at 473 K. Results for the hydrogenolysis of n-pentane which was followed on five of the seven catalysts are given in table 4.As with n-butane, the reaction was approximately first-order, and the main process was stepwise hydrogenolysis with cleavage of one of the central C-C bonds to give ethane and propane as the principal products. The course of the hydrogenolysis of cyclopentane at 473 K exhibited approximate zero-order kinetics as the percentage of reactant decreased steadily with time, The main reaction over platinum was ring-opening to form n-pentane, and there was little further hydrogenolysis to smaller hydrocarbons. However, as the iridium content of the catalysts increased the extent of the multiple hydrogenolysis increased and rose sharply for the 100% iridium catalyst. For cyclopentane the parameter M , is defined simply as C, S, and not as in eqn (1). The rates of reaction and the selectivities are given in table 5.103-231 18 Reactions of Hydrocarbons 75 on Pt-Ir Catalysts 50 - 1 100 200 300 t/min Fig. 1. Hydrogenolysis of a mixture of cyclopentane and n-pentane on 100% iridium at 473 K: 0, cyclopentane; 0, n-pentane; a, total of smaller hydrocarbons. The approximate zero-order kinetics observed with cyclopentane suggested that it was strongly adsorbed on the catalysts at 473 K. In orders to investigate this point further an experiment was carried out on the hydrogenolysis of a mixture of cyclopentane and n-pentane with a combined pressure of 800 Pa and the usual 10: 1 ratio of hydrogen to hydrocarbon on the 100% iridium catalyst. The results are shown in fig. 1 . The presence of cyclopentane effectively inhibited the reaction of n-pentane, which only became apparent after the cyclopentane had fallen to < 5% of the total carbon content of the reaction vessel.Since the rate for n-pentane was four times the rate for cyclopentane in the usual runs with a single hydrocarbon reactant (see tables 4 and 5 ) , it is clear that the cyclopentane is more strongly adsorbed than n-pentane at 473 K on the 100% iridium catalyst. Discussion Previous workQ has established that the catalysts used in this work were highly dispersed and therefore the size of the metal crystallites was small. The form and position of the first peak in the temperature-programmed reduction experiments suggested that the bimetallic catalysts consisted of a single metal phase, with properties intermediate between those found for the two monometallic catalysts.Phase separation occurred if the catalysts was calcined at 773 K. The single-phase character of the catalysts could be restored by reduction and subsequent calcination at 523 K, but some loss of metal area was associated with phase separation and its reversal. It is not possible apriori to exclude the possibility of some surface enrichment by platinum in the bimetallic catalysts. The heats of sublimation of platinum and iridium are 5 1 1 and 692 kJ molt1, respectively, and so platinum is the component which should segregate to the surface.'* Such enrichmentD. Garden, C. Kernball and D. A . Whan 31 19 was detected by Auger electron spectroscopy for the alloy catalysts used by Rasser et aE.8 and which contained large ( > 25 nm) metal particles.High-temperature l6 of the gasification or the hydrogenation of graphite by bimetallic particles of platinum and iridium have provided evidence for platinum surface enrichment in a reducing atmosphere but iridium enrichment in an oxidising atmosphere ; these results were obtained for metal particles of substantial size. For small metal crystallites, where the proportion of surface atoms to bulk atoms is high, the extent of surface enrichment may be limited by the significant depletion of the concentration of the surface-active component in the 'bulk' phase, as has been discussed in connection with bimetallic rhodium-iridium cata1ysts.l' For these reasons we believe that any surface enrichment by platinum for the present catalysts may not be appreciable.Exchange Reactions The results given in table 1 for the turnover numbers for the two exchange reactions across the series of catalysts show little change in catalytic activity with metal composition. These results confirm that we have a consistent series of catalysts but provide no evidence in support of the bimetallic character of the catalysts or for the factors which might determine their activity. The fact that similar rates of exchange were found for cyclopentane at 303 K and for methane at 473 K is in broad agreement with earlier work13 on the relative ease of exchange of the two hydrocarbons with deuterium. The main factor which accounts for the large difference in reactivity is undoubtedly the difference in the C-H bond dissociation energies, which are 397 and 431 kJ mol-l for cyclopentane and methane, respectively.ls The gradual increase in the multiplicity of the methane exchange with iridium content shows that the part played by adsorbed methylene species (a,a-diadsorbed methane) increases with iridium content.This change is in the expected direction, since iridium is a better hydrogenolysis catalyst than platinum and there tends to be a correlation between hydrogenolysis activity and the ability to form a,a-diadsorbed species.lg The initial product distributions for cyclopentane exchange, set out in table 2, exhibit a number of different trends which are influenced by catalyst composition. Before considering these results it is helpful to recall the various mechanisms which have been established for the exchange of cyclopentane with deuterium.For palladium catalysts five distinct mechanisms were identified by Schrage and Burwell,20 and four of these are relevant to the present work. They are as follows. (a) The reversible formation of adsorbed cyclopentyl radicals which return to the gas phase without appreciable side reactions and give rise to the D, product. ( b ) The reversible formation of adsorbed cyclopentene species which return to the gas phase from part of the catalyst as D, product. (c) A mechanism which takes place on other parts of the catalyst and involves the interconversion of the adsorbed species described in (a) and (b). This is the so-called @-process (interconversion of a-adsorbed and a,p-diadsorbed alkane), and when it takes place readily it will give rise to D, and D, products involving the replacement of the hydrogen atoms on one side of the C , ring.