|
1. |
Front cover |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 85,
Issue 11,
1989,
Page 041-042
Preview
|
PDF (539KB)
|
|
摘要:
Contents 4259 4269 4277 4287 4295 431 1 4321 4335 Protonation Constant of Caffeine in Aqueous Solution M. Spiro, D. M. Grandoso and W. E. Price Ionic Equilibria in Acetonitrile Solutions of 2-, 3- and 4-Picoline N-oxide Perchlorates, studied by Potentiometry and Conductometry L. Chmurzynski, A. Wawrzyn6w and Z. Pawlak Liquid-phase Adsorption of Binary Ethanol-Water Mixtures on NaZSM-5 Zeolites with Different Silicon/Aluminium Ratios W-D. Einicke, M. Heuchel, M. v.Szombathely, P. Brauer, R. Schollner and 0. Rademacher Influence of Oxidation/Reduction Pretreatment on Hydrogen Adsorption on Rh/TiO, Catalysts. An lH Nuclear Magnetic Resonance Study J. P. Belzunegui, J. M. Rojo and J. Sanz Volumetric Properties of Mixtures of Simple Molecular Fluids A. C. Colin, E. G. Lezcano, A.Compostizo, R. G. Rubio and M. D. Peiia Study of Ultramicroporous Carbons by High-pressure Sorption. Part 4.-Iso- thems and Kinetic Transport in Activated Carbons J. E. Koresh, T. H. Kim, D. R. B. Walker and W. J. Koros Kinetic and Equilibrium Studies associated with the Solubilisation of n- Pentanol in Micellar Surfactants G. Kelly, N. Takisawa, D. M. Bloor, D. G. Hall and E. Wyn-Jones The effect of Carboxylic Acids on the Dissolution of Calcite in Aqueous Solution. Part 1 .-Maleic and Fumaric Acids R. G. Compton, K. L. Pritchard, P. R. Unwin, G. Grigg, P. Silvester, M. Lees and W. A. House 130-2Contents 4259 4269 4277 4287 4295 431 1 4321 4335 Protonation Constant of Caffeine in Aqueous Solution M. Spiro, D. M. Grandoso and W. E. Price Ionic Equilibria in Acetonitrile Solutions of 2-, 3- and 4-Picoline N-oxide Perchlorates, studied by Potentiometry and Conductometry L.Chmurzynski, A. Wawrzyn6w and Z. Pawlak Liquid-phase Adsorption of Binary Ethanol-Water Mixtures on NaZSM-5 Zeolites with Different Silicon/Aluminium Ratios W-D. Einicke, M. Heuchel, M. v.Szombathely, P. Brauer, R. Schollner and 0. Rademacher Influence of Oxidation/Reduction Pretreatment on Hydrogen Adsorption on Rh/TiO, Catalysts. An lH Nuclear Magnetic Resonance Study J. P. Belzunegui, J. M. Rojo and J. Sanz Volumetric Properties of Mixtures of Simple Molecular Fluids A. C. Colin, E. G. Lezcano, A. Compostizo, R. G. Rubio and M. D. Peiia Study of Ultramicroporous Carbons by High-pressure Sorption. Part 4.-Iso- thems and Kinetic Transport in Activated Carbons J. E. Koresh, T. H. Kim, D. R. B. Walker and W. J. Koros Kinetic and Equilibrium Studies associated with the Solubilisation of n- Pentanol in Micellar Surfactants G. Kelly, N. Takisawa, D. M. Bloor, D. G. Hall and E. Wyn-Jones The effect of Carboxylic Acids on the Dissolution of Calcite in Aqueous Solution. Part 1 .-Maleic and Fumaric Acids R. G. Compton, K. L. Pritchard, P. R. Unwin, G. Grigg, P. Silvester, M. Lees and W. A. House 130-2
ISSN:0300-9599
DOI:10.1039/F198985FX041
出版商:RSC
年代:1989
数据来源: RSC
|
2. |
Back cover |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 85,
Issue 11,
1989,
Page 043-044
Preview
|
PDF (921KB)
|
|
摘要:
THE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY ASSOCIAZIONE ITALIANA DI CHIMICA FlSlCA DEUTSCHE BUNSEN-GESELLSCHAFT FUR PHYSIKALISCHE CHEMIE KONINKLIJKE NEDERLANDS CHEMISCHE VERElNlGlNG SOCIETE FRANGAISE DE CHIMIE, DIVISION DE CHlMlE PHYSIQUE FARADAY DIVISION GENERAL DISCUSSION No. 90 Colloidal Dispersions University of Bristol, 10-12 September 1990 Orga nising Com mitte e Professor R. H. Ottewill (Chairman) Professor P. Botherol Professor E. Ferroni Or J. W. Goodwin Professor H. Hoff mann Professor A.L. Smith Professor P. Stenius Dr Th. F. Tadros Professor A. Vrij Dr D. A. Young The joint meeting of the Societies will be directed towards examining current understanding of the behaviour of colloidal dispersions. In particular, stability and instability, short range interactions, dynamic effects, non-equilibrium interaction, concentrated dispersions and order-disorder phenomena will form topics for discussion.The preliminary programme is now availablemay 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 SYMPOSIUM No. 26 Molecular Transport in Confined Regions and Membranes Oxford, 17-18 December 1990 Experimental, theoretical and simulation studies which address fundamental aspects of molecular transport will be discussed in the following main areas: a) Transport of atoms and molecules in pores, zeolite networks and other cavities; time-dependent statistical mechanics of small systems in confined geometries b) Molecular transport through synthetic membranes, biological membranes, smectic liquid crystalline phases and Langmuir Blodgett films; the dynamics of the molecules forming the membrane c) Diffusion, reorientation, conformational dynamics, viscosity and conductivity of polymer melts, to include papers dealing with bulk systems since the segments of the polymer will move in the anisotropic environment of the complete chain d) Applications of Brownian dynamics to the study of diffusion in porous media and across membranes including the transport of flexible aggregates such as microemulsions e ) The growth of crystals, colloidal aggregates and droplets on irregular surfaces and in pores Contributions for consideration by the Organising Committee are invited and abstracts of about 300 words should be sent by 31 December 1989 to: Dr D.J. Tildesley, Department of Chemistry, The University, Southampton SO9 SNH. Full papers for publication in the Symposium Volume will be required by August 1990.THE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY ASSOCIAZIONE ITALIANA DI CHIMICA FlSlCA DEUTSCHE BUNSEN-GESELLSCHAFT FUR PHYSIKALISCHE CHEMIE KONINKLIJKE NEDERLANDS CHEMISCHE VERElNlGlNG SOCIETE FRANGAISE DE CHIMIE, DIVISION DE CHlMlE PHYSIQUE FARADAY DIVISION GENERAL DISCUSSION No. 90 Colloidal Dispersions University of Bristol, 10-12 September 1990 Orga nising Com mitte e Professor R. H. Ottewill (Chairman) Professor P. Botherol Professor E. Ferroni Or J. W. Goodwin Professor H. Hoff mann Professor A.L. Smith Professor P. Stenius Dr Th.F. Tadros Professor A. Vrij Dr D. A. Young The joint meeting of the Societies will be directed towards examining current understanding of the behaviour of colloidal dispersions. In particular, stability and instability, short range interactions, dynamic effects, non-equilibrium interaction, concentrated dispersions and order-disorder phenomena will form topics for discussion. The preliminary programme is now availablemay 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 SYMPOSIUM No. 26 Molecular Transport in Confined Regions and Membranes Oxford, 17-18 December 1990 Experimental, theoretical and simulation studies which address fundamental aspects of molecular transport will be discussed in the following main areas: a) Transport of atoms and molecules in pores, zeolite networks and other cavities; time-dependent statistical mechanics of small systems in confined geometries b) Molecular transport through synthetic membranes, biological membranes, smectic liquid crystalline phases and Langmuir Blodgett films; the dynamics of the molecules forming the membrane c) Diffusion, reorientation, conformational dynamics, viscosity and conductivity of polymer melts, to include papers dealing with bulk systems since the segments of the polymer will move in the anisotropic environment of the complete chain d) Applications of Brownian dynamics to the study of diffusion in porous media and across membranes including the transport of flexible aggregates such as microemulsions e ) The growth of crystals, colloidal aggregates and droplets on irregular surfaces and in pores Contributions for consideration by the Organising Committee are invited and abstracts of about 300 words should be sent by 31 December 1989 to: Dr D.J. Tildesley, Department of Chemistry, The University, Southampton SO9 SNH. Full papers for publication in the Symposium Volume will be required by August 1990.
ISSN:0300-9599
DOI:10.1039/F198985BX043
出版商:RSC
年代:1989
数据来源: RSC
|
3. |
Contents pages |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 85,
Issue 11,
1989,
Page 143-144
Preview
|
PDF (158KB)
|
|
摘要:
ISSN 0300-9599 JCFTAR 85(11) 3609-3900 (1 989) JOURNAL OF THE CHEMICAL SOCIETY Faraday Transactions I Physical Chemistry in Condensed Phases 3609 3623 3629 3645 3663 3675 3687 3695 3709 3717 3725 3733 3747 3757 3767 3785 3797 CONTENTS Membrane Potential and Ion Transport in Inhomogeneous Ion-exchange Membranes A. Higuchi and T. Nakagawa Zeolite Synthesis in the SO,-Al,O,-Na,O-Pyridine-H,O System W. J. Smith, J. Dewing and J. Dwyer Influence of Potassium on the Catalytic Properties of V,O,/TiO, Catalysts for Toluene Oxidation J. Zhu and S. L. T. Andersson Influence of Phosphorus on the Catalytic Properties of V,O,/TiO, Catalysts for Toluene Oxidation Structure of Vanadium Oxides on ZrO, and the Oxidation of Butene H. Miyata, M. Kohno, T. Ono, T. Ohno and F. Hatayama Initial Cracking Properties and Physicochemical Characterization of Acid- leached Small-port (SP) and Large-port (LP) Mordenites by Pulse n-Hexane Cracking, Infrared and 27Al Magic Angle Spinning Nuclear Magnetic Resonance Spectroscopy F. Goovaerts, E.F. Vansant, J. Philippaerts, P. De Hulsters and J. Gelan Structural Vibrations of Acid-leached Mordenites. Determination of Structural Aluminium by Wavenumber and Intensity Analysis F. Goovaerts, E. F. Vansant, P. De Hulsters and J. Gelan Particle-Metal Interactions. A Raman and Electrochemical Study of a Compacted Electrode of Copper Phthalocyanine and Silver Metal A. J. Bovill, A. A. McConnell and W. E. Smith Conductivity Study of NaI Solutions in n-Propanol-n-Butanol Mixtures at 298.15 K. The Effect of Ion Pairing on the Standard Dissolution Enthalpies of NaI Ring-Disc Electrodes.Part 24.-Studies of Counterion Fluxes at a Thionine- coated Electrode W. J. Albery and A. R. Mount Effect of Phosphate and Polyphosphonates on the Dissolution of Barium Fluoride Kinetics of Aquation of [Fe(S-Br-phen),l2+ Ions in Aqueous Solutions as a Function of Temperature and Pressure. The Isochoric Controversy and Analysis of Equilibrium and Kinetic Data M. J. Blandamer, J. Burgess, H. J. Cowles, I. M. Horn, J. B. F. N. Engberts, S. A. Galema and C. D. Hubbard Unusual Thermodynamic Behaviour on Complexation of Cobalt(1r) with Chloride, Bromide and Iodide Ions in Hexamethylphosphoric Triamide Y, Abe, K. Ozutsumi and S. Ishiguro Pore Structure of Electrochemically Pretreated Glassy Carbon and Uptake of Lithium Ions into Micropores Hydrogenolysis of A1 kanes.Part 4.-Hydrogenolysis of Propane, n-Butane and Isobutane over Pt/A1,0, and Pt-Re/Al,O, Catalysts G. C. Bond and M. R. Gelsthorpe Influence of Support on the Availability of Nickel in Supported Catalysts for Hydrogen Chemisorption and Hydrogenation of Benzene S. Narayanan and G. Sreekanth Topological Investigations of the State of a Salt in some Binary Mixtures of Non-electrolytes J. Zhu, B. Rebenstorf and S. L. T. Andersson S. Taniewska-Osinska, A. Piekarska, A. Bald and A. Szejgis S. M. Hamza and S. H. El-Hamouly T. Nagaoka, Y. Uchida and K. Ogura P. P. Singh and M. BhatiaContents 3807 3813 3819 3825 3833 3845 3853 386 1 3871 3879 389 1 3899 Energetics of Molecular Interactions in Binary Mixtures of Non-electrolytes containing a Salt P.P. Singh and M. Bhatia General Isotherm Equation for Adsorption of Surfactants at SolidlLiquid Interfaces. Part 1. Theoretical B-Y. Zhu and T. Gu General Isotherm Equation for Adsorption of Surfactants at Solid/Liquid Interfaces. Part 2. Applications B-Y. Zhu, T. Gu and X. Zhao Interlayer Water Molecules of Vanadium Pentaoxide Hydrate. Part 1 .-Phase Equilibrium with Water Vapour at a Relative Pressure higher than 0.05 S. Kittaka, Y. Ayatsuka, K. Ohtani and N. Uchida Stepwise NO Chemisorption Processes on Synthetic Chrysotile Asbestos : Tubular Crystals with Acidic and Basic Surfaces H. Uchiyama, K. Kaneko and S. Ozeki Haem Peptide-Protein Interactions. Part 2.--Kinetics and Mechanism of the Interaction of Microperoxidase-8 with Apomyoglobin P. A. Adams, R. D. Goold and A. E. Thumser Droplet Formation and Contact Angles of Liquids on Mammalian Hair Fibres B. J. Carroll Surface Characterization of the Active RuO, xH,O Catalyst Supported on Teflon G. Morea, L. Sabbatini, P. G. Zambonin, N. Tangari and V. Tortorella Travelling Waves in the Acidic Nitrate-Ferroin Reaction G. Pota, I. Lengyel and G. Bazsa Catalytic Activity of SAP05 for Cracking of Butane and Hexane C. Halik, S. N. Chaudhuri and J. A. Lercher EXAFS Investigation of Structural Changes induced during the Pretreatment of a Titania-supported Iron-Ruthenium Catalyst F. J. Berry, X. Changhai, S. Jobson and R. Strange Corrigendum to Solid-state Nuclear Magnetic Resonance Study of a Series of Phosphonic and Phosphinic Acids R. K. Harris, L. H. Merwin and G. Hagele
ISSN:0300-9599
DOI:10.1039/F198985FP143
出版商:RSC
年代:1989
数据来源: RSC
|
4. |
Back matter |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 85,
Issue 11,
1989,
Page 145-154
Preview
|
PDF (747KB)
|
|
摘要:
JOURNAL OF THE CHEMICAL SOCIETY Faraday Transactions II, Issue 11 Molecular and Chemical Physics For the benefit of readers of Faraday Transactions I , the contents list of Faraday Transactions II, issue 11, is reproduced below 1713 Spin-coupled Valence Bond Study of the Lithium Hydride Anion M. J. Ford, D. L. Cooper, J. Gerratt and M. Raimondi 1721 Coordination Structures of Cu Ions in the Mixed-solvent System of Formamide and Ammonium Formate M. Miyake, S. Nagahara, Y. Yoshikawa and T. Suzuki 1729 Dark and Photovoltaic Properties of Doped Tetraphenylporphyrin Sandwich Cells. Part 1.-Doping Effects and Dark Electrical Properties W. A. Nevin and G. A. Chamberlain 1765 Evidence for the Marcus Inverted Region of Back-electron-transfer in Solution by Means of Chemically Induced Dynamic Nuclear Polarization H.G. 0. Becker, D. Pfeifer and K. Urban 1771 AM1 Study of Intramolecular Hydrogen Bonding in the Dithio Analogues of Malondialdehyde and Acetylacetone G. Buemi 1779 Photophysics of Molecular Rotors of Functionalised Surface-active Styrylcyanine Dyes M. S. A.Abde1-Mottaleb, A. M. K. Sherief, L.F.M.Ismae1, F. C.de Schryver and M. A. Vanderauweraer 1789 CNDO/VS-SOS Calculations of Second-order Hyperpolarkabilities J. 0. Morley, P. Pavlides and D. Pugh 1799 Addition of Cyclohexane to Slowly Reacting H2-02 Mixtures at 480 "C S. K. Gulati and R. W. Walker 181 3 Gas-phase Proton-transfer Reactions in Xylene-Dimethyl Ether Mixtures. Further Evidence for Mobile Protons M. T. Fernandez, K. R. Jennings and R. S. Mason 1827 Structure and Vibrational Frequencies of [HCNKrFI'.An A6 Initio Study Including Electron Corelation M. A. Vincent and I. H. Hillier 1831 A Two-step Model for Cubic Autocatalysis in an Open System C. Kaas-Petersen, J. K. McGarry and S. K. Scott 1837 Kinetics of the Gas-phase Pyrolysis of Pentamethyldisilane. A Re-investigation R. Becerra, J. S. Bertram, R. T. Walsh and I. M. WattsThe following papers were accepted for publication in Faraday Transactions during August, 1989. 8103579B 8/04885A 9/00134D 9/00374F 9/00727J 910 1 505A 9101568.J 910 17 18F 9/01778J 9/01870K 91019181 9/02030F 9m57F 9/02071C 9102113B 9n2221J 9/00730J 9/01713E 9/02M6H Synthesis of ZSM-20: Comparison of Properties with m l i t e Y Dwyer, J., Millward, D., O’Malley, P. J. and Araya, A. Hydrogen Isotope Effects in the Absorption of Water by Perfluorosulphonate (Nafion- 11 7) and Polystyrene-Divinylbenzene Sulphonate (Dowex 50W) Ion-exchange Resins Iyer, R.M., Pushpa, K. K. and Nandan, D. An Inverse Kirkwood-Buff treatment of the Thermodynamic Properties of DMSO-Water and Cyanomethane-Water Binary Liquid Mixtures at 298.2K Blandamer, M. J., Blundell, N. J., Burgess, J., Cowles, H. J. and Horn, I. M. Ultrasonic Absorption Propeties and Kirkwood-Buff Integral Functions for Propanone-Water Mixtures at 298.2K Blandamer, M. J., Blundell, N. J., Burgess, J., Cowles, H. J. and Horn, I. M. Adsorption of Isobutene on Partially Hydrophobized Aerosil Valencia, E. and Maldonado, A. Kinetics of N20 Decomposition on Fe3+ Supported on Pure and Li-modified A1203 Lycourghiotis, A., Kordulis, Ch., Latsim, H.and Pornonis, P. Hydrophobic Interactions of Alkanols. A Calorimetric Study in Water at 298.15K Elia, V., Cascella, C., Castronuovo, G., Sartorio, R. and Wurzburger, S. Infrared Study of Hydrogen Adsorbed on 23-02 Onishi, T., Maruya, K.-I., Kondo, J., Sakata, Y. and Domen, K. A SIMS Study of the Oxidation of Zirconium at High Temperatures and Low Oxygen Pressures Yamamoto, M., Naito, S., Mabuchi, M. and Hashino, T. Smectite Molecular Sieves. Part 3.-Theoretical Considerations of Sorption Kinetics Barrer, R. M. and Craven, R. J. B. Structural Aspects of the Chlorocyclohexane/Thiourea Inclusion System Harris, K. D. M. and Thomas, J. M. Study of Peroxide-modified Titanium Dioxide (Anatase) Hadjiivanov, K., Klissurski, D., Kantcheva, M.and Gyurova, L. Solvent Polarity of Aqueous Polymer Solutions as measured by the Solvatochromic Technique Zaslavsky, B. Yu., Miheeva, L. M., Masimov, E. A., Djafarov, S. F. and Reichardt, C. Forces between Mica Surfaces Bearing Adsorbed Homopolymers in Good Solvents: A Reappraisal in the Light of Recent Theoretical Advances Luckham, P. F. and Klein, J. NMR Studies of Complexes Formed by Adsorption of Ammonia on Nickel-Y Type Zeolite Fraissard, J., Bonardet, J. L. and Gedeon, A. Investigation of the Structure of Adsorbed Polymer Layers in the Good Solvent Case by Optical Evanescent Waves Caucheteux, I., Jerome, R., Rondelez, E and Hervet, H. A Theoretical and Topological Study on the Electroreduction of Chlorobenzene Derivatives Fontanesi, C., Benedetti, G. and Battistuzzi, G.On the Molecular Dynamics of Nucleosides and Nucleotides Sterk, H., Konrat, R. and Kalcher, J. A Molecular Mechanics Study of Hindered Phenols used as Antioxidants Mitchell, P. C. H., Drew, M. G. B. and Hopkins, W. A. (ii)9/02285F 9/02383F 9/02453K 9/02505G 9/02540E 9/02541C 9/02571E 9/02707F 9/027931 9/028 10B 9/02822F 9/02835H 9/028827 9/03199E 9/03500A 9/03 5 25G 9/03589C 9m3633D 9/032501 9/03630J 9/0363 1H NH4 Exchange: A Sensitive Tool for the Characterization of Structurally Modified Zeolites De Hulsters, P. and Vansant, E. F. Absorption of Hydrogen by Pd-Nb(Ta) Solid Solution Alloys Sakamoto, Y., Kajihara, K., Kikumura, T. and Flanagan, T. B. CIDEP and CIDNP Studies on Hydrogen Abstraction of 9,lO-Anthraquinone from Xanthene Terazima, M., Maeda, K., Sugawara, M., Takahashi, S.and Azumi, T. Dielectric Properties of Urea and Acetamide in Aqueous Solution Gabriel, C., Bateman, J. B., Evans, G. F. and Grant, E. H. Conformation and Ring Inversion in y-Butyrolactone. Part 1 .-Microwave Spectrum Lister, D. G., Lopez, J. C., Alonso, J. L., Cervellati, R., Degli Esposti, A. and Palmieri, P. Conformation and Ring Inversion in y-Butyrolactone. Part 2.-Ab Initio and Flexible Model Computations Lister, D. G., Degli Esposti, A., Alonso, J. L., Cervellati, R., Lopez, J. C. and Palmieri, P. A Mean Field Analysis of an Ion-Dipole Mixture Against a Charged Hard Wall with Specific Adsorption. Part 2.ANon-linear Results Outhwaite, C. W. and Molero, M. E.m.f. Studies of Cellulose Acetate Membranes: Binding of Divalent Cations and Membrane Potentials of KCI/KF Mixtures Sorensen, T.S., Skacel, F., Malmgren-Hansen, B. and Jensen, J. B. Tris(bipyridine)ruthenium(n) Photosensitized Reduction of a Com-Schiff Base Complex in a Network of Gelatin Hydrogel and in an Aqueous Gelatin Solution Kurimura, Y., Hiraizumi, K.-I., Harakawa, T., Yamashita, M., Osada, Y., Shigehara, K. and Yamada, A. Influence of Hydrogen Chloride Addition on the Catalytic Isomerization Activity of Chlorinated Alumina and Chlorinated Platinum/Alumina Solids. Superacid Behaviour Bernard, P.