摘要:

Contents 3663 3669 3675 3683 3693 370 1 3709 3717 3725 3737 Normal and Abnormal Electron Spin Resonance Spectra of Low-spin Cobalt(r1) IN,]-Macrocyclic Complexes. A Means of Breaking the Co-C Bond in B12 Co-enzyme M. Green, J. Daniels and L. M. Engelhardt The Interaction between Superoxide Dismutase and Doxorubicin. An Electron Spin Resonance Approach V. Malatesta, F. Morazzoni, L. Pellicciari-Bollini and R. Scotti Biomolecular Dynamics and Electron Spin Resonance Spectra of Copper Complexes of Antitumour Agents in Solution. Part 2.-Rifamycins R. Basosi, R. Pogni, E. Tiezzi, W. E. Antholine and L. C. Moscinsky An Electron Spin Resonance Investigation of the Nature of the Complexes formed between Copper(I1) and Glycylhistidine D. B. McPhail and B. A. Goodman A Vibronic Coupling Approach for the Interpretation of the g-Value Temperature Dependence in Type-I Copper Proteins M.Bacci and S. Cannistr aro The Electron Spin Resonance Spectrum of Al[C,H,] in Hydrocarbon Matrices J. A. Howard, B. Mile, J. S. Tse and H. Morris N; and (CN); Spin-Lattice Relaxation in KCN Crystals H. J. Kalinowski and L. C. Scavarda do Carmo Single-crystal Proton ENDOR of the SO, Centre in y-Irradiated Sulphamic Acid N. M. Atherton, C. Oliva, E. J. Oliver and D. M. Wylie Single-crystal Electron Spin Resonance Studies on Radiation-produced Species in Ice 1,. Part 1.-The 0- Radicals Single-crystal Electron Spin Resonance Studies on Radiation-produced Species in Ice I,. Part 2.-The HO, Radicals J. Bednarek and A. Plonka J. Bednarek and A. PlonkaContents 3663 3669 3675 3683 3693 370 1 3709 3717 3725 3737 Normal and Abnormal Electron Spin Resonance Spectra of Low-spin Cobalt(r1) IN,]-Macrocyclic Complexes.A Means of Breaking the Co-C Bond in B12 Co-enzyme M. Green, J. Daniels and L. M. Engelhardt The Interaction between Superoxide Dismutase and Doxorubicin. An Electron Spin Resonance Approach V. Malatesta, F. Morazzoni, L. Pellicciari-Bollini and R. Scotti Biomolecular Dynamics and Electron Spin Resonance Spectra of Copper Complexes of Antitumour Agents in Solution. Part 2.-Rifamycins R. Basosi, R. Pogni, E. Tiezzi, W. E. Antholine and L. C. Moscinsky An Electron Spin Resonance Investigation of the Nature of the Complexes formed between Copper(I1) and Glycylhistidine D. B. McPhail and B. A. Goodman A Vibronic Coupling Approach for the Interpretation of the g-Value Temperature Dependence in Type-I Copper Proteins M. Bacci and S. Cannistr aro The Electron Spin Resonance Spectrum of Al[C,H,] in Hydrocarbon Matrices J. A. Howard, B. Mile, J. S. Tse and H. Morris N; and (CN); Spin-Lattice Relaxation in KCN Crystals H. J. Kalinowski and L. C. Scavarda do Carmo Single-crystal Proton ENDOR of the SO, Centre in y-Irradiated Sulphamic Acid N. M. Atherton, C. Oliva, E. J. Oliver and D. M. Wylie Single-crystal Electron Spin Resonance Studies on Radiation-produced Species in Ice 1,. Part 1.-The 0- Radicals Single-crystal Electron Spin Resonance Studies on Radiation-produced Species in Ice I,. Part 2.-The HO, Radicals J. Bednarek and A. Plonka J. Bednarek and A. Plonka

ISSN:0300-9599    

DOI:10.1039/F198783FX013

出版商:RSC    

年代:1987

数据来源: RSC

ISSN:0300-9599    

DOI:10.1039/F198783BX015

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

ISSN 0300-9599 JCFTAR 83(4) 943-1 346 (1 987) 943 957 967 985 1007 1029 1041 I055 1063 1081 1089 1101 1109 1119 1127 1137 1149 32 JOURNAL OF THE CHEMICAL SOCIETY Faraday Transactions I Physical Chemistry in Condensed Phases CONTENTS Dielectric Behaviour of Adsorbed Water. Part 1 .-Measurement at Room Temperature on TiO, Dielectric Behaviour of Adsorbed Water. Part 2.-Measurement at Low Temperatures on TiO, T. Iwaki and T. Morimoto A Phenomenological Approach to Micellisation Kinetics D. G. Hall The Kinetics of Solubilisate Exchange between Water Droplets of a Water-in-oil Microemulsion P. D. I. Fletcher, A. M. Howe and B. H. Robinson Fluorescence Quenching as a Probe of Size Domains and Critical Fluctuations in Water-in-oil Microemulsions A. M. Howe, J. A. McEonald and B.H. Robinson Setschenow Coefficients for Caffeine, Theophylline and Theobromine in Aqueous Electrolyte Solutions P. Phrez-Tejeda, A. Maestre, M. Balcin, J. Hidalgo, M. A. Muiioz and M. Sanchez Electron Spin Resonance Studies of Free and Supported 12-Heteropoly Acids. Part 5.-Elucidation of Structures of Hydrated and of Dehydrated Heteropoly Acids by Quantum Chemical Calculations of Electron Spin Resonance Par- ameters F. Ritschl and R. Fricke Phosphorus-3 1 Nuclear Magnetic Resonance Studies of Solid Diphosphine Disulphides. Crystallographic Considerations R. K. Harris, L. H. Merwin and G. HHgele Experimental Evaluation of Adsorption Behaviour of Intermediates in Anodic Oxygen Evolution at Oxidized Nickel Surfaces B. E. Conway and T. Liu The Effect of Gasification by Air (623 K) or CO, (1098 K) in the Development of Microporosity in Activated Carbons J.Garrido, A. Linares-Solano, J. M. Martin-Martinez, M. Molina-Sabio, F. Rodriguez-Reinoso and R. Torregrosa Theoretical Analysis of Activation Parameters in Mixed Solvents Involving Various Chemical Equilibria. Reaction of Ethyl Iodide with Bromide Ion in N-Met h ylace tamide-Acet oni t rile and N-Me thy lace tamide-N,N-Dime thy lace ta- mide Mixtures Y. Kondo and S. Kusabayashi Light-induced Hydrogen Formation and Photo-uptake of Oxygen in Colloidal Suspensions of a-Fe,O, Formation of Carbocations from C, Compounds in Zeolites of Different Acidities H. Forster, J. Seebode, P. Fejes and I. Kiricsi Kinetics of the Solvolysis of the cis-Dichlorobis( 1,2-diarninoethane)cobalt(111) Ion in Water and in Water-t-Butyl Alcohol Mixtures G.S. Groves, A. F. M. Nazer and C. F. Wells Corrosion Processes in Quantized Semiconductor Colloids studied by Pulse Radiolysis M. T. Nenadovic, J. M. Nedeljkovic and 0. I. Micic Effects of the Oxidation State of Cerium on the Reactivity and Sulphur Resistance of Nio/Zeolite, Nio/Silica and Nio/CeO, Catalysts in Butane Hydro- genolysis I. Akalay, M-F. Guilleux, J-F. Ternere and D. Delafosse Fourier-transform Infrared Study of the Role of Zeolite Lewis Sites in Methanol Reactions over HZSM-5 Surfaces M. B. Sayed T. Morimoto and T. Iwaki J. Kiwi and M. Gratzel FAR 11159 1169 1179 1189 1193 1203 1213 1227 1237 1253 1261 1269 1281 1293 1307 1315 1323 Con tents Iron(I1) Cobalt Ferrites. Preparation and Interfacial Behaviour S.Ardizzone, A. Chittofrati and L. Formaro Electrogenerated R4N(Hg), Films E. Kariv-Miller, V. SvetlilSic' and P. B. Lawin Adsorption of Thiophene and Model Hydrocarbons on MoO,/y-Al,O, Cata- lysts studied by Gas-Solid Chromatography D. Vattis, H. Matralis and .4. Lycourghiotis Identification of the Structural Transition Temperature in the Solutions of Dialkyl Sulphoxides S. A. Markarian, K. R. Grigorian and L. K. Simonian Electron-transfer Reactions on CdSe Colloids as studied by Pulse Radiolysis N. M. Dimitrijevic' Properties of Capillary-condensed Benzene W. D. Machin and P. D. Golding Hydrogenation of Conjugated Dienes and Isomerization of Butenes on CdO, c0@4 and Cr,O, T. Suzuki, K. Tanaka, I. Toyoshima, T. Okuhara and K. Tanaka Structure of the Catalytic Site on a Silica-supported Catalyst derived from Copper(I1) Acetate M.Nomura, A. Kazusaka, N. Kakuta, Y. Ukisu and K. Miy ahara Ultraviolet-Visible Reflectance Studies of Hydrogen Adsorption and CO-H, Interaction at MgO and CaO Surfaces E. Garrone and F. S. Stone Free Energies, Enthalpies and Entropies of the Monothiocyanate Complex Formation of Trivalent Metals in Dimethyl Sulphoxide Solutions D. Puchal- ska, W. Grzybkowski and D. W6jcik Rotating-disc Electrodes. ECE and DISPl Processes R. G. Compton, R. G. Harland, P. R. Unwin and A. M. Waller Origin of Idealized Static Thermodynamic Forces inducing Solid Particle Motion, Orientation and Related Effects in a Solute Concentration Gradient E. A. Boucher Ionic Solvation in Water-Cosolvent Mixtures. Part 14.-Free Energies of Transfer of Single Ions from Water into Water-Ethylene Carbonate and Water-Propylene Carbonate Mixtures G. S. Groves, K. H. Halawani and C. F. Wells Reaction of Neopentane with Hydrogen over Pd, Pt, Ir and Rh Z. Karpinski, W. Juszczyk and J. Pielaszek Micellar Effects upon the Reactions of Complex Ions in Solution. Retardation of the Base Hydrolysis of cis-(Azido)(imidazole)bis(ethylenediamine)cobalt(nI) by Sodium Dodecyl Sulphate A. C. Dash and R. C. Nayak Oxidative Coupling of Methane over Samarium Oxides using N,O as the Oxidant K. Otsuka and T. Nakajima Tin Dioxide Gas Sensors. Part 1.-Aspects of the Surface Chemistry revealed by Electrical Conductance Variations J. F. McAleer, P. T. Moseley, J. 0. W. Norris and D. E. Williams

ISSN:0300-9599    

DOI:10.1039/F198783FP045

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

JOURNAL OF THE CHEMICAL SOCIETY Faraday Transactions II, Issue 4,1987 Molecular and Chemical Physics 579 587 60 1 607 619 627 639 647 655 663 675 683 For the benefit of readers of Faraday Transactions I, the contents list of Faraday Transactions II, Issue 4, is reproduced below. Kinetics and Mechanism of the Thermal Gas-phase Reaction between Bis(trifluoromethyl)trioxide, CF,O,CF,, and 1,l -Difluoroethylene, CF,CH, J. Czarnowski Theoretical Analysis of Conformational Transitions in Model Compounds of Natural Teichoic Acids M. Litowska and A. Tempczyk The Thermal Unimolecular Decomposition of 3-Methylcyclobutanone H. M. Frey, H. P. Watts and I. D. R. Stevens Reaction Probabilities and Cross-sections in the Reaction of Isotopically Pure Hydrogen Atoms and Propane J. E. Nicholas and G.Vaghjiani Dielectric Relaxation Spectroscopy of an Acetamide-Sodium Thiocyanate Eutectic Mixture A. Amico, G. Berchiesi, C. Cametti and A. Di Biasio Chemiluminescent Reactions of Fluorine Atoms with Organic Iodides in the Gas Phase. Part 1.-Iodomethanes H. S. Braynis, D. Raybone and J. C. Whitehead Chemiluminescent Reactions of Fluorine Atoms with Organic Iodides in the Gas Phase. Part 2.-Aliphatic and Aromatic Iodides H. S. Braynis, D. Raybone and J. C. Whitehead Vibrational Dephasing in Liquid 1,4-Dioxane M. F. R. M. Ferreira Marques and A. M. Amorim da Costa Ultrasonic Studies of Rotational Isomerism in 2,3-Dimethylbutane and 2,2,4- Trimethylpentane at Elevated Pressures R. Ledwig and A. Wiirflinger A Monte Carlo Study of the Effects of Temperature and of Wall Separation on the Structure and Properties of Water between Metal Walls N.G. Parsonage and D. Nicholson Collisionally Induced Rotational Energy Transfer within the A2A-State of CH R. N. Dixon, D. P. Newton and H. Rieley Potential-energy Functions for the Ground States of CO,, CS, and OCS, and Dynamical Calculations on the Reaction O(l0) + CS(lI:+) -+ S(l0) + CO(lC+) J. N. Murrell and H. Guo ' The following papers were accepted for publication in J . Chem. Soc., Faraday Trans. I during January 1987. 6/996 Non-ideal Behaviour of Benzene Solutions of Tri-N-Octylammonium Bromide and Cyclohexanol C. Klofutar and S. Paljk 6/ 1 102 Thermally activated Ruthenium Dioxide Hydrate: A Reproducible, Stable Oxygen Catalyst A. Mills, S. Giddings and I. Pate1 (1)6/ 1 375 The Mechanism of Hydrogenolysis and lsomerization of Oxacylcloalkanes on Metals. Part 8.-New Results on the Mechanism of Hydrogenolysis of Oxiranes on Platinum and Palladium F.Northeisz, A. G. Zsigmond, M. Bortok and G. V. Smith 6/ 1414 Primary Processes of Stabilizer Action in Radiation-induced Alkane Oxidation 0. Brede, R. Hermann and R. Mehnert 6/ 1543 Adsorption of Organic Molecules on a Titanium Dioxide (Rutile) Surface N. Suda and M. Nagao 6 / 1562 On the Vapour Pressure of Benzene. Part 1 .-An Assessment of Some Vapour Pressure Equations P. D. Golding and W. D. Machin 6/ 1563 On the Vapour Pressure of Benzene. Part 2.-Saturated Vapour Pressures from the Triple Point to 300 K 6/ 1780 Kinetic and Equilibrium Studies associated with the Agrregation of Non-ionic Surfactants in Non-polar Solvents P.Jones, E. Wyn-Jones and G. J. T. Tiddy 6/1819 Mechanism of Deuterium Addition and Exchange of Propene over Ni/SiO, and Pt/SiO, at Lower Temperatures S. Naito and M. Tanimoto 6/ 1828 Time-resolved Reflection Spectrum Study of Ordered Structure of Monodis- persed Polystyrene Spheres in an Electric Field and Role of Debye Screening Length T. Okubo 6/ 829 Ordered Solution Structure of a Monodispersed Polystyrene Sphere in the Presence of Neutral Polymers as studied by the Reflection Spectrum Method T. Qkubo 6/ 834 Solvent Effect on the Reactions of Coordination Complexes. Part 1.-Kinetics of Solvolysis of the Cis-bromobenzimidazole bis(ethylenediamine)cobalt(rrr) in Methanol-Water Media A. C. Dash and N. Dash 6/ 1849 Towards a Complete Configurational Theory of Non-equilibrium Polymer Adsorption 6/ 1880 Solvent and Substituent Effects on Intramolecular Charge Transfer of Selected Derivatives of 4-Trifluoromethyl-7-aminocoumarin C.Guo and Y. B. Feng 6/ 1920 The 1,2-Ethanedi01-2-Methoxyethanol Solvent System. The Dependence of the Dissociation Constant of Picric Acid on the Temperature and on the Compo- sition of the Solvent Mixture 6/ 1954 Coupled 13C Longitudinal and Transverse Magnetic Relaxation in Micellar and Non-micellar Sodium Octanoate F. Heatley 6/ 1955 Reactive Intermediates for the Ethene Homologation Reaction on Molybdena Silica Catalysts K. Tanaka and K-I. Tanaka 6/1967 Interaction of Enzymes with Surfactants in Aqueous Solution and in Water- in-oil Microemulsions B.H. Robinson and P. D. I. Fletcher 6/ 1974 EXAFS Study of the Reaction between Silica-supported Copper@) oxide Catalysts and Acetic Acid M. Nomura, A. Kaxusaka, Y. Ukisu and N. Kakuta 6/2049 The Study of Aluminium Deposition from THF Solutions of AlCl, + LiAlH, using Microelectrodes. Part 1.-1 : 1 AlCl, to LiAlH, J. N. Howarth and D. Pletcher 6/2050 The Study of Aluminium Deposition from THF Solutions of AlCl,+LiAlH, using Microelectrodes. Part 2.-The Influence of Solution Composition J. N. Howarth and D. Pletcher 6/2052 Complexing and Transfer Free Energies of Metal Ion Dibenzocryptates A. F. Danil de Namor, F. F. Salazar and P. Greenwood P. D. Golding and W. D. Machin W. Barford and R. C. Ball G. G. Franchini, L. Tassi and G. Tosi (ii)6/2082 Hemimicelle Formation of Cationic Surfactants at the Silica Gel/Water Inter- face Y.Gau, J. Du and T. Gu 6/2084 Electron Spin Resonance Studies of HPt(CN)L2 and Pti- formed by Irradiation of KPt(CN,) in Solvents J. L. Wyatt, M. C. R. Symons and A. Hasegawa 6/2155 Study of the Influence of the Impregnation Acidity on the Structure and Properties of Molybdena/Silica Solids H. M. Ismail, C. R. Theocharis and M. I. Zaki 6/22 15 The Hydration of Aliphatic Aldehydes in Aqueous Micellar Solutions V. R. Hanke, W. Knoche and E. Dutkiewicz 6/2313 Cluster Size Distribution in a Monte Carlo Simulation of the Micellar Phase of an Amphiphile and Solvent Mixture C. M. Care 6/2362 Surface Reactivity and Spectroscopy of Alkaline Earth Oxide Powders. Part 3.-H/D and l80/l6O Exchange on Specpure CaO J.Cunningham and C. P. Healy 6/2384 A Study of Proton Transfer extracted by 2,2’-Bipyridine from Water to Nitrobenzene using Chronopotentiometry with Cyclic Linear Current-scanning and Cyclic Voltammetry Y. Liu and E. Wang 6/2493 Heterogeneous Decomposition of Trichlorofluoromethane on Carbonaceous Surfaces A. J. Colussi and V. T. Amorebieta (iii)Agnel, J-P. L., 225 Akalay, I., 1137 Alberti, A., 91 Allen, G. C., 925 Anderson, A. B., 463 Anderson, J. B. F., 913 Antholine, W. E., 151 Ardizzone, S., 11 59 Atherton, N. M., 37, 941 Axelsen, V., 107 Balon, M., 1029 Barratt, M. D., 135 Barrer, R. M., 779 Basosi, R., 151 Bastl, Z., 51 1 Bateman, J. B., 841 Battesti, C. M., 225 Becker, K. A., 535 Berclaz, T., 401 Berleur, F., 177 Berry, F. J., 615 Bertagnolli, H., 687 Berthelot, J., 231 Beyer, H.K., 511 Bianconi, A., 289 Blandamer, M. J., 865, 559 Blyth, G., 751 Borbkly, G., 51 1 Boucher, E. A., 1269 Braquet, P., 177 Brazdil, J. F., 463 Briscoe, B. J., 938 Bruce, J. M., 85 Brustolon, M., 69 Budil, D. E., 13 Burch, R., 913 Burgess, J., 559, 865 Burke, L. D., 299 Busca, G., 853 Buscall, R., 873 Cairns, J. A., 913 Carley, A. F., 351 Cassidy, J. F., 231 Celalyan-Berthier, A., 401 Chalker, P. R., 351 Chandra, H., 759 Chieux, P., 687 Chittofrati, A., 11 59 Clark, B., 865 Clifford, A. A., 751 Colin, A. C., 819 Coller, B. A. W., 645, 657 Coluccia, S., 477 Compostizo, A., 819 Compton, R. G., 1261 Chu, D-Y., 635 Cumulative Author Index 1987 Conway, B. E., 1063 Corvaja, C., 57 Couillard, C., 125 Courbon, H., 697 Craven, J.B., 779 Crossland, W. A., 37 D’Alba, F., 267 Dash, A. C., 1307 Daverio, D., 705 Davoli, I., 289 Dawber, J. G., 771 De Doncker, J., 125 De Laet, M., 125 De Ranter, C. J., 257 Declerck, P. J., 257 Delafosse, D., 1137 Delahanty, J. N., 135 Di Lorenzo, S., 267 Diaz Pefia, M., 819 DimitrijeviC, N. M., 1193 Dodd, N. J. F., 85 Ducret, F., 141 Dudikova, L., 51 1 Dusaucy, A-C., 125 Elbing, E., 657, 645 Empis, J. M. A., 43 Endoh, A., 41 1 Engberts, J. B. F. N., 865 Evans, J. C., 43, 135 Fan, G., 323 Fatome, M., 177 Fejes, P., 1109 Fletcher, P. D. I., 985 Flint, N. J., 167 Fonnaro, L., 1159 Fonnosinho, S. J., 431 Forrester, A. R., 21 1 Forster, H., 1109 Fraissard, J., 45 1 Fricke, R., 1041 Fujii, K., 675 Galli, P., 853 Garrido, J., 1081 Garrone, E., 1237 Geoffroy, M., 401 Gervasini, A., 705 Gilbert, B.C., 77 Golding, P. D., 1203 Gottschalk, F., 571 Gozzi, D., 289 Grampp, G., 161 Gratzel, M., 1101 Gray, P., 751 Greci, L., 69 Grieser, F., 591 Grigorian, K. R., 1189 Grossi, L., 77 Groves, G. S., 1281, 11 19 Grzybkowski, W., 281, 1253 Guardado, P., 559 Guilleux, M-F., 1 137 Hagele, G., 1055 Hakin, A. W., 559, 865 Halawani, K. H., 1281 Hall, D. G., 967 Hall, M. V. M., 571 Halpern, A., 219 Hamada, K., 527 Harland, R. G., 1261 Harrer, W., 161 Harris, R. K., 1055 Hartland, G. V., 591 Hasegawa, A., 759 Hatayama, F., 675 Heatley, F., 517 Hemminga, M. A., 203 Herold, B. J., 43 Hertz, H. G., 687 Hidalgo, J., 1029 Higgins, J. S., 939 Holden, J. G., 615 Howe, A. M., 985, 1007 Howe, R. F., 813 Hudson, A., 91 Hunter, R., 571 Hutchings, G.J., 571 Imamura, H., 743 Imanaka, T., 665 Ito, T., 451 Iwaki, T., 943, 957 Jackson, S. D., 905 Jaenicke, W., 161 Janes, R.. 383 Juszczyk, W., 1293 Kakuta, N., 1227 Kanno, T., 721 Kariv-Miller, E., 1 169 Karpinski, Z., 1293 Kazusaka, A,, 1227 Kerr, C W., 85 Kiricsi, I., 1109 Kiwi, J., 1101 Kobayashi, M., 721 Koda, S., 527 Kondo, Y., 1089 Konishi, Y., 721 Kordulis, C., 627 Korth, H-G., 95 Kowalak, S., 535 Kubelkova, L., 51 1 Kclxkawa, Y., 675 Kusabayashi, S., 1089AUTHOR INDEX La Ginestra, A., 853 Lambelet, P., 141 Lavagnino, S., 477 Lawin, P. B., 1169 Leaist, D. G., 829 Lecomte, C., 177 Lin, C. P., 13 Linares-Solano, A., 1081 Lindgren, M., 893 Liu, R-L., 635 Liu, T., 1063 Loliger, J., 141 Lorenzelli, V., 853 Loretto, M. H., 615 Lund, A., 893 Lycourghiotis, A., 627, 1179 Lyons, C.J., 645 Lyons, M. E. G., 299 McAleer, J. F., 1323 McDonald, J. A., 1007 Machin, W. D., 1203 Maestre, A., 1029 Maezawa, A., 665 Makela, R., 51 Maniero, A. L., 69, 57 Marchese, L., 477 Marcus, Y., 339 Mari, C. M., 705 Markarian, S. A., 1189 Martin-Martinez, J. M., 1081 Masiakowski, J. T., 893 Masliyah, J. H., 547 Matralis, H., 1179 Matsuura, H., 789 McCarthy, S. J., 657 McLauchlan, K. A., 29 Mehandru, S. P., 463 Merwin, L. H., 1055 Micic, 0. I., 1127 Miyahara, K., 1227 Miyata, H., 675 Molina-Sabio, M., 1081 Monk, C. B., 425 Morazzoni, F., 705 Morimoto, T., 943, 957 Moseley, P. T., 1323 Moyes, R. B., 905 Mozzanega, M-N., 697 Muiioz, M. A., 1029 Nair, V., 487 Nakajima, T., 1315 Narayanan, S., 733 Narducci, D., 705 Nayak, R. C., 1307 Nazer, A. F.M., 11 19 Nedeljkovic, J. M., I127 Nenadovic, M. T., 1127 Nomura, H., 527 Nomura, M., 1227 Norris, J. 0. W., 1323 Norris, J. R., 13 Nukui, K., 743 Nuttall, S., 559 O’Brien, A. B., 371 Ohno, T., 675 Ohshima, K., 789 Okabayashi, H., 789 Okamoto, Y., 665 Okuhara, T., 1213 Ono, T., 675 Otsuka, K., 1315 Pallas, N. R., 585 Parry, D. J., 77 Patrono, P., 853 Pedersen, J. A., 107 Pedulli, G. F., 91 Pirez-Tejeda, P., 1029 Pethica, B. A., 585 Pethrick, R. A., 938 Pichat, P., 697 Pielaszek, J., 1293 Pilarczyk, M., 281 Pizzini, S., 705 Pogni, R., 151 Pomonis, P., 627 Priolisi O., 57 Puchalska, D., 1253 Purushotham, V., 21 1 Radulovic, S., 559 Raffi, J. J., 225 Ritschl, F., 1041 Riviere, J. C., 351 Roberts, M. W., 351 Robinson, B. H., 985, 1007 Rodriguez-Reinoso, F., 1081 Roman, V., 177 Romiio, M.J., 43 Rosseinsky, D. R., 245, 231 Rowlands, C. C., 43, 135 Rubio, R. G., 819 Sakai, T., 743 Sanchez, M., 1029 Sangster, D. F., 657 Saraby-Reintjes, A., 271 Saucy, F., 141 Savoy, M-C., 141 Sayed, M. B., 1149 Seebode, J., 1109 Segal, M. G., 371 Segre, U., 69 Seyedmonir, S., 813 Sidahmed, I. M., 439 Simonian, L. K., 1189 Steenken, S., 113 Stevens, D. G., 29 Stone, F. S., 1237 Suppan, P., 495 Sustmann, R., 95 Suzuki, T., 1213 Svetlieid, V., 1169 Swartz, H. M., 191 Symons, M. C. R., 1, 383, 759 Szostak, R., 487 Tabner, B. J., 167 Taga, K., 789 Takaishi, T., 41 1 Tan, W. K., 645 Tanaka, K., 1213, 1213 Tempkre, J-F., 1137 Tempest, P. A., 925 ThiCry, C. L., 225 Thomas, T. L., 487 Thurai, M., 841 Tilquin, B., 125 Tomellini, M., 289 Tonge, J. S., 231, 245 Torregrosa, R., 1081 Toyoshima, I., 1213 Trabalzini, L., 15 1 Tsuchiya, S., 743 Tsukamoto, K., 789 Turner, J.C. R., 937 Tyler, J. W., 925 Ukisu, Y., 1227 Uma, K., 733 Unwin, P. R., 1261 Vachon, A., 177 van de Ven, T. G. M., 547 Vattis, D., 1179 Vincent, P. B., 225 Vink, H., 801, 941 Vordonis, L., 627 Vuolle, M., 51 Waddicor, J. I., 751 Waller, A. M., 1261 Wells, C. F., 439, 939, 11 19, Wells, P. B., 905 White, L. R., 591, 873 Whyman, R., 905 Williams, D. E., 1323 Williams, J. O., 323 Williams, W. J., 371 Wilson, I. R., 645, 657 Wbjcik, D., 1253 Yamada, K., 743 Zhang, Q., 635 1281THE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY GENERAL DISCUSSION N o . 84 Dynamics of Elementary Gas-phase Reactions University of Birmingham, 14-16 September 1987 Org anising Committee : Professor R.Grice (Chairman) Dr M. S. Child DrJ. N. L. Connor Dr M. J. Pilling Professor I. W. M. Smith Professor J. P. Simons The Discussion will focus on the development of experimental and theoretical approaches to the detailed description of elementary gas-phase reaction dynamics. Studies of reactions at high collision energy, state-to-state kinetics, non-adiabatic processes and thermal energy reactions will be included. Emphasis will be placed on systems exhibiting kinetic and dynamical behaviour which can be related to the structure of the reaction potential- energy surface or surfaces. The preliminary programme 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 SYMPOSIUM N o .23 Molecular Vibrations University of Reading, 15-16 December 1987 Organising Committee: Professor I. M. Mills (Chairman) Dr J. E. Baggott Professor A. D. Buckingham Dr M. S. Child Dr N. C. Handy Dr B. J. Howard The Symposium will focus on recent advances in our understanding of the vibrations of polyatomic molecules. The topics to be discussed will include force field determinations by both ab initio and experimental methods, anharmonic effects in overtone spectroscopy, local modes and anharmonic resonances, intramolecular vibrational relaxation, and the frontier with molecular dynamics and reaction kinetics. The preliminary programme may be obtained from: Mrs. Y. A. Fish, The Royal Society of Chemistry, Burlington House, London W1V OBNTHE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY GENERAL DISCUSSION N o .85 Solvat ion University of Durham, 28-30 March 1988 Organising Committee: Professor M. C. R. Symons (Chairman) Professor J. S. Rowlinson Professor A. K. Covington Dr I. R. McDonald The purpose of the Discussion is to compare solvation of ionicand non-ionicspecies in the gas phase and in matrices with corresponding solvation in the bulk liquid phase. The aim will be to confront theory with experiment and to consider the application of these concepts to relaxation and solvolytic processes. Contributions for consideration by the organising Committee are invited in the following areas: (a) Gas phase non-ionic clusters (b) Liquid phase non-ionic clusters (c) Gas phase ionic clusters (d) Liquid phase ionic solutions (e) Dynamic processes including solvolysis Abstracts of about 300 words should be sent by 31 May 1987 to: Professor M.C. R. Symons, Department of Chemistry, The University, Leicester LE17RH. Dr J. Yarwood Dr A. D. Pethybridge Professor W. A. P. Luck Dr D. A. Young ~~~ ~ ~ THE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY GENERAL DISCUSSION No. 86 Spectroscopy at Low Temperatures University of Exeter, 13-15 September 1988 Organising Committee: Professor A. C. Legon (Chairman) Dr P. 6. Davies Dr B. J. Howard Dr P. R. R. Langridge-Smith Dr R. N. Perutz Dr M. Poliakoff The Discussion will focus on recent developments in spectroscopy of transient species (ions, radicals, clusters and complexes) in matrices or free jet expansions.The aim of the meeting is to bring together scientists interested in similar problems but viewed from the perspective of different environments. Contributions for consideration by the Organising Committee are invited. Titles should be submitted as soon as possible and abstracts of about 300 words by30 September 1987 to: Professor A. C. Legon, Department of Chemistry, University of Exeter, Exeter EX4 4QD. Full papers for publication in the Discussion volume will be required by May 1988. (vii)JOURNAL OF CHEMICAL RESEARCH Papers dealing with physical chemistry/chemical physics which have appeared recently in J. Chem. Research, The Royal Society of Chemistry's synopsis + microform journal, include the following: Is the PhCH2(N20)' Cation likely to be a Reaction Intermediate? A Theoretical Study Peter Conformational Structure of Bipyridine Radical Cations Hans-Jorg Hofmann, Renzo Solid-Liquid Equilibria in the Quarternary System Monomethylhydrazine - Sodium Chloride Alexandrina Salas-Padron and Marie-Therese Helge M.W. Gill, Howard Maskill, Dieter Poppinger, and Leo Radom (1 987, Issue 2) Cimiraglia, and Jacopi Tomasi (1 987, Issue 2) - Sodium Hydroxide - Water at 298.1 K Saugier Cohen-Adad (1 987, Issue 2) Egsgaard and Lars Carlsen (1 987, Issue 1 ) Gas-phase Pyrolysis of Nitroethene. Surface-promoted Formation of Nitrosoethene Unified Theory of Metal-ion-complex Formation Constants Paul L. Brown and Ronald N. Sylva (1 987, Issue 1 ) Empirical Relationships for Estimation of Enthalpies of Formation of Simple Hydrates, Part 1.Phillippe Hydrates of Alkali-metal Cations, of Hydrogen, and of Monovalent Cations Vieillard and H. Donald B. Jenkins (1 986, Issue 12) Empirical Relationships for Estimation of Enthalpies of Formation of Simple Hydrates, Part 2. Phillippe Vieillard and H. Donald B. Jenkins Hydrates of Alkaline-earth-metal Cations (1 986, Issue 12) Empirical Relationships for Estimation of Enthalpies of Formation of Simple Hydrates, Part 3. Hydrates of Transition Metal Cations (Cr2+, Fe2+, Mn2+, Co2+, Ni2+, Cu2+, Zn2+, Cd2+) and of U02+ Phillippe Vieillard and H. Donald B. Jenkins (1 986, Issue 12) (viii)FARADAY DIVISION INFORMAL AND GROUP MEETINGS Division Full-day Endowed Lecture Symposium on Intramolecular Dynamics and Chemical Reactivity including the Centenary Lecture by S.A. Rice and the Tilden Lecture by M. S. Child To be held at the Scientific Societies Lecture Theatre, London on 6 May 1987 Further information from Mrs Y. A. Fish, The Royal Society of Chemistry, Burlington House, London W1V OBN Polymer Physics Group Electroactive Polymers To be held a t the Geological Society, London on 14 May 1987 Further information from Dr G. C. Stevens, CERL, Kelvin Avenue, Leatherhead KT22 7SE Electrochemistry Group with the Electrochemical Technology Group of the SCI Graduate Students’ Meeting To be held at Imperial College, London on 10 June 1987 Further informationfrom: Dr G. Kelsall, Department of Mineral Resources Engineering, Imperial College, London SW7 2AZ Electrochemistry Group with Macro Group UK Polymer Electrolytes To be held at the University of St.Andrews on 18-19 June 1987 Further information from Dr C. A. Vincent or Dr J. R. MacCallum, Department of Chemistry, University of St. Andrews, St. Andrews KY16 9ST Gas Kinetics Group Thermally and Photochemically Activated Reactions To be held at the University of Edinburgh on 9-10 July 1987 Further information from Professor R. Donovan, Department of Chemistry, University of Edinburgh, West Mains Road, Edinburgh EH9 3JJ Division Xlth International Symposium on Molecular Beams To be held at the University of Edinburgh on 13-17 July 1987 Further information from Dr J. F. Gibson, The Royal Society of Chemistry, Burlington House, London WlVOBN Industrial Physical Chemistry Group The Physical Chemistry of Small Carbohydrates (as part of the International Symposium on Solute-Solute-Solvent Interactions) To be held at the University of Regensburg, West Germany on 10-14August 1987 Further information from Dr F.Franks, Pafra Ltd, 150 Science Park, Milton Road, Cambridge CB4 4GG Industrial Physical Chemistry Group The Interaction of Biologically Active Molecules and Membranes To be held a t Girton College, Cambridge on 8-10 September 1987 Further information from Dr T. G. Ryan, ICI New Science Group, PO Box 11, The Heath, Runcorn WA7 4QE Polymer Ph ysics Group Biennial Meeting To be held at University of Reading on 9-1 1 September 1987 Further information from Dr D. Bassett, Department of Physics, University of Reading, Reading RG7 2AD Neutron Scattering Group Applications of Neutron and X-Ray Optics To be held at the University of Oxford on 14-15 September 1987 Further information from Dr R.K. Thomas, Physical Chemistry Laboratory, South Parks Road, Oxford OX1 302Surface Reactivity and Catalysis Group New Methods of Catalyst Preparation and Characterization To be held at Brunei University on 14-16 September 1987 Further information from: Dr M. Bowker, ICI New Science Group, PO Box 11, The Heath, Runcorn, Cheshire WA7 4QE Colloid and Interface Science Group Polydispersity in Colloid Science To be held at the University of Nottingham on 15-16 September 1987 Further information from Dr. R. Buscall, ICI plc, Corporate Colloid Science Group, PO Box 11, The Heath, Runcorn WA7 4QE Polymer Physics Group New Materials To be held at the University of Warwick on 22-25 September 1987 Further information from Dr M. J. Richardson, Division of Materials Applications, National Physical Laboratory, Queens Road, Teddington, Middlesex TW11 OLW Division Autumn Meeting Spectroscopy of Gas-phase Molecular Ions and Clusters To be held at the University of Nottingham on 22-24 September 1987 Further information from Professor J. P. Simons, Department of Chemistry, University of Nottingham, Nottingham NG7 2RD Polymer Physics Group with the Institute of Marine Engineers Polymers in a Marine Environment To be held in London on 14-16 October 1987 Dr G. J. Lake, MRPRA, Brickendonbury, Herts SG13 8NL Electrochemistry Group with the SCI Electrosynthesis To be held at the University of York on 15-17 December 1987 Further information from Dr G. Kelsall, Department of Mineral Resources Engineering, Imperial College, London SW7 2AZ Neutron Scattering Group Scattering from Disordered Systems To be held a t the University of Bristol on 16-18 December 1987 Further information from: Dr R. J. Newport, Physics Laboratory, The University, Canterbury, Kent CT2 7NR