( d ) The interchange reaction21 which permits the molecules to turn over on the catalyst surface through a vertically oriented alkene intermediate, possibly only physically adsorbed, and gives rise to products with from six to ten deuterium atoms. This mechanism can occur on the same sites as those involved in (c). The results in table 2 show that all of these four processes take place on all the catalysts in the series, and this fact is evidence that there are probably three different types of active sites on each catalyst. The relative importance of the four processes is shown by plotting the appropriate data from table 2 in fig.2. Processes ( c ) and ( d ) occur to substantial extents on 100% platinum and decline more or less linearly (ignoring some experimental error) with increasing iridium content. Conversely, processes ( a ) and (b),3120 Reactions of Hydrocarbons on Pt-Ir Catalysts 0 50 Ir (%) 100 Fig. 2. Variations of initial products for the exchange of cyclopentane with catalyst composition : 0, D,-cyclopentane; 0, D,; U, D,+D,; m, D,-D,,. although operative over 100% platinum, occur to increasing extents as the percentage of iridium increases. In none of the four cases is there any substantial departure from a linear variation with metal composition. So while the catalytic properties are varying in a recognisable and relatively uniform manner with metal composition, it is not possible to make any useful deductions about the relative importance of electronic factors or metal ensembles from these exchange processes with cyclopentane.There is no evidence to suggest the need for sites involving more than one or two metal atoms for any of the processes. Changes in metal composition might be causing variations in the proportions of the different types of sites or changes in their relative activity. There has been some uncertainty about the nature of the process (b) giving rise to the D,-product. At one stage the product was thought to be 1,l -dideuterocyclopentane, resulting from the reversible formation of an apdiadsorbed species; this is the mechanism favoured by van Broekhoven and Ponec2, in a recent study of exchange of cyclopentane and of methane on a number of supported noble-metal catalysts.believe that the D,-product is 1,2-dideuterocyclopentane formed through the reversible dissociation of the alkane to alkene on sites which do not permit the usual a,p-process to occur. The strongest evidence in support of this interpretation is from the ,D n.m.r. spectra of the exchanged cyclopentanes formed over supported rhodium and iridium catalyst^.,^ The spectra showed the formation of substantial quantities of 1,2- dideuterocyclopentane as an initial product over rhodium and significant amounts over iridium. We agree with van Broekhoven and Ponec that both rhodium and iridium are metals which can form a,a-diadsorbed alkane species, but only at higher temperatures and not at the comparatively low temperatures used for cyclopentane exchange in the present work.0.Garden, C. Kemball and D. A . Whan 3121 Ir (%) hydrogenolysis of n-butane ; 0, hydrogenolysis of cyclopentane. Fig. 3. Variation of turnover numbers (molecule s-l site-') with catalyst composition: 0, Hydrogenolysis Reactions The rates of all three hydrogenolysis reactions increased regularly with the iridium content of the catalysts. The largest difference between the two metals was found for n-butane, which reacted more than 100 times faster over iridium than over platinum, but the increase in activity was approximately proportional to iridium content, as shown in fig. 3. There was similar behaviour with n-pentane, for which iridium was some 60 times more active than platinum.In contrast, the increase in activity across the series was only 10-fold for cyclopentane; the plot of activity against composition was slightly curved as shown in fig. 3, and there was no evidence for a maximum in the curve such as that reported by Leclercq ert aL8 for their a-alumina-supported catalysts at a composition of Ir/Pt of 75/25. These results, while confirming the expected greater activity of iridium as a hydrogenolysis catalyst, do not provide any evidence about the nature of the bimetallic catalysts. Little isomerisation was found with any of the systems, and the main reaction with the two straight-chain hydrocarbons was stepwise hydrogenolysis, except for the 100 % iridium catalysts, which showed some evidence for multiple carbonxarbon bond rupture as indicated by the values of M , in tables 3 and 4.Preferential central-bond rupture occurred with n-butane on all catalysts, but it was enhanced for those of ca. 50: 50 Composition. This was indicative of some bimetallic interaction. Foger and Anderson,, have demonstrated that for supported iridium catalysts the degree of preferential central-bond rupture with n-butane is markedly dependent on crystallite size, being greater with small metal particles. The high preferential central-bond rupture observed in our work was to be expected, since all our catalysts were highly dispersed and hence of small crystallite size. The selective formation of ethane and propane from n-pentane on all catalysts is consistent with the results for n-butane and shows that, to use the terminology suggested by Foger and Anderson, CII-CIT bond rupture is easier than CI-CII bond rupture for both hydrocarbons.We have no reason to doubt that the mechanism is essentially that described2, as the C,-unit mode. The values for R, and R, in tables 3 and 4 provide evidence that on most of the catalysts the rate-determining step is associated with carbon-carbon bond-breaking and3122 Reactions of Hydrocarbons 5 - rn 0 0 on Pt-Ir Catalysts 50 Ir (%> 100 Fig. 4. Variation of the turnover number for multiple hydrogenolysis of cyclopentane with catalyst composition: the line is the turnover number for 100% Ir x (XI,)3. not with product desorption. The rate of methane exchange provides a measure of the rate of desorption of chemisorbed methane from the surface.We can be reasonably certain that the larger hydrocarbon molecules can desorb more rapidly than methane, since they are known to exchange with deuterium more rapidly on most metals.13 It follows that values of R < 1 indicate methane production rates for the hydrogenolysis reactions which are smaller than the rates of desorption in the exchange reaction and not therefore rate-determining. For the 100% iridium catalysts the values of R, and K, are ca. 1, and the rate of methane desorption may be no longer faster than carbon<arbon bond rupture. However, it is unlikely that product desorption is the rate-controlling reaction even with 100 % iridium. When product desorption is rate-controlling, values of R > 10 are usually found and the extent of the multiple hydrogenolysis becomes pronounced, as observed for the hydrogenolysis of n-butane at 545 K over a supported rhenium catalyst.25 The selectivity of the cyclopentane hydrogenolysis, shown in table 5, is interesting. On 100% Pt the reaction was limited to ring-opening, with negligible formation of products other than n-pentane. However, the extent of further hydrogenolysis increased with iridium content, particularly towards 100% Ir. The values of MB were used together with the turnover numbers to derive turnover numbers, NCM = NcH(MB- l), from the multiple hydrogenolysis reaction. The values of Nc3T are plotted in fig. 4, together with a curve which is proportional to (X1r)3, where XI, is the mole fraction of iridium in the catalysts. The general agreement between the points and the curve suggests that a simple interpretation of the results is to assume that multiple