-M. and Primet, M. Ground-state Energies of Three-body Coulomb Systems Calculated by the Short-time Green Function Monte Car10 Method Ball, R. G. J. and Wells, B. H. Ring Inversion in Tetrahydrothiophen-3-one: A Microwave Study Lopez, J.C., Villamanan, R. M., Munoz, J. M. and Alonso, J. L. Use of Convolution Voltammetry for the Determination of Single Ionic Gibbs Energies of Transfer Kontturi, K., Kontturi, K., Murtomaki, L. and Schiffrin, D. J. Far-ultraviolet Solution Spectroscopy of Cyanide, Cyanate, Selenocyanate and Tellurocyanate. The Solution Spectroscopy of the NCX Series Fox,M. F. and Hayon, E. Photodissociation of Metal Cluster Ions: Dissociation Energies and Optical Spectroscopy Jarrold, M. F., Ray, U., Bower, J. E. and Creegan, K. M. Structure of Carbon Clusters as Studied by the Coulomb Explosion Method Feldman, H., Kella, D., Malkin, E., Miklazky, E., Naaman, R., Vager, Z. and Zajfman, J. Infrared Spectroscopy and Dimer Formation at the Surface of Medium-Large Argon Clusters Scoles, G., Goyal, S., McCombie, J., Pate, B.and Levandier, D. J. Correlated Walk Model of the Melting Transition in Small Clusters Schlag, E. W. and Selzle, H. L. Photoionization of Na-, Cs-, Ca- and Ba-Oxide Clusters Martin, T. P., Bergmann, T. and Malinowski, N. Non-resonant and Resonant Multiphoton Ionization of (NO)n and Ar&O Clusters with Picosecond Laser Pulses Barton Smith,D. and Miller, J. C. Infrared Spectroscopy of Large C02 Clusters Ewing, G. E. and Disselkamp, R. (iii)Cumulative Author Index 1989 Abe, M., 1493 Abe, Y., 3747 Adachi, K., 1065, 1075, 1083 Adams, P. A., 3845 Agathonos, P., 1357 Aguilella, V. M., 223 Aiello, R., 2749 Akitt, J. W., 121, 2035 Alaiion, M. R. L., 3425 Albery, W. J., 1181, 1189, Al-Bizreh, N., 1303 Albuquerque, L.M. P. C., 207 Allen, G. C., 55 Almeida, B. S., 1217 Amodeo, P., 621 Anderson, J. A., 11 17, 1129, Anderson, M. W., 1945 Anderson, S. L. T., 3629, 3645 Anpo, M., 609 Anzai, S., 2499 Aouali, L., 2771 Aoyama, T., 3353 Apelblat, A., 373 Arai, T., 929, 1451 Arai, Y., 2369, 2809 Archer, M. D., 1027 Aruga, T., 2597 Asakura, K., 441, 2021 Attwood, D., 3011 Austin, J. C., 1159 Ayatsuka, Y., 3825 Azenha, M. E. D. G., 2625 Bahra, G. S., 1979 Baiker, A., 999 Bakshi, M. S., 2285, 2297 Bald, A., 479, 3335, 3709 Balej, J., 3327 Barlow, M. T., 1945 Barone, G., 621, 2087 Barone, V., 621 Bartle, K. D., 2347 Bazsa, G., 3273, 3871 Beckett, M. A., 727 S e , M., 2525 Bellotto, M., 895 Bengtsson, L., 305, 317, 2917 Berleur, F., 3587 Berry, F. J., 467, 3891 Bertoldi, M., 237 Bertran, J., 1207 Beyer, G.K., 2737 Beyer, H. K., 2127 Bhatia, M., 3797, 3807 Bicelli, L. P., 1685 3717 2983, 2991, 3505 Bielanski, A., 2847 Bird, R., 2173 Black, S. N., 1795 Blandamer, M. J., 735, 1809, Bloor, D., 2099 Bodart, P., 2749 Bolis, V., 855, 1383 Bolshakov, G. F., 3119 Bolton, J. R., 1027 Bond, G. C., 168, 3767 Bonnet, P-A., 3587 BorEly, G., 2127 Borbely, G., 2737 Borowko, M., 343 Bosch, H., 1425 Boss, R. D., 11 Boucher, E. A., 2963 Bovill, A. J., 3695 Bowker, M., 165, 2635 Boyd, S. A., 2953 Brandreth, B. J., 3579 Bremer, L. G. B., 3359 Brimblecome, P., 157 Brookes, B. I., 2173 Brown, P., 2099 Brown, R., 2159 Bruce, J. M., 2647 Bulow, M., 1501 Burch, R., 3561, 3569 Burgess, J., 735, 1809, 3733 Burkhardt, I., 21 13 Burrows, H. D., 2625 Busca, G., 137, 237 Buxton, G.V., 3513 Byfield, M. P., 2713 Caceres, M., 3425 Cai, F. X., 1991 Calado, J. C. G., 1217 Camillen, P., 3385 Campbell, J. A., 843 Campelo, J. M., 2535 Carbonara, M., 1257 Cargill, R. W., 2665 Carlsen, L., 3403 Carlstrom, G., 1049 Caro, J., 1501 Carroll, B. J., 3853 Cartmell, D. W., 3513 Cascella, C., 3289 Cassol, A., 2445 Castronuovo, G., 2087, 3289 Cattania, M. G., 801 Chadwick, A. V., 166, 1979 Chandra, H., 1801 Changhai, X., 3891 3733 Chappell, R. J., 3569 Chaudhuri, S. N., 3879 Che, M., 609 Chen, J., 829 Chen, L-f., 33 Chien, S-H., 2199 Chiou, C. T., 2953 Chuvylkin, N. D., 3233 Claesson, P. M., 1933 Cleaver, B., 2453 Clegg, S. L., 157 Clifford, A. A., 2347 Cohen, H., 1169 Cole-Hamilton, D. J., 3385 Collette, H., 2749 Colling, C. N., 1303 Coluccia, S., 609, 1655 Comninos, H., 633 Compton, R.G., 761, 773, 977, CondlyfTe, D. H., 2453 Conway, B. E., 2355 Conway, S. J., 71, 79, 1841 Cooper, J., 1365 Copperthwaite, R. G., 633 Costas, M., 2211 Cottrell, M. R., 1809 Coudurier, G., 1607, 2615 Coding, S. B., 3033 Couves, J. W., 1979 Covington, A. K., 2827, 2835 Cowles, H. J., 3733 Cox, B. G., 187 Cristiani, C., 895 Cristinziano, P., 621 Cruz, M. J., 2071, 2079 Cucinotta, V., 2445 Czapkiewicz, J., 2669 da Silva Pereira, M. I., 2473 da Costa, F. M. A., 2473 da Costa, M. A., 907 Dainty, C., 3385 Dalas, E., 2465, 3159 Danil de Namor, A. F., 2705 Das, I., 201 1 Das, P. K., 2405 Das, S., 1531 Dash, A. C., 2405, 2797 Datka, J., 47, 837, 3079 Davey, R. J., 1795 Davis, K. G., 2901 Davis, M. I., 2723 Dawber, J. G., 727 de Vizcardo, Y.F., 2705 de Acosta, V. D., 2705 De Giglio, A., 23 1821, 2255, 2273, 3451AUTHOR INDEX De Hulsters, P., 3087, 3095 3675, 3687 DCkany, I., 3373 Delafosse, D., 2771 Dell’Atti, A., 23 Del Vecchio, P., 2087 Dennison, P. R., 3537 Dereigne, A., 2771 Derouiche, A., 3301 Deschaux, M., 2605 Deshmukh, R. D., 2675 Dewing, J., 3623 Di Bernardo, P., 2445 Dianoux, A. J., 2525 Dickel, G., 1463, 1671 Ding, J., 1599 Di Quarto, F., 3309 Dmitrieva, Z. T., 31 19 Domen, K., 929, 1451 Dong, S., 1575, 1585, 1599 Donini, J. C., 91 Douheret, G., 2723 Downs, G. W., 1841 Drummond, C. J., 521, 537, 551, Duatti, A., 3107 Dunn, M., 2827, 2835 Dwyer, J., 3623 Easteal, A. J., 1091 Eaton, G., 3257 Eden, J., 991 Egawa, C., 2597 Egsgaard, H., 3403 Elders, J. M., 2035 El-Hamouly, S.H., 3725 Elia, V., 3289 Elisei, F., 1469 El Jamal, M., 2615 el Torki, F. M., 349 Endoh, A., 1327 Engberts, J. B. F. N., 3733 Ernst, S., 2127 Espinos, J. P., 1279 Esteso, M. A., 2575 Fahim, R. B., 1723 Falconer, J. W., 71, 79, 1841 Fatome, M., 3587 Fernandez, A., 1279 Fernandez, C., 2749 Fernandez-Merida, L., 2575 Fernandez-Rneda, C., 1019 Ferranti, F., 2241 Finch, J. A., 91 Finegold, L., 2945 Fink, P., 3079 Flanagan, T. B., 1787 Fletcher, P. D. I., 147, 3075 Foerch, R., 1139 Foo, C. H., 65 Forissier, M., 1607, 2615 Formosinho, S. J., 2625 Forster, H., 1149 Forzatti, P., 895 Franchini, G., 1697 56 1 Frankel, R. B., 3033 Franks, F., 2417, 2945 Frey, H. M., 167 Frood, D. G., 3045 Frost, V. L., 2713 Fubini, B., 237, 855, 1383 Fujiwara, T., 2931 Fulop, V., 2127 Gabelica, Z., 2749 Gabriel, C.J., 11 Gabrys, B., 168 Gadzekpo, V. P. Y., 1027 Galema, S. A., 3733 Gallardo-Jiminez, M. A., 2901, Gans, P., 1835 Garcia, A., 2535 Garrone, E., 585, 1373, 1383 Garst, J. F., 1245 Gasser, D., 999 Gavish, B., 1199 Gelan, J., 3675, 3687 Gelsthorpe, M. R., 2641, 3767 Gervasini, A., 801 Geus, J. W., 269, 279, 293, 1267 Giamello, E., 237, 855, 1373 Giancola, C., 2087 Gilbert, P. J., 147 Gill, D. S., 2285, 2297 Gill, J. B., 1835 Gillette, G., 2369 Girault, H. H., 843 Goatly, M. B., 3074 Golding, P. D., 2229 Golunski, S. E., 3569 Gonzalez-Diaz, 0. M., 2575 Gonzilez-Elipe, A. R., 1279 Gonzalez-Lafont, A., 1207 Goodwin, J. W., 2785 Goold, R. D., 3845 Goovaerts, F., 3675, 3687 Gomer, H., 1469 Gottschalk, F., 363 Grech, E., 3187 Grieser, F., 521, 537, 551, 561 Griffiths, J.F., 3059, 3471 Grzybkowski, W., 3395 Gu, T., 3813, 3819 Guardado, P., 735 Guest, A., 1897 Guil, J. M., 1775 Gutierrez, C., 907 Gutschick, D., 21 13 Guy, P. D., 1795 Guyan, P. M., 2647 Habeeb, M. M. M., 3187 Habibullah, M., 3045, 3145 Hagele, G., 1409 Hakin, A. W., 1809 Halawani, K. H., 2185 Halawani, K. H. M., 2999 Halik, C., 3879 Hall, D. G., 188 1, 2099 Halle, B., 1049 2909 Hamnett, A., 3071 Hampton, S., 773 Hamza, S. M., 3725 Han, S., 829 Handreck, G. P., 645, 3195, Harkin, V. S., 2857 Harland, R. G., 761, 2273 Harris, K. R., 3281 Harris, R. K., 1409, 1853 Harrison, P. G., 1897, 1907, Hasselaar, M., 1267 Hasted, J. B., 99 Hatano, M., 199 Hatayama, F., 3663 Hatley, R. H. M., 2945 Hattori, T., 3135 Hayamizu, K., 2973 Hayashi, A., 2931 Hayashi, K., 3353 Hayashi, S., 2973 Hazra, D. K., 1531 Healy, T.W., 521, 537, 551, 561 Heatley, F., 917 Hegarty, B. F., 1861 Herder, C. E., 1933 Herder, P. C., 1933 Hernandez-Luis, F. F., 2575 Herrington, T. M., 3529 Hesselink, W. H., 389 Hester, R. E., 171, 1159 Hey, M. J., 1743 Hibino, T., 2327 Higgins, J. S., 170 Higuchi, A., 127, 3609 Hill, W., 691 Hirai, T., 969 Hiratsuka, H., 2809 Hisada, O., 2555 Holmberg, B., 305, 317, 2917 Holz, M., 1257 Hong, C. T., 65 Hori, Y., 2309 Horn, I. M., 1809, 3733 Howard, J., 1233 Howarth, 0. W., 121, 2035 Hubbard, C. D., 735, 3733 Humeniuk, L., 3045 Hummel, A., 991 Hunger, M., 1501 Hunter, R., 363, 633, 2875 Hussein, G. A. M., 1723 Hutchings, G. J., 363, 633, 2507, Hwang, L-P., 2335 Ichikawa, K., 175 Ikeda, R., 111 Ikeda, S., 1619 Ikeda, Y., 1099 Imamura, H., 1647 Imanishi, Y., 1065, 1075, 1083 Indelli, A., 2241, 3107 Inoue, Y., 1765 321 5 1921 2875AUTHOR INDEX Ishida, H., I l l Ishiguro, S ., 2587, 3747 Itaya, K., 1351 Ito, K., 2555 Ito, T., 2381 Itoh, N., 493 Iwasawa, Y., 441, 2021, 2597 Jackson, S . D., 3579 James, J., 2683 Jeanjean, J., 277 1 Jensen, J. B., 2649 Jiang, R., 1575, 1585 Jin, T., 175 Jobic, H., 2525 Jobson, S., 3891 Johnson, G. R. A., 677 Johnston, C., 11 11 Jonkers, G., 389 Jorgensen, N., 1111 Jozwiak, M., 2141 Juillard, J., 1337, 1709 Jutson, J. A., 55 Kalenik, J., 3187 Kaneko, K., 869, 3437, 3833 Kanno, T., 579 Karaiskakis, G., 1357 Karger, J., 1501 Kato, S., 1619 Kato, T., 2499 Katoh, T., 127 Keeler, J. H., I63 Kelebek, $., 91 Kemball, C., 2159, 2173 Kermarec, M., 1991 Khan, S.U. M., 2001 Kim, T. H., 1537, 1545, 1557 Kinjo, K., 2555 Kiraly, Z., 3373 Kishi, R., 655 Kishimoto, S., 1787 Kitajima, K., 1647 Kittaka, S., 3825 Kiwi, J., 1043 Kliachko, A. L., 3233 Klinowski, J., 1945 Klinszporn, L., 3395 Knijff, L. M., 269, 293 Kobayashi, M., 579 Koda, S., 957 Kohno, M., 3663 Kondo, Y . , 2931 Koresh, J. E., 1537, 1545, 1557 Koros, W. J., 1537, 1545, 1557 Korsunov, V. A., 3233 Kosugi, N., 869 Kotaka, T., 1065, 1075, 1083 Koutsoukos, P. G., 2465, 3159, Kozlowski, Z . , 479 Kubelkova, L., 2847 Kucherov, A. V., 2737, 3233 Kurdziel, M., 2695 Kuriacose, J. C . , 2249 JOU, F-Y., 2675 3165 Kuroda, H., 869 Kusabayashi, S., 293 1 Kuwabata, S., 969 Lamotte, J., 2397 Lancaster, N. M., 1303, 1315 Land, E. J., 2647 Lanza, P., 3107 Larcombe, M.C., 3033 Larramona, G., 907 Laschi, F., 601 Lavalley, J. C., 2397 Lawrence, D. G., 1365 Lawrence, K. G., 23 Leach, H. F., 2173 Lee, C . M., 3343 Lee, J-F., 2953 Lelj, F., 621 Lengyel, I., 3273, 3871 Lepetit, C., 1991 Lercher, J. A., 3879 Levy, O., 373 Lewis, T. J., 1009 Leyendekkers, J. V., 663 Lhermet, C., 1709 Li, C., 929, 1451 Lilley, T. H., 2901, 2909 Lillford, P. J., 2417 Linert, W., 3273 Liu, J.-Y., 1027 Liu, T., 1607 Lluch, J. M., 1207 Longdon, P. J., 1835 Lorenzelli, V., 137 Loudon, R., 169 Louis, C., 1655 Lowe, B. M., 945 Lubetkin, S . D., 1753 Luna, D., 2535 Lund, A,, 421 Machin, W. D., 2229 MacPhee, D. E., 2665 Maeda, M., 2555 Mafe, S., 223 Maffi, S., 1685 Mahmudov, A. U., 2857 Maignan, A., 783 Majerz, I., 3187 Malecka, A., 2847 Malet, P., 1279 Malik, N.A., 3245 Malitesta, C., 1685 Manchado, M. C., 1775 Mandal, H., 3045 Mann, S., 3033 Manzurola, E., 373 Marcantonatos, M. D., 2481, Marchese, L., 1655 Marchetti, A., 1697 Marchettini, N., 2149 Marcus, Y., 381, 3019 Marinas, J. M., 2535 Markovits, G., 373 Marshall, L., 2785 2605 Martin-Martinez, J. M., 3 125 Maruya, K., 929, 1451 Mashkina, A. V., 2819 Masiakowski, J. T., 421 Mastikhin, V. M., 2819 Mather, A. E., 2675 Matsuhashi, N., 111 Matsui, H., 957 Matsumoto, A., 3431 Matsumoto, H., 2369, 2809 Matsumoto, T., 175 Matthews, R. W., 1291 Maxwell, V., 3385 Mayagoitia, V., 2071, 2079 Mazzucato, U., 1469 McAleer, J. F., 783 McConnell, A. A., 3695 McLure, I. A., 1217 McMurray, N., 2047, 2055 Meima, G . R., 269, 279, 293, Melo, M-J. B. V., 2473 Menon, M. P., 2683 Merwin, L.H., 1409 Meyerstein, D., 1169 Midmore, B. R., 3529 Miessner, H., 691, 2113 Miguel, M. de G. M., 2625 Miheeva, L. M., 2857 Miller, D. G., 3343 Mills, A., 503, 2047, 2055 Mills, D., 2347 Mitani, T., 1485 Mitchell, B., 1795 Miyata, H., 3663 Miyoshi, H., 1873 Mizoe, K., 1327 Mizuno, K., 1099 Molina-Sabio, M., 3125 Morazzoni, F., 801, 258 1 Mordente, M. G. V., 2983, 2991, Morea, G., 3861 Morel, J-P., 1709, 3461 Morel-Desrosiers, N., 1709, Moreno. M. S., 2535 Mori, T., 3135 Mori, Y . , 3135 Morrison, C., 1043 Morterra, C., 1383, 2113 Mortland, M. M., 2953 Morton, J. R., 1963 Moseley, P. T., 783 Mosier-Boss, P. A., 11 Mosquera, V., 301 1 Mouaddib, N., 3413 Moulder, R., 2347 Mount, A. R., 1181, 1189, 3717 Mousset, G., 1337 Mudrakovsky, I. L., 2819 Mukherjee, T., 2647 Munuera, G., 1279 Murakami, Y ., 2327, 3135 1267 3495 346 1AUTHOR INDEX Murata, A., 2309 Murata, K., 2369, 2809 Nagai, Y., 2369, 2809 Nagaoka, T., 3757 Nagy, J. B., 2749 Nagy, 0. B., 2891 Nakagawa, T., 127, 3609 Nakamura, D., 111 Nakamura, T., 493 Nandi, D., 1531 Narayanan, S., 3785 Nastro, A., 2749 Natarajan, P., 813 Nazhat, N. B., 677 Neagle, W., 429, 719 Neilson, G. W., 1365 Newman, K. E., 485 Nicholas, A., 773 Nicol, J. M., 1233 Nikishenko, S. B, 3233 Niwa, M., 2327 Nomura, H., 957, 1619 Northing, R. J., 2273, 3451 Nosov, A. V., 2819 Nowak, R. J., 11 Nowicka, B., 479 Nunes, M. R., 907, 2473 Ndiiez, J., 3425 Ogura, K., 3757 Ohlmann, G., 691 Ohno, T., 3663 Ohtaki, H., 2587 Ohtani, K., 3825 Ohyama, Y., 749 Okubo, T., 455, 749 Olier, R., 2615 Oliva, A., 1207 Oliveira Jr, 0.N., 1009 Olivier, D., 1991 Onishi, H., 2597 Onishi, T., 929, 1451 Ono, T., 3663 Orchard, S. W., 363 Osada, Y., 655 Otsuka, K., 199 Otto, F. D., 2675 Ozeki, S., 3833 Ozutsumi, K., 3747 Packer, K. J., 3537 Pacynko, W. F., 1397 Padmaja, S., 2249 Pal, A,, 2723 Pandey, J. D., 331 Paniego, A. R., 1775 Patterson, D., 221 1 Pawlowska, M. M., 2481 Pellicer, J., 223 Pemberton, J. L. J., 2713 Pereira, I., 907 Perrichon, V., 3413 Peter, L. M., 2473 Petersen, R. L., 2435 Pethrick, R. A., 2867, 3221 Petrakis, D. E., 3173 Philippaerts, J., 3675 Phillips, C. H., 3471 Piazza, S., 3309 Piekarska, A., 3709 Pilarczyk, M., 3395 Pilkington, M. B. G., 2255 Piwowarska, Z., 47, 837 Pornonis, P. J., 3173 Pope, C. G., 945 Portanova, R., 2445 Portugal, J. M., 2705 Portwood, L., 711, 1801 Pota, G., 3871 Pradhan, J., 2797 Pratt, J.M., 2713 Preston, K. F., 1963 Preti, C., 1697 Price, W. E., 415, 1091, 3281 Primet, M., 3413 Pritchard, T. N., 1853 Pudney, P, 2635 Quarrell, R. E. L., 3451 Rai, R. D., 331 Rajaram, J., 2249 Ramakrishnan, V., 2249 Ramaraj, R., 813 Ramis, G., 137 Rao, K. J., 251 Rard, J. A., 3343 Rastogi, P. P., 3257 Rastogi, R. P., 2011 Raven, C. I., 1743 Rebenstorf, B., 3645 Reed, W. F., 349 Rees, L. V. C., 33, 1501 Reis, J. C. R., 207 Reller, A., 855 Reschetilowski, W ., 294 1 Rhodes, C. J., 711 Roberts, G., 2635 Robinson, G., 2417 Robinson, J. N., 3385 Rochester, C. H., 71, 79, 429, 719, 1111, 1117, 1129, 1841, 2983, 2991, 3495, 3505 Rodnikova, M. N., 2857 Rodriguez-Mellado, J. M., 156 Rodriguez-Reinoso, F., 3125 Rojas, F., 2071, 2079 Roman, V., 3587 Rooney, J.J., 1861 Rosen, D., 99 Rosseinsky, D. R., 3073 Rossi, C., 601, 2149 Rowlinson, J. S., 171, 172 Rubio, R. G., 3425 Ruddick, A. J., 1795 Ruiz, J. J., 1567 Ryzhikova, I. G., 3119 Saadalla-Nazhat, R. A., 677 Sabbatini, L., 1685, 3861 Sacco, A., 23, 1257 Said, M., 99 Sakata, Y., 929, 1451 Salazar, F. F., 2705 Salvagno, S., 1009 Sanchez, F., 1809 Sarkany, A., 151 1, 1523 Sartorio, R., 3289 Sato, Kazunori, 1765 Sato, Kiyoshi, 1765 Saussey, J., 2397 Sayari, A., 1963 Schmehl, R. H., 349 Schmidt, J. A., 1027 Schneider, H., 187 Schneider, I., 187 Schumann, M., 1149 Sciotto, D., 2445 Scotti, R., 801, 2581 Scurrell, M. S., 2507 Sdoukos, A. T., 3173 Seimiya, T., 2499 Sellers, R. M., 3513 Selvaraj, U., 251 Sharma, A., 201 1 Sheppard, N., 1723 Sherwood, J. N., 2867, 3221 Shido, T., 441 Shindo, Y., 1099 Shizuka, H., 2369, 2809 Shukla, R.K., 331 Sijpkes, A. H., 2563 Singh, P. P., 3797, 3807 Sinot, P. J., 1425 Slinkin, A. A., 2737, 3233 Sloth, P., 2649 Smith, E. A., 3245 Smith, E. G., 1853 Smith, G. W., 91 Smith, J. J., 1 1 Smith, M. R., 467 Smith, T. D., 645, 3195, 3215 Smith, W. E., 3695 Smith, W. J., 3623 Soares, V. A. M., 1217 Sobczyk, L., 3187 Somsen, G., 2563 Song, S., 1575 Sorek, Y., 1169 7 Ssrensen, T. S., 2649 Sozzani, P., 2581 Sparks, N. H. C., 3033 Spencer, M. S., 3537 Spivak, G. V., 2857 Spoto, G., 2113 Squire, G. D., 3561 Sreekanth, G., 3785 Stevens, A. D., 1439 Stewart, A. A., 843 Stirling, C. J. M., 1009 Strange, R., 3891 Streat, M., 3075 Stroka, J., 187 Strumolo, D., 801 Sugawara, S., 1351 Sundar, H.G. K., 251 Sunseri, C., 3309Sutton, H. C., 883 Suzuki, H., 2587 Suzuki, K., 2973 Swallow, A. J., 2647 Sykes, A. F., 3059 Symons, M. C. R., 71 1, 1439, Szczepaniec-Cieciak, E., 2695 Szejgis, A., 479, 3709 Szpak, S., 11 Szpakowska, M., 2891 Taiwo, F. A., 2427, 2435 Takagi, T., 1099 Takagi, Y., 493 Takahashi, R., 2309 Takaishi, T., 1327 Takano, S., 2499 Takisawa, N., 2099 Tamura, K., 1493 Tanabe, T., 1787 Tanaka, H., 2369 Tangari, N., 3861 Taniewska-Osinska, S., 479, 2141, 3335, 3709 Taniguchi, S., 3135 Tashiro, T., 2381 Tassi, L., 1697 Taylor, D. M., 1009 Terai, M., 1493 Thamm, H., 1 Themistocleous, T., 633 Theocharis, C. R., 2641 Thomas, J. M., 1945 Thumser, A. E., 3845 Tiddy, G. J. T., 1397 Timmermann, E. O., 1631 Tissier, M., 1337 Toi, K., 2381 Tolazzi, M., 2445 Tomat, G., 2445 Tondre, C., 3301 Tonokura, K., 2369, 2809 Tortorella, V., 3861 Tosi, G., 1697 Traboulssi, R., 2705 Trickett, K.A., 3281 1801, 2427, 2435, 3245, 3257 AUTHOR INDEX Tsang, S. C., 3561 Tsuchiya, S., 1647 Tsutsumi, K., 1327 Tsyganenko, A. A., 2397 Uchida, N., 3825 Uchida, Y., 3757 Uchiyama, H., 3833 Ugliengo, P., 585, 1373 Ulgiati, S., 2149 Ulman, L., 2695 Unger, B., 2941 Unwin, P. R., 1821 Urch, D. S., 1139 Vaccari, A,, 237 Valencia, E., 3481 van Buren, F. R., 269, 279, 293, Van-Den-Begin, N., 1501 van der Riet, M., 2875 van Dillen, A. J., 269, 279, 293, van Leur, M. G. J., 279 van Lith, D., 991 van Rensburg, L. J., 633 Vansant, E. F., 3087, 3095, van Veen, J. A. R., 389 van Vliet, T., 3359 Vazquez-Gonzilez, M.I., 1019 Vedrine, J. C., 1607 Vedrine, J. C., 2615 Verbiest, J., 3087, 3095 Villar, V. P., 3011 Vink, H., 699 Vis, R. J., 269, 279 Vuilleumier, J-J., 2605 Wacker, T., 33 Walker, D. R. B., 1545, 1557 Walker, P. A. M., 1365 Walker, S., 3045, 3145 Waller, A. M., 773, 977 Walstra, P., 3359 Wang, P-L., 2335 Wang, Y-P., 2199 Warman, J. M., 991 Watanabe, T., 2381 1267 1267 3675, 3687 Waugh, K. C., 163 Weale, K. E., 165 Weitkamp, J., 2127 Wells, C. F., 2185, 2999 Wendlandt, K-P., 2941 West, R., 2369 Wilkinson, D. P., 2355 Willett, M. J., 1907, 1921 Williams, D. E., 783 Williams, G., 503 Woodhouse, J. R., 2507 Woolf, L. A., 1091 Wormald, C. J., 1303, 1315 Woinicka, J., 1709, 3335 Wright, J. D., 1979 Wurzburger, S., 3289 Wyn-Jones, E., 2099 Xyla, A. G., 3165 Yamada, Y., 609 Yamamoto, Y., 3353 Yan, Y., 3087, 3095 Yao, Z., 2211 Yarwood, J., 1397 Yeates, D., 2641 Yeh, C., 2199 Yeh, C-t., 65 Yoneyama, H., 969, 1873 Yoon, C.S., 2867, 3221 Yoshida, N., 1787 Yoshioka, H., 1485 Yoshitake, H., 2021 You, X., 829 Young, D. A,, 173 Zaki, M. I., 1723 Zambonin, P. G., 1685, 3861 Zanonato, P., 2445 Zaslavsky, B. Yu., 2857 Zecchina, A., 609, 1655, 2113 Zhan, R., 1599 Zhao, X., 3819 Zhu, J., 3629, 3645 Zielinski, R., 1619 Zukoski IV, C. F., 2785 Zhu, B-Y., 3813, 3819 (viii)THE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY SYMPOSIUM No. 25 Large Gas Phase Clusters University of Warwick, 12-14 December 1989 Organising Committee: Professor K. R. Jennings (Chairman) Professor P. J. Derrick Professor D. Phillips Dr N. Quirk Dr R. P. H.Rettschnick Dr A. J. Stace The Symposium will focus on recent developments in the rapidly expanding field of large gas phase clusters, including the preparation, structure and reaction of both neutral and ionic dusters. It is hoped that the meeting will bring together scientists working on many different types of duster, e.g. rare gas atoms, metals, inorganic and organic species, and biomolecules, to discuss the chemistry and physics of clusters from different viewpoints. The final programme and applic2ition form are now available and 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. 89 Structure of Surfaces and Interfaces as Studied using Synchrotron Radiation University of Manchester, 4-6 April 1990 Organising Committee: Professor J.N. Sherwood (Chairman) Professor D. A. King Dr G. King Dr C. Norris Dr R. Oldman Dr G. Thornton The Discussion will focus on the wealth of novel information which can be obtained on the nature and structure of surfaces using the full spectral range of synchroton radiation. Emphasis will be placed on the scientific results of recent investigations rather than on technical aspects of experimentation. Papers will be welcome which shed new light on the structure of the complete range of interfaces: solid/solid, solid/gas, solidlliquid, gaslliquid and "dean" surfaces including both static and dynamic in sifu examinations. It is hoped that the discussion will define the utility of synchroton radiation examinations in surface science studies at a time of expansion of the availability of such sources and the inauguration of new and more powerful sources.The preliminary programme is now available and 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 ASSOCIAZIONE ITALIANA DI CHIMICA FlSlCA DEUTSCHE BUNSEN-GESELLSCHAFT FUR PHYSIKALISCHE CHEMIE KONINKLIJKE NEDERLANDS CHEMISCHE VERElNlGlNG SOCIETE FRANGAISE DE CHIMIE, DIVISION DE CHlMlE PHYSIQUE FARADAY DIVISION GENERAL DISCUSSION No. 90 Colloidal Dispersions University of Bristol, 10-12 September 1990 Orga nising Com mitte e Professor R. H. Ottewill (Chairman) Professor P.Botherol Professor E. Ferroni Or J. W. Goodwin Professor H. Hoff mann Professor A.L. Smith Professor P. Stenius Dr Th. F. Tadros Professor A. Vrij Dr D. A. Young The joint meeting of the Societies will be directed towards examining current understanding of the behaviour of colloidal dispersions. In particular, stability and instability, short range interactions, dynamic effects, non-equilibrium interaction, concentrated dispersions and order-disorder phenomena will form topics for discussion. The preliminary programme is now availablemay 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 SYMPOSIUM No. 26 Molecular Transport in Confined Regions and Membranes Oxford, 17-18 December 1990 Experimental, theoretical and simulation studies which address fundamental aspects of molecular transport will be discussed in the following main areas: a) Transport of atoms and molecules in pores, zeolite networks and other cavities; time-dependent statistical mechanics of small systems in confined geometries b) Molecular transport through synthetic membranes, biological membranes, smectic liquid crystalline phases and Langmuir Blodgett films; the dynamics of the molecules forming the membrane c) Diffusion, reorientation, conformational dynamics, viscosity and conductivity of polymer melts, to include papers dealing with bulk systems since the segments of the polymer will move in the anisotropic environment of the complete chain d) Applications of Brownian dynamics to the study of diffusion in porous media and across membranes including the transport of flexible aggregates such as microemulsions e ) The growth of crystals, colloidal aggregates and droplets on irregular surfaces and in pores Contributions for consideration by the Organising Committee are invited and abstracts of about 300 words should be sent by 31 December 1989 to: Dr D. J. Tildesley, Department of Chemistry, The University, Southampton SO9 SNH. Full papers for publication in the Symposium Volume will be required by August 1990.