ISSN:0300-9599    

DOI:10.1039/F198783BP047

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. 1, 1387,83,943-956 Dielectric Behaviour of Adsorbed Water Part 1.-Measurement at Room Temperature on TiO, Tetsuo Morimoto" Department of Chemistry, Faculty of Science, Okayama University, Tsushima, Okayama 700, Japan Tohru Iwakif Department of Chemistry, Faculty of Science, Hiroshima University, Naka-ku, Hiroshima 730, Japan The dielectric behaviour in the Ti0,-H,O adsorption system has been investigated at frequencies from 0.1 Hz to 5 MHz at room temperature. A large dielectric dispersion has been found from 0.1 to lo4 Hz, which shifts to a higher frequency region when the coverage of physisorbed H,O increases. A simultaneous increase in conductance has been observed when the coverage increases. Thus, the dielectric relaxation observed at room temperature has been interpreted in terms of the interfacial polarization, which obeys the two-layer model and increases with increasing conductance. A mechanism is postulated for an enhancement of the conductance which takes place when H,O is adsorbed on TiO,.Recently, the properties of chemisorbed H,O, i.e. of surface hydroxyls, on metal oxides have been investigated as distinguished from those of physisorbed H20,1-4 and it has been clarified that the surface hydroxyls have significant effects on the surface properties of the s o l i d ~ . ~ - l ~ Dielectric studies provide important information on the structure of adsorbates and the strength of their binding on a solid surface.ll9 l2 A number of investigators, using this technique, have dealt with different kinds of adsorbed H,Q, such as physisorbed H,O on solid s ~ r f a c e s , ~ ~ - l ~ capillary-condensed H,O in pores17-21 and interlayer H,O in crystalline minerals.,, The same technique was applied to the a-Fe,O,-H,O adsorption system at room temperature by McCafferty et u/.239 24 They observed a large dielectric dispersion in the low-frequency region and interpreted the phenomenon in terms of a Debye-type relaxation.Kurmakil7 also reported three kinds of relaxation for the adsorbed H,O on silica from dielectric measurements at 298 K. On the other hand, Kondo et al.25 observed three relaxation processes attributable to adsorbed H,O and surface hydroxyls on y-Al,03, from their measurements at temperatures from 77 to 390 K. The purpose of the present investigation is to clarify the dielectric behaviour of the Ti0,-H,O adsorption system and the effect of surface hydroxyls on the dielectric properties of the system.With TiO,, the properties of surface hydroxyls have been established by various techniques such as the adsorption isotherm,l? heat of immersion4~ 2 6 ~ 27 and i.r. spectra.' However, no investigation has been published on the dielectric properties of adsorbed H,O on TiO,, despite the practical importance of the substance not only for pigments but also for photocatalysts in H,O t Present address: Hiroshima Technical Institute, Mitsubishi Heavy Industry, Hiroshima 733, Japan. 943 32-2944 Dielectric Behaviour of Adsorbed Water Experiment a1 Materials The sample used in this study was TiO, (rutile) produced by hydrolysing titanium sulphate (Teikoku Koko Co.).The sample was washed repeatedly with 2 mol dm-, aqueous ammonia, with 2 mol dm-3 HNO, and thoroughly with H,O; finally it was electrodialysed until the electric conductance of the supernatant liquid approached that of H,O. The sample thus purified was outgassed in a vacuum of Torrt for 4 h at 873 K, treated in 100 Torr 0, for 1 h at 723 K and the temperature was lowered to room temperature in the 0, atmosphere. After outgassing, the sample was exposed to saturated H,O vapour for 12 h at room temperature to accomplish surface hydration and was then outgassed at room temperature to remove physisorbed H,O. The specific surface area of the sample was measured by the N, adsorption method and was found to be 8.35 m2 g-l.H,O used for the adsorption measurements was purified by freeze-thaw cycles. Procedure The dielectric cell consisted of two concentric stainless steel cylinders 100 mm in length, 0.5 mm in thickness, 20 mm in outer diameter and 2 mm apart. In order to facilitate attainment of the adsorption equilibrium, the electrodes were perforated, the pores being 1.0 mm in diameter. Two kinds of electrodes were employed, viz. blocking and non-blocking ones. The former was covered with a Teflon film of 30 pm thickness and the capacitance of the film was 1600 pF. The capacitance of the cell with non-blocking electrodes was measured at 298 K using cyclohexane, dichloroethane and nitrobenzene as the standard substances and was found to be 24.60 pF; the stray capacitance was 2.40 pF.The cell was packed with the sample with a packing density of 28% and it was connected to a conventional volumetric adsorption apparatus equipped with an oil manometer. The sample was packed into the cell in a dry box filled with N, to av@d contact with the atmosphere. After the sample was fully hydroxylated and outgassed at room temperature, the adsorption isotherm of H,O was measured at temperatures of 278,288,298 and 308 K, where the attainment of the adsorption equilibrium was confirmed by following the change in capacitance. It took more than 24 h for the accomplishment of equilibrium in the initial pressure range. The dielectric permittivity E’ and dielectric loss E” were measured simultaneously over the frequency region from 0.1 Hz to 5 MHz with impedance bridges (TRlOC and TR4) made by Ando Electric Co.Alternatively, the sample was outgassed at higher temperatures (373, 473 and 773 K) to remove surface hydroxyls successively and the dielectric measurements were made after cooling to room temperature. The amount of hydroxyls remaining on the surface after outgassing at higher temperatures was determined by means of the successive ignition loss method.3* Results Dielectric Measurement with Non-blocking Electrodes The adsorption isotherms of H,O on TiO, measured at various temperatures are illustrated in fig. 1. By applying the B.E.T. method to the isotherm the monolayer volume of H,O can be calculated to be 0.183f0.002 cm3 (s.t.p.) m-,, i.e. 4.92 molecule nrn-,. If we assume that the crystal planes (1 lo), (100) and (1 11) of rutile are equally exposed on the sample, the surface density of Ti atoms can be computed to be 10.36 atom nm-2.29* 30 Therefore, the ratio of the number of H,O molecules in the first physisorption t 1 Torr = 101 325/760 Pa.T.Morimoto and T. Iwaki 945 0.7 0.6 0.5 N E 0.4 a n -. W m 5 0.3 2 0.2 0.1 c I 1 I I 1 4 8 12 16 20 PIP, Fig. 1. Adsorption isotherms of H,O on TiO, at 278, 288, 298 and 308 K. Dotted lines indicate the monolayer volume obtained by the B.E.T. method. layer to that of underlying hydroxyls is found to be ca. 1 : 2, in accordance with the result reported previously.2 In fig. 2 the dielectric permittivity E’ of H,O adsorbed on TiO, measured with non-blocking electrodes at 298 K is plotted against the frequency. After the sample was outgassed at 773 K the number of remaining surface hydroxyls was found to be 0.15 OH groups nm-2 .2 Dielectric dispersion is hardly observed on this sample, where the E’ values of ca.5.0 are observed. On the fully hydroxylated sample with 10.4 OH groups nm-2 and without physisorbed H20 E’ increases slightly at frequencies below 10 Hz, while it remains almost constant at frequencies higher than 10 Hz. However, E’ increases markedly with increasing coverage 8 of physisorbed H20, especially at low frequencies, and thus a large dielectric dispersion appears, e.g. the E’ value measured at 30 Hz on the sample with 8 = 2.26 is ca. 100 times larger than that on the sample having only surface hydroxyls. Fig. 3 shows the dielectric loss E” of H 2 0 adsorbed on TiO,, measured under the same conditions as those for E’ in fig.2. The E” value in fig. 3 indicates the same tendency as that of the E’ value at the point that the measured values of E’ and E” increase in the logarithmic scale with increasing 8, particularly at low frequencies. However, the En vs. fcurve does not reveal a straight line on the surface with only chemisorbed H20, but it gives a maximum near lo5 - lo6 Hz. When a small amount of H 2 0 is physisorbed, not only does the E” value increase markedly, but also a maximum appears in the curve at a low frequency near 1 Hz. Further increase in 8 gives rise to the enhancement of the E” value and the extinction of the maximum at 105-106 Hz.946 I Dielectric Behaviour of Adsorbed Water f/Hz Fig. 2, Dielectric permittivity for various 8 of adsorbed H,O on TiO,, measured at different frequencies with non-blocking electrodes at 298 K.The number in the figure indicates that of samples with different 0 as shown in table 1. 1 0' 10' €" 1c 1 1 0 Fig. 3. Dielectric loss for various 0 of adsorbed H,O on TiO,, measured at different frequencies with non-blocking electrodes at 298 K. The number in the figure indicates that of samples with different 0 as shown in table 1.T. Morimoto and T. Iwaki 947 Table 1. Coverage 8 of adsorbed H,O on TiO, at 298 K no. PI& 8 1 2 3 4 5 6 7 8 9 10 1 1 12 13 14 15 16 0 0 0.001 0.003 0.022 0.058 0.094 0.156 0.235 0.336 0.430 0.533 0.648 0.750 0.839 0.905 OU Ob 0.104 0.205 0.44 1 0.623 0.749 0.940 1.130 1.369 1.604 1.884 2.260 2.787 3.525 4.678 a The fully hydroxylated sample was evacuated at 773 K for 4 h, treated in 100 Torr 0, at 773 K for 1 h, cooled to 298 K and re-evacuated at 298 K; the sample has 0.15 OH nm-2.Fully hydroxylated sample, having 10.4 OH nrn-,. The dielectric isotherm of adsorbed H,O on TiO, at 298 K is illustrated in fig. 4, being replotted from fig. 2. It is seen from fig. 4 that E’ at low frequency, e.g. at 100 Hz, increases appreciably when a small amount of H,O is physisorbed. At higher frequencies, however, the increment of E’ due to the increase in 0 becomes small, e.g. at 100 kHz, the E’ value remains almost unchanged until PIP, = 0.9, after which it increases steeply. Although the data are not shown here, it has been found that when the temperature is raised, every dielectric isotherm has a similar curve shape and the absolute value of E’ increases.McCafferty et aZ.23924 have observed that E’ of H,O on a-Fe,03 rises sharply at the beginning of the second layer, while that on TiO, reveals a sharp rise at the beginning of the first layer. The electrical conductance G of adsorbed H,O on TiO, measured with non-blocking electrodes at 298 K is plotted against the relative pressure of H,O in fig. 5. Here, DC indicates the direct current conductance measured just after the application of the direct current voltage. Fig. 5 shows that G increases with increasing 0 of H,O as well as with increasing frequency, where the conductance isotherm expressed by the logarithmic scale reveals the same tendency as that of the adsorption isotherm and also of the dielectric isotherm at low frequencies.When the frequency becomes higher G increases in contrast to the decrease in E ’ . In addition, it is seen from fig. 5 that the d.c. conductance determines the whole shape of the conductance isotherms, especially at frequencies lower than lo4 Hz and that G reveals a large frequency dependence in the low-frequency region at 0 < 0.4 and it becomes weak at 0 > 1. A similar phenomenon has been observed with organic ~ e m i ~ ~ n d ~ ~ t o r ~ . ~ ~ The conductivity 0 of the samples can be computed by multiplying the cell constant 4.49 x lom3 to the observed G value; the result of the calculation shows the sample to be an insulator. However, in the presence of small amount of adsorbed H,O, the d.c.conductivity amounts to a value corresponding to a semiconductor, 1 OW5- 1 0-1 S m-’ .32948 Dielectric Behaviour of Adsorbed Water 1 1 I I 0 0.2 0.4 0.6 0.8 PIP, 0 Fig. 4. Dielectric isotherms for adsorbed H,O on TiO, at 298 K, measured at several frequencies with non-blocking electrodes. Dielectric Measurement with Bloc king Electrodes The blocking electrodes were also employed for the measurement of the frequency dependence of E’ and E” at 298 K and the results are given in fig. 6. In comparison with the data obtained by the non-blocking electrodes (fig. 2), the E’ value in fig. 6 is remarkably small, especially in the low-frequency region, and at 8 < I it converges to a limiting value when the frequency decreases. On the sample covered only with surface hydroxyls, the values of E’ and E” are small and independent of the frequency, i.e.no dispersion occurs. When the physisorption of H,O proceeds, the dielectric dispersion first appears at lower frequency, it extends to the next higher-frequency region, and finally the relaxation frequencyf, reaches 4.8 x 1 O4 Hz at 8 = 6.4. The E” curve has a Debye-type shape, having a peak height of 22 at 8 < 2; when 8 > 2, the height increases slightly. Although the data are not shown here, if we calculate tan6 from the E’ and E” values in fig. 6 and plot the value against frequency, we get a Debye-type curve. Furthermore, the Cole-Cole arc plots can also be given from the E’ and E” values, as illustrated in fig. 7. The size of the Cole-Cole arc is found to remain unchanged up to 6 = 1 , but it increases siightly when 6 > 1.These arcs can be expressed by the following Cole-Cole eauation :33 Here, E* is the complex dielectric permittivity, and E , and E , are the dielectric permittivities at f= 0 and co, respectively. cu = 2 4 and t is the mean value of the dielectric relaxation time, which is equal to (27cfm)-l. a is a parameter indicating the degree of the distribution of the relaxation time, being 0 < a < 1. From fig. 7, a can be estimated to be 0.14, which represents the distribution of the relaxation time to be small. The value, e0 - E,, is the chord length of the arc, being constant (60) until 0 = 1 is reached, but increasing gradually with increasing 8 when 8 > 1, amounting to 71 at 8 = 6.4; it also increases with rising temperature.T.Morimoto and T. Iwaki 949 I I I I I I I I 1 I I I I I I I I 0.2 0.4 0.6 0.8 pip, Fig. 5. Conductance isotherms for adsorbed H,O on TiO, at 298 K, measured at different frequencies from 30 Hz to 5 MHz with non-blocking electrodes. DC indicates the direct current conductance obtained just after the application of direct current voltage. Discussion Dielectric Dispersion in the Low-frequency Region Johnson and Cole34 have found with formic acid that the relationship E’ ~ c f - ” , where n = 2, holds in the frequency region below lo4 Hz, which is interpreted in terms of the interfacial polarization due to the d.c. conduction. The present system shows that the relationship E’ ~ f - ~ , where n = 1, holds when 8 -= 1 . However, at 8 > 1, the relationship E‘ ~ f - ~ does not hold, because another small relaxation appears over the frequency region from 30 Hz to 100 kHz and overlaps with the principal relaxation (fig.2). The dielectric relaxation, which appears at room temperature in the low-frequency region when H20 is physisorbed, was discovered by Bailey35 on proteins as well as on inorganic crystals such as NaHCO,. Furthermore, a similar relaxation has been observed on synthetic polymers, clays and living cells,36 and is explained not through the orientation of the molecules themselves, but through the time-dependent surface conductance. Schwartz3’ has accounted for the phenomenon in terms of the movement of the counter ions. McCafferty et ~ 1 . ~ ~ 9 24 have found a similar dielectric relaxation at room temperature in the low-frequency region on the a-Fe20,-H20 adsorption system and elucidated the phenomenon in terms of the relaxation of H20 itself, which is strongly adsorbed on surface hydroxyls. However, several contradictions can be found in their conclusion if950 30 - - Dielectric Behaviour of Adsorbed Water 100 80 60 e' 40 20 0 I I 1 I I I 1 1 1 30 Fig.6. Dielectric permittivity (a) and dielectric loss (6) for various 8 of adsorbed H,O on TiO,, measured at different frequencies with blocking electrodes at 298 K. The 8 values are 0 (l), 0.01 ( 9 , 0.05 (3), 0.07 (4), 0.10 (5), 0.19 (6), 0.25 (7), 0.28 (S), 0.34 (9), 0.41 (lo), 0.55 ( l l ) , 0.75 (12), 1.2 (13), 1.9 (14), 2.3 (15), 2.9 (16), 3.8 (17) and 6.4 (18). €'I 20 10 0 20 40 60 80 100 Fig.7. Cole-Cole plots for various 8 of adsorbed H,O on TiO, at 298 K. The 8 values are 0 (a), 0.05-0.94 (O), 1.2 (x), 1.4 (A), 1.6 (o), 1.9 (+), 2.3 (m), 2.9 (A), 3.8 (a) and 6.4 (a).T. Morimoto and T. Iwaki 95 1 . . - . - c1 . . . . . . . . 9 * . . * . . . . . . . . . . -52 . . . . , . . . . . . I ' - _ 62 - ~ - _ - . - - Rl R2 Fig. 8. Two-layer model and equivalent circuit. C, and C, are electrical capacitance, and R, and R, electrical resistance. their experimental data are examined in detail. (1) Despite the well known fact that the E' values of both bulk ice and liquid H 2 0 have negative temperature coefficients, that of the a-Fe20,-H20 system gives a positive one, as in the case of the present Ti0,-H,O system. (2) The chord length, E, - E * , of the Cole-Cole arc should increase with increasing t?, but their actual data have no change with 8.(3) The activation energy of the relaxation at 288 K increases extraordinarily with increasing 8, e.g. at 8 = 3.5 it amounts to 1.66 times larger than the value of ice, 56.5 kJ mol-l. (4) If we estimate the dipole moment of adsorbed H20 on a-Fe20, on the basis of the Cole-Cole plots, the value of 30 D t can be obtained for H20, which is anomalously large compared with the value 1.84 D for liquid H 2 0 . These discrepancies suggest that the dielectric dispersion observed at room temperature on the systems such as a-Fe20,-H20 and Ti0,-H20 is caused by a relaxation other than the rotational oscillation of adsorbed H 2 0 itself. Two-layer Model For the reason stated above, it is necessary to postulate an idea for the interpretation of dielectric relaxation at room temperature.The key to solving this problem lies in the change in the G value of the system Ti0,-H20 (fig. 5), i.e. the fact that the conductance isotherm reveals a tendency similar to the adsorption isotherm and also to the dielectric isotherm at low frequencies, as described above. as illustrated in fig. 8. The left-hand side of fig. 8 implies a layer containing both the electrode-particle and particle-particle interfaces in the system, where the interfacial polarization can appear through the formation of the electrical double 40 when the charges migrate on We therefore propose the use of the two-layer 7 I D = 3.33564 x C m.952 Dielectric Behaviour of Adsorbed Water applying a field to the sample.The right-hand side of fig. 8 is a layer containing the bulk and surface of all the particles, where the quantity of adsorbed H,O changes. If the conductance of the sample is measurable, the interfacial polarization will occur in such a system, which results in the appearance of the dielectric relaxation of a Maxwell- Wagner type.38 The dielectric permittivities E , and E , and the dielectric relaxation frequencyf, in this system can be expressed by the following equations:38 1 Here, E , and E, are the dielectric permittivities in the two layers, o1 and 0, are the electrical conductivities, dl and d, are the thickness of the two layers, respectively, and d is the sum, dl + d,. If we block the electrodes with an insulating film, 0, will become so small that the condition 0, < a, holds.Since usually dl 4 d,, eqn (2)-(4) can be simplified as in the following equations: E , = ( 2 ) d em = ( 2 ) d As can be understood from eqn (7), when o2 varies significantly on adsorption of H,O compared with the change in E, and e,, the fm value is affected remarkably by 0,. As shown in fig. 5, the experimental G value increases pronouncedly with increasing 8, which naturally leads to an increase inf, according to eqn (7). Sincef, is also the frequency at the inflection point of the E’ us. f curve, or at the maximum point of the e” us.fcurve, the increase in f, gives rise to a shift of the E’ and E” curves to a higher-frequency region as a whole. Thus, the experimental data in fig. 2 and 6 can really be elucidated on the basis of the two-layer model.Effect of Blocking Electrodes As shown in fig. 6, the E , value converges to 68 when 8 = 1 is reached. This value was proved to be equal to the capacitance of the blocking Teflon film, by measuring the capacitance of the cell filled with H,O instead of the sample. The E , value as well as the e’ values at low frequencies increase when 8 increases; this becomes more conspicuous when the non-blocking electrodes are employed (fig. 2). On the other hand, at higherT. Morirnoto and T. Iwaki 953 10 5* I I I I I I I I I I 0 0.2 0.4 0.6 0.8 PIP, 3 Fig. 9. Dependence off, on relative pressure, measured at various temperatures, 278 (Q), 288 (O), 298 (O), 308 (0) and 323 K (a). frequencies the increase in E’ is not so prominent regardless of whether the electrodes are blocked or 42 For instance, at a high frequency of 10 kHz the E’ value is only ca.30-40 at 0 = 2.9 when the electrodes are either blocked or not, though at the low frequency of 100 Hz it amounts to 78 when blocked, and to 300 when non-blocked. Thus, it follows that the results shown in fig. 6 are owing to a small capacitance and a large resistance of the blocking electrodes. When the non-blocking electrodes are utilized, the G value increases remarkably with increasing 0 (fig. 5), which leads to a large interfacial polarization and accordingly an extraordinary enhancement of e’ at low frequencies Thef, value can be read from the top of the E” curve in fig. 6, and plotted against the relative pressure PIP, of H,O as shown in fig. 9.It is found from fig. 9 that thefm curve reveals a sharp knee up to 6’ = 1 and then a slow increase. The shape of this curve is very similar to the conductance isotherm, especially around 1 kHz, as shown in fig. 5. Furthermore, fig. 9 shows that the fm value increases with rising temperature, in accordance with the result shown in the temperature dependence of the conductance isotherm, though the latter is not given here. In other words, thef, value increases in parallel with the G value as 0 increases; this really manifests the validity of eqn (7) of the two-layer model. Under this condition, E , and E , can be determined as the dielectric permittivities of the blocking film and the sample, respectively, being the values at the both ends of the E’ us.fcurve in fig. 6, as predicted by eqn (5) and (6). Thus, we can conclude from the theory of the two-layer model that the shift of the E’ curve to a higher frequency region due to the increase in 0 can be ascribed to the increase in 0. (fig. 2). Effect of Non-blocking Electrodes Since the conductance of TiO, itself is essentially small, the contact zone at the electrode-particle or particle-particle interface may operate as a blocking layer ; this is true for TiO, when no physisorbed H,O is present. However, when H,O is adsorbed, the interfaces are deblocked to facilitate the electrical conduction which results in a drastic enhancement of E’ and also of E” in the low-frequency region. In spite of such954 Dielectric Behaviour of Adsorbed Water 0.1 0 ‘A 0.2 0.4 0.6 0.8 PPO Fig.10. Dependence off,, on relative pressure, plotted at various dielectric permittivities, 30 (o), 50 (a), 100 (@) and 200 (a). complicated phenomena, we can infer the nature of the E’ curve obtained by non-blocking electrodes. In view of the fact that an inflection point appears around a constant E’ value of 34 in the E’ vs. f curve when the blocking electrodes are used (fig. 6), we can further try to plot the f value near a midway value of E’, e.g. at E’ = 100, of the curve in fig. 2, where non-blocking electrodes are used. Fig. 10 shows such plots,f,, us. PIP,, together with those for E’ values other than 100. From fig. 10, it can be found that the shape of thef,, curve obtained by non-blocking electrodes is similar to that of thef, curve obtained by the blocking electrodes, as shown in fig.9. Furthermore, all of these curves resemble the conductance isotherm illustrated in fig. 5. This suggests that the dielectric dispersion observed by non-blocking electrodes (fig. 2) is also caused by the same reason as that for the dispersion observed by blocking electrodes (fig. 6), though the apparent shape of the E’ curve is quite different from each other in both cases. The difference lies in the fact that the use of the non-blocking electrodes permits the charge injection into and the charge extraction from the space charge region, which results in a larger interfacial polarization. The dielectric dispersion obtained on the system a-Fe203-H20 at room temperature by McCafferty et al.239 24 may be elucidated by the same reason as that discussed here.Dielectric Activation Energy at Room Temperature As discussed above, it can be concluded that the dielectric relaxation observed at room temperature in the low-frequency region on the Ti02-H20 system is ascribed to the interfacial polarization which is set up by the charge accumulation when the conductance is measurable. This suggests that the dielectric relaxation of adsorbed H 2 0 itself will appear at lower temperatures or at higher frequencies. The activation energy E, of conduction can be estimated from the temperature coefficient of G through the equation: G = Go exp (-EJRT). (8) Here, Go is a constant. From the conductance data in fig. 5, E, can be calculated to be 37.7 kJ mol-l on the surface covered with hydroxyls and 26.4 kJ mol-1 in the presenceT. Morimoto and T.Iwaki 955 ( a ) (6) Fig. 11. Schematic representation of adsorbed H,O on TiO,. Surface hydroxyl layer (a) and the first physisorption layer of H,O on surface hydroxyls (b). of physisorbed H,O, the latter being almost unchanged even when 8 increases. Thef, value obtained by the blocking electrodes can be expressed by the following equation according to Eyring’s theory :43 1 /z = Znf, = A exp (- &/RT). (9) Here, z is the dielectric relaxation time. This equation permits the evaluation of the activation energy of dielectric relaxation, Ed, on the basis of thef, data. From the temperature coefficient off, in fig. 9, Ed can be calculated to be 27.2 kJ mol-1 over a wide range of 8, in accordance with the activation energy of conductance E,, which shows that the activation process is reduced to a single phenomenon, i.e.the conduction of the system. Mechanism of Conduction in TJ0,-H,O Adsorption System As discussed above, when the surface is covered singly with surface hydroxyls, G is very small and at the same time E, is large, but when H,O is physisorbed, G increases and at the same time E, decreases. Since the activation energy approximates to the energy of hydrogen bonding, the conduction can be inferred to be due to the proton hopping between the two neighbouring H 2 0 molecules as shown in fig. 11 (a). On the surface of TiO,, each H,O molecule in the first physisorption layer is adsorbed on two hydroxyls.2 Therefore, the proton hopping which is accelerated by the H,O physisorption may occur from surface hydroxyls to the adsorbed H 2 0 molecules, as demonstrated in fig.11 (b). Thus, the surface conductance increases as 8 increases. The system is composed of small but a large number of contact points between particles, through which the transfer of a finite amount of charge is permitted. This will produce a giant dipole in each particle as well as the polarization at electrode-particle interfaces. All these effects serve to enhance the interfacial polarization. This work was partly supported by a Grant-in-Aid for Scientific Research, No. 57470007, from the Ministry of Education, Science and Culture of Japanese (3 over n men t . References 1 T. Morimoto, M. Nagao and F. Tokuda, Bull. Chem. SOC. Jpn, 1965, 41, 1533.2 T. Morimoto, M. Nagao and F. Tokuda, J. Phys. Chem., 1969, 73, 243. 3 T. Morimoto and H. Naono, Bull. Chem. SOC. Jpn, 1973, 46, 2000. 4 M. Nagao, K. Yunoki, H. Muraishi and T. Morimoto, J. Phys. Chem., 1978, 82, 1032. 5 T. Morimoto and M. Nagao, J. Phys. Chem., 1974, 78, 11 16. 6 T. Morimoto, Y. Yokota and S. Kittaka, J. Phys. Chem., 1978, 82, 1996. 7 T. Morimoto and K. Morishige, J. Phys. Chem.. 1975, 79, 1573. 8 M. Nagao and T. Morimoto, Bull. Chem. SOC. Jpn, 1976, 49, 2977. 9 T. Morimoto and Y. Suda, Langmuir, 1985, 1, 239.956 Dielectric Behaviour of Adsorbed Water 10 M. Nagao, K. Matsuoka, H. Hirai and T. Morimoto, J. Phys. Chem., 1982, 86, 4188. 11 R. L. McIntosh, Dielectric Behaviour of Physically Adsorbed Gases (Marcel Dekker, New York, 1966). 12 G.Jones, Dielectric and Related Molecular Processes, ed. M. Davies (The Chemical Society, London, 1977), vol. 3, p. 176. 13 S. M. Nelson, A. C. D. Newman, T. E. Tomlinson and L. E. Sutton, Trans. Faraday Soc., 1959, 55, 2186. 14 G. Ebert and G. Langhammer, Kolloid Z., 1961, 174, 5. 15 M. G. Baldwin and J. C. Morrow, J. Chem. Phys.. 1962, 36, 1591. 16 K. Kaneko and K . Inoue, Bull. Chem. Soc. Jpn, 1974, 47, 1139. 17 S. Kurosaki, J. Phys. Chem., 1954, 58, 320. 18 K. Kamiyoshi and J. Ripoche, J. Phys. Radium, 1958, 19, 943. 19 M. Freymann and R. Freymann, J . Phys. Radium, 1954, 15, 165, 20 N. K. Nair and J. M. Thorp, Trans. Faraday SOC., 1965, 61, 975. 21 B. Morris, J. Phys. Chem. Solids, 1969, 30, 73. 22 P. Hoekstra and W. T. Doyle, J. Colloid Interface Sci., 197 1, 36, 5 13. 23 E. McCafferty, V. Pravdic and A. C. Zettlemoyer, Trans. Faraday SOC., 1970, 66, 1720. 24 E. McCafferty and A. C. Zettlemoyer, Discuss. Faraday Soc., 1971, 52, 239. 25 S. Kondo, M. Muroya, H. Fujiwara and N. Kamauchi, Bull. Chem. SOC. Jpn, 1973,44, 1362. 26 T. Morimoto, M. Nagao and T. Omori, Bull. Chem. SOC. Jpn, 1969, 42, 943. 27 T. Omori, J. Imai, M. Nagao and T. Morimoto, Bull. Chem. SOC. Jpn, 1969, 42, 2198. 28 R. I. Bickley, Chemical Physics of Solids and Their Surfaces, ed. J. M. Thomas and M. W. Roberts (The 29 P. Jones and J. A. Hockey, Trans. Faraday Soc., 197 1, 67, 2679. 30 P. Jones and J. A. Hockey, Trans. Faradav Soc., 1972, 68, 907. 31 C. M. Huggins and A. H. Sharbaugh, J. Chem. Phys., 1963, 38, 393. 32 J. Volger, Progress in Semiconductors (Heywood, London, 1960), vol. 4, p. 205. 33 K. S. Cole and R. H. Cole, J. Chem. Phys., 1941, 9, 341. 34 J. F. Johnson and R. H. Cole, J . Am. Chem. SOC., 1951, 73,4536. 35 S. T. Bayley, Trans. Faraday SOC., 1951, 47, 509. 36 H. P. Schwan, G. Schwartz, J. Maczuk and H. Pauly, J. Phys. Chem., 1962,66, 2626. 37 G. Schwarz, J. Phys. Chem., 1962, 66, 2636. 38 L. K. H. van Beek, Progr. Dielectrics, 1967, 7, 69. 39 0. W. Johnson, J. W. DeFord and S. Myhra, J. Appl. Phys., 1972, 43, 807. 40 J. J. Fripiat, A. Jelli, G. Poncelet and J. Andre, J. Phys. Chem., 1965, 69, 2185. 41 G. Jones and M. Davis, J. Chem. SOC., Faraday Trans. 1, 1975, 71, 1971. 42 G. Jones, J . Chem. SOC., Faraday Trans. 1 , 1975,71, 2085. 43 S. Glasstone, K. J. Laidler and H. Eyring, Theory of Rate Processes (McGraw-Hill, New York, 1941), Chemical Society, London, 1978), vol. 7, p. 118. p. 548. Paper 61009; Received 2nd January, 1986