hydrogenolysis depends on having an ensemble of three iridium atoms, and that platinum ensembles or bimetallic ensembles have negligible activity.The variation of the extent of the multiple hydrogenolysis of cyclopentane with catalyst comparison is the only evidence in the present work which seems to require a definition of the catalyst sites in terms of an ensemble of atoms. Finally, the evidence both from orders of reaction and the results of the simultaneous reaction of cyclopentane and n-pentane shows that the cyclic molecule is more strongly adsorbed than the alkanes at 473 K. It is possible that a strongly adsorbed cyclopenta- dienyl species may form at the higher temperatures required for hydrogenolysis but not at room temperature, at which the catalysts show good activity for cyclopentane exchange with no evidence of self-poisoning due to strongly held species.D.Garden, C. Kemball and D. A . Whan 3123 D. Garden held a CASE studentship. We are grateful to the Petrochemicals and Plastics Division of ICI plc for assistance with the preparation and characterisation of the catalysts. We had helpful advice from Drs M. A. Day, A. C . Far0 Jr and R. J. Sampson. References 1 J. H. Sinfelt, U.S. Patent 3,953,368 (1976). 2 J. H. Sinfelt, Adv. Catal., 1973, 23, 91. 3 C. Betizeau, G. Leclercq, R. Maurel, C. Bolivar, H. Charcosset, R. Frety and L. Tournayan, J . Catal., 4 J. H. Sinfelt, Bimetallic Catalysts-Discooeries, Concepts and Applications (Wiley, New York, 1983), 5 J. C. Rasser, Ph.D. Thesis (Delft, 1977). 6 J. C. Rasser, W. H. Beindorff and J. J. F. Scholten, J . Catal., 1979, 59, 21 1. 7 J. R. Bernard, J. Bousquet and P. Turlier, in Proc. 7th Znt. Congr. Catal.. ed. T. Seiyama and K. Tanabe 8 G. Leclercq, H. Charcosset, R. Maurel, C. Bertizeau, C. Bolivar, R. Frety, D. Jaunay, H. Mendez and 9 A. C. Far0 Jr, M. E. Cooper, D. Garden and C. Kemball, J . Chem. Res., 1983, (S) 110, (M) 11 14. 1976, 45, 179. pp. 86-120. (Elsevier, Amsterdam, 1981), vol. A, pp. 149-1 57. L. Tournayan, Bull. Soc. Chim. Belg., 1979, 88, 577. 10 J. H. Sinfelt and G. H. Via, J . Catal., 1979, 56, 1. 1 1 A. G. Graham and S. E. Wanke, J. Catal., 1981, 68, 1. 12 K. Foger and K. Jaeger, J. Catal., 1981, 70, 53. 13 C. Kemball, Adv. Catal., 1959, 11, 223. 14 J. R. Anderson, Structure of Metallic Catalysts (Academic Press, New York, 1979), p. 446. 15 R. T. K. Baker, R. D. Sherwood and J. A. Dumeric, J . Catal., 1980, 62, 221. 16 R. T. K. Baker, R. D. Sherwood and J. A. Dumeric, J . Catal., 1980, 66, 56. 17 I. H. B. Haining, C. Kemball and G. L. Haller, J . Chem. Soc., Faraday Trans. 1, 1981, 77, 2519. 18 J. G. Calvert and J. N. Pitts, Photochemistry (Wiley, New York, 1966), p. 824. 19 C. Kemball, Catal. Rev., 1972, 5, 33. 20 K. Schrage and R. L. Burwell Jr, J . Am. Chem. Soc., 1966,88,4549. 21 R. L. Burwell Jr, Acc. Chem. Res., 1969, 2, 289. 22 E. H. van Broekhoven and V. Ponec, J . Mol. Catal., 1984, 25, 109. 23 A. C. Far0 Jr, C. Kemball, R. Brown and I. H. Sadler, J . Chem. Res., 1982, ( S ) 342, (M) 3735. 24 K. Foger and J. R. Anderson, J . Catal., 1979, 59, 325. 25 I. H. B. Haining, C. Kemball and D. A. Whan, J . Chem. Res., 1977, (S) 342, (M) 3735. Paper 5/2 120; Received 4th December, 1985

ISSN:0300-9599    

DOI:10.1039/F19868203113

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. 1, 1986, 82, 3125-3139 Reactions of Hydrocarbons on Alumina-supported Pt-Ir Bimetallic Catalysts Part 2.-Exchange of Benzene with Deuterium, and Exchange and Hydrogenolysis of 2,2-Dime t hylpropane Arnaldo C. Far0 Jr and Charles Kernball" Department of Chemistry, University of Edinburgh, West Mains Road, Edinburgh EH9 3JJ Exchange reactions of benzene at 263 K and the exchange and hydrogenolysis of 2,2-dimethylpropane at a range of temperatures have been studied over a series of alumina-supported Pt-Ir catalysts. Neither the rate nor the character of the exchange reactions varied greatly with metal composition. The rates of hydrogenolysis of 2,2-dimethylpropane increased markedly with iridium content, and the shapes of the activitysomposition curves were interpreted in terms of active sites consisting of ensembles of four metal atoms.Mixed metal ensembles had an activity intermediate between those of the pure metals. The present results, together with those from the preceding paper, show that, for all hydrocarbons, the depth of hydrogenolysis increased with iridium content, probably owing to a rising strength of adsorption. With platinum, similar types of C-C bonds in different hydrocarbons underwent cleavage at similar rates ; secondary-secondary bonds were most easily broken and primary-quaternary were the least reactive. In Part 1, exchange reactions of cyclopentane and methane, and the hydrogenolysis of n-butane, n-pentane and cyclopentane were studied on a series of alumina-supported Pt-Ir cata1ysts.l Metal composition had little effect on the activity of the catalysts for the exchange reactions, but influenced the character of the reactions in a regular manner.In all cases, rates of hydrogenolysis increased with iridium content. With cyclopentane at 473 K, the main reaction was ring-opening, but the extent of further hydrogenolysis increased sharply with iridium content in a fashion that suggested the possible requirement of ensembles of three iridium atoms. Reactions with 2,2-dimethylpropane (2,2 DMP) have now been carried out over a range of temperatures in order to explore further the properties of these Pt-Ir catalysts. The exchange of this molecule with deuterium may occur by one or more of three processes which have been described in earlier Process I involves stepwise exchange through the reversible formation of an adsorbed alkyl radical.Process I1 gives multiple exchange, but limited to a single methyl group through interconversion between adsorbed alkyl radicals and a,a-diadsorbed (alkylidene) species. Process 111, which only becomes significant at higher temperature, leads to the formation of initial products containing deuterium atoms in two or more methyl groups; it requires the interconversion of adsorbed alkyl radicals and a, y-diadsorbed species. The hydrogenolysis of 2,2 DMP is also worthy of examination because the presence of the quaternary carbon atom necessitates a so-called iso-unit mode of reaction5 with the initial formation of a,y- diadsorbed species; such a species if attached to a single metal atom is a metallocyclobutane.6 The molecule cannot undergo hydrogenolysis by the C,-unit mode, which involves dissociation of hydrogen atoms from adjacent carbon atoms and is a common mechanism for the reaction of straight-chain hydrocarbons. Thus the reactions 31253126 Reactions of Hydrocarbons on Pt-Ir Catalysts of 2,2DMP provide a means of examining the catalytic properties of bimetallic systems in a way which is essentially different from the study of the corresponding reactions of straight-chain hydrocarbons.When this research was in progress, information was obtained7 about another study of the hydrogenolysis of 2,2DMP on a