摘要:

physicochemical topics, thereby encouraging scientists of different disciplines to contribute their varied viewpoints to a coiiinion theme. A recent Discussion is :- The Royal Soci- of Chemistry- No.75 lntraamolecwlar Kinetics No. 75 in the series, this publication is the result of a general discussion held at the University of Warwick in April 1983. Contents: The Spiers Meniorlal Lecture; Vibrational Redistribution within Excited Electronic States of Polyatomic Molecules Inrraniolecular R e h u t i o n o f 1. vcited States lsomerization of Intcrnal~ncrgy-selected Ions Kinetics of Ion-Molecule Collision Coinple\es in the Gas Phase, E\periinent and Theory lntrainolccular Decay 01' Soinc Open-shell Pulya t o niic Ca lions On tlic Theory u i Iiitrdniolccul~r I n e r g y Transfer Pulsed Laser Preparation and Ouaniuin Superposition Statc Evolution in ReguLtr and Irregular Systems A Ouantuiii-iiicclianical Internal-collision Model for State-sclcctcd Uniinolccular Decoiiiposilio n The Correspondence Principle and Intramolecular Dynamics lntrainoleculdr Dcphasiiig.t'icusecond Evolution of Wavepacket States in a Molecule with Int erinediate-casc level Struct urc Energy Conversion in van der Waals C'u~~iplc\c\ ol s-Tetrarine and Argon Tim-dependent Processes in Polyatuinic Molecules During and After Intense Intrarcd Irradiation Energy Distributions in tlic (.N(X'L+) bragnient froiii tlie Infrared Multiplepholun Dissociation ol' CI. ICN. A Coinparison between 1:xperiiiiental Results and the Predictions ot Statistical Theories of ChFO + Product Energy Partitioning in the Decoiii- position of State-selectively Excited HOON and IIOOD Low-power Inl-rarcd Laser I'hoiolysis o f Tetramethy ldioxetan Uniinolecular Reactions lnduccd by Vibrational Overtone Excitation Uniiiiolecular Decomposition of t-Butylhydro- peroxide by Direct Excitation of the 6-0 0-11 Stretching Overtone I'icosecond-jet Spectroscopy and Photoclieinistry.Energy Redistribution and its Iiiipact'on Coherence, Isoincrization, Ihssociatiun and Solvalioii knergy Redistribution in Large Molecules. Duect St ud y o f In1 rainolucular Rehxa lion in the Gas Phase with Picosecond Gating Rotation-dependent Intrainolecuhr I'r~)cessc.sofSO:(A'A.) in a Superwnic Jct Role of Rotation-Vibration Interaction in Vibrational Keh\ation. Energy Kcdistribution in k,xcitcd Singlet I~'ornialdc1iyde Sub-lhppler.Spectroscopy of Benrcnc in the "('liaiinel-lliree" Region Intraiiiulccular 1:lectronic Kclau~tion and I'liotois~)iiieruati[)n Processes in tlie lsuhted Azabenrene Molecules Pyridinc, Pyrazinc and I'yriiiiidinc Softcover 434pp 0 85186 658 1 Price f25.00 ($48.00) Rest of the World f26.00 RSC Members f 16.25 Faraday Discussions of the Chemical Society 7< lnrruniolei u h r Kincrit I Faraday Symposia are usually held annually and are confined to more specialiscd topics than Discussions, with particular reference to recent rapidly developing lines of rescuch. A recent Symposium is :- No.l?The Hydrophobic Interadion No. 17 in the series, this publication is the result of a symposium on The Hydrophobic Interaction held at the Uiiiversity of Reading in December 1982.Contents: Hydrophobic Interdctionr a llistaric.11 Per spect ivr llydrupliobic Ilydration Geometric Kelaution in Water. Its Role in Precise Vapour-pressure Measureiiients of the SolubilkdtiorI of Benzene by Aqueous Sodiuiii Octylsulphate Solutions Nuclear Magnetic Resonance R e b u t i o n Investigation of Tetrahydrofuran and Methyl Iodide Clathrdtes Infrared and Nuckar Magnetic Kcwnance Studies Pertaining to the (age Model t o r Solutions oS Acetone in Water Irothernial Transport Properties in Solutions o f S y mmet r ica I Tet ra-alk y hmnioniuiii Bromides Thermodynamics of Cavity I'oriiiaiion in Water. A Molecular Dynamics Study Molecular Librations and Solvent Oricnt- ational Correlations in Hydrophobic Phenomena Monte Carlo Computer Siniulation Study of the Hydrophobic Effect.Potential ot Mean Force for ECfir)gaq at 25 and SOv C Hydroplicibic Moments and Protein Structure Application 01' the Kirkwood-Buff Theory to the t'roblcin 01 Hydrophobic Interactions Ihentangleinent of Ilydrophubic and IFlcctrosi~tic Contributions t o the I.ilni Pressures O i Ionic Surfactants llydrophobir. Intcracliuns in Dilute Su lut io ns u t 1'0 1 y (vin y I a Ico lio I) ('onioriii;tiionaI and 1:unc.i ional I'ropertics of tiaeiiwglobin in Water+Alcohol Mixtures. Dependence o f Bull. Electrostatic and tlydrupliohic I n t c r x t i o n s upon ptl and KCI concentrations Softcover 24Opp 0 85186 668 9 Price f36.50 ($70.00) Rest of the World f38.50 RSC Members f 23.75 ORDERING RSC Members should send their orders to: The Royal Society of Chemistry.The Membership Officer. 30 Russell Square, Non-RSC Members The Royal Society of Chemistry, Distribution Centre, Blackhorse Road, L London WC1 B 5DT. Letchworth, Herts SO6 IHN, England. Faradaj Symposia of the Chemical Society hGi 17 I hc HI drophohr' Inrcrm rron 1 9 X ? (viii)physicochemical topics, thereby encouraging scientists of different disciplines to contribute their varied viewpoints to a coiiinion theme. A recent Discussion is :- The Royal Soci- of Chemistry- No.75 lntraamolecwlar Kinetics No. 75 in the series, this publication is the result of a general discussion held at the University of Warwick in April 1983. Contents: The Spiers Meniorlal Lecture; Vibrational Redistribution within Excited Electronic States of Polyatomic Molecules Inrraniolecular R e h u t i o n o f 1.vcited States lsomerization of Intcrnal~ncrgy-selected Ions Kinetics of Ion-Molecule Collision Coinple\es in the Gas Phase, E\periinent and Theory lntrainolccular Decay 01' Soinc Open-shell Pulya t o niic Ca lions On tlic Theory u i Iiitrdniolccul~r I n e r g y Transfer Pulsed Laser Preparation and Ouaniuin Superposition Statc Evolution in ReguLtr and Irregular Systems A Ouantuiii-iiicclianical Internal-collision Model for State-sclcctcd Uniinolccular Decoiiiposilio n The Correspondence Principle and Intramolecular Dynamics lntrainoleculdr Dcphasiiig. t'icusecond Evolution of Wavepacket States in a Molecule with Int erinediate-casc level Struct urc Energy Conversion in van der Waals C'u~~iplc\c\ ol s-Tetrarine and Argon Tim-dependent Processes in Polyatuinic Molecules During and After Intense Intrarcd Irradiation Energy Distributions in tlic (.N(X'L+) bragnient froiii tlie Infrared Multiplepholun Dissociation ol' CI.ICN. A Coinparison between 1:xperiiiiental Results and the Predictions ot Statistical Theories of ChFO + Product Energy Partitioning in the Decoiii- position of State-selectively Excited HOON and IIOOD Low-power Inl-rarcd Laser I'hoiolysis o f Tetramethy ldioxetan Uniinolecular Reactions lnduccd by Vibrational Overtone Excitation Uniiiiolecular Decomposition of t-Butylhydro- peroxide by Direct Excitation of the 6-0 0-11 Stretching Overtone I'icosecond-jet Spectroscopy and Photoclieinistry. Energy Redistribution and its Iiiipact'on Coherence, Isoincrization, Ihssociatiun and Solvalioii knergy Redistribution in Large Molecules.Duect St ud y o f In1 rainolucular Rehxa lion in the Gas Phase with Picosecond Gating Rotation-dependent Intrainolecuhr I'r~)cessc.sofSO:(A'A.) in a Superwnic Jct Role of Rotation-Vibration Interaction in Vibrational Keh\ation. Energy Kcdistribution in k,xcitcd Singlet I~'ornialdc1iyde Sub-lhppler. Spectroscopy of Benrcnc in the "('liaiinel-lliree" Region Intraiiiulccular 1:lectronic Kclau~tion and I'liotois~)iiieruati[)n Processes in tlie lsuhted Azabenrene Molecules Pyridinc, Pyrazinc and I'yriiiiidinc Softcover 434pp 0 85186 658 1 Price f25.00 ($48.00) Rest of the World f26.00 RSC Members f 16.25 Faraday Discussions of the Chemical Society 7< lnrruniolei u h r Kincrit I Faraday Symposia are usually held annually and are confined to more specialiscd topics than Discussions, with particular reference to recent rapidly developing lines of rescuch.A recent Symposium is :- No.l?The Hydrophobic Interadion No. 17 in the series, this publication is the result of a symposium on The Hydrophobic Interaction held at the Uiiiversity of Reading in December 1982. Contents: Hydrophobic Interdctionr a llistaric.11 Per spect ivr llydrupliobic Ilydration Geometric Kelaution in Water. Its Role in Precise Vapour-pressure Measureiiients of the SolubilkdtiorI of Benzene by Aqueous Sodiuiii Octylsulphate Solutions Nuclear Magnetic Resonance R e b u t i o n Investigation of Tetrahydrofuran and Methyl Iodide Clathrdtes Infrared and Nuckar Magnetic Kcwnance Studies Pertaining to the (age Model t o r Solutions oS Acetone in Water Irothernial Transport Properties in Solutions o f S y mmet r ica I Tet ra-alk y hmnioniuiii Bromides Thermodynamics of Cavity I'oriiiaiion in Water.A Molecular Dynamics Study Molecular Librations and Solvent Oricnt- ational Correlations in Hydrophobic Phenomena Monte Carlo Computer Siniulation Study of the Hydrophobic Effect. Potential ot Mean Force for ECfir)gaq at 25 and SOv C Hydroplicibic Moments and Protein Structure Application 01' the Kirkwood-Buff Theory to the t'roblcin 01 Hydrophobic Interactions Ihentangleinent of Ilydrophubic and IFlcctrosi~tic Contributions t o the I.ilni Pressures O i Ionic Surfactants llydrophobir. Intcracliuns in Dilute Su lut io ns u t 1'0 1 y (vin y I a Ico lio I) ('onioriii;tiionaI and 1:unc.i ional I'ropertics of tiaeiiwglobin in Water+Alcohol Mixtures. Dependence o f Bull. Electrostatic and tlydrupliohic I n t c r x t i o n s upon ptl and KCI concentrations Softcover 24Opp 0 85186 668 9 Price f36.50 ($70.00) Rest of the World f38.50 RSC Members f 23.75 ORDERING RSC Members should send their orders to: The Royal Society of Chemistry. The Membership Officer. 30 Russell Square, Non-RSC Members The Royal Society of Chemistry, Distribution Centre, Blackhorse Road, L London WC1 B 5DT. Letchworth, Herts SO6 IHN, England. Faradaj Symposia of the Chemical Society hGi 17 I hc HI drophohr' Inrcrm rron 1 9 X ? (viii)