ISSN:0300-9599    

DOI:10.1039/F19878300943

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. I , 1987, 83, 957-966 Dielectric Behaviour of Adsorbed Water Part 2.-Measurement at Low Temperatures on TiO, Tohru Iwakit Department of Chemistry, Faculty of Science, Hiroshima University, Naka-ku, Hiroshima 730, Japan Tetsuo Morimoto* Department of Chemistry, Faculty of Science, Okayama University, Tsushima, Okayama 700, Japan On TiO, samples with controlled amounts of chemisorbed and physisorbed H,O, the dielectric permittivity and dielectric loss have been measured at frequencies from 0.1 Hz to 5 MHz at temperatures from room temperature to 77 K. As a result, three kinds of relaxations have been discovered and the crude Cole-Cole plots have been analysed into three arcs I, I1 and 111; relaxation I is due to interfacial polarization as reported in our previous paper.The arcs I1 and I11 are assigned to the relaxations due to surface hydroxyls and physisorbed H,O, respectively. It is found that arc I1 decreases with increasing coverage 8 of physisorbed H,O and becomes extinct at 6 > 1. Furthermore, both the chord length of arc I11 and the dielectric activation energy of the relaxation 111 reveal a similar variation as 6 increases: first they increase linearly until 8 = 2 is reached and then the slope of each curve becomes more gradual. Finally, the dielectric activation energy of adsorbed H,O approaches that of ice. These phenomena are explained on the basis of the proposed models of H,O adsorbed on TiO,. H 2 0 molecules are chemisorbed on the surface of metal oxides to form surface hydroxyls in the atmosphere, on which the physisorption of H,O OCCU~S.~-~ We have developed a method for determining the amount of surface hydroxyls and a technique for controlling the concentration of them,4~ promoting the investigation of the surface properties of metal oxides.Dielectric investigation is one of the most effective techniques used to ascertain the adsorbed state and binding force of adsorbate molecules.6 Hitherto, a number of studies have been made on the dielectric properties of adsorbed H 2 0 on metal oxides such as Si0,,7v Al,0397 lo and Fe20,.11-13 In a previous paper14 the dielectric behaviour of the TiO,-H,O adsorption system has been investigated at room temperature. As a result, it has been found that a large dielectric dispersion appears in the low-frequency region from 0.1 to lo4 Hz and that the dispersion originates in an interfacial polarization of the Maxwell-Wagner type which increases (when H,O is adsorbed) owing to an enhance- ment of electrical conductance.This suggests that the dielectric relaxation of the adsorbed molecules themselves will appear in a higher frequency region at lower temperatures. Kamiyoshi and Ripoches reported three kinds of dielectric absorption of adsorbed H,O on silica gel from measurements at low temperatures. Kondo et al.1° also observed three kinds of relaxation processes of adsorbed H,O on y-Al,03 from low-temperature measurements. The present investigation has been undertaken to clarify the dielectric relaxation of surface hydroxyls and physisorbed H,O by measuring the dielectric permittivity and dielectric loss in the Ti0,-H,O adsorption system at low temperatures. 7 Present address : Hiroshima Technical Institute, Mitsubishi Heavy Industry, Hiroshima 733, Japan.957958 Dielectric Behaviour of Adsorbed Water Experiment a1 The TiO, (rutile) sample used in this study was the same as that in the previous paper,14 having a specific surface area of 8.35 m2 g-l. The structure of the dielectric cell was also described in detail in the previous paper; the cell was connected to a volumetric adsorption apparatus. Here, stainless-steel electrodes were employed after blocking with Teflon film 30 pm in thickness. The packing density of TiO, was 28%. Prior to the dielectric measurements, the population of adsorbed H,O on TiO, was controlled on the basis of the adsorption isotherm.After the adsorption equilibrium of H,O was first achieved at a given pressure at 273 K, the temperature of the adsorption system was reduced gradually to 253 K and again equilibrium was attained at this temperature. After that the temperature was lowered to 232, 195, 178, 159 and 77 K successively. At each stage the dielectric permittivity and dielectric loss were measured at frequencies from 0.1 Hz to 5 MHz using impedance bridges TR4 and TRlOC (Ando Electric Co.). The attainment of equilibrium was also confirmed by measuring the dielectric permittivity. It took ca. 2-6 h for equilibration at every temperature. Taking account of the vapour pressure of H,O at a given temperature, the adsorbed amount was corrected; as a result, the correction was found to be negligibly small, e.g.0.13% at 8 = 0.38. After evacuating the sample in a vacuum of Torrt at 373 K, the remaining amount of surface hydroxyls was measured by means of the successive ignition loss meth0d.~9 l5 Results The dielectric permittivity E’ and dielectric loss E” of H,O adsorbed on TiO,, at 8 = 2.56 as an example, measured by the blocking electrodes at various temperatures, are illustrated in fig. 1 as a function of the frequencyf. It is seen from fig. 1 that the curves of E’ and E” at 273 K are typical Debye curves. The E’ curve gives a limiting value E, at the lowest frequency region, passes a broad transition region extending from 10, to lo6 Hz and finally arrives at another limiting value e , at the highest frequency region.Corresponding to this change in E ’ , the E” curve represents the peak centred on the inflection point in the E’ curve. This dielectric dispersion I can be assigned to the interfacial polarization which appears at the electrode-particle and particle-particle interfaces through the migration of charges due to proton hopping in the adsorption layer of H,O; this can be interpreted in terms of a two-layer model of the Maxwell-Wagner type, as discussed previ0us1y.l~ When the temperature is lowered, each of the E’ and E” curves shifts as a whole to a lower frequency region. At 195 K, the flat part E, of the curve disappears and at 178 K the inflection point of the E’ curve as well as the peak point of the E” curve disappear from the figure. In the previous work,14 the shift of this absorption peak took place towards the higher-frequency side when 8 increased at a constant temperature, whereas fig.1 shows that the same shift occurs when the temperature is raised at a constant 8. Both findings can be ascribed to a single reason, viz. both parameters can facilitate proton hopping between neighbouring H,O molecules. In fig. 1 (b), a small peak of relaxation I11 appears near the higher-frequency end of the absorption peak I at 195 K; the relaxation I11 also shifts to the lower-frequency side when the temperature is lowered. Since this peak is very small, the corresponding step in the E’ curve is also very faint; thus, only the inflection point is indicated by arrows in fig. 1 (a). The dielectric losses measured at 273 K on the sample covered singly with surface hydroxyls are also plotted in fig.1 (b). As is seen from this curve, the absorption peak I does not appear on this sample at 273 K because of a small electrical conductance. t 1 Torr = 133.322 Pa.T. Iwaki and T. Morimoto 959 Fig. 1. Typical variation in dielectric permittivity E’ (a) and dielectric loss E” (b) with frequency for adsorbed H,O on TiO, at 8 = 2.56, measured with blocking electrodes at various temperatures (marked on the figure). The broken line indicates the E” curve of TiO, covered singly with surface hydroxyls, measured at 273 K. Instead, a new absorption peak I1 appears near 30 kHz, although it is not discernible in the absorption curve on the sample with physisorbed H,O with 0 = 2.56, measured at the same temperature.As will be clarified later, this kind of peak has been found to appear at low temperatures only when 0 < 8 < 1. Thus, it is understood that three kinds of relaxation I, I1 and 111 are contained in the E’ and E” curves in fig. 1 and they overlap. Therefore, it is preferable to analyse the original curves of E’ and E” into the three relaxations for the detailed investigation of each relaxation. In fig. 2 an example of the graphical analysis is illustrated on the curves of E’ and E” at 8 = 0.73, measured at 178 K. In this analysis every relaxation is assumed to obey the following equations:I6 sinh ( 1 - a) 2 cosh(1 -a)Z+sin- 1 2 E’ = E , + - ( E o - E , ) an 1 2 2 an cosh(1 -cc)Z+sin- 2 cos - E N = - ( & ( ) - E m ) z = log(cuz).960 Dielectric Behaviour of Adsorbed Water 1.0 0.8 0.6 0.4 0.2 0 E 4.4 4.6 4.8 5.0 5.2 5.4 5.6 c1 Fig.2. Analysis of dielectric permittivity E' (a), dielectric loss E" (6) and Cole-Cole plot (c) for adsorbed H,O on TiO, at 8 = 0.73, measured at 178 K. Frequencies (in Hz) marked on curve in ( c ) .T. Iwaki and T. Morimoto 96 1 c’ Fig. 3. Cole-Cole plots for several coverages of adsorbed H,O on TiO,: (a) 6 OH nmP2 outgassed at 373 K; (b) 10 OH nm-,, 8 = 0; (c) 8 = 0.38; ( d ) O = 0.75; (e) 8 = 1.23; d f ) 0, 8 = 2.59; n , 8 = 4.0; 0 , O = 5.0. Measurements at temperatures: 159 (O), 178 (O), 195 (0) and 273 K (A) in (a)-(d), and 178 K all in (f). Frequencies in kHz on figure. Broken lines indicate the plots analysed into three arcs. Here, E, and E, are the dielectric permittivities a t f = 0 and co, respectively.a represents the distribution of the dielectric relaxation time z, usually having a value between 0 and 1. w is the angular frequency (co = 271f). The mean value of the relaxation time z can be related to the dielectric relaxation frequencyf,, i.e. the frequency at the top of the Cole-Cole arc, by the following equation : 1 zfm = - 27c ’ (4) - First, the largest part, corresponding to relaxation I, is subtracted from the whole relaxation curve, then the remaining curve is analysed according to the method employed by von Hippel et aZ.,17 which leads to separation into relaxations I1 and 111. Fig. 2(a) and ( b ) show the analysis of E’ and E”, respectively, and fig. 2(c) gives the Cole-Cole arcs I, I1 and I11 obtained therefrom.The complex dielectric permittivity E* for each relaxation is expressed by the following ( 5 ) equation :Is E O - E , 1 + (iwz)l-a. & * = E m +962 Dielectric Behaviour of Adsorbed Water 0.5 1.58 0 4.5 5.0 5.5 E’ Fig. 4. Cole-Cole plots for the coverages, 0 = 0.75 (A) and 2.59 (B), of adsorbed H,O on TiO,, measured at (a) 159, (b) 232 and (c) 298 K. Broken lines indicate the plots analysed into three arcs. Frequencies on curves (M = MHz, K = kHz). For instance, the complex dielectric permittivity at 30 Hz is obtained from the synthesis of the vectors indicated by arrows in arcs I and 11, where the contribution of E” of arc I11 is negligible at this frequency. In fig. 3, the crude Cole-Cole plots and three kinds of arcs separated therefrom are given for the samples with various coverages of hydroxyls and physisorbed H20.Fig. 3(b) illustrates the original Cole-Cole plots on the sample fully covered with surface hydroxyls (10 OH nrn-,) and without physisorbed H20. Hence, only relaxation I1 appears, and it seems to be composed of two arcs at 159 and 178 K, but they are united into the single smooth arc at 273 K. This fact suggests that there are two kinds of surface hydroxyls on TiO,, which differ in their states of motion in the alternating field, and that they are reduced to the single state when the temperature is raised. Fig. 3(a) shows the Cole-Cole plots on a partially dehydroxylated sample, i.e. with 6 OH nm-,, which was made by evacuating the fully hydroxylated sample at 373 K for 4 h.In this figure, the whole arc of relaxation I1 is smaller than that on the fully hydroxylated surface and the original two arcs are split into three arcs, suggesting the existence of three kinds of hydroxyls. When the temperature is raised the three arcs unite to a smooth one as in the case of the fully hydroxylated surface. Though all the Cole-Cole plots on the samples evacuated at different temperatures are not illustrated here, it has been found that the whole arc becomes smaller, leaving the triply split state as it is, when the dehydroxylation proceeds. This undoubtedly provides evidence that the arc I1 is attributable to the relaxation of surface hydroxyls themselves. In addition, it should be noted that relaxation I does not appear under the conditions in fig.3(a) and (b) because the conductance is extremely small. The feature of the Cole-Cole plots varies remarkably when 6 increases, as shown inT. Iwaki and T. Morimoto 963 fig. 3(c)-(f); some of them are analysed into three kinds of relaxations, I, I1 and 111. First, it should be pointed out that the size of arc 111 increases with increasing 8, which shows that arc I1 can be ascribed to the relaxation of physisorbed H 2 0 itself.l8, l9 Secondly, it is also interesting to see that arc I1 decreases with increasing 8 and becomes extinct when 8 > 1. Thirdly, at constant 8, the original arc as well as the separated ones enlarge when the temperature is raised. The dielectric permittivity at the higher-frequency end of arc 111, E,, which is considered to be composed of the sum of the dielectric permittivity of TiO, and that of the intrinsic one of adsorbed H20, increases slightly with lowering the temperature of measurement, in agreement with the Curie-Weiss rule.2o Fig.4 demonstrates the variation of the three kinds of arcs, observed at 8 = 0.75 and 2.59 when the temperature is widely changed. As shown from the plots at 8 = 0.75, the size of arc 111 remains almost unchanged, in accordance with the theory in which the rotational oscillation of adsorbed molecules is independent of temperature.6 However, the frequency, which gives the same value of E', becomes higher when the temperature is raised and the frequency region at which arc 111 appears at 298 K is beyond 5 MHz, the uppermost frequency measurable in the present apparatus.Unlike arc 111, the size of arc I1 depends largely on the temperature and increases with rising temperature, suggesting that relaxation I1 involves an activation process. When 8 = 2.59, arc I1 becomes extinct as described above, despite the fact that arcs I and I11 remain larger than those observed when 8 = 0.75. At 232 K, the frequency region at which arc 111 appears is beyond 5 MHz. At 298 K, the higher-frequency end of relaxation I also exceeds 5 MHz. This also comes from the fact that relaxation I is caused by interfacial polarization and itsf, value shifts to higher frequencies when the temperature is raised.14 In conclusion, relaxation 111 of adsorbed H 2 0 itself at room temperature is outside the range of frequencies measurable in the present study; therefore the dielectric dispersion observed at low frequencies at room temperature is only the relaxation of the interfacial polarization, as described previously.l4 Discussion Onsager2' deduced theoretically that the chord length of the Cole--Cole arc can be expressed by the following equation: where N is the number of molecules in the unit volume, p the dipole moment of the molecule, E the dielectric permittivity, n the refractive index, k the Boltzmann constant and T the absolute temperature. Eqn (6j shows that the chord length of the Cole-Cole arc, E ~ - E , , is proportional to the number of molecules. Fig. 5 gives the plot of AE (= E , - E,) of arc I11 us. 8. Eqn (6) has so far been corroborated indirectly by plotting the increment of the dielectric permittivity against 8.On the contrary, the linearity in fig. 5 is direct evidence for eqn (6), although the whole curve is composed of two linear parts: a steeper straight line and a shallower one before and after 8 = 2, respectively. This change in the slope of the curve suggests that the state of motion of adsorbed H,O varies at this coverage. Furthermore, A & may depend upon temperature according to eqn (6), but the experimental data in fig. 5 do not support the prediction, i.e. the AE curves measured at different temperatures agree with each other. The absence of the temperature dependence of AE is rather in accord with McIntosh's theory that the dielectric permittivity of adsorbed molecules moving as rotational oscillators is independent of the temperature.6 On the other hand, the dielectric relaxation frequencyf, also gives information on the movement of adsorbed H 2 0 in the alternative field.Fig. 6 shows the plots off, vs. T-l on relaxation 111 at different values of 8. For comparison, the data are cited on964 Dielectric Behaviour of Adsorbed Water 0.8 0.6 W Q 0.4 0.2 0 I I 1 I I 1 2 3 4 5 e Fig. 5. Chord length of Cole-Cole plots as a function of coverage at 159 K (a), 178 K (0) and 195 K (0). 108 106 G --. G 104 lo2 c,O , 2:O , 20,O 18: 16: , 4 5 6 1 0 3 KIT Fig. 6. Dielectric relaxation frequency as a function of temperature, at 8 = 0.38 (O), 0.75 (a), 1.23(A), 2.59 (A), 4.0 (H) and 5.0 ( x ) . Dotted and broken lines indicate the data for A1,0, (20% H,O)y and SO, (23.3% H,O),s respectively.Al,0,9 and with 20 and 22.3 wt % H,O, respectively, by replotting the original data. The& values for ice Ihz2 are also cited in fig. 6 . Since the dielectric relaxation frequency of liquid H,O at 273 K is known to be > 1 O 1 O Hz, it is found that the state of motion of adsorbed H,O on TiO, is between those of liquid H,O and solid H,O. Moreover, the f m value of adsorbed H,O decreases when the temperature is lowered as well as whenT. Iwaki and T. Morimoto 965 40 d I - h 30 a 2 . 4 20 I I I I I 1 1 2 3 4 5 101 e Fig. 7. Dielectric activation energy as a function of coverage. . . 0 0 0 0 I I I I ,Ti\ /Ti\ /Ti, ,Ti\ 0 0 0 (a) (b 1 Fig. 8. Schematic representation of surface hydroxyl layer (a) and the first physisorption layer of H,O on surface hydroxyls (6).9 increases and approaches that of ice. Thus, we can understand that the rotational motion of adsorbed H,O is strongly restricted by the surface. Another observation from fig. 6 is that the slope of the straight line off, us. T-l becomes greater as 9 increases. The slope of this plot permits the calculation of the dielectric activation energy AE of relaxation 111, as shown in fig. 7. As fig. 7 shows, the 9-dependence of AE is similar to that of A& in fig. 5. The H,O initially adsorbed is strongly bound to the surface, but its oscillational mobility is as large as that of liquid H,O. When 9 increases, AE increases linearly as does A&, implying that the movement of adsorbed H,O is increasingly disturbed through mutual interaction. Finally at 8 > 2, the slope of the AE curve decreases and approaches the value of ice, suggesting that the movement of adsorbed molecules in the alternating field is as disturbed as that in ice.Thus, it is reasonable to infer that at small 9 the H,O molecules are strongly adsorbed on surface hydroxyls through hydrogen bonding1 which permits easy orientation for the alternating field and difficult participation in the ice nucleation. When 9 > 2, three-dimensional ice crystals begin to grow and the orientation polarization of adsorbed H,O becomes as hard as that of ice. It has been reported that at 9 = 1 the ratio of the number of physisorbed H,O molecules to that of the underlying surface hydroxyls is ca. 1 : 2 on the surface of TiO,,' which indicates that one H,O molecule is bonded to two hydroxyls through hydrogen bonding.The models of the fully hydroxylated (9 = 0) and H,O-physisorbed (9 > 1) surfaces on TiO, are illustrated in fig. 8. When the alternating field is applied to the surface hydroxyls which are free from physisorbed H,O, they will be oriented to one966 Dielectric Behaviour of Adsorbed Water direction at the moment when the hydrogen-bonded structure is formed more strongly between the neighbouring hydroxyls than when no field is applied [fig. 8(a)]. This may facilitate proton hopping between the neighbouring hydroxyls. As stated above, arc I1 (i.e. the relaxation of surface hydroxyls) involves an activation process; undoubtedly this is proton hopping. As pointed out in fig. 3(b), the Cole-Cole plot of surface hydroxyls gives two arcs, suggesting the existence of slightly different modes in relaxation 11, probably because of the variety of the underlying crystal planes.If some surface hydroxyls are removed by degassing the fully hydroxylated surface at higher temperature, the length of hydroxyl chains will become short, which may produce another mode of relaxation process [fig. 3 (a)]. As stated above, the size of arc I1 is reduced as H 2 0 is adsorbed (fig. 3); this indicates that surface hydroxyls lose the characteristic relaxation through the physisorption of H20. It may be considered that the adsorption of H 2 0 on the top of surface hydroxyls break the hydrogen bonding between neighbouring hydroxyls, and at the same time forms the hydrogen bonding of the H 2 0 molecule with the underlying hydroxyls, as expressed schematically in fig.8 (b). This would account for the simultaneous occurrence of the decrease of arc I1 and the increase of arc 111. This work was partly supported by a Grant-in-aid for Scientific Research, no. 57470007, from the Ministry of Education, Science and Culture of the Japanese Government. References 1 T. Morimoto, M. Nagao and F. Tokuda, J. Phys. Chem., 1969, 73, 243. 2 T. Morimoto and M. Nagao, J. Phys. Chem., 1974, 78, 11 16. 3 T. Morimoto, Y. Yokota and S . Kittaka, J. Phys. Chem., 1978,82, 1996. 4 T. Morimoto and H. Naono, Bull. Chem. SOC. Jpn, 1973, 46, 2000. 5 T. Morimoto and Y . Suda, Langmuir, 1985, 1, 239. 6 R. L. McIntosh, Dielectric Behaviour of Physically Adsorbed Gases (Marcel Dekker, New York, 1966). 7 N. K. Nair and J. M. Thorp, Trans. Faraday SOC., 1965,61, 975. 8 K. Kamiyoshi and J. Ripoche, J. Phys. Radium, 1958, 19,943. 9 K. Dransfeld, H. L. Frisch and E. A. Wood, J. Chem. Phys., 1962, 15, 1574. 10 S. Kondo, M. Muroya, H. Fujiwara and N. Yamauchi, Bull. Chem. SOC. Jpn, 1973,46, 1362. 11 E. McCafferty, V. Pravdic and A. C . Zettlemoyer, Trans. Faraday SOC., 1970, 66, 1720. 12 E. McCafferty and A. C. Zettlemoyer, Discuss. Faraday SOC., 1971,52, 239. 13 K. Kaneko and K. Inoue, Bull. Chem. SOC. Jpn, 1974,47, 1 139. 14 T. Morimoto and T. Iwaki, J. Chem. SOC., Faraday Trans. 1 , 1987, 83, 943. 15 M. Nagao, K. Yunoki, H. Muraishi and T. Morimoto, J. Phys. Chem., 1978, 82, 1032. 16 K. S. Cole and R. H. Cole, J. Chem. Phys., 1941, 9, 341. 17 A. von Hippel, D. B. Knoll and W. B. Westphal, J. Chem. Phys., 1971, 54, 134. 18 J. J. Windle and T. M. Shaw, J. Chem. Phys., 1954, 22, 1752. 19 J. J. Windle and T. M. Shaw, J. Chem. Phys., 1956,25,435. 20 S . Oka and 0. Nakata, Theory of Solid Dielectrics (Iwanami Book Co, Tokyo, 1960). 21 L. Onsager, J. Am. Chem. SOC., 1936, 58, 1486. 22 S. Kawada, J. Phys. SOC. Jpn, 1978, 44, 1881. Paper 6/010; Received 2nd January, 1986