series of Pt-Ir catalysts not unlike those used in the present work but in a flow system at a temperature of 538 K.Benzene exchange with deuterium was also examined to see whether there were any special features about the activation of the aromatic C-H bond on the Pt-Ir catalysts. Some experiments were also included on the hydrogenolysis of 2-methylpropane (2 MP) and 2-methylbutane (2MB) on the monometallic catalysts in order to aid the interpre- tation of the results from Part 1 and the present investigation. The main objective of the work was to obtain data on the variations of activity and selectivity with metal composition suitable for determining the relative importance of electronic factors or ensemble size for the catalytic properties of the Pt-Ir bimetallics. Experimental The preparation and characterisation of the Pt-Ir catalysts containing ca. 0.5 wt % of metal supported on y-Al,O, have been described,8 and the pretreatment was given in Part 1 .l Deuterium, 2-methylpropane and 2,2-dimethylpropane were obtained from the Matheson Co., 2-methylbutane from B.D.H. Ltd and benzene from the Aldrich Chemical Co. Ltd; all were of 99% or greater purity. In general the apparatus and procedure were similar to those described in Part 1, and all experiments were carried out in static systems. For exchange, the reaction vessel (207 cm3) was linked to a Micromass MM-601 mass spectrometer by a capillary leak. Benzene was ionised using 16 eV electrons and 2,2DMP was analysed in terms of the fragment ions C,X;S (X representing H or D) formed by 70 eV electrons. Peak heights were corrected for naturally occurring isotopes and for fragmentation, and standard -methods2 were used to obtain the initial rate of exchange and the mean number of deuterium atoms, M, entering each reacting molecule under initial conditions.Gas- chromatographic analyses in the hydrogenolysis experiments were carried out using a column (length 3 m, diameter 3.2 mm) containing n-octane on 80-100 mesh Porasil C operated at room temperature with nitrogen at 50 p.s.i.g. (450 kPa) as carrier gas. Under these conditions all alkanes from methane to cyclopentane were separated in ca. 20 min. Results Exchange Reactions Reaction mixtures consisted of a 10: 1 ratio of deuterium to hydrocarbon and the pressure of the hydrocarbon was 0.58 kPa, corresponding to ca. 2.2 x 1019 molecules in the reaction vessel. Rates of exchange of benzene at 263 K are given in fig.1, and increased slightly with iridium content. The exchange was mainly stepwise, and product distributions on all catalysts were close to 93% D,, 6% D2 and 1 % D, with A4 = 1.1. Some deuteration giving cyclohexanes in the range C,H7D, to C&2 occurred simul- taneously with exchange on all catalysts. The rate could not be measured accurately, but it was ca. two orders of-magnitude smaller than the rate of exchange. The most prominent product of the addition reaction was C,H,D,, and the average number of deuterium atoms per molecule was between 7 and 8. The exchange of 2,2DMP was followed at 354 and 393 K on all catalysts and over a wider range of temperatures on the two monometallic samples. In every case the course of the reaction followed the usual reversible first-order plot expected for exchange,2 except that a slight amount of self-poisoning occurred leading to a decrease of rate whichA .C. Faro and C. Kemball 3127 0 50 100 Ir (%) Fig. 1. Rates of exchange of benzene at 263 K (0) and of 2,2DMP at 393 K (a). was normally some 10 to 20% over 100 min. A decrease of ca. 30% was observed with RC244 (Ir/A1203) at 452 K, the highest temperature used. The results at 354 and 393 K for all catalysts are given in table 1, and the corresponding data for the monometallic iridium catalyst at the higher temperatures in table 2. Turnover numbers at 354 K were not influenced by catalyst composition, but at 393 K platinum-rich catalysts were more active than the iridium catalyst by a factor of ca.3 (see fig. 1). In every case the main reaction was stepwise exchange (process I) but some initial products containing two or three deuterium atoms were also formed and attributed to methyl-group exchange (process 11). The contribution of process I1 increased with temperature on all catalysts, but at a fixed temperature was slightly more significant on platinum-rich than on iridium-rich catalysts. Products with four or more deuterium atoms, owing to multiple exchange involving two or more methyl groups (process 111), were only found on the monometallic iridium catalyst at temperatures of 393 K or above; the proportion of these rose sharply with increase of temperature. Good Arrhenius plots (based on the C,Xi ions formed in the mass spectrometer) were obtained from the results on the two monometallic catalysts, and the derived activation energies are given in table 3 as Ei, corresponding to products leading to the ion C,H,+D$ in the mass spectrometer.With both metals the lowest activation energy was found for E, (process I) and the overall activation energy, El-,, was not appreciably greater because process I made the greatest contribution to the exchange under all conditions studied. The activation energy which was greater than El by some 33 or 34 kJ mol-1 on both metals, was attributed to process 11. The highest values observed were for E4-, and E, on iridium, and these were associated with process 111; the agreement between the two values confirmed that all products giving rise to the D, to D, ions were formed by similar mechanisms.3128 Reactions of Hydrocarbons on Pt-Ir Catalysts Table 1.Turnover numbeqa product distributionsb and values of M for the exchange of 2,2-dimethylpropane on the Pt-Ir catalysts reaction at 354 K reaction at 393 K catalyst Ir(%) NE/lO-, D, D, D, M NE/lO-, D, D, D, A4 RC238 0 0.67 88.6 10.8 0.6 1.12 4.5 82.3 14.5 3.2 1.21 RC239 19 0.67 86.6 12.6 0.8 1.14 5.3 79.3 16.9 3.8 1.24 RC240 39 0.61 89.2 10.1 0.7 1.11 4.3 84.5 13.2 2.3 1.17 RC241 46 0.59 89.7 9.6 0.7 1.11 4.1 86.9 11.1 2.0 1.15 RC242 54 0.64 91.7 7.5 0.8 1.09 3.9 86.3 11.5 2.2 1.16 RC243 73 0.63 93.4 5.9 0.3 1.06 3.3 90.5 8.0 1.5 1.11 RC244 100 0.51 96.1 3.3 0.6 1.04 1.5 93.0 5.4 1.4' 1.09 a Turnover numbers, N,, are given in molecule s-l site-'. to C,X; ions. Also 0.2% D, and 0.1 % D,.Distributions and values of M refer Table 2. The exchange of 2,2-dimethylpropane on RC244 (0.39% Ir/A1203) at higher temperatures T/K NE/lO-, D, D, D, D, D, D, D, D, D, M ~- 420 2.9 84.7 10.2 3.0 0.8 0.4 0.3 0.1 0.1 0.4 1.26 452 7.3 73.6 14.6 5.9 1.6 0.8 0.8 0.4 0.8 1.4 1.57 Table 3. Activation energies for the exchange processes for 2,2-dimethylpropane on the monometallic catalysts temperature activation energiesa/kJ mol-1 range catalyst /K L J b 4 E2-3 E4-9 E, - - RC238 (0.5 1 % Pt/A1,0,) 323-393' 56 54 87 RC244 (0.39% Ir/A1203) 30 38 35 69 117 109 a Ei refers to the activation energy for the production of compounds leading to the formation of the ion C,H,-iDl in the mass spectrometer. Approximate values of El-9 for the bimetallic catalysts based merely on