ISSN:0300-9599    

DOI:10.1039/F198480FX017

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

AUTHOR INDEX xxxv Sabbatini, L., 1029 Sacco, A., 2669 Sanders, J. V., 571 Sangster, D. F., 291 1 Sarka, K., 521 Sasahira, A., 473 Sasse, W. H. F., 571 Satchell, P. W., 2395 Sato, K., 841 Sato, Y., 341 Savino, V., 759 Sayers, C. M., 1217 Schiller, R. L., 1257 Schmidt, J., 1 Schmidt, P. P., 2017 Schneider, H., 3275, 3285 Schulz, R. A., 489, 1323 Scott, J. M. W., 739, 1651, 2287, Scott, S. K., 3409 Segall, R. L., 2609 Sehested, K., 2929, 2969 Seidl, V., 1367 Sem, P., 297 Serratosa, J. M., 2225 Seyama, H., 237 Seyedmonir, S. R., 87, 2269 Shanahan, M. E. R., 37 Sheppard, A., 2999 Sherwood, P. M. A., 135, 2099, Shindo, Y., 879, 2199 Shiotani, H., 2145 Shizuka, H., 383, 341 Siekhaus, W. J., 61 Sircar, S., 1101, 2489 Smart, R. St C., 2957, 2609 Smith, I. M., 3021 Smith, R., 3233 Snow, R.L., 3463 Solar, S., 2929 Solar, W., 2929 Solymosi, F., 1841 Soma, M., 237 Soupart, J-B., 3209 Sourisseau, C., 3257 Spink, J. A., 3469 Spoto, G., 1875, 1891 Spotswood, T. M., 3147 Staricco, E. H., 2631 Stassinopoulou, K., 3095 Stedman, D. H., 285 Stout, D. R., 3481 Strohbusch, F., 1757 Strumolo, D., 1479 Struve, P., 813, 2167 Styring, M. G., 3051 Subramanian, R., 2405 2881, 3359 2549, 2867 Sundar, H. G. K., 3491 Sutcliffe, L. H., 669, 3021 Sutton, H. C., 2301 Sutton, L. E., 635 Suzuki, H., 803 Suzuki, T., 1925, 3157 Symons, M. C. R., 423, 1005, Szamosi, J., 1645 Szczepaniak, W., 2935 Takagi, Y., 1925 Takahashi, K., 803 Takahashi, N., 629 Takanaka, J., 941 Takao, S., 993 Takasaki, S., 803 Takegami, H., 1221 Tam, S-C., 2255 Tamamushi, R., 2751 Tamaru, K., 29, 1567, 1595 Tamilarasan, R., 2405 Tanabe, S., 803 Tanaka, K., 2563,2981 Tanaka, T., 119 Taniewska-Osinska, S., 1409 Tascon, J.M. D., 1089 Teo, H. H., 981, 1787 Tetenyi, P., 3037 Thomas, J. K., 1163 Thompson, L., 1673 Thomson, M., 1867 Thomson, S. J., 1689 Tiddy, G. J. T., 789, 3339 Tittarelli, P., 2209 Tominaga, T., 941 Tomkinson, J., 225 Tonelli, C., 1605 Toprakcioglu, C., 13,413 Tran, T., 1867 Trasatti, S., 913 Tripathi, A. D., 1517 Tronc, E., 2619 Troncoso, G., 2127 Truscott, T. G., 2293 Tsurusaki, T., 879 Tuck, J. J., 309 Turner, P. S., 2609 Tusk, M., 1757 Tvarbikova, Z., 2639 Tyrrell, H. J. V., 1279 Ueki, Y.. 341 Ueno, A., 803 Unno, H., 1059 Valencia, E., 2127 van de Ven, T. G. M., 2677 van Ommen, J. G., 2479 van Truong, N., 3275, 3285 Vargas, I., 1947 2767, 2803, 21 1, 1999 Vedrine, J.C., 1017 Veith, J., 2313 Velasco, J. R., 3429 Vesala, A., 2439 Vickerman, J. C., 1903 Vincent, B., 2599 Vinek, H., 1239 Vink, H., 507, 1297 Waghorne, W. E., 1267 Wagley, D. P., 47 Walker, R. W., 435, 3187, 3195, Wallington, T. J., 2737 Wang, G-W., 1039 Watkins, P. E., 2323 Watkiss, P. J., 1279 Watt, R. A. C., 489 Webb, G., 1689 Webster, B. C., 255, 267 Weiner, E. R., 1491 Wells, C. F., 2155. 2445 Wells, J. D., 1233 Whang, B. C. Y., 291 1 Whittle, E., 2323 Wichterlova, B., 2639 Wiesner, S., 3021 Wilhelmy, D. M., 563 Williams E. H., 3147 Williams, P. A., 403 Williams, R. J. P., 2255 Wokaun, A., 1305 Wolff, T., 2969 Wood, S. W., 3419 Woolf, L. A., 549, 1287 Wright, C. J., 1217 Wu, D. C., 1795 Wiirflinger, A., 3221 Wyn-Jones, E., 1915 Yamabe, M., 1059 Yamamoto, S., 941 Yamashita, H., 1435 Yamauchi, H., 2033 Yamazaki, A., 3245 Yariv, S., 1705 Yasumori, I., 841 Yeates, S.G., 1787 Yide, X., 969, 3103 Ylikoski, J., 2439 Yokokawa, T., 473 Yoneda, N., 879 Yonezawa, T., 1435 Yoshida, S., 119, 1435 Zambonin, P. G., 1029 Zanderighi, L., 1605 Zecchina. A., 2209, 2723, 1875, Zipelli, C., 1777 Zundel, G., 553 348 1, 2827 1891AUTHOR INDEX xxxv Sabbatini, L., 1029 Sacco, A., 2669 Sanders, J. V., 571 Sangster, D. F., 291 1 Sarka, K., 521 Sasahira, A., 473 Sasse, W. H. F., 571 Satchell, P. W., 2395 Sato, K., 841 Sato, Y., 341 Savino, V., 759 Sayers, C. M., 1217 Schiller, R. L., 1257 Schmidt, J., 1 Schmidt, P. P., 2017 Schneider, H., 3275, 3285 Schulz, R. A., 489, 1323 Scott, J. M. W., 739, 1651, 2287, Scott, S.K., 3409 Segall, R. L., 2609 Sehested, K., 2929, 2969 Seidl, V., 1367 Sem, P., 297 Serratosa, J. M., 2225 Seyama, H., 237 Seyedmonir, S. R., 87, 2269 Shanahan, M. E. R., 37 Sheppard, A., 2999 Sherwood, P. M. A., 135, 2099, Shindo, Y., 879, 2199 Shiotani, H., 2145 Shizuka, H., 383, 341 Siekhaus, W. J., 61 Sircar, S., 1101, 2489 Smart, R. St C., 2957, 2609 Smith, I. M., 3021 Smith, R., 3233 Snow, R. L., 3463 Solar, S., 2929 Solar, W., 2929 Solymosi, F., 1841 Soma, M., 237 Soupart, J-B., 3209 Sourisseau, C., 3257 Spink, J. A., 3469 Spoto, G., 1875, 1891 Spotswood, T. M., 3147 Staricco, E. H., 2631 Stassinopoulou, K., 3095 Stedman, D. H., 285 Stout, D. R., 3481 Strohbusch, F., 1757 Strumolo, D., 1479 Struve, P., 813, 2167 Styring, M. G., 3051 Subramanian, R., 2405 2881, 3359 2549, 2867 Sundar, H.G. K., 3491 Sutcliffe, L. H., 669, 3021 Sutton, H. C., 2301 Sutton, L. E., 635 Suzuki, H., 803 Suzuki, T., 1925, 3157 Symons, M. C. R., 423, 1005, Szamosi, J., 1645 Szczepaniak, W., 2935 Takagi, Y., 1925 Takahashi, K., 803 Takahashi, N., 629 Takanaka, J., 941 Takao, S., 993 Takasaki, S., 803 Takegami, H., 1221 Tam, S-C., 2255 Tamamushi, R., 2751 Tamaru, K., 29, 1567, 1595 Tamilarasan, R., 2405 Tanabe, S., 803 Tanaka, K., 2563,2981 Tanaka, T., 119 Taniewska-Osinska, S., 1409 Tascon, J. M. D., 1089 Teo, H. H., 981, 1787 Tetenyi, P., 3037 Thomas, J. K., 1163 Thompson, L., 1673 Thomson, M., 1867 Thomson, S. J., 1689 Tiddy, G. J. T., 789, 3339 Tittarelli, P., 2209 Tominaga, T., 941 Tomkinson, J., 225 Tonelli, C., 1605 Toprakcioglu, C., 13,413 Tran, T., 1867 Trasatti, S., 913 Tripathi, A.D., 1517 Tronc, E., 2619 Troncoso, G., 2127 Truscott, T. G., 2293 Tsurusaki, T., 879 Tuck, J. J., 309 Turner, P. S., 2609 Tusk, M., 1757 Tvarbikova, Z., 2639 Tyrrell, H. J. V., 1279 Ueki, Y.. 341 Ueno, A., 803 Unno, H., 1059 Valencia, E., 2127 van de Ven, T. G. M., 2677 van Ommen, J. G., 2479 van Truong, N., 3275, 3285 Vargas, I., 1947 2767, 2803, 21 1, 1999 Vedrine, J. C., 1017 Veith, J., 2313 Velasco, J. R., 3429 Vesala, A., 2439 Vickerman, J. C., 1903 Vincent, B., 2599 Vinek, H., 1239 Vink, H., 507, 1297 Waghorne, W. E., 1267 Wagley, D. P., 47 Walker, R. W., 435, 3187, 3195, Wallington, T. J., 2737 Wang, G-W., 1039 Watkins, P. E., 2323 Watkiss, P. J., 1279 Watt, R. A. C., 489 Webb, G., 1689 Webster, B. C., 255, 267 Weiner, E. R., 1491 Wells, C. F., 2155. 2445 Wells, J. D., 1233 Whang, B. C. Y., 291 1 Whittle, E., 2323 Wichterlova, B., 2639 Wiesner, S., 3021 Wilhelmy, D. M., 563 Williams E. H., 3147 Williams, P. A., 403 Williams, R. J. P., 2255 Wokaun, A., 1305 Wolff, T., 2969 Wood, S. W., 3419 Woolf, L. A., 549, 1287 Wright, C. J., 1217 Wu, D. C., 1795 Wiirflinger, A., 3221 Wyn-Jones, E., 1915 Yamabe, M., 1059 Yamamoto, S., 941 Yamashita, H., 1435 Yamauchi, H., 2033 Yamazaki, A., 3245 Yariv, S., 1705 Yasumori, I., 841 Yeates, S. G., 1787 Yide, X., 969, 3103 Ylikoski, J., 2439 Yokokawa, T., 473 Yoneda, N., 879 Yonezawa, T., 1435 Yoshida, S., 119, 1435 Zambonin, P. G., 1029 Zanderighi, L., 1605 Zecchina. A., 2209, 2723, 1875, Zipelli, C., 1777 Zundel, G., 553 348 1, 2827 1891