ISSN:0300-9599    

DOI:10.1039/F19878300957

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

J. Chem SOC., Faraday Trans. I, 1987,83, 967-983 A Phenomenological Approach to Micellisation Kinetics Denver G. Hall? Unilever Research Port Sunlight Laboratory, Quarry Road East, Bebington, Wirral, Merseyside L63 3J W A general phenomenological treatment of micellisation kinetics applicable to solutions containing mixed micelles is developed. Large and small perturbations from equilibrium are considered with the main emphasis on the latter where linear methods apply. It is argued that the micelle-monomer exchange kinetics of each micellar species can be described in terms of a single driving force, namely the chemical potential difference between micellar and monomeric forms. The approximations inherent in this view- point are discussed for single-component micelles. The particular case of alcohol solubilisation is considered and an expression is obtained for the relaxation time which enables equilibrium thermodynamic data to be used in the interpretation of the kinetic results.The driving force for the slow process due to changes in micelle number is shown to be the subdivision potential both for the stepwise association and micellar association mechanisms. Also, for single-component micelles the phenomenological coefficient is shown to be the equilibrium rate of micelle formation and breakdown divided by RT. Expressions are derived which allow for coupling between the fast and slow processes. The treatment leads to the same results as previous work based mainly on mass-action kinetics when the particular features of the latter are fed into it.However, it is more general and in some respects more convenient to use. In a recent paper1 the stepwise association model of micellisation developed by Aniansson and coworker^^-^ was modified by allowing for the changes in rate coefficients with solution composition that occur in solutions of ionic surfactants. This was done by postulating that the specific rates of the reactions concerned depend only on a small number of quantities which can be varied independently at equilibrium. The resultant treatment was also applied to mixed micelles. In the same paper a brief outline was given of an alternative approach based on the thermodynamics of irreversible processe~.~ This approach is considered more fully in the present paper. It gives results which are identical to those obtained previously and has two major advantages.It is more general and requires less tedious algebraic manipul- ation. A number of new problems are tackled formally. Among these are (1) a more complete description of the fast process for mixed micelles than that given previously, (2) a more general description of the slow process, (3) coupling between the two processes and (4) an extension to deviations from equilibrium beyond the linear regime. Whenever possible the thermodynamic aspects of the treatment are handled generally. As a consequence the treatment appears somewhat formal. However, this drawback is offset by the advantage that any particular thermodynamic treatment can be used. t Also at : Department of Chemistry and Applied Chemistry, University of Salford, Salford M5 4WT.967968 Phenomenological Approach to Micellisation Kinetics Basic Kinetic and Thermodynamic Equations We consider a system consisting of a solvent, component 0, a series of micellar species, r, s, t , etc., and a series of non-micellar species i, j , k , etc., which include among them the non-micellar forms of the micellar constituents. According to the stepwise association model the only reactions of any significance in the formation and breakdown of micelles are the uptake and release of monomeric material. A typical reaction may be written as r+i's where in this instance s is the micellar species formed by the addition of an i molecule or ion to a micelle of type r. According to linear phenomenological theory the net rate of the above reaction, Jri, is given by Jri = lri(O, + 0i- 0,) (1) where the subscript ri denotes the particular reaction concerned, lri is the appropriate phenomenological coefficient and the corresponding driving force (0, + Oi - 0,) is the affinity of the above reaction.For uncharged species the Bi are chemical potentials as usually understood. For charged species Qi is the chemical potential of the species concerned plus the chemical potential of enough counterions to neutralise its charge. Further details concerning this notation and usage are given in ref. (6)-(8), and it is sufficient for our present purposes to note that in systems containing ions the ei play the same role as the chemical potentials pi in systems where all species are uncharged.Since the above reaction is an elementary process we may writeg* lo lri = C i / k T where Ti is the equilibrium forward or backward rate of the above reaction. Eqn (2) is equally valid in ideal and non-ideal systems and does not rely on the reactions conforming to transition-state theory.ll Eqn (1) and its analogues for all other micellar and monomeric species are the basic kinetic equations on which the analysis of both fast and slow processes for the stepwise association model is based. To describe the thermodynamics of states undergoing reaction we use a generalised form of the procedures outlined in the appendix of ref. (1). We begin with the fundamental equation of the above system, which at constant T and p may be written as dg = 0i dmi + x 0, dC, (3) a r where g denotes the Gibbs free energy per mole of solvent, mi is the concentration of non-micellar i and C, is the concentration of micellar species r .The units of mi and C, are moles per mole of solvent and differ from molalities only by a constant multiplying factor. Let denote the number of i molecules or ions present in a micelle of type r . Let Cif" = C, 3 C, denote the concentration of micellar i and let C, = C, C, denote the concentration of micelles. When g has its minimum value consistent with given values of the mi, Cr and C , it is readily shown by the method of undetermined multipliers that the 0, are given by 0, = C Ni 0 Y - k ~ (4) i where the multipliers 0r and E are the same for all micellar species.lg l2 Eqn (4) enables eqn (3) to be rewritten as dg = x t9,dmi+C Brdq+EdC,.i iD. G. Hall 969 It is evident from this equation that we may regard 02 as the chemical potential of micellar i and also that E is the subdivision potential defined by Hill.13 When eqn (4) and ( 5 ) apply the micelles are in equilibrium among themselves but not necessarily with monomer. Also C, does not in general have the value which for given mi and Cy minimises g . We use the term ‘partial equilibrium’ to describe such states. When Oy = Oi for all i the system is in chemical equilibrium and when E = 0 we have subdivision equilibrium. When both conditions are satisfied the system is in complete equilibrium. When a micellar system in complete equilibrium is perturbed as in a T-jump or p-jump experiment both chemical and subdivision equilibria are disturbed.According to the stepwise association model the return to equilibrium takes place in essentially two stages. The faster stage is the re-equilibration of monomeric material with micelles and takes place at approximately constant C,. The slower process is the establishment of the new equilibrium value of C, which takes place under the condition that chemical equilibrium is maintained throughout. Thus whereas eqn (1) suggests that there should be a relaxation time corresponding to each independent elementary reaction, in practice one observes at most a single relaxation process corresponding to the equilibration of each micellar constituent and a slower relaxation process corresponding to the equilibration of C,. This suggests that for a given micellar constituent those relaxation processes which contribute significantly to the overall amplitude are kinetically alike.We proceed as follows. Treatment of the Fast Process Derivation of General Equations for Relaxation Times For a given micellar constituent i we sum eqn (1) over all micellar species to obtain dmi -- - CI iri (8, - e, - ei). dt r We now rewrite this expression as dm, - = C [(Oi - 0:) - (6; - 6;) - (0; - 6;)] + C I,, [(O, - 0:) - (6, - 0;) - (Oi - Oi)] (7) dt r r in which superscript * denotes the state at the end of the fast process in which we have chemical but not subdivision equilibrium and superscript ’ denotes the partial equilibrium state with the same mi and C, as the perturbed state of interest. It is apparent from eqn (4) that (0:-0;) = OF is the same for all micellar species and that (O,*-O;) = a;.Consequently when the second term on the right-hand side of eqn (7) is negligible compared with the first this equation simplifies to give which is of linear phenomenological form with a driving force (OF-0:) and a phenomenological coefficient given by &. l,,. We note that RTC, I,, is the equilibrium rate at which i monomers enter or leave the micelles. Now according to eqn (5) the differences (OF-6;) and (O;-O,*) in eqn (8) can be expressed in terms of the perturbations in the mi and the q. Hence we have970 Phenomenological Approach to Micellisa t ion Kinetics where the prefix A denotes a quantity in the perturbed state of interest minus the corresponding quantity at the end of the reaction.The derivatives are taken at constant Cm because Cm is assumed not to change significantly during the fast process. Since the perturbations occur in a closed system we have for each constituent Consequently eqn (8) becomes There is an analogue of eqn (1 1) for each micellar constituent and the resultant set of equations lead to a set of relaxation times whose reciprocals are the eigenvalues of the matrix laik I, where aik is the coefficient of Am, in eqn (1 1). In general the derivatives in eqn (1 1) are not obtainable from equilibrium experimental data. They can, however, be estimated if we have detailed expressions for the dependence of chemical potentials on solution composition. For solutions of non-ionic surfactants and ionic surfactants plus salt it can be shown, albeit by some rather tedious manipulation, that the same expressions for the relaxation times result as were obtained previously1 when the same expressions for the chemical potentials are used.However, the present treatment is more general because it can accommodate situations which the previous treatment cannot. An example is the case where monomeric surfactant ions behave differently from inorganic ions of the same charge type and may specifically influence the behaviour of the micelles. The only significant approximation made above is that the second term on the right-hand side of eqn (7) is much smaller than the first. Whether or not this is so will depend to some extent on the perturbation concerned. For single-component non- interacting micelles Almgren et a1.4 have derived the conditions whereby this term is zero for arbitrary perturbations.A more general discussion of the issues involved for single-component micelles is outlined in the appendix. Application to Alcohol Solubilisation As a particular application of the above treatment we consider the kinetics of alcohol exchange between micellar and aqueous environments in solutions of a longish chain- length ionic surfactant (C16 or thereabouts) and a shortish chain-length alcohol such as pentanol. For solutions of this kind one expects two well separated relaxation times associated with the micelle-monomer equilibria with the faster time attributable primarily to the alcohol exchange and the slower time associated with the surfactant exchange. When dealing with the faster process it will be a good approximation to assume that the amount of micellar surfactant as well as the concentration of micelles remains constant.On the other hand one might expect the alcohol to be effectively in equilibrium with micelles during the slower process of surfactant exchange. Moreover, since the equilibrium concentration of surfactant monomer will be rather small the effect of the alcohol thereon can almost certainly be ignored. Under these circumstances the alcohol exchange kinetics are governed approximately by the expressionD. G . Hall 97 1 where i denotes the alcohol and the derivatives are taken at constant CF, mk and Cm, with k denoting the surfactant. In general the derivatives on the right-hand side of eqn (1 2) cannot be evaluated from equilibrium measurements.However, if the chemical potential of monomeric alcohol is given by 0; = 8:(T,p)+RTlnmd then evidently (aO~/Xr),i and (i30~/Clmi)c~ are both zero. Under these circumstances eqn (1 2) simplifies to give (13) which leads to a relaxation timel/zi given by Consider now the quantity (a@/aC ?),-p, cm. Although this quantity is not obtainable directly from equilibrium measurements what can be measured at equilibrium, for example from vapour pressures, is (aOF/aCF) Cz & because at equilibrium E = constant = 0. However, since and and according to eqn (5) it follows that Where the final term is necessarily positive for the system to be thermodynamically stable. C, lri is in effect the equilibrium rate of alcohol exchange between micellar and aqueous environments and can be expected to depelid on the amount of alcohol present in the system.Thus it makes sense to write E l r i = k q / R T r where the rate coefficient k may be regarded as the average time interval spent by an alcohol molecule in a micelle. Eqn (20) has already been used successfully to discuss the kinetics of pentanol exchange in solutions of cetyl pyridinium ch10ride.l~ When used in conjunction with the phenomenological treatment of relaxation data outlined previo~slyl~ it may be possible in favourable cases to estimate both the kinetic term Cr lri and the thermodynamic term (al:/aC,>,,, cp (aC,/aq),p, &. This latter term is closely related to the effect of the alcohol on the number of surfactant ions per micelle and vanishes when no effect of this972 Phenomenological Approach to Micellisation Kinetics kind occurs.When the kinetics of surfactant exchange are much slower than those of the alcohol it will be reasonable to treat these by using the formalism applicable to single-component micelles and regarding the influence of the alcohol as a medium effect. This viewpoint has much in common with the treatment of counterion effects described in ref. (1). The advantages of the present approach over alternative published treatments are clear. No assumptions concerning the form of the micelle population are required. Nor need it be assumed that the micelles are non-interacting. This contrasts with the treatment of Aniansson16 as applied by Yiv et al." and subsequently extended by Wall and Elvingson.ls Also, no particular forms need be assumed for the law describing the partitioning of the alcohol between micellar and aqueous environments or for the kinetic equations describing the exchange between the two environments.This contrasts with previous work based on a two-state m0de1.~~~ 2o Moreover, the present treatment does not invoke any assumptions which are not either explicitly or implicitly present in the alternatives. The above application is but a single example illustrating the power of the phenom- enological approach. Similar methods are applicable to more complex micellar systems, to binding by macromolecules and possibly also to macromolecular conformation changes. However, the treatment remains approximate and it is clear that further evaluation of the approximation inherent in eqn (8) is desirable.Linear Treatment of the Slow Process It is generally accepted that the slow relaxation process is concerned with equilibration of C, the concentration of micelles whilst the micelles remain effectively in equilibrium with monomer. In thermodynamic language this corresponds to the attainment of subdivision equilibrium under conditions of chemical equilibrium. According to the stepwise association model this process can only occur through the breakdown or formation of micelles by loss or gain of monomers. More recently evidence has been forwarded which suggests that micellar association and disproportionation are important in some For the stepwise association model we follow closely the arguments given by Aniansson and We begin by dividing the micelle population into three regions 1-111 where region I refers to small aggregates such as dimers, trimers, etc., region I11 refers to micelles proper and region I1 refers to rare intermediate aggregates whose concentrations are sufficiently small that their contribution to the overall stoichiometry of the system is negligible.For region I we write 22 In the discussion below both mechanisms will be considered. 0, = C Nr 0i (22) 0, = Z N ; ~ , + E . (23) i whereas for region 111 i For region I1 we write where e, = 0 at the boundary between regions I and I1 and E, = E at the boundary between regions I1 and 111. Together with eqn (1) eqn (24) gives for reactions in region I1 There is a strong analogy between this equation and Ohm's law as applied to a resistance between two points Y and s..Iri corresponds to the current, E, and E , correspond to the electrical potentials at r and s and corresponds to the resistance. Moreover,D. G. Hall 973 when the steady-state approximation applies to region 11, so that it is legitimate to write for species in this region then the rate of formation of micelles of type r from smaller micelles is equal to the rate at which they react to give bigger micelles. This gives us a series of equations between the Jri analogous to Kirchoff's laws. It is clear therefore that for multicomponent micelles the behaviour of region I1 during the slow process is analogous to that of a network of resistances connecting a region of potential E with a region of zero potential. Hence we may write where - J corresponds to the total current through the network, E corresponds to the total potential drop across it and 1 is the overall resistance of the network calculable by standard methods from eqn (25) and (26).The negative sign in eqn (27) arises because a positive E favours a decrease in C,. Each micellar species in the network corresponds to a junction of resistances, and if there are c micellar constituents then each species connects to c larger species and n smaller species where n, its number of constituents, is < c. dC,/dt = 0 (26) J = dC/dt = - I & (27) The case of single component micelles is very straightforward. We have simply J , = - E,(E, - E , .- ~ ) (28) but by virtue of eqn (28) J , = J is the same for all r in region 11. Hence J/lr = - ( E , - E , - ~ ) and we obtain on summing over all r in the region concerned J = - [z (1,)-']-' E = (%). r This expression gives for the slow relaxation time. which is a general form of the result obtained previously for ionic surfactants in the presence of supportingelectrolyte. It is easily shown that eqn (3 1) reduces to the expression derived by Aniansson and Wall for the particular case they consider, namely ideal systems which conform to mass-action kinetics with rate coefficients that depend only on T and p . However, the above derivations do not rely on any particular form for either the kinetic equations or the equations describing the dependence of chemical potentials on solution composition.In particular, no assumptions concerning ideality are required. Consider now eqn (30). This equation is of the linear phenomenological form with overall driving force E and a phenomenological coefficient which is a combination of the various I,. The overall coefficient ( X , l ; l ) is in effect the net resistance of a set of resistances in series. However, it also turns out that RTE, l;l)-l may be regarded as the equilibrium rate of micelle formation or breakdown. This latter statement may be justified as follows. Consider an overall reaction which proceeds through a series of intermediates via a single route with each step having a stoichiometry number of unity. When the steady-state approximation applies the overall forward and backward rates for such a reaction may be expressed in terms of the rates of the individual steps by n yf.(-1 D r,=n" (32 a) 33 (32 b) FAR 1974 where i refers to the ith step, n is the total number of steps and Phenomenological Approach to Micellisation Kinetics n n n-1 i-2 i-3 i-1 D = n r:+rF n r: ... n r t . At equilibrium when re = rp for all i we find that rf and rb become (33) Clearly, since RTl,. is the equilibrium forward or backward rate of the reaction (r - 1)-mer + monomer s r-mer then R T P (1;l)I-l may be taken as the overall equilibrium rate of micelle formation or breakdown as suggested above. In some cases it may be legitimate to apply the expressions derived above for single-component micelles to mixed micelles. Consider for example a mixture of two surfactants i and j in which i has a much greater c.m.c.than j . The equilibrium micelle-monomer exchange rate of i can be expected to be correspondingly greater than that o f j in both single-component and mixed micelles. Moreover, this should apply for all aggregation numbers and micelle compositions, including the rare intermediate region. Consequently, for a given flux between two adjacent species in this region the driving force will be much less when the two species differ by a molecule of i than when they differ by a molecule ofj. Under these circumstances micelles with the same number o f j molecules all have approximately the same value of E irrespective of the amount of i they contain. In other words the change in E between regions I11 and I is associated almost entirely with the loss o f j molecules.In effect this is equivalent to treating i as a component of a mixed solvent which is assumed either not to participate in the processes under study or to be in complete equilibrium throughout the system during the timescales of interest. To deal with the situation where C, can change through the association or disproportionation of aggregates we may proceed as follows. Consider the reaction r-mer + s-mer + (r + s)-mer where r and s refer to micellar species in region 111. Let the net rate of the reaction be J,,. At chemical equilibrium we have, according to linear phenomenological theory, where RTl,., is the equilibrium rate of the above reaction and E is the driving force. In the case that r and s are species in region I and ( r + s ) is a species in region I11 we have J,., = -lrs E However, whereas the reaction described by eqn (35) corresponds to the loss of a micelle by region I11 the reaction described by eqn (36) corresponds to the gain of a micelle by region 111.Hence the overall rate of change in C , due to aggregate association and breakdown may be written as dC,/dt = -x I,, E rs (37) where the rs pair may be both in region I or both in region 111. Reactions for which r is in region I and s in region I11 have zero driving force under quasi-equilibrium conditions, and in any event do not contribute to changes in C,, the concentration of aggregates in region 111. Reactions involving micelles in region I1 are considered as sufficiently rare to ignore.If we now combine eqn (27) and (37) we obtain for the overall rate of change in C , due to both mechanisms acting together dC,/dt = -(l+x lrs) E . rsD. G . Hall 975 The corresponding relaxation time is given by (39) 1 a& - = ( l + ~ 1,s) (-) - r2 rs a c m ~i Eqn (38) and (39) apply to single- and multi-component micelles alike. For the former eqn (39) is a general and properly justified form of the expression presented previously by Kahlweit and 22 For reasons which are not altogether clear Kahlweit and coworkers do not consider reactions between micelles in region 111. However, there seems to be no obvious reason why they should not be included. Eqn (38) suggests strongly that the driving force for changes in C , at chemical equilibrium is the subdivision potential E irrespective of the mechanism whereby the changes in C, take place.Moreover, for single-component micelles at least the coefficient of E is the equilibrium rate of micelle formation and breakdown. However, it is debatable whether or not talking about reactions such as r-mer + s-mer (r + s)-mer is strictly sensible in the present context. Since micelles are in dynamic equilibrium with monomer and the exchange processes concerned are rapid the contents of a given micelle may change appreciably during a reactive encounter with another micelle, especially if the timescale of such encounters is as slow as the values of zz suggest. Coupling between Fast and Slow Processes In the discussion so far we have assumed, in common with Aniansson and coworkers and others, that C , is strictly constant during micelle-monomer exchange and that micelles and monomers are strictly in chemical equilibrium throughout the slow process during which subdivision equilibrium is attained.Clearly these results are not strictly valid and the equations which result from making them are approximate. On intuitive grounds one expects these approximations to be valid when the relaxation times concerned are widely separated. Under these circumstances any changes in Cm which take place during the fast process should be much smaller than those which occur during the slow process. Also the deviation from chemical equilibrium during the slow process should be much smaller than that typically associated with the fast process. These intuitive expectations are supported by some recent model calculations of Kegeles.2s However, there may be situations in which the above approximations are no longer valid.Two possible examples are (a) solutions of shorter chain-length surfactants in which average aggregation numbers are small and species in region I1 are not that rare and (b) solutions of surfactants in apolar media in which small reversed micelles can form. In this section we derive expressions, albeit approximate, which do allow for coupliag between the two processes. For simplicity we consider only single-component micelles and as before take as our starting point J , = - lr(8, - 8,-, - 8,) (40) where J , is the net rate of the process ( r - 1)-mer + monomer --+ r-mer. For the micelles proper in region 111 we suppose as above that ( Q r - @ r - A = ern is the same for all r and that e, = r0, + E .For region I we suppose that premicellar aggregates such as dimers and trimers are in equilibrium with monomer so that 8, = re,. (43) 33-2976 For region I1 we suppose that Phenomenological Approach to Micellisation Kinetics (8, - er-, -el) = A,. (44) Let this region consist of (n+ 1) species and let t denote that species with the highest aggregation number. We suppose as before that the steady state approximation holds in this region so that dCm= J = J r = l r A r (45) dt where Cm denotes the concentration of aggregates in region 111. Summing over the n steps within region I1 we find that since 8, = t8, +E and Ot--n = (t -n) 8, eqn (46) gives If, in this context, we regard m, as the total concentration in region I then we have for the rate of change of monomer concentration dm, = - C J,-nJ-(t-n)J dt 111 where the summation includes all the species in region 111.The second term on the right-hand side of eqn (48) accounts for the n steps in region I1 and the final term allows for the fact that when a member of species (t-n) breaks down this contributes (t-n) monomers to region I. Substituting for Jr as given by eqn (40) and (41) and for J a s given by eqn (47) we obtain Eqn (47) and (49) constitute two coupled linear phenomenological equations expressing dml/dt and dC,/dt in terms of the two driving forces (8, - 8,) and e. It is noteworthy that the cross-phenomenological coefficients are equal and thus satisfy the Onsager reciprocal relati~nship.~ Clearly the two equations are approximate and it is not immediately obvious whether or not the approximations involved could lead to discrepancies between theory and experiment which may outweigh the effects that arise from the coupling.The equations lead to two relaxation _times. To obtain expressions for these we write (8,--8,) and E in terms of the perturbations in ml and C,, i.e. substituting these expressions into eqn (48) and (49) we obtain dAm, dt dACm dt = -all Am, -alrn AC, - -aml Am1-ammACm --For solutions of non-interacting aggregates calculation of the thermodynamic terms in the above expressions is straightforward; also, when a,, > a,, eqn (53a) and (53b) become equivalent to the expressions which result when coupling between the fast and slow processes is ignored.Non-linear Treatment of Fast and Slow Processes For many systems of practical interest the perturbations from equilibrium are too large for the linear theory described above to apply. Hence it is important to establish to what extent one can develop fairly general arguments in the non-linear regime. A suitable starting point for such procedures is the expression where Jfi and J:, are the forward and backward rates of the reaction r + i e s . Eqn (54) is equally applicable to ideal and non-ideal solutions.ll It leads to the result Eqn ( 5 5 ) is a general non-linear form of eqn (1) and becomes identical to this expression when it is legitimate to linearise the exponential. Fast Process As before which together with eqn (55) gives (56) (57)978 When (@, - e,) = Oy eqn (57) becomes Phenomenological Approach to Micellisa t ion Kinetics For single-component micelles in ideal systems it is probably quite a good approximation to put where k, is the same for all micellar species.This gives J:i = k, Cr-,m1 (59) dm, dt = kf Cmm, ( e x p e - RT 1). However, and in this case (em - el) = (em - e*) - (0, - e*) (6, - O*) = RT In m,/m,* where m;" is the quasi-equilibrium value of m,, which may be regarded as a known quantity. Together with eqn (60) eqn (62) gives now and e -e* = k C m , * e x p L - dt m( RT ml) (lm-~*=~'' (%) dCf dm, cfn* aCf cm Thus if o2 is known as a function of at constant Cm the integral on the right-hand side of eqn (64) can be expressed in terms of the departure from equilibrium and we obtain a non-linear first-order differential equation relating dm,/dt to some function of m,.However, it should be stressed that this equation is approximate because of the assumptions that (8, - Or-,) and k$ are the same for all micellar species. Slow Process Stepwise Association Mechanism As above we divide the aggregate population into three regions. In region I1 during the slow process we have J',i/J:i = exp [(E,-&,)/RT] (66) where so that It follows that When the steady-state approximation is valid eqn (18) is in some ways analogous to eqn (25). Again JTi is analogous to a current and exp &,/kTis analogous to a potential at some junction Y. The term J:% exp-&,/kT plays the same role as lri. For solutions in which allD.G. Hall 979 monomer and aggregate species behave ideally JFi = kFi C, where k,b, is the rate constant for the loss of an i molecule by species s and C, is the concentration of species s given (69) by where C; is the concentration of species s in the solution at complete equilibrium in which the monomer concentrations are the same as those in the reacting system of interest. Hence J:i exp (-&,/kT) has the significance of an equilibrium rate. Clearly the non-linear case can be regarded as analogous to a network of resistances in much the same way as the linear case. C, exp - (&,/RT) = CE For micelles of a single component eqn (68) becomes dCm = J = J , = J ; exp - ( E , / R T ) [exp (&,-,/RT) - exp (&,./RT)I (70) dt which is the non-linear analogue of eqn (28).Dividing both sides by J : exp-(&,/RT) and summing overall r in region I1 we obtain J C (J,b)-l exp (&,/RT) = [ 1 - exp (e/RT)] r d C m (71) (72) so that J = - - [C (JF)-l exp (&,/RT)]-l[ 1 - exp (&/RT)]. By use of eqn (31), (33) and (66) it is straightforward to show that the term Er (J;)-l exp (&,/RT)]-l exp ( E / R T ) is the rate of micelle breakdown. If we denote this quantity by J k and the corresponding quantity for the rate of micelle formation by JL then it also follows that dt r = exp (&/RT). (73) J‘, Indeed these results are particular cases of a more general relationship described elsewhere .117 25 For ideal systems eqn (72) gives J=---- dCm = [c (k: C;)-l]-l[ 1 - exp (&/RT)] (74) exp (&/RT) = Cm/C& (75) dt r but in this case where C& is the concentration of micelles in a system at complete equilibrium with the same monomer concentration as in the reacting system.Hence However, since Cm = (C, -m,)/N and N = X(m,) under the conditions of interest it follows that the right hand side of eqn (76) is a function of m, only. The left-hand side may now be written as (aCm/am,)cl dm,/dt and it is straightforward to show that Hence we obtain finally where C, is the value of C in the system at complete equilibrium with the m, concerned.980 Phenomenological Approach to Micellisation Kinetics Eqn (78) expresses dm,/dt as a function of rn, only and can be integrated to obtain rn, and hence C, as functions o f t when explicit expressions in terms of rn, for the various quantities on the right-hand side are available.When micelle association and disproportionation can occur we have instead of - - eqn (36) J&/JPs = exp(&/RT) which gives Jrs = J,", [exp (c/RT)- 13. Eqn (79) and (80) apply to reactions between micelles in For association of micelles in region I to give micelles Jf,,/J& = exp ( - e/RT). (79) (80) region 111. in region I11 we have (81) The contribution of micellar association to dC,/dt is thus given by I l l I - C Jrs +C J r s = C J:s [exp (&/RT) - 11. rs rs rs For ideal systems Hence JFs = kFs C(r+s) = k;s Cyr+s) ~ X P (&/RT)- dCm = - Z kps C&+s) exp (&/RT) [exp (&/RT) - 11 dt TS = -C kf,, CF CE exp ( E / R T ) [exp (&/RT) - 11 (84) TS where CF C; and C&+,, are the concentrations of the species concerned in a system at total equilibrium where the monomer concentrations are the same as in the reacting system of interest.Combining eqn (72) and (82) we find that when both mechanisms for changes in C, where J k is the rate of micelle disappearance. In this context it must be remembered that the reaction rmer + Smer + (Y +s)-mer constitutes micelle disappearance when proceeding from left to right. Concluding Remarks The above examples illustrate well the power and generality of phenomenological methods in describing the kinetics of micellisation. A particular advantage of the treatment is that the thermodynamic and kinetic terms can be handled separately. Hence the treatment is not restricted to any particular model for the equilibrium thermodynamic behaviour.In some respects the approach is more convenient than traditional methods based on mass-action kinetics. This arises primarily from the fact that the rates of the individual steps may be written as a single term rather than as the difference between two terms. The approach is also more general than previous treatments, and indeed all the significant results obtained previously follow straightforwardly when appropriate expressions for the thermodynamic and kinetic terms are inserted into the general equations. A number of important advances have been made. These include (i) the treatment of mixed micelles which leads to expressions that enable equilibrium thermodynamicD. G . Hall 98 1 data to be used effectively in the interpretation of relaxation experiments, (ii) the demonstration that the subdivision potential is the driving force for changes in the micelle number during the slow process both for the stepwise association mechanism and the mechanism of aggregate association and disproportionation, (iii) the recognition that the phenomenological coefficient for the slow process may be regarded as the equilibrium rate of micelle formation or breakdown divided by RTand (iv) the derivation of equations which allow for coupling between the fast and slow processes.In the last section perturbations from equilibrium which are too large to handle using the linear formalism were considered. As might have been expected the general formal expressions obtained are difficult to handle without making considerable simplifications.However, they do provide a basis for further work. Appendix The Approximation in eqn (8) The approximation in eqn (8) consists of replacing the micelle population in the real relaxing system with that population which minimises g for the same values of the mk, Cp and C,. For single-component micelles we may argue quantitatively as follows. Eqn (6) takes the form where 1, is the phenomenological coefficient for the reaction (r - 1)-mer + monomer - A r-mer. Eqn (Al) may be written in the form where All derivatives are taken at constant T, p and C, and all refer to that state of the system at the end of the fast process. C, denotes that all C, other than the differentiating variable are held constant. Let us regard X , as a continuous function of r, the number of surfactant molecules per micelle.We may replace the summation in eqn (A2) by an integral so that m C, X , ACr = 1 X(r) (CT- C:) dr r rmin where the lower limit of integration corresponds to the smallest aggregation number we may reasonably call micellar. We may now expand X(r) as a Taylor series about some value of r , r+ which we may if we wish take as the most probable value of r. This gives982 so that together with eqn (A4) we obtain Phenomenological Approach to Micellisation Kinetics +!(? 2 dr2 rPr+ s(r--r+)2(Cr-Cf)dr+. . .. (A6) Since Cm = C, C, is constant it follows that the first term on the right-hand side of eqn (A6) is zero. Also [r(Cr - Cf) dr = C, AN (A71 where Axis the perturbation in the mean aggregation number and P J r2(Cr - Cf) dr = C, AN2 where A F is the perturbation in the mean-square aggregation number.Hence we may now write Z X r A C r = [ ( $ ) , r -r+ -r+(g) r-r+ ]C,AX + 1. (ET 2 dr2 ,.,,.+ Cm A p + higher-order terms. (A9) Consider now the second term on the right-hand side of eqn (7). For the case of interest this becomes lr[(6r-%)-(or-i -%-i)-(6i -K)I and may be written as I= xr AC; r where AC; = (C, - Ci). Using the same procedures as above we may now write I: X,.AC; = X+ s (Cr-Ci)dr+(9 ar r-T+ [(r-r+)(C,.-Ci)dr +-!(?? 2 ar2 r-r+ [(r-r+)2(Cr-Ci)dr. (A10) However, in this case Cm and XC, are the same for both populations so that (Cr- C;)dr = 0 s and Hence s (r - r+)(C,. - Ci) dr = 0. 1 a2x X,. AC; = - (-) s r z (C,. - C;) dr + higher-order terms r 2 ar2 ,=,.+ = 1 (E) C , [F - (F)] + higher-order terms.(A1 1) 2 ar2 ,.-,.+D. G. Hall 983 It follows therefore that eqn (8) is a good approximation for single-component micelles when a2X/i3r2 and higher-order derivatives are all very small. To show the equivalence of this discussion with that of Almgren et aL4 we note that for non-interacting micelles (aO,/aC,) = 0 unless r=s, in which case (aO,/aC,) = RT/C,. Thus RT RT xr = 1, - - 1,+, -. Cr Cr k,b Cr - kf, Cr-1 m, 1, = - RT RT However where kf, and k,b, respectively, are forward and backward rate coefficients for the reaction ( r - 1)-mer + monomer r-mer. Hence X , = k: - kf,,, m,. (A131 Since m,, the equilibrium value of the monomer concentration, depends on the total concentration, the condition that the second and higher derivatives of A',.with respect to r vanish under all circumstances (i.e. for arbitrary perturbations at all concentrations) is that kb and kf are both linear functions of r . This is the result given by Almgren et a1.4 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 D. G. Hall, J. Chem. Soc., Faraday Trans. 2, 198 1, 77, 1973. E. A. G. Aniansson et al., J. Phys. Chem., 1976, 80, 905. E. A. G. Aniansson and S. N. Wall, J. Phys. Chem., 1974,78, 1024; 1975,79, 857. M. Almgren, E. A. G. Aniansson and K. Holmoker, Chem. Phys., 1977, 19, 1. R. Haase, Thermodynamics of Irreversible Processes (Addison Wesley, Reading, Mass., 1969). D. G. Hall, Trans. Faraday Soc., 1971, 67, 2516. D. G. Hall, J . Chem. Soc., Faraday Trans. 2, 1978, 74, 405. D. G. Hall, J. Chem. Soc., Faraday Trans. 2, 1981,77, 1121. D. G. Hall, Z. Phys. Chem. N.F., 1982, 129, 109. G. W. Castellan Ber. Bunsenges. Phys. Chem. 1963, 67, 731. D. G. Hall, J. Chem. SOC., Faraday Trans. 2, 1986, 82, 1305. D. G. Hall, C , of Aggregation Processes in Solution, ed. E. Wyn-Jones and J. Gormally (Elsevier, Amsterdam, 1983). T. L. Hill, Thermodynamics of Small Systems (Benjamin, New York 1963 and 1964), vol. 1 and 2. J. Gormally, B. Sztuba, E. Wyn-Jones and D. G. Hall, J. Chem. SOC., Faraday Trans. 2, 1985,81,000. D. G. Hall, J. Gormally and E. Wyn-Jones, J. Chem. Sac., Faraday Trans. 2, 1983, 79, 645. E. A. G. Aniansson, in Techniques and Applications of Fast Reactions in Solution, ed. W. J. Gittins and E. Wyn-Jones (D. Reidel, Dordrecht, 1979), p. 249; E. A. G. Aniansson, in Aggregation Processes in Solution, ed. J. Gormally and E. Wyn-Jones (Elsevier, Amsterdam, 1983), chap. 3. S. Yiv, R. Zana, W. Ulbricht and H. Hoffman, J. Colloid Interface Sci., 1981, 80, 224. S. Wall and C. Elvingson, J. Phys. Chem., 1985, 89, 2695; C. Elvingson and S. Wall, J. Phys. Chem., 1986, 90, 5250. D. G. Hall, P. L. Jobling, J. E. Rassing and E. Wyn-Jones, J. Chem. SOC., Faraday Trans. 2, 1977, 73, 1582. J. Gittins, D. G. Hall, P. L. Jobling, J. E. Rassing and E. Wyn-Jones, J. Chem. Soc., Faraday Trans. 2, 1978, 74, 1957. E. Lessner, M. Teubner and M. Kahlweit, J. Phys. Chem., 1981, 85, 3167. M. Kahlweit, J Colloid Interface Sci., 1982, XX, 90, 92. M. I. Temkin. Dokl. Akad. Nauk SSSR, 1963, 152, 156. M. I. Temkin, Int. Chem. Eng., 1971, 11, 709. M. Boudart, J. Phys. Chem., 1976,80, 2869. G. Kegeles. J. Colloid Interface Sci., 1984, 99, 153. Paper 61024; Received 3rd January, 1986