the results in table 1 were 61, 58, 58, 54 and 49 kJ mol-l for RC239 to RC243, respectively. ' Turnover number at 323 K was 0.001 1.Turnover numbers at 301 and 322 K were 0.00045 and 0.001 1, respectively. Hydrogenolysis Reactions Experiments on the hydrogenolysis of 2,2 DMP were carried out in the temperature range 461-544 K using pressures of 4.2 kPa of hydrogen and 0.56 kPa of alkane, the latter corresponding to ca. 1.6 x 1019 molecules in the reaction vessel. Selectivities for the reactions were expressed in terms of the parameters Si, the number of molecules of hydrocarbon with i carbon atoms produced initially from each molecule of reactant used up. Products involving isomerisation were expressed as S;. Thus, for 2,2 DMP as reactant S4 corresponds to 2-methylpropane, Si to n-butane and S; to 2-methylbutane (n-pentane was not observed).A .C. Faro and C. Kernball 3129 The definition used for M,, which indicated the depth of hydrogenolysis or the number of C-C bonds broken in each reacting molecule of 2,2DMP, was MB = (s;+csi)/(l-s;)- 1. i A further quantity, yR, was defined in order to estimate the overall extent of carbon skeletal rearrangement with 2,2 DMP. Eight reactions had to be considered to account for the range of products formed: 2,2 DMP + H, + (CH,),CH + CH, 2,2DMP + 2H, -+ C,Hs + 2CH, 2,2DMP + 3H2 -+ C2H, + 3CH, 2,2 DMP + 4H, + 5CH, (3) (4) 2,2 DMP + H, -+ n-C,H,, + CH, ( 5 ) 2,2DMP+ H, -+ C3H, + C2H, 2,2DMP+2H2 -+ 2C,H6+CH, 2,2DMP +(CH,),CHC,H5. (8) Reactions (1)-(4) can occur without a skeletal rearrangement of 2,2 DMP, but reactions (5)-(8) require such a process.If yi represents the fractional contribution of reaction i to the overall reaction we have the following relations between the products and the fractional contributions: s; = Y s (9) We define the total contribution of rearrangement processes as y,, where YR = Y.5 +Y6 +Y7 +Y8 and by use of eqn (9)-(14) this becomes YR = (2s; + S , + 2s; + S3 + s, - 1) +y1. It is not possible to obtain an exact solution for y , as'the number of products is less than the number of processes. However, if we assume that y4 = 0, i.e. that the complete degradation to methane is negligible, the known quantities in eqn (16) set a lower limit to the contribution of the skeletal rearrangement processes. Table 4 summarises the results for 2,2DMP over all the catalysts at 480 K and also at the highest temperature used with each catalyst. The results include turnover numbers, product distributions and the derived parameters MB and yR.Rates of reaction over a range of temperatures were obtained on all catalysts except RC240, and the data are shown as Arrhenius plots in fig. 2. Derived parameters are given in table 5, together with some pressure-dependence results for three catalysts from experiments using either one-third the normal pressure of hydrogen of three times the normal pressure of hydrocarbon.3130 Reactions of Hydrocarbons on Pt-Ir Catalysts Table 4. Results for the hydrogenolysis of 2,2-dimethylpropane products catalyst Ir(%) T/K NH/10-4 S, S, S, S, Si S; M , RC238 0 480 544 RC239 19 480 522 RC240 39 480 RC241 46 480 523 RC242 54 480 520 RC243 73 480 520 RC244 100 480 520 0.13 0.54 0.76 0.95 1.9 3.4 9.9 14.1 13.8 30 28 28 67 1.07 0.10 0.09 0.73 0.05 0.99 0.18 0.12 0.70 0.07 1.07 0.24 0.08 0.80 - 1.16 0.20 0.08 0.80 - 1.21 0.33 0.08 0.73 - 1.34 0.43 0.03 0.67 - 1.52 0.41 0.11 0.58 - 1.29 0.44 0.08 0.64 - 1.62 0.34 0.08 0.61 - 1.36 0.48 0.09 0.60 - 1.74 0.44 0.10 0.52 ~ 1.73 0.60 0.05 0.50 - 2.56 0.53 0.09 0.28 - 0.07 1.19 0.05 1.17 1.19 - 1.24 1.35 - 1.47 - 1.62 1.46 - 1.65 - 1.53 1.80 1.88 _.2.46 - - - _. - I 1 I 1.9 2.0 2.1 lo3 KIT 0.16 0.24 0.12 0.08 0.14 0.13 0.10 0.16 0.03 0.17 0.06 0.15 -0.10 Fig. 2. Arrhenius plots for the hydrogenolysis of 2,2DMP: 0, RC238; A, RC239; 0, RC241; ., RC242; A, RC243 and 0, RC244. Results for the reactions of 2-methylpropane (2 MP) and 2-methylbutane (2 MB) on the two monometallic catalysts are given in table 6.Discussion Exchange Reactions The small variation in the rate of exchange of benzene with metal composition (fig. I ) and the similar activities shown by all catalysts for the exchange of 2,2DMP at 354 KA . C. Faro and C. Kemball 3131 Table 5. Arrhenius parameters and kinetic data at 480 K for the reaction of 2,2-dimethylpropane 1% ( A kinetic dataa /molecule -_______ catalyst Ir(%) E/kJ mol-l s-l sitep1) n m RC238 0 155 12.0 -1.2 1.1 RC239 19 160 13.1 - - RC241 46 1 64 13.8 - - RC242 54 135 11.0 RC243 73 108 8.3 -0.7 0.3 RC244 100 102 8.1 -0.5 0.2 - - a Results fitted to a power rate law, N , = l ~ ( p , , ) ~ (pHC)m. Table 6. Results for the hydrogenolysis of 2-methylpropane and 2-methylbutane on the monometallic catalysts products reactant catalyst T/K NH/10-4 S, S , S, S,.Sia Ska MB 2MP Pt/A1203 480 0.44 0.93 0.06 0.93 - 0.05 - 1.02 2MP Ir/A1203 480 16 1.46 0.64 0.35 - 0.06 - 1.61 2MB Pt/A1,0, 473 0.48 0.96 0.04 0.04 0.64 0.32 trace 1.00 2MB Ir/A1203 472 16 0.97 0.47 0.21 0.51 0.10 0.04 1.35 a In this table, S, refers to 2-methylpropane, ,S; to n-butane and S; to n-pentane. support the conclusion in Part 1 that we were working with a consistent series of catalysts. Two features about the results for the exchange of 2,2DMP at 393 K suggest possible surface enrichment by platinum. First, the activity of the catalysts remained more or less constant as the content of iridium was increased to 54% and then decreased by a factor of three (fig.1). Secondly, as may be seen from data in table 3, the pattern of activation energies across the series showed a similar variation, with values of 55-60 kJ mol-1 for compositions up to 54% iridium and then a decrease for RC243 and RC244. Wong7 reported some evidence of platinum surface enrichment from chemi- sorption of hydrogen and of carbon monoxide for the alumina-supported catalysts he used. The results for the exchange of 2,2 DMP on RC244 were comparable to those reported for a highly dispersed silica-supported iridium catalyst.3 Our activation energy of 38 kJ mol-l compares with the previous value of 37 kJ mol-l, and the initial products from both catalysts were similar. The evidence that the activation energies for the three exchange processes with 2,2DMP are in the order E,,, > E,, > E, for the present iridium catalyst conforms to the patterns reported earlier3 for silica-supported rhodium, where the corresponding values were 76, 55 and 37 kJ mol-l, respectively.3132 Reactions of Hydrocarbons on Pt-Ir Catalysts Table 7.Ratios of rates of exchange: hydrogeno- lysis for the monometallic catalysts with 2,2DMP catalyst 480 K 520 K Pt /A120, s x 104 1 x 104 Ir/A12Q3 1 x 102 40 H y drogenol y sis Rate-determining Step for the Reaction of 2,2 DMP The main possible rate-determining steps for hydrogenolysis are ( a ) chemisorption of the alkane, (b) carbon-carbon bond cleavage or (c) desorption of the products. Comparison of the rate of exchange of 2,2DMP, which gives a measure of the rate of reversible chemisorption of the alkane, and the rate of hydrogenolysis shows that (a) is unlikely to be slow.The estimated ratios of these reactions for the two monometallic catalysts are given in table 7. However, it is interesting to focus attention not so much on the overall rate of exchange but more specifically on the rate of process 111, which involves the reversible formation of the a,y-diadsorbed species, also believed to be involved in the mechanism of carbon+arbon bond cleavage.6? 