ISSN:0300-9599    

DOI:10.1039/F198480BX019

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

JOURNAL OF T H E CHEMICAL SOCIETY FARADAY TRANSACTIONS, PARTS I AND I 1 The Journal of the Chemical Society is published in six sections, of which five are termed Transactions; these are distinguished by their subject matter, as follows: Dalton Transactions (Inorganic Chemistry). All aspects of the chemistry of inorganic and organometallic compounds; including bioinorganic chemistry and solid-state inorganic chemistry; of their structures, properties, and reactions, including kinetics and mechanisms ; new or improved experimental techniques and syntheses. Faraday Transactions I (Physical Chemistry). Radiation chemistry, gas-phase kinetics, electrochemistry (other than preparative), surface and interfacial chemistry, heterogeneous catalysis, physical properties of polymers and their solutions, and kinetics of polymerization, etc.Faraday Transactions II (Chemical Physics). Theoretical chemistry, especially valence and quantum theory, statistical mechanics, intermolecular forces, relaxation phenomena, spectroscopic studies (including i.r., e.s.r., n.m.r., and kinetic spec- troscopy, etc.) leading to assignments of quantum states, and fundamental theory. Studies of impurities in solid systems. Perkin Transactions I (Organic Chemistry). All aspects of synthetic and natural product organic, organometallic and bio-organic chemistry, including aliphatic, alicyclic, and aromatic systems (carbocyclic and heterocyclic). Perkin Transactions II (Physical Organic Chemistry). Kinetic and mechanistic studies of organic, organometallic and bio-organic reactions.The description and application of physicochemical, spectroscopic, and theoretical procedures to organic chemistry, including structure-activity relationships. Physical aspects of bio-organic chemistry and of organic compounds, including polymers and biopolymers. Authors are requested to indicate, at the time they submit a typescript, the journal for which it is intended. Should this seem unsuitable, the Editor will inform the author. The sixth section of the Journal of the Chemical Society is Chemical Communications, which is intended as a forum for preliminary accounts of original and significant work, in any area of chemistry that is likely to prove of wide general appeal or exceptional specialist interest. Such preliminary reports should be followed up eventually by full papers in other journals (e.g.the five Transactions) providing detailed accounts of the work. NOTES I t has always been the policy of the Faraday Transactions that brevity should not be a factor influencing acceptability for publication. In addition however to full papers both sections carry at the end of each issue a section headed ‘Notes’, which are short self-contained accounts of experimental observations, results, or theory that will not require enlargement into ‘full’ papers. The Notes section is not used for preliminary communications. The layout of a Note is the same as that of a paper. Short summaries are required. The procedure for submission, administration, refereeing, editing and publication of Notes is the same as for full papers.However, Notes are published more quickly than papers since their brevity facilitates processing at all stages. The Editors endeavour to meet authors’ wishes as to whether an article is a full paper or a Note, but since there is no sharp dividing line between the one and the other, either in terms of length or character of content, the right is retained to transfer overlong Notes to the full papers section. As a guide a Note should not exceed 1500 words or word-equivalents. (i)NOMENCLATURE AND SYMBOLISM Units and Symbols. The Symbols Committee of The Royal Society, of which The Royal Society of Chemistry is a participating member, has produced a set of recommendations in a pamphlet ‘Quantities, Units, and Symbols’ (1975) (copies of this pamphlet and further details can be obtained from the Manager, Journals, The Royal Society of Chemistry, Burlington House, London W 1 V OBN).These recommendations are applied by The Royal Society of Chemistry in all its publications. Their basis is the ‘Systeme International d’Unites’ (SI). A more detailed treatment of units and symbols with specific application to chemistry is given in the IUPAC Manual of Symbols and Terminology for Physicochemical Quantities and Units (Pergamon, Oxford, 1979). Nomenclature. For many years the Society has actively encouraged the use of standard IUPAC nomenclature and symbolism in its publications as an aid to the accurate and unambiguous communication of chemical information between authors and readers. In order to encourage authors to use IUPAC nomenclature rules when drafting papers, attention is drawn to the following publications in which both the rules themselves and guidance on their use are given: Nomenclature of Organic Chemistry, Sections A , B, C, D, E, F, and H (Pergamon, Oxford, 1979 edn).Nomenclature of Inorganic Chemistry (Butterworths, London, 197 1, now published by Pergamon). Biochemical Nomenclature and Related Documents (The Biochemical Society, London, 1978). A complete listing of all IUPAC nomenclature publications appears in the January issues of J. Chern. Soc., Faraday Transactions. It is recommended that where there are no IUPAC rules for the naming of particular compounds or authors find difficulty in applying the existing rules, they should seek the advice of the Society’s editorial staff.(ii)THE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY GENERAL DISCUSSION NO. 78 Radicals in Condensed Phases University of Leicester, 4-6 September 1984 Organising Committee Professor M. C. R. Symons (Chairman) Dr G. 6. Buxton Dr T. A. Claxton Dr K. A. McLauchlan Professor Lord Tedder Dr R. L. Willson The discussion will be primarily concerned with the structure and reactions of radicals in liquids and solids. It is designed to bring together theoretical work on structure, environmental effects and reactivity with spectroscopic and mechanistic studies directly concerned with radicals. Fundamental aspects will be stressed, and particular attention will be given to new developments including measurement at short time intervals, special solvent effects, and the effects of external fields.A special area for inclusion will be electron gain and loss processes including trapped and solvated electrons, electrochemical reactions, and specific electron capture and electron loss in low-temperature systems. Photochemical charge-transfer processes will also be included. The final programme and application form may be obtained from Mrs Y. A. Fish, The Royal Society of Chemistry, Burlington House, London W1V OBN THE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY D E UTSCH E B U N S EN G ES E LLSC HAFT FU R PHYS I KALl SC H E CH EM I E SOClETE DE CHlMlE PHYSIQUE ASSOCIAZIONE ITALIANA D I CHIMICA FlSlCA Joint Discussion Meeting on: Laser Studies in Reaction Kinetics Evangelische Akademie, Tutzing, West Germany, 24-27 September 1984 Organising Committee R.Ben Aim (Gif sur Yvette) G. Giacometti (Padova) P. Rigny (Gif sur Yvette) E. W. Schlag (Munchen) I. W. M. Smith (Cambridge) J. Troe (Gottingen) K. Welge (Bielefeld) The aim of this meeting is the discussion of the latest experiments and related theories in the field of laser studies of elementary chemical reactions in molecular beams, in the gas phase, and in the condensed phase. The discussion will include oral contributions and poster presentations. Contributions are invited and a very brief preliminary abstract (less than 50 words) should be submitted by 15 May 1984 to: Professor Dr J. Troe, lnstitut fur Physikalische Chemie, Universitat Gottingen, Tammannstrasse 6, 03400 Gottingen, West Germany. Authors of accepted contributions will be required to provide a manuscript for publication in a special issue of the Berichte der Bunsengesellschaft fur Physikalische Chemie.The Faraday Division has a small fund to assist members with the expenses of attending this conference. Applications for a grant should be submitted to Mrs Y. A. Fish, The Royal Society of Chemistry, Burlington House, London WIV OBN, by 31 July 1984. (iii)THE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY SYMPOSIUM NO. 19 Molecular Electronic Structure Calcu lat ions- Met hods and Applications University of Cambridge, 12-1 3 December 1984 N.B. Please note change of date Molecular electronic structure calculations have now developed into a powerful predictive tool and are necessary in several different fields to aid the understanding and interpretation of experimental observations.The meeting will review the current state of this rapidly developing discipline and will bring together experts on some of the most advanced methods and their applications. The meeting will provide an opportunity for discussion and comparison of the various techniques currently in use. It will therefore not only be a valuable forum for discussion among research workers in the field, but should also show the non-specialist what theoretical calculations can be expected to achieve now and in the near future. 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 GENERAL DISCUSSION NO.79 Polymer Liquid Crystals University of Cambridge, 1-3 April 1985 The object of the meeting will be to discuss all aspects of the developing subject of polymeric liquid crystals. The hope is to bring together scientists from the fields of conventional polymer science and monomeric liquid crystals who are active in this field. The discussion is aimed at understanding the following facets: (a) The chemical characteristics that give rise to polymer liquid crystalline behaviour. (b) The nature of the high local anisotropy of these systems and their structural organisation at the molecular, micron and macroscopic levels. (c) The physical properties and their industrial exploitation, with particular reference to the influence of external force fields such as flow, electric and magnetic fields.(d) The inter-relations of polymer liquid crystals with small-molecule mesophases, conventional flexible polymers and biopolymers which exhibit liquid-crystalline behaviour. Contributions are invited for consideration by the Organising Committee. A title and 300- word abstract should be submitted as soon as convenient and not later than 31 May 1984 to : Professor B. R. Jennings. Electro-optics Group, Department of Physics, Brunel University, Uxbridge UB8 3PH.THE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY GENERAL DISCUSSION NO. 80 Physical Interactions and Energy Exchange a t the Gas-Solid Interface McMaster University, Hamilton, Ontario, Canada, 23-25 July 1985 Organising Committee : Professor J.A. Morrison (Chairman) Dr M. L. Klein Professor G. Scoles Professor W. A. Steele Professor F. S. Stone Dr R. K. Thomas The discussion will be concerned with certain aspects of current research on the gas-solid interface: elastic, inelastic and dissipative scattering of atoms and molecules from crystal surfaces, and the structure and dynamics of physisorbed species, including overlayers. Emphasis will be placed on the themes of physical interactions and energy exchange rather than on molecular-beam technology or the phenomenology of phase transitions on overlayers. The interplay between theory and experiment will be stressed as they relate to the nature of atom and molecule surface interaction potentials, including many-body effects. Contributions for consideration by the Organising Committee are invited; titles should be submitted as soon as possible, and abstracts of about 200 words by 31 May 1984, to: Professor J. A.Morrison, Institute for Materials Research, McMaster University, Hamilton, Ontario, Canada L8S 4M1 THE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY SYMPOSIUM NO. 20 Phase Transitions in Adsorbed Layers University of Oxford, 17-1 8 December 1985 Organising Committee : Professor J. S. Rowlinson (Chairman) Dr E. Dickinson Dr R. Evans Mrs Y. A. Fish Dr N. Parsonage Dr D. A. Young The aim of the meeting is to discuss phase transitions at gas/liquid, liquid/liquid and solid/fluid interfaces, and in other systems of constrained geometry or dimensionality less than three. Emphasis will be placed on molecularly simple systems, whereby liquid crystal interfaces and chemisorption phenomena are excluded.Contributions for consideration by the Organising Committee are invited and abstracts of about 300 words should be sent by 12 October 1984 to: Professor J. S. Rowlinson, Physical Chemistry Laboratory, University of Oxford, South Parks Road, Oxford OX1 3QZ. Full papers for publication in the symposium volume will be required by August 1985.FARADAY DIVISION INFORMAL AND GROUP MEETINGS Industrial Physical Chemistry Group The Metal-Polymer Interface To be held at Girton College, Cambridge on 10-12 July 1984 Further information from Dr T. G. Ryan, ICI New Science Group, PO Box 11, The Heath, Runcorn, Cheshire WA7 4QE Gas Kinetics Group 8th International Symposium on Gas Kinetics To be held at the University of Nottingham on 16-20 July 1984 Further information from Professor J.P. Simons, Department of Chemistry, University of Nottingham, Nottingham NG7 2RD Industrial Physical Chemistry Group Bacterial Adhesion To be held at Girton College, Cambridge on 11-1 3 September 1984 Further information from Dr I. D. Robb, Unilever Research, Port Sunlight, Wirral, Mersevside L63 3JW Electrochemistry Group with the SCI Electrochemical Technology Group Electrolytic Bubbles To be held at Imperial College, London on 13-1 4 September 1984 Further information from Dr G. H. Kelsall, Imperial College, London SW7 Polar Solids Group with British Ceramic Society, Institute of Physics and Mineralogical Society of Great Britain Kinetics and Mass Transport of Silicate.and Oxide Systems To be held at The Geological Society, London on 13-1 4 September 1984 Further information from Dr R.Freer, Department of Electrical and Electronic Engineering, North Staffordshire Polytechnic, Beaconside, Stafford ST1 8 OAD Division with the Gas Kinetics Group Autumn Meeting : Combustion Chemistry in the Gas Phase To be held at the University of Hull on 18-20 September 1984 Further information from Dr R. W. Walker, Department of Chemistry, The University, Hull HU6 7RX Colloid and Interface Science Group with the SCI Foaming, Aeration and Dynamic Surface Tension To be held at Imperial College, London on 24-25 September 1984 Further information from Professor A. Bailey, Imperial College, London SW7 Electrochemistry Group with Statistical Mechanics and Thermodynamics Group The Electrical Double Layer To be held at the University of Southampton on 25-26 September 1984 Further information from Dr E.Dickinson. Procter Department of Food Science, University of Leeds, Leeds LS2 9JT Carbon Group Refactory Applications of Carbon To be held at the University of Sheffield on 2 5 2 7 September 1984 Further information from Dr B. Rand, Department of Ceramics, Glasses and Polymers, University of Sheffield, Sheffield S10 2TZNeutron Scattering Group Neutron Scattering from Aqueous Solutions To be held at the University of Bristol on 27-28 September 1984 Further information from Dr G. W. Neilson, H. H. Wills, Physics Laboratory, University of Bristol, Tyndall Avenue, Bristol BS8 1TL Polymer Physics Group with the Institute of Marine Engineers Polymers in a Marine Environment To be held at the Institute of Marine Engineers, London on 30 October and 1 November 1984 Further information from Dr G.J. Lake, MRPRA, Brickendonbury, Hertford SG13 8ML Neutron Scattering Group Neutrons in Magnetism To be held at the University of Southampton in December 1984 Further information from Dr R. B. Rainford, Department of Chemistry, University of Southampton, Southampton SO9 5NH Division Annual Congress: Solid-state Chemistry To be held at the University of St Andrews on 26-28 March 1985 Further information from Professor P. A. H. Wyatt, Department of Chemistry, University of St Andrews, The Purdie Building, St Andrews KY16 9ST (vii)physicochemical topics, thereby encouraging scientists of different disciplines to contiibute their varied viewpoints to a common theme.A recent Discussion is :- The Royal Society No.7” Intramolecular Kinetics of Chemistry No.75 in the series, this publication is the result of a general discussion held at the University of Warwick in April 1983. Contents: The Spiers Memorial Lecture; Vibrational Redistribution within Excited Electronic States of Polyatomic Molecules Intramolecular Relaxation of Excited States lsomerization of Internal-energy-selected Ions Kinetics of Ion-Molecule Collision Complexes in the Gas Phase, Experiment and Theory Intramolecular Decay of Some Open-shell Polyatomic Cations On the Theory of Intramolecular Energy Transfer Pulsed Laser Preparation and Quantum Superposition State Evolution in Regular and Irregular Systems A Quantum-mechanical Internal-coUsion Model for State-selected Unimolecular Decomposition The Correspondence Principle and Intramolecular Dynamics Intramolecular Dephasing.Picosecond Evolution of Wavepacket States in a Molecule with Intermediate-case level Structure Energy Conversion in van der Waals Complexes of s-Tetrazine and Argon Time-dependent Processes in Polyatomic Molecules During and After Intense Infrared Irradlation Energy Distributions in the CN(X’Z+) Fragment from the Infrared Multiplephoton Dissociation of CF3CN. A Comparison between Expermental Results and the Predictions of Statistical Theories Of C6Fb + Product Energy Partitioning in the Decom- position of State-selectively Excited HOON and HOOD Low-power Infrared Laser Photulysis of Tetramethyldioxetan Unimolecular Reactions Induced by Vibrational Overtone Excitation Unimolecular Decomposition of t-Butylhydro- peroxide by Direct Excitation of the 6-0 0-H Stretching Overtone Picosecond-jet Spectroscopy and Photochemistry.Energy Redistribution and its Impact.on Coherence, Isomerization, Dissociation and Solvation Energy Redistribution in Large Molecules. Direct Study of Intramolecular Relaxation in the Gas Phase with Picosecond Gating Rotation-dependent Intramolecular Processesof SO?(AIA:) in a Supersonic Jet Role of Rotation-Vibration Interaction in Vibrational Relaxation. Energy Redistribution in Excited Singlet Formaldehyde Sub-Doppler, Spectroscopy of Benzene in the “Channel-three” Region Intramolecular Electronic Relaxation and Photoisomerization Processes in the Isolated Azabenzene Molecules Pyridine, Pyrazine and Pyrimidine Softcover 434 p 0 85186 658 1 Price f25.00 ($48.00) Rest of the World f26.00 RSC Members f16.25 Faraday Discussions of the Chemical Society No J 5 Inrrarnolecular Kinerrrs ‘983 Faraday Symposia are usually held annually and are confined to more specialised topics than Discussions, with particular reference to recent rapidly developing lines of research.A recent Symposium is :- NO.~? The Hydrophobic Interaction No. 17 in the series, this publication is the result of a symposium on The Hydrophobic lnterxtion held at the University of Reading in December 1982. Contents: Hydrophobic Interactions-a Historical Perspective Hydrophobic Hydration Geometric Relaxation in Water.Its Role in Precise Vapour-pressure Measurements of the Solubilization of Benzene by Aqueous Sodium Ortylsulphate Solutions Nuclear Magnetic Resonance Relaxation Investigation of Tetrahydrofuran and Methyl Iodide Clathrates Infrared and Nuclear Magnetic Resonance Studies Pertaining to the Cage Model for Solutions of Acetone in Water Isothermal Transport Properties in Solutions of Symmetrical Tetraakylammonium Bromides Thermodynamics of Cavity Formation in Water. A Molecular Dynamics Study Molecular Librations and Solvent Orient- ational Correlations in Hydrophobic Phenomena Monte Carlo Computer Simulation Study of the Hydrophobic Effect. Potential of Mean Force forC(CH4)iJaq at 25 and 50° C Hydrophobic Moments and Protein Structure Application of the Kirkwood-Buff Theory to the Problem of Hydrophobic Interactions Disentanglement of Hydrophobic and Electrostatic Contributions to the Film Pressures of Ionic Surfactants Hydrophobic Interactions in Dilute Solutions of Poly(viny1 alcohol) Conformational and Functional Properties of Haemoglobin in Water+Alcohol Mixtures. Dependence of Bulk Electrostatic and Hydrophobic Interactions upon pH and KCI concentrations Softcover 24 p 0 85186 668 9 Price f36.50%70.00) Rest of the World f 38.50 RSC Members f23.75 ORDERING RSC Members should send their orders to: The Royal Society of Chemistry, The Membership Officer, 30 Russell Square. London WC1B 6DT. Non-RSC Members The Royal Society of Chemistry, Distribution Centre, Blackhorse Road, Letchworth, Hem SG6 IHN, England. Faraday Symposia of the Chemical Society No 17 The Hvdrophabrc Interactron 1982 (viii)