ISSN:0300-9599    

DOI:10.1039/F19878300967

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. I , 1987, 83, 985-1006 The Kinetics of Solubilisate Exchange between Water Droplets of a Water-in-oil Microemulsion Paul D. I. FletcherJ Andrew M. Howet and Brian H. Robinson9 Chemical Laboratory, University of Kent, Canterbury, Kent CT2 7NH Exchange rates of aqueous solubilisates between water droplets in a water-in-oil microemulsion stabilised by sodium bis(2-ethyl-hexyl) sulpho- succinate (AOT) have been measured as a function of droplet size, tempera- ture and the chain length of the oil. The effects of additives (e.g. alcohols) on the exchange kinetics have also been investigated. Exchange rates were measured using very fast chemical reactions as indicators for exchange. Three types of reaction were investigated : proton transfer, metal-ligand complexation and electron transfer.Similar exchange rates were found for all three reactions. The indicator reactions involve the exchange of reactant ions of differing size and charge type; exchange rates were, however, independent of the ion transferred, but dependent on droplet size and temperature. For AOT as dispersant, exchange occurs with a second-order rate constant of lo6-los dm3 rno1-I s-l, two to four orders of magnitude slower than the droplet encounter rate as predicted from simple diffusion theory. The apparent activation enthalpy is high (and increases with droplet size) but is largely compensated by a positive activation entropy. Exchange, on balance, is a relatively facile process which typically takes place on a millisecond timescale (depending on the droplet concentration).The exchange mechanism involves transient water droplet coalescence and separation. This is the dynamic process whereby the equilibrium properties of the microemulsion, e.g. droplet size and polydispersity, are maintained. There is a correlation between the exchange rate constants and the stability of the single-phase microemulsion. This relationship between the kinetic and equilibrium properties is discussed in terms of the ‘natural curvature’ of the surfactant interface and inter-droplet interactions. Much of the interest in microemulsions is concerned with their unique properties as a reaction medium. Reaction rates and equilibria may be altered by several orders of magnitude as compared with the corresponding values in homogeneous solution.1-3 In particular, the use of microemulsion water droplets as a novel environment for enzyme-catalysed reactions has attracted much interest4 and the potential for controlled and selective synthesis is now being ~ealised.~-’ For reaction in a water-in-oil micro- emulsion involving reactant species totally confined within the dispersed water droplets, a necessary step prior to their chemical reaction is transfer of reactants into the same droplet. When chemical reaction is fast (close to diffusion-controlled), the overall reaction rate is likely to be controlled by the rate of inter-droplet transfer of reacting species (so-called ‘ solubilisate exchange ’). In this paper data for exchange kinetics involving a range of reactant ions in water droplets stabilised by sodium bis(2-ethyl-hexyl) sulphosuccinate (AOT) in hydrocarbon solvents are reported.The inter-droplet transfer rate is studied as a function of droplet size and temperature. In addition, the hydrocarbon oil (dispersion medium) has been systematically varied, and the effects of surface-active additives have been studied. t Present address: Department of Chemistry, University of Hull, Hull HU6 7RX. 1 Present address: AFRC Food Research Institute, Colney Lane, Norwich NR4 7UA. 4 Present address: School of Chemical Sciences, University of East Anglia, Norwich NR4 7TJ. 985986 Kinetics of Solubilisate Exchange The AOT-oil-water microemulsion system is particularly suitable for systematic investigation since no cosurfactant is required. AOT itself forms a reverse hexagonal liquid-crystalline phase.With water and oil, an extensive L2 phase is present. The microemulsions are thermodynamically stable (they form spontaneously and droplet size is independent of variations in preparation method). Relatively large amounts of water (ca. 100 mol of water per mol of AOT) may be solubilised in n-heptane at room temperature. This figure does not change significantly with AOT concentration over the range 0.01-0.5 mol dm-3. Water is dispersed in the form of microdroplets; the sizes have been measured using a variety of techniques including ultracentrifugation,8 photon correlation spectroscopy (PCS),g~ lo small-angle neutron scattering (SANS),11-15 membrane diffusion,16 fluorescence dep~larisationl~ and time-resolved fluoresence quenching.ls The thickness of the AOT interfacial layer has been determined from SANS data to be 0.9 nm, which indicates only slight penetration of the AOT alkyl tail region by the alkane s01vent.l~ Shape fluctuations may occur, but the assumption of sphericity gives consistent results by different experimental techniques.Water droplet radii can be systematically varied from 0 to 20 nm, being proportional to the molar ratio water : AOT (R). Altering the amount of water and AOT but keeping R constant changes propor- tionally the droplet concentration but not droplet size. The droplet size may be calculated to a first approximation using the equation:1° (1) where r (hydrodynamic radius)/nm = 0.175R+ 1.5 R = [H,O]/[AOT]. Single-phase microemulsions are stable over a restricted temperature range which depends on composition.At high R values the range is only a few degrees, whereas low R value systems are much more stable.12 The temperature corresponding to the maximum water uptake at constant [AOT] may be equated with the phase inversion temperature (p.i.t.),,O The value of the p i t . is dependent upon the oil used as continuous solvent, the presence of surface-active additives and the ionic strength of the aqueous component. For the single-phase AOT microemulsion systems, decreasing the temperature below the lower temperature boundary produces a reduced water content microemulsion phase (i.e. lower R value) and a conjugate (almost pure) water phase. Increasing the temperature beyond the upper temperature stability limit does not produce separation of an 0-W microemulsion and a conjugate oil phase as observed for mixtures containing comparable volumes of oil and water.20 The W-0 microemulsion phases contain only small amounts of water, which leads to a more complex phase separation. For systems containing short-chain alkanes (e.g.heptane) phase separation produces a liquid- crystalline phase (probably lamellar in structure containing AOT and water) with a conjugate oil phase. For higher alkanes ‘critical’ phase separation into two W-0 microemulsion phases of different droplet concentrations is observed. This phase separation is driven by increasingly attractive inter-droplet interactions as the phase boundary is approached. l4 The phase separation sequence (i.e.water + AOT-oil below the p.i.t. and oil + AOT-water above the p.i. t.) reflects increasing hydrophobicity of AOT as the temperature is decreased. In aqueous micellar solutions, surfactant molecules exchange rapidly between micelles and bulk aqueous solution and the micelles disintegrate/reform over a longer timescale.21 In contrast, very little is known about the dynamic processes which occur in water-in-oil microemulsions. Eicke et al. demonstrated exchange of hydrophilic solutes between droplets by a droplet collision mechanism on a timescale shorter than a few seconds22 and this was later confirmed.8 Atik and Thomas23 and Fletcher and Robinson24 measured the second-order rate constant for exchange in AOT microemulsions and obtained rate constants two to three orders of magnitude less than the diffusion-controlled limiting value.A mechanism for solute exchange was postulated whereby, in an initial step, twoP. D. I. Fletcher, A . M. Howe and B. H. Robinson 987 droplets collide to form an encounter pair. In a small fraction of the encounters, droplets fuse to form a short-lived droplet dimer which then decoalesces. The two droplets then separate with concomitant randomisation of solute occupancy together with surfactant and water.24 Such a mechanism forms a basis for understanding the mechanism whereby microemulsions spontaneously form and break. (A microemulsion system that does not exchange cannot form spontaneously.) Equilibrium properties such as particle size and polydispersity are maintained in rapid dynamic equilibrium via this mechanism, which is therefore expected to be intimately related to the phase behaviour of the system.In view of the fundamental nature of exchange in determining microemulsion properties, this paper is concerned with a systematic investigation of the kinetics of the inter-droplet exchange process and its relation to phase behaviour. Experiment a1 AOT (Sigma) (supplied as dioctyl sulphosuccinate, sodium salt) was used without further purification. AOT is a di-ester and some samples (from other suppliers) contain hydrolysis Batches used in this work were titrated using an acid-base indicator25 and were found to contain negligible acidic impurities. Interfacial tension measurements with Sigma AOT show it to be free of an impurity (presumably the alcohol hydrolysis product) which causes a minimum in interfacial tension vs.surfactant concentration plots for other AOT samples.? n-Heptane (Fison’s SLR grade) was distilled over sodium metal, stored over type 4A molecular sieve and filtered before use. Water was de-ionised and doubly distilled. Murexide (ammonium purpurate) was a B.D.H. laboratory reagent. Purity was checked spectrophotometrically (extinction coefficient at 522 nm = 1.4 x lo4 dm3 mol-1 cm-l).* Murexide solutions were used im- mediately after preparation since they fade slowly. Decomposition is accelerated in acidic solutions and provides a further check on acid impurity levels. 4-Nitrophenol-2- sulphonate (disodium salt) (NPSA) was obtained from the Alfred Bader Library of Rare Chemicals (Aldrich). Ammonium hexachloroiridate(1v) (99.9 % ) was obtained from Aldrich and potassium ferrocyanide from Fisons.Bis-bipyridyl bis-cyanoiron(I1) was generously supplied by Dr J. Holzwarth (Fritz Haber Institute, Berlin). All other reagents were of the highest grade available. U.v.-visible absorption spectra were recorded using a Cary 219 instrument. The sample chamber was thermostatted to within 0.1 K of the desired temperature. Kinetic measurements were made using a small-volume stopped-flow instrument designed and built in this laboratory. The dead time was 1 to 2ms.19 Some kinetic measurements involving electron-transfer reagents were obtained using the continuous-flow method with integrating observation (CFIO). This equipment was used at the Fritz-Haber Institute, Berlin and was made available to us by Dr J.F. Holzwarth. The technique is valuable for kinetic studies in the sub-millisecond time range.2s Microemulsion-phase stability maps were determined as follows : mixtures of AOT, oil and water in which the R value was changed systematically, were contained in tightly stoppered flasks, which were stored with occasional shaking in thermostatted baths for several days. Solutions which were, by visual inspection, transparent or faintly bluish (at high R values), but which showed no sign of sedimentation, were taken to be one-phase microemulsions. Theory of Kinetic Measurements Solubilisate exchange rates were measured using fast chemical reactions where the product formation rate is limited by the inter-droplet exchange rate.To use the observed kinetics of the reaction to measure the exchange rate the following criteria must be t Interfacial tensions were measured by Dr J. Mead, University of Hull.988 Kinetics of Solubilisate Exchange obeyed. First, reactant species must be confined within the water droplets to exclude the possibility of passage through the continuous oil solvent. This is ensured by the use of hydrophilic, charged reagents which show no solubility in the oil. Secondly, solubilisate exchange must be the rate-determining step of the reaction. This may be ensured by (i) the use of intrinsically fast reactions such that kchem is close to being diffusion-controlled and (ii) by decreasing the solubilisate exchange rate by lowering the concentration of water droplets.For reactants A and B confined within the water droplets the reaction scheme may be described by the limiting cases (a) and (6). Scheme (a). Reaction in dimer Scheme (b). No reaction in dimer 'Ikchem 0 The circles around the species symbols indicate that the species are located within the water droplets. The larger circles represent a short-lived droplet dimer of lifetime 1 /kdiss. ke, is the second-order rate constant for transfer of species A and B into the same droplet and kchem is the rate constant for chemical reaction [expressed as a second-order rate constant (dm3 of water mol-l s-l)]. When conditions (i) and (ii) above pertain, then d [ o ] / d t = k[@][@] where k = k,, in units of dm3 (of total solution) mol-1 s-l. The essential difference between scheme (a) and scheme (b) is the following.If chemical reaction occurs as in scheme (a), then 0*7/kdiss ' 2/(kchem [Alwd) (2) where [AIwd is the aqueous concentration of one molecule of A inside a single droplet (i.e. [A]wd = ( N x the volume of water in one droplet)-'). This is equal to 0.03 mol dm-3 for one reactant molecule in an R = 10 droplet. In eqn (2) the factor 2 arises since the volume of water in the droplet dimer is twice that of a separated droplet. Also in eqn (2), reaction involves one molecule of A reacting with one molecule of B inside a droplet. Hence, [AIwd is equal to [BIwd. Alternatively, if 0*7/kdiss < /(kchem[Alwd) (3) reaction occurs in a monomer droplet following decoalescence of the transient dimer [scheme (b)]. The chemical reactions we have used as indicators for the exchange process are sufficiently fast that scheme (b) is unlikely to operate, but there is a maximum statistical factor of two on the overall exchange kinetics depending on which scheme is applicable.Three types of fast reaction were investigated as shown below. Electron Transfer Ir(C1):- + Fe(CN)t- 3 Ir(C1):- + Fe(CN)g- (9 (ii)P. D. I. Fletcher, A . M. Howe and B. H . Robinson 989 (bp is the ligand bipyridyl.) Values of k, in aqueous solution have been measured by Holzwarth et al.27 For reaction (i) at I(NaC1) = 0.1 mol drn-,, k, = 5 x lo7 dm3 mol-l s-l; at For reaction (ii) at I(NaC1) = 0.5 mol dm-3, k , is 1 x lo9 dm3 mol-l s-1.2s Both electron- transfer reactions are essentially irreversible. Reactants were assumed to be exclusively located inside the water droplets.I(NaC1) = 1.0 mol drn-,, k, = 5 x lo8 dm3 mol-1 s-l. Proton Transfer 0- OH NO2 NO2 The pK, of HNPSA- (measured spectrophotometrically) is 6.80f0.05 at 25.0 "C in aqueous solution at I = 0. Using the joule-heating temperature-jump method the observed relaxation time [at pH 6.15, 25 "C, ionic strength 0.3 mol dm-, (NaNO,), indicator concentration mol drn-,] was faster than the heating time of the instru- ment (ca. 5 ps). Hence, k, must be greater than 1O1O dm3 mo1-l s-l (i.e. close to the diffusion-controlled limit). (k, for 4-nitrophenolate is 3.8 x 1O1O dm3 mol-l s-l and k, is lo3 s - ~ . ~ ~ ) The solubilities of NPSA2- and HWPSA- are negligible in heptane but greater than mol dmd3 in water, so that the reactants are likely to be confined within the droplets.Metal-Ligand Complexation Complex formation between zinc(aq) and murexide (ammonium purpurate) was studied. Zn2++ Mu- \ I / Zn+ Maas reported a value of (2.8f0.3) x lo7 dm3 mol-1 s-l for k, and (4.5k0.2) x lo4 s-l for k , in water at 20 OC.,O Again, both reagents show no solubility in heptane, but a high solubility in water, so they are confined within the droplets. Calculations of k,, from Kinetic Data A microemulsion solution of reactant A was rapidly mixed in a fast-flow apparatus with a microemulsion solution containing reactant B. The reaction was monitored spect rop ho tometrically. For electron-transfer reactions the kinetic analysis is straightforward. These reactions are irreversible, so low concentrations of reactants can be used and only singly occupied droplets need be considered.Thus, the reaction scheme is: A1 + Bl -+ products990 Kinetics of Solubilisate Exchange where A1 and B1 represent droplets containing single molecules of A and B. The experimentally determined second-order rate constant may then be equated directly with k,,. A knowledge of droplet concentrations is not required to obtain values of k,, provided that : (4) For proton transfer and metal-ligand complexation, analysis is more complex since these reactions are not irreversible. In general, for a process A + B -P C occurring within the water droplet, the situation may be represented by the following scheme: overall concentrations of A and B + [water droplets]. x, Y, 2-0, 00 Since the reactions proceed to equilibrium, higher reactant concentrations are required to achieve a detectable level of product formation such that each water droplet contains a low integral number of reactant molecules. (A droplet containing n molecules of species A is represented by An.) The reaction scheme represented by eqn (5) includes all product-formation reaction steps such as A1 + BI -+ Cl + 0 (where 0 represents an ‘empty’ droplet), A2+B1 -+ C1 +Al, A2+B1 -+ A, C1+ 0 etc.and also the mechan- istic steps that do not form product such as A2+ 0 + A1 +Al. For proton transfer and metal-ligand complexation reactions, concentrations were limited such that the simplified eqn (6) was applicable: This equation was used to generate all the elementary transfer steps.It was assumed that reactant species are distributed randomly (Poisson distribution) throughout the droplets at equilibrium. This appears justified from fluorescor/quencher studies in the same AOT microemulsion system.ls Furthermore, this assumption is equivalent to the statement that all elementary transfer steps proceed with the same second-order rate constant (k,,) modified only by calculable statistical factors. Initial reactant concen- trations (Al, B1, B2 etc.) were calculated from the known overall reactant concentrations and experimentally determined droplet concentrations using the Poisson distribution equation : P(n) = P / n ! exp ( - R ) ~ (7) P(n) is the probability that a droplet contains n species whose average occupancy is A. The mechanism [eqn (6)] contains 74 elementary steps.This, together with initial concentrations, was used as input for a large numerical integration computer prog~am.~~f 31 The program was run to simulate the kinetic course of the reaction. It was checked that running the program with zero concentration of A (i.e. with an initial non-Poisson distribution of B alone) yielded the correct Poisson distribution of species B amongst the droplets in the limit of long simulation times. Only two parameters in the procedure were unknown; the rate constant k,, and an equilibrium constant K (since these reactions are not irreversible). The experimental data are close to a single- exponential decay and this decay is correctly simulated by the program. Values of k,, and K were varied until coincidence between experiment and simulation was obtained.Details of the analysis procedure are given in ref. (19). The use of different reactions as indicators for droplet exchange enabled us to investigate ion charge and size effects on the exchange process.P. D. I . Fletcher, A . M . Howe and B. H. Robinson 99 1 Table 1. Values of k,,/106 dm mol-l s-l in AOT-stabilised water-in-n-hep tane-microemulsions" 10 20 10 15 20 30 10 15 20 30 10 15 20 30 1 Ob 15 20b 30b* - 5 - 10 4.2k0.5 3.1 f0.5 1.7 k 0.3 - 15 7.4f 1 4.3 k 0.7 2.9 f 0.5 - 20 1 0 f 2 7.3 f 1 6.6& 1 - 25 - 1 6 f 3 11 +2 - 1.8k0.2 1.0 fO.1 3.1 f0.3 2.0 & 0.2 1.4 +O. 1 4.9 & 0.3 3.5 k 0.3 2.7f0.1 - - - - 7.5 f 1.5 6.6f 1.1 1 4 f 4 1 4 f 2 14f 1 - " Data are shown for various R, temperatures and reactions. The reactions are; (a) H+/NPSA2-; (b) Zn2+/murexide and (c) Ir(Cl):-/Fe(CN)t-. The values of the droplet hydrodynamic radii are 3.25, 4.13, 5.00 and 6.75 nm for R = 10, 15, 20 and 30, respectively.Values of lie,/ lo6 dm3 mol-l s-' of 14,lO and 10 for R = 11, 22 and 33, respectively, have been measured using the quenching of an excited stateof pyrene tetrasulphonate by copper ions or Fremy's salt at 25 0C.23 A value of k,, of (lo+ 1) x lo6 dm mol-l s-' was measured for the electron- transfer reaction between Ir(C1):- and Fe(bipyridyl),(CN),. Results and Discussion Exchange in Microemulsions containing n-Heptane as a Continuous Phase Values of k,, for different indicator reactions, microemulsion compositions and tem- peratures are shown in table 1.A priori, there is no reason to expect k,, to be related for the different reactions. However, for a particular R value and temperature, k,, values determined with different chemical reactions are remarkably similar. This has clear implications for the exchange mechanism. It should be noted that values of k,, from the proton-transfer and metal-ligand complexation reactions were determined at low AOT concentrations (varied in the range 1-10 mmol dmV3) and rely on experimentally measured droplet concentrations. Results obtained with the electron-transfer reactions cover higher AOT concentrations and a knowledge of droplet concentrations was not required. Included in table 1 are time-resolved triplet excited-state quenching data which also provide k,, data for comparison purposes.23 Again, the same values of k,, are obtained.Rather higher values of k,, (> los dm3 mol-l s-l) have, however, been reported for propyl viologen sulphonate exchange in quenching experiments with magnesium tetraphenyl porphyrin triplet.32 In this case, quenching might be possible by992 Kinetics of Solubilisate Exchange Table 2. Comparison of k,, values for AOT-heptane-water microemulsions using different concentrations of reactants and 3 3.5 4.5 3 3.5 4.5 3.5 3 2.5 2 3 3.5 2.5 2 3.5 2.5 3.5 2 200 200 1 00 100 100 100 100 100 100 50 50 50 50 5.4 10.8 21.6 2.7 5.4 10.8 16.2 8.1 12.6 6.3 18.9 25.2 8.4 4.2 16.8 4.2 8.4 8.4 25 5 57 52 25 14 10 10 5 5 5 5 5 k,, conditions 2.84 2.84 2.84 2.84 2.84 2.84 2.84 2.84 3.26 3.26 3.26 3.26 3.26 3.26 3.26 3.26 3.26 2.17 7 8 3 3 5 2 2.8 58 3 4.6 5.6 3.6 5.8 4.1 A = Zn2+ 4.3 B = murexide 4.3 R = 15 3.7 T = 15°C 3.5 4.5 4.2 3.6 4.0 A = Hf 4.5 B = NPSA2- 4.3 R = 15 4.4 T = 15°C 4.1 4.5 4.0 4.2 4.2 5.0 20 A = Fe(CN):- 16 B = IrC1:- 12 R = 10 14 T = 24°C 16 17 14 14 14 21 16 13 14 a AOT concentrations are in mmol dm-3, the concentrations of A and B are in pmol dm-3 and k,, values are in lo6 dm3 mol-' s-l.Proton-transfer and ligand-exchange results were obtained using the stopped-flow method ; electron-transfer results by the CFIO technique. a different mechanism. Owing to the large size of the porphyrin, quenching may take place during an encounter between droplets without exchange. Then, higher values of k,, might be expected. The results in table 2 show k,, is independent of the reactant or AOT concentrations. The independence of k,, on the reactant concentrations and the different (but always rapid) kchem values indicates that chemical reaction is never the rate-determining step.This also supports scheme (a) as the predominant mechanism for the overall chemical reaction. Data in fig. 1 and table 1 were used to calculate apparent enthalpies and entropies of activation for the exchange process. It is known from SANS measurements that droplet size for a particular R is virtually independent of temperature and AOTP. D. I. Fletcher, A . M. Howe and B. H. Robinson 993 Fig. 1, Ir(C1): 1 I I 3.3 3.4 3.5 3.6 103 KIT Arrhenius plots for the AOT-water-heptane microemulsion system, determined using the -/Fe(CN)i- reaction. 0, R = 10; A, R = 20; +, R = 30.Am = E,-RT; where d In k,,/d( 2-l) = - EJR. Table 3. Apparent activation parameters for the droplet exchange process in AOT-heptane microemulsions" 10 70f10 93f13 - 6 7 f 7 116+20 15 83+15 140f20 - - - 20 95+20 180f20 86f11 87+9 180_+40 30 - - - 108f 11 250f60 a AH/kJ mol-l and A S / J K-l mol-l. The different reactions are (a) proton transfer, (b) ligand exchange and (c) electron transfer. concentration for microemulsion compositions not too close to phase b0~ndaries.l~ Plots of Ink,, us. 1 / T provide activation enthalpies from the slopes; some typical plots are shown in fig. 1 and activation parameters in table 3. Large positive activation enthalpies and entropies for exchange are indicated which increase with increasing R (and hence droplet size).All three types of reaction have low activation enthalpies (of the order of 10-30 kJ mol-l) in a bulk water medium but show the same (greatly altered) activation parameters in the microemulsion. This provides additional evidence that the rate- determining step has changed from the chemical reaction step to that involving communication of the reactants [scheme (a)]. Two extreme mechanisms for inter-pool solubilisate exchange may be postulated. Mechanism A involves initial diffusion together of two droplets to form an encounter pair. While the droplets are in contact, the solubilisate species may diffuse through the994 Kinetics of Solubilisate Exchange surfactant bilayer at the point of contact of the droplets. The droplets then separate without having coalesced.Droplet encounters can be ' sticky' (i.e. experience short-range attractive interactions) and the contact time of an encounter increases as the upper temperature limit for microemulsion stability is a~pr0ached.l~ This should be reflected in k,, values (see later discussion). Mechanism B involves fusion of two droplets to form a transient, unstable droplet dimer of some indeterminate shape insofar as the interface may have fluctuating curvature on a fast (sub-microsecond) timescale. However, if the dimer species is sufficiently long-lived, the aqueous contents of the coalesced droplet will distribute randomly by diffusion. Simultaneous redistribution of AOT and water, previously identified with the separate droplets, will also occur. This process is not predicted by mechanism A.The dimer, which we can anticipate to have a lifetime in the microsecond time range, then breaks down into two separate droplets with random- isation of solubilisates between droplets. Mechanism B is thought to operate in AOT- stabilised droplets for the following reasons. First, k,, is independent of the exchanging species. This is true for H+, Zn2+, the electron-transfer reagents (negatively charged), Cu2+ and Fremy's Rates of permeation of ions through the surfactant bilayer (mechanism A) would be expected to be highly species dependent, whereas randomisation within the transient dimer (mechanism B) should be species independent as observed. Secondly, we have observed that mixing microemulsions of large droplet size (large R ) with microemulsions of small droplet size (low R ) produces, within a few seconds, a microemulsion containing droplets of intermediate size.Intuitively, it seems unlikely that such a major reorganization of the droplet population can occur without droplet transient coalescence and re-separation. An intermediate mechanism may be proposed involving generation of water 'channels' during the lifetime of the encounter pair. This may be achieved by cooperative movements of a surfactant 'block' and such processes have been postulated to occur in membrane^.^^ Such channels would have to be large to obviate charge effects at the surfactant-water interface on the migrating ionic solubilisate. If large channels are postulated, the mechanism is essentially that given by B.If every encounter between droplets resulted in solubilisate exchange, k,, should be comparable with the collision frequency of droplets. The diffusion-controlled rate constant, kDc, for encounters between droplets of overall radius Y and diffusion coefficient D (in units of m2 s-l) is given by the Smoluchowski equation: (8) = 8000RT/3q (9) k,,/dm3 mol-l s-l = 8N(2r) 103D where R is the gas constant, Tis the absolute temperature and q is the solvent viscosity (in units of kg m-l s-l). For n-heptane at 25 "C, q = 3.86 x lop4 kg m-l s-l and hence kDc is ca. 1.7 x 1O1O dm3 mol-l s-l. For a droplet concentration of mol dmp3, the time between encounters is ca. 60 ns (= {kDc[droplets]}-l). Comparison of kDc with k,, (table 1) shows that 1 in 1000 to 10000 encounters results in solubilisate exchange.There is, therefore, a free-energy barrier to the exchange process. The apparent enthalpy of activation is substantial, but this is largely offset by a high, positive entropy of activation (table 3). However, the implication of these results is that only the fastest chemical reactions in AOT-stabilised microemulsion systems are likely to be controlled by transport of reactants between droplets. Furthermore, generalising from these results and those we have obtained for other surfactant-stabilised systems, we believe AOT-stabilised droplets are among the most kinetically stable droplet dispersions so far investigated. Thus k,, values we have determined tend to reflect lower limits in microemulsion systems. Increasing k,, and/or prolonging the lifetime (k;&s) of the dimer (or higher-order aggregates) results in a progressive transition from a well defined droplet dispersion to a bi-continuous structure.The rate constant k,, may be considered to be made up of at least two steps, the firstP. D. I . Fletcher, A . M. Howe and B. H. Robinson 995 of which represents association of two droplets to give an encounter pair (a fast pre-equilibrium with equilibrium constant Ken). The fusion step (characterised by a first-order rate constant kfUs) is a subsequent slow step during which solubilisate exchange occurs. Thus we have: . I fast encounter slow fused pair dimer k,, (dm3 mol-1 s-l) = Ken kfus and Ken = k,n/k-en. The process described in eqn (10) is apparently reversible, but in practice the collapse of the dimer to reform the encounter pair is expected to involve randomisation both of the water core contents and of the surfactant shell.Hence the product (k,,[droplets])-l essentially defines the lifetime of an individual discrete droplet in the system. Encounter-pair formation is expected to be diffusion-controlled and hence ken is equal to k,, [eqn (9)] and depends only on viscosity. The equilibrium constant Ken (and hence k,,) is affected by the presence of inter-droplet interactions through the rate of encounter-pair dissociation k-en. For ' hard-sphere ' droplet behaviour kPen = 6D/(2r)2 (1 1) and Ken = (4/3~)N(2r)~. (12) Ken [eqn (1 2)] is temperature-independent, consistent with a ' hard-sphere ' interaction. A typical value of Ken (R = 20 system, r = 5 nm) is 2500 dm3 mol-l.For interactions between AOT-stabilised droplets in n-decane a SANS structure-factor analysis suggests that an additional short-range attractive interaction is present which increases sharply in the vicinity of the upper-temperature phase transition.'