ISSN:0300-9599
DOI:10.1039/F198985BP145
出版商:RSC
年代:1989
数据来源: RSC
|
5. |
Membrane potential and ion transport in inhomogeneous ion-exchange membranes |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 85,
Issue 11,
1989,
Page 3609-3621
Akon Higuchi,
Preview
|
PDF (789KB)
|
|
摘要:
J. Chem. SOC., Faraday Trans. I, 1989, 85(11), 3609-3621 Membrane Potential and Ion Transport in Inhomogeneous Ion-exchange Membranes Akon Higuchi" and Tsutomu Nakagawa Department of Industrial Chemistry, Faculty of Engineering, Meiji University, 1-1-1 Higashi-mita, Tama-ku, Kawasaki, Kanagawa 214, Japan The Nernst-Planck equation is modified for the transport of uni-univalent ions in inhomogeneous membranes, which contain continuous gradients of partition and activity coefficients of the ions and/or fixed charge density. Standard chemical-potential gradients in the membranes are employed in the equation as a new force. Numerical solutions of membrane potential and ion fluxes for 10 model membranes are obtained from simulation at the condition of zero current. Various potential and concentration profiles are observed, which depend on the partition coefficients and/or the fixed charge density in the membranes.The membrane potential, which arises between two aqueous solutions of an electrolyte separated by a charged membrane, has been the subject of theoretical and experimental studies. Theoretical treatment of the membrane potential and ion transport in a homogeneously charged membrane was developed by Teorell,' Meyer and Severs2 (TMS theory) and other investigator~.~-~~ There are also some fundamental ~ t u d i e s l ~ - ~ ~ for inhomogeneously charged membranes. Gierke and Hsu's'~ proposed fixed charge of Nafion was described as a cluster-channel network, and inhomogeneities in the fixed charge density in the membrane gave rise to a potential within the membrane which impeded co-ion transport relative to that of counterions.Selvey and Reissl' considered fluctuations in the fixed charge density due to ion clustering and solved the Nernst-Planck and Poisson equations using perturbation theory. Sonin & Grossman" studied membrane potential and ion transport in layered ion-exchange membranes, although their primary interest is the current-voltage characteristics across the membranes. 21 observed an unusual asymmetric potential, between two identical electrolyte solutions separated by composite mem- branes, having different charge density, or by asymmetrically charged membranes. The asymmetric potential is explained by the existence of the water phase between the mem- branes, which has lower concentration of ions than the external ~ o l u t i o n s .~ ~ The potential is generated by the permeation of ions into the membranes at the steady state, but not in the equilibrium. Takagi and Nakagaki22 reported interesting results of the facilitated and reverse transport of ions in inhomogeneous membranes, asymmetric with respect to the partition coefficients of ions and/or the fixed charge density. They tried to verify their results from their theory. Their equation, however, leads to permanent ion fluxes even for a situation in which two external solutions separated by the asymmetric membrane are identical and without consideration of a water phase in the membrane. It is obvious that this violates the first law of thermodynamics. However, their results are a product of the nature of the asymmetric potentials and should be regarded as transient in character.Nevertheless, the study has prompted us to try to develop an exact theory of membrane potential and ion transport in the inhomogeneous membranes. Liquori and BOtre,lg and other 36093610 In homogeneous lon-exchange Membranes we developed the theory for a multilamellar series array of ion- exchange membranes having independent partition coefficients and charge densities. The goal of this study is to propose a modified Nernst-Planck equation for inhomogeneous membranes that have continuous gradients in the partition coefficients of ions and/or fixed charge density. In our previous Model Membranes considered in this study have continuous gradients of ionic partition and activity coefficients and/or of the fixed charge density along the x axis (taken to be perpendicular to membrane surfaces).The membrane is defined to occupy the region 0 6 x 6 1. On either side the membrane is in contact with aqueous electrolyte solutions having concentrations C,, at x = 0 and C, at x = 1. In order to emphasize the essential points of an ideal system, the following assumptions were made. (a) All charges are considered to be point charges and ionic dimensions are neglected. (b) Gradients of the partition coefficients, activity coefficients, the fixed charge density, diffusion coefficients, ion concentration and electric potential only exist along the x axis in the membrane. (c) Hydrostatic pressure gradients and volume movements of the fluid are regarded as negligible. ( d ) Electroneutrality is observed in all parts of the membrane (current is not generated in the system).( e ) Anion and cation fluxes are equal in any part of the membrane. (f) The total membrane potential Aq5 is given as the sum of the Donnan potential at the two interfaces between the membrane and the external solution, A#Don, and potential generated inside the membrane, (g) Equilibrium conditions are maintained at the surfaces between the external solution phase and the membrane surfaces. (h) The external solution is a uni-univalent salt solution. (i) Activity coefficients of ions in the external solution are unity. ( j ) The diffusion and partition coefficients of ions and the fixed charge density in the membrane are constant or dependent on ion concentration, Theory General Equation Assumptions (d) and (f) lead to A# = '#Don +'$in C-(x) = C+(x) + wC,(x) where C+(x) and C-(x) are the concentration of cation and anion at x in the membrane, cu has a value of + 1 or - 1 for a positively or negatively charged membrane and C,(x) is the fixed charge density at x in the membrane.Electrochemical potentials of cation and anion at x in the membrane, ~ + ~ ( x ) and pTm(x), are given by (3) (4) where pYrn(x) and ,uU",(x) are the standard chemical potentials of cation and anion at x in the membrane, y+ and y- are the activity coefficients of cation and anion at x in the membrane, qhrn(x) is the electric potential at x in the membrane, F is the Faraday constant, R is the gas constant and T is the absolute temperature.p+rn(x) = ~Yrn(x) + RTln Y+ C+(X) + f'$rn(x) P-~(X) = p?,(x) + RTln y- CJx) - Fq5,(x)A . Higuchi and T. Nakagawa 361 1 Donnan Potential When the membrane is assumed to be at equilibrium in the external solution with concentration C+soln and CPsoln, the electrochemical potential is given as ( 5 ) (6) Where pLo+soln and pZsoln are the standard chemical potentials of cation and anion, respectively, in the external solution, and Qsoln is the electric potential in the solution. We define (7) (8) po+soln + C+soln + F$soln = po+m(x) + RTln Y+ C+(x) + F # m POsoln + RTln C-soln-F#soln = p:m(X) + RTln y- C - ( X ) - F # ~ . pu",m(x> -p;soln = - RTln K+ pO,(x) -pOsoln = - RTln K - . Qm - #soln = ( - RT/F) In [Y+ C+(X)/K+ C+solnl = ( - RT/F) In [ K - C-soln/~- C-(x)].Then Qm - $soln is (9) (10) From this we obtain C + s o l n C-soln(K+ K-/Y+ Y-1 = C+(X> C-(X>* With K+(x) = K+/Y+ and K-(x) = rc-/y-, and C+soln CPsoln = C : or C:, we obtain Finally, by combining eqn (2), (1 1) and (1 2), the concentration at the surface is given as Flux Equation The fluxes of cations and anions, J+ and L, respectively, are given by J+ = - D+(x) C+(x) dpu,,(x)/L dx = - D+(x) C+(x) [RTd In y+ C+(x)/dx + Fd#,/dx + dpo+,(x)/dx]/l (1 6) J- = - D-(x) C-(x) [RTd In y- C-(x)/dx - Fd#,/dx + dpu",(x)/dx]/l (1 7) where D+(x) and D-(x) are the mobilities of the cations and anions at x in the membrane and L is membrane thickness. Since the fixed charge density and the partition coefficients of ions give different values at different places in the inhomogeneous membranes, the standard chemical potential will not be constant but will be a function of position x in the membrane.dp;,(x)/dx and dpO,(x)/dx cannot therefore, be regarded as zero for the imhomogeneous membranes, while eqn (1 6) and (1 7) reduce to the conventional Nernst-Planck equation for homogeneous membranes neglecting dp;,(x)/dx and dpu" (x) / dx.3612 Inhomogeneous Ion-exchange Membranes Eqn (22) and (23) are applicable in cases where the diffusion coefficients, the partition coefficients and the fixed charge density are the function of not only the position x, but the concentration from the derivation processes of the equations. The permeability coefficient, P, is finally calculated from P = J+ L/ABS(Co - Cl).(24) Note that at steady state P should not depend on x. Computational Method To obtain the membrane potential and the flux analytically we must integrate eqn (22) and (23), which is, at present, difficult. However, a numerical solution for the flux and the integration of eqn (22) can be achieved using a computer. If we know the flux, concentration profiles in the membrane can be built up stepwise, via eqn (23), from dC+(x)/dx = {J+ L / [ D + ( X ) RTI + [m2 - K:(x)/K+(x)l C+(x)>/(m, - 1). (25) Computational procedures are shown as follows. (1) Set the control data on the situation to be studied and define the membrane model [the definition of the functions of C,(x), K+(x), K-(x), D+(x) and D-(x), and read T, L, u, Co and C,]. (2) Set Ax which is small compared with unity (0 < Ax $ 1).(3) Calculate C+(O) and C+(l) from eqn (13). (4) Input J+ as an initial value, and set n = 1.A . Higuchi and T. Nakagawa 3613 Table 1. Model membranes for the calculation. D+(x) = 5.382 x lo-’ cm2 rnol s-’ J-l, D-(x) = 8.201 x lo-’ cm2 mol s-l J-l, T=298Kandc;o=-l membranes C,(x)/equiv. dm-3 K(x) model I a model I b model IIa model I1 b model IIIa model IIIb model IV model V model VIa model VIb 0.0 1 0.0 1 0.0 199 - 0.0198~ 0.0001 +0.0198~ 0.0 1 0.0 1 0.005 + 0.03~-0.03~* 0.015-0.03~+0.03~~ 0.002+0.016x 0.018-0.016~ 1 0.5 1 1 1.8-1.6~ 0.2 + 1 . 6 ~ 0.25 + 1 . 5 ~ + 1 .5x2 0.1 +0.8x 0.75 - 1 . 5 ~ + 1 .5x2 0.9 - 0 . 8 ~ ( 5 ) Calculate dC+[(n - 1) Ax]/dx from eqn (25). (6) Calculate C+(nAx) from eqn (26): C+(nAx) = C+[(n - 1) Ax] + AxdC+[(n - 1) Ax]/dx.(26) (7) Set n = n+ 1 and repeat ( 5 ) and (6) until n = l/Ax. (8) If C+(nAx) at n = l/Ax > C+(l), new 4. is generated under the condition that (9) If C+(nAx) at n = l / A x < C+(l), new J+ is generated under the condition that (10) If C+(nAx) at n = l/Ax is approximately equal to C+(l), J+ and C+(x) are (1 1) Integrate eqn (22) from Simpson’s equation and calculate Aq5in(A$in = Ji Qm dx). (12) Calculate ADon and A$ from eqn (1) and (15). (13) Print out all results. J+(new) = J+(old)a (1 < a at C, > C,, 0 < a < 1 at C, < C,). Go to ( 5 ) with n = 1. J+(new) = J+(old)b(O < b < 1 at Co > C,, 1 < b at C, < C,). Go to ( 5 ) with n = 1. obtained from the above procedures. Go to (1 1). Results and Discussion Model Membranes The method proposed above was used to calculate the membrane potential and the flux for some model membranes.In order to emphasize the essential points, the calculations performed in this study were confined to the following conditions : co = - 1, T = 298 K, dD+(x)/dx = 0, dD-(x)/dx = 0, K+(x) = K-(x), Ji K+(x) dx = 0.5 or 1.0, j’: C,(x) dx = 0.01 equiv. dm-3 and Ax = 0.001. The values of D+(x) and D-(x) for sodium ion and chloride ion in bulk water, were chosen to be 5.382 and 8.201 x cm2 mol s-’ J - I . C,(x) and K+(x) for the 10 models addressed in this study are summarized in table 1 . Calculations were performed using a 16-bit personal computer (PC-980lVX, NEC Corp.) with N88 B~sIc(86) language on MS-DOS ver. 3.10 (Microsoft Corp.).Membrane Potential Fig. 1 shows membrane potentials of models Ia, IIa and IIb. Model I a represents a homogeneous cation-exchange membrane with C,(x) = 0.01 equiv. dm-3 and K(x) = 1. Models I1 a and I1 b are inhomogeneous membranes having asymmetric charge density, and model IIa is identical to model IIb with values of Co and C, interchanged. The three models give different membrane potentials at C, > low3 mol dm-3, although the models have the same values of Ji C,(x) dx and JiK(x) dx.3614 120 80 40 -40 In homogeneous Ion-exchange Membranes -80 -120 . 1 L 1 o - ~ 1 o - ~ lo-* lo-' 1 C, /mol dm-3 Fig. 1. Membrane potentials calculated for models I (-), IIa (----), IIb (--) and TMS fitting curve (..-....... ) with K(x) = 1, C,(x) = 0.0307 equiv. dmP3 and D+/D- = 0.0325.C , = mol dm-3. We can judge whether a given membrane is inhomogeneous by observing two separate membrane potentials under the same conditions; for example, if Co is varied from lop3 to 1 mol dm-3 with C, = mol dm-3 and likewise C, is varied with C, = 10-3 mol dm-3, TMS theory predicts the two potentials to be identical. Model IIb is also concluded to be the inhomogeneous membrane from the fact that the membrane potential observed in model I1 b can not be explained by TMS theory. The best fit of the membrane potential from eqn (1) of ref. (13) (TMS theory) using a non- linear least-squares method gives a negative value of the fixed charge density ( - 7.80 x 10-3eq/l). The fixed charge density and ion-mobility ratio (D+/D-) can also be estimated in the TMS theory from maximum membrane potential and C, at A+ex (C0-e,).13 Membrane potential calculated by this procedure with A#ex = 28.84 mV, CO-ex = 5.725 x mol dm-3, C,(x) = 0.0307 equiv.dm-3 and D+/D- = 0.0325 is also shown in fig. 1. The figure suggests that TMS theory explains the membrane potential at Co < 0.03 mol dmP3, but the discrepancy of the membrane potential between TMS theory and model I1 b gradually increases with the increase of C, at Co 2 0.03 mol dm-3. This is due to an inflection point at C, ca. 0.1 mol dmF3 observed in the membrane potential of model IIb. It is known that the membrane potential for a bipolar membrane, which consists of juxtaposed cation and anion exchange membranes, also shows the inflection point in some cases.23 Since the bipolar membranes can also be regarded as the inhomogeneous membrane, the inflection point found in the membrane potential should be a characteristic aspect for the inhomogeneous membrane. Fig.2 shows the membrane potentials of models IIIb, IV, V, VIa and VIb. The membrane potential of model IIIa, which is not shown in the figure, is estimated to be ca. identical to that of model V at Co < 0.5 mol dm-3. Models IIIa and I11 b are the inhomogeneous membranes with respect to the partition coefficients and not to the charge density, and model I11 a is identical to model I11 b with the values of C, and C, interchanged. Models IV or V are symmetrical for the charge density and the partition coefficients about the line of x = 0.5 in the membranes. Models VIa and VIb are constructed on the conditions that K(x) = H(x) and 1 -H(x) is linearly proportional toA .Higuchi and T. Nakagawa 361 5 120 80 40 $ 0 \ a d - 4 0 -80 -120 /----- -/*- .A- /----- . I 1 lo-& 1 o - ~ lo-* lo-’ 1 c,, /mol dni3 Fig. 2. Membrane potentials calculated for models 111 b (-), IV (----), V (---), VIa (-a*-) and VIb (.......... ). C, = mol dmV3. CJx) where H(x) is water content at x in the membrane. The model VIa is identical to model VIb with the values of Co and C, interchanged. It is found in fig. 2 that the membrane potentials of models IIIa and VIa also show different values estimated from the potentials of models IIIb and VIb, although membrane properties between models IIIa and IIIb, or models VIa and VIb are identical except in the direction of ion fluxes.Models I a, I1 a, I1 b, I11 a and I11 b have the properties of Ji K(x) dx = 1 and Ji CJx) dx = 0.01 equiv. dm-3, and models I b, IV, V, VI a and VI b have the conditions of JiK(x) dx = 0.5 and Ji CJx) dx = 0.01 equiv. dmP3. The models, however, give different potentials according to the membrane model. It is found that the macroscopic properties of the membrane such as Ji K(x) dx and Ji C,(x) dx cannot be estimated solely from the results of the inhomogeneous membrane potential, whereas the properties can be determined if the membrane is homogeneous. This is because the profiles of the partition coefficients and charge density in the membrane significantly influence the membrane potentials, as found in fig. 1 and 2 . Permeability Difference Another index which characterizes the given membrane as an inhomogeneous one is the variation of the ion flux with direction.Fig. 3 shows the ratios of Pb to Pa for models 11, I11 and VI where Pa and Pb are the permeability coefficients of models Y, and yb ( Y = 11, I11 or VI). Since models IV and V are identical owing to the symmetry about the line of x = 0.5, permeability differences with direction are not observed although models IV and V describe inhomogeneous membranes. Maxima were observed in the plots of Pb/Pa against C, for models 11, I11 and VI. p b / p a for model VI was found to be higher than that for models I1 and III. Similar maxima were reported for the bipolar membranes in a previous It was suggested that the permeability difference due to the direction of flux is caused by the diffusion potential difference resulting from the flux direction in the bipolar rnembrane~.~~ The permeability difference was, however, observed even in3616 4 0 20 Inhomogeneous Ion-exchange Membranes .I 1.4 1.3 s 1.2 1.1 1.0 CL" 10-3 1 o-2 10" 1 C , /mol dm-3 Fig. 3. Permeability ratios of Pb to Pa calculated for models I1 (-), I11 (----) and VI (---). C, = mol dm-3. -0 3 -20 -40 -60 > 3 a d h - 8 0 t -100 0 0.2 0.4 0.6 0.8 1.0 Fig. 4. Membrane potential profiles calculated for models I a (-), I1 a (----), I1 b (---), I11 a (_.._) and IIIb (.......... ). the case of D+(x) = D-(x) = 5.382 x lo-' cm2 mol J-l s-' for model VI (e.g. Pa = 2.546 x cm2 s-l and Pb = 3.535 x lo-* cm2 s-l on the conditions of Co = 0.06 mol dm-3 and C, = 0.001 mol dm-3).It is found that the potential generated inside the membrane such as the standard chemical potential gradient also plays an important role for the permeability difference in the inhomogeneous membranes.A . Higuchi and T, Nakagawa 3617 60 40 20 - 0. 3 -20 - 4 0 -60 > h v Y a -a -80 -100 0' \ 1 I I . 0 0.2 0.4 0.6 0.8 1.0 Fig. 5. Membrane potential profiles calculated for models I b (-), IV (----), V (--.-), VIa (-. .-) and VI b (. . . . . . . . . . ). C, = 0.01 rnol dm-3 and C, = mol dm-3. Table 2. Patterns of C,(x), K(x), C+(x), CJx) and A$(x) in model membranesa membranes C,(x) K(x) C+jx) CJx) A$(x) model I a model Ib model IIa model IIb model IIIa model IIIb model IV model V model VIa move1 VIb I L I 7 L 7 u n I f a -+, Constant; 7, increase from Co side to C, side; I , decrease from C, side to C, side; n, convex upward from C, side to C, side; u, convex downward from C,, side to C, side.Potential Profiles Membrane potential profiles for 10 model membranes are calculated by integrating eqn (22) over distance, x, in the membranes. The results at C, = 0.01 mol dm-3 and C, = 0.001 mol dm-3 are depicted in fig. 4 and 5. The potentials of models IV and V show convex profiles in the membrane where the profiles of the partition coefficients and the charge density also show convex function in the membrane (see table l), and are completely different profiles from those estimated from TMS theory.3618 Inhomogeneous Ion-exchange Membranes 2.8 2.4 0.8 0.4 0 0 0.2 0.4 0.6 0.8 1.0 Fig. 6.Concentration profiles of cation calculated for models Ia (-), I1 a (----), I1 b (-*-), IIIa (--..-) and IIIb (..-...-... ). C, = 0.01 mol dm-3 and C, = mol dm-3. 0 0.2 0.4 0.6 0.8 1.0 Fig. 7. Concentration profiles of cation calculated for models I b (-), 1V (----), V (-.--), VI a (_.._) and VIb (.......... ). C,, = 0.01 mol dm-3 and C, = mol dm-3.A . Higuchi and T. Nakagawa 3619 1.4 1.2 CO m -1.0 'E 0.8 k 2 0.6 U - \ I L, 0.4 0.2 0 0 0.2 0.4 0.6 0.8 1.0 Fig. 8. Concentration profiles of anion calculated for models I a (-), I1 a (----), I1 b (-.--), X I11 a (-a*-) and I11 b (-.......-. ). C,, = 0.01 mol dm-3 and C, = mol drnp3. The inhomogeneous membranes in this study suggest that the membrane potential increases with decreasing fixed charge density or with increasing partition coefficients along the distance, x, and the potential decreases with increasing fixed charge density or with decreasing partition coefficient along the x axis (see table 2).The potential profiles are, therefore, directly related to the functions of the fixed charge density and the partition coefficients, and may be valuable properties to estimate the functions of CJx) and K(x) if the potential profiles can be directly obtained from the experiments. Concentration Profiles Concentration profiles of cation and anion for 10 model membranes are calculated by eqn (2) and (26). The concentration profiles at Co = 0.01 mol dmP3 and C, = 0.001 mol dm-3 are shown in fig. 6-9. Several patterns of concentration profiles are observed in the figures.Convex profiles were observed for C+(x) in models IIIb, IV and V, or C-(x) in models I11 b and V. It was also found that the concentration in the membrane increases from x = 0 to x = I for C+(x) in models I1 a and VIa, or C-(x) in model VI b, although these profiles are opposite to the flux direction. Table 2 summarizes the patterns of C,(x), K(x), C+(x), C-(x) and A$(x) in the membranes. There seems to be some relationship between C+(x) and the function of the fixed charge density in the membranes, while any relationship between CJx) and Cz(x) or K(x) is not observed in table 2. Evidently, the relationship between C+(x) and Cz(x) should be observed if the fixed charge density is in a high-concentration region, or C, and C, are in low- concentration regions. This is because the counter-ion concentration, C+(x) in this case, is mainly determined by C,(x) from eqn (2). The flow direction for the 10 membranes of this study having various concentration profiles was observed to be from high concentration to low concentration, although some concentration profiles were against the flow direction.3620 Inhomogeneous Ion-exchange Membranes 1.4 1.2 CO -1.0 'E ,.,' 0.8 k 0.6 d 0.4 0.2 0 U M \ h 0 0.2 0.4 0.6 0.8 1.0 Fig.9. Concentration profiles of anion calculated for models I b (-), IV (----), V (----), VI a (_.._) and VIb (.......... ). C, = 0.01 mol dm-3 and C, = mol dm-3. Comments on Reverse Transport Takagi and Nakagaki22 studied theoretically and found experimentally ' the reverse transport ' of NaCl through inhomogeneous membranes such as NaOH- treated collodion-collodion, where this reverse transport refers to solutes permeating in an opposite direction to the concentration gradient between the two external solutions of the membranes.Their experimental results can be related to the phenomena of the asymmetric p~tentiall'-~l and should be regarded as of temporary nature, since for such reverse transport not to require any energy is obviously against the first law of thermodynamics. One reason why their theory led to such a misleading result is that they applied the conventional Nernst-Planck equation to the inhomogeneous membranes, and the standard chemical potential gradient in the membranes, which is considered in this study, was not considered in their theory. They also demonstrated reverse transport even for a membrane having K(x) = constant and asymmetric charge distribution.In this case the modified Nernst-Planck equation reduces to the conventional Nernst- Planck equation because of dK(x)/dx = 0. Therefore, in this case, we cannot attribute the reverse transport solely to a disregard of the standard chemical potential gradient. There was also the assumption presented below according to in order to solve flux equation analytically, uiz. C+(x) = c1 - w1 C+(O) + w C+( 1) C-(x) = [ 1 - 6(x)] C ( 0 ) + 6(x) c-( 1) (27) (28) where d(x) is a continuous function of x that does not depend on the concentration, and is the same for C+(x) and C-(x). By combining eqn (27) and (28), we obtainA . Higuc hi and T. Nakaga w a 362 1 On the other hand, the differential of eqn (2) is represented by eqn (30).dC+(x)/dx = dC-(x)/dx - dC,(x)/dx. (30) There is no gradient of fixed charge density in eqn (29), whereas there is such a gradient in eqn (30). Eqn (29) and (30) reduce to the same equation in the special case of dC,(x)/dx = 0 : dC+(x)/dx = dC-(x)/dx. (31) In our study, the membrane potential and ion fluxes were calculated by the simulation method, and eqn (27) and (28) were not used. It is concluded that eqn (27) and (28), the validity of which is not clear, lead to the misleading equations in their theory. Although we found that Takagi and Nakagaki22 treated their equation inadequately for the transport of ions in the inhomogeneous membranes, it should be noted that it was their studies which stimulated and prompted us to develop the present investigation. References 1 T. Teorell, Proc. SOC. Exptl. Biol. Med., 1935, 33, 282. 2 K. H. Meyer and J. F. Sievers, Hehi. Chim. Acta, 1936, 19, 649. 3 K. S. Spiegler, J. Electrochem. Soc., 1953, 100, 303. 4 R. Schlogl, Z . Elektrochem., 1952, 56, 644. 5 G. Schmid, Z. Elektrochem., 1950, 54, 424; G. Schmid, Z. Phys. Chem., N.F., 1954, 1, 305. 6 M. Nagasawa and 1. Kagawa, Discuss. Faraday Soc., 1956, 21, 52. 7 N. Kamo, Y. Toyoshita, H. Nozaki and Y. Kobatake, Kolloid-Z. Z. Polym., 1971, 248, 914. 8 H. U. Demisch and W. Pusch, J. Colloid Interface Sci., 1979, 69, 247. 9 N. Minoura and T. Nakagawa, Koubunshi Ronbunshu, 1980, 37, 761. 10 T. Kinoshita, T. Yamashita, T. Iwata, A. Takizawa and Y. Tsujita, J. Macromol. Sci., Phys., 1983, B22, 1 . 1 1 H. Vink, Acta Chem. Scand., 1979, A33, 547. 12 Y. Kimura, H-J. Lim and T. Iijima, J. Membrane Sci., 1984, 18, 285. 13 A. Higuchi and T. Iijima, J. Appl. Polym. Sci., 1986, 31, 419. 14 N. Kamo and Y. Kobatake, J. Colloid Interface Sci., 1974, 46, 85. 15 T. D. Gierke and W. Y. Hsu, in Perfluorinated Ionomer Membranes, ACS Symposium Series, No. 180 (American Chemical Society, Washington, D.C., 1982), p. 283. 16 C. Selvey and H. Reiss, J. Membrane Sci., 1985, 2, 1 1 . 17 A. A. Sonin and G. Grossman, J. Phys. Chem., 1972, 76, 3996. 18 W. H. Koh and H. P. Silverman, J. Membrane Sci., 1983, 13, 279. 19 A. M. Liquori and C. Botre, J. Phys. Chem., 1967, 71, 3765. 20 N. Lakshminurayanaiah and F. A. Siddiqi, Biophys. J., 1971, 11, 617. 21 F. de Korosy, J. Phys. Chem., 1968, 72, 2591. 22 R. Takagi and M. Nakagaki, J. Membrane Sci., 1986, 27, 285. 23 A. Higuchi and T. Nakagawa, J. Membrane Sci., 1987, 32, 267. 24 P. Henderson, Z. Phys. Chem., 1907, 59, 118. Paper 8/03684E ; Received 21st September, 1988
ISSN:0300-9599
DOI:10.1039/F19898503609
出版商:RSC
年代:1989
数据来源: RSC
|
6. |
Zeolite synthesis in the SiO2–Al2O3–Na2O–pyridine–H2O system |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 85,
Issue 11,
1989,
Page 3623-3628
W. J. Smith,
Preview
|
PDF (866KB)
|
|
摘要:
J. Chem. Soc., Furaday Trans. I , 1989, 85( 1 l), 3623-3628 Zeolite Synthesis in the Si0,-Al,O,-Na,O-Pyridine-H,O System W. J. Smith, J. Dewing and J. Dwyer Department of Chemistry, UMIST PO Box 88, Manchester The crystallisation of zeolites from the SO,-Al,O,-Na,O-pyridine-H,O system has been investigated. Three zeolitic phases were isolated : mordenite, ferrierite and ZSM-5. Only the ferrierite phase is specifically directed by the pyridine molecules. Depending upon the composition of the reaction mixture, mordenite or ZSM-5 can be synthesised in the absence of the organic additive. The ferrierite phase is formed over a wide range of gel compositions although the yield and purity of the sample is dependent upon the aluminium content of the gel. The ferrierite phases formed are relatively siliceous and are all within the composition range SiO,/Al,O, = 25-38.Cu. 2-3 pyridine molecules are occluded per unit cell within the ferrierite channels. Since the early 1960s the addition of organic species to zeolite reaction systems to promote the growth of crystalline materials with novel structures or compositions has been of continued interest. Many reviews concerned with the role of such organic bases in the crystallisation of zeolites are available in the literat~re.'-~ As the number of exotic organic compounds claimed in the synthesis of novel or known zeolite structures increases it is pertinent, from an industrial viewpoint, to investigate the more commercially available and less expensive organic synthesis systems. One such system which has received little attention in the zeolite literature involves the use of pyridine.The use of pyridine as an organic additive, originally reported by van Erp et al.,5 reveals the existence of several crystalline phases within the synthesis system. A more recent patent by Morimoto et al. details the synthesis of a novel zeolite ISI-6 using pyridine and ethane- 1,2-diol as organic additives.6 The present paper reports the crystalline phases obtained in the pyridine synthesis system as reaction parameters such as the SiO,/Al,O, and OH/SiO, reaction ratios are varied. Additionally the role of pyridine in the crystallisation of the separate phases is discussed. Experimental Zeolite syntheses were performed in 140 cm3 capacity stainless steel autoclaves at a temperature of 175 "C, the autoclaves were not agitated during the reactions.The reagents used were aqueous sodium silicate (Pyramid no. 1 sodium silicate, 29.25 wt YO SiO,, 8.85 wt YO Na,O, 61.9 wt % H,O), aluminium nitrate (BDH), pyridine (Raught Ltd), nitric acid (May & Baker, 70 wt %) and distilled water. Reaction mixtures were prepared as follows. The aluminium nitrate was dissolved in water and then nitric acid and pyridine were added to form a solution A. The remaining distilled water was added to the sodium silicate and the mixture was added dropwise to solution A over a period of 20-40 min with agitation. The pH of the resulting gel was recorded before it was transferred to the autoclave. At various stages during the reactions the autoclaves were removed from the oven and cooled to room temperature, a sample of the reaction mixture was taken from the reaction vessel.The solids were filtered thoroughly, washed with distilled water and 36233624 Zeolite Synthesis Table 1. Reaction details for syntheses performed in the Si0,-Al20,-Na2O-pyridine-H,O systema reaction no. SiO,/Al,O, pyridine/SiO, H,O/SiO, time/h productsb SiO,/Al,O, 1 2 3 4 5 6c 7d 8 9 10 16.3 16.3 16.3 29.2 58.6 54.6 58.6 58.6 73.3 97.7 0.99 0.67 0.00 0.67 0.67 0.67 0.67 0.00 0.67 0.67 47.3 39.8 39.8 25.1 39.8 24.8 47.0 27.3 39.8 39.8 112 166 216 216 194 256 165 165 166 194 MOR MOR MOR FER FER, trace Q FER MORD, Q FER, Q ZSM-5, Q FER, ZSM-5 magadiite 19.8 13.8 12.2 30.1 28.6 38.6 29.9 a OH/SiO, = 0.05, Na,O/SiO, = 0.3. FER: ferrierite, MOR: mordenite, Q : a-quartz. OH/SiO, = 0.44.Sodium silicate source 27.6 wt YO SO,, 8.63 wt YO Na,O, 63.77 wt YO H,O. F F FZ 0.0 LM-Z- 0 20 40 60 80 100 120 Si@/&Q Fig. 1. Crystallisation ranges of zeolites produced in the pyridine system. R: pyridine, M: mordenite, F: ferrierite, Z: ZSM-5, Na,O/SiO, = 0.3, OH/SiO, = 0.05. dried at 120 “C. The products were analysed by X-ray powder diffraction (Philips diffractometer, Cu Ka radiation) and scanning electron microscopy (Philips scanning electron microscope model 505). Results and Discussion The most important reaction variables in the synthesis of high-silica zeolites are the ratios, SiO,/Al,O,, OH/SiO, and ‘organic’/SiO,.l. 7 , 8 To investigate the different crystalline phases produced in the pyridine system the SiO,/Al,O,, pyridine/SiO, and OH/SiO, reaction ratios were varied systematically.The reactions performed and the crystalline phases produced are shown in table 1. Fig. 1 shows the crystallisation fields of the different phases with respect to the SiO,/Al,O, and pyridine/SiO, reaction ratios. At present three zeolites have been isolated within this system, mordenite, ferrierite and ZSM-5, along with silica phases such as a-quartz and magadiite [NaSi,O,,(OH), . H,O]. At relatively low SiO,/Al,O, values [SiO,/Al,O, = 16.3; reactions (1)-(3), table 11 mordenite was the only crystalline phase formed. The presence of pyridine in theJ . Chem. SOC., Faraday Trans. I , 1989, Vol. 85, part 1 1 Plate 1 Plate 1. Scanning electron micrographs of the mordenite samples revealing the effect of the pyridine/SiO, ratio on crystal morphology; (a) pyridine/SiO, = 0.0, (b) pyridine/SiO, = 0.67, ( c ) pyridine/SiO, = 0.99.W. J. Smith. J. Dewing and J. Dwyer (Fucing p . 3624)J . Chem. SOC., Faraday Trans. I , 1989, Vol. 85, part 11 Plate 2 Plate 2. Scanning electron micrographs of the ferrierite samples produced in the pyridine system; (a) reaction (4), (b) reaction ( 5 ) , ( c ) reaction (6). W. J. Smith, J. Dewing and J. DwyerJ . Chem. SOC., Faraday Trans. I , 1989, Vol. 85, part 1 1 Plate 3 Plate 3. Scanning electron micrograph of the ZSM-5 sample produced from reaction (8). W. J. Smith, J. Dewing and J. DwyerW. J. Smith, J. Dewing and J. Dwyer 3625 10 1s 20 2 5 30 Fig. 2. X-Ray powder diffractogram of the ferrierite sample produced from reaction (4).2810 aluminium-rich reaction mixture does not influence the type of crystalline phase formed; mordenite is produced even in the absence of pyridine [reaction (3)], although the reaction time required for maximum mordenite crystallinity decreases with increasing pyridine/SiO, ratio (table 1). The unit-cell composition of the mordenite sample produced from reaction (1) (pyridine/SiO = 0.99) is [43.6 SO,, 4.4 AlO,] 3.1 Na,O, 0.3 pyridine. Very little pyridine is present in the sample after washing and drying. Somewhat less pyridine is retained in the mordenite structure than might be expected for other organic materials. Clearly the relatively small pyridine molecules are not acting as void-filling/structure-stabilising organic species within the 12-ring large pore mordenite structure and the framework anionic charge is more efficiently balanced by sodium cations than by protonated pyridine molecules.This situation is not the case for the more siliceous medium-pore ferrierite structure (discussed later). If the composition of this mordenite sample (SiO,/Al,O, = 19.8) is compared with that produced in the pyridine-free reaction [reaction (3), SiO,/Al,O, = 12.21 it is apparent that more siliceous crystals are produced in the presence of pyridine. The effects of the pyridine/SiO, ratio on crystal morphology can be seen in plate 1. All the mordenite crystals consist of needle-like crystallites aggregated together and, as the pyridine content in the reaction mixture increases, these aggregates tend to increase in size from 15 pm [reaction (3)] to 50 pm [reaction (l)].A bimodal distribution of aggregates is apparent in the pyridine-free reaction with the occurrence of regularly shaped aggregates 4pm in size as well as the larger aggregates [plate 1 (a)]. The actual size of the needle-like crystals appears to go through an optimum value with increasing pyridine content in the reaction mixture from 3-6 pm [pyridine/SiO, = 0.0, plate 1 (a)] to 20 pm [pyridine/SiO, = 0.67, plate 1 (b)] to 5 pm [pyridine/SiO, = 1 .O, plate 1 (c)]. It is clear from these results that pyridine does not act as a structure-directing agent in the synthesis of mordenite. However, the relationship between the composition of the final product and the pyridine content of the reaction mixture suggests that the pyridine is influencing the crystallisation in a manner additional to its capacity as a base.At this stage it is difficult to determine the exact nature of the influence of pyridine but reaction parameters such as the surface energy of the system, the dispersion of the initial amorphous gel and the solvation of the aluminosilicate precursors in solution will be affected by the presence of pyridine, and all of these are likely to influence the crystallisation of mordenite. Increasing the SiO,/Al,O, ratio in the gel from 16.3 to 29.2 [reaction (4)] results in the crystallisation of a ferrierite phase. A typical X-ray diffractogram of the ferrierite phase is shown in fig. 2. The X-ray diffractogram is somewhat diterent from that of natural ferrierite;9' lo in particular, the X-ray diffraction line at 1 1.3 A (28 = 7.8") is significantly3626 Zeolite Synthesis Table 2.Unit-cell data on ferrierite samples synthesised in the pyridine system reaction no. unit-cell composition 4 5 6 [33.8 Si, 2.2 All 0.9 Na,O, 2.6 pyridine [33.6 Si, 2.4 All 0.4 Na,O, 2.1 pyridine [34.2 Si, 1.8 All 1.6 Na,O, 2.9 pyridine weaker for the ferrierite samples synthesised in this work. These differences in the X-ray diffractograms of natural ferrierite and the ferrierite-type materials synthesised using organic additives have been reported previously,l'T l2 and attributed to differences in cation content.'l The ferrierite phase is formed over the range SiO,/Al,O, = 29.2 [reaction (4)] to 73.3 [reaction (9)) This SiO,/A1,0, range for ferrierite crystallisation in the pyridine system is wider than those reported for other organic systems.'* l3 At higher SiO,/Al,O, values ZSM-5 co-crystallised with the ferrierite phase [SiO,/Al,O, = 97.7, reaction (lo)].Although the ferrierite phase can be synthesised from gels with relatively wide SiO,/A1,0, compositions, the composition of the final phase is always similar (SiO,/Al,O, = 30) and the product yield decreases with increasing gel SiO,/Al,O, ratio. The ferrierite phases produced in the pyridine system are much more siliceous than those produced in inorganic reaction systems and are similar in composition to samples produced using a variety of organic additives.13-15 The existence of a maximal SiO,/ Al,O, composition for the as-synthesised ferrierite phase [reaction (6), SiO,/A1,0, = 38.61 and the co-crystallisation of ZSM-5 from more siliceous gels [reaction (lo)] suggests that a minimum amount of framework aluminium is required in the ferrierite precursors for viable nucleation and crystal growth.As outlined by Jacobs and Martens,12 an upper limiting composition having SiO,/Al,O, = 34 for the ferrierite structure represents one aluminium atom per six-ring. The present results seem to be in good agreement with this proposal, the most siliceous ferrierite sample synthesised in the pyridine system having a composition of SiO,/Al,O, = 38.6 (see tables 1 and 2). The small discrepancy in the value of the limiting compositions reported previously12 and those observed here may arise from small amounts of siliceous impurities in the present sample.Although these results reveal a limiting composition for the ferrierite samples produced in the pyridine system, recent papers by Gies and Gunawardane'' and Gunawardane et a1.l' report the synthesis of completely siliceous ferrierite samples using a combination of 1,2-diaminoethane and boric acid as the template. The unit-cell compositions of the ferrierite phases are shown in table 2. In contrast to the mordenite samples synthesised at lower SiO,/A1,0, values, the ferrierite structures contain substantial amounts of occluded organic suggesting that pyridine is an efficient pore-filling material for the 1 O-ring medium pore ferrierite structure. The pyridine molecules are not degraded during crystallisation and are occluded intact into the ferrierite channels.The discrepancy between the charge-balancing Na+ and the framework A10, for the ferrierite samples synthesised from reactions (4) and (5) indicates that some of the cation exchange sites are occupied by pyridine molecules, presumably in the protonated form. This is confirmed by preliminary F.t.i.r. studies'* which reveal bands at 1437, 1483 and 1541 cm-' in the infrared spectra of as-synthesised ferrierite samples. The first two bands indicate that the pyridine molecules are intact whilst the latter band indicates that a portion of the pyridine molecules is in the protonated Plate 2 shows the crystal morphology of the ferrierite phases produced from reactions (4), ( 5 ) and (6). All the ferrierite crystals are of a plate-like form which is an unusual morphology for synthetic ferrierite-type zeolites which tend to consist of fibrous needles.13 Plate-like ferrierite crystals are also reported by RollmanZ0W.J . Smith, J . Dewing and J . Dwyer 3627 and Gies and Gunawardanel‘ Plate 2(b) and ( c ) shows the effect of decreasing the water content of the gel, from H,O/SiO, = 39.8 [reaction (5)] to 24.8 [reaction (6)], on the crystal morphology of the ferrierite phase. The plate-like crystals become less regular with the ‘plates’ orientated in various directions. At this stage no sorption experiments have been performed on the mordenite and ferrierite samples synthesised in this work. However, it is known that ferrierite-type materials synthesised using organic additives have increased sorption capacities for certain hydrocarbons as compared to natural ferrierite.l4 These differences in sorption capacities presumably arise from differences in the degree of faulting in the ferrierite structures. If the pyridine is omitted from the initial gel, of composition SiO,/Al,O, = 58.6 [reaction (S)], then the final product consists of ZSM-5 and a small amount of a-quartz. Plate 3 shows the crystal morphology of the ZSM-5 phase produced, the regular-shaped crystals ranging up to 40 pm in size. The ‘ bar-bell’ morphology is similar to that of other ZSM-5 phases synthesised from solely inorganic systems. 21 If the OH/SiO, ratio of the initial gel is increased from OH/SiO, = 0.05 [reaction (6)] to 0.44 [reaction (7)] then rriordenite and a-quartz are formed instead of ferrierite. The crystallisation of mordenite at higher OH/SiO, values of the reaction mixture is a common feature in high-silica zeolite synthesis.22.23 At high OH/SiO, values a relatively large proportion of the silica will remain in solution during the reaction and the formation of aluminous precursors suitable for mordenite nucleation will be preferred. It is clear from these results that the pyridine system favours the formation of zeolite structures containing five-ring units. The product series mordenite -, ferrierite -, ZSM-5 with increasing gel SiO,/A1,0, has been observed in solely inorganic albeit with SiO,/Al,O, crystallisation ranges different from those shown in fig. 1, and it is apparent that pyridine molecules are not essential in the generation of zeolite precursors containing five-membered rings.Such species are thought to be generated via interactions with sodium cations21 and the final structure formed is dependent on the aluminium content of the reaction mixture. The presence of pyridine extends the SiO,/Al,O, crystallisation range of the ferrierite phase, up to gel compositions of SiO,/Al,O, = 73.3, via occlusion of pyridine into the growing crystal structure both as a charge- balancing cation and as an agent for structure stabilisation by pore occupancy.25 At values of SiO,/Al,O, greater than 73.3 there is insufficient aluminium to generate a ferrierite structure containing one aluminium per six-ring and associated counter- balancing organic species. Under these circumstances the more stable siliceous ZSM- 5 is formed as a co-crystallising phase.The presence of silica phases such as magadiite and the transformation of the ZSM-5 phase to a-quartz at these high SiO,/Al,O, values suggests that pyridine is not an efficient pore-filler for the ZSM-5 structure. Conclusions Siliceous ferrierite samples were synthesised from gels containing pyridine. Although ferrierite samples were produced from gels of varying SiO,/Al,O, composition the final product appears to have a limiting composition of SiO,/Al,O, = 38.6. Mordenite and ZSM-5 were co-crystallising phases in the pyridine system; in both these cases the pyridine is not thought to be acting as an efficient pore-filling agent. We thank Shell Research B.V., Postbus 3003,1003AA Amsterdam for financial support of W.J.S. and in particular W. H. J. Stork and D. M. Clark for assistance during the course of this work.3628 Zeolite Synthesis References 1 B. M. Lok, T. R. Cannan and C. A. Messina, Zeolites, 1983, 3, 283. 2 E. Moretti, S. Contessa and M. Padovan, La Chimeca. el Industria, 1985, v67, gen, Feb. 21. 3 H. Robson, Chemtech, 1978, March, 176. 4 L. D. Rollman, ‘Zeolites: Science and Technology’, NATO Advanced Study Institute on Zeolites, Portugal, 1983, ed. F. R. Ribeiro (Martinus Nijhoff, Kluwer Academic, The Netherlands, 1984), p. 109. 5 W. A. van Erp, J. A. van der Griend, I. E. Maxwell and J. M. Nanne, Eur. Patent 01001 15, 1983. 6 N. Morimoto, K. Takatsu and M. Sugimoto, Eur. Patent 0121730, 1984. 7 A. Araya and B. M. Lowe, J. Chem. Res. (S), 1985, 192. 8 L. D. Rollman, Eur. Patenf 0021675, 1981. 9 L. W. Staples, Am. Mineral., 1955, 40, 105. 10 R. M. Barrer and D. J. Marshall, Am. Mineral., 1965, 50, 484. 11 G. T. Kokotailo, J. L. Schlenker, F. G. Dwyer and E. W. Valyocsik, Zeolites, 1985, 5, 349. 12 P. A. Jacobs and J. A. Martens, Stud. Surf. Sci. Catal., 1987, no. 33, 217. 13 K. Suzuki, Y. Kiyommi, S. Shin, K. Fujisawa, H. Watanabe, K. Saito and K. Noguchi, Zeolites, 1986, 14 C. J. Plank, E. J. Rosinski and M. K. Rubin, U.S. Patent 4046859, 1977. 15 L. D. Rollman, U.S. Patent 4296083, 1981. 16 H. Gies and R. P. Gunawardane, Zeolites, 1987, 7 , 443. 17 R. P. Gunawardane, H. Gies and B. Marler, Zeolites, 1988, 8, 127. 18 W. J. Smith, unpublished data. 19 J. W. Ward in Catalysis by Zeolites, ed. J. A. Rabo, ACS. Monograph, 1977, 171, 118. 20 L. D. Rollman, US. Parent 4107195, 1978. 21 A. Nastro, C. Collelo and R. Aiello, Stud. Surf. Sci. Catal., ed. B. Dnaj, S. Hocevar and S. Pejovnik, 22 E. W. Valyocsik and L. D. Rollman, Zeolites, 1985, 5, 123. 23 V. N. Romannikov, V. M. Mastikhin, S. Hocevar and B. Drzaj, Zeolites, 1983, 3, 31 1. 24 G. Bellussi, G. Perago, A. Carati, U. Cornaro and V. Fattore, Stud. SurJ Sci. Catal., ed. P. J.. Grobert, 25 A. Araya and B. M. Lowe, Zeolites, 1986, 6, 11 1. 6, 290. 1985, 24, 39. W. J. Mortier, E. F. Vansant and G. Schulz-Ekloff, 1988, 37, 37. Paper 8/04726J ; Received 29th November, 1988
ISSN:0300-9599
DOI:10.1039/F19898503623
出版商:RSC
年代:1989
数据来源: RSC
|
7. |
Influence of potassium on the catalytic properties of V2O5/TiO2catalysts for toluene oxidation |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 85,
Issue 11,
1989,
Page 3629-3644
Jingmin Zhu,
Preview
|
PDF (1213KB)
|
|
摘要:
J . Chem. SOC., Faraday Trans. 1, 1989, 85( I I), 3629-3644 Influence of Potassium on the Catalytic Properties of V,O,/TiO, Catalysts for Toluene Oxidation Jingmin Zhu? and S. Lars T. Andemson* Department of Chemical Technology, Chemical Center, Lund Institute of Technology, P.O. Box 124, S-22100 Lund, Sweden Monolayer-type vanadium catalysts, 2 wt YO vanadium on TiO, (Degussa P25), were prepared with potassium additions ranging from a potass- ium : vanadium atom ratio of 0.125 up to 10.1. The catalytic performance in the oxidation of toluene and of mixed oxidation products was investigated and correlated with features shown by X.r.d., ESCA and i.r. studies. Spectroscopic and activity results suggest predominant formation of monolayer vanadium with adsorbed potassium species at low K/V ratios.ESCA data indicate a very high dispersion of the major part of the vanadium in all samples. At K/V 1 small quantities of KVO, were detected, and at K/V = 10.1 the additional presence of external crystalline potassium compounds were found. Potassium in excess of the amount bonded to vanadium results in a reduced surface area and an increased anatase to rutile transformation. The activity for toluene oxidation decreases rapidly and linearly with potassium addition. It is < 2 % of the original rate at a K/V atom ratio of 0.4. At higher K/V ratios the further decrease is low. The activity changes may be explained by an increasing coverage of adsorbed potassium species, exerting both a geometric and an electrostatic effect. Selectivities also show great changes, but in a more complex manner.With increasing K/V ratio oxidative coupling products, and all acidic products, disappear whereas carbon oxide formation increases. The selectivity for benzaldehyde is much less affected. The selectivity changes may be explained by altered acid-base properties affecting the oxidation of products. For the formation of phthalic anhydride and anthraquinone different mechanisms involving base or acid catalysed oxidation of benzoylbenzoic acid, a proposed intermediate, are suggested. Many investigations have been devoted to the characterization of the vanadia-titania catalysts.lq2 During the last years much work has been done on the structure and the chemical and physical properties of vanadium ' monolayer ' catalysts, because of their excellent catalytic performance in oxidation of hydrocarbons.37 ' For V,O,/TiO, catalysts one observes a clear support enhancement of the active and some discussion exists concerning the importance of the crystalline structure of TiO, in determining the catalytic behaviour of the supported phase. Generally, anatase has been reported to be the preferred ' but a better performance of V,O, supported on rutile has also been fo~nd.~.'O Recently, it was noted that at a low concentration of vanadium, as in the monolayer catalysts, contaminations on the support, such as phosphorus and potassium present in many commercial preparations of TiO,, may significantly influence the catalytic behaviour." It was also shown that these contaminants are mainly present in the TiO, surface, which would strongly affect the formation of ideal monolayer catalysts." Furthermore, it has been shown that such t Permanent address: East China Institute of Chemical Technology, 130 Meilong Road, Shanghai 201 107, People's Republic of China.36293630 Potassium Catalytic Properties for Toluene Oxidation impurities have a large negative influence on the catalytic properties of the vanadium oxide phase in oxidation of both toluene and o-xylene.'2*13 The effect of these contaminants is strongly dependent on the type of hydrocarbon oxidized. Toluene oxidation is more strongly influenced by these additions than o-xylene oxidation. l3 It is notable that both phosphorus and potassium are used as promoters in some vanadium-type catalysts, mainly for oxidation of other hydrocarbon^.'^ The role of both these additions are obscure considering the negative effects mentioned above.The present investigation was made to study the influence of potassium on the structure and catalytic properties of a well studied V,O,/TiO, ~ata1yst.l~ Toluene oxidation was chosen as a model reaction. The catalysts were prepared from pure starting materials without these impurities, and vanadium loading was fixed at 2 wt % V. This loading is enough to cover the TiO, surface and gives maximum toluene oxidation rate.15 Experiment a1 Catalyst Preparation The catalysts were prepared by adding aqueous solutions of NH,VO, (Merck) and K,CO, (B.D.H.) to titanium dioxide. The TiO, used was Degussa P-25 and is specified as having a surface area of 50 T 15 m2 g-l and consists predominantly of anatase.The surface area of the batch used in this work was measured to be 47.8 T0.5 m2 g-l. The contaminations are specified as < 0.2% SO,, < 0.3 % A1,0,, < 0.01 YO Fe,O, and < 0.3 O/O HCl. The samples were dried at 393 K in air and calcined at 773 K for 1 h in air. The vanadium concentration for all samples was 2 wt YO [defined as g(vanadium) g-'(titania) x 1001. The atomic ratios of potassium to vanadium, K/V, were 0.125, 0.25, 0.64, 1.0, 1.985, 3.0 and 10.1 for catalysts KV 0.125, KV 0.25, KV 0.64, KV 1.0, KV 2.0, KV 3.0 and KV 10.1, respectively. The catalyst without potassium addition is referred to as 2V/TiO,. The sample with potassium supported on TiO, is K/TiO, and contained 7.65 wt % K [defined as g(potassium) g-'(titania) x 1001.Surface Area Measurements The B.E.T. method was used for calculation of the surface area. The method for measuring the adsorption of N, at 77 K has been reported ear1ier.l' X.R.D. Measurements The catalysts were analysed by X-ray diffraction using a Philips X-ray diffraction instrument with a PW 13 10/01/01 generator and Cu Ka radiation. Quantitative analysis of the anatase-rutile mixtures was performed according to the method described by Spurr and Myers." S.E.M. Measurements Scanning electron microscopic investigations were performed with an ISI- 1 OOA instrument equipped with EDAX (energy dispersive analysis of X-rays). I.R. Measurements Infrared spectra were recorded on a Perkin-Elmer 297 spectrophotometer.The KBr disc method was used, 1-3 mg of sample being added to 200 mg KBr.J. Chem. SOC., Faraday Trans. 1, 1989, Vol. 85, part I 1 Plate 1 Plate 1. Electron micrographs of K-V/TiO, catalysts (magnification lo4 x ) . (a) TiO,; (6) 2% V/TiO,; (c) KV 10.1. J. Zhu and S. L. T. Andersson (Facing p . 363 1)J . Zhu and S. L. T. Andersson 363 1 ESCA Measurements ESCA measurements were performed on an AEI ES200B electron spectrometer equipped with an A1 anode (1486.7 eV). The full width at half maximum of the Au 4f; line was 1.8 eV. Charging effects were corrected for by referring to the 0 1s signal, being assigned the value 529.6 eV." For quantitative estimations, peak areas above a linear base line were measured with a planimeter and corrected for the various count-rate settings.The V 2p2 intensity was obtained by subtracting the estimated 0 1s K,3.4 intensity from the 'partly overlapping peaks of V 2p: and 0 1s K,3,4. The 0 1s Kz3,4 intensity was estimated from the 0 1s KZ1,, intensity by multiplying with the 0 1s KZ3,JO 1s K,,,, ratio measured for the pure support. Activity Measurements A conventional flow apparatus operated at atmospheric pressure was used as described elsewhere." The gas flow was 15-55 dm3 h-l. Toluene was introduced with saturators to give 1.2 vol YO toluene and 15.7 vol YO of oxygen balanced with argon. The temperature was varied between 300 and 400 "C. The catalyst bed, one layer or two layers according to the required purpose, contained 0.05-0.25 g of catalyst of 10Ck200 pm particle size and was operated isothermally.All catalysts were diluted with inert quartz to avoid adverse thermal effects. The on-line analytical methods have been described in detail elsewhere. 2o Special care was taken for collecting samples and choosing analysis conditions to avoid depositing some high boiling point products. The reactor was operated with conversions between 0.2-10 YO and reaction rates were directly calculated from these differential data. Results and Discussion Characterization of Catalysts The surface areas, shown in fig. 1, were calculated as m' g;it and m2 g;:o,. At low levels of additive constancy of the latter would indicate that the added phase does not contribute any additional surface area to that of the support itself. However, at low K/V ratios, < 1, there is a small increase in surface area, both per g,,, and per gTiO,, with increasing addition of K.Earlier it was observed that pure TiO, undergoes a larger decrease in surface area under the present preparation conditions than preparations with small additions of ~anadium.'