ISSN:0300-9599    

DOI:10.1039/F198480FP037

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

J . Chem. SOC., Faraday Trans. I , 1984, 80, 1017-1028 Kinetic Study of the Partial Oxidation of Propene and 2-Methylpropene on Different Phases of Bismuth Molybdate and on a Bismuth Iron Molybdate Phase BY DOUGLAS CARSON,? MICHEL FORISSIER AND JACQUES C. VEDFUNE* Institut de Recherches sur la Catalyse, CNRS, 2 Avenue Albert Einstein, 69626 Villeurbanne, France Received 18th February, 1983 Kinetic studies of the partial oxidation of propene to acrylaldehyde and 2-methylpropene to methacrylaldehyde have been carried out on various bismuth molybdates [Bi,Mo,O,, (a), Bi,Mo,O, (a + y or B) and Bi,MoO, (y)] and Bi,FeMo,O,, (phase X). The partial pressures of 0, and alkene have been varied and a maximum in the yield of aldehyde has been observed for a value of the alkene pressure which depends on the sample.In contrast, no maximum has been observed for carbon dioxide. A simple mechanistic model has been proposed involving as a first step the rapid adsorption of the alkene. This is followed by the rate-determining step, the formation of a n-ally1 intermediate with a surface site S; the methacrylaldehyde is then formed relatively fast. Next 2-methylpropene reacts with site S to give an oxygenated hydrocarbon species which is strongly adsorbed and which blocks the site active for aldehyde formation. Finally the blocking hydrocarbon is oxidized into CO,, liberating the active surface sites. The model fits the experimental data well and allows one to differentiate different catalysts by the rate constant of step (1): the rate constants of the other steps are approximately equal.The study clearly shows that a comparison between catalyst performances is misleading, particularly for 2-methylpropene oxidation, under the same experimental conditions, since the relative ordering depends on the pressure of alkene with respect to that of oxygen. Carbon dioxide does not result from a total oxidation of the aldehyde under our conditions of low conversion levels. The initial activity for aldehyde formation is much higher in the case of 2-methylpropene than for propene and follows the order X % a+y > a > y > p. The alkane adsorption equilibrium may play the major role in catalytic activity. The discovery of catalysts based on bismuth molybdatel for the selective (amm-)- oxidation of propene and butenes has led to extensive research into such materials.A series of catalysts composed of Bi, Fe, 0 and one of the elements P, Mo and W has been patented.2*3 In a previous paper4 we have shown that, although all three phases Bi2M03012 (a), Bi,Mo,O, (j7) and Bi2Mo06 ( y ) of bismuth molybdate are selective alkene (amm-)oxidation catalysts, there are appreciable differences in their activity. A synergy effect occurring due to an intimate mixture of a and y phases was observed to result in enhanced selectivity for propene oxidation. This was in- terpreted as an appropriate addition of the useful properties of both phases through the various steps of the reaction, particularly for promoting the redox mechanism proposed earlier.5$ It has been shown' that maximum selective utilization of reactive lattice oxygen occurs after partial reduction.The role of various additives is mostly unknown since such additives may affect several parameters implied in the reaction. The role of Fe has been postulated to regulate the number of defects in a solid,8 to t Present address : Laboratoire de Recherches, CERCHAR, 60550 Verneuil en Halatte, France. 10171018 OXIDATION OF ALKENES ON BISMUTH MOLYBDATES enhance the Sb dissolution in SnO, for an SbSnO catalysta or even to form different phases with Bi and/or M o . ~ J * ~ ~ Most of the studies reported in the literature concerning the selective oxidation catalyst Bi-Mo-4 deal with the oxidation of propene or but-1-ene. There has recently been a growing interest in using other molecules as probes to elucidate the mechanism of the reactions mentioned above.Grassellil* has used azopropene and allyl alcohol to study the allyl intermediate, and Gryzbowska et aZ.lg have used C,H,I to study the functionality of Bi and Mo. On the other hand little attention has been paid to the selective oxidation of 2-methylpropene except in the patent literature. 2-Methylpropene is an ideal molecule to study allyl intermediates because the structure and symmetry of the intermediate is presumed to be the same as for the propene, except for the addition of a methyl group which has electron-donating properties and can stabilize or destabilize the intermediate. Thus by comparing the reaction ratio of propene and 2-methylpropene partial oxidations one might hope to tell whether the intermediate is stabilized or not by the addition of electron density.The effect of alkyl substituents on the oxidation of a series of alkenes has been studied20 at 733 K over bismuth molybdate. The relative reactivities of the alkenes are related to the type of allylic H abstraction: tertiary > secondary > primary. Using different substituents for the oxidation of p-substituted alkylbenzenes, Burrington et suggested that the n-ally1 intermediate in the rate-determining step is radical-like in character. However, an inherent danger in comparing different reactants is that they must be compared under equivalent conditions. A question immediately arises : what exactly are equivalent conditions, such as amount of adsorbed reactants, state of surface reduction, strength of adsorption etc.? In spite of the small variation in the partial pressure of reactant which may occur in an industrial installation, a kinetic study may show important differences between catalysts.It is a powerful way to obtain information on reaction mechanisms. This kind of study is rare for the reaction of propene and we do not know of such a study in the case of 2-methylpropene. Moreover, non-stationary kinetic studies are useful to prove experimentally that adsorption takes place and to distinguish between fast and slow steps of the reaction. The purpose of the present work was to compare the kinetic and catalytic behaviour of various bismuth molybdates and of a BiMoFeO phase for the partial oxidation reactions of both propene and 2-methylpropene.EXPERIMENTAL PREPARATION OF CATALYSTS The a, /3 and y phases were prepared as described in ref. (4): an acidic solution A of bis- muth nitrate (PH 1) and a basic solution B of ammonium paramolybdate, in relative concentrations .calculated to yield the stoichiometry chosen for the phases, were prepared. Coprecipitates were obtained by adding B dropwise into A while stirring and by adjusting the final pH to 7 by adding NH,OH. The Bi,FeMo,O,, phase was prepared by coprecipitation of the different salts at a pH automatically controlled as described in ref. (22). The starting salts were bismuth and ferric nitrates and ammonium paramolybdate of RP grade supplied by Prolabo. An HNO, solution of the three salts was prepared to ensure a Bi:Fe:Mo = 3: 1 : 2 stoichiometry and was introduced at a constant rate into the preparation vessel.Ammonium hydroxide solution was simultaneously introduced but the rate was automatically controlled to maintain the pH at 2.7. This value was chosen by preliminary studies of the precipitation conditions of Bi and Fe hydroxides separately as a function of the pH by adding ammonium hydroxide solution. The samples were dried in an oven at 383 K overnight and calcined in air at 723 K to eliminate NH; and NO; ions.D. CARSON, M. FORISSIER AND J. C. VEDRINE 1019 Table 1. Catalytic data for the partial oxidation of propene and 2-methylpropene at 633 K into acrylaldehyde and methacrylaldehyde, respectively. po, = 80 Torr, 1 Torr = 133.3 Pa sample Bi3FeMo20,,, X 2.3 106 80 763 2 24 Bi2M03012, a 1.3 15 200 200 7 20 Bi,Mo,O,, a + y 3.2 29 50 270 3 20 Bi,MoO,, y 2.7 5 300 82 26 20 - Bi,Mo,O,, D 0.5 3 300 65 19 (A) Activity for low propene pressure (133 Pa or 1 Torr) (in mol rn-, s-l).(B) Propene pressure at maximum acrylaldehyde yield (in Torr). (C) Activity for low 2-methylpropene pressure (1 Torr) (in mol m-2 s-l). (D) 2-Methylpropene pressure at maximum methacryl- aldehyde yield (in Torr). (E) Number of adsorbed carbon atoms per nm2 of surface. The amount of C per nm2 area was determined by complete cut-off of alkene feed at steady state as described in the text in experiment (i) and amount of CO, formed by complete combustion of the adsorbed hydrocarbon species. The specific surface areas are given in table 1. The various samples were characterized by X-ray diffraction and i.r.spectro~copy.~ The Bi3Fe,Mo0,, phase, yellow in colour, presents an X-ray diffractogram very similar to that of the Bi3FeMo20,, phase obtained by Sleight13 and close to the phase X identified by Lo Jacono et ~ 1 1 . ~ ~ as scheelite in nature. As described in ref. (4), the sample Mo:Bi:O = 2:2:9 was found by i.r. spectroscopy and X-ray diffraction to be an equimolar mixture of a and y phases. The true /3 phase could be prepared by further calcining the sample above 825 K. CATALYTIC REACTIONS Two similar flow reactors were used for the oxidations of propene and 2-methylpropene with adapted chromatographic analysis. The flow rate was generally 1 cm3 s-l, the weight of catalyst was 100 or 200 mg and the conversion level was < 4%.For kinetic measurements the mixing of reactants was controlled with a three-way Brooks flow regulator. A steady state was attained before each measurement. In order to be sure that no aging or modification of the catalyst occurred when varying the relative ratio of alkene to oxygen, successive values of this ratio were arbitrarily varied from lower to higher values and vice versa. Non-stationary experiments were also carried out in order to obtain information on surface species. (i) Under steady conditions the 2-methylpropene feed was suddenly and completely cut off. (ii) Under the same steady conditions, and with the same experimental set-up, the 2-methylpropene feed was suddenly dropped to a fifteenth of its value. This was obtained by the rotation of a three-way value which cuts out part of the 2-methylpropene flow.A rapid analysis (3 min) could be performed to observe the change in concentration of the products. RESULTS The catalytic data concerning the partial oxidation of propene and 2-methylpropene at 633 K are given in fig. 1 and 2. A maximum in the yield of methacrylaldehyde (fig. 2) from 2-methylpropene was observed, while a near plateau was obtained for CO,. The effect of varying the oxygen pressure is shown in fig. 3. It is clear that the phenomenon mentioned above does not depend on the oxygen pressure. This kind of curve is generally observed when there is competition for adsorption between reactants and products on the catalyst.1020 OXIDATION OF ALKENES ON BISMUTH MOLYBDATES 0 a rn - W W - 10 20 30 40 I t pressure of propene/ 1 O3 Pa Fig, 1.Variation in the rate of formation of acrylaldehyde (0) and CO, (0) as a function of the partial pressure of propene at 633 K for a+y Bi,Mo,O,. po, = 80 Torr x 10600 Pa. I I - 5 10 pressure of 2-methylpropene/ 1 O3 Pa Fig. 2. Variation in the pressure of methacrylaldehyde (H) and CO, (0) produced as a function of 2-methylpropene partial pressure at 633 K on the a+y sample. The full line corresponds to the calculated best fit to the experimental points following steps (1)-(4) in the proposed mechanism. po, = 80 Torr x 10600 Pa.D. CARSON, M. FORISSIER AND J. C. VEDRINE E - I ; 2 $ E 2 W h 1021 1 I I t I 8 12 c I 8 12 pressure of 2-methylpropene/l O3 Pa Fig. 3. Rate of formation of metacrylaldehyde (upper curves) and CO, (lower curves) as a function of 2-methylpropene pressure for various pressures of 0, for the a + y sample : (a) po, = 150 Torr x 20000 Pa, (6) po2 = 85 Torr x 11 300 Pa and (c) po2 = 48 Torr z 6400 Pa.The results of non-stationary experiments give more information about this adsorption. Indeed the variations of product concentration as a function of time are plotted in fig. 4. In case (i) when the supply of 2-methylpropene is completely cut off only CO, could be detected. The amount of CO, formed can be evaluated from the area under the curve (the bent part). In case (ii) the 2-methylpropene flow dropped to a 15th of its value and the amount of methacrylaldehyde formed dropped drastically at time t = 0 but increased progressively again with time. Experiment (i) shows that some adsorbate compounds limit the formation of methacrylaldehyde and are converted into CO,, whereas experiment (ii) shows that the elimination of such1022 OXIDATION OF ALKENES ON BISMUTH MOLYBDATES I t 0 0.5 1 tlh 0 1 tlh 2 Fig. 4.Variation in the pressures of (a) 2-methylpropene, (b) methacrylaldehyde and (c) CO, with time when the 2-methylpropene pressure is suddenly dropped to zero (upper curves) and to a fifteenth of its value (lower curves). adsorbates liberate sites which can again oxidize 2-methylpropene from the feed to methacrylaldehyde. Further experiments of type (i) were made with the catalyst placed in a thermal analysis calorimeter under continuous flow of reactant mixture at the reaction temperature. The microcalorimetric measurements show that the heat does not vanish when the admission of 2-methylpropene stops but rather follows the formation of CO,.This heat does not come from CO, desorption or COT decomposition but from an exothermic process, probably oxidation of a hydrocarbon species adsorbed on the surface of the solid. Experiments using infrared spectroscopy of phase X, treated in situ under conditions close to those of the catalytic reaction, exhibit C=O and C-H vibrational bands at 1725, 1795 and 1385, 1455 cm-l, respectively. This result shows that the adsorbed hydrocarbon is an aldehyde, a ketone or an acid, and is possibly polymeric in nature.D. CARSON, M. FORISSIER AND J. C. VEDRINE! 