9-11 Our results permit us to compare the rates of process 111 and hydrogenolysis for the iridium catalyst, but extrapolation of the exchange data suggests that over the temperature range 480-520 K the rate of process I11 is only 6 7 times faster than the rate of hydrogenolysis.So the extent of reversibility associated with the formation of the a,?-species is relatively high for iridium and greater than for r h ~ d i u m , ~ where process 111 was less than three times the rate of hydrogenolysis at 427 K. Clearly, the degree of reversibility of the species varies with the metal, and for iron films12 at 486 K the formation was essentially irreversible, as hydrogenolysis occurred but no exchange by process I11 was observed. There is good evidence that (c), the desorption of products, is not rate-determining in the hydrogenolysis of 2,2DMP on any of the catalysts. Faster rates of production of methane, ethane and propane were found from the reactions of n-butane at 473 K (Part 1) than from the hydrogenolysis of 2,2DMP at 480 K; likewise 2-methylpropane was produced more rapidly from 2MB than from 2,2DMP on the two monometallic catalysts.Also, the exchange of methane at 473 K, which is likely to be the slowest of the hydrocarbons to desorb, was faster than its rate of production from the hydrogenolysis of 2,2DMP at 480 K on each of the catalysts; the ratio decreased from lo2 for RC238 (Pt/A1203) to 4 for RC244 (Ir/Al2O3). The elimination of (a) and ( c ) as possible rate-determining steps means that (b), the cleavage of carbon-carbon bonds, is the slowest step in the overall process of hydrogenolysis of 2,2DMP. The rise in the values of M,, the mean number of carbon-carbon bonds broken, with iridium con tent does suggest a greater tendency for hydrocarbon species to be retained on the surface as the percentage of iridium increased.Conversely, the predominance of the cleavage of only one carbon<arbon bond for the platinum-rich catalysts is an indication of relatively easy desorption of hydrocarbons from such catalysts.A . C. Faro and C. Kernball 3133 Isomerisation and Rearrangement Processes Isomerised products from the reaction of 2,2 DMP were observed only on Pt/A120, (table 4) and the amounts of 2-methylbutane and n-butane formed were smaller than commonly found with other platinum ~ata1ysts.l~- l5 This confirms suggestions that highly dispersed platinum catalysts, such as those in the present series, are selective for hydrogen01ysis.l~~ l6 Despite the absence of isomerisation products from 2,2DMP on all catalysts containing iridium, the values of yR suggest that considerable skeletal isomerisation accompanies hydrogenolysis.The results indicate that isomerised intermediates are degraded to yield small molecules instead of leaving the surface as isomerised products, and substantiate the proposal that desorption of hydrocarbon from iridium is more difficult than from platinum. The values of yR confirm recent findings7 that iridium has an isomerisation activity comparable to that of platinum, provided that the products are measured at very low conversions. The use of this parameter provides a means of estimating rearrangement processes at conversions of a few per cent as long as the complete degradation to methane is relatively unimportant. In the present case this is true for all catalysts at lower temperatures, and for the platinum-rich catalysts at the higher temperatures. With iridium-rich catalysts the depth of hydrogenolysis (MB) increases markedly with temperature.At the same time, the corresponding values of yR decrease, which may be attributed to the fact that the assumption of negligible complete degradation no longer holds. This is clearly the case for RC244 at 520 K, where the negative value of yR results from the neglect of y4. The combination of rearrangement processes on the iridium-rich catalysts is demon- strated mainly by the presence of large amounts of ethane in the initial products. If the ethane arose solely by successive demethylation, there should be three molecules of methane formed for each molecule of ethane; in general, the amounts of methane, after deducting the quantity associated with the formation of 2-methylpropane, were insufficient to confirm succesive demethylation.We do not accept the interpretation that Yao and SheleP6 proposed for their results for the reaction of 2,2DMP on a catalyst with a low loading of platinum at 525 K. They proposed that propane and ethane were formed by stepwise demethylation. However, the results in their fig. 1 show (a) insufficient production of methane and (b) the appearance of ethane and propane as initial products of the reaction. Pressure Dependence of the Reactiom of 2,2 DMP We consider a possible simplified reaction scheme in order to interpret the kinetic results, to determine the relationship between the activation energies observed in this work and others in the literature, and also to obtain some guidance about mechanisms and the number of surface atoms involved in the active sites.We assume the following scheme: and slow Nads + products where N represents the hydrocarbon, z* is the number of sites (*) required for formation of the appropriate intermediate and the other symbols have their usual meaning. We assume that hydrogen does not participate in the rate-determining step;17* l8 it is involved in the establishment of the dissociative equilibria (17) and (19), and also in the3134 Reactions of Hydrocarbons on Pt-Ir Catalysts hydrogenative desorption of products which may become rate-determining at very low hydrogen pressure. Let e,, ON and OF represent the fractional coverages of the surface by hydrogen atoms, the relevant hydrocarbon intermediate and free sites.The following expressions relate the coverages to the partial pressures of the gases and Kl and K3, the equilibrium constants for reactions (1 7) and (1 9), respectively: and The initial rate of reaction, R, is given by R=k26N where k, is the rate constant for reaction (1 8). Three limiting situations are considered. (1) Surface coverage is small, so that OF z 1 . The rate expression is (2) Surface coverage by hydrogen is large, so that eH z 1 . The corresponding rate expression is (3) Surface coverage by the hydrocarbon species is large, so that ON = 1. The rate is then R = k,. We note that the order of reaction with respect to alkane was close to 1 for Pt/Al,O,, so that either situation (1) or (2) applies. Furthermore, the t.p.d.profile reported for this catalyst, obtained under a hydrogen pressure comparable to that used for hydrogenolysis, indicated that appreciable surface