~ For pure TiO, this is probably due to a small fraction of very fine TiO, particles, possibly amorphous, sintering during calcination. The data suggest that at low additions of K the surface species inhibit sintering by preventing contact between TiO, primary particles. At high K/V ratios, > 1, the surface area decreases considerably with increasing addition of K. This effect can be due to sintering of the support with K present in excess and to support pore blocking by the added compounds.The X.r.d. line widths, however, do not change very much. Investigations with electron microscopy (see plate 1) revealed that external phases are formed in sample KV 10.1. The large agglomerates of crystallites seen near the middle of plate 1 ( c ) were found by EDAX analysis to contain potassium and traces of vanadium. From X.r.d. data (vide infra) it was found that various crystalline potassium and vanadium compounds exist in the KV 10.1 catalyst. For this sample the linewidth of the anatase (101) reflection had decreased by 0.03" (28) indicating sintering to some extent, whereas the rutile (1 10) line was not affected. These results suggest that pore blocking by deposited potassium and vanadium compounds is largely responsible for the decrease in surface area at high K/V ratios.The X-ray powder diffraction patterns of the various catalysts confirm that the major constituents are anatase and rutile in all samples. In table 1, some additional lines3632 Potassium Catalytic Properties for Toluene Oxidation I I I I I 2.0 4.0 6.0 80 10 12 20' K/V (atom ratio) Fig. 1. Surface areas of K-V/TiO, catalysts as a function of K/V atomic ratio. (a) m2 g;:t; (b) m2 g,&; observed in samples KV 1 .O to KV 10.1 are shown. From these lines KVO, is identified and is seen to increase in concentration on going from KV 1 .O to KV 10.1. Many new lines appeared for catalyst KV 10.1. Owing to the low intensity and the overlap of lines from various compounds some lines are difficult to assign. However, the lines match with K,CO, 1 .5H20, KOH, KOH - H20 and K,O.Thus, in the KV 10.1 catalyst a large amount of the potassium is present in a phase separate from that of the vanadium and titanium oxides. The presence of K2C03 in KV 10.1, but not in the other samples, was expected from phase diagrams.22 The other results, however, show differences with that expected from the phase diagram for the K20-Ti0,-V205 system as seen in table 2.23 From X.r.d. data the weight percent of anatase in the titanium dioxide (A',) was calculated from the intensity ratio of the anatase (101) and rutile (110) peaks.,' The variation of the anatase content with the K/V ratio is shown in fig. 2. A slight increase in anatase content can be seen with spreading V on the Ti02(P25) in agreement with the previous report.15 The anatase content is almost constant with K additions up to K/V = 1.At higher K/V ratios a decrease in the wt YO of anatase starts. This change is probably accompanied by a sintering of primary particles, which partly explains the loss in surface area (vide supra). A comparison of KV 10.1 and K/TiO,, which was prepared without vanadium loading, shows that the changes in the anatase-rutile transition are similar. This clearly shows that a strong interaction between potassium and TiO, induces the anatase transition to the rutile form. Phase diagrams, see table 1, suggest the presence of K,Ti,O,,, so formation of surface potassium titanate is very likely. The fact that the Xa is unchanged up to K/V = 1 illustrates the very poor interaction between TiO, and K at lower potassium loading with vanadium in excess.The surface composition of the 10&200 pm catalyst particles was studied by ESCA. The V 2p3,, peak can clearly be seen in spite of the overlap with the Ka3,4 satellite of the 0 1s peak. Table 3 provides the estimated binding energies. The V 2p,,, binding energy estimated for K-V/TiO, catalysts is ca. 516.8 eV, indicating the predominance of the pentavalent state. The choice of 0 Is as a reference level is perhaps not so good in this system where overlayers of oxygen-containing potassium compounds are formed. TheseTable 1. X.r.d. reflections for 15" c 28 < 45' for the K-V/TiO, catalysts. (Lines for anatase and rutile are not included.) line intensity" assignmentsb -_ 20" d / A KV 1.0 KV 2.0 KV 3.0 KV 10.1 KVO, K,CO,' KOH KOHd K,O 17.1 5.18 22.7 3.92 28.6 3.12 29.7 3.01 30.79 2.90 31.5 2.84 32.3 2.77 32.45 2.76 32.65 2.74 33.44 2.60 39.8 2.27 44.1 2.05 46.07 1.97 48.97 1.86 4 6 1 1 3 4 6 15 - - - - - - - 2 3 9 18 16 18 14 20 6 20 5 8 3 4 e - X - 9 s! R x a Ratio ( x 100) to the anatase (101) intensity.From ref. (21). K,CO; 1.5H20. KOH -H,O. Increased intensity compared to the other rutile lines. w m w w3034 Potassium Catalytic Properties for Toluene Oxidation Table 2. Phase composition according to X.r.d. analysis and phase diagrams of the K,&TiO,-V,O, systema phase composition catalyst X.r.d. phase diagramb KV 0.125 TiO, TiO,, V,0,/K,V,02, = 8 KV 0.64 TiO, TiO,, K,V,O,,/KVO, = 7.5 KV 1.0 TiO,, KVO, TiO,, KVO, KV 2.0" TiO,, KVO, TiO,, K,V,O,, KVO,d KV 3.0 TiO,, KVO, TiO,, K,V,O,/K,Ti,O,, = 1 KV 10.1 TiO,, KVO,, TiO,, K,V,O,/K,Ti,O,, = 0.12, (K,CO,) KV 0.25 TiO, Ti027 K2V8O20.8 ('y) K,CO, 1.5H,O, KOH, K,O, KOH*H,O a Ref.(22), ref. (23). Ref. (21). " Exact composition was 1.985. Very low quantity. Table 3. Binding energies of 0 Is, Ti 2p3,,, V 2p,,, and K 2p for the K-V/TiO, catalysts binding energy/eV 2V/TiO, KV 0.125 KV 0.25 KV 0.64 KV 1.0 KV 2.0 KV 3.0 KV 10.1 0 0.125 0.25 0.64 1 .o 1.985 3 .O 10.1 458.4 516.5 458.6 516.5 458.5 516.8 458.6 516.8 529*6 458.8 516.8 458.7 516.8 458.7 517.0 458.6 516.8 58.1 - 57.9 292.7 58.3 292.8 58.2 292.9 58.0 293.0 58.1 293.0 58.3 292.9 58.2 293.0 60' I I I I I 0 2 4 6 8 10 12 K/V (atom ratio) Fig. 2. Anatase content {X, = [ A / ( A + R)] x 10,) of the titanium dioxide measured by X.r.d.as a function of the K/V ratio in K-V/TiO, catalysts. 0, TiO,; ., K/TiO,; a, K-V/TiO, catalysts.J . Zhu and S. L. T. Andersson 3635 " 0 2 4 6 8 10 12 K/V (atom ratio) Fig. 3. Intensity ratio I\. 2p:l,2/ITi 2p3,2 as a function of K/V atomic ratio in K-V/TiO, catalysts. would contribute to the 0 1s level and may result in a small shift with increasing K/V ratio. The difference between V 2p,,, and Ti 2p,,,, B.E. (Av-Ti), is 58.1 T0.2 eV, of the order of the error limit with no special trend. The intensity ratio, Iv 2p3,,/ITi 2p3,, as a function of the K/V ratio is shown in fig. 3. At additions up to K/V = 1, a larger decrease in the intensity ratio than at higher K/V ratios is observed. This is caused by an increased agglomeration producing a loss in exposed V for ESCA analysis with increased K/V ratio up to 1.0, where in fact KVO, was identified with X.r.d.These crystallites contain a large proportion of V not seen by ESCA. In the same interval an increase of ca. 4 m2 g-l in the surface area was obtained (see fig. I), which supports the agglomeration. The decrease in Iv2p31,/ITi2p,/2 is ca. 16% from K/V = 0 up to K/V = 1, which indicates that only a small fraction of the vanadium forms KVO,. The major part is present in a monolayer or a highly dispersed form, with associated potassium. From K/V = 1 up to K/V = 10.1, the decrease in the intensity ratio is only 5 YO, indicating a very small change in the vanadium dispersion at potassium additions larger than K/V = 1.0. The intensity ratios Of IK 2p/ITi 2p,,2 and IK 2p/Iv 2p3,2 are shown in fig.4. Both curves are similar due to the relatively small change in the Iv 2p3i2/ITi 2p3,2 ratio. The relative increase in the intensity ratios is largest at additions up to 0.25, indicating that low additions of K result in a very high dispersion, and that above K/V = 0.25 some agglomeration of K compounds occurs. X.r.d. data identified KVO, in samples of K/V = 1.0 and K/V = 2.0. When extrapolated the data for K/V = 0 and K/V = 0.25 shows that the IK2p/IL72p3,, ratio measured for sample KV 2.0 is ca. half of that expected. This large discrepancy does not correlate with the relatively small decrease in I , 2p3,2/ITi 2p3/, (see fig. 4), which indicates that the loss in the IK zp/ITi 2p,,z ratio is mainly due to agglomeration of potassium compounds in addition to a small amount of agglomeration of K-V-0 compounds.Thus, it seems very likely that a considerable proportion of monolayer vanadium is present even up to K/V = 2. The relative increase in IK2p/ITi2p, between K/V = 2 and 3 is higher than the initial increase. Thus, in this region the added potassium is strongly confined to the surface with a high dispersion. At K/V ratios higher than 3.0 there is a marked agglomeration of these compounds. X.r.d. and S.E.M. data indicated (vide supra) very large agglomerates of crystalline and amorphous K compounds in sample KV 10.1. The3636 Potassium Catalytic Properties for Toluene Oxidat ion Fig. 4. Intensity ratio IK 2p/ITi 2p, and IK 2p/Iv 2p3,2 as a function of K/V atomic ratio in K-V/Ti02 cata'ysts' @, IK 2p/'Ti 2pSl2; 0 7 IK 2p/IV 2p,/,' Iv 2p3/2/1Ti 2p,/, ratio is almost constant between K/V = 3.0 and 10.1, indicating that the K compounds cover underlying V compounds and titania with no preference for either of these.Fig. 5 shows i.r. spectra for some K-V/TiO, catalysts of various K/V ratios. All samples show the usual band at 1636 cm-l for coordinated water. With potassium additions this increases in intensity with an asymmetric broadening towards lower wavenumbers, probably due to a component at ca. 1570 cm-l which is joined by a band at ca. 1350 cm-l at a K/V ratio of 2.0. These additional bands correspond to the v, mode of carbon dioxide adsorbed on the potassium species, probably in a bidentate mode.24 With increasing potassium loading both bands increase in intensity and move towards higher wavenumbers to indicate the presence of K,CO,.1 SH,O at a K/V ratio of 10.1. The broad band above 3600 cm-l increases in a similar manner. In the region 900-1000 cm-' several bands evolve with increasing potassium loading. At K/V = 1 .O [fig. 5 (c)] a weak band at ca. 975 cm-l and a stronger band at ca. 930 cm-l appear. These cannot be distinguished at K/V = 0.64 and are assigned to vv,,(sym) of KV03.25 Both bands are more intense and sharper at K/V ratios of 2.0 and 3.0 [fig. 5 ( d ) and (e)], and a third weak band at 900 cm-l has appeared, also due to KVO,. In sample KV 10.1 slightly different bands have evolved at positions of ca. 1016, 990, 920(sh) and 840 cm-l. KVO, is present in this sample according to X.r.d.data, but these bands do not fit with this compound, which then must be due to some other potassium polyvanadate. The shift upwards in wavenumber indicates a higher vanadium content, and the band at 1016cm-' indicates V=O species. As a comparison, K,V,O,, gives bands26 at 1012, 994, 986, 945 and 883 cm-l. Activity and Selectivity Measurements Catalytic Behaviour in Toluene Oxidation Fig. 6 shows the toluene oxidation rate at 380 "C as a function of the K/V ratio. A sharp decrease in the rate of toluene oxidation is observed with increasing amount of potassium under these conditions. It is evident that potassium is not only inactive on its own for toluene oxidation but is also a poison to the active vanadium phase. TheJ . Zhu and S.L. T. Andersson 3637 I 1 I I I I I I I I I I I I I I I I 4000 3000 2000 1600 1200 800 wavenumber/cm- ' Fig. 5. 1.r. spectra for K-V/TiO, catalysts of various K/V ratios. (a) TiO,; (b) KV 0.25; (c) KV 1.0; (d) KV 2.0; (e) KV 3.0 and ( f ) KV 10.1. extrapolated lines suggest that the main decrease in activity is up to the K/V ratio of 0.4. For samples with K/V = 0.64 or lower no K-V-0 compounds could be detected by any of the techniques used (uide supra), and a very high dispersion of most of the vanadium was indicated. The breakpoint at K/V = 0.4 corresponds to 0.8 wt % V which is close to the monolayer coverage 1.1 wt % determined earlier.15 Thus, potassium seems to be very well dispersed. The selectivity patterns for the catalysts are shown in fig. 7. It has been shown previo~sly,~~, 28 that catalytic oxidation of toluene may yield products of side-chain oxidation and oxidative coupling.Examples of the former group include benzaldehyde (BA), benzoic acid (BzA), maleic anhydride (MA) and benzoquinone (BQ) and the latter group phthalic anhydride (PA), methyldiphenylmethane (MDPM), benzophenone (BP), diphenylethanone (DPE) and anthraquinone (AQ). An interesting difference in the product distribution of the K-V/TiO, catalyst system is found in the oxidative coupling products. Such products are much more pronounced over pure V,O, crystallites. The oxidative coupling products disappear gradually with increasing potassium loading. Thus, for KV 0.125 and KV 0.25 only PA is found and for KV 0.64 none. This may be caused by potassium deactivating the sites for oxidative coupling reactions or it may be3638 Potassium Catalytic Properties for Toluene Oxidation I 3 Fig.6. Rate of oxidation of toluene at 380 "C as a function of K/V atomic ratio. 0, V,O,; x , TiO,; 0, K-V/TiO, catalysts. 20 ::L 0 Fig. 7. Selectivities for the various products in oxidation of toluene over the K-V/TiO, catalysts. (a) 2V/TiO,; (b) KV 0.125; (c) KV 0.25; ( d ) KV 0.64; (e) KV 1.0; (f) KV 2.0. (MA) maleic anhydride ; (BQ) benzoquinone ; (BA) benzaldehyde ; (BzA) benzoic acid ; (PA) phthalic anhydride; (MDPH) methyl-diphenylmethane ; (BP) benzophenone ; (DPE) diphenylethanone ; (AQ) anthraquinone.J . Zhu and S. L. T. Andersson 3639 6 4 2 0 - 4 -2 2 & 4 I c - \ o l 2 0 4 2 0 12 8 4 0 8 % 4," O N 8 s 4 0 8 4 0 CE - v I 1 o n Fig.8. Reaction rates at 380 "C of toluene and of the formation of the various products in the two layer catalyst tests. (a) V,O,; (b) V,O, +TiO,; (c) V,O, + K/TiO,; ( d ) V,05 + KV 1 .O, (V,O, in first layer). Notation as in fig. 7 except : (o-MDPM) o-methyl-diphenylmethanone ; (p-MDPM) p- methyl-diphenylmethanone and (DPED) diphenylethanedione. that the combustion of these products or their intermediates is facilitated. The latter is more likely considering the decrease in selectivity for benzoic acid with increasing potassium loading. The selectivity for benzaldehyde, on the other hand, shows only a relatively small decrease with increasing K/V ratio, except for higher ratios such as K/V = 2.0. These results show that potassium not only poisons the active sites for toluene oxidation, but also produces new reactive sites for the oxidation of benzoic acid and other intermediates.Catalytic Behaviour in Oxidation of Mixed Products The effects of potassium on the oxidation of various products were investigated by using layers of different catalysts to form two catalyst beds in the reactor. In the first bed 0.7 m2 of V,O, was placed to give a stable production of relatively large quantities of the oxidation products [see fig. 8(a)]. The second layer was 2.3 m2 of KV 1 .O, K/TiO, or TiO,. The reaction rates at 380 "C are shown in fig. 8. The change in the rate of toluene oxidation obtained with the second layer present is rather small since KV 1.0, K/TiO, and TiO, are much less active per m2 for toluene oxidation than V,O,.The differences observed are in the overall rates of product formation. With TiO, as the second layer, [see fig. 8(b)], not much effect on the product I20 FAR I3640 Potassium Catalytic o m / CHO Properties for Toluene Oxidation &I-& Fig. 9. Reaction network for formation of anthraquinone in toluene oxidation [from ref. (27)]. distribution was obtained. This result shows that TiO,, having undergone the same preparation procedure as the other catalysts, is relatively inactive for the oxidation of the products as well as for the oxidation of toluene. In the case of K/TiO, as the second layer [see fig. 8(c)], the greatest difference is that maleic anhydride disappears completely and an increase of ca. 35% in the rate of benzoic acid formation is obtained.At the same time an increase of ca. 24% in the CO formation rate, 50 % in the CO, rate and a 30 YO decrease in the phthalic anhydride rate is obtained, [cf. fig. 8(a) and (b)]. The rates for the formation of the other products are almost unchanged. This result clearly shows that the potassium species in the K/TiO, catalyst induces activity for the oxidation of MA and PA to CO and CO,, but has no such effect for the oxidation of benzoic acid to CO and CO,. The CO,/CO ratio also increases. The effect of potassium, in the K-V/TiO, system, on the oxidation of benzoic acid is remarkable [see fig. 8(d)], with a 90% decrease compared with V,O, alone [fig. 8(a)]. Note that the rates for all products produced from the oxidative coupling route are strongly suppressed, except for anthraquinone.In this relatively simple product spectrum MA, benzophenone (BP and diphenylethanedione (DPED) have disappeared and PA has been reduced by a factor of 100. The KV 1.0 sample is evidently a very efficient catalyst for oxidation of the products of toluene oxidation. Only small amounts of potassium in the K-V/TiO, catalysts were required to affect the activity strongly and the characterization of the catalysts suggested a very high dispersion of vanadium at low potassium loadings. Based on this it is tentatively suggested that at lower potassium loadings, potassium species are placed on or in the vanadium oxide monolayer. The two-layer catalyst test showed that potassium on titania induces activity for complete oxidation of MA and PA.For potassium on the 2V/TiO, catalyst our model for the surface composition implies that potassium on the vanadia monolayer induces activity for complete oxidation of the acids and several of the oxidative coupling products. Considering correlations between activity and acid-base proper tie^,^' it is suggested that potassium, increasing the basicity of the surface, enhances the oxidation of acidic organic compounds by facilitating their adsorption or preventing their desorption from the surface. It will, however, have a much smaller effect on compounds that are neither basic nor acidic, which is supported by the very small effects on benzaldehyde. It is interesting to note that anthraquinone was not affected in any of the three two- layer tests, but for the K-V/TiO, catalysts anthraquinone disappears at a K/V ratio of 0.125 and in addition phthalic anhydride disappears first at a K/V ratio of 0.64 (see fig.7). It is well known that catalytic oxidation of anthraquinone produces phthalicJ . Zhu and S . L. T. Andersson 364 1 K '\A -- I- - Fig. 10. Reaction mechanism for base-catalysed reaction of o-benzoylbenzoic acid to phthalic anhydride. (Illustrated with single tetrahedral vanadium ~pecies.)'~ - -H+ Fig. 11. Reaction mechanism for acid-catalysed reaction of o-benzoylbenzoic acid to anthra- quinone. (Illustrated with single tetrahedral vanadium species.)15 anh~dride,~' but our results strongly indicate that in toluene oxidation phthalic anhydride is not produced via anthraquinone alone. In a previous a fairly detailed reaction network for the formation of phthalic anhydride in four semi-parallel routes was suggested (see fig.9). Among these only the fourth route is desirable if anthraquinone is to be avoided. It is tentatively suggested (see fig. 10) that at first adsorption of o-benzoylbenzoic acid on the potassium sites occurs, followed by deprotonation and rearrangement. Then one carboxyl oxygen reacts via nucleophilic attack on the keto group, splitting off one phenyl species and forming phthalic anhydride. The phenyl species reacts further to benzoquinone, maleic anhydride, CO and CO,. We can also infer that most of the anthraquinone produced in the oxidation of toluene over 2V/TiO, follows the third route as detailed in fig. 11. In this route Bronsted acidic V-OH groups present in the 2V/Ti02 ~atalyst'~ catalyse the formation of anthraquinone from o-benzoylbenzoic acid by protonation of the carboxylic OH group. Water leaves and the remaining group reacts by an electrophilic attack on the ring.Thus, over a basic catalyst phthalic anhydride is formed directly from o- benzoylbenzoic acid in a base-catalysed step, whereas over a Bronsted acidic catalyst anthraquinone is formed instead in an acid-catalysed step. The Temperature Dependence of the Activity The effect of temperature on the activities for toluene oxidation over various catalysts is shown in fig. 12. The decrease in apparent activation energy (measured between 28MOO "C) with increasing potassium loading is in agreement with the results reported by van Hengstum et a1.13 Note that the crossing points between the first three catalysts are at reasonable temperatures.The isokinetic temperature for 2V/TiO, compared with 120-23642 Potassium Catalytic Properties for Toluene Oxidation 3.5 3.0 2.5 fl 2.0 -i 1.5 c - 0.5 -1.0 -1.5 1. lo3 K/T Fig. 12. Arrhenius plots for reaction rates in toluene oxidation. (a) 2V/TiO,; (b) KV 0.125; (c) KV 0.25; ( d ) KV 0.64; (e) KV 1.0. 30 25 20 - 15 k - T S en \ i : 10 5 n V 0 20 40 60 80 100 120 140 E~~~ /H mol-' Fig. 13. Logarithm of the apparent pre-exponential factor as a function of the apparent activation energy for the various K-V/TiO, catalysts. The error bars indicate the standard deviation in the measured data points. [The line obeys the equation In A = - 1.67( T 0.9) + 0.231( T 0.01 1) Eapp.] Notation as in fig.12.J . Zhu and S. L. T. Andersson 3643 the K-V/TiO, catalysts shows a linear correlation up to K/V = 0.64 where it is 248 “C. The extrapolated value for K/V = 0 is 317.5 “C. Thus, at low temperatures potassium addition results in an increased activity. Utilization of this effect is, however, limited since temperatures which are too low give very low rates and this also leads to an accumulation of high boiling point products. Fig. 13 shows the logarithm of the apparent pre-exponential factor as a function of the apparent activation energy, EaPP. A linear dependence, commonly called the compensation effect, is obtained. Since the data was calculated from the reaction rates, it is possible that other terms in the rate equation, except for the rate constant, are temperature dependent.These are not accessible since the detailed mechanism is unknown. Alkali and alkaline earth metals at very low concentrations on iron have long been known as electron More recent studies have shown that for metal catalysts potassium may increase the rate of dissociative adsorption of molecules such as C03’ and N233 on metals and 0,34 on carbon. For the same reason, the rate of reoxidation of the reduced vanadium cations should increase with potassium addition and the increased supply of dissociated oxygen species explains the increased production of carbon oxides with increased K/V ratio. It also agrees with the less pronounced oxygen partial pressure dependence of the toluene oxidation rate with KV 0.125 than with 2V/TiO,. Potassium may act by injection of electron density into the valence band, thereby changing the Fermi l e ~ e l , ~ * - ~ ’ which is a non-local effect.Potassium would also give rise to a cocsiderable increase in the local electrostatic potential over a distance of at least ca. 3.7 A from the potassium In consequence, each potassium ion will stroFgly affect at least two vanadium sites. If we consider a close packing of circles of 3.7 A radius, each m2 of the surface will be covered by ca. 0.4 x 10”. The packing density in a close packed vanadium monolayer is ca. 1 x lo1’ VO, units per m2.15 This gives a K/V surface ratio at complete effective coverage of ca. 0.4, in agreement with the break point in the activity data at K/V = 0.4.This result indicates that at low K/V ratios, potassium is adsorbed on the vanadium monolayer. Thus, the concentration of the surface vanadium species will decrease with increasing K/V ratio and thereby also the apparent pre-exponential factor, which is assumed to be proportional to the active site concentration. The effect on the apparent activation energy, on the other hand, must be due to the electronic interactions of potassium. It is possible that the rate determining step, which usually is assumed to be the first hydrogen abstraction, is preceded by an adsorption step. The effect of potassium is then to increase the adsorption strength of this intermediate, which effectively reduces the apparent activation energy, since in this case Eapp = Etrue + AHads.The increase in AHads required is then 1 10 kJ mol-’. It is tentatively suggested that the compensation effect arises as a result of potassium making some V sites more efficient, thereby decreasing EapP. Simultaneously, the potassium covers other V sites thus decreasing the pre-exponential factor. The Swedish Board for Technical Development and the National Energy Administration are acknowledged for financial support. References 1 M. S. Wainwright and N. R. Foster, C u d . Rev. Sci. Eng., 1979, 19, 21 I . 2 See for example various papers in; Cutul. Today, 1987, 1. 3 G. C. Bond and P. Konig, J . Curd., 1982, 77, 309. 4 M. Inomata, K. Mori, A. Miyamoto, T. Ui and Y . Murakami, J. Phys. Chem., 1983, 87, 754. 5 R. Kozlowski, R. F. Pettifer and J. M. Thomas, J. Phys.Chem., 1983, 87, 5172. 6 P. Cavalli, F. Cavani, I. Manenti, F. Trifiro and M. El-Sawi, Ind. Eng. Chern. Res., 1987, 26, 804. 7 K. Mori, M. Miura, A. Miyamoto and Y. Murakami, J . Phys. Chem., 1984, 88, 5232.3644 Potassium Catalytic Properties for Toluene Oxidat ion 8 M. Gasior, B. Grzybowska and M. Czerwenka, in Proc. Vth Int. Symp. Het. Catal., Varna, ed. D. Shopov, A. Andreev, A. Palazov and L. Petrov (Publ. House. Bulg. Acad. Sci., Sofia, 1983), part 1, p. 75. 9 W. E. Slinkard and P. B. de Groot, J . Catal., 1979, 68, 423. 10 G. C. Bond, A. J. Sarkany and G. D. Parfitt, J. Catal., 1979, 57, 476. 1 1 S. L. T. Andersson, J. Chem. SOC., Faraday Trans. I , 1986, 82, 1537. 12 A. V. van Hengstum, J. G. van Ommen, H. Bosch and P. J. Gellings, Appl. Catal., 1983, 8, 369. 13 A. V. van Hengstum, J. Pranger, J. G. van Ommen and P. J. Gellings, Appl. Cafal., 1984, 11, 317. 14 D. B. Dadyburjor, S. S. Jewur and E. Ruckenstein, Catal. Rev. Sci. Eng., 1979, 19, 293. 15 B. Jonson, B. Rebenstorf, R. Larsson and S. L. T. Anderson, J. Chem. SOC., Faraday Trans. I , 1988, 16 B. Jonson, B. Rebenstorf, R. Larsson and S. L. T. Andersson, J. Chem. SOC., Faraday Trans. I, 1988, 17 R. A. Spurr and H. Myers, Anal. Chem., 1957, 29, 760. 18 S. L. T. Anderson, J. Chem. SOC., Faraday Trans. 1, 1979, 75, 1356. 19 B. Jonson, B. Rebenstorf, R. Larsson, S. L. T. Andersson and S. T. Lundin, J. Chem. SOC., Faraday 20 S. L. T. Andersson, J. Chromatogr. Sci., 1985, 23, 17. 21 JCPDS, Powder Diffr. File, Inorg. Phases, Int. Centre Diffr. Data, Swarthmore, 1985. 22 F. Holtzberg, A. Riesman, M. Berry and M. Berkenblit, J. Am. Chem. SOC., 1956, 78, 1536. 23 A. A. Fotiev, A. P. Palkin, A. A. Soboleva and L. A. Perelyaeva, Russ. J. Inorg. Chem. (Engl. Transl.), 24 G. Busca and V. Lorenzelli, Mat. Chem., 1982, 7 , 89. 25 S. Onodera and Y. Ikegami, Inorg. Chem., 1980, 19, 615. 26 V. N. Krasil’nikov, M. P. Glazyrin, A. A. Ivakin, L. A. Perelyaeva and A. P. Palkin, Russ. J. Inorg. 27 S. L. T. Anderson, J. Catal., 1986, 98, 138. 28 J. E. Germain and R. Laugier, Bull. SOC. Chim. Fr., 1972, 541. 29 M. Ai, Bull. Chem. SOC. Jpn., 1976, 49, 1328. 30 E. I. Andreikov, Kinet. Katal., 1971, 12, 776. 31 S. Brunauer and P. H. Emmett, J . Am. Chem. SOC., 1940, 62, 1732. 32 C. T. Campbell and D. W. Goodman, SurJ Sci., 1982, 123, 413. 33 G. Ertl, M. Weiss and S. B. Lee, Chem. Phys. Left., 1979, 60, 391. 34 P. Sjovall, B. Hellsing, K-E. Keck and B. Kasemo, J. Vac. Sci. Technol. A, 1987, 5, 1065. 35 E. L. Garfunkel, J. J. Maj, J. C. Frost, M. H. Farias and G. A. Somorjai, J. Phys. Chem., 1983, 87, 36 J. K. Nerrskov and P. Stolze, SurJ Sci., 1987, 189/190, 91. 84, 3547. 84, 1897. Trans. 1, 1986, 82, 767. 1981, 26, 576. Chem. (Engl. Transl.), 1983, 28, 417. 3629. Paper 8/04786C; Received 5th December, 1988
ISSN:0300-9599
DOI:10.1039/F19898503629
出版商:RSC
年代:1989
数据来源: RSC
|
8. |
Influence of phosphorus on the catalytic properties of V2O5/TiO2catalysts for toluene oxidation |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 85,
Issue 11,
1989,
Page 3645-3662
Jingmin Zhu,
Preview
|
PDF (1129KB)
|
|
摘要:
J . Chem. SOC., Furaduy Trans. 1, 1989, 85(11), 3645-3662 Influence of Phosphorus on the Catalytic Properties of V,O,/TiO, Catalysts for Toluene Oxidation Jingmin Zhu,? Bernd Rebenstorfl and S. Lars T. Andersson" Department of Chemical Technology, Chemical Center, University of Lund, P.O. Box 124, S-22100 Lund, Sweden Monolayer-type vanadium catalysts, 2 wt % V on TiO, (Degussa P25), have been prepared with phosphorus additions of a P/V atom ratio of 0.25-5.0. The catalytic performance in the oxidation of toluene was investigated and correlated with features shown by X.r.d., ESCA and F.t.i.r. studies of adsorbed CO. Spectroscopic and activity results suggest that phosphorus species are highly dispersed on the monolayer vanadium structure at low P/V ratios. Above P/V = 1.25 agglomeration of P-V-0 phases occurs, and crystalline vanadium phosphates are detected at a P/V ratio of 5.0.The anatase/rutile ratio changes very little with increasing P/V ratio, Surface areas are constant, except for a decrease at a P/V ratio of 5.0, probably due to P-V-0 phases. ESCA data suggest the presence of V4+ in calcined catalysts at a P/V ratio higher than 1 .O, which is supported by the i.r. studies of adsorbed CO. Furthermore, the i.r. studies suggest the formation of V3+ in CO-reduced samples without phosphorus additions, whereas V4+ seems to be stabilized with increasing P/V ratio. Phosphorus shows a strong negative influence on the activity in toluene oxidation. The rate correlates linearly with the surface V/(V+P) atom ratio determined by ESCA.Apparent activation energies do not change, indicating that the negative effect of phosphorus on the activity is dominated by a steric effect. Selectivities also change with phosphorus addition. The selectivity for side- chain oxidation products decreases, and those for carbon oxides and oxidative coupling products increase with increasing P/V ratio. A model for surface P-V-0 species is presented and the change in selectivities are suggested to be caused by an increased life time of the initial intermediate, probably the benzylic radical. In a previous paper1 we reported on the influence of potassium on the catalytic properties of a monolayer- type catalyst [vanadium oxide supported on TiO, (Degussa P25)] in toluene oxidation. In this contribution the effect of phosphorus is presented.Vanadium-phosphorus oxide catalysts have found an important application in the oxidation of C , hydrocarbons, and much effort has been devoted to the understanding of the nature and structure of these The properties of these catalysts supported on various carriers, however, have attracted considerably less attention, especially for monolayer-type cataly~ts.~-~ A large influence on the catalytic properties of the vanadium-oxide phase supported on TiO, was found with small amounts of phosphorus and potassium. ' 3 Several commercial preparations of TiO, contain impurities, especially phosphorus, potassium and sulphur. It was found that phosphorus is strongly bonded to the surface of Ti0,8 and that it prevents the formation of ideal monolayer-type catalysts. A surface enrichment of the impurities was observed after calcination, which is mainly due to the decreased surface area and a strong confinement People's Republic of China.t Permanent address: East China Institute of Chemical Technology, 130 Meilong Road, Shanghai 201 107, $ Department of Inorganic Chemistry 1. 36453646 Influence of Phosphorus on V,O,/TiO, of impurities to the surface. It could be possible that preparations of different phases of TiO, from various laboratories give products varying in degree of contamination. This aspect is relevant to the debate on the effect of the crystalline form of TiO, as support for vanadium oxide. Results based on very pure Ti0,9310 suggest that the crystalline form of TiO, has no great influence on the activity and selectivity of these catalysts.The work presented here was performed to further our understanding of these effects. Catalysts containing phosphorus-vanadium oxide supported on TiO, were prepared from pure components and characterized by both catalytic and spectroscopic studies. Experiment a1 Catalyst Preparation The catalysts were prepared by adding aqueous solutions of NH,VO, (Merck P.A.) and H,PO, (Merck Krist.) to TiO, (Degussa P25). The samples were dried at 393 K in air and calcined at 773 K for 2 h in air. The vanadium concentration was for all catalysts kept at 2wtY0, defined as gV/gTiO,. Catalysts were prepared with an atom ratio of phosphorus to vanadium, P/V, of 0,0.25,0.5, 1 .O, 1.25, 2.0 and 5.0. These catalysts were named PV 0, PV 0.25, PV 0.5, PV 1.0, PV 1.25, PV 2.0 and PV 5.0, respectively.For comparison one sample containing 2.37 wt YO P, defined as gP/gTiO,, was prepared without vanadium and was named P/TiO,. Activity Measurements A conventional flow apparatus operated at atmospheric pressure was used as described elsewhere.*' Standard conditions used were : 1.2 vol % toluene and 15.7 vol YO oxygen in argon, 34.6 dm-3 h-l total flow rate and temperature 300-600 "C. The catalyst bed contained 0.05-0.25 g of catalyst of particle size 100-200pm and was operated isothermally. All catalysts were diluted with inert quartz to avoid adverse isothermal effects. The on-line analytical methods have been described in detail elsewhere.', The reactor was operated with conversions between 0.2-10 % and reaction rates were directly calculated from these differential data.Surface-area Measurements A gravimetric B.E.T. apparatus was used for measuring the adsorption of N, at 77 K. Prior to adsorption the samples were outgassed overnight at a temperature of 573 K and a pressure of 6 x lo-' Torr.? The B.E.T. method was used for calculation of the surface area. X.R.D. Measurements The catalysts were analysed with X-ray diffraction using a Philips X-ray diffraction instrument with a PW 1310/01/01 generator and Cu Kcc radiation. Quantitative analysis of the anatase-rutile mixtures was performed according to the method described by Spurr and Myers.13 ESCA Measurements ESCA measurements were performed on a AEI ES200B electron spectrometer equipped with an A1 anode (1486.6 eV).The full width at half maximum of the Au 4f5 7 1' me was 1.8 eV. Charging effects were corrected for by referring to the 0 1 s signal, being assigned the value 529.6 eV.14 For the quantitative estimations, peak areas above a linear baseline t 1 Torr z 101 325/760 Pa.J. Zhu, B. Rebenstorf and S. L. T. Andersson 3647 50 t 30 t l a ~ " ~ ' " " ' ' " " ' ~ ' ' ' ' ' 0 1 2 3 4 5 P/V atom ratio Fig. 1. Surface area of the P-V-O/TiO, catalysts as a function of the P/V atom ratio. (a) m2 g-' of catalyst; (b) m2 g-I of TiO,. x , TiO,; A, P/TiO, (m2 g-'); 0, P/Ti02 (m2 g&). were measured with a planimeter and corrected for the various count-rate settings. The V 2p; intensity was obtained by subtracting the estimated 0 1s Ka,,, intensity from the partly overlapping peaks of V 2p;+O 1s Ku,,,.The 0 1s Ka,,, intensity was estimated from the 0 1s Ku,,, intensity by multiplying with the 0 1s Ka,,,/O 1s Ka,,, ratio measured for the pure support. For calculation of P/(P + V) and V/(P + V) atom ratios sensitivity factors of V 2p; = 2.1714 and P 2p = 0.315 were used. I.R. Measurements Infrared measurements of adsorbed CO were performed with self-supporting discs placed in an i.r. cell, described previously." Spectra were taken (128 scans) on a Nicolet 20 SXC Fourier-transform i.r. spectrometer. The resolution was 2 cm-'. The samples were degassed at Torr and heated in oxygen at 773 K for 30 min (oxidized form) and reduced at 773 K with CO for 30 min (reduced form). 1.r. spectra were recorded under the conditions given in the text.Results and Discussion Characterization of the Catalysts Surface Areas The surface areas of the different catalysts with different P/V ratios are shown in fig. 1. The surface areas per gram of TiO, were calculated and compared with those observed per gram of sample. With increasing P/V ratio up to 2.0, there appears to be a small decrease in the former and a small increase in the latter. The effects, though very small may be attributed to formation of a small amount of external V-P-0 phases. A much bigger difference is found when comparing with the pure TiO, treated in the same manner as the V-P-O/TiO, catalysts. It is suggested that this discrepancy is due to the sintering of some very small particles, that are present in the original TiO,, which is inhibited by surface P and V species.An increase had already been observed at an addition of 0.1 wt % V as reported ear1ier.l' However, this inhibition of sintering is only3648 I A Influence of Phosphorus on V,O,/TiO, A R 1 Fig. 2. X-Ray powder diffractograms for some V-P-O/TiO, catalysts. (a) P/TiO,; (b) PV 5.0. A, anatase; R, rutile; B, B-VOPO,; V, (VO),P,O,. 75 * O 2 65 t I I I I 1 2 3 4 P/V atom ratio Fig. 3. Anatase content [X, = 100 x A/(A + R)] of the titanium dioxide measured by X.R.D. as a function of the P/V ratio in P-V-O/TiO, catalysts. 0, TiO,; A, P/TiO,; e, P-V-O/TiO, catalysts.J . Zhu, B. Rebenstorf and S. L. T. Andersson 3649 520 51 5 51 0 binding energyIeV 525 Fig. 4. Photoelectron spectra of the V 2p binding-energy region for the P-V-O/TiO, catalysts.The P/V ratios are: (a) 0; (b) 0.25; (c) 0.5; ( d ) 1.0; (e) 1.25; (f) 2.0; (g) 5.0. observed also with phosphorus addition, since the P/TiO, catalyst gives a similar surface area at low P/V ratios. When increasing the P/V ratio up to 5, however, a considerable decrease in area occurs. XRD All samples were investigated by XRD analysis and gave almost identical diffractograms with the pure TiO, support. These are represented by the diffractogram of P/TiO, showing lines for anatase and rutilel' (see fig. 2). Additional lines were only observ5d with sample PV 5.0 [see fig. 2(b)]. The lines at 28 = 15.82' and 28.51" (5.60 and 3.12 A) are assigned to (VO),P,O, and the lines at 28 = 17.05' and 29.58" (5.20 and 3.02 A) to j3-VOP0,.19 The presence of these crystalline materials, and additional amorphous phosphorus materials, on the surface of the catalysts suggests that the decrease in surface area for the PV 5.0 catalyst is at least partly caused by pore blocking.The amount of anatase in the support, Xa in wt %, was calculated from the intensities of the anatase (101) peak and the rutile (1 10) peak13 (see fig. 3). The amount of anatase increases with vanadium addition up to 2 wt % V (compare TiO, with PV 0), which has previously been assigned to isolated V species inhibiting the anatase-rutile trans- f0rmation.l' With the addition of phosphorus, no effect is seen at P/V = 0.25 due to the presence of isolated vanadium species. With increasing P/V ratio up to 1.0, a decrease is obtained since P-V-0 species are formed and fewer vanadium species are left.For the same reason, the amount of anatase is constant for P/V = 1.0-2.0. At the highest3650 Influence of Phosphorus on V,O,/TiO, Table 1. Binding energies' of Ti 2p3/,, V 2p3/, and P 2p for the P-V-O/TiO, catalysts binding energy/eV catalyst PV 0 0 458.4 516.5 - PV 0.25 0.25 458.4 516.6 132.8 PV 0.5 0.5 458.5 516.6 132.9 PV 1.0 1.0 458.6 516.6 132.9 PV 1.25 1.25 458.5 516.5 132.8 PV 2.0 2.0 458.4 516.4 132.9 PV 5.0 5.0 458.4 - 132.7 a Referenced to 0 1s = 529.6 eV.14 0.5 1 .o bulk P/(P + V) (atom%) "0 Fig. 5. Surface concentration of phosphorus and vanadium measured by ESCA as a function of the bulk composition of P-V-O/TiO, catalysts. (a) P/(P + V) ; (b) V/(P + V). phosphorus addition, however, crystalline compounds, p-VOPO, and (VO),P,O, as seen by XRD analysis (vide supra), are present and these are suggested as facilitating the transition as for V20,.20 When the latter is crystallised on the surface of anatase, a certain misfit and strain appear which induces a topotactic transition to rutile.,' The P/Ti02 catalyst shows only a marginal difference from the support itself, indicating an inhibiting effect on the anatase-rutile transformation in agreement with literature data.,, This is in contrast to potassium which accelerates this transition.' The data indicate a very high dispersion of the vanadium and phosphorus species on these catalysts.ESCA The surface composition of the 100-200 pm catalyst particles was determined by ESCA measurements. As seen in fig.4, the V 2p,,, can be clearly seen in spite of the overlap withJ. Zhu, B. Rebenstorf and S. L. T. Andersson 81 10.2 365 1 P/V atom ratio Fig. 6. The intensity ratios Ip2p/ITi2p3,2 and Ip2p/Iv2p3/2 as functions of the P/V ratio in P-V-O/TiO, catalysts. the 0 1s peak (at ca. 520 eV) originating from the Ku,,, satellite in the X-ray radiation. The V 2p3/, peak hardly changes in form, position and intensity up to a P/V ratio of 1.0. The binding energy (B.E.) values for V 2p,,,, Ti 2p,,, and P 2p are given in table 1. The binding energies indicates the predominant presence of pentavalent vanadium. l4 As seen in the spectra at P/V ratios of 1.25 and higher, the V 2p,,, peak is broadened towards lower B.E. indicating the presence of VIV. This seems to be quite apparent at P/V ratios of 2.0 and 5.0 and is in agreement with the XRD results with the presence of (VO),P,O,. The ESCA data were quantified and fig.5 shows the P/(P + V) and V/(P + V) surface atom ratios determined by ESCA, as a function of the bulk P/(P+V) atom ratio. The surface concentration of phosphorus is clearly much higher than the corresponding bulk concentration. This is most apparent at low phosphorus additions, and for PV 0.25 and PV 0.5 catalysts the surface concentration is twice as high as the bulk concentration. It is thought to depend on the bonding ‘affinity’ between P/V as compared to P/Ti and V/Ti interfaces. The P/V ratio measured by ESCA is in the following taken as an approximate value of the P/V ratio in the outermost surface layer. It is suggested that the rapid increase in the surface P/(P+V) ratio up to a P/V bulk ratio of 0.5 corresponds to saturation of the vanadium monolayer with surface phosphorus, and thereafter various V-P-0 compounds are formed as noted for P/V = 5.0 (vide supra).Assuming that each surface phosphorus species bonds to one vanadium, the surface P/(P + V) ratio for P/V = 0.5 corresponds to the same coverage as 1.1 wt % V, obtained for the monolayer coverage earlier.17 The intensity ratios Ip2p/Iv2p3,, and Ip2p/ITi2p312 are shown in fig. 6 as functions of the P/V ratio. Both curves increase linearly up to a P/V ratio of 1.25. The slope is different, however, mainly due to the difference in sensitivity factors forV 2p,,, and Ti 2p3,,. Between P/V ratios of 1.25 and 5.0 both curves also increase, but with slopes different from those at lower P/V ratios.The explanation for this is that at P/V ratios higher than 1.25, the supposed P-V-0 monolayer phase starts to agglomerate, giving a reduction in the amount of vanadium seen in the ESCA analysis and a larger increase in Ip2p/Iv2p312. At the same time, the mean thickness of the overlayers on the titania surface is decreased, resulting in a larger Ti 2p,,, contribution in the ESCA analysis, and the IPZP/ITiZpaln ratio decreases. It seems that an excess of phosphorus promotes the aggregation, which is in agreement with findings for alumina supported vanadyl phosphate catalyst^.^3652 F Influence of Phosphorus on V,O,/TiO, 30 Fig. 7. 1.r. spectra of CO (80 Torr) adsorbed at 293 K on the oxidized form of the catalysts.(a) TiO,; (b) PV 0; (c) PV 0.25. I 1 I I I I 2250 2200 2150 2100 2050 2( wavenumber/cm- ' 00 Fig. 8. 1.r. spectra of CO (80 Torr) adsorbed at 163 K on the oxidized form of the catalysts. (a) PV 0; (b) PV 0.25; (c) PV 0.5; ( d ) PV 1.0; (e) PV 1.25; (f) PV 2.0; (g) TiO,; (h) P/TiO,.J . Zhu, B. Rebenstorf and S. L. T. Andersson 3653 t ( g ) ( h ) / I I 1 I w avenumber/cm- ' 50 2200 2150 2100 2050 2000 Fig. 9. 1.r. spectra of CO (80 Torr) adsorbed at 293 K on the reduced form of the catalysts. Notation as in fig. 8. F. T. I . R. The catalysts were investigated by F.t.i.r. studies of CO adsorption at room temperature (293 K) and at low temperature (163 K) on oxidized and on reduced forms of the catalysts. Fig. 7-10 illustrate the spectra recorded under different conditions.On the oxidized form of the catalysts, adsorbed CO could be detected on TiO,, by a very strong band, and on PV 0 and PV 0.25, by much weaker bands (see fig. 7). The vc0 band found at 2184 cm-l identifies the presence of coordinatively unsaturated Ti ions Ti,,, on the surface of the oxidized form of Ti02.239 24 The very weak band at 2194 cm-l observed for PV 0 and PV 0.25 is assigned to exposed surfaces of the TiO, A comparison between the spectrum of TiO, and that of PV 0 indicates that the uncovered part of the support is rather small, less than 5 %. The shift of the band from 2184 to 2194 cm-' indicates the influence of the different surroundings. No adsorption of CO on the oxidised vanadium or phosphorus species can be detected.The results indicate that most of the TiO, surface is covered by these species already at low phosphorus loading. Low-temperature measurements of adsorbed CO are shown in fig. 8. Two main bands at ca. 2164 and 2141 cm-l were observed for all spectra, and assigned to CO adsorbed on surface hydroxyl groups and CO clusters, re~pective1y.l~ The former band also occurs at 2164 cm-' for P/TiO,, which should expose mainly P-OH groups. In the spectra for TiO,, at least three different OH sites, giving bands at 2164, 2156 and 2149 cm-l can be distinguished. The former is protiably due to adsorption on singly bonded OH groups. With increasing P/V ratio, the band at 2164 cm-l decreases in intensity. In earlier where the vanadium concentration was varied, a band at 2149cm-l was assigned to Ti-OH groups and a band at 2162 cm-l to V-OH.It should be noted that a third band at 2181 cm-l was observed in the spectrum of P/TiO, and as a shoulder in3654 Influence of Phosphorus on V20S/Ti02 I I 1 I wavenumber/cm- ' 2250 2200 2150 2100 2050 2C 30 Fig. 10. 1.r. spectra of CO (80 Tom) adsorbed at 163 K on the reduced form of the catalysts. Notation as in fig. 8. all spectra of the catalysts containing phosphorus. With decreasing P/V ratio from 2 to 0, it shifts position from 2181 to 2190 cm-'. This band is assigned to GO adsorbed on Ti,,, ions. Coordinatively unsaturated P species are expected at higher wavenumbers26 and VIrl surface species are not present in the oxidized form of the catalyst. The data indicate that the support surface is not fully covered in any of the catalysts.It should be noted that this band increases in intensity with decreasing P/V ratio, in accordance with a lower coverage at low P/V ratios. At higher P/V ratios than 1.0, traces of a band are seen at ca. 2200 cm-'. This band could be due to Ti,,, ions of a higher Lewis developed due to the influence of phosphorus, but it is more probably due to the presence of Vrv species, indicated by both XRD and ESCA in calcined catalysts of higher P/V ratios. Fig. 9 shows the vco bands that appear upon adsorption of CO at room temperature on the reduced form of the catalysts. One main band appears to be present and the position is shifted from 2183 cm-' in the spectrum of PV0.25 continuously with increasing P/V ratio up to 2204 cm-l for PV 2.0.In an earlier study'? of similar catalysts with different loading of vanadium, the reduced catalysts all showed a band at 2187 cm-', independent of loading, and the support itself gave a band at 2184 cm-l. Due to this effect it was difficult to assign the band to either V or Ti,,, sites. The new results suggest that it is mainly due to vanadium sites of different valences (vide infra) with a contribution from Ti sites. A weak band observed at 2145 cm-l for PV 1.0 could be due to adsorption on V" site~,~' because this band is already apparent at room temperature. CO adsorbed on surface hydroxyl groups or as surface clusters gives bands in this region only at low temperatures. A weak band at 2195 cm-' is observed in the spectrum ofJ.Zhu, B. Rebenstorf and S. L. T. Andersson 3655 1 2 P/V atom ratio 2 2 1 0 ~ ' " ' " ' I ' I ' Fig. 11. The CO vibration frequency, vco, of CO adsorbed on surface vanadium sites on the reduced form of the P-V-O/TiO, catalysts as a function of the P/V atom ratio. Adsorption temperature: (a) 163 K ; (b) 293 K. P/TiO,. It is at the same position as a very weak band observed for the oxidized form of PV 0 and PV 0.25, and is possibly due to a particular Lewis acidic Ti,,, site. The support itself gives rise to an intense band at 2185 cm-l due to Ti,,, ions of the same type as in the oxidized form, and in addition a weak band at 2206cm-', which must be ascribed to a different Ti,,, site of higher Lewis This band appears already after some degree of dehydroxylation.28 The 2185 cm-l band is due to CO adsorbed on sites formed at a high degree of dehydration.28 It is very sensitive to inductive effects and shifts downwards with decreasing dehydration. A band at 2183 cm-' appears at CO adsorption on H,-reduced Ti0,/Si0,,29 where Ti3+ cations are clearly formed.At higher degree of reduction, and consequently, higher CO coverage, the band shifts downwards to 2178 cm-l. Our shift between low- and room-temperature spectra is also ca. 5 cm-'. These data suggest that Ti3+ is already formed at low-temperature evacuation. The spectra obtained at low-temperature adsorption of CO on reduced samples are shown in fig. 10. The band at 2141 cm-l is due to CO clusters and the band at 2164 cm-l to CO adsorbed on surface OH groups. It is worth noting that the vco band for CO adsorbed on P/TiO, at 2186 cm-l becomes stronger after reduction of the catalysts and that it has shifted from 2181 cm-l for the oxidized sample.Simultaneously, the band for CO adsorbed on OH groups shifts to 2167 cm-l. These shifts are influenced by the overlap and different relative intensities of the two bands. These results indicate dehydroxylation with a decrease in OH groups and increase in Ti,,, species. The main peak in the spectra for P-V-0 samples shows a linear shift of over 24 cm-l from PV 0.25 with a band at 2177.8 cm-l to PV 2.0 with a band at 2202 cm-l. This shift and that for room-temperature spectra are shown in fig. 11 as a function of the P/V ratio of the different P-V-O/TiO, catalysts. The main cause for the shift of this band is attributed to the vanadium species with3656 InJluence of Phosphorus on V,O,/TiO, II I \ \ ( h ) I 1 I I I I I 10 3700 3600 3500 31 wavenumber/cm- ' 30 Fig.12. 1.r. spectra of OH bands at 293 K for oxidized and reduced P-V-O/TiO, catalysts. A, Catalysts in the reduced form; B, catalysts in the oxidized form; C, support and P/TiO, in oxidized and reduced forms. Notations (a)+) as in fig. 8. In addition; (g) oxidized TiO,; (h) reduced TiO,; (i) oxidized P/TiO,; ( I ] reduced P/TiO,. change of both their oxidation state and environment. The effect of phosphorus as a more electronegative element would be to increase the positive charge on neighbouring cations. Long-range inductive effects do not seem to be of importance since no shift at the lowest addition of phosphorus could be observed. In earlier studies30 we found that there is an almost linear correlation between vco and the valence state of vanadium supported on silica, with a shift of 28 cm-l between CO adsorbed on V3+ and on V4+, and 35 cm-' between V2+ and V3+.Bearing in mind earlier assignments2' of CO adsorbed on V3+ sites to bands at ca. 2180-2190 cm-', the band at 2202 cm-l in the spectrum of the reduced form of PV 2.0 should be assigned to V4+ sites. Moreover, phosphorus is considered to stabilize V4+ in square pyramidal coordination. 2, Spectra from the OH stretching region of both oxidized and reduced samples are shown in fig. 12. Note that the main OH band at 3654 cm-' for oxidized PV 0 shifts to 3672 cm-l for reduced PV 0. The difference in the shift between the oxidized and the reduced catalyst decreases gradually with increasing P/V ratio up to PV 2.0, where no shift is observed and both oxidized and reduced forms give an OH band at 3662 cm-'.The P-OH groups on TiO, give a band at 3665 cm-' [spectra (i) and (51, fig. 121. Since the treatment with CO mainly affects the vanadium species, the shift of the OH band is correlated to the oxidation state of the vanadium species. In fig. 13, the wavenumber ofJ . Zhu, B. Rebenstorf and S. L. T. Andersson 3657 0 3674 2 1 P/V atom ratio Fig. 13. The wavenumber of v,, as a function of the P/V ratio of different P-V-O/TiO, catalysts. - ~~ (a) Oxidized form; (b) reduced form. 20 A P/V atom ratio Fig. 14. The rate of oxidation of toluene at different temperatures as a function of the P/V atom ratio in P-V-O/TiO, catalysts.(a) 360 "C; (b) 340 "C; (c) 320 "C; (a) P/TiO, at 360 "C. the OH stretching frequency is shown as a function of the P/V ratio of the reduced and the oxidized forms of the P-V-O/TiO, catalysts. If the valence state of vanadium is taken as + 5 for the oxidized form of PV 0 and as + 3 for the reduced form of PV 0 the data in fig. 13 suggest that the oxidation state of the oxidized and the reduced forms of PV 2.0 is + 4. If this consideration is correct, the band shifts in vco for adsorbed CO (vide supra) can be reasonably explained by the variation in the oxidation state of vanadium species in different catalysts caused by the added phosphorus forming different surface phases. The phosphorus, when present in high amounts seems to stabilize strongly the vanadium in the +4 valence state.It prevents both oxidation and reduction of the vanadium species under present conditions. In the low-temperature spectra of the oxidized catalysts a weak band around 2000 cm-l, assigned to V4+, was observed at PV > 1, where ESCA also indicates V4+.3658 Influence of Phosphorus on V,O,/TiO, 100 80 - 60 ' 40 20 5 9 P . X n " 0 20 40 60 80 100 V/(P + V) by ESCA (atom%) Fig. 15. The relative activities rpvx/rpvo, where Xis the P/V atom ratio, shown as a function of the vanadium surface concentration, V/(P + V), as determined by ESCA. Activity and Selectivity Measurements Fig. 14 shows the reaction rate of toluene oxidation at different temperatures as a function of the amount of phosphorus added to the P-V-O/TiO, catalysts.A decrease in the rate of oxidation of toluene is observed with increasing P/V ratio, which agrees with literature data for vanadium phosphate catalyst^.'^^*^^ At the highest P/V ratio the rate decreased to almost as low a rate as measured for P/TiO,. The activity for toluene oxidation is strongly suppressed at low levels of phosphorus addition. This indicates that the first phosphorus bonds preferentially to the vanadium species and not to the titanium dioxide. Similar effects observed earlier for potassium additions are stronger. It is interesting to note that the rate does not uxry with the P/V ratio in a perfectly smooth manner (see fig. 14). A comparison with the surface composition as analysed by ESCA, see fig.5, shows that the surface V/(P+V) ratio varies in a similar manner. Fig. 15 shows a correlation between the activity of the various P-V-O/TiO, catalysts of different P/V ratios, divided by the activity of PV 0, and the surface vanadium composition calculated from ESCA measurements. This linear correlation strongly suggests that the activity for oxidation of toluene is mainly dependent on the amount of vanadium species exposed on the catalyst surface. The activity of the P/TiO, catalyst is very low and the effect of the phosphorus seems mainly to be due to a steric effect, by covering the surface and making the vanadium species less accessible to the reactants. This also follows from the effects on the activation energies. Fig. 16 shows an Arrhenius plot for the rate of oxidation of toluene over the various P-V-O/TiO, catalysts. It is evident that all lines are parallel, except a smaller deviation for P/TiO,, and the apparent activation energies calculated from the slopes are all within 132 & 6 kJ mol-l.The activity is thus only determined by the apparent pre-exponential factor. Most likely, it may be interpreted in terms of the number of active sites, stressing that the main effect of phosphorus is a steric influence. This is in contrast to potassium, where both activation energies and pre-exponential factors changed. The selectivity pattern for the P-V-O/TiO, catalysts is more complex than wasJ . Zhu, B. Rebenstorf and S. L. T. Andersson 3659 3 2 - 1 - 2 Fig. 16. Arrhenius plot for reaction rates in toluene oxidation.