1023 2.5 5.0 7.5 pressure of 2-methylpropene/103 Pa Fig. 5. Variation in the conversion level at 633 K as a function of the 2-methylpropene pressure for the various samples.po, = 80 Torr x 10600 Pa. c 100- Y B X a a + Y ; 5 10 15 pressure of 2-methylpropene/l O3 Pa Fig. 6. Variation in the selectivity for methacrylaldehyde formation at 633 K as a function of 2-methylpropene pressure for the various samples. po, = 80 Torr x 10600 Pa. The variations of the conversion of 2-methylpropene and its selectivity with regard to methacrylaldehyde are plotted in fig. 5 and 6, respectively, for the various samples as a function of 2-methylpropene pressure, the oxygen pressure being kept equal to 80 Torr, the temperature to 623 K and the flow rate to 1 cm3 s-l. These curves show great differences between the catalysts. A comparison of the catalytic properties of the various samples shows a relative order between them, showing that the catalytic properties are very dependent on the reaction conditions.For instance, at a1024 OXIDATION OF ALKENES ON BISMUTH MOLYBDATES 2-methylpropene pressure of 80 Torr the activity follows the order y > X > p > a+y > a, whereas at 5 Torr 2-methylpropene pressure the order is X > a+y > a > y > p. In the case of propene oxidation the maximum in activity occurred at too high a relative pressure to allow the curves to cross each other, and the ordering remained the same over our range of pressures. These results may well explain the discrepancies observed in such an ordering by other workers. In the case of propene the above features were not so striking and a maximum in acrylaldehyde yield was not obtained for the p and y phases.However, since a maximum was observed for the X, a and cx+y phases the same type of behaviour probably occurred for all samples and may reasonably be generalized to propene. In order to compare the different samples, the activities at very low alkene pressure (133 Pa) are given in table 1 as the alkene pressure corresponding to the maximum acrylaldehyde or methacrylaldehyde yield, together with the alkene: oxygen ratio and the amount of CO, formed by total combustion of adsorbed carbonaceous compounds [experiment (i)]. Note that the values in the last column, i.e. the number ofcarbon atoms adsorbed per nm2, is about the same for all samples. DISCUSSION The fact that the activity of the catalysts at low alkene pressure is ca.10 times greater for 2-methylpropene than for propene may be interpreted as the effect of the electron-donating properties of the extra methyl group, giving a more stable reaction intermediate.18-,l However, our experimental results show that extensive adsorption of reaction products occurred. Moreover, the reactants may also be adsorbed and therefore the difference in the adsorption constants of 2-methylpropene and propene might also explain the difference in activity between these alkenes. We will discuss this in more detail below. Even in the case of adsorption of the reactant it must be remembered that, since hydrogen abstraction is the rate-limiting step, the reactant is assumed to be in equilibrium with the gas phase and therefore is not involved in a rate process and is influenced by adsorption equilibrium rather than by adsorption kinetics.Thus both adsorption and desorption are fast compared with the rate-determining step. The maximum in activity for formation of acrylaldehyde or methacrylaldehyde and the plateaus reached for CO, indicate that there is probably competition for the active surface sites. This may involve the aldehyde or oxygen which is denied access to the surface, the species denying access undoubtedly being a reaction product strongly adsorbed on the surface. This can be seen more clearly from the non-stationary experiments (i) and (ii). These results show that the adsorbed 2-methylpropene, which ultimately reacts to form methacrylaldehyde, is easily described since methacryaldehyde was no longer detected when the 2-methylpropene was suddenly removed from the feed stream.In contrast, the amount of CO, in the reaction products diminished slowly after the 2-methylpropene feed was cut, until it finally reached zero. When the pressure of 2-methylpropene was reduced from 80 to 5 Torr the CO, level descended slowly as previously, but stopped at a new steady-state level. The amount of methacryl- aldehyde diminished to ca. 5-10% of its value at 80 Torr and then rose slowly to a new steady-state value. This rise coincided with the reduction in CO,. These results indicate that a hydrocarbon is strongly adsorbed on the surface and blocks surface-active sites. It oxidizes slowly and then liberates sites to permit the formation of methacrylaldehyde to proceed more rapidly.The fact that the rate of formation of carbon dioxide is neither a function of theD. CARSON, M. FORISSIER AND J. C. VEDRINE 1025 rate of methacryaldehyde formation nor a function of the presence of methacrylalde- hyde in the product stream indicates that, under the present experimental conditions, carbon dioxide does not come from the total oxidation of the methacrylaldehyde. That being said, note that the activity curves for acrylaldehyde or methacrylaldehyde and C02 are related to each other in the same way for all the catalysts studied; i.e. the 2-methylpropene pressure which produces the maximum activity for methacrylaldehyde also produces 63% of the maximum activitity for carbon dioxide for the various samples. It then appears that there is a common intermediate in the formation of methacrylaldehyde and CO,, but this intermediate is not absorbed methacrylaldehyde. The simplest mechanistical model that one may reasonably suggest is as follows: k3 Rads Bads where R is the reactant (2-methylpropene or propene), Rads and Bads are adsorbed species of reactant other than from the x-ally1 complex, S an active site, BS is a blocked active site, A is the aldehyde (methacrylaldehyde or acrylaldehyde), k,, kd and Kl are, respectively, the rate constants for adsorption and desorption and the adsorption equilibrium constant, ki is the rate constant of the ith reaction, and Ox, rx and px are, respectively, the coverage ratio, rate of formation and partial pressure of component x.No assumption is made about the active site, which may be unique or may correspond to several sites necessary for the reaction. The partial pressure of 0, is taken to be constant. At the stationary state, rBS = 0 and OS + 8,, = 1. The rates of formation of aldehyde and C02 may be written as and where A = k2, B = k,/k4, C = l/Kl and D = k,. The reaction mechanism can be explained as follows. The first step is the rapid adsorption of the alkene, considered to be virtually at equilibrium. Secondly, the adsorbed alkene reacts with a surface site to form the z-ally1 intermediate: this is the rate-determining step. After the n-ally1 intermediate is formed the aldehyde is formed relatively fast such that its rate of formation is limited by the x-ally1 concentration.In the third step 2-methylpropene reacts to become a strong adsorbed species which blocks site S from participating in step (2). Finally, in step (4) the blocking species is oxidized to form CO,, liberating an active surface site. Step (4) is the only mechanism postulated in which the strongly adsorbed 2-methylpropene leaves the surface.1026 OXIDATION OF ALKENES ON BISMUTH MOLYBDATES Table 2. Constants of steps (1x4) applied to experimental data for the partial oxidation of 2-methylpropene to methacrylaldehyde on various bismuth molybdates. For comparison purposes the assumption is made that the concentration of active surface sites is the same for all catalysts sample ~~ X 0.56 2.1 0.24 0.28 a 0.15 2.2 0.15 0.15 a+Y 0.33 1.3 0.12 0.13 0.053 1.9 0.16 0.15 0.038 3.4 0.4 0.5 Y P The species which poisons the active sites is not an intermediate in the formation of methacrylaldehyde as discussed above [experiments (i) and (ii)] and is strongly adsorbed on the surface.Infrared spectra of a spent catalyst revealed the presence of C=O and C-H bonds, whereas the microcalorimetry results showed that this adsorbed species is oxidized to CO, before it is liberated. Indeed the only way the blocking species can leave the surface is by total oxidation in the presence of gaseous oxygen, whereas the substantial heat of combustion of the blocking species indicates that a hydrocarbon-like species is involved. Grzybowska et have observed that carbonaceous compounds on the surface of spent bismuth molybdate catalysts do not desorb at 700 K in uacuo.Therefore it may be suggested that the blocking species is an oxygenated hydrocarbon strongly adsorbed on the catalyst surface and not an intermediate species. It must be remembered that the present data are initial activities measured at low conversion levels (< 4%). The statement that CO, does not come directly from methacryaldehyde or acrylaldehyde does not preclude the total oxidation of the aldehyde at higher conversion levels. It is possible from the experimental curves in fig. 1 and 2 to fit experimental and calculated values. The full lines in fig. 1 and 2 correspond to the best fit with the experimental points. The values of k obtained are given in table 2. The values of k2-k4 do not differ significantly with the sample. However, the value of K,, i.e.the adsorption equilibrium constant of the alkene, varies drastically with the sample. The same model can be used for the selective oxidation of propene. As seen in the Experimental section, there is no well defined maximum in acrylaldehyde activity. Therefore one may be less confident in the model since the data easily fit a three-parameter equation. Nonetheless, one can comment that, in general, the maximum or plateau occurs at a much higher pressure of propene than in the case of 2-methylpropene. According to our model, this indicates that the former is less readily adsorbed than the latter. Such a difference in adsorption equilibria may explain the difference in activity between 2-methylpropene and propene at low alkene pressure. The model and its good fit with experimental values show that adsorption of alkene may be the important step of the reaction.K, is the constant which changes most significantly with the nature of the catalyst and with the nature of the alkene. We have discussed above the other possibility that the difference in reaction rate between propene and 2-methylpropene may be explained as being due to the donating property of the extra methyl group giving a more stable intermediate. We have no experimental data either to prove this or to rule it out definitely. Experiments were carried out using a Texas gauge to try to determine the adsorption capacity of the samples with respectD. CARSON, M. FORISSIER AND J . C. VEDRINE 1027 to propene and 2-methylpropene. We did not succeed, first because the adsorption was too weak to be detected and secondly because reaction occurred too easily, which precluded any differentiation with adsorption of reactants.This model is very simple and does not provide any details concerning the nature of the oxygenated hydrocarbon species that poisons the surface, or of the nature of the active site(s). Also, the role of oxygen has not been taken into account. However, since the oxygen pressure has no influence on the shape of the curves in fig. 1 and 2, it can be considered to be included in the constants in the kinetic rate equations; the role of oxygen could infer the blocking mechanism. It is indeed possible that hydrocarbon adsorption decreases the rate of reoxidation of the catalyst by molecular oxygen and therefore results in a more reduced catalyst which is more adsorbent and less selective.The virtue of this model is that it fits the data well. Also, even if the model is not entirely correct, the mathematical description characterizes the data well. Introduction of a small amount of aldehyde into the inlet feed might be thought to give more information concerning the mechanism of reaction as proposed. However, it is obvious that such an aldehyde is oxidized into CO, via step (4) or any other step. Moreover (and more importantly), the decrease in the oxidation rate of the alkene upon introduction of aldehyde was experimentally impossible to follow with success with our experimental set-up and low conversion conditions. CONCLUSIONS The kinetic studies of the partial oxidation of propene and 2-methylpropene at low levels of conversion (< 4%) allow one to draw the following conclusions.(i) The electron-donating properties of the methyl group may be related to an increase in the catalytic rate of the partial oxidation properties of all bismuth molybdate samples when comparing propene and 2-methylpropene. (ii) The catalytic kinetic behaviour may well be described by a simple model in four steps which includes (a) rapid adsorption on the alkene, (b) the formation of a 71-ally1 intermediate (the rate-determining step) and its transformation into the aldehyde, ( c ) the reaction of the alkene and oxygen with active surface sites resulting in an oxygenated hydrocarbon which blocks and thus poisons the active site and, finally, ( d ) the total oxidation of this blocking species into CO,, liberating the active sites.(iii) Catalytic behaviour only has a basis for comparison if one considers the initial activity at very low pressure of alkene or the rate constant of step (1). The ordering is then X % a + y > a > y > /I. The ordering varies, particularly for 2-methylpropene, if it is determined at different relative pressures of alkene and oxygen, i.e. the ordering is invalid in general, except under precise conditions. (iv) The equilibrium of the alkene adsorption may be the major parameter in the catalytic activity and may be particularly sensitive to the atomic arrangement at the bulk surface. (v) The blocking of active sites is due to an oxygenated hydrocarbon whose presence in mass per unit surface area is the same whatever the sample.This blocking hydrocarbon is oxidized to CO,, liberating the active sites, but does not give the aldehyde. The contributions of D. Foujols, J. C. Volta (IRC), A. Laarif and F. Theobald (Universite de Franche Comti, BesanCon) are gratefully acknowledged.1028 OXIDATION OF ALKENES ON BISMUTH MOLYBDATES 1 J. D. Idol, U.S. Patent, 2,094,580, 1959; R. K. Grasselli and H. F. Hardman, U.S. Patent, 3,642,930, 1972. H. H. Voge, W. E. Armstrong and L. B. Ryland, U.S. Patent, 3,110,746, 1963. D. Carson, G. Coudurier, M. Forissier, J. C. VCdrine, A. Laarif and F. Theobald, J. Chem. SOC., Faraday Trans. I , 1983, 79, 1921. P. Mars and D. W. Van Krevelen, Chem. Eng. Sci., 1954, 3, 41. F. Weiss, J. Marion, J. Metzger and J. M. Cognion, Kinet. Katul., 1973, 14, 32. J. F. Brazdil, D. D. Suresh and R. K. Grasselli, J. Cutul., 1980, 66, 347. K. Yu. Adzhamov, F. M. Poladov and T. G. Alkhazov, Kinet. Kutal., 1979, 20, 1350. B. Benaichouba and J. M. Herrmann, React. Kinet. Catal. Lett., 1983, 22, 209. lo I. B. Anenkova, T. G. Alkhazov and M. S. Belenskii, Kinet. Katul., 1969, 10, 1305. l1 Ph. A. Batist, C. G. Moesdijk, I. Matsuura and G. C. A. Schuit, J. Catal., 1971, 20, 40. l2 G. W. Keulks, J. L. Hall, C. Daniel and K. Suzuki, J. Catal., 1974, 34, 79. l3 A. W. Sleight, Muter. Res. Bull., 1974, 9, 951. l4 D. V. Tarasova, T. V. Andrusmkevich, V. A. Dzisko, Ye. Sazonova, T. A. Nikoro and E. G. Ismailov, l5 L. Ya. Margolis, in Oxidation of Hydrocarbons on Heterogeneous Catalysts (Khimiya, Moscow, 1977) la P. L. Villa, A. Szabo, F. Trifiro and M. Carbucicchio, J. Catal., 1977, 47, 122. 3 J. C. Daumas, J. Y. Derrien and F. Van Den Bussche, French Patent, 2,364,061, 1976. React. Kinet. Catal. Lett., 1976, 4, 287. and references herein. I. V. Nicolescu and I. Sandulescu, Rev. Roum. Chirn., 1981, 26, 217. J. D. Burrington, C. T. Kartisek and R. T. Grasselli, J. Catal., 1980,63,235; R. K. Grasselli and J. D. Burrington, Adu. Catal., 1981, 30, 133. ID B. Grzybowska, J. Harber and J. Janas, J. Catal., 1977, 49, 150. 2o C. L. Adams, in Proc 3rd Znt. Congr. Catal. (North Holland, Amsterdam, 1965), p. 240. 21 J. D. Burrington, C. T. Kartisek and R. T. Grasselli, J. Org. Chem., 1981, 46, 1877; J. Catal., 1981, 22 J. C. Volta et al., to be published. 23 M. Lo Jacono, T. Noterman and G. W. Keulks, J. Catal., 1975, 40, 19. 24 B. Grzybowska, J. Haber, W. Marczewski and L. Ungier, J. Catal., 1976, 42, 327. 69, 495. (PAPER 3/261)