*. l5 This attractive interaction would have the effect of increasing Ken (and hence kex). Decreasing the temperature more than ca. 15 "C below the upper transition or decreasing the chain-length of the alkane solvent (which raises the upper transition temperature; see fig. 2) causes the attractive interaction to become negligible. For the measurements in n-heptane, where the upper temperature transition is greater than 60 "C for the R range studied, the kinetic results apply to temperatures more than 35 "C below the transition.Hence, the attractive interactions are likely to be negligible. Also, there is no correlation between k,, and r3 [eqn (1 2)] so variations in k,, are thought to be dominated by the factors determining the magnitude of kfus. This is not necessarily true for higher alkane solvents for which the upper transition temperature is reduced (fig. 2). Clustering of AOT-stabilised droplets in dodecane has been observed using PCS at temperatures around 25 0C,34 indicating attractive interactions are present at lower temperatures for this alkane. However, all the kinetic results for the different oils and additives are generally confined to temperatures more than 20 "C below the upper transition. Hence the conclusion that variations in k,, are dominated by the kfus term is likely to be valid for all the results presented here.For k,, = lo7 dm3 mol-l s-l, kfus for an R = 20 system is 4 X lo3 s-l [eqn (lo)]. The tendency of droplets to fuse is given by kfus/k-fus. There is evidence from SANS form factor analysis for the existence of these higher fused aggregates,12 but kfus/k-fus is clearly less than 1, and is probably typically 0.1, suggesting a dimer lifetime of ca. 25 ps. The fusion of two droplets in contact to give a transient droplet dimer involves some decrease in oil/water interfacial area. Area contraction may be associated either with desorption of surfactant from the interface (into the continuous oil solvent) or a compression of the surfactant film and a corresponding decrease in the interfacial area occupied per AOT molecule.In addition, the dimer may be considered a non-spherical, irregular, disordered entity with rapid fluctuations in local interfacial curvature and surfactant density. In any case, fusion of two droplets must occur through a series of996 Kinetics of Soluhilisate Exchange R 5( C 1 I 15 30 45 TI" C Fig. 2. One-phase microemulsion stability region maps for 0.1 mol dm3 AOT in various straight-chain alkane solvents. The numbers indicate the number of carbon atoms in the solvents. The ringed and un-ringed numbers indicate the low- and high-temperature phase boundaries, respectively. The bar on the upper temperature transition line for n-pentane marks the boiling point of the solvent. structural states containing local interfacial regions of unfavourable surfactant film curvature.It is likely that most if not all of the factors mentioned above contribute to the enthalpy and entropy of activation for the droplet fusion step. Apparent activation enthalpies and entropies both increase with increasing droplet size (increasing R). As a result, k,, becomes independent of droplet size at a particular temperature. Since, in the absence of water, AOT in alkane solvents readily forms reversed micelles, the concentration of monomeric AOT in alkanes is ~ub-millimolar.~~ Hence desorbed surfactant in the microemulsion system is likely to be in the form of reversed micelles in the continuous oil medium. There is some evidence for the existence of such reversed micelles in AOT water-in-oil microemulsions from small-angle X-ray ~ c a t t e r i n g ~ ~ and kinetic data.25~37 In summary, a droplet dispersion may be considered to consist of the following entities maintained in dynamic equilibrium: 2 droplets +encounter pair f transient fused dimer (+monomeric surfactant or reversed micelles ejected from droplets as a consequence of fusion).Oil Solvent Variation Exchange is a fundamental dynamic process and it is reasonable to expect a correlation between exchange and the thermodynamic stability. The simplest way to explore this correlation is to change the oil component, which shifts the microemulsion stability map on the temperature axis. Fig. 2 shows the single-phase microemulsion regions for differentp. D. I. Fletcher, A . M. Howe and B.H. Robinson 997 n-alkane solvents. Higher chain-length alkanes shift the region of microemulsion stability to lower temperatures. The microemulsion phase boundaries are described by plots of maximum R (or water solubilisation) us. temperature. The temperature at which the highest R is achieved may be equated with the phase inversion temperature (pit.). The positions of the phase boundaries are essentially independent of AOT concentration (at fixed R ) since the AOT does not significantly partition from the water-oil interface to the oil. (This is not true for many microemulsion systems, e.g. with non-ionic surfactants or when cosurfactants are present. In these systems, surfactant and cosurfactant partition between the interface and the oil and the droplet composition alters on dilution.A pseudo-binary representation of these systems would not be useful.) The value of the p.i.t. is sensitive to the oil, the salt concentration in the water and the presence of surface-active additives (cosurfactants). The effect of these variables may be rationalised in terms of the surfactant film curvature. For mixtures containing comparable volumes of oil and water, and a fixed quantity of AOT, increasing the temperature through the p.i.t. causes transfer of AOT from the oil phase to the water 39 This transfer is associated with a change in surfactant film curvature from that of a W-0 aggregate structure (i.e. surfactant polar headgroups on the interior of the aggregate and the apolar tails on the exterior surface, defined here as negative curvature) to that of an 0-W aggregate structure (positive curvature).At temperatures close to the p.i.t. the AOT film curvature is in the region of zero. The third, surfactant-rich phase which is often observed near the p.i.t. has a 'bicontinuous' structure with zero net surfactant film curvature.*O A distinction must be made between natural curvature and actual curvature of the surfactant film. Natural (or spontaneous) film curvature (which is a function of temperature, alkane chain-length, salt concentration etc.) is the observed surfactant film curvature when the droplet phase is in equilibrium with a conjugate phase of the dispersed component. At the p.i.t. the natural radius increases to infinity. The actual surfactant film curvatures observed in single-phase microemulsions are, however, very much determined by the constraints of the microemulsion composition.For the AOT microemulsions studied in this paper, the droplet size (= 1 /actual curvature) is deter- mined by R as described in eqn (1) and shows only a very slight dependence on temperature and alkane s o l ~ e n t . ~ ~ ! ~ * Thus the tendency of the AOT to locate at the oil-water interface dominates over resulting unfavourable curvature effects. However, at the lower-temperature boundary, the actual curvature is equal to the natural curvature. Since the actual curvature in the AOT one-phase microemulsions is virtually independent of temperature, whereas the natural curvature becomes less negative with increase in temperature, the discrepancy between the natural and actual curvatures increases as the upper boundary is approached.Table 4 shows k,, data obtained in different alkane solvents. k,, increases with increasing carbon number of the solvent at a fixed temperature ( 5 "C) as shown in fig. 3. It is an interesting point that a 50/50 molar ratio of hexane-decane gives the same kc?x as octane. All data in alkane solvents may be represented on a common graph by using a temperature scale relative to the relevant microemulsion stability maps. The temperature dependence of all the data (including alkane solvent variation, additives and heavy-water substitution) is shown in fig. 4 as a plot of Ink,, us. ( T - TPIT)/(TPIT)z. TPIT is the p.i.t., taken as equal to the lower temperature boundary at R = 50.TpIT values were corrected for addition of reactants (typically by + 3 "C). Fig. 4 is a differential form of the normal Arrhenius plot with a slope equal to -E,/R. The data are coincident for all systems of equal R. Therefore, these are corresponding states (related by TPIT or p.i.t.) in the different AOT-stabilised microemulsions. Hence, changes in k,, correspond closely to the shifts in the microemulsion stability maps. The analysis in fig. 4 implies a common mechanism for exchange in all the solvents. Values of Ea are 61 +4 kJ mol-l (R = lo),998 Kinetics of Solubilisate Exchange Table 4. Values of k,,/106 dm3 mol-l s-l for water-oil micro- emulsions formed by 0.1 mol dm-3 AOT in various oil solventsa R solvent T/OC 10 20 30 Cn5 Cn6 Cn7 Cis08 Cn8 Cn9 CnlO Cnll Cn12 cyclohexane cycloheptane cyclo-octane decalin oct- 1-ene 5 10 15 20 5 10 15 20 5 10 15 20 25 5 10 15 20 5 10 15 5 10 5 10 5 5 15 15 15 15 5 10 15 25 1 .Of 0.1 1.8f0.3 2.8 f0.3 4.3 f 0.4 1.3kO.l 2.1 f0.2 3.3 f0.4 5.1 f 0.6 1.8 k 0.2 3.1 f0.3 4.9 k 0.3 14+2 2.3 f 0.3 4.1 kO.5 6.6 f 0.9 - - 2.7 f0.2 4.6k0.5 7.1 f0.7 4.4 k 0.7 - 5.8 f0.8 - 7.9 f 0.7 11 f 4 0.80 f 0.05 1.1 fO.1 2.1 f0.3 2.1 & 0.3 1.4f0.1 2.5 f 0.2 3.3 f 0.4 - - 0.80 & 0.08 1.6f0.2 3.2 f0.3 0.66 k 0.04 1.4f0.1 2.2 f0.2 3.7 f 0.4 1,OfO.l 2.0 k 0.2 3.5 f0.3 7.5k1.5 14f2 1.2 fO.1 2.3 f 0.2 4.2 +_ 0.4 - 1.7 f 0.1 3.0 f 0.2 5.3 & 0.4 2.7 f 0.3 3.8 f 0.3 4.1 f0.7 - 6.3 f0.5 7.0 f 0.5 - - - - - - - 2.1 & 0.2 - - - - - - 1.5f0.2 2.6f0.3 - 1.4 & 0.1 2.7 f 0.1 6.6f 1.1 14& 1 0.75 f 0.08 1.9 & 0.2 3.5 f0.3 8.1 & 1.2 1.3kO.l 2.6 & 0.2 4.7 f 0.1 2.2 f 0.3 3.4 f 0.2 3.1 fO.l 4.3 k0.3 4.8 f 0.7 - - - - - - - - - a Cnx indicates a straight-chain alkane of x carbon atoms and Cis08 is iso-octane.P.D. I . Fletcher, A . M . Howe and B. H . Robinson 999 I 20 5 6 7 8 9 1 0 1 1 12 alkane solvent chain length Fig. 3. Plot of k,, (log scale) at 5 "C against chain length of solvent. Upper line, R = 10; lower line, R = 20. The triangular symbols refer to solvent mixtures of hexane and decane. 77+7 kJ mol-l ( R = 20) and 94+ 12 kJ mol-l ( R = 30). The figure also shows the temperature (on this reduced scale) at which k,, becomes independent of R. The exchange rate is slowest ( 105-10s dm3 mol-1 s-l) at the low-temperature boundary and rises to lo8 to lo9 dm3 mol-1 s-l as the upper boundary is approached (estimated by extrapolation of the linear Arrhenius plot to high temperatures).The droplets are most stable (i.e. show the slowest exchange rate) at temperatures close to the lower-temperature phase boundary where the actual surfactant film curvature is equal to the natural film curvature. Droplet fusion proceeds through a series of states in which the surfactant film of the coalescing (exchanging) droplets has to adopt high positive curvatures. A likely structure for the transition state is shown below. At * there is a very high local positive curvature (i.e. an expanded head region and compressed tail region). From the discussion of phase behaviour, increasing the temperature above the lower-temperature boundary favours increased positive (or less negative) natural curvature.Hence, regions of local positive curvature become energetically less unfavourable facilitating exchange. Increas- ing temperature therefore reduces the free-energy barrier to droplet fusion [kfus step of1000 Kinetics of Solubilisate Exchange Fig. 4. Plot of the k,, (log scale) us. ( T - TPIT)/(TPIT):! for AOT-stabilised W-0 microemulsions in a range of oils and with a variety of additives. 0, R = 10; +, R = 20; 0, R = 30. eqn (lo)]. Hence the temperature dependence of k,, does not correspond to the normal situation of a greater fraction of species achieving sufficient thermal energy to overcome a fixed energy barrier to reaction. The situation for droplet exchange is better described as one in which the energy barrier reduces with increasing temperature.For this reason the activation parameters derived from the Arrhenius plots are referred to as apparent values. Changes in the free-energy profile for exchange are illustrated in fig. 5, which also shows interaction energy changes which occur in the close vicinity of the upper- temperature transition. An interesting corollary to this discussion is that the rate of exchange between hydrophobic species in dispersed oil droplets in an 0-W microemulsion stabilised by an ionic surfactant should decrease with increasing temperature. 0-W microemulsion droplets possess interfacial surfactant films with positive curvature. If the exchange process proceeds by a fusion mechanism then 0-W droplet fusion would require the formation of regions of local negative curvature.Since increasing temperature favours positive curvature the energy barrier to fusion in this case is expected to increase with increasing temperature. Increasing the temperature also increases the degree of counter- ion dissociation for charged droplets which would also lead to increased repulsive electrostatic inter-droplet interactions. Both these effects would result in a decreased rate of exchange with increasing temperature. Fig. 6 shows the microemulsion stability regions using various hydrocarbons based on eight carbon atom (a) and also various cyclic alkanes (b). Cyclisation, chain-branchingP . D. I . Fletcher, A . M. Howe and B. H. Robinson 1001 distance Fig. 5. Schematic free-energy profiles of the droplet fusion process at ( i ) a temperature close to the lower temperature phase boundary, (ii) an intermediate temperature and (iii) a temperature close to the upper temperature boundary.The separation distances correspond to (a) separated droplets, (b) a droplet encounter, (c) the energy maximum in the fusion process (the transition state) and ( d ) the fused droplet dimer species. 101 R 51 101 5 65 15 30 TIo C B 20 LO 60 Fig. 6. Microemulsion stability maps for 0.1 mol dmP3 AOT in various solvents. A, Various eight- carbon solvents; (a) n-octane; (b) iso-octane; (c) oct-1-ene. B, Various cyclic alkanes: C6, cyclohexane ; C7, cycloheptane; C8, cyclo-octane; D, decalin. and non-saturation of the hydrocarbon solvent all cause a shift of the stability region to higher temperatures.This implies that these solvents are all more effective than the straight-chain oil in favouring increased negative curvature of the surfactant interface by packing in the alkyl-tail region. Table 4 shows some values of k,, for these microemulsion systems. A comparison of these data with those for the corresponding straight-chain alkanes shows k,, is decreased for these oils as expected.1002 Kinetics of Solubilisate Exchange 0 1 2 15 30 45 [cholesterol]/ 1 O-* mol dm-3 T/" C Fig. 7. Effect of cholesterol on the exchange rate and microemulsion stability map. (a) k,, us. cholesterol concentration for 0.1 mol dm-3 AOT in n-heptane at 10.0 "C and R = 10. (b) Phase stability map for 0.1 mol dm-3 AOT in n-heptane.The numbers indicate the concentrations (mol dmd3) of cholesterol. R I 1 0 0.1 0.2 0.3 0 15 30 45 [ benzyl alcohol]/mol dm-3 T / O C Fig. 8. Effect of benzyl alcohol on the exchange rate and microemulsion stability map. (a) k,, us. benzyl alcohol concentration for 0.1 mol dm-3 AOT, 24f 1 "C: +, R = 10; A, R = 20; 0, R = 30. (b) Microemulsion phase stability map for 0.1 mol dm-3 AOT in n-heptane with (0.1) and without (0) 0.1 mol dmW3 benzyl alcohol. Effect of Cosurfactants and H,O/D,O Substitution Toluene, benzyl alcohol and cholesterol are known to have dramatic effects on the stability and permeability of lipid bilayers. Many single-tail surfactants, such as sodium dodecyl sulphate, do not form single-phase water-in-oil microemulsions in the absence of a medium-chain-length alcohol, e.g.pentanol. The effect of alcohol cosurfactants (of medium to long chain-length) is to produce increased negative curvature of the surfactant film so as to favour the formation of reversed micelles or water-in-oil microemulsions. The alcohols screen inter-headgroup electrostatic repulsions and pack the alkyl tail regions of the surfactant films, both factors favouring increased negative curvature. (The ' wedge-shaped' molecular geometry of AOT makes it a favourable surfactant for water-in-oil microemulsion stabilisation without a cosurfactant.) There- fore, these additives play a fundamental role in determining phase behaviour.P. D. I . Fletcher, A . M . Howe and B. H. Robinson 1003 1 1 0 l - 2 0.5 1 0 o? 15 30 L5 0- 0 0.5 1 fraction D 2 0 TI" C fraction D20 Fig.9. Effect of H,O/D,O substitution on the exchange rate and microemulsion stability map. (a) k,, us. fraction of D,O for 0.1 mol dmP3 AOT in n-heptane at 10 "C: +, R = 10; 0, R = 20. (b) Microemulsion phase stability map in n-heptane with H20 (H) and D20 (D). (c) k,, us. fraction of D,O for 0.1 mol dm-3 AOT in n-decane at 10 "C: 0, R = 10; A, R = 20; +, R = 30. Fig. 7 and 8 show representative effects of cholesterol and benzyl alcohol on the exchange rate and microemulsion stability. Cholesterol and also toluene decrease the exchange rate and shift the stability region to higher temperatures. This behaviour is consistent with these additives favouring increased negative curvature of the interface. Benzyl alcohol has the opposite effect.The exchange rate is increased and the micro- emulsion region is shifted to lower temperatures. Benzyl alcohol must be located at the interface in such a way as to swell the headgroup region of the surfactant shell more than the tail region, thus favouring decreased negative curvature. The hydroxy group of the alcohol presumably serves to locate it close to the AOT head; the aromatic group then expands the headgroup region more than the tail region. Fig. 9 shows the effect of substituting D20 for H20. A shift in stability to higher temperatures with D20 is accompanied by a decrease in exchange rate. It appears, therefore, that D,O favours the increased negative natural curvature of the interface. As for alkane variation, additives which increase the p.i.t.also decrease the exchange rate. Both effects are associated with an increased tendency of the surfactant film to favour negative curvatures. The temperature dependence of all the data (for alkane solvent variation, additives and heavy water substitution) is shown in fig. 4 and it can be seen that the k,, data for all the variables which affect the microemulsion system fall on coincident lines for equal R. Independent measurements have confirmed that droplet sizes do not change significantly for the concentrations of additives It appears, therefore, that the corresponding states of the microemulsions (related by the p i t . ) may be reached either by changing the alkane solvent or by addition of additives. Droplet Exchange, Microemulsion Structural Properties and Phase Behaviour Experiments, using quasi-elastic incoherent neutron scattering, have shown that addition of toluene or benzyl alcohol has no significant effect upon the rate of short-range local diffusion of AOT within the microemulsion interfacial layer.42 The fast local diffusion of the AOT is also found to be independent of droplet size.43 It appears, therefore, that k,, is not correlated with surfactant mobility but rather with the energy required to produce localised regions of high positive curvature in the interface.Clarke and co-workers used dynamic light scattering to study concentrated AOT1004 Kinetics of Solubilisate Exchange microemulsions. They observed a biexponential decay of the intermediate scattering function for droplet volume fractions greater than 0.25 and assigned the slower relaxation to polydispersity fluctuation^.^^ They conclude that the timescale of the polydispersity fluctuations is at least three orders of magnitude slower than the particle collision frequency.These polydispersity fluctuations are predicted from the exchange results reported in this paper to occur on this timescale. Attempts using the stopped-flow method to measure pool exchange kinetics with other microemulsion systems [e.g. hexadecyltrimethylammonium bromide (CTAB)-water 50/50 heptane+hloroform] have shown that these systems all exchange faster than the system studied here (and too fast for stopped-flow study). Preliminary data for the AOT-glycerol-heptane system indicate that exchange is very rapid and comparable to that of the AOT-water-heptane system.These results for glycerol as the dispersed component are particularly interesting since the kinetics indicate that the viscosity of the dispersed component does not determine the exchange rate. (The viscosity of glycerol is some 1500 times that of water.) Measurements of AOT local diffusion in glycerol dispersions using quasi-elastic incoherent neutron scattering show that AOT diffusion is slowed by a factor of two as compared with the corresponding water di~persion.~~ Using a fast laser photolysis method, Thomas and co-workers measured k,, values of ca. lo8 dm3 mol-1 s-l for microemulsions stabilised by potassium oleate and hexanol. Substitution of hexanol by pentanol increases k,,,459 46 which is consistent with the exchange rate being dominated by the energy needed to form regions of local positive curvature in the surfactant film since the shorter chain-length alcohol is expected to be more effective in favouring positive film curvature. Lindman and c ~ - w o r k e r s ~ ~ - ~ ~ measured, using an n.m.r.technique, the self-diffusion coefficients (over macroscopic distances) of all components in a variety of water-in-oil microemulsion systems. Microemulsions with a ‘ discrete ’ droplet structure are expected to show slow diffusion of surfactant and dispersed water relative to the continuous oil component since these are contained within the slow-moving, large aggregates for the major time (10-100 ms) of the measurement. Only for AOT-water-p-xylene and oleate plus decanol cosurfactant systems are relatively slow diffusion of water and surfactant observed, which implies that these droplets retain their discrete nature on the millisecond timescale.Oleate plus short-chain alcohols, the non-ionic surfactant CI2E4, sodium dodecyl sulphate plus pentanol and octyl benzene sulphonate plus pentanol surfactant systems all show fast diffusion of all components on this timescale. These studies indicate that many microemulsions are ‘ bicontinuous’ as probed by this technique; i.e. exchange is fast on the timescale of the n.m.r. technique used. This conclusion is consistent with our observation that the AOT-stabilised microemulsions are amongst the most kinetically stable (i.e. show the lowest k,, values). The :discreteness’ of the droplets is, of course, related to the timescale of the exchange process as compared with the timescale probed by the particular technique employed.The concentration of droplets depends on the AOT concentration and R. For 0.1 mol dm-3 AOT and R = 10, the droplet concentration is 1.0 mmol dm-3.18 Hence, for this solution, the time of independent existence of a droplet ranges from 1-10 ms close to the lower phase boundary to 1-lops at the upper boundary. On timescales slower than the exchange rate, the water in the microemulsion can be considered, from the point of view of a reaction medium, as effectively continuous, since a long timescale would allow the sampling of many water droplets via exchange. On these slow timescales, the analysis of kinetic data for water-soluble reactants in microemulsions can be based on concentrations per volume of dispersed water (i.e. the dispersed water component may be treated as a separate continuous phase, a ‘pseudo-phase’).On a nanosecond timescale, as probed by fluorescence lifetime measurements, exchange does not have time to occur. On this timescale, the droplets may be considered to be discrete entities. A time-resolved fluorescence quenching study of the acridiniumP. D. I . Fletcher, A . M . Howe and B. H . Robinson 1005 ion has shown that exchange does not occur in the AOT system over 100 ns.18 In this situation, local reactant concentrations within the droplet must be considered and so only certain concentrations are permitted, according to the statistical occupancy of the droplets. The exchange process is fundamental to an understanding of many phenomena in microemulsions, including chemical reaction kinetics, electrical conductivity and self- diffusion processes.The microemulsion structure may appear to consist of discrete droplets or to be bicontinuous depending on whether the timescale probed is slow or fast relative to the exchange process. The concept of a ‘ bicontinuous’ microemulsion structure is understandable in terms of a system of rapidly exchanging droplets. It is a pleasure to thank Dr J. Holzwarth and his co-workers (Fritz-Haber Institute, Berlin) for kindly making available to us the CFIO instrument. We also wish to thank the S.E.R.C. for a studentship and DAAD for a travel/support grant (A.M. H.). References 1 J. H. Fendler, Membrane Mimetic Chemistry (Wiley, New York, 1982).2 C. J. O’Connor, T. D. Lomax and R. E. Ramage, Adv. Colloid Interface Sci., 1984, 20, 21. 3 P. D. I. Fletcher and B. H. Robinson, J . Chem. Soc., Faraday Trans. I , 1984,61, 1594. 4 P. L. Luisi, Angew. Chem., 1985, 24, 439. 5 R. Hilhorst, C. Laane and C. Veeger, FEBS Lett., 1983, 159, 31. 6 P. Luthi and P. L. Luisi, J . Am. Chem. Soc., 1984, 106, 7285. 7 P. D. I. Fletcher, R. B. Freedman, B. H. Robinson and R. Schomacher, Biochim. Biophys. Acta, 1986, 8 B. H. Robinson, D. C. Steytler and R. D. Tack, J. Chem. Soc., Faraday Trans. 1, 1979, 75, 481. 9 M. Zulauf and H. F. Eicke, J . Phys. Chem., 1979, 83, 480. in press. 10 J. D. Nicholson and J. H. R. Clarke, Proc. Int. Symp., Surfactants in Solution, ed. K. Mittal and 11 C.Cabos and P. Delord, J . Appl. Crystallogr., 1979, 12, 502. 12 B. H. Robinson, C. Toprakcioglu, J. C. Dore and P. Chieux, J. Chem. Soc., Faraday, Trans. I, 1984, 13 C. Toprakcioglu, J. C. Dore, B. H. Robinson, A. M. Howe and P. Chieux, J . Chem. SOC., Faraday 14 M. Kotlarchyk, S. H. Chen, J. S. Huang and M. W. Kim, Phys. Rev. A , 1984,29, 2054. 15 J. S. Huang, S. A. Safran, M. W. Kim, G. S. Grest, M. Kotlarchyk and N. Quirke, Phys. Rev. Lett., 16 S. I. Chou and D. 0. Shah, J . Colloid Interface Sci., 1980, 78, 249. 17 E. Keh and B. Valeur, J . Colloid Interface Sci., 1981, 79, 465. 18 N. J. Bridge and P. D. I. Fletcher, J . Chem. Soc., Faraday Trans. I, 1983, 79, 2161. 19 P. D. 1. Fletcher, Ph.D. Thesis (University of Kent, 1982). 20 H. Kunieda and K.Shinoda, J . Colloid Interface Sci., 1980, 75, 601. 21 E. A. G. Aniansson, S. N. Wall, M. Almgren, H. Hoffmann, I. Kielmann, W. Ulbricht, R. Zana, 22 H. F. Eicke, J. C. W. Shepherd and A. Steinmann, J . Colloid Interface Sci., 1976, 56, 168. 23 S. S. Atik and J. K. Thomas, J . Am. Chem. SOC., 1981, 103, 3543. 24 P. D. I. Fletcher and B. H. Robinson, Ber. Bunsenges. Phys. Chem., 1981, 85, 863. 25 P. D. I. Fletcher, N. M. Perrins, B. H. Robinson and C. Toprakcioglu, in Reverse Micelles, ed. P. L. Luisi and B. E. Straub (Plenum Press, New York, 1984), p. 69. 26 J. F. Holzwarth, in Techniques and Applications of Fast Reactions in Solution, NATO ASI Symp. Ser., ed. W. J. Gettins and E. Wyn-Jones (Elsevier, Amsterdam, 1979), p. 509. 27 H. Bruhn, S. Nigam and J. F. Holzwarth, Faraday Discuss. Chem. Soc., 1982, 74, 129. 28 H. Bruhn and J. F. Holzwarth, Ber. Bunsenges. Phys. Chem., 1978, 82, 1006. 29 M. Eigen, W. Kruse, G. Maas and L. DeMaeyer, Prog. React. Kinet., 1964, 2, 286. 30 G. Maas, Z. Phys. Chem. N. F., 1968, 60, 138. 31 D. DeTar, Computer Programs for Chemistry (Benjamin, New York, 1969), vol. 2. 32 M. P. Pileni, J. M. Furois and B. Hickel, in Surfactants in Solution, ed. K. Mittal and B. Lindman 33 P. Fromherz, personal communication. 34 A. M. Howe, J. A. McDonald and B. H. Robinson, J . Chem. Soc., Faraday Trans. I , 1987, 83, 1007. B. Lindman (Plenum Press, New York, 1984), vol. 3, p. 1663. 80, 13. Trans. I , 1984,80,413. 1984, 53, 592. J. Lang and C. Tondre, J. Phys. Chem., 1976,80,905. (Plenum Press, New York, 1984), p. 1471. 34 FAR 11006 Kinetics of Solubilisate Exchange 35 B. Djermouni and H. J. Ache, J. Phys. Chem., 1979,83, 2476. 36 J. C . Dore, A. North, J. McDonald, A. M. Howe, R. K. Heenan and B. H. Robinson, Colloid Surf., 1986, 19, 21. 37 P. D. I. Fletcher, A. M. Howe, N. M. Perrins, B. H. Robinson, C. Toprakcioglu and J. C. Dore, in Surfactants in Solution, ed. K. Mittal and B. Lindman (Plenum Press, New York, 1984), vol. 3, 1745. 38 R. Aveyard, B. P. Binks, S. Clark and J. Mead, J . Chem. SOC., Faraday Trans. 1, 1986, 82, 125. 39 R. Aveyard, B. P. Binks and J . Mead, J. Chem. SOC., Faraday Trans. 1, 1986,82, 1755. 40 E. W. Kaler, H. T. Davis and L. E. Scriven, J. Chem. Phys., 1983, 79, 5685. 41 A. M. Howe, C . Toprakcioglu, J. C. Dore and B. H. Robinson, J . Chem. SOC., Faraday Trans. I , 1986, 42 P. D. I. Fletcher, B. H. Robinson and J. Tabony, J . Chem. Soc., Faraday Trans. I , 1986, 82, 231 1. 43 J. Tabony, A. Llor and M. Drifford, Colloid Polym. Sci., 1983, 261, 938. 44 J. H. R. Clarke, J. D. Nicholson and K . N. Regan, J. Chem. SOC., Faraday Trans. I , 1985, 81, 1173. 45 S. S. Atik and J . K. Thomas, J . Am. Chem. SOC., 1981, 103, 7403. 46 S. S. Atik and J. K . Thomas, J . Phys. Chem., 1981, 85, 3921. 47 B. Lindman, P. Stilbs and M. E. Moseley, J . Colloid Interface Sci., 1981, 83, 569. 48 P. G. Nilsson and B. Lindman, J. Phys. Chem., 1982,86, 271. 49 T. Warnheim, E. Sjoblom, U. Henriksson and P. Stilbs, J . Phys. Chem., 1984, 88, 5420. 82, 2411. Paper 6/21 1 ; Received 30th January, 1986