coverage by hydrogen is to be expected under conditions used for the reaction of 2,2DMP. The apparent activation energy of 155 kJ mol-l for reaction of 2,2DMP on Pt/Al,O, is in good agreement with the values of 134-154 kJ mol-l reported by Foger and Anderson15 for highly dispersed Pt/Y-zeolite catalysts. Also, our order of - 1.2 with respect to hydrogen compares with their value of - 1.5 obtained at considerably higher pressures (27-100 kPa). The results of Foger and Anderson probably corresponded to conditions of high hydrogen coverage, while ours reflected a regime intermediate between situations (1) and (2).On this basis, the hydrogen order reported by Foger and Anderson is related to the total number of sites involved in the chemisorption of the alkane, including those necessary to accommodate the dissociated hydrogen atoms. An intermediate which fits the observed value, z = 3, is the metallocyclobutane species shown in scheme 1 . Scheme 1. Intermediate for hydrogenolysis of 2,2 DMP on platinum. With Ir/A1203 reaction orders were closer to zero than for Pt/Al,O,, indicating that the reaction on iridium was taking place under appreciable hydrocarbon coverage. So in this case our results differ from those reported by Foger and Anderson5 for supported iridium catalysts (first-order in alkane, minus third-order in hydrogen).Since their results were obtained with much higher hydrogen pressures than ours, they may have corresponded to a regime of high hydrogen coverage. Our activation energy of 102 kJ mol-l for Ir/Al,O, was much smaller than their values of 230-250 kJ mol-l. We believe that the explanation may again be a consequence of the differences in surfaceA . C. Faro and C. Kemball 3135 coverage for the two sets of experiments. The relationship between apparent activation energy, E,, and the true activation energy, E,, for reaction (1 8) will be E, = E,-mQ,-nQ, where m and n are the orders with respect to alkane and hydrogen, and the Qs represent the respective heats of adsorption. Since a large negative order in hydrogen was obtained by Foger and Anderson their reported apparent activation energy may have contained a much larger contribution from the heat of hydrogen chemisorption than would apply in our case.The fact that our experiments were carried out under appreciable hydrocarbon coverage does not allow us to draw any direct conclusions as to the nature of the intermediate involved in the hydrogenolysis of 2,2DMP on iridium. However, if we accept the -3 order in hydrogen reported by Foger and Anderson, the intermediate must be more dehydrogenated and/or bonded to a larger number of metal atoms than the one which operates on highly dispersed platinum. A possibility is the species shown in scheme 2, which is similar in nature to the a,a,y-adsorbed intermediate proposed previo~sly.~~ lo Scheme 2.Intermediate for hydrogenolysis of 2,2 DMP on iridium. Comparative Reactivity of Diflerent Hydrocarbons From the data presented in Part 1 and in this paper, it is possible to devise a rate of hydrogenolysis per bond in the various molecules on Pt/A1,0, from the overall rates and selectivities. The derived reactivities are given in table 8 and show that the order, using P, S, T and Q for primary to quaternary carbon atoms, is S-S > S-P > P-T > P-Q z S-T. The similar reactivities for the S-S bonds in n-butane, n-pentane and cyclopentane may imply that all involve similar intermediates which might be a,a,P,P-tetra-adsorbed species. Assumption of such a mechanism would explain the large difference in activity found between S-S and S-T bonds, a difference which could not be easily interpreted on the basis of a mechanism involving a,?-adsorbed intermediates.Table 8. Reactivities for hydrogenolysis on Pt/A1,0, at 473 or 480 K in terms of different types of C-C bonds n,/lO-* s-l (per bond)" reactant T/K P-S P-T P-Q S-S S-T 1.3 - n-butane 473 0.3 - 2-methylpropane 480 - 0.14 - - 2,2DMP 480 - n-pentane 473 0.65 - - - - 0.025 - _- 2-methylbutane 473 0.3 0.08 - - 0.02 cyclopentane 473 - 2.3 - 1.8 - - - - a P, S, T and Q represent primary, secondary, tertiary and quaternary carbon atoms, respectively; nB is a turnover number per bond.3136 Reactions of Hydrocarbons on Pt-Ir Catalysts The cleavage of P-T, P-Q and S-T bonds, which were all relatively slow and cannot involve a,a,b,B-adsorbed intermediates, may proceed through a, y-adsorbed species.Certainly the pressure dependence results with 2,2 DMP suggest that the metallocyclobutane species shown in scheme 1 is a possibility. The significant difference in reactivity of P-T and S-T bonds may result from a preference of the metallocyclo- butane species to decompose to give adsorbed methylene and olefin rather than adsorbed alkylidene and olefin as illustrated in scheme 3. The difference in reactivity between P-T and P-Q bonds may arise if the formation of the metallocyclobutane intermediate is easier when it takes place through an adsorbed alkene species. This suggestion was made by Clarke and Rooney19 to explain why 2-methylpropane but not 2,2 DMP was isomerised on palladium. t c c’ ‘c’ Scheme 3. Modes of reaction of a metallocyclobutane from 2 MB : (A) favoured, giving P-T bond cleavage; (B) not favoured, giving S-T bond cleavage. Primary-secondary bonds were more reactive than other bonds involving primary atoms, although the difference between P-S and P-T was not large enough to allow us to conclude with certainty what kind of intermediate is involved in P-S hydrogenolysis.It may be a,a,P,/?-adsorbed but reacting more slowly than for S-S cleavage. A less likely alternative is a metallocyclobutane species, but reacting more rapidly than for P-T or P-Q cleavage. Detailed comparison of bond reactivities on Ir/A1203 was not possible due to the sizeable contribution of multiple hydrogenolysis with this catalyst. The iridium catalyst was slightly more selective than platinum for S-S cleavage.The order of reactivity of the various hydrocarbons was similar on both catalysts except that cyclopentane was less reactive than n-butane or n-pentane on iridium. Most hydrocarbons, with the exception of 2,2 DMP, displayed a little more selectivity for isomerisation with iridium than with platinum. In general the behaviour of the two metals showed broadly similar selectivities for the cleavage of different bonds and for skeletal rearrangement processes. The main difference from the standpoint of selectivity was the greater tendency of iridium to promote multiple hydrogenolysis. The Injluence of Metal Composition on the Rate of Hydrogenolysis of 2,2 DMP The sharp fall in the rate of reaction from RC244 to RC243 (73% Ir) suggested that ensembles of possibly 3 or 4 metal atoms might be needed to form the active site for the hydrogenolysis of 2,2DMP.Conversely, the sharp rise in activity from RC238 to RC239 (19% Ir) by a factor of 4 at 480 K indicated that mixed-metal ensembles must be significantly more active than platinum ensembles. In order to calculate the variationA . C . Faro and C. Kernball 3137 5 4 I 0 50 100 Ir (%) Fig. 3. Experimental and calculated activities for the hydrogenolysis of 2,2DMP: results at 0, 480; A, 500 and 0, 520 K ; lines calculated from expression (20) for ensemble size of (--) s = 3, (-) s = 4 and (---) s = 5. lo3 KIT Fig. 4. Arrhenius plot for the derived activity for mixed ensembles consisting of four metal atoms. in activity with metal composition we adopted the following simplifying assumptions.