ISSN:0300-9599    

DOI:10.1039/F19848001017

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. 1, 1984, 80, 1029-1038 Differential Pulse Voltammetry as an in situ Monitoring Technique for the Thermal-decomposi tion Kinetics of Nitrate Melts BY FRANCESCO PALMISANO, LUIGIA SABBATINI AND PIER GIORGIO ZAMBONIN* Dipartimento di Chimica, Laboratorio di Chimica Analitica, Universita degli Studi, Via Amendola 173, 70126 Bari, Italy Received 14th April, 1983 Differential pulse voltammetry has been used to detect NO; at trace level in fused nitrates. A detection limit of 10ppb was estimated. The technique was employed for studying the thermal-decomposition kinetics of molten nitrates under controlled experimental conditions. A pseudo-zero-order kinetic process was found, characterized by a kinetic constant of 7.6 x mol kg-' s-l at 538 K and by an apparent activation energy of 39 & 2 kJ rno1-l in the temperature range 510-620 K.The suggested reaction mechanism is 2 NO; s NO; + NO;* NO;* -+ NO, ++O, where the first reaction seems to represent an equilibrium prior to the second rate-determining reaction. The equilibrium constant and the value of AH for the reaction NO; = NO; +to, were also calculated from equilibrium data and compared with previous literature parameters estimated by indirect or approximate methods. Alkali-metal nitrate melts have been ~ s e d l - ~ for many years as heat-transfer fluids in chemical processing industries or metal heat-treatment plants and represent one of the most extensively studied fused-salt systems. At present, interest in these solvents is due mainly to their potential application as high-temperature heat-transfer fluids and thermal-storage media5-' in energy conversion and energy storage plants (see, for example, solar thermal electric power plants).Mixtures of NaNO, and KNO, are particularly suited for these purposes because of their chemical stability (even on contact with air), low melting point, high specific heat, low vapour pressure, low corrosive properties, convenient viscosity and low cost. Furthermore, these properties and our knowledge of the basic chemistry and electrochemistry of oxygen8 and hydrogeng systems in these melts, make molten NaNO, + KNO, mixture potentially suited as solvent for medium temperature H,-0, fuel cells and they could offer a good compromise between sufficiently fast electrode kinutics and low maintenance costs.Knowledge of the thermal stability of these melts and of their decomposition products is therefore of fundamental importance both from a theoretical and a practical point of view. For a long time sodium and potassium nitrates were considered to be stable in air up to 773 and 873 K, respe~tively,l~-~~ and consequently most of the published papers 10291030 THERMAL DECOMPOSITION OF NITRATE MELTS concerning their thermal stability refer to high-temperature values where the main decomposition reaction is paralleled by secondary processes which can lead to the formation of several species such as 02-, OH-, N, etc. More recently, thermal decomposition has been assumed14~ l 5 to proceed to a certain extent at lower temperatures, thus accounting for the presence of ‘unavoidable’ NO; impurities in melts which have been prepurified and carefully melted to avoid local overheating.The thermodynamics16 of reaction (1) and the kinetics17 of the reverse of reaction (1) have been studied in this laboratory in the temperature range 500-700 K. However, at present no information is available on the kinetics of reaction (1). This is ascribed in part to the absence of a simple, precise and sensitive technique suitable for monitoring in situ the decomposition rate of the nitrate ions. In the present work differential pulse voltammetry (d.p.v.) has been used to detect NO; in the low ppb range. The technique proved to be good enough for following reaction (1) even at temperatures as low as 500 K, where the melt is usually considered thermally stable.The thermal-decomposition kinetics of (K, Na)NO, melts maintained under an oxygen atmosphere (Po, = 1 atm) have been studied in the temperature range 510-620 K, and the thermodynamics of reaction (1) have been re-evaluated and compared with literature data. NO, = NO; +$02 (1) EXPERIMENTAL The chemicals used and experimental set-up have been described el~ewhere.~ NO; impurities were removed from the melt by exhaustive potential-controlled electrolysis at +0.65 V on a large platinum anode (e.g. the Pt beaker used as melt container). An Amel model 551 potentiostat was used for this purpose. All potentials are referred to an Ag,Ag+ 0.07 mol kg-l reference electrode. The counter electrode used during the exhaustive electrolysis for the purification of the melt was a large piece of Pt foil placed in a separate compartment.At the end of the pre-electrolysis it was necessary to remove the fritted counter electrode compartment (enriched in NO; during the exhaustive electrolysis step) from the cell in order to avoid recontamination of the melt. A Pt wire directly immersed in the melt was then used as a counter electrode for the voltammetric investigations. Differential pulse voltammograms were run on a Pt disc microelectrode (standing or rotating) using a PAR 174A polarographic analyser in conjunction with a Philips PM 8 121 X Y recorder. Oxygen or nitrogen of UPP grade, presaturated with water at a vapour pressure of 15 Torr,* were used to maintain the melt under an appropriate atmosphere and to keep its water content constant. When necessary, the NO; content in the melt could be varied by adding small droplets of a solid solution of NO; in the same solvent.For temperatures > 580 K control experiments were performed in which the kinetics were followed by the same ex situ spectrophotometric method used by Kust and Burke.14 RESULTS AND DISCUSSION DIFFERENTIAL PULSE VOLTAMMETRY The voltammetric behaviour of the NO;/NO, system in anhydrous and wet fused nitrates has been the subject of previous investigations.l89 l9 In water-saturated melts * 1 Torr E ( 10 I 325/760) Pa.F. PALMISANO, L. SABBATINI AND P. G. ZAMBONIN 1031 +0.2 +0.4 t0.6 V/V us Ag, Ag' Fig. 1. Nitrite oxidation d.p. voltammogram obtained in water-saturated (Na, K)NO, melt at 533 K.[NO,] = 1.9 x mol kg-l; pulse repetition time, 1 s; pulse amplitude, 50 mV; scan rate, 5 mV s-l. (e.g. the condition in this work) a catalytic mechanism with partial regeneration of the depolarizer is operating, the overall electrode process being represented by (2) NO; + H,O --+ NO; + 2H+ + 2e. In dried melts NO; is oxidized according to the simple one-electron reversible process NO; -+NO,+e (3) while in partially wet melts a mixture of reactions (2) and (3) operates. The differential pulse voltammetric response was found to be consistent with the previous results. The conditions for optimum sensitivity and resolution were : pulse amplitude 50 mV, pulse repetition time 1 s and scan rate 2-5 mV s-l. Fig. 1 shows a typical differential pulse voltammogram obtained in a water-saturated melt containing FJO;] = 1.9 x mol kg-l; a detection limit (SIN = 2) of 2 x lo-' mol kg-l (ca.10 ppb) was estimated for the specified experimental conditions used. The differential pulse voltammetric response under conditions other than pure diffusion was also measured since in molten salt studies it is often necessary to use solid electrodes under conditions of convective mass transport (e.g. when rotation of the electrode is necessary in order to remove gaseous reaction products etc.). A typical plot of the d.p.v. peak current against the square root of the electrode rotation rate (04) is shown in fig. 2. There is a region, up to ca. 600 r.p.m., in which the d.p.v. current is insensitive to the convective mass transport. However, after a short transition region, a linear dependence between the peak current and O: is verified.In the first region of the plot the Nernstian diffusion-layer thickness is small in comparison with the diffusion convective-layer thickness so that the independence of1032 THERMAL DECOMPOSITION OF NITRATE MELTS < 4 i CI c & 3 u 2 / I 1 I I 0 5 10 15 20 uv2/(rad s-')v2 Fig. 2. Plot of d.p.v. peak current against mi. the peak current from the rotation speed is readily understood. On the other hand, the formulation of d.p.v. current given by Parry and Osteryoung20 nFACD (a- 1) Aim,, = ~ - 6 (a+l) where 0 = exp (nFAE/2RT), may be considered of general validity, assuming that both in unperturbed solutions and under diffusive control, 6 in eqn (i) is represented by the Nernst diffusion-layer thickness 6 = .\/(nDt) (ii) while under the opposite conditions, i.e.pure convective control, 6 in eqn (i) is represented by 1.6116Di6 s = (iii) where v is the kinematic viscosity. Combination of eqn (i) and (iii) accounts for the linear dependence of peak current on cut. The experimental slope was found to be ca. 10% lower than that calculated according to eqn (i) and (iii), the discrepancy being ascribed to some uncertainty in determining the electrode area and the NO, diffusion coefficient. Between these two extreme conditions a region of mixed convective-diffusive control exists which can be described by the treatment of Osteryoung et aL21 An improvement in sensitivity, induced by the convective mass transport, is apparent from fig.2 but it is not sufficiently advantageous for analytical purposes when considering the technical problems usually associated with high rotation speeds. On the other hand, it was found that the reproducibility of d.p.v. measurements on stationary electrodes became worse when the temperature was > ca. 570 K. Typical measurements performed at 655 K with standing and rotating Pt disc electrodes are reported in table 1 . The d.p.v. current at the stationary electrode is progressively lowered as the number of measurements increases; this behaviour was not observed when a rotating electrode was used. This phenomenon may be accounted for by electrode poisoning (e.g. by oxidation products) which becomes noticeable at high temperatures and which can be significantly reduced by electrode rotation.Thus allF. PALMISANO, L. SABBATINI AND P. G. ZAMBONIN 1033 Table 1. Reproducibility of the nitrite oxidation d.p.v. peak current on both standing and rotating disc electrodes; [NO,] = 3.62 x lop4 mol kg-'; T = 655 K; pulse repetition time, 1 s; pulse amplitude, 50 mV; scan rate, 5 mV s-I peak current/,uA rotating disc electrode experiment number standing electrode (600 r.p.m.) 4.45 4.25 3.85 3.45 3.20 mean 3.85 r.s.d. (%) 13.7 4.55 4.65 4.55 4.70 4.65 4.62 1.4 the measurements at temperatures > 570 K should be performed using a rotating disc electrode. The d.p.v. current was also found to be insensitive to vigorous gas purging of the melt at a suitable distance from the electrode. This was particularly useful when continuous purging was required, as in the experiments described below.A more thorough investigation on