(a) PV 0; (b) PV ( d ) PV 1.0; (e) PV 1.25; (f) PV 2.0; (g) PV 5.0; (h) P/TiO,. 0.25; (c) PV 0.5; 15 10 g - 3 >r > d .I .- m 5 I I I I 1 2 3 4 5 P/V atom ratio Fig. 17. Selectivities for oxidation products in oxidation of toluene at 340 "C over P-V-O/TiO, catalysts as a function of the P/V ratio. (a) Side-chain oxidation products (SPI); (b) oxidative coupling products (SPII); (c) methyl-diphenylmethane (MDPM) ; ( d ) carbon oxides (CO + CO,). Selectivities for P/TiO,: (0) CO+CO,; (V) SPI. (MDPM is also contained in SPII.)3660 Influence of Phosphorus on V,O,/TiO, SPII toluene - SPI CO+C@ Fig. 18. Reaction network in oxidation of toluene over P-V-O/TiO, catalysts. Notations as in fig. 17. Fig. 19. Model for phosphorus vanadium compounds formed on titania and the reaction mechanism.observed for the K-V-O/TiO, catalysts. A lot of products formed in the oxidative coupling route31 are produced : methyl-diphenylmethane (MDPM), o-methyl-diphenyl- methanone (OMDPM), p-methyl-diphenylmethanone (PMDPM), benzophenone (BP), diphenylethanone (DPE), phthalic anhydride (PA) and anthraquinone (AQ). To simplify the discussion, we add all these products together as one component, named SPII, since all are formed in the second reaction route. MDPM is also presented separately. In the first reaction route, i.e. side-chain oxidation, the products benzaldehyde (BA) and benzoic acid (BzA) are added together as one component, named SPI. The selectivity for these and for carbon oxides at the oxidation of toluene are shown in fig. 17 as a function of the P/V ratio.It is clear that the selectivity of SPI decreases .with increasing P/V ratio, whereas the selectivity for SPII, on the contrary, increases with increasing P/V ratio, which is also the case for the selectivity for carbon oxides. Note the linear increase in selectivity for MDPM, the first product along the oxidative coupling route. The carbon oxides increase only with increasing P/V ratio up to 1.25 and are thereafter constant. TiO, was earlier observed to have a very low activity for the oxidation of toluene and of the products as compared with V,O,. Any effects on selectivities from covering Ti,,, sites with P-species cannot therefore be detected. Considering that the present measurements were performed at low conversions (< 5 %), the selectivity pattern clearly show that the oxidation of toluene, over these catalysts, proceeds through three parallel routes (see fig.18). It is evident that phosphorus mainly promotes the production of oxidative coupling products and carbon oxides.J . Zhu, B. Rebenstorf and S. L. T. Andersson 366 1 Catalyst Structure and Reaction Mechanisms As shown above, the rate of oxidation of toluene was strongly influenced already at relatively low additions of phosphorus, and the surface area, XRD, ESCA and i.r. measurements indicated that these phosphorus species were highly dispersed over the vanadium layer. Only at the highest P/V ratios could separate phases be identified in very small amounts. A good linear correlation between the rate of oxidation of toluene and the vanadium surface concentration was shown, further indicating a high dispersion and surface enrichment of phosphorus.The active sites for the oxidation of toluene in these catalysts are doubtless vanadium sites and their activity does not seem to change with phosphorus addition, but only the number of those exposed on the surface to the reactants. This seems to support the current idea31-37 that the terminal oxygens on vanadium are important species for the initial hydrogen abstraction step in hydrocarbon oxidation. The selectivities are, on the contrary, affected, which points to some effect of phosphorus. In an earlier study,17 it was suggested that tetrahedral vanadium species are formed on the TiO, surface, doubly bonded to the titanium dioxide via oxygen bridges. It is suggested here that the OH group of these have to some extent reacted with the phosphorus species to give the structure A of pentavalent vanadium shown in fig.19. This species may be formed at low P/V ratios and may be converted upon reduction to the trivalent state. At higher P/V ratios aggregation of P-V-0 species occurs, probably with structures more similar to those found in vanadium phosphate^.^ With this structure, the steric effect of phosphorus in toluene oxidation is evident. Owing to the rather large volume of this species, the approach of the toluene molecule to the V=O species will be restricted, reducing the overall rate of reaction. After the initial hydrogen abstraction a second V=O site has been proposed as a requisite for a second hydrogen abstraction and oxygen i n ~ e r t i o n .~ ~ It has also been suggested that step sites at 'rough' surfaces are important in restricting the molecular motion and that the complete process occurs at the same site by diffusion of lattice oxide ions.33 Another viewpoint was obtained from i.r. studies of adsorbed toluene, suggesting a four-site The selectivity pattern presented here strongly suggests that a second active site is required for further selective oxidation of the intermediate formed after the first hydrogen abstraction. The intermediate is considered to be a benzyl radi~al,~'. 32, 38 probably surface bonded by some other interaction, because if diffusing out to the gas phase, carbon oxides would probably be formed.The environment of the active site may influence such an effect, which may be of crucial importance for the selectivity. It is suggested that the phosphorus species plays such a role and interacts with the aromatic nucleus producing species (b) in fig. 19. After hydrogen abstraction, this will give a relatively long-lived benzylic radical [species (c), fig. 191 which may interact with a second toluene molecule, producing methyl-diphenylmethanes, before it gets the opportunity to continue the reaction at another site. The phosphorus species could also influence the possibility of interaction with another site by steric hindrance. At higher P/V ratios P-V-0 agglomerates are also formed. These contain bulk oxide ions that may diffuse to the first active site, supplying the second oxygen, wherefore there will always be an opportunity for producing the side-chain products.This model suggests that the dynamics of the catalytic process are important. The intermediate may have various reaction routes to follow, and the distribution between these will be determined by potential energies and statistics. The Swedish Board for Technical Development and the National Energy Administration are acknowledged for financial support.3662 Injluence of Phosphorus on V,O,/TiO, References 1 J. Zhu and S. L. T. Anderson, J. Chem. SOC., Faraday Trans. 1, in press. 2 B. K. Hodnett, Catal. Rev. Sci. Eng., 1985, 27, 373. 3 Selective Catalytic Oxidation of C,-Hydrocarbons lo Maleic Anhydride, ed. B. K. Hodnett, Catal. Today 4 G.Centi and F. Trifiro, Chim. Znd., 1986, 68, 74. 5 M. Nakamura, K. Kawai and Y. Fujiwara, J. Cat!!., 1974, 34, 345. 6 R. Fricke, H. G. Jerschkewitz, G. Lischke and G. Ohlmann, 2. Anorg. Allg. Chem., 1979, 448, 23. 7 A. J. van Hengstum, J. Pranger, J. G. van Ommen and P. J. Gellings, Appl. Catal., 1984, 11, 317. 8 S. L. T. Anderson, J. Chem. Soc., Faraday Trans. 1, 1986, 82, 1537. 9 P. Cavalli, F. Cavani, 1. Manenti and F. Trifiro, Ind. Chem. Eng. Res. Dev., 1987, 26, 639. (Elsevier, Amsterdam, 1987), vol. 1. 10 F. Cavani, G. Centi, E. Foresti, F. Trifiro and G. Busca, J. Chem. SOC., Faraduy Trans. I , 1988,84,237. 11 B. Jonson, B. Rebenstorf, R. Larsson, S. L. T. Anderson and S. T. Lundin, J. Chem. SOC., Faraday 12 S. L. T. Anderson, J. Chromatogr. Sci., 1985, 23, 17. 13 R. A. Spurr and H. Myers, Anal. Chem., 1957, 29, 760. 14 S. L. T. Anderson, J. Chem. Soc., Faraday Trans. 1, 1979, 75, 1356. 15 W. J. Carter, G. K. Schweitzer and T. A. Carlson, J. Electron Spectrosc. Relat. Phenom., 1974, 5, 827. 16 B. Rebenstorf and R. Larsson, 2. Anorg. Allg. Chem., 1979, 453, 127. 17 B. Jonson, B. Rebenstorf, R. Larsson and S. L. T. Andersson, J. Chem. SOC., Faraduy Trans. I , 1988, 18 JCPDS, Powder Diffr. File, Inorg. Phases, Int. Centre Diffr. Data, Swarthmore, 1985. 19 E. Bordes and P. Courtine, J. Catal., 1979, 57, 236. 20 D. Cole, C. Cullis and D. Hucknall, J. Chem. Soc., Furuduy Trans. I , 1976, 72, 2185. 21 A. Vejux and P. Courtine, J. Solid State Chem., 1978, 23, 93. 22 J. Criado and C. Real, J. Chem. Soc., Faraduy Trans. I , 1983, 79, 2765. 23 G. Busca, H. Saussey, 0. Saur, J. C. Lawalley and V. Lorenzelli, Appl. Catul., 1985, 14, 245. 24 M. I. Zaki, B. Vielhaber and H. Knozinger, J. Phys. Chem., 1986, 90, 3176. 25 G. Busca, Langmuir, 1986, 2, 577. 26 M. I. Zaki and H. Knozinger, Spectrochim. Acta, Part A , 1987, 43, 1455. 27 B. Jonson, Thesis (University of Lund, Sweden, 1988). 28 C. Morterra, J. Chem. SOC., Faraday Trans. I , 1988, 84, 1617. 29 A. Fernandez, J. Leyrer, A. R. Gonzalez-Elipe, G. Munuera and H. Knozinger, J. Catal., 1988, 112, 489. 30 B. Jonson, B. Rebenstorf, R. Larsson and S. L. T. Andersson, J. Chem. SOC., Faruday Trans. I , 1988, 84, 1897. 31 S. L. T. Anderson, J. Catal., 1986, 98, 138. 32 G. Busca, F. Cavani and F. Trifiro, J, Catal., 1987, 106, 471. 33 K. Mori, A. Miyamoto and Y. Murakami, J. Chem. SOC., Faraday Trans. I, 1987, 83, 3303. 34 J. C. Otamiri and A. Anderson, Catal. Today, 1988, 3, 21 1. 35 G. C. Bond, J. P. Zurita and S. Flamerz, Appl. Catal., 1986, 22, 361. 36 J. Haber, in Proc. 8th Znt. Congr. Catal., Berlin, F.R.G., July 2-6, 1984, Verlag Chemie, Weinheim, 37 A. J. van Hengstum, J. Pranger, S. M. van Hengstum-Nijhuis, J. G. van Ommen and P. J. Gellings, 38 J. E. Germain and R. Laugier, Bull. SOC. Chim. Fr., 1971, 2, 750. Trans. I , 1986, 82, 767. 84, 3547. 1984, vol. 1, p. 85. J. Catal., 1986, 101, 323. Paper 9100251K; Received 14th January, 1989
ISSN:0300-9599
DOI:10.1039/F19898503645
出版商:RSC
年代:1989
数据来源: RSC
|
9. |
Structure of vanadium oxides on ZrO2and the oxidation of butene |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 85,
Issue 11,
1989,
Page 3663-3673
Hisashi Miyata,
Preview
|
PDF (641KB)
|
|
摘要:
J. Chem. SOC., FaradajJ Trans. 1, 1989, 85( 1 I), 3663-3673 Structure of Vanadium Oxides on ZrO, and the Oxidation of Butene Hisashi Miyata,* Mitsuru Kohno and Takehiko Ono* Department of Applied Chemistry, University of Osaka Prefecture, Sakai, Osaka 591, Japan Takashi Ohno and Fumikazu Hatayama School of Allied Medical Sciences, Kobe University, Suma, Kobe 654-01, Japan The oxidation of butene on V-Zr oxides prepared by the gas-phase and the impregnation method has been studied by F.t.i.r. spectroscopy as well as the microcatalytic method. V-Zr catalysts above 6.0 wt % vanadium loading show almost constant activity. V-Zr catalysts at low vanadium loadings show high selectivities to furan and buta- 1,3-diene, while high vanadium loading catalysts show high selectivities to acetaldehyde and acetic acid.The formation of a dihydrofuran-like intermediate before the formation of furan has been proposed. The oxyhydrative scission mechanism is proposed in the formation of acetaldehyde and acetic acid. This formation is attributed to the presence of Brnrnsted acidity. Vanadium oxide catalysts in combination with various promoters are widely used for selective oxidation of hydrocarbons. We have recently reported the characterization of various vanadium/metal oxides prepared by a gas-phase and have proposed that new species are formed by the reaction of vanadium oxytrichloride with surface OH groups on carrier oxides. Materials of this class are of growing interest in a wide range of practical applications. In previous papers,"*5 we have also reported the catalyst structures and catalytic activities of such oxide catalysts. Few studies for the selective oxidation of alkene on V-Zr catalyst have appeared in the literature.In the present work. by the selective immobilization of V,O, mono- and multi-layers on ZrO,, the influence of the support on the oxidation selectivity of alkene and intermediates in the oxidation has been investigated using Fourier- transform infrared spectroscopy as well as the microcatalytic method. Experimental The zirconium hydroxide was prepared from zirconium oxydichloride. The hydroxide was calcined at 383 K, followed by decomposition at 723 K. Two preparation methods were used, The first was a wet impregnation method with ammonium metavanadate (VZr-2.0-VZr-24, 2.0-24 wt % vanadium oxide as V,O,).The second was a gas-phase preparation method using vanadium oxytrichloride. In order to obtain samples with different vanadium layers, the circulation of vanadium oxytrichloride vapour was repeated several times. The vanadium contents ranged from 1.8 to 6.3 wt % vanadium oxide as V,05 (GVZr- 1.8-GVZr-6.3). The concentration of the surface OH groups on the zirconia, which controls the formation of the vanadium oxide surface phase, was fixed by appropriate choice of the pretreatment temperature of the zirconia. Details of the prepraration methods of catalysts were described previously.' The catalyst was heated to 723 K under evacuation and kept at that temperature under flow of oxygen (ca. 4 kPa) for 2 h. This treatment was repeated several times before each experiment.The physical parameters of those catalysts are listed in table 1 . 36633664 Structure and Butene Oxidation on V,O,/ZrO, Table 1. Physical properties of GVZr and VZr catalysts preparation catalyst methoda V,O, wt O h surface area/mz g-I VZr-2.0 VZr- 3.7 VZr-6.0 VZr-7.6 VZr-8.4 VZr- 15.6 VZr-24.0 GVZr- 1.8 GVZr-2.0 GVZr-6.3 2.0 3.7 6.0 7.6 8.4 15.6 24.0 1.8 2.0 6.3 67 68 60 44 53 49 39 62 61 53 a Gas = gas-phase preparation; wet = wet impregnation method. cycles of VOC1, circulation. Parentheses show number of Two types of apparatus were used. The first was a conventional closed-circulation system equipped with an i.r. cell in the circulation loop. The second was a pulse microreactor system directly connected with a gas chromatograph for analysis of the reaction products.Details of the apparatus and procedures were described in previous Fourier-transform infrared spectra were recorded on a Shimadzu FTIR-4000. After 100 accumulations had been stored the spectral data were transferred on an IEEE-488 bus line to a master computer (PC-9801 VX2, NEC). In order to obtain quantitative information on the spectral behaviour, some of the spectra were subtracted and deconvoluted by using a data acquisition system. The details of the data acquisition and analysis system were described previously. ' 7 ' 9 For a pulse microreactor study, the dried catalysts were tested in a fixed-bed reactor. The catalyst charge of 30 mg was preheated with flowing 0, for 1 h at 673 K, followed by flowing He for 1 h at reaction temperature (373-573 K).Following pretreatment, a gaseous mixture of Z-but-2-ene, oxygen, and helium (0.04,0.08, 1.7 mmol) was fed over the catalyst. Gaseous reaction products were measured by gas chromatography with a column of MS- 13X or GCPAC-54. Conversion and product selectivity were calculated as follows. Conversion (%) = 100 x (mole of butene reacted)/(mole of butene fed) Selectivity (YO) = 100 x (mole of product reduced to C,)/(mole of butene reacted). The characterization of the catalyst structures was carried out using XRD, F.t.i.r. and laser Raman techniques.'v Results and Discussion Microcatalyic Results Catalysts were tested for Z-but-2-ene oxidation in the temperature range 373-573 K. A fresh sample was used for each reaction temperature.As shown in fig. 1, the conversion of Z-but-2-ene is increased with increasing vanadium loadings and reached saturated value at vanadium loadings above ca. 6.0 wt %, suggesting that the active sites of catalysts are almost constant above 6.0 wt YO. Reaction products varied with the reaction temperatures and with vanadium loadings of catalysts. A trace of formic acid and maleic anhydride was formed on the catalysts at high vanadium loadings. Fig. 2 shows the selectivities of butene oxidation at various temperatures. The catalysts at lowH. Miyata et al. 3665 Fig. 1. Total v*o, (wt%) 473, (c) 573 K. Fig. 2. Products selectivity of 2-but-2-ene oxidation on V-Zr catalysts. (a) Furan, (b) buta-1,3- diene, (c) CH,CHO, ( d ) CH,COOH, CO+CO, (dotted lines); 0, 573 K; A, 473 K; a, 373 K.3666 Structure and Butene Oxidation on V,05/Zr0, 0.1 rl % 1'7 - 15 ' * 13 ' * 1 1 ' wavenumber/ lo2 cm- ' Fig.3. F.t.i.r. spectra of Z-but-2-ene adsorbed on GVZr-6.3. (a) Background, (b) after introduction of 2-but-2-ene (ca. 1 . 1 kPa for 30 min at 293 K) followed by 30 min evacuation at 293 K, (c) followed by 30 rnin at 353 K in oxygen, ( d ) 30 rnin at 413 K in oxygen, (e) 30 min at 473 K in oxygen. The spectra of (b)-(e) below 1800 cm-' shown after subtraction of background. vanadium loadings show high selectivity to furan and buta- 1,3-diene, while the catalysts at vanadium loadings above 6.0 wt YO show high selectivity to acetaldehyde and acetic acid. At high temperature the catalysts above 6.0 wt % show moderate selectivity to furan.Above 473 K total oxidation, as CO and CO, formation, occurred. The results showed that the nature of surface vanadia species on the catalysts at low vanadium loadings are different from those of high vanadium loading catalysts. It should be noted that pure zirconia showed little or no activity for the oxidation of Z-but-2-ene at these temperatures. F.t.i.r. Spectra of cis-But-2-ene on GVZr-6.3 and GVZr-2.0 As described above, V-Zr catalysts at vanadium loadings above 6.0 wt YO show almost the same activities as, but different selectivities from, low vanadium loading catalysts. Therefore, we focused on the catalysts with mono- and multi-layered vanadia (GVZr- 2.0 and GVZr-6.3) at vanadium loading below ca. 6 wt YO. Fig. 3(b) shows the spectrum of 2-but-2-ene adsorbed on GVZr-6.3 at room temperature.The spectrum exhibits the OH band around 3665 cm-l, the CH stretching band near 3000 cm-' and bands in the bending region. Sharp absorption bands at 1463 and 1385 cm-l are due to the bending modes of CH, group of adsorbed butene. The broad band at 1645 cm-' is assigned to v(C=C) of n-bonded b ~ t e n e , ~ suggesting that Z-but-2-ene is adsorbed as a n-complex on GVZr-6.3 at room temperature. Oxygen of 2.5 kPa was admitted to the GVZr-6.3 containing Z-but-2-ene, the temperature of the catalyst being raised in stages under circulation of oxygen [fig. 3(c)-(e)]. With increasing disc temperature the intensities of the bands due to the n-complex decreased ; simultaneously, new bands appeared at 1565, 1450, 141 5 and 1355 cm-l and were intensified.Surface acetate species have been found around 1550 and 1440 cm-l in the cases of various metal oxide^,^^^^-'^ which have been attributable to the asymmetric and symmetric stretching vibrations of acetate ions. Thus, the bands at 1565 and 1415 cm-' are assigned to va,(COO) and v,(COO) of surface acetate species, respectively. Considering that the asymmetric v(C00) may be overlapped, the bands at 1565 and 1355 cm-l are attributable to same modes of surface formate species.',H . Miyata et al. 3667 I , I . I . I 17 15 13 11 wavenumber/ 1 O2 cm- ' Fig. 4. F.t.i.r. spectra of Z-but-2-ene adsorbed on GVZr-1.8. (a) Background, (b) after introduction of Z-but-2-ene (ca. 1.1 kPa for 30 min at 293 K) followed by 30 min evacuation at 293 K, (c) followed by 30 min at 353 K in oxygen, ( d ) 30 min at 413 K in oxygen, (e) 30 rnin at 473 K in oxygen, (f) after introduction of small amount of dihydrofuran followed by oxidation at 353 K for 30 min, (g) followed by 30 rnin at 413 K in oxygen.The spectra of (b)-(g) below 1800 cm-' shown after subtraction of background. I k;; -' - 35 30 ' 1 I 1 1 17 15 13 11 I k;: - 35 30 ' wavenumber/ 1 O2 cm- ' Fig. 5. F.t.i.r. spectra of Z-but-2-ene adsorbed on ZrO,. (a) Background, (b) after introduction of Z-but-2-ene (ca. 1.1 kPa for 30 rnin at 293 K) followed by 30 rnin evacuation at 293 K, (c) followed by 30 rnin at 353 K in oxygen, ( d ) 30 rnin at 413 K in oxygen, (e) 30 rnin at 473 K in oxygen. The spectra of (6)-(e) below 1800 cm-' shown after subtraction of background.3668 Structure and Butene Oxidation on V,O,/ZrO, d / / 5 0 5 1( 00 i0 YO5 (wt%> Fig.6. Products selectivity of butan-2-one oxidation on GVZr-6.3 and on GVZr-2.0. (a) CH,CHO, (b) CH,COOH; 0, 573 K; A, 473 K; 0 , 373 K. 0.1 I I ,1 1 I 1 I , 1 , 35 30 17 15 13 11 wavenumber/102 cm-' Fig. 7. F.t.i.r. spectra of butan-2-01 adsorbed on GVZr-6.3. (a) Background, (b) after introduction of butan-2-01 (ca. 1.1 kPa for 30 min at 293 K) followed by 30 min evacuation at 293 K, (c) after 14 h of adsorption at 298 K, ( d ) followed by 30 min at 353 K in oxygen, (e) 30 min at 413 K in oxygen, (f) 30 min at 473 K in oxygen. The spectra of (b)-(f) below 1800 cm-' shown after subtraction of background. Similar experiments were carried out with GVZr-1.8 (fig.4). Only weak bands appeared at almost the same positions as those in the case of GVZr-6.3 at room temperature [fig. 4(b)], suggesting that the surface vanadia provides active sites for 2-but-2-ene adsorption. On oxidation with increasing temperature the spectral behaviour was observed. At 353 K new bands, which were not observed with GVZr-6.3,H. Miyata et al. 3669 1 "-R: 17 15 13 11 35 30 / wavenumber/lO? cm-' Fig. 8. F.t.i.r. spectra of butan-2-one adsorbed on GVZr-6.3. (a) Background, (b) after introduction of butan-2-one (ca. 1.1 kPa for 30 min at 293 K) followed by 30 min evacuation at 293 K, (c) after 15 h of adsorption at 298 K, (d) followed by 30 min at 353 K in oxygen, (e) 30 min at 413 K in oxygen. The spectra of (b)+e) below 1800 cm-' shown after subtraction of background. appeared at 1666, 1610, 1392, and 1356 cm-l [fig.4(c)]. As reported previ~usly,~ at low temperature the adsorbed 2-but-2-ene molecule on V,O,/TiO, incorporates 0.5 oxygen molecules. Thus, the formation of oxygen-containing intermediates is suggested in a similar reaction. Above 413 K, bands at 1561, 1446, 1383, and 1363 cm-' are intensified [fig. 4 ( 4 , (e)], suggesting that a similar reaction with GVZr-6.3 took place. In order to clarify new intermediate species formed above 353 K on GVZr-1.8, the spectra of oxygen-containing compounds adsorbed on catalysts and their spectral behaviour were studied. Fig. 4(f) shows the spectrum of 2,3-dihydrofuran adsorbed on GVZr-1.8, at 353 K. Bands were observed at 1662, 1615, 1453, 1390, and 1360 cm-l.Only this compound was found to give essentially the same spectra and spectral behaviour as the bands of Z-but-2-ene. Therefore the species formed at ca. 353 K are tentatively assigned to a dihydrofuran-like structure. Similar experiments were carried out with ZrO, alone. The resulting spectra are shown in fig. 5. Above 413 K the acetate and formate bands are observed at 1548 and 1447 cm-l, and 1570 and 1361 cm-l, respectively. These bands are much weaker than those with GVZr-6.3 and GVZr-1.8. In the oxyhydrative scission mechanism'? l4 of butene, the formation of enole-type intermediates is pr~posed.~ Thus, the adsorption of butan-2-01 and butan-2-one, and their oxidation over GVZr catalysts were investigated. Fig. 6 shows the results with the oxidation of butan-2-one over the GVZr-6.3 and GVZr- 1.8.The products were almost the same as those on butene oxidation. The GVZr-6.3 favours the formation of acetaldehyde and acetic acid. The spectrum of butan-2-01 adsorbed on GVZr-6.3 exhibits bands due to alkoxide species [fig. 7(b)]. On increasing the period of adsorption the 1660 cm-l band was intensified [fig. 7(c)], which can be assigned to the C=O stretching vibration of coordinately adsorbed ketone on Lewis-acid sites. After oxygen of 2.5 kPa was introduced onto the catalyst the temperature of the disc was raised in stages. At 353 K the 1550 cm-l band was intensified. In addition the OH band at 3660 cm-l was intensified, suggesting that the coordinated adsorbed ketone was3670 Structure and Butene Oxidation on V,O,/ZrO, I I ,# I .I . I . I 35 30 " 17 15 13 11 wavenumber/102 cm-' Fig. 9. F.t.i.r. spectra of butan-2-one adsorbed on GVZr-1.8. (a) Background, (b) after introduction of butan-2-one (ca. 1.1 kPa for 30 min at 293 K) followed by 30 min evacuation at 293 K, (c) after 15 h of adsorption at 298 K, ( d ) followed by 30 min at 353 K in oxygen, (e) 30 min at 413 K in oxygen. The spectra of (b)-(e) below 1800 cm-' shown after subtraction of background. Table 2. Characterization of the catalysts fraction of surface conc. amorphous of amorphous Raman catalyst v,o, (W oxide/pmol m-* i.r. band/cm-' band/cm-' VZr-2.0 VZr-3.7 VZr-6.0 VZr-7.6 VZr-8.4 VZr- 15.6 VZr-24 GVZr-2.0 GV Zr- 6.3 100 100 60 75 40 20 100 100 - 3 6 12 12 13 14 15 4 13 972 - - - 1005, 986 - 1008, 986 - 1018, 1002 - 1011, 983 1020 - 1025 - 978 1011, 995, 979 - - - - 1010-998 1023, 995 1032, 996 1032 - 1032 - 994 997 1020, 1000 1032 - - - dehydrogenated to enole-type species.The spectrum above 413 K resembles that of butene at 473 K, suggesting that a similar reaction occurred. Fig. 8 shows the spectra of butan-2-one on GVZr-6.3. On increasing the time of adsorption, the 1550 cm-l band, which can be attributed to the enol-type of adsorbed ketone, appeared. The spectra at higher temperatures are similar to those for butene and butan-2-01. In the case of GVZr-1.8, the enole-type bands appeared at rather low temperatures [fig. 9 (41. Structure Characterization of Catalysts The structure of vanadia species dispersed on zirconia was studied by XRD, F.t.i.r. and laser Raman techniques.XRD results showed that the VZr catalysts at vanadium loadings below 6.0 wt YO together with GVZr catalysts have no diffraction lines due toH. Miyata et al. 367 1 1 1 10 9 1 1 10 9 wavenumber/ 10' cm- ' Fig. 10. F.t.i.r. spectra of Z-but-2-ene adsorbed on (A) GVZr-6.3 and (B) GVZr-1.8. (a)-(e) As for fig. 3 and 4. crystalline V205, indicating that the most of the vanadia species are highly dispersed on zirconia. The amounts of crystalline V,O, and amorphous vanadia species on catalysts were estimated by using the method applied previously." The concentrations of amorphous surface vanadia species per surface area are listed in table 2. Although the calculation of amorphous species from XRD lines may include some errors, the concentrations of this species are almost constant for the catalysts at vanadium loadings above 6.0 wt %.Table 2 also summarizes the observed or calculated V=O bands of i.r.l and Raman spectra. The mono-layered sample (GVZr-2.0), which was prepared by a single cycle of VOCl, circulation, exhibits a single i.r. band at ca. 980 cm-', whereas the multi-layered catalyst (GVZr-6.3) shows complex bands, which were separated into three bands.l Generally, the i.r. band of V=O in crystalline V205 shows at 1020-1025 cm-' and the Raman band at 995 cm-l. On the other hand, the i.r. band at 980 cm-' and the Raman band at 1026 cm-l are attributable to those of polyvanadate. In the present study, in situ i.r. spectra of V=O region were observed. The spectra below 1100 cm-l corresponding to those in fig.3 and 4 are shown in fig. 10. The V=O band appeared at 1040cm-l after oxidation of GVZr-6.3 at 723 K [fig. IOA]. Introduction of butene on the catalyst caused the 5 cm-l shift to lower wavenumber and the slight reduction in intensity. The intensity and the position of this band were not changed appreciably on oxidation of butene above 473 K. A similar band appeared at 1030 cm-l for GVZr-1.8 [fig. lOB]. Thus, the V=O band shifts to a higher wavenumber in the reaction condition than that in the KBr disc.' Although further discussion will not be given in the present paper, those bands seem to be composed of two or more bands. As regards the character of V=O species of surface vanadate, the different reactivity would be expected with mono- or multi-layered vanadia species.121 F A R I3672 Structure and Butene Oxidation on V,O,/ZrO, butadiene furan p 3 w\ I w-!-cHsH3 r) O$"CHCH3 I? /-A C \ -v-0-v-0-v- + -v-0-v-0-4- + -0-0-0-0- * acetaldehyde acetic acid Fig. 11. Reaction scheme. Catalyst Structure and Catalytic Activities From the above considerations we propose that the oxidation of Z-but-2-ene proceeds as shown in fig. 11. Spectral behaviour together with high selectivity (77%) to acetaldehyde confirms that the cleavage of C-C bond takes place above 353 K. According to the proposal by Takita et aZ.,14 the butenes are converted to acetaldehyde and acetic acid in the presence of water on vanadium oxide catalysts by the oxyhydrative scission mechanism, where Brarnsted-acid sites play an important role in the hydration step.On the GVZr-2.0,Z-but-2-ene is mainly dehydrogenated to buta- 1,3-diene, and furan is formed at 373 K, although at high temperatures the formation of acetaldehyde is appreciable. This feature, together with the presence of the dihydrofuran-like species on this catalyst, suggests that the dehydrogenation proceeds via 71-ally1 species of butene, although the band due to 71-ally1 species (ca. 1600 ~ m - l ) ~ is not clear in fig. 3. Therefore furan formation on the GVZr catalyst at low vanadium loading occurs through the same pathway as that reported on the V-Ti As described above, the oxidation activities correlate well with the concentration of amorphous vanadia species. The difference of the selectivity between GVZr-2.0 (GVZr-1.8) and GVZr-6.3 seems to arise from the nature of surface vanadate, being accompanied by the concentration of amorphous vanadate or by the vanadium layers covered on zirconia.As described in the structural characterization by i.r. and Raman spectroscopies, the V=O species in a mono-layered vanadia is more weakened than that in a multi-layered vanadia by the interaction or bonding with ZrO,. Thus, the mono- layered vanadia favours the formation of furan. Low vanadium loading catalyst showed little or no concentration of Brarnsted acidity, while high vanadium loading catalyst showed high Brarnsted acidic concentration. l6 Considering high activity for acetaldehyde and acetic acid formation on GVZr-6.3, the C-C bond scission is closely related to Brarnsted acidity. Thus, a situation similar to that reported for the V,O,/TiO, system5 would be expected.References 1 H. Miyata, K. Fuji, T. Ono, Y. Kubokawa, T. Ohno and F. Hatayama, J. Chem. SOC., Faraday Trans. 2 H. Miyata, H. Nishiguchi and T. Ono, Chem. Express, 1988, 3, 243. I, 1987, 83, 675.H . Miyata et al. 3673 3 H. Miyata, K. Fujii and T. Ono, J. Chem. SOC., Faraday Trans. I , 1988, 84, 3121. 4 H. Miyata, T. Mukai, T. Ono, T. Ohno and F. Hatayama, J. Chem. SOC., Faraday Trans. I , 1988, 84, 5 T. Ono, T. Mukai, H. Miyata, T. Ohno and F. Hatayama, Appl. Catal., 1989, 49, 273. 6 T. Nakajima, T. Sonoda, H. Miyata and Y. Kubokawa, J. Chem. SOC., Faraday Trans. I , 1982,78,555. 7 H. Miyata, K. Fujii, S. Inui and Y. Kubokawa, Appl. Spectrosc., 1986, 40, 1177. 8 H. Miyata, S. Tokuda and T. Yoshida, Appl. Spectrosc., 1989, 43, 522. 9 A. Ramstetter and M. Baerns, J. Catal., 1988, 109, 303. 2465. 10 H. Miyata, T. Nakajima and Y. Kubokawa, J. Catal., 1981, 69, 292. I 1 T. Nakajima, H. Miyata and Y. Kubokawa, J. Chem. SOC., Faraday Trans. I , 1985, 81, 2409. 12 S. J. Puttock and C. H. Rochester, J. Chem. SOC., Faraday Trans. I , 1986, 82, 3013. 13 M. Y. He and J. G. Ekerdt, J. Catal., 1984, 87, 381. 14 Y. Takita, T. Maehara, N. Yamazoe and T. Seiyama, J. Catal., 1987, 104, 359; Y. Takita, K. Nitta, T. Maehara, N. Yamazoe and T. Seiyama, J. Catal., 1977,50, 364; T. Seiyama, K. Nitta, T. Maehara, N. Yamazoe and Y. Takita, J. Catal.. 1977, 49, 164. 15 T. Ono, H. Miyata and Y. Kubokawa, J. Chem. SOC., Faraday Trans. I , 1987, 83, 1761. 16 H. Miyata, M. Kohno, T. Ono, T. Ohno and F. Hatayama, unpublished results. Paper 8/05054F ; Received 29th December, 1988 121-2