ISSN:0300-9599    

DOI:10.1039/F19878300985

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

J . Chern. SOC., Faraday Trans. I , 1987,83, 1007-1027 Fluorescence Quenching as a Probe of Size Domains and Critical Fluctuations in Water-in-oil Microemulsions Andrew M. Howe,? Julie A. McDonald and Brian H. Robinson*$ Chemical Laboratory, University of Kent, Canterbury, Kent CT2 7NH A static fluorescence quenching technique has been used to determine the discrete droplet concentration, and by implication droplet size, in oil- continuous microemulsions stabilised by Aerosol OT (AOT). In particular (i) the nature and structure of water-in-oil microemulsions in the critical region have been investigated for the H,O-AOT-dodecane system and (ii) the effect of additives (benzyl alcohol and toluene) on the size of droplets in the single-phase system has been studied. From these studies it is demonstrated that the static fluorescence quenching technique is particularly appropriate for the investigation of changes in size domains. The results for the critical system suggest that there is no significant increase in droplet size in the single-phase system as the critical point is approached. However, evidence is obtained for facilitated solute exchange between droplets in the critical region.Small systematic increases in droplet concentration (decreases in droplet size) are readily detected on addition of benzyl alcohol and toluene to a single-phase system at compositions far removed from the critical (phase-transition) region. The static fluorescence quenching method with Ru(bp)g+ as fluorescer and Fe(CN)i- as quencher is shown to be optimal for the determination of size domains in the 1-2 nm region.The quenching rate constant within a droplet is slower than in bulk aqueous solution at high ionic strength. At high concentrations of benzyl alcohol the fluorescer partitions to the interfacial region which allows facilitated exchange of Ru(bp)g+ between droplets. Single-phase water-in-oil (w/o) microemulsions formed in the presence of the surfactant AOT [Aerosol-OT, sodium 1,4-bis(2-ethylhexyl) sulphosuccinate] are thermo- dynamically stable. They contain essentially discrete droplets which may be stable over a considerable range of composition and temperature. 1-3 The extent of the temperature stability range (which may well exceed 50 "C) depends on droplet size, which is pro- portional to the water/surfactant mole ratio, R.The chemical composition of the oil phase and the presence of additives influence stability and the temperature corresponding to the maximum solubilisation of water.3* Above the upper temperature limit of instability, single-phase microemulsions gen- erally separate into two phases in equilibrium; for the H,O-AOT-dodecane system both phases are oil-continuous and contain droplets of similar size but at different concentrations. There has been much recent interest in phase transitions in microemul- sion ~ystems.~-l~ Along the upper-temperature phase-transition boundary a critical point may be located for a particular R value and AOT concentration. At the critical point, defined in terms of a critical composition Xc and a critical temperature T,, phase separation occurs into two phases of equal volume and composition.We have been able to identify the critical composition for the H,O-AOT-dodecane w/o microemulsion dispersion at 25 "C. This system is particularly convenient for study because the critical point at ambient temperatures corresponds to a low water content (< 5% by volume) t Present address: AFRC Food Research Institute, Colney Lane, Norwich NR4 7UA. 1 Present address: School of Chemical Sciences, University of East Anglia, Norwich NR4 7TJ. 1007 34-21008 Fluorescence Quenching in Water-in-oil Microemulsions and a rather low R value. This implies that the droplets in the single-phase microemulsion are small at temperatures far below c. As the carbon number of the alkane is decreased, the critical point is associated with a larger R value and water content.The corresponding critical point in n-decane has been reported to be 5 cm3 H,O-3 g AOT-95 cm3 decane ( R = 40.8 and [AOT] = 0.067 mol dm-3) at 43 OC.14 (% H,O by volume = 4.9%, based on pAoT = 1.15 kg m-3 and pdecane = 0.75 kg m-3 at 43 “C.) For both w/o microemulsions and the corresponding aqueous micellar systems there has been controversy as to whether there are size changes as the temperature of the single-phase system is increased in the region of Xc and q.5’ 14-21 Convenient methods for size characterisation, such as photon correlation spectroscopy (PCS), small-angle neutron scattering (SANS) and small-angle X-ray scattering (SAXS), present difficulties since they provide information simultaneously concerning both inter-droplet interactions and droplet size.In the critical region the droplets experience ‘clustering’ as a result of a dramatic increase in droplet correlations. Using techniques such as PCS a ‘correlation length’ for the system can be evaluated which is related to the effective extent of a ‘cluster’. However, in such a situation it is impossible to distinguish between the two limiting cases illustrated schematically as (a) and (b) in scheme 1. If (a) applies then PCS (a) coalescence/growth 1 (b) “sticky encounters” increasing temperature m Scheme 1. can provide useful information in dilute systems about the size of the aggregated species via the determination of the translational diffusion coefficient. If (b) applies, then only information concerning the extent and lifetime of the aggregated structural domains is determined; for the latter, values are typically in the ps time region.The techniques of static fluorescence quenching (SFQ) and dynamic fluorescence quenching (DFQ) have previously been used to measure droplet sizes and exchange kinetics in water-in-oil micr~emulsions.~~-~~ In addition, Pileni et al.25 have utilised hydrated electrons, formed by pulse radiolysis, as a probe of water pool sizes. In this paper we report an SFQ study based on a system containing Ru(bp):+ as fluorescer and Fe(CN)i- as quencher. This combination probes the discrete nature of droplets in the ps time domain, and ideally provides information on the mean droplet concentration and hence droplet size.Other fluorescer-quencher combinations could be used which would probe the system in a different time and structure domain; e.g. methyl viologen22 could be used as a more effective quencher. There is some evidence24 that viologens are preferentially located in the interface region and can exchange rapidly between droplets. Fe(CN)i- was therefore preferred as kinetic exchange data for the closely related Fe(CN):- are available for the microemulsion under study, and such negatively-charged complex ions are known to be located inside the core of the water droplets.’ The critical system based on H,O-AOT-dodecane has small droplets away from the phase transition region with a water core radius of 3.2 nm (by SANS). This is most important because the interpretation of the SFQ data with Fe(CN)i- as quencher is then simplified.Other studies have used the R = 40.8 system in decane which contains droplets of water-core radius 4.6 nm. l4 For a simple analysis, a different fluorescer/quencher combination would be required for that system.A . M . Howe, J . A. McDonald and B. H. Robinson 1009 Experiment a1 Materials AOT was obtained from Sigma as the sodium salt of dioctylsulphosuccinate and was used without further purification. It had a quoted purity of >99% and was found to contain insignificant amounts of the alcohol and carboxylic acid impurities which form as a result of self-hydrolysis. Purity was confirmed by interfacial tension measurements at the heptane-water interface (we are grateful to Dr J. Mead, Chemistry Department, University of Hull, for making these measurements). In addition, the sample showed similar phase behaviour to a purified sample of AOT kindly provided by Professor L.Magid. Toluene, n-heptane (both SLR grade) and benzyl alcohol (>98%) were obtained from Fisons and n-dodecane (99 % ) from Aldrich. Water was triply distilled. Ru(bp),Cl, [tris(2,2'-bipyridyl)ruthenium(rr) chloride] was a gift from Shell (Thornton Research Centre) and K,Fe(CN), [potassium hexacyanoferrate(rr1) ( > 99% )] was obtained from Fisons. Microemulsions were prepared as follows: the required quantity of a (0.50 mol dm-3) AOT-in-alkane stock solution was introduced into a 10 cm3 volumetric flask; the fluorescer and various quencher stock aqueous solutions were then added followed by the additional amount of water and finally any additive, all with a microlitre syringe.The mixture was made up to the mark with alkane and shaken vigorously until a visually clear, single-phase liquid was obtained. Phase Diagrams The phase behaviour was determined by visual inspection of samples. Each sample was prepared by weighing AOT into a stoppered, graduated flask and adding the required volume of dodecane. The sample was mildly sonicated until all of the AOT had dissolved then water was added by means of a microlitre syringe. All solutions were thoroughly shaken and thermostatted to f O . 1 "C in the water bath of a Haake F3-C thermostat. Single-phase microemulsions formed optically clear dispersions without sonication within a few minutes. Phase separation into two optically clear oil-continuous micro- emulsions normally required ca.1 h for completion. However, in some cases two- or three-phase systems required up to 24 h to reach phase equilibrium. To calculate the percentage by volume of the various components, we took the density of AOT to be 1.15 kg m-3. The critical point was determined according to the following procedure: to a series of thermostatted samples, containing varying concentrations of AOT in dodecane, water was gradually added until the sample just became turbid when shaken. At the onset of turbidity these samples were left for phase separation to occur. The sample which separated into two clear phases of equal volumes was taken to be the critical composition at that chosen temperature. SFQ Measurements The fluorescence intensity measurements were made using an Aminco-Bowman spectro- photofluorimeter with a xenon light source.An Aminco 4-8912 ratio photometer was used to eliminate fluctuations due to instability of the lamp output. Rectangular 10 mm Hellma quartz fluorescence cuvettes were used. It was essential for the sample to be oxygen-free, as oxygen rapidly quenches *Ru(bp)i+ by an energy-transfer reaction with a second-order rate constant of 3.3 x lo9 dm3 mol-1 s-'.,~ Oxygen was removed from the sample solution in the cuvette by steady bubbling with oxygen-free nitrogen that had been previously passed through a jar containing the appropriate alkane. The procedure was carried out until the luminescence intensity was unchanged and took < 2 min.Prior1010 Fluorescence Quenching in Water-in-oil Microemulsions 1c n Y .- 1 d l !a W 4 X/nm Fig. 1. Emission spectrum of *Ru(bp):+ in a microemulsion (R = 10, 0.1 mol dm-3 AOT in heptane) in the absence (a) and presence (b) of 0.5 mol dm-3 benzyl alcohol at 25 "C. Excitation wavelength 470 nm. to each measurement of the fluorescence intensity, the cuvette to be used was filled with a microemulsion solution, without fluorescer or quencher, but with the same AOT and water concentrations to be used in the experiment. This was placed in the fluorimeter, and the intensity reading was set to zero to eliminate any long-time instrument drift. Fluorescence intensity measurements of the same microemulsion system containing fluorescer and quencher were made on several occasions, and when oxygen-free, the intensity remained constant over a period of 24 h.The excitation wavelength was 470 nm [not 450 nm, which is the absorption maximum of the Ru(bp)g+, since Fe(CN)i- also absorbs slightly at this wavelength]. The emitted light was detected in the region of the emission maximum, 610-620 nm (605 nm in the presence of benzyl alcohol). The luminescence spectra of *Ru(bp)i+ in an AOT-stabilised microemulsion and the same microemulsion system in the presence of benzyl alcohol are shown in fig. 1.A . M . Howe, J . A . McDonald and B. H . Robinson 101 1 DFQ Measurements The time-resolved fluorescence measurements were made using an Ortec single-photon- counting equipment with a thyratron-gated spark lamp.23 Samples contained in a standard 10 mm x 10 mm quartz fluorescence cuvette were mounted in a block thermo- statted to f 0.5 "C.Oxygen-free nitrogen, saturated with the appropriate alkane, was bubbled through the solution during the experiment. The maximum in the emission signal was at 580 nm using the RCA 8850 photomultiplier. The photomultiplier was cooled with solid carbon dioxide in order to reduce thermal noise. A minor fluorescent impurity present in the AOT had a decay rate of ca. 2 x lo8 s-l. At 580 nm the fluorescence intensity due to the impurity was always <3% of that from the *Ru(bp):+ and was unaffected by the presence of water or Fe(CN)g-. The first 10 ns of the decay was rejected and the effect of the impurity was then ignored. The data analysis procedure has been previously described. 23 PCS Experiments The photon correlation spectrometer and data analysis procedures have been described previo~sly.~~ For temperature-variation measurements at fixed scattering angle (0 = 90°) samples were contained in rectangular 10 mm Hellma quartz fluorescence cells.Angular dependence measurements were made using a 10 mm diameter cylindrical cell made from precision-bored glass tubing. The cell was mounted in a transparent dish containing water which was accurately thermostatted (to f O . l "C) by circulation through an immersion coil in a Haake thermostat. Measurements were possible over an angular range of 50 to 130" using this equipment. Theory for the Fluorescence Quenching Experiments The theory, in its simplest form, enables the concentration of discrete droplets in a microemulsion to be evaluated.It is based on the theory developed by Infelta28 for aqueous micellar systems and subsequently used by Atik and Thomas22 for AOT- stabilised water-in-oil microemulsions. There is no detectable solubility, as determined by u.v.-visible absorption, of Ru(bp)i+ or Fe(CN)i- in AOT-alkane mixtures in the absence of water, and so it is assumed that the fluorescer and quencher ions are partitioned entirely into the aqueous core of the droplets. To avoid complications due to partitioning to the interface, and subsequent facilitated transfer, Fe(CN)g- was preferred over methyl viologen. The quencher is assumed to be distributed between the water-droplet cores according to a Poisson (or random) distribution.The probability P(n) that a water droplet contains n quencher ions is then given by - (1) nn exp ( - R ) n! P(n) = where A is the average number of quencher ions per droplet. The quenching process within a droplet may be represented as in scheme 2, where k,, (s-l) is the first-order Scheme 2.1012 Fluorescence Quenching in Water-in-oil Microemulsions rate constant for quenching of a fluorescer by a single quencher ion in a droplet. The first-order rate constant, k,,, for the quenching process is dependent on droplet size according to k~~ = k ~ [ Q l ~ where [QIM is the concentration of a quencher ion in an individual droplet (expressed as moles of Q per dm3 of water inside a droplet) and kQ (dm3 mol-1 s-l) is the second-order rate constant for the quenching reaction inside the droplet, which may be compared with that in bulk aqueous solution.The back electron-transfer reaction to regenerate the starting materials F and Q is a very rapid process. It might be expected that k, would be close to the value of 6.5 x lo9 dm3 mol-1 s-l obtained in 0.5 mol dm-3 aqueous NaCl at 25 0C.29 This value is close to the diffusion-controlled limit and would suggest efficient quenching within small droplets such as are present in the H,O- AOT-dodecane system. The experimental conditions are such that, in general, only a small percentage of the droplets will contain a fluorescer molecule, whereas the concentration of quenchers is consistent with occasional multiple occupancy. The fluorescence intensity in the absence of quencher I, is given by (2) a I, = I(t = 0) exp (-k, t)dt (3) I-, and in the presence of n quencher ions per droplet I , by 02 r a IQ = I(? = 0) c Pn J exp{-(kl+nkQM+k,,[D]~)t}dt 12-0 t=o (4) where kyl is the fluorescence lifetime in the absence of quencher ions, k,, is the second-order rate constant for the exchange of solubilisates (quencher) between the aqueous droplet cores,l [D] is the concentration of droplets and P is the probability that a droplet contains at least one quencher ion.The ratio of intensities in the presence and absence of quencher is given by If the following condition holds : kQM %' k1 % kEXIDIP (A) then quenching is very efficient, there is no fluorescence from pools containing one or more quencher ions, and the distribution of the fluorescer and quencher ions between the droplets is static on the timescale of the fluorescence decay.The terms in eqn ( 5 ) are then equal to zero for n > 0 and exp(-A) for n = 0. The ratio of intensities in the presence and absence of quencher then reduces to ln(I,/I,) = ~i (6) where = [QI/Pl (7) and [Q] is the total concentration of quencher ions in the microemulsion (expressed as mol dm-3 of total solution). Therefore a plot of ln(I,/I,) against [Q] should be linear, the reciprocal gradient being equal to the discrete droplet concentration [D]. The radius of the droplet aqueous core r may now be calculated using andA . M . Howe, J . A . McDonald and B. H . Robinson 1013 ... HZO 80 90 dodecane \ Fig. 2. Phase diagram (plotted to scale as volume fraction) for the H,O-AOT-dodecane system at 25 "C.cp is the critical point. where p is the density of water, rn is the molar mass of water and N is Avogadro's number. An average value for the radius of the water core is then obtained. (Note that the method is essentially independent of inter-droplet interactions.) However, the simple theoretical approach used is restricted by condition (A) to the situation where there is no detectable fluorescence from droplets containing one or more quencher ions and there is no exchange of quencher ions between droplets within the fluorescence lifetime of *Ru(bp)i+ (p). It is further assumed that the droplet size range is not so polydisperse tha,t condition (A) breaks down and also that any polydispersity is essentially independent of quencher concentration.Results Phase Diagrams The phase diagram for the H,O-AOT-dodecane system at 25 "C in the oil-rich region of the phase diagram is shown in fig. 2. A single-phase w/o microemulsion region, denoted L,, is located towards the AOT-dodecane axis of the phase triangle. The two-phase w/o microemulsion region 2L2 is observed at higher volume fractions of water, adjacent to the single-phase region. Further increase in the water content results in the separation of a water-continuous, liquid-crystalline phase to form a three-phase region (2L2 + LC). At low volume fractions of surfactant a single w/o microemulsion coexists with a liquid-crystalline phase (L, + LC). Addition of high concentrations of quencher ions (ca. mol dm-3) to a mixture of composition Xc at 25 "C in the absence of quencher is found to increase T, (to ca.40 "C). The shift in T, at lower concentrations of quencher ( 6 2 x mol dm-3) is insignificant ( < 3 "C). PCS Measurements Measurements were carried out on a microemulsion containing 4.9 % v/v H,O ( R = 18), 5.8 % AOT (0.15 mol dm-3) and 89.3 % dodecane, this composition corresponding to the1014 Fluorescence Quenching in Water-in-oil Microernulsions Table 1. PCS measurements on a microemulsion at the critical composition X, (4.9% H,O v/v, 5.8% AOTand 89.3% dodecane) approaching the critical temperature T, = 25 "C T/"C q/10m3 kg m-' s-l t/103 s-l c/nm K( 5.3 6.1 7.1 8.2 9.5 10.9 12.1 14.6 18.1 20.1 21.1 21.9 23.2 24.1 3.18 3.18 3.18 3.18 3.188 3.192 3.207 3.255 3.365 3.445 3.486 3.525 3.592 3.647 11.40f0.15 1 1.33 f 0.09 11.18f0.11 10.48 f 0.07 10.28 f 0.08 9.008 f 0.08 8.953 f 0.081 7.786 & 0.066 6.138 k 0.056 5.068 & 0.03 1 4.541 k0.034 4.135 f 0.026 3.566 f 0.012 3.063 0.01 1 ~~ 7.70 7.77 7.48 8.51 8.72 10.06 10.13 11.72 15.02 18.63 21.32 24.08 29.87 39.72 0.1 0.1 0.1 0.2 0.2 0.2 0.2 0.3 0.3 0.4 0.5 0.6 0.7 1 .o critical composition at 25 "C.The dynamic correlation length c was measured as a function of increasing temperature approaching T,. Correlation lengths were obtained using the Kawasaki equation,30 which relates the decay rate r, in units of s-l, of the intensity autocorrelation function to the scaled correlation length K5. r = 2DPH0(K5) (9) (10) where K is the scattering wavevector: K = (4nn/A) sin (8/2) and n is the refractive index of the medium, ;3.is the wavelength of the incident radiation and 8 is the scattering angle. The function Ho(Kc) is given by (1 1) Ho(X) = 0.7531 /X+ 1 / X + (1 - 1 /X4) arctan x] with X = K t . The collective diffusion coefficient, D, may be related to by where k is the Boltzmann constant and 71, is the shear viscosity. In our analysis we have made the approximation qs = qsoln where qsoln is the macroscopic solution viscosity. (Note that the solution viscosity increases with temperature, as shown in table 1.) The data were analysed in the non-local hydrodynamic region, i.e. for Kc < 1, and are summarised in table 1 . Note that for K l 4 1 (the hydrodynamic region) the term H,(K<) tends to unity and < can be equated with the hydrodynamic radius of a single droplet. We observe an increase of approximately a factor of five in the correlation length on increasing the temperature to T, over the range 5-24 "C.Above 24 "C, in the critical region, K5 > 1 and eqn (9) then reduces30 to = 2 A P (1 3) where A = kT/16qs. (14) According to eqn (1 3) the decay rate should have a strong angular dependence in the critical region. Fig. 3 shows a plot of the observed decay rate against K3 for the critical sample at 25 "C; good agreement with eqn (13) is found.A . M . Howe, J . A . McDonald and B. H . Robinson 1015 J 10 20 30 40 K3/ 1 O6 nm-3 Fig. 3. Plot of the observed decay rate z of the intensity correlation function against k3 for the microemulsion at the critical point (Xc, TJ. SFQ and DFQ Measurements Distribution of Quencher Ions Results are reported for an H,O-AOT-heptane microemulsion at 25 "C.The essential difference between the microemulsions in the different alkanes is that the heptane system, at the compositions studied, is far from a phase boundary and critical point. If the quencher ions are distributed amongst the water pools according to a Poisson distribution and condition (A) holds, then eqn (6) follows. A linear plot of ln(Io/IQ) against quencher concentration is obtained for an R = 10,O. 1 mol dm-3 AOT-in-heptane microemulsion as shown in fig. 4. This indicates that the quencher ions are distributed according to Poisson statistics, in good agreement with the results of other workers.22 The pool concentration, from fig. 4 for an R = 10, 0.1 mol dm-3 AOT-in-heptane microemulsion, is (6.5k0.4) x lov4 mol dmW3 at 25 "C.The water pool radius r, calcu- lated from eqn (9, is then 2.2 nm; this value compares reasonably well with other determinations of 2.4 nm2, and 2.0 nm23 by fluorescence quenching and 2.6 nm by SANS studies.2 A higher value for the radius of the droplet core may be expected from SANS measurements on a D,O-(H)AOT-(H) alkane system. In the SANS contrast profile the sulphonate head groups of the AOT were included in the core,, but the volume occupied by the AOT head groups may not be available to the fluorescer or quencher. The pool1016 Fluorescence Quenching in Water-in-oil Microemulsions 2.0 - 1.5 - n < 3 fr 1.0 - ov I I I 0 0.5 1 .o 1.5 [ Fe(CN):-] / 1 0-3 mol dm-3 Fig.4. Plot of ln(Io/IQ) us. quencher concentration from SFQ experiments in a microemulsion (R = 10, 0.1 mol dm-3 AOT in heptane) at 25 "C. concentration increases to (8.010.5) x rnol dm-3 and (12.0+0.6) x lop4 mol dm-3, and r decreases to 1.9 and 1.4 nm as R is decreased to 8 and 5, respectively, for 0.1 mol dm-3 AOT in heptane at 25 "C (see fig. 5 for the radii determinations and fig. 8 for the concentration determinations). From the droplet size data at low R the sizes of larger droplets may be calculated using the simple geometrical ' head-and-shoulders ' model proposed by Bridge and Fletcher,23 providing the proportion f of AOT molecules at the interface remains constant at a constant temperature in a given oil. (Recent SAXS studies have demonstrated the presence of small numbers of reversed micelles of AOT coexisting with the surfactant- coated d r o p l e t ~ .~ ~ ) The model assumes that if the AOT molecules are close packed at the droplet surface then this close packing will occur not at the sulphonate head group but rather at the branched hydrocarbon tail, corresponding to the 'shoulders' of cross-sectional area s per AOT molecule. If the shoulders are at a distance (r + d ) from the centre of the droplet the value of d may be calculated by A value of 0.73 kO.01 nm is obtained for d, slightly higher than the 0.50 nm reported by Bridge and Fletcher.23 The value of sf(fis the fraction of AOT partitioned into the surfactant shell around the droplet) is calculated to be 0.73 kO.01 nm2 from sf= 47~(r+d)~[Dl/[AOT]. (16) The value of sfagrees well with that obtained by Bridge and Fletcher23 of 0.76 nm2 and is close to the value of the interfacial area per molecule at the planar heptane-aqueousA .M. Howe, J. A . McDonald and B. H . Robinson I 1017 20 40 60 R Fig. 5. Plot of water pool radius against R, predicted from application of the ' head-and-shoulders ' model to SFQ measurements at low R (R = 5-10, 0.1 mol dmP3 AOT in heptane at 25 "C). Also included are core-size data obtained by other workers (0, SANS;2 A, SANS14 and 0, SANS38). NaCl ( >/ 0.5 mol dmP3) interface.13 The values obtained for d and sfallow us to calculate the water core radius r (in nm) as a function of the water-to-surfactant ratio, R (fig. 5). Combining eqn (8) and (16): 1 1. 1sfr3 R = - n(r + d)2 ' The predictions agree well with the results of other workers using SANS measurements (fig.5) and a range of other techniques.22* 2 5 v 3 1 9 37 (17) Addition of Toluene and Benzyl Alcohol Toluene was added, to 40% v/v (3.8 mol dmP3), to the R = 10, 0.1 mol dm-3 AOT- in-heptane microemulsion system. Toluene has no effect on the luminescence lifetime of *Ru(bp)i+; in the presence of Fe(CN)i- DFQ experiments indicate that condition (A) holds (fig. 6). The SFQ measurements then yield the droplet concentration directly. Plots of In (I,JIQ) against quencher concentration are given in fig. 7. The droplet concentration appears to increase by ca. 30% from (6.5 & 0.4) x loP4 mol dm-3 in the absence of toluene to (8.7 kO.4) x lop4 mol dmP3 at 40% toluene. This corresponds to a decrease in the water pool radius from 2.2 to 2.0 nm and an increase in water-AOT interfacial area of ca.20 % . The decrease in droplet radius on addition of toluene has been previously reported from ultracentrifuge, viscosity and PCS 33 The results suggest that toluene penetrates between the tails of the AOT molecules at the AOT-oil interface, a result in good agreement with the conclusion of recent SANS studies4 and n.m.r. e~periments.~~3 35 Benzyl alcohol was added at concentrations up to 0.5 mol dm-3 (5.2% v/v) to the R = 10, 0.1 mol dm-3 AOT-in-heptane microemulsion. This leads to changes in the absorption and emission spectra of Ru(bp)z+. The 452 nm peak in the absorption maximum is shifted to 454 nm, the extinction coefficient is increased by 5% and the shoulder at 430 nm is decreased in the presence of 0.5 mol dm-3 benzyl alcohol. The1018 4 3 n h c1 .L.2 .E 2 W M - 1 Fluorescence Quenching in Water-in-oil Microemulsions a Fig. 6. Luminescence decay of *Ru(bp):+ in the presence of a range of concentrations of quencher ion in a microemulsion ( R = 10, 0.1 mol dm-3 AOT, 2.82 rnol dmP3 toluene in heptane) at 25 "C. (a) No quencher, (b) [Fe(CN):-] = 4.2 x lo-* mol dm-3, (c) [Fe(CN)i-] = 1.32 x mol dm-3. emission spectrum of *Ru(bp)z+ in the presence and absence of 0.5 mol dm-3 benzyl alcohol is given in fig. 1. The emission peak is doubled in intensity and shifted from 618 to 606 nm in the presence of 0.5 mol dmU3 benzyl alcohol. In addition, the luminescence lifetime increases from 5 15 to 720 ns on addition of 0.3 mol dme3 benzyl alcohol to the R = 10 system.Although quite soluble in water and heptane, benzyl alcohol is very soluble in a microemulsion or an AOT-heptane mixture, which suggests that the benzyl alcohol is preferentially partitioned into the interfacial region (in good agreement with recent SANS studies on the effect of benzyl alcohol on droplet size4). Ru(bp):+ is much more soluble in benzyl alcohol than concentrated salt solutions, and may therefore be expected to be partitioned into the interfacial region on addition of benzyl alcohol to the microemulsion. [Ru(bp):+ is not soluble in AOT-heptane mixtures, but it is on addition of benzyl alcohol.] At a concentration of 0.3 mol dm-3 benzyl alcohol in an R = 10, 0.1 mol dm-3 AOT-in-heptane microemulsion, DFQ results indicate that k,, $= k,, but the long-time decay rate is increased from 1.38 x lo6 s-l (k,) in the absence of quencher to 1.91 x lo6 and 2.67 x lo6 s-l (k,) in the presence of 4.2 x lop4 and 1.23 x mol dmP3 quencher ions, respectively. The quencher-dependent rate constant probably indicates that the distribution of fluorescer and/or quencher ions is no longer static on the timescale of the experiment.A value for the second-order rate constant for inter-droplet quenching kIT (i.e. for quenching of a fluorescer by a quencher ion for the case in which the two species are not originally in the same droplet) may then be estimated using (18) k , = k1 +kIT[D] P. The droplet concentration for the R = 10,O. 1 mol dmP3 AOT-in-heptane microemulsion in the presence of 0.3 mol dmP3 benzyl alcohol is ca.1.1 x mol dm-3 from SFQ experiments. If a Poisson distribution of quencher ions between droplets is assumed (see above) then a value of ca. 1.5 x lo9 dm3 mol-1 s-l is obtained for kIT.A . M . Howe, J . A. McDonald and B. H. Robinson 1019 2.c 1.5 h 5 3 ti 1.0 0 0.5 1 .o 1.5 [Fe(CN)i-l/10-3 mol dm-3 Fig. 7. Plot of In (l,,/l,) us. quencher concentration from SFQ experiments in a microemulsion ( R = 10, 0.1 mol dm-3 AOT in heptane) with a range concentrations of added toluene at 25 "C. [toluene]/mol dmP3 = 0, 0.47; 0, 0.94; +, 1.88; 0, 2.82; 0, 3.76. Variation of R Plots of In (IO/IQ) against quencher concentration from static fluorescence experiments for R = 5 to 50 are shown in fig. 8. The gradients of these plots exhibit a maximum, and the corresponding apparent pool concentration a minimum at R = 20.For R > 10 the plots of In (IO/IQ) against [Q] start to exhibit downward curvature at high quencher concentration, which is an indication that the intra-droplet quenching efficiency is significantly less than unity when there is only one quencher in the pool. Hence the simple interpretation of SFQ data, based on eqn (6), cannot be used. The pool concentration, as measured by this technique, is therefore increasingly overestimated and the radius underestimated as R is increased. At the limit eqn ( 5 ) then reduces to1020 Fluorescence Quenching in Water-in-oil Microemulsions 0.5 0 Fig. 8. Plot of In (I,,/I,) us. quencher concentration from SFQ experiments in microemulsions, with a range of R values from 5 to 50, stabilised by 0.1 mol dmP3 AOT in heptane at 25 "C.R = 0, 5 ; +, 8; 0, 20; U, 30; 0, 40; a, 50. A correction term C [which increases with B, hence the curvature in In (lo/lQ) us. [Q] plots, and rapidly converges at high n / 4 must be introduced into eqn (6) to take account of the incomplete quenching within the droplets. Then where and if an accurate value of kQM is known from DFQ experiments the droplet concentration may be determined from the SFQ data. At R = 30, kQ, z k,22 and for FZ = 1, C z 0.75 while C z 2.18 for z = 2, thus the curvature in the plots of In (Io/IQ) against [Q] at high [Q] for large R values are explained. Using eqn (21), a value of ca. 1.8 x mol dm-3 at 25 "C is obtained for [D] at R = 30,A .M. Howe, J . A . McDonald and B. H . Robinson 1021 which corresponds to a water-core radius at 4.9 nm, in good agreement with the findings of other workers and the 'head-and-shoulders' prediction (fig. 5). A value of 5.6 x lo8 dm3 mol-l s-l may be calculated for k,, which is an order of magnitude lower than in aqueous solution at 0.5 mol dm-3 NaCl. This lower value in the microemulsion droplet may be a result of the high viscosity of the water in the droplet,23 which is more significant for small droplets, and partitioning of the oppositely charged fluorescer and quencher species within the droplet as a result of the negatively charged AOT,23 an effect which may be more significant at high R. The limiting droplet size at which the SFQ technique is strictly valid for the fluorescer/quencher couple used here may now be estimated to be when quenching within a droplet is very efficient [and inter-droplet exchange is slow, condition (A)], i.e.thus where k , is the rate of luminescence decay in the absence of quencher. The value of k , has been determined by the DFQ technique to be 1.88 x lo6 s-l at 25 "C in an R = 20 AOT-in-heptane microemulsion. [QIM is the concentration of a single quencher ion within a droplet and k , is the second-order rate constant for the quenching reaction (which has a value close to 5.6 x lo8 dm3 mol-l s-l for R = 30 at 25 "C). Therefore the maximum droplet volume Vmax for the application of the simple theory above is given by Vmax = k,/ 10k, N r3 < 3kQ/40nk,N (25) (24) and so the values of the droplet water-core radius that may be determined by the chosen couple using SFQ is given by and Y < 2.3 nm.It can be concluded therefore that the SFQ method, using the Ru(bp)i+/Fe(CN)i- couple is not valid for the larger-droplet systems, i.e. for R > 12, unless the correction term which allows for incomplete quenching within a droplet is considered. Approaching a Critical Point SFQ studies were carried out on a microemulsion system of fixed composition 4.9 % H 2 0 ( R = 1 8), 5.8 % (0.15 mol dm-3) AOT and 89.3 % alkane (% volume) for the n-alkanes heptane and dodecane. Measurements were therefore made as a function of temperature approaching the critical point for the dodecane system at 25 "C. The same composition in the heptane system serves as a control, since the composition of the system is far removed from a critical point or phase boundary over the same temperature range.In order not to change T, for microemulsions of the above composition in dodecane the SFQ experiments were restricted to low concentrations of quencher (< 2 x 1 0-4 mol dmP3). Plots of In (I,,/I,) against quencher concentration for the dod- ecane system are given in fig. 9(a) (with expansion) and the heptane system in fig. 9(b). As the temperature is increased towards 25 "C the gradients of both plots increase corresponding, on the basis of condition (A), to an apparent reduction in pool concentration from 7.4 x lop4 mol dm-3 at 5.4 "C to 3.4 x loh4 mol dm-3 at 25.0 "C in dodecane and from 9.1 x lop4 mol dm-3 at 5.4 "C to 7.0 x mol dm-3 at 25.0 "C in heptane.The corresponding apparent increase in r, calculated using eqn (8), with temperature is from 3.0 nm at 5.4 "C to 3.8 nm at 25.0 "C in dodecane and 2.8 nm at 5.4 "C to 3.0 nm at 25.0 "C in heptane. The values of apparent pool concentration and apparent size of the aqueous core are given in table 2 for all the microemulsion systems1022 Fluorescence Quenching in Water-in-oil Microemulsions I [ Fe(CN)i-]/ lo-' mol dm-3 i (a) 5 10 [Fe(CN)2-]/10-4 mol dm-3 Fig. 9. Plot of In (&/IQ) against quencher concentration for a microemulsion of composition 4.9% H,O, 5.8% AOT and 89.3% alkane on increasing the temperature from 5.4 (O), through 15.0 (A) to 25.0 (0) "C. (a) Dodecane approaching T, (lower plot shows results at high [Q]; (6) heptane (control) 5.4 (A) to 25 (0).A .M . Howe, J. A . McDonald and B. H . Robinson 1023 Table 2. Apparent droplet concentration [D] and apparent water core radius r for microemulsions of composition 4.9% v/v H,O, 5.8% AOT and 89.3% dodecane or heptane dodecane heptane T/"C [D]/ loP4 mol dmF3 r/nm [D]/10P4 moi dm-3 r/nm 5.4 7.4 (5.6) 3.0 (3.3) 9.1 (6.7) 2.8 (3.1) 25.0 3.4 (2.7) 3.8 (4.1) 7.0 (5.3) 3.0 (3.3) -1 5.0 4.8 (3.7) 3.4 (3.7) a Values in the table are based on analysis using eqn (6) and values in parentheses are based on eqn (20). studied. A value of ca. 3.3 nm is predicted for r at 25 "C from fig. 5 for the heptane system, the apparently smaller value obtained directly from the SFQ measurements is a result of insufficiently rapid intra-droplet quenching. It may be concluded that the radius of the droplets in the heptane-continuous system is essentially unchanged over the temperature range studied.As lower quencher concentrations were used in the dodecane systems any corrections for incomplete quenching will be less than for the heptane system. A value of r greater than expected for the dodecane system may be explained as an increase in droplet size (table 2), but may also be an indication that the distribution of F and Q is no longer static on the ps timescale, and the implications of this will be discussed in the next section. A slight complication is that the critical composition in the dodecane system is sensitive to the quencher ion concentration. There is a shift in the phase boundary for microemulsion stability with quencher in such a manner that, for an increase in quencher concentration, the value of T, is higher at a given Xc.For low quencher concentrations (2 x mol dm-3) the shift in the phase boundary is insignificant (< 3 "C). On examination some curvature may be noted in the ln(Z,,/ZQ) plot at high (> 5 x mol dm-3 quencher ion concentration the phase boundary is at 41 "C, the apparent value of [D] is 4.5 x lop4 mol dm-3 and r is 3.5 nm. Since the presence of high quencher concentrations shifts the microemulsion phase boundary and its associated critical point further away from the chosen microemulsion composition this result is perhaps to be expected. As the quencher concentration is increased the microemulsion is increasingly ' non-critical '.Note that the SFQ method provides direct information on the apparent droplet concentration. SANS provides information on polydispersity and inter-droplet inter- actions but is not especially effective for the determination of droplet concentrations. Using PCS an apparent correlation length is determined (table 1) which in the limit (at infinite dilution) may be identified with the hydrodynamic radius. However, this is found to increase by a factor of five from 5.4 to 24.1 "C for the dodecane system. mol dmP3) quencher concentration [insert to fig. 9(a)]. At 1 x Discussion The droplet size measurements as a function of the molar ratio of water to AOT (R) at constant temperature and AOT concentration agree well with those reported by other workers using SFQ and SANS for the low-R, non-critical microemulsions (fig.5). The radii of larger droplets (R > 10) are accurately predicted from the low-R data by the ' head-and-shoulders ' model. Addition of toluene was found to increase the droplet concentration, probably as a result of penetration of toluene molecules between the shoulders of the interfacially bound AOT. From considerations of phase behaviour and kinetics toluene is thought1024 Fluorescence Quenching in Water-in-oil Microemulsions . I / \I/ -9-f \ kl T Fig. 10. Two quenching mechanisms, (a) and (b), for quenching in the *Ru(bp)i+/Fe(CN):- system in AOT-in-heptane microemulsions at high concentrations of benzyl alcohol. to make the surfactant shell more convex to the heptane continuous phase.l* The SFQ measurements support this idea.Benzyl alcohol is thought to act as a cosurfactant, hence reducing R and droplet size ( i e . droplet concentration increases). At high concentrations benzyl alcohol has a more specific effect on the FQ experiments because the Ru(bp)i+ is now partitioned into the interfacial region. The second-order rate constant for transport of the excited fluorescer between droplets (kIT) was ca. 1.5 x lo9 dm3 mol-' s-l, a factor of 50 greater than the pool-exchange rate constant obtained with reactants Ir(C1):- and Fe(CN):-, which are confined to the droplet water core.' A possible mechanism for this reaction involves the transfer of *Ru(bp):+ between the surfactant shells of different droplets during the contact between droplets following collision.The proposed mechanism is given in fig. 10 and is similar to that which has been proposed for transfer of some viologens between For the microemulsions of composition 0.15 mol dmP3 AOT, R = 18 in dodecane, the SFQ measurements indicate an apparent increase in droplet size as the critical point is approached. There is little change in the apparent droplet size observed for the non-critical heptane system over the same temperature range. At R = 18 it is necessary to allow for incomplete quenching within the droplets. At this relatively low R value the correction is found to be small (ca. 10% to Y) and plots of ln(Io/IQ) against [Q] do not appear to exhibit any downward curvature over the range of quencher concentration used. However, the effect of the change in temperature on k , and k,, must also be considered. Both rate constants will decrease as the temperature is decreased, and if a larger activation energy is associated with the quenching process then an apparent decrease in droplet size on decreasing temperature can be explained. However, SANS measurements indicate that there is little change in droplet size with increase in tem~erature.~3 l4A .M. Howe, J. A . McDonald and B. H . Robinson 1025 3 0 'dimer' k$2 (fusion) +o+-oo -EN encounter pair f0 k-EN k 2 - k-2 0 't r i rner ' further growth 000 - further clustering Fig. 11. Possible dynamic processes representing clustering or growth in a microemulsion. A further assumption involved in taking the droplet concentration directly from the SFQ measurements is that the rate of transfer of quencher ions between droplets is assumed to be slow on the timescale of the fluorescence decay, i.e. k , > lOk,,[D]P.For the non-critical heptane systems k,, has been determined to be ca. 1 x lo7 dm3 mol-l s-l at 25 OC.' On substitution into condition (C) it can be seen that this condition is fulfilled for all measurements in heptane, since [D] < 1 x mol dmP3. For the dodecane system k,, has been determined to be 7 x lo6 dm3 mol-1 s-l at 5 "C for an R = 20 microemu1sion.l The activation energy for the exchange process in R = 20 microemulsions far from the upper-temperature phase transition is 77 kJ mol-l, and a value of ca. 7 x lo7 dm3 mo1-1 s-' may be predicted for k , , at 25 "C for the dodecane system, in the absence of strong inter-droplet interactions.However, the situation for the critical microemulsions is less clear-cut, since the onset of critical behaviour may radically increase the rate of droplet exchange. If inter-droplet exchange of quencher were to occur on the timescale of the luminescence decay, then from eqn ( 5 ) and (20): Therefore, as the value of k,,[D] P approaches k,, the value of In (Io/IQ) increases with increasing [Q], which could apparently indicate a decrease in [D] and an increase in r. We cannot investigate pool exchange in the critical region by simple fast-flow kinetic methods, such as were used in ref. (I), as the reactions are too fast (ca. ps), although a full DFQ study should yield values for both the exchange and quenching rate constants on approach to the critical point.A value of k,, can, however, be estimated. If exchange of quencher ions occurs on the p s timescale (i.e. comparable with k ~ l ) then k , , 2 lo9 dm3 mol-l s-l, at least an order of magnitude greater than predicted from the results determined far from the critical point. The onset of the various growth and clustering processes approaching the critical point are represented schematically in fig. 1 1 . Exchange will occur during transient fusion of the drop1ets.l The growth process is only favoured if the fused species are relatively1026 Fluorescence Quenching in Water-in-oil Microemulsions long-lived. If this is the case then observed In (I,/I,) values would be expected to increase with temperature owing to incomplete droplet quenching [from eqn (20)].This would indicate an apparent decrease in droplet size in direct contrast with the apparent increase observed experimentally. For the heptane system (cf. fig. 11) where KEN = kEN/k-EN, KEN (or more particularly k-EN) will depend on the nature of the interaction potential between the droplets. SANS studies of the structure factor S(Q) in the heptane system suggest that there is a small net attractive interaction (of ca. 20 kJ m ~ l - l ) . ~ ~ We can then estimate KEN to be of the order 103-104 dm3 mol-l, and then k , can be calculated to be 103-104 s-l. In the case of the dodecane system close to the critical point, if extensive clustering does occur then KEN and, in effect, [D], would be expected to increase within the localised environment of a droplet cluster.Hence, from eqn (27), k,, approaches k , and the observed k,, is determined by the first-order process Scheme 3. of fused droplet formation. Then the kinetic scheme can be effectively simplified to scheme 3. Then the condition which is relevant in our size analysis is k, > IOk, P. (D) If this condition does not hold, then the interpretation based on a slight size increase in the dodecane system as T, is approached would be incorrect, since in reality k E , can reasonably be expected to increase in the critical region. A droplet exchange rate on a ps timescale ( k , M lo6 s-l) as T, is approached would be consistent with the apparent observed increase in droplet size observed in the dodecane system over the temperature range 5.4-25 "C.An interesting supporting observation is that the electrical conductivity increases dramatically for the dodecane system as the critical point is approached, with a frequency cut-off in the MHz region.4o This implies a percolation-type phenomenon in the ~ystem.~-l, Since the water phase content of the system, at ca. 5 % , is relatively low it is likely that open, network-like rather than close-packed, spherically based superstructures must be present. Recent modelling of percolation phenomena in microemulsion systems suggests that the depth and range of the attractive inter-droplet interactions are important as well as the volume fraction of surfactant and water.1°-12 The conductivity has been reported to be frequency dependent at the percolation threshold for the H,O-AOT-decane system, with a characteristic time of ca.0.5 ps.l0 The implication is, once again, that temporal fusion of adjacent droplets in the critical regime could occur on a ps timescale. We conclude that droplet exchange is facilitated in the critical region and the change in size of the droplets is probably negligible. This is supported by analysis of the SANS form factor at high Q, which does not provide evidence for any significant size in~rease.~ The apparent increase in r for the dodecane system, which is greater than for the heptane system, is probably the result of a real increase in k,, in the critical region either as a result of a more facile fusion of droplets (an increase in k,) or as a result of 'sticky' collisions between droplets (an increase in KEN).What is clear is that in fig. 11 the clustering pathway to form open networks dominates over growth and is the prime factor determining the enhanced correlation length which is observed experimentally by scattering methods as the critical point is approached.A . M. Howe, J . A . McDonald and B. H . Robinson 1027 We are grateful to Dr N. J. Bridge for the use of the DFQ equipment and analysis of the DFQ data. We thank the S.E.R.C. for providing funds for the support of the PCS work. J. A. McD. and A. M. H. acknowledge support by S.E.R.C. studentships and A. M. H. also thanks Shell Research Ltd (Thornton Research Centre) for CASE support. We thank the S.E.R.C. (Biotechnology) for the funds to purchase the PCS equipment. We are grateful to A.Katsikides for making the viscosity measurements reported in table 1 and we also acknowledge useful discussions with R. Schomacker, University of Bielefeld. References 1 P. D. I. Fletcher, A. M. Howe and B. H. Robinson, J . Chem. SOC., Faraday Trans. I , 1987,83, 985. 2 B. H. Robinson, C. Toprakcioglu, J. C. Dore and P. Chieux, J. Chem. SOC., Faraday Trans. I , 1984, 3 H-F. Eicke, J. Colloid Interface Sci., 1979, 68, 440. 4 A. M. Howe, C. Toprakcioglu, J. C. Dore and B. H. Robinson, J . Chem. Soc., Faraday Trans. I, 1986, 5 C. Toprakcioglu, J. C. Dore, B. H. Robinson and A. M. Howe, J . Chem. SOC., Faraday Trans. 1, 1984, 6 S. A. Safran and L. A. Turkevich, Phys. Rev. Lett., 1983, 50, 1930. 7 M. J. Grimson and F. Honary, Phys. Lett. A, 1984, 102, 141. 8 B. Widom, J.Chem. Phys., 1984, 81, 1030. 9 M. A. van Dijk, Phys. Rev. Lett., 1985,55, 1003. 80, 13. 32, 241 1 . 80, 413. 10 S. Battacharya, J. P. Stokes, M. W. Kim and J. S. Huang, Phjis. Rezi. Lett., 1985, 55, 1884. 1 1 S. A. Safran, Phys. Rev. A, 1985, 32, 506. 12 A. L. R. Bug, S. A. Safran, G. S. Grest and I. Webman, Phys. Rev. Lett., 1985, 55, 1896. 13 R. Aveyard, B. P. Binks, S . Clark and J. Mead, J . Chem. Soc., Faraday Trans. I, 1986, 82, 125. 14 M. Kotlarchyk, S-H. Chen and J. S. Huang, J. Phys. Chem., 1982, 86, 3273. 15 J. S. Huang and M. W. Kim, in Proc. SPE-DOE Symp. Enhanrecl Oil Recovery, Tulsa, 1982 (SPE- 16 J. S. Huang and M. W. Kim, Phys. Rev. Lett., 198!. 47. 1462. 17 J. S. Huang and M. W. Kim, in Scattering Techniques Appiicd to Supramolecular and Non-equilibrium 18 M. W. Kim and J. S. Huang, Phys. Rev. B, 1382, 20, 2 X 3 . 19 J. Tabony, A. Drifford and A. de Guyer. Chem. Phvs. Lett., 1983, 96, 119. 20 M. Corti and V. Degiorgio, Phys. Ret.. Lett., 1980, 45, 1045. 21 M. Corti and V. Degiorgio, J . Phja. Chem.. 1981, 85, 1442. 22 S. S. Atik and J. K. Thomas, J . Am. Chem. SOC., 1981, 103, 3543. 23 N. J. Bridge and P. D. I. Fletcher, J . Chem. SOC.. Faraday Trans. I , 1983, 79, 2161. 24 J. M. Furois, P. Brochette and M-P. Pileni, J . Colloid Interface Sci., 1984, 97, 552. 25 M. P. Pileni, P. Brochette, B. Hickel and B. Lerebours, J . Colloid Interface Sci., 1984, 98, 549. 26 J . N. Demas, D. Diemente and E. W. Harris, J. Am. Chem. SOC., 1973, 95, 6864. 27 P. D. I. Fletcher, M. F. Gala1 and B. H. Robinson, J . Chem. Soc., Faraday Trans. I , 1984, 80, 3307. 28 P. P. Infelta, Chem. Phys. Lett., 1979, 61, 88. 29 C-T. Lin and N. Sutin, J. Phys. Chrm., 1976, 80, 97. 30 K . Kawasaki, Phys. Reg. A , 1970, 1, 1750. 31 A. N. North, J. C. Dore, J. A. McDonald, B. H. Robinson, R. K. Heenan and A. M. Howe, Colloids 32 R. A. Day. B. H. Robinson, J. H. R. Clarke and J. V. Doherty, J . Chem. SOC., Faraday Trans. I, 1979, 33 J. D. Nicholson, J. V. Doherty and J. H. R. Clarke, in Microemulsions, ed. I. D. Robb (Plenum Press, 34 C. A. Martin and L. J. Magid, J. Phys. Chem., 1981, 85, 3938. 35 A. N. Maitra, G. Vasta and H-F. Eicke, J . Colloid Interface Sci., 1983, 93, 383. 06 P. D. I . Fletcher and B. H. Robinson, Chem. SOC. Rev., in press. 37 E. Key and B. Valeur, J . Colloid Interface Sci., 1981, 79, 465. 38 C. Cabos and P. Delord, J. Appl. Crystallogr., 1979, 12, 502. 39 J. S. Huang, J . Chem. Phys., 1985, 82, 480. 40 A. Katsikides (University of Kent), unpublished results. ASME, Dallas, 1982), p. 901. Systems, ed. S-H. Chen, B. Chu and R. Nassal (Plenum Press, New York, 1982), p. 809. Surf., 1986, 19, 21. 75, 132. New York, 1981), p. 33. Paper 61458; Received 6th March, 1986