the oxidation mechanism of NO; under the present experimental conditions is out of the scope of the present investigation, which is only concerned with the conditions required for using this electroanalytical technique (d.p.v.) as a tool for in situ monitoring of the NO; concentration in melts. PURIFICATION OF THE MELT FOR THE THERMAL-DECOMPOSITION STUDY It is well known that, whatever the purification procedure of the starting salts, a (Na, K)NO, melt invariably contains NO; impurities. Furthermore, as observed by several authors, the ageing of the melt causes an increase in the nitrite content. This observation suggests that the presence of NO; derives not only from nitrite impurities in the nitrates, or from other impurities capable of reducing nitrate ion to nitrite ion when in a molten state,,, but also from the thermal decomposition of the melt even at temperatures around the melting point.In order to follow the decomposition kinetics, purification of the melt from the NO; impurities was necessary. Purging,, with NO,, which reacts with NO; according to the reaction NO, +NO; -+ NO; +NO (4) was not effective for the present work since the NO; content could be lowered to only ca. mol kg-l. The same results were obtained with direct injections in the melt of concentrated HNO,, which decomposed according to the reaction 2 HNO, -+ +0, + H,O + 2 NO, i.e. with in situ production of nitrogen dioxide. Massive potential-controlled electrolysis on a Pt anode revealed the most efficient method for the removal of NO;.This purification procedure had an additional advantage in that the efficiency of the electrolysis could be followed by monitoring the decay of the electrolysis current and/or by recording the d.p. voltammogram of1034 THERMAL DECOMPOSITION OF NITRATE MELTS tO.2 +0.4 +0.6 +0.8 V/V us Ag, Ag’ Fig. 3. D.p. voltammograms recorded at different extents of electrolysis: t = (1) 0, (2) 2, (3) 6, (4) 12 and (5) 30 h. Electrolysis potential, +0.65 V; T = 533 K; pulse repetition time, 1 s; pulse amplitude, 50 mV; scan rate, 5 mV s-l. Current sensitivity: curve (l), 100 nA cm-l; curves (2) and (3), 50 nA cm-l; curve (9, 10 nA cm-l. the residual nitrite. Typical examples of such voltammograms are shown in fig.3. After a certain time a ‘steady state’ was reached in which the NO; consumed by the electrolysis process was balanced by the NO; produced by thermal decomposition and by the NO; diffusing from the fritted counter-electrode compartment. The ultimate NO; concentration was found to be dependent on the melt temperature, e.g. the higher the temperature the higher the residual NO; concentration. For temperatures around the melting point (ca. 500 K) the NO; content could be lowered below the detection limit of the d.p.v. technique, while for a temperature around 600 K the residual NO; concentration was only of the order of In all the experiments concerning the decomposition of NO; the melt was maintained under an oxygen atmosphere (with vigorous and continuous purging of wet O,,po, = 745 Torr, pH,O = 15 Torr) in order to calculate the equilibrium constant of reaction (1) and to compare it with previous data16 obtained by an ‘indirect’ met hod.mol kg-l. KINETICS AND THERMODYNAMICS OF REACTION (1) Once the purification had been accomplished the kinetics of NO; production were followed by d.p. voltammetry. After a certain time, which varied from several days to several weeks depending on the melt temperature and on the ‘initial’ NO;F. PALMISANO, L. SABBATINI AND P. G. ZAMBONIN 10 - I M Y, - & In I . 13 3 g 5 - I 0 . 1035 - ..-..* 0 . . 0 . . . . & .. . . 0 . . . +. f .# I I I I 0 5 10 15 20 time/days Fig. 4. Plot of [NO;] against time at T = 538 K. concentration, the oxidation current of nitrite ions attained a ' steady-state ' value, which changed if the composition of the atmosphere over the melt was changed. For instance, if the oxygen partial pressure over the melt was lowered from ca.1 atm to 0.2 atm, a new plateau was reached, thus indicating that the steady-state situation was a true equilibrium. The steady-state NO; concentration was determined by the standard addition method, in order to eliminate uncertainties in the electrode area and the diffusion coefficient of NO;. This value was used in the calculation of the equilibrium constant for reaction (1) and also permitted the conversion of the peak current against time data to a set of [NO;] against time data. A typical plot obtained at 538 K is shown in fig. 4. The linear relationship between NO; concentration and time indicates that the main decomposition process is actually reaction (l), which appears as a pseudo-zero-order kinetic process.The relevant kinetic constant was evaluated as 7.6 x mol kg-l s-l at 538 K. Similar kinetic measurements were performed in the temperature range 510-620 K. Additional experiments in which the NO; formation rate was followed by the ex situ spectrophotometric method were performed at temperatures > 580 K where the NO, concentrations were in the dynamic range of this technique. Results agreed well with those obtained by d.p.v. experiments. From a log K against 1/T plot in the above temperature range an apparent activation energy of 39 2 kJ mol-1 was calculated. A simple mechanism which can be proposed on the basis of the observed kinetic behaviour is the following: slow NO;---, NO;+O (a) fast go+o-+o,) NO, = NO;+&O,.Alternative mechanisms involving species such as 0;- and O;, which are known1' to catalyse reaction (l), are unlikely since the catalysed reaction, which parallels the uncatalysed one, could predominate only at catalyst concentrations several orders of1036 THERMAL DECOMPOSITION OF NITRATE MELTS magnitude higher than those which can be calculated (on the basis of the known24 thermodynamic data) under the present experimental conditions (water concentration in the range 10-3-10-2 mol kg-l). Recently Br~oker,~ found that in a mixture of fused KN1603 and KN180, species such as N180i60- and N1s0160; are formed with the l 8 0 , l60 interchange occurring concurrently with the decomposition of the nitrate melt. This finding reduces the role of the simple autodissociation (6) NO; = NO: + 0’- of the melt anions; at the same time the possibility of melt decomposition (and isotope exchange) via reactions such as those proposed by Zambonin26 in which 0,- and 0;- are the ‘active species’, is also greatly reduced since, as specified above, the concentrations of these oxide species are exceptionally low.It follows that NO; arises predominantly by thermally activated decomposition of the melt anions and that the more obvious active species (which accounts for the isotopic exchange) is another NO; ion. Moreover, the quite low value found for the apparent activation energy suggests that the NO bond in the NO; ion is broken after a certain number of activation steps so that sufficient energy is accumulated in the NO bond to permit its disruption.The simplest way to represent this situation is NO; +NO; e (NO;); + NO; + NO;* (4 slow NO;* ---+ NO; + 0 fast go +o- 0,) NO, = NO,+BO, where NO;* represents a critically energized molecule, derived from an activated complex such as (NO;):, which can undergo the decomposition step. The relatively fast rate of isotopic exchange found by Brooker leads us to consider step (c) as an equilibrium prior to step (d), which is the rate-determining step. (This assumption can also rationalize the observed pseudo-zero-order kinetic law.) Of course, in this case the measured kinetic constant is not ‘purely kinetic’, since it contains the equilibrium constant for step (c).The log Kes against 1 /Tplot of the experimental equilibrium constants for reaction (1) obtained ‘directly’ from the experimental findings of the present work is reported in fig. 5 and compared with literature data. While the slight discrepancy with the results given in ref. (16) can be ascribed to the ‘indirect method’ (involving an approximate AH value for the reaction 0, + 0;- + 2 0;) previously used for calculating the equilibrium constants of reaction ( l ) , the data in ref. (14) appear to be unreliable. The anomalous results obtained in ref. (14) could be ascribed to imperfect control of the experimental conditions, especially the (unmentioned) purification of the melt. In the absence of a suitable purification procedure the melt could be buffered in NO; and would reach a true equilibrium only at extremely long times, since the reverse of reaction ( I ) is very slow in the absence of catalysts.F. PALMISANO, L.SABBATINI AND P. G. ZAMBONIN 1037 1.4 1.6 1.8 2.0 lo3 KIT Fig. 5. Comparison between the equilibrium constants for reaction (1) found in the present work [curve (l)] and those from ref. (16) [curve (2)] and ref. (14) [curve (3)]. We thank Mr S. Giacummo for technical assistance. This work was carried out partly with the financial support of the Italian National Research Council (C.N.R. - Roma) and partly with that of the Minister0 Pubblica Istruzione. W. E. Kirst, W. M. Nagle and J. B. Castner, A. I. Chem. Trans., 1941, 36, 371. H. P. Vomich and V. W. Uhl, Chem. Eng., 1963, 70, 135.P. Lloyd and E. A. C. Chamberlain, J. Iron Steel Inst., London, 1945, 142, 1 1 . R. J. Box and B. A. Middleton, J. Iron Steel Inst., London, 1945, 151, 71. L. N. Tallerico, Sandia Laboratories report SAND, 1979, 79-8015. M. C. Silverman and J. R. Engel, Oak Ridge National Laboratories Report ORNL/TM - 1977 - 5682. ' L. Radoseviche, Sandia Laboratories report SAND, 1978, 78-8221. P. G. Zambonin, Anal. Chem., 1971, 43, 1571. P. G. Zambonin, E. Desimoni, F. Palmisano and L. Sabbatini, in Ionic Liquids, ed. D. Lovering and D. Inman (Plenum Press, London, 1980), p. 249. lo K. H. Stern, J. Phys. Chem. Re$ Data, 1972, 1, 747. l 1 E. S. Freeman, J. Phys. Chem., 1956, 60, 1487. l 2 E. S. Freeman, J. Am. Chem. Soc., 1957, 79, 838. l 3 R. F. Bartolomew, J. Phys. Chem., 1966, 70, 3442. l 4 R. N. Kust and J. D. Burke, Inorg. Nucl. Chem. Lett., 1970, 6, 333. l5 A. F. J. Goeting and J. A. A. Katelaar, Electrochim. Acta, 1974, 19, 267. l6 F. Paniccia and f. G. Zambonin, J. Chem. Soc., Faraday Trans. 1, 1976, 72, 1512. l8 E. Desimoni, F. Palmisano and P. G. Zambonin, J. Electroanal. Chem., 1977, 84, 3 15. l9 F. Palmisano, L. Sabbatini, E. Desimoni and P. G. Zambonin, J. Electroanal. Chem., 1978, 89, 31 1 . 2o E. P. Parry and R. A. Osteryoung, Anal. Chem., 1965, 37, 1634. F. Paniccia and P. G. Zambonin, J. Phys. Chem., 1973, 77, 1810.1038 THERMAL DECOMPOSITION OF NITRATE MELTS 21 D. J. Myers, R. A. Osteryoung and J. Osteryoung, Anal. Chem., 1974, 46, 2089. 2 2 H. S. Swofford and P. G. MacCormick, Anal. Chem., 1965, 37, 970. p 3 L. E. Topol, R. A. Osteryoung and J. H. Christie, J. Phys. Chern., 1966, 70, 2837. 24 P. G . Zambonin, J. Electroanal. Chem.. 1970, 24, 365. 25 M. H. Brooker, J. Electrochem. Soc., 1979, 126, 2095. 26 P. G . Zambonin, J. Electroanal. Chem., 1971, 33, 243. (PAPER 3/595)