( a ) The composition of the surface can be expressed in terms of the overall composition of the metal catalysts, and possible platinum-enrichment of the surface is ignored. ( b ) The same ensemble size of 3, 4 or 5 metal atoms is required on all catalysts. (c) All mixed-metal ensembles exhibit the same activity. These assumptions lead to an expression for the turnover number, NX, for a catalyst with a mole fraction, x, of iridium (20) Nz = (1 -xX)"pt+xSn1,+[1 -xS-(l -X)S]nptIr3138 Reactions of Hydrocarbons on Pt-Ir Catalysts in which s is the ensemble size, npt and nIr are the turnover numbers for the monometallic catalysts and nptIr the turnover number for a hypothetical surface containing only mixed ensembles. Fig. 3 shows the fit obtained with expression (20) for three ensemble sizes at three different temperatures using, in each case, a value of nPtIr selected to give the smallest average deviation from the experimental results.A satisfactory fit was obtained at all temperatures with an ensemble size of s = 4, the largest deviation being only 20% of the experimental rate. Further evidence for the validity of this approach was obtained by plotting an Arrhenius equation for the selected values of nptIr, as shown in fig. 4. The plot was linear and gave an activation energy of 162 kJ mol-l, which accords with the results in table 5. Although this simple model reproduces the experimental results, some objections can be raised to it. There is no theoretical justification for the assumption of equal activities for all types of mixed ensembles.Next, the ensembles responsible for the hydrogenolysis of 2,2DMP need not be the same size for platinum, bimetallics or iridium. In fact, the rather different reaction orders with respect to hydrogen for the hydrogenolysis of 2,2DMP in this work and reported elsewhere for platinum15 and iridium5 suggest a variation in ensemble size with composition. Finally, the statistical approach for the estimation of the proportions of the various ensembles may not be accurate. The concept of a mixed-metal ensemble with an activity different from those for pure-metal ensembles implies that electronic effects are important. Further evidence for bimetallic interaction in the present series of catalysts is provided by the gradual change in selectivity with metal composition (table 4).If the metals had been acting independently one would have expected to see iridium-like product distributions across much of the composition range because of the much greater activity associated with iridium compared with platinum. Conclusions A comparison of the results in this paper and in Part 1 indicates a number of general points about the behaviour of the highly dispersed platinum-iridium catalysts. For exchange reactions, i.e. the activation of C-H bonds, there was comparatively little variation in activity across the series. However, the amount and the character of the multiple exchange, particularly with cyclopentane and to a lesser extent with methane or 2,2 DMP, varied regularly with metal composition.The variation in the activity of the catalysts for C-C bond cleavage was much greater than for the activation of C-H bonds, and the shape of the activityxomposition curve was hydrocarbon-dependent. With the straight-chain hydrocarbons and cyclopentane, reported in Part 1, the activity increased more or less linearly with the iridium content of the catalysts. With 2,2DMP the activity was more strongly influenced by iridium content, and this behaviour was interpreted as evidence for active sites consisting of ensembles of some four metal atoms. Mixed-metal ensembles showed an activity intermediate between those composed of the pure metals. A probable explanation for the difference between the shapes of the activitysomposition curves is that n-butane, n-pentane and cyclopentane can react by the C,-unit mode of C-C bond cleavage, whereas the hydrogenolysis of 2,2 DMP necessarily requires an iso-unit mode mechanism. With all hydrocarbons, the depth of hydrogenolysis tended to increase with iridium content, rising sharply as the percentage of iridium approached 100%. This was attributed to an increasing strength of hydrocarbon adsorption with iridium content, a conclusion supported by kinetic data and by comparison of rates of hydrocarbon exchange and hydrogenolysis.Similar types of C-C bonds in different hydrocarbons underwent cleavage at similar rates on platinum. Secondary-secondary bonds were most easily broken, with primary-A . C. Faro and C. Kernball 3139 quaternary bonds being the least reactive. The former probably involved a,a,P,P- tetra-adsorbed intermediates, and the latter either a metallocyclobutane or some other form of a,y- or a,a,y-adsorbed species. A. C . F. acknowledges financial support from Petrobras, Brazil, and we are grateful to the Petrochemicals and Plastics Division of ICI plc for assistance with the preparation and characterisation of the catalysts. References 1 Part 1 : D. Garden, C . Kemball and D. A. Whan, J . Chem. SOC., Faraday Trans. I , 1986,82, 31 13. 2 C. Kemball, Adv. Catal., 1959, 11, 223. 3 I. H. B. Haining, C . Kemball and G. L. Haller, J. Chem. SOC., Faraday Trans. I , 1981, 77, 2519. 4 R. Brown, C . Kemball, J. A. Oliver and I. H. Sadler, J . Chem. Res., 1985, ( S ) 274, ( M ) 3201. 5 K. Foger and J. R. Anderson, J . Catal., 1979, 59, 325. 6 J. J. Rooney, J. Catal., 1979, 58, 334. 7 G. L. Haller, personal communication; T. C . Wong, Ph.D. Thesis (Yale University, 1982). 8 A. C. Far0 Jr, M. E. Cooper, D. Garden and C. Kemball, J . Chem. Res., 1983, ( S ) 110, ( M ) 11 14. 9 J. R. Anderson and N. R. Avery, J. Catal., 1966, 5, 446. 10 G. Leclercq, L. Leclercq and R. Maurel, J . Catal., 1977, 50, 87. 11 F. Gavin and F. G. Gault, in Chemistry and Chemical Engineering of Catalytic Processes, ed. R. Prins 12 R. S. Dowie, C . Kemball, J. C . Kempling and D. A. Whan, Proc. R. SOC. London, Ser. A , 1972, 327, 13 M. Boudart and L. D. Ptak, J . Catal., 1970, 16, 90. 14 M. Boudart, A. W. Aldag, L. D. Ptak and J. E. Benson, J . Catal., 1968, 11, 35. 15 K. Foger and J. R. Anderson, J . Catal., 1978, 54, 318. 16 H. C. Yao and M. Shelef, J . Catal., 1982, 73, 76. 17 B. S. Gudkov, L. Guczi and P. Tetenyi, J. Catal., 1982, 73, 76. 18 H. C. Yao and M. Shelef, J . Catal., 1979, 56, 12. 19 J. K. A. Clarke and J. J. Rooney, Adv. Catal., 1975, 25, 125. and G. C. A. Schuit (Sijthoff and Noordhoff, Alphen aan den Rijn, 1980), p. 351. 491. Paper 512121; Receitled 4th December, 1985

ISSN:0300-9599    

DOI:10.1039/F19868203125

出版商:RSC    

年代:1986

数据来源: RSC