ISSN:0300-9599    

DOI:10.1039/F19878301007

出版商:RSC    

年代:1987

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. 1, 1987, 83, 1029-1039 Setschenow Coefficients for Caffeine, Theophylline and Theobromine in Aqueous Electrolyte Solutions Pilar PCrez-Tejeda, Alfredo Maestre," Manuel Bal6n, JosC Hidalgo, Maria A. Muiioz and Miguel Sanchez Departamento de Fisicoquimica Aplicada, Facultad de Farmacia, Tramontana sln, 41012 Sevilla, Spain Setschenow coefficient measurements have been made on methylated xan- thines (caffeine, theophylline and theobromine) in several inorganic and organic aqueous salt solutions at 298 K. In general, the order of salting effect is similar for the three solutes and can be rationalized as a consequence of structural salt-water interactions. Scaled particle theory and the Cross and McTigue equation have been employed to compare theoretical and experimental salting constants.The methylated xanthines caffeine (1,3,7-trimethylxanthine), theophylline (1,3-dimethyl- xanthine) and theobromine (3,7-dimethylxanthine) (Scheme 1) are important biochemi- cal solutes because they present pharmacological propertied and are also stabilizers of some pharmaceutical preparations in aqueous solutions. R i = R Z = R , = C H j theophylline: R1 = R, = CH,; R, = H theobromine: R1 = H; R2 = R3 = CH, 0 I R, Scheme 1. Methylated xanthines. In recent years their physicochemical properties have received considerable atten- t i ~ n . ~ - ~ Caffeine is reported to form 1 : 1 mixed molecular complexes with different organic compounds containing aromatic ring ~ystems.~ Recently De Taeye and Zeegers6 have reported caffeine-water interactions by i.r.spectroscopy. However, studies of transfer parameters from water to aqueous salt solutions involving these methylxanthines are scarce. Cesaro et al.' have taken these properties into account to discuss the possible evidence for a direct interaction of caffeine with urea and with the guanidinium ion. However, the transfer parameters may also reveal other interactions besides that between a non-electrolyte and an electrolyte, such as non-electrolyte-solvent and electrolyte-solvent interactions. In this sense the Setschenow coefficient, or salting constant, k,, is a measure of the free energy of transfer from water to aqueous salt solutions and it reflects the modification of solute-solvent interactions by an added salt.8 For these reasons we believe that it may be interesting to study the effects of inorganic and organic salts on methylxanthines by measurements of Setschenow coefficients.Moreover, the experimental data have been considered from a theoretical point of view, given that these methylxanthines behave as non-electrolytes because they are nearly neutral in water and, with a pK, of ca. 9-14, are practically un-ionized in aqueous solutions. 10291030 Se tscheno w Coeficien ts Experimental Reagents Anhydrous caffeine, theophylline and theobromine were of the purest grade available from commercial sources (Sigma) and were used after drying at 85 "C and were kept in a vacuum desiccator with silica gel prior to use. The salts employed (LiC1, NaCl, KC1, KBr and KF) were reagent grade (Merck) and were used after drying overnight at 110 "C.Tetra-alkylammonium bromides were reagent grade (Fluka) and their water content was checked by passing through a cation-exchange resin (Fluka, Dowex 50WX8, p.a.) and titrimetric determination with NaOH, using phenolphthalein as indicator. The extractant employed, chloroform, was reagent grade (Merck) and was used as received. Doubly distilled water was used in every experiment. Distribution Measurements Distribution ratios of a known amount of solute (9 x lop5 mol dme3) among water, or salt solution, and chloroform were determined by vigorous agitation of small aliquots. After equilibrium was reached (ca. 15 min), the agitation was stopped and the systems were left in a water bath at 25.0 f 0.1 "C for 4 h.Aliquots from the aqueous phase were taken and analysed spectrophotometrically in a Perkin-Elmer Lambda 5 spectrophoto- meter (caffeine, 273 nm; theophylline, 271 nm and theobromine, 272 nm). At these wavelengths the molar absorption coefficients were 9.74 x lo3, 1.02 x lo4 and 1.01 x lo4 dm3 mol-l cm-l, respectively. The distribution ratios were taken as the average of at least three concordant determinations and the distribution constants in water, Do, for caffeine, theophylline and theobromine were 17.5, 0.33 and 0.53, respectively . Solubility Measurements The solubilities of caffeine, theophylline and theobromine in water and salt solutions were determined in the following manner. An excess amount of solute added to 0.025 dm3 of salt solution was vigorously shaken for at least 48 h in a water bath at 25.0k0.1 "C.After equilibrium was reached (2-3 days), the samples were filtered and after suitable dilutions the concentrations of the solutes were analysed spectrophoto- metrically as before. By this method, values of 0.106,0.035 and 2.57 x lop3 mol dmP3 for caffeine, theophylline and theobromine, respectively, were obtained in water and are in agreement with those found in a previous Results The salting effect at high dilution on a non-electrolyte can be estimated by solubility or distribution methods from the Setschenow equation : (1) where So and S are, respectively, the solubilities of the non-electrolyte in water and salt solutions, Do and D are distribution ratios in water and aqueous electrolyte solutions, c, is the concentration of salt and k, represents the salt-non-electrolyte interaction (salting constant or Setschenow coefficient), which has a characteristic value for a given salt-non-electrolyte pair. Values of logD/Do or log So/S are plotted against c, (fig.1-3) and the slopes and initial slopes of these lines give k, values on the molar scale (table 1). We have obtained values of k, for the NaCl-caffeine pair from solubility and distribution methods and the results are the same within experimental error (table 1). log So/S = logD/D, = k, C,P. Pkrez-Tejeda et al. 1031 I I I I 0.5 1 .o 1.5 2 .o c,/mol dm-3 - 0.80' Fig. 1. Variation of logD/D, or logS,/S with salt molarity, c,, for caffeine in LiCl (n), NaC1, ( X >, KC1 (A), KBr (O), KF (V), Me,NBr (H), Et,NBr (O), Pr,NBr (+) and Bu,NBr (A) at 25 "C. A comparison of the k, values in table 1 shows that the salting effect order is: caffeine theophylline t heo bromine Li+ < Na+ z K+ Na+ z K+ < Li+ (as chlorides) (as bromides) (as potassium) Bu4N+ < Pr,N+ < Et,N+ < Me,N+ Br- < C1- < F- the caffeine being more salted-out by alkine halides (except in LiCl) and more salted-in by R,NBr than are theophylline and theobromine. The influence of alkaline halide on the three methylxanthines is preferently due to a halide effect.In turn, k, values for the two isomers theophylline and theobromine are in general rather similar. The action of KF on theophylline at the lowest salt concentration produces a net salting-out and a salting-in at the higher ones.Usually, organic solutes (non-polar and polar) are systematically salted-out in inorganic aqueous salt solutions,1° but in the case of polyfunctional solutes the opposite effect may be observed, as is the case of the KBr-theobromine pair (table 1). Koreman and Bortnikovall have found similar effects on isomer series (nitrophenols, p-n-alkylphenols, monohalogenophenols and hydroxy- benzoic acids) with alkaline halides. In these isomer series the values and sign for k, depend on the location and kind of substituent.1032 Se tsc h en o w Coe,fjc ien t s 0.4r 0.21 h ? 2 0" 0.or 9 9 M - n 0 M - -0.2 -0.41 0 5 1 .o 1.5 2.0 c,/mol dm-3 Fig. 2. Variation of log DID,, or log S,/S with salt molarity, c,, for theophylline in LiCl (o), NaCl ( X ), KC1 (A), KBr (01, KF (V), Me,NBr (m), Et,NBr (O), Pr,NBr (+) and Bu,NBr (A) at 25 "C. Discussion The salting effect on a non-electrolyte by a given salt in water is governed by the overall effect of the mutual interaction among salt, non-electrolyte and water.The observed salting-out effect by alkali halides on the three methylxanthines may be explained by considering the combined action of two factors.12 On the one hand, water molecules are usually more polarized by the ionic field than non-electrolyte molecules are. Therefore, water will be more attracted by ions, resulting in an increase of the methylxanthine activity coefficient (ynel > 0 or salting-out). This attraction follows the order F- > Cl- > Br-, since the ionic field increases in this order. On the other hand, hydrophilic ions have a characteristic influence on water structure, their structure-making order being F- > Cl- > Br-.13 The overlap between the cospheres around ions and non-electrolytes may produce a repulsive interaction, also resulting in salting-out.l4 The decrease in k, with increasing size of tetra-alkylammonium bromide is caused by interactions different from those for alkali halides. Whereas polarization effects by the ionic field are negligible with R4N+, van der Waals dispersive forces become important with large non-electrolytes and these ions bring about a favorable interaction (salting-in). However, k, > 0 and k, z 0 for Me,N+ and Et,N+ on the three methyl- xanthines cannot be explained in these terms. Secondly, the sharing of R4N+ and methylxanthine cospheres may result in an attractive or repulsive interaction because of the amphiphilic nature of these non-electrolytes.There is probably a repulsive interaction in this overlap with Me,N+ and Et4N+ ions, which compensates for and is even greater than the dispersive forces. In short, it seems that the magnitude of the salting effect under study is connected withP. Pe'rez-Tejeda et al. 1033 -0.40' I I I I 0.5 1.0 1.5 2.0 c,/mol dm-3 Fig. 3. Variation of log DID,, or log S , / S with salt molarity, c,, for theobromine in LiCl (n), NaCl ( X 1, KCl (A), KBr (O), KF (V), Me,NBr (M), Et,NBr (O), Pr,NBr (+) and Bu,NBr (A) at 25 "C. Table 1. Salting constants k,(mol-l dm3) for caffeine, theophylline and theobromine at 25 "C salt caffeine theoph ylline theobromine LiCl NaCl KCl KBr KF Me,NBr Et,NBr Pr,NBr Bu,NBr 0.086 0.0 1 2" 0.142+0.004" 0.137f0.01 I' 0.148 0.009" 0.032 & 0.008" 0.358 Ifr 0.009" ca.0.000' - 0.056 & 0.012' - 0.13 1 2 0.034' - 0.428 & 0.01 6' 0.099f0.013" 0.071 kO.01 la 0.068 0.0 12" O.OOOb ca. 0.230' 0.075 f 0.01 1 ' ca. 0.000' - 0.103 0.025' - 0.41 6 & 0.016' 0.104 & 0.036" 0.063 f 0.027" 0.074f0.019" 0.196 f 0.007' 0.063 f0.017' ca. 0.00Ob - 0.028 f 0.007' - 0.065 & 0.005' - 0.320 f 0.027' " Calculated from distribution method. Calculated from the solubility method.1034 m E ‘D - - I E =l * 0.00 -0.20 -0.40 Setschenow Coeficients I I I 6o ( b ) 40 50 - - - I 1 I I -&(s)/10-7 dm3 mo1-l bar-’ Fig. 4. Plots of experimental k, with partial molar adiabatic compressibilities at infinite dilution &(,) for caffeine (a), theophylline (m) and theobromine (0): (a) KX (X = C1-, Br-, F-) from ref.(27) and (b) R,NBr (R = Me, Et, Pr, Bu) from ref. (15) at 25 “C. the way those salts modify the water structure. Partial molar adiabatic compressibilities at infinite dilution, dkcs,, give a sensitive indication of this modification.l5? l6 In fig. 4 plots of k, us. d&,, for the systems under study have been made. Except for theobromine with potassium halides, a linear correlation is found. Theoretical Prediction of Salting Coefficients In order to calculate quantitatively the action of electrolytes on non-electrolytes, several theories have been deve10ped.l~ The McDevit and Long approach has been extended by Cross and McTigue to enable the calculation of activity coefficients of polar non-electrolytes in concentrated solutions of electr01ytes.l~~ la The relative mole fraction activity coefficient, ynel, for the distribution of non-electrolyte between an organic and an aqueous phase is given by: limcnel-o (log YneJ = I &el cs(K -ds)1/2.303RTP where Cel is the molar volume of the non-electrolyte, c, is the molar concentration of the salt, is the intrinsic volume, &s is the apparent molar volume of salt in cs molarTable 2.Experimental and theoretical values of k , for caffeine and theophylline at 25 "C _ _ - _ _ _ _ _ _ _ ~ LiCl NaCl KCl KBr KF Me,NBr Et,NBr Pr,NBr Bu,NBr expt/mole fraction SPT Cross expt/mole fraction SPT Cross 0.094 0.287 3.150 0.107 0.268 0.133 0.151 0.396 0.194 0.080 0.365 0.173 0.I52 0.352 0.158 0.072 0.324 0.139 caffeine 0.033 0.370 0.306 0.402 0.125 0.364 theophylline 0.000 0.240 0.283 0.373 0.111 0.321 -0.031 -0.116 -0.219 -0.194 -0.158 -0.322 -0.373 -0.546 -0.607 0.041 -0.066 -0.191 -0.186 -0.162 -0.319 -0.195 -0.282 -0.302 -0.544 - 0.526 - 0.046 - 0.532 -0.514 - 0.024 P 11036 Setschenow Coeflcients Table 3. Parameters used in SPT caffeine : 92Sa 7.72' 19.3c theophylline : 93.1a 7.34' 16.2c Li+ Na+ K+ F- c1- Br- Me,N+ Et,N+ Pr,N+ Bu,N+ 1 .8gd 2.34d 2.9gd 2.32d 3.2gd 3.60d 5.02f 6.16f 6.98f 7.62f 0.079 0.2 1 " 0.87" 1 .04" 3.02" 4.1 7" 9.559 16.909 24.209 31.509 a Calculated from ref. (35). Calculated from its molar volume. Obtained by adding bond refraction given in ref. (20) and (36). Values taken from Gouray and Adrian in ref. (13), p.65. " From ref. (31). f From ref. (37). 9 From ref. (38). salt solution, p is the isothermal compressibility of a c, molar salt solution, rnel is the distance of closest approach to the non-electrolyte molecule and P, is the distance of closest approach to the average solvated ion present in solution. The parameters used in eqn (2) were: Gel = 0.1442 dm3 mol-1 for caffeine and Kel = 0.1240 dm3 mol-' for the~phylline.~? l9 The V, values were taken from ref. (1 8) for LEI, NaCl, KCI, KBr and KF, and for tetra-alkylammonium bromides the I/, values were calculated from their molecular weights and crystal 21 and 4, taken from ref. (22). The rnel values were calculated from (Vne, = &rr;,J. The T;, were taken from the individual ion radii tabulated in ref.(23) for alkali halides and from those in ref. (24) for tetra-alkylammonium bromides. The /3 values were calculated using several methods ; for LiC1, NaCl and KC1 salts from ref. (25) and for KBr from ref. (26). The adiabatic compressibilities, p,, were calculated for KF from parameters tabulated in ref. (27). These values were taken instead of isothermal compressibilities, b, since they are quite similar for 1 : 1 electrolytes. However, this approximation cannot be made with tetra- alkylammonium bromides, for which reason the isothermal compressibilities were calculated from /?, and 4, found in ref. ( I 5). In order to carry out this computation the apparent expansibilities, &, and the apparent heat capacities, q5cp are needed.The bE values were calculated from ref. (16) and (27), and 4cp from ref. (27). We have calculated logy,,, from eqn (2) at various concentrations of salt for the electrolytes indicated above and we have then obtained the theoretical values of k, from the plot of log ynel us. c,. The results obtained are compiled in table 2, where experimental k, is transformed from the molar scale to the mole fraction scale. The agreement between k, from the Cross equation and experimental k, is acceptable for alkaline halides with caffeine and theophylline. However, for R,NBr the salting-in is predicted only qualitatively, excepting the Me,NBr-theophylline pair. Scaled particle theory (SPT) has been used by some authors to foretell salting effects of non-polar so1utes.2s Nevertheless, their application to polar solutes has been more restricted, with varying degrees of 29-32Table 4.Comparison of k, and k, terms from SPT theory with experimental k, LiCl NaCl KCI KBr KF Me,NBr Et,NBr Pr,NBr Bu,NBr 0.51 1 0.327 0.233 - 0.658 0.421 .0.270 - caffeine 0.618 0.601 0.749 -0.230 -0.527 -1.07 - 1.65 0.270 -0.296 -0.360 0.069 0.423 0.836 1.24 0.422 0.329 0.730 -0.100 -0.539 -1.06 - 1.78 theophylline CD 0.563 0.548 0.682 -0.207 -0.472 -0.968 - 1.49 rr p k , 0.465 0.599 -0.206 -0.243 -0.244 -0.269 -0.322 0.055 0.369 0.737 1.10 0.3 13 0.323 0.3 16 0.269 0.562 -0.014 -0.435 -0.928 -1.63 ka k , - k, ~ _ _ _ _ _ _ _ .~1038 Setschenow Coeficients electrolyte solutions. From these, k, may be expressed in condensed forrn by: Masterton and Lee33 developed equations specifically for calculating k , in aqueous 1 : 1 where k, is the cavity term, kp the interaction term and k , transforms k, from the molar scale to the mole fraction scale.In order to find k, for a given salt-non-electrolyte pair it is necessary to know the diameters, 0, and polarizabilities, a, of the solute, cation, anion and water, the apparent molal volume of the salt at infinite dilution, #O, and the energy parameter of the non-electrolyte, Elk. For the three methylxanthines Elk and o values are not available, which is particularly important in the situation under consideration. Some have determined the Elk parameter for non-electrolytes from the Mavroyannis-Stephen equation35 and reasonable fits were obtained with the experim- ental values.In the present work we have employed this equation to calculated Elk for caffeine and theophylline. All parameters are shown in table 3. The k, values calculated from SPT are given in table 2. From this it can be seen that the agreement with experimental k, is better for tetra-alkylammonium bromides than for alkali halides. .It is interesting to compare the relative contributions of the k, and kp terms to the theoretical k,. This comparison is made in table 4. From it we observe that the k, terms are positive for alkali halides and negative for R4NBr, while opposite signs are obtained for k terms. The sign of k, will, of course, depend upon the relative magnitudes of k, and /,+ In our case, as can be seen from tables 3 and 4, the cavity terms (k,) dominate the interaction terms (kp).In table 4, we have also enclosed the difference, k, (exp) - kp, and the agreement between this and k, is remarkable for caffeine and theophylline with KF and R,NBr. Therefore, the k, term from SPT could explain our results quantitatively for those salts that increased the water structure, but this is not the case with the other salts. Therefore, it can be concluded that the experimental results cannot be completely explained by either the SPT or the Cross theories. It seems evident that much more theoretical work must be done in order to be able to predict salting effects on polyfunctional solutes. References 1 D. T. Hurst, An Introduction to the Chemistry and Biochemistry of Pyrimidines, Purines and Pteridines 2 S. J. Gill, M. Downig and G.F. Sheats, Biochemistry, 1967, 6, 272. 3 A. Cesaro, E. Russo and V. Crescenzi, J. Phys. Chem., 1976,80, 335. 4 J. H. Stem and L. R. Beeniga, J. Phys. Chem., 1975,79, 582. 5 J. H. Stern, J. A. Devore, S. L. Hansen and 0. Yawz, J. Phys. Chem., 1974,78, 1922. 6 J. De Taege and Th. Zeegers-HugsKens, J . Pharm. Sci., 1985,74,660. 7 A. Cesaro, E. Russo and D. Tessarotto, J. Solution Chem., 1980, 9, 221. 8 G. Perron, D. Joly, J. E. Desnoyers, L. Avedikian and J. P. Morel, Can. J. Chem., 1978, 56, 522. 9 S. H. Yalokowsky, S. C. Valvani and T. J. Rooseman, J. Pharm. Sci., 1983, 72, 866. 10 C. Treiner, Can. J . Chem., 1981,59, 2518. 11 Ya. 1. Korenman and R. N. Bortnikova, Russ. J. Phys. Chem., 1979, 53, 545. 12 B. E. Conway, Pure Appl. Chem., 1980, 57, 263. 13 B.E. Conway, Ionic Hydration in Chemisfry and Biophysics (Elsevier, Amsterdam, 1981), chap. 23. 14 J. E. Desnoyers, M. Arel, G. Perron and C. Jolicoeur, J. Phys. Chem., 1969, 73, 3346. 15 B. E. Conway and R. E. Verrall, J. Chem. Phys., 1966, 70, 3952. 16 J. G. Mathieson and B. E. Conway, J. Chem. SOC., Faraday Trans I , 1974, 70, 752. 17 R. F. Cross, J. Phys. Chem., 1975,79, 1822. 18 R. F. Cross and P. McTigue, J. Phys. Chem., 1976, 80, 814. 19 A. Martin, A. N. Paruta and A, Adjei, J. Pharm. Sci., 1981, 70, 1 15. 20 Handbook of Chemistry and Physics, ed. R. C . West (C.R.C. Press, Boca Raton, Florida, 1982/83), 21 R. McMullan, G. A. Jeffrey, J. Chem. Phys., 1959,31, 1231. 22 F. J. Millero, Water and Aqueous Solutions: Structure, Thermodynamics, and Transport Processes, ed. (John Wiley, New York, 1980), chap. 3. pp. E-377. E. A. Home (Wiley, New York, 1972), pp. 565-592.P. Perez-Tejeda et al. 1039 23 B. E. Conway, J. E. Desnoyers and A. C. Smith, Philos. Trans. R. SOC. London, Ser. A, 1964,256,389. 24 P. S. Ramanathan, C. V. Krishnan and H. L. Friedman, J. Solution Chem., 1972, 1,237. 25 H. S. Harned and B. B. Owen, The Physical Chemistry of Electrolytic Solutions (Reinhold, New York, 26 R. E. Gibson, J. Am. Chem. SOC., 1935,57,284. 27 F. J. Millero, Activity Coefficients in Electrolyte Solution, ed. R. M. Pytkowicz (CRC Press, Florida, 28 R. A. Pierotti, Chem. Rev., 1976, 76, 717. 29 M. Aveyard and R. Heselder, J. Chem. SOC., Faraday Trans. 1, 1975, 71, 3121. 30 C. Treiner and M. Fromon, J. Chem. Soc., Faraday Trans. I , 1980, 76, 1062. 31 M. Balon, A. Maestre, E. M. de Miguel and E. S. del Hoyo, Can. J. Chem., 1983,61, 244. 32 M. H. Sagert and D. W. P. Lau, Can. J. Chem., 1985,62, 81. 33 W. L. Masterton and T. P. Lee, J. Phys. Chem., 1970, 74, 1776. 34 Yuh-Loo Tien Cheng, M. Y. Schrier and E. Schrier, J. Phys. Chem., 1974,78, 165. 35 C. Mavroyannis and M. J. Stephen, Mol. Phys., 1962,5, 629. 36 R. C. Reid and T. K. Sherwood, The Properties of Gases and Liquids, their Estimation and Correlation 37 N. Desrosiers and J. E. Desnoyers, Can. J. Chem., 1976, 54, 3800. 38 B. E. Conway, D. M. Novak and L. Laliberte, J. Solution Chem., 1974, 3, 683. 1958), pp. 267-271. 1979), VO~. 11, pp. 63-94. (Utema, Mexico, 1968), p. 423. Paper 61642; Received 1st April, 1986 35-2

ISSN:0300-9599    

DOI:10.1039/F19878301029

出版商:RSC    

年代:1987

数据来源: RSC