ISSN:0300-9599    

DOI:10.1039/F19848001029

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

J . Chem. SOC., Faraday Trans. I , 1984, 80, 1039-1047 Reaction of Adsorbed Carbon Monoxide with Hydrogen on Magnesium Oxide BY GONG-WEI WANG AND HIDESHI HATTORI* Department of Chemistry, Faculty of Science, Hokkaido University, Sapporo 060, Japan Received 3rd May, 1983 The reactivity of carbon monoxide adsorbed on magnesium oxide surfaces towards hydrogen has been examined by the temperature-programmed desorption (t.p.d.) technique and by i.r. spectroscopy. The surface species resulting from the reaction of adsorbed CO with H, in the temperature range 343-583 K gave t.p.d. peaks for CO and H, at 773 K. The t.p.d. profiles were in agreement with that of adsorbed formaldehyde but were totally different from that of adsorbed formic acid. The surface species showed i.r. features in the C-H stretching region and in the region 1700-1000 cm-l.One set of the features in the latter region appearing at 1604 and 1370 cm-I could be attributed to forrnate. Based on the t.p.d. profiles and i.r. spectra it is concluded that CO adsorbed on the MgO surface reacts with H, to form formyl (HCO) which is then adsorbed on the surface oxygen atoms to give formate (HCOO). The formation of various types of adsorbed CO on well degassed MgO surfaces has been observed by several groups.1-6 The proposed surface species include anions such as [O-C-C-O]2-, [(cO)6l2- and (CO);-. It has also been reported that H, is dissociatively adsorbed on the MgO surface to form H+ and H- pair~.~-lO Because of the anionic nature of the adsorbed CO it is expected that the latter may react with a proton on the MgO surface.We have reported preliminary results of the reaction of adsorbed CO with H2.11 In the present work a more detailed study of the reactivity of CO adsorbed on the MgO surface towards hydrogen has been carried out by the temperature-programmed desorption (t.p.d.) technique and by i.r. spectroscopy. It is suggested that adsorbed CO reacts with H, to form formyl groups on the surface which are then adsorbed onto surface oxygen atoms. EXPERIMENTAL MATERIALS Magnesium oxide was prepared from Mg(OH), (Kanto Chemical Co. Ltd) by decomposition at I273 K for 2 h under a vacuum. Cylinder carbon monoxide was purified by passage through well degassed MgO kept at liquid-nitrogen temperature followed by contact with zinc vapour.Formaldehyde was prepared by decomposition of paraformaldehyde at 373 K, and purified by passage through a trap kept at 195 K. Formic acid was purified by distillation over boric anhydride under reduced pressure. TEMPERATURE-PROGRAMMED DESORPTION PROCEDURE Magnesium hydroxide (0.50 g) was decomposed in an adsorption vessel to give MgO. After cooling to 273 K, the MgO sample was exposed to 2660 Pa of CO for 30 min followed by outgassing for 30 min at 273 K. The sample was then exposed to ca. 4 kPa of H, and was heated at different temperatures for 30 min. The sample was cooled to 273 K, outgassed for 30 min, 1039REACTION OF CO+H, ON MgO 300 400 500 600 700 800 900 1000 1100 desorption temperature/K Fig. 1. T.p.d. profiles from the surface species formed on MgO by the reaction of adsorbed CO with H, at different temperatures.CO: A, 343; 0 , 4 2 3 and 0, 483 K; CO,: A, 343; 0, 423 and ., 483 K; H, (shown in the inset): 0, 423 and 0, 483 K. The dotted line denotes the evolution of CO when H, was not