ISSN:0300-9599
DOI:10.1039/F19898503663
出版商:RSC
年代:1989
数据来源: RSC
|
10. |
Initial cracking properties and physicochemical characterization of acid-leached small-port (SP) and large-port (LP) mordenites by pulse n-hexane cracking, infrared and27Al magic angle spinning nuclear magnetic resonance spectroscopy |
|
Journal of the Chemical Society, Faraday Transactions 1: Physical Chemistry in Condensed Phases,
Volume 85,
Issue 11,
1989,
Page 3675-3685
Filip Goovaerts,
Preview
|
PDF (677KB)
|
|
摘要:
J. Chem. Suc., Faraday Trans. I , 1989, 85(1 l), 3675-3685 Initial Cracking Properties and Physicochemical Characterization of Acid-leached Small-port (SP) and Large- port (LP) Mordenites by Pulse n-Hexane Cracking, Infrared and 27Al Magic Angle Spinning Nuclear Magnetic Resonance Spectroscopy Filip Goovaerts, Etienne F. Vansant, Jos Philippaerts and Peter De Hulsters University of Antwerp (U.I.A.), Laboratory of Inorganic Chemistry Universiteitsplein I , B-2610 Wilrijk, Belgium Jan Gelan Limburg University, Department SBM, University Campus, B-3610 Diepenbeek, Belgium Small- and large-port mordenite zeolites have been characterized by 27Al m.a.s.n.m.r. and infrared spectroscopy to evaluate the observed changes in catalytic properties as a function of the dealumination degree.As a test reaction, n-hexane cracking was used. A dual Lewis-Brmsted cracking mechanism, together with a specific structural change upon dealumination, is proposed to be responsible for the high activity of some of these H- mordenites. Recently a number of papers involving studies of dealuminated mordenites were p~b1ished.l-l~ This increased interest probably results from the excellent reactivity and stability of high-silica-content ZSM-5-type zeolites in hydrocarbon reactions.l* Since mordenite can be dealuminated to a high degree without losing its crystallinity,15-17 this zeolite is an extremely useful substrate in studying the effect of the residual aluminium content on the catalytic performance. Previous reports have shown that extracting aluminium from the mordenite framework by acid leaching results in large and frequently favourable changes in catalytic properties.18-20 Weller and Bauer discovered that the cracking of n-hexane increased with Si/Al ratio." A maximum was observed near an Si/A1 ratio of 9. Also in cyclohexane and n-pentane hydroisomerisation, the activity was found to increase after extracting a certain amount of structural aluminium from the f r a m e ~ 0 r k . l ~ This effect was also observed in cumene cracking2' and in the transalkylation of o-xylene.2 This interesting property of H-mordenite is not yet fully understood and further work in this region is necessary. Another peculiar property of H-mordenite was discovered by Minachev2' who noticed that the activity in n-pentane isomerisation is inversely proportional to the hydrogen pressure used, indicating the presence of hydrogenation-dehydrogenation functions even in the absence of noble metals such as Pd or Pt.Recently, substantial interest in characterizing acid-leached and thermally treated mordenites by 29Si, "A1 and 'H m.a.s.n.m.r. can be noted in the l i t e r a t ~ r e . ~ - ' ~ 27A1 m.a.s.n.m.r. spectroscopy is extremely useful in determining the localisation and the relative amount of both structural and non-structural aluminium species in zeolites. The study of the latter species in particular has opened up the discussion of whether these entities take part in catalytic hydrocarbon reactions together with the structural hydroxyl groups. 13, 22 The OH functional groups can be characterized by infrared spectroscopy.In H- mordenites three types of hydroxyl groups can be distinguished: OH I (3740 cm-') is 36753676 Acid-leached Small- and Large-port Mordenites assigned to hydroxyls on the outer zeolite crystal surface; OH I1 (3640 cm-l) is attributed to hydroxyls attached to non-structural aluminium species ; OH I11 (3605 cm-l) is assigned to acidic hydroxyls localised near structural aluminium. 23-26 27Al m.a.s.n.m.r. and infrared OH-stretching data were used to evaluate dealuminated mordenites together with their catalytic properties. Experimental Materials The starting materials, sodium small-port and sodium large-port mordeni te, supplied, respectively, by 'La Societk Chimique de la Grand Pavoisse' and the Norton Co., were acid-leached under various conditions of acid concentration, treatment time and temperature.The total aluminium contents of the mordenites were analysed by classical chemical analysis and are listed in table 1. Catalytic Evaluation Conversion and selectivity data on n-hexane cracking were obtained in a flow reactor containing 0.2g of wet catalyst, crushed to 60-100 mesh size. Before the catalytic measurements, the catalyst was dehydrated at 673 K for 1 h under an N, flow. Afterwards, a helium stream was passed through an n-hexane saturator at 273 K. A 2 cm3 pulse of this mixture was passed over the dehydrated catalyst at 573 K, resulting in an apparent contact time of 0.8s. The products, sampled in a liquid N, trap, were identified by an FID gas chromatograph. Conversion, reaction products and coke data were calculated relative to a blank run with (conversion O h ) = (reaction products %) + (coke O h ) .Selectivity data are defined as: Si = 100 AJAAC,, with Ai = product area in wt YO, AAC,, = conversion area in wt YO. 1.r. Spectroscopy Apparatus and Specifications The i.r. spectra were recorded on a Nicolet 5 DXB FTIR spectrometer using an MTEC 100 photoacoustic detector (PAS). The spectra were recorded with 4 cm-' resolution, a mirror velocity of 0.16 cm s-' and 1000 scans were averaged, Happ Genzel apodization was used and the single-beam spectra obtained in this way, were ratioed against the background photoacoustic spectrum of carbon black. After dehydration, the samples were stored in a glove box purged with dry nitrogen. The photoacoustic cell was purged with ultra-dry helium.Deu terat ion Procedure To study the hydroxyl group population of the mordenite samples in more detail, liquid D,O was used to deuterate the hydroxyls. By this technique, interference of residual water could be avoided. The zeolite samples were dehydrated at room temperature and deuterated for 2 h. Before dehydration at higher temperatures, liquid D,O was evacuated until vacuum conditions were reached, After this procedure, the absorptionF. Gooilaerts et al. 3677 Table 1. A1 contents determined by chemical analysis and 27Al m.a.s.n.m.r. sample acid-leaching procedure A I S / u . ~ . AINS/u.c Na MSP HMSPl HMSP2 HMSP3 HMSP4 HMSP5 HMSP6 Na MLP HMLPl HMLP2 HMLP3 Na MSP+2x0.5 mol dm-3 HCI; 3 h ; 298 K Na MSP+2 x 0.5 mol dm-3 HCI; 24 h ; 298 K Na MSP+5 mol dmP3 HCl; 1 h ; 373 K Na MSP + 5 mol dmP3 HCI; 3 h; 373 K Na MSP+5 mol dm-3 HCl; 10 h; 373 K Na MSP+9 mol dm-3 HCI; 20 h; 373 K Na MLP+2 x0.5 mol dmP3 HCl; 3 h; 298 K Na MLP+5 mol dm-3 HCI; 3 h; 373 K Na MLP+9 rnol dm-3 HCl; 20 h ; 373 K 2.4 2.4 2.3 2.2 2.0 1.5 0.8 2.4 2.4 2.0 1.7 7.4 5.9 5.8 4.9 3.8 1.8 7.4 6.3 4.7 3.5 - 0.0 1.5 1.3 1.1 0.9 0.7 0.0 1.1 1.6 1.5 - AIT = total aluminium content after dehydration in mmol g-l; Als,u,c, = number of structural aluminium atoms per unit cell; AINS,u.c, = number of non-structural aluminium atoms per unit cell; MSP = small-port mordenite: MLP = large-port mordenite.Table 2. Absorption bands of hydroxyl groups tY Pe v(OH)/cm-I v(OD)/cm-l I 3738 2758 I1 3660 2700 I11 3613 2665 bands of the reacted hydroxyl groups shifted to lower wavenumbers with an isotopic shift ratio of 0.738 (table 2).N . M. R. Spectroscopy The n.m.r. experiments were performed using a Varian LX 200 apparatus operating in the Fourier- transform mode. The 27Al spectra were measured by single-pulse excitation at 52.1 MHz (pulse width = 4 ,us; acquisition time = 0.1 5 s) and a magic-angle spinning frequency of 6 kHz. The chemical shifts were referenced to that of an AlCl, aqueous solution. Before measurement at room temperature, the samples were stored in a desiccator at constant water vapour pressure. Results and Discussion In fig. 1 (a) and (b) the n-hexane conversion, total reaction products and coke data are plotted against the total aluminium contents of the small-port (SP) and large-port (LP) mordenite samples.For both mordenite types, a sharp increase in conversion and reaction products is observed in a narrow dealumination range. Further extraction of aluminium results in a significant drop in catalytic activity. Structural dealumination results in a decrease of the acidic type I11 hydroxyl groups [v(OD) = 2665 cm-’1 with a parallel increase of type I OH groups [v(OD) = 2758 cm-l],3678 Acid-leached Small- and Large-port Mordenites 80. 70 * 60 - .- VJ 50. 8 ' 40 30 20 l o t I I I I 1 2 3 total AI content/mmoI g-' total AI content/mmol g-' Fig. 1. (a) n-Hexane cracking over acid-leached SP-mordenites. 0, YO conversion; A, YO reaction products; ., % coke. (b) n-Hexane cracking over acid-leached LP-mordenites. 0, YO conversion ; A, YO reaction products; W, YO coke.F.Goovaerts et al. 3679 2800.0 2760.0 2720.0 2680.0 2640.0 2600.0 2800.0 2760.0 2720.0 2680.0 2640.0 2600.0 wavenurnter/crn-‘ wavenumberkm- ’ Fig. 2. OD stretching bands of acid-leached SP- and LP-mordenites dehydrated at 723 K in uacuo. (a) SP-form: (i) HMSPl, (ii) HMSP2 (iii) HMSP3, (iv) HMSP4, (v) HMSPS, (vi) HMSP6. (6) LP- form (i) HMLPl, (ii) HMLP2, (iii) HMLP3. as shown in fig. 2(a) and (6). This result is in agreement with the dealumination mechanism proposed by Kerr.,’ \ 2 I I I / ‘\ Si 0 \ I / \ -- Si-0-Al .... / HO- S.iO + 3HC1 + 4 ( - - Si- OH) + A3+ + 3C1- \ / type 111 type I 0 Si Scheme 1 Considering the structural OH groups of type I11 to be the only activity-determining factor in n-hexane cracking, one should expect a constant or a decreased conversion level with increasing structural dealumination.However, to explain the observed conversion increase, other controlling factors must be considered. Recently, Meyers et a1.l indicated that non-structural aluminium species are also important in determining the accessibility and acidity of acid sites in H-mordenite. These results suggest that these species can play an important part in catalytic hydrocarbon reactions. Indeed, the decrease of these entities (sample HMSPl to sample HMSP4) must be considered as one of the reasons for the sharp increase in n-hexane conversion. For the LP-mordenites, however, the amount of non-structural aluminium determined by ‘7Al m.a.s.n.m.r. increases after dealumination at reflux temperature.This result disagrees with the i.r. spectra shown in fig. 3(a) and (b), recorded after dehydrating the samples in N, flow, a procedure identical with that used before catalytic evaluation. A more intense OD absorption band at 2700 cm-‘ can be observed in sample HMLPl3680 Acid- leached Small- and Large-por t Morden ites 1 I I I I I I 1 I 2800.0 2760.0 2720.0 2680.0 2640.0 2600.0 2800.0 2760.0 2720.0 2680.0 2640.0 2600.0 w avenumber/cm-' wavenumber/cm- ' Fig. 3. OD stretching bands of: (i) HMLP1, (ii) HMLP2, (iii) HMLP3. (a) Dehydrated at 473 K in N,; (6) dehydrated at 673 K in N,. I I I 1 I 2800.0 2760.0 2720.0 2680.0 2640.0 2600.0 w avenumber/m- ' I I I t 2800.0 2760.0 2720.0 2680.0 2640.0 2600.0 wavenurnber/cm-' Fig. 4. OD stretching bands of: (i) HMLPl, (ii) HMSPI.(a) Dehydrated at 473 K in N,; (b) dehydrated at 673 K in N,. compared to the LP samples 2 and 3. In the l i t e r a t ~ r e , ~ ~ this OD absorption band (OH : 3660 cm-l) is ascribed to OH groups attached to non-structural aluminium. Most probably, some non-structural aluminium present in sample HMLPl cannot be detected by 27Al m.a.s.n.m.r. so that the actual amount of these species is higher than the values listed in table 1. This statement is supported by the comparable intensities of the type I1 OD absorption in HMLPl and HMSPl [fig. 4(a) and (b)]. An extensive study of the hydroxyl group population was made as a function of dehydration temperature. The i.r. spectra are shown in fig. 5 . As the dehydration temperature increases (between 473 and 773 K), the 2700 cm-' OD absorption decreases.F.Goovaerts et al. 368 1 3000.0 2900.0 2800.0 2700.0 2600.0 2500.0 2400.0 2300.0 2200.0 wavenumber/cm-' Fig. 5. OD stretching bands of HMLPl as a function of dehydration temperature: (i) 473, (ii) 573, (iii) 673, (iv) 773 K. Assuming non-structural aluminium present in the hydrated samples to be octahedral [A1(H,0)J3+, the following dehydration reactions can be assumed : or T [A1(H20)J3+ -+ Al(OH), + 3H,O + 3H+ Al(OH), -+ AlOOH + H 2 0 T T AlOOH + H+ -+ AlO+ + H 2 0 T AI(OH), + aH+ + [Al(OH),J" + aH20 (3) (4) where a = +1, +2, +3. Reaction (1) explains the intense 2700 cm-l OD absorption after dehydrating at 473 K. A further increasing dehydration temperature up to 673 K results in the formation of AlOOH species. At 773 K or higher, A10+ cations are probably formed.Apart from the reactions (1)-(3), explaining the observed intensity changes, reaction (4) can not be ruled out. When [Al(OH),-,]"+ cations are present, some structural hydroxyl groups will be neutralised. The highest dealuminated sample HMSP6, still produces a significant n- hexane conversion (ca. 20%) so that at least one of the two remaining hydroxyl groups must be free for reaction. Therefore, if reaction (4) occurs, entities such as Al(0H); must exist so that only one of the two remaining structural hydroxyl groups is neutralized. Based on these assumptions, AlOOH or Al(0H); species can be expected in samples dehydrated at 673 K, the temperature used before catalytic evaluation. Additional information about the short-range environment of aluminium, is provided by the 27Al m.a.s.n.m.r.spectra shown in fig. 6 ( a ) and (b). Dependent on the acid- leaching procedure, two types of non-structural aluminium were detected. The mildly acid-leached samples, HMSPl and HMLP1, are both characterised by a broad n.m.r. band of non-structural aluminium in the 0 ppm region [fig. 6 (a), spectra (1) and (4)]. In HMSPl clearly two non-structural aluminium species are present. The broad n.m.r. shoulder is due to a non-mobile aluminium entity with a relatively small relaxation time.3682 Acid-leached Small- and Large-port Mordenitds - 100 50 0 -50 12) 100 50 0 -50 ( 3 ) r 100 50 0 -50 100 50 0 100 50 0 50 0 100 50 0 - 5 0 100 50 0 -50 Fig. 6. (1)-(6) 27A1 m.a.s.n.m.r.spectra of acid-leached SP- and LP-rnordenites. (1) HMSPl ; (2) HMSP4; (3) HMSP6; (4) HMLPl; ( 5 ) HMLP2; (6) HMLP3. (7) and (8) "A1 m.a.s.n.m.r. spectra of the thermally treated (1 h, 673 K, N,) samples HMSPl and HMSP4. The narrow band species have a less restricted motion and are located in ion-exchange positions as A13+. Severe acid treatment at reflux temperature favours the formation of this narrow-band aluminium, as shown in spectra (2), (3), ( 5 ) and (6). The fact that a thermal treatment (673 K in an N, flow) followed by a rehydration step, transforms all the narrow-band species to broad-band aluminium entities indicates a change in localisation [fig. 6, spectra (7) and (S)]. This broadening effect cannot be due to a further extraction of structural aluminium because the amount of non-structural species present in the thermally treated samples is the same as in the parent samples.The migration of Ca2' ions from the large 12-ring channels to the smaller side-pockets upon dehydration has been reported in the literature.,' This suggests that, hydrated A13+ cations originally present in the large 12-ring channels lose their coordination waterF. Goouaerts et al. 3683 upon thermal treatment and move to the side-pockets. The possibility of non-structural aluminium being present in these smaller cages also explains that this effect can be observed in mild acid-leached samples. As suggested by Bodart" the first structural aluminium atoms to be attacked by acid leaching are those in the four-membered rings.These rings interconnect the large 12-ring channels with the side-pockets so that a mild dealumination results in depositing non-structural aluminium in these holes. Increased dealumination at reflux temperature extracts more four-ring aluminium which results in an opening of the side-pockets. In these samples the residual non-structural aluminium entities are present in the large 12-ring channels as completely hydrated A13+. These cations possess a high symmetric octahedral coordination which results in the observed sharpness of their n.m.r. peaks. The coordination water decreases the importance of the specific Coulomb interaction with the zeolite framework so that mobile aluminium entities can be expected (large &). The band-broadening upon dehydration is the result of a relocalisation in the side-pockets where the number of water ligands is lowered and the coordination symmetry of the complex ions is less.Based on the n.m.r. and i.r. data, we can identify the non-structural aluminium in dehydrated samples as AlOOH or Al(0H); entities located in the side-pockets. These data are in agreement with the observed catalytic results. Indeed, zeolite samples with a dealumination degree which resulted in an opening of the side-pockets are the most active catalysts. In these pockets one non-structural aluminium entity is present and acts as a dehydrogenation-hydrogenation site (table 1, HMLP2 and HMSP4). In this manner the incoming n-hexane molecules are transformed to the more active hexene, which reacts easily with the remaining structural hydroxyl groups present in the large 12- ring channels.Severe dealumination (increased acid concentration and treatment time) lowers the number of structural hydroxyl groups in the zeolite with a drop in conversion as a consequence [fig. 1 (a): HMSP4 -+ HMSP61. The activity of non-structural aluminium in hydrogen distribution is supported by the high amount of saturated reaction products as shown in table 3. All samples produce a high ratio of aliphatic to olefinic cracking products with maximum values for the most active samples. This latter effect proves that the strongest adsorption, and consequently the highest hydrogenation efficiency, is reached when the non-structural aluminium is located in accessible side-pockets. Also the data of Minachev21 support the de- hydrogenation-hydrogenation properties of H-mordenite.An increased n-hexane conversion results in dramatic changes in the product selectivity as shown in fig. 7(a) and (b). The isobutane selectivity attains maximum values for the most active SP and LP samples. Assuming the following cracking pattern : c, -K,+c2 and C,+C5+C, with C, = iso-C,+n-C,, C5 = iso-C5 + n-C, C,, C, = coke+small amount of detected C,, C, the theoretical coke production can be calculated, proposing the C, and C , products to be strongly adsorped on the catalyst giving coke formation. Since the amount of individual reaction products were calculated in weight percent, the following relations are valid:3684 Acid-leached Small- and Large-port Mordenites Table 3.Ratio of aliphatic to olefinic products as a function of the conversion level sample conversion (%) C,,/C,, c 4 ~ / c 4 ~ C t ~ t . ~ / ~ t ~ t . o HMSPl HMSP2 HMSP3 HMSP4 HMSP5 HMSP6 HMLPl HMLP2 HMLP3 36 70 79 80 72 19 17 78 79 4.4 7.5 9.6 9.9 9.7 4.1 5.7 25.6 24.6 14.0 7.6 27.4 11.6 64.0 18.5 67.0 18.5 81.0 20.0 17.1 9.2 30.7 4.3 128.8 27.2 124.8 27.0 A = aliphatic; 0 = olefinic; Ctot = total amount of reaction products. 50 - 40 - Si 30 - 20 - 50 - 40 - Si 30 - 20 - total AI content/mmoI g-' total A content/mmol g-' Fig. 7. Product selectivity S, after n-hexane cracking over acid-leached SP- and LP-mordenites. (a) SP-mordenites ; (b) LP-mordenites. where A(X) are the detected individual reaction products (wt %), A(X)coke are the undetected theoretical coke amounts (wt YO) and M(X) = molecular weight.Based on these relations the theoretical coke amount can be calculated: Acoke theoretical = A(C2)coke + A(Cl)coke' The practical amount of coke is defined as the difference between the total amount of C , hydrocarbons present in a blank run and the total amount of hydrocarbons analysed after reaction. The ratio of the theoretical to experimental coke formation is listed in table 4. The least active SP samples HMSPl and HMSP6, are characterized by a ratio close to one, indicating a strong C,, C , adsorption on the catalyst, as predicted by the proposed model. The highly active SP and LP samples, however, produce a ratio close to two, indicating that C , and C , fragments are now detected as condensation products, especially in isobutane. In the samples HMSPl and HMSP6, the amount of cracking products per unit cell is low because of the inaccessibility of the side-pockets and a small hydroxyl population, respectively.Therefore, condensation products with or between C,, C , fragments cannot be formed. In the highly active SP and LP samples, however, dehydrogenation results in a large number of highly active olefins which are able to condensate with preadsorbed C,, C2 carbenium ions. The relatively high amount of coke produced by sample HMLPl results in the ratioF. Goovaerts et al. 3685 Table 4. Coke formation after n-hexane crack- ing for different acid-leached SP and LP mor- denite samples co ke(t heor) HMSPl HMSP2 HMSP3 HMSP4 HMSPS HMSP6 HMLPl HMLP2 HMLP3 36 70 79 80 72 19 17 78 79 1.1 2.1 2.4 1.9 2.0 0.9 0.01 1.6 1.4 of 0.01 as shown in table 4.This observation is the result of highly active non-structural aluminium species on which a strong hydrocarbon adsorption can occur. As discussed before, these Lewis-acid sites are not detected by 27Al m.a.s.n.m.r. spectroscopy. References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 B. L. Meyers, T. H. Fleisch, G. J. Ray, J. T. Miller and J. B. Hall, J . Catal., 1988, 110, 82. H. K. Beyer, I. V. Miskin and F. Hange, React. Kinet. Catal. Lett., 1984, 26, 9. A. L. Klyachko, G. I. Kapustin, T. R. Brueva and A. M. Rubinstein, Zeolites, 1987, 7 , 119. N. G. Keats, G. Curthoys, I . Catal., 1985, 96, 288. F. Raatz, E. Freund and C. Marcilly, J . Chem. SOC., Faraduy Trans.I , 1983, 79, 2299. F. Raatz, C. Marcilly and E. Freund, Zeolites, 1985, 5, 329. D. Freude, E. Brunner, H. Pfeifer, D. Prager, H. H. Jerschkewitz, U. Lohse and G. Oehlmann, Chem. Phys. Lett., 1987, 139, 325. D. Freude, M. Hunger and H. Pfeifer, 2. Phys. Chem. Neue Fofge, 1987, 152, 171. D. Freude, M. Hunger, H. Pfeifer, Chem. Phys. Lett., 1986, 128, 62. P. Bodart, J. B. Nagy, G. Debras, Z. Gabelica and P. A. Jacobs, J. Phys. Chem., 1986, 90, 5183. P. Fejes, I. Hannus, I. Kiricsi, H. Pfeifer, D. Freude and W. Oehme, Zeolites, 1985, 5, 45. G. R. Hays, W. A. Van Erp, M. C. M. Alma, P. A. Couperus, R. Huis and A. E. Wilson, Zeolites, 1984, 4, 377. V. G. Stepanov, A. A. Shubin, K. G. Ione, V. M. Mastikhin and K. I. Zamaraev, Kinet. Katal., 1984, 25, 1043. I. Wang, T. J. Chen, K. J. Chao and T. C. Tsai, J. Catal., 1979, 60, 140. W. L. Kranich, Y. Ma, L. B. Sand, A. H. Weiss and I. Zweibel, International Conference on Molecular Sieve Zeolites, Worcester, Mass., 1970. P. E. Eberly and C . N. Kimberlin, Ind. Eng. Chem. Prod. Res. Dev., 1970, 9, 335. M. M. Dubinin, G. M. Federova, G. M. Plavnik, L. I. Piguzova and E. N. Prokofeva, Izu. Akad. Nuuk SSSR, Ser. Khim, 1968, 11, 2429. S. W. Weller and J. M. Bauer, AlChE 62nd Annual Meeting, Washington D.C., Nov. 16-20th, 1969. J. R. Hooper, A. Voorhies Jr, Ind. Eng. Chem. Prod. Res. Dev., 1972, 11, 294. K. V. Topohieva, B. V. Romanovsky, L. I. Piguzova, Ho Si Thoang and Y. W. Bizreh, Proc. 4th Int. Cong. Cataf., Moscow, 1968, ed. B. A. Kazansky (Adler, New York, 1968). K. H . Minachev, Adv. Chem. Ser., 1971, 102, 441. F. Fajula and F. G. Fault, J . Cutal., 1981, 68, 291. J. Philippaerts, E. F. Vansant, G. Peeters and E. Vanderheyden, Anal. Chim. Acta, 1987, 195, 237. H. Karge and K. Klose, 2. Phys. Chem. N . F., 1973, 83, 100. H. Karge, Z . Phys. Chem. N . F., 1971, 76, 133. H. Karge, 2. Phys. Chem. N . F., 1980, 122, 103. G. T. Kerr, J . C a d . , 1969, 15, 200. J. Elsen, G. S. D. King, W. J. Mortier, J . Phys. Chem., 1987, 91, 5800. Paper 8/04837A; Received 7th December, 1988
ISSN:0300-9599
DOI:10.1039/F19898503675
出版商:RSC
年代:1989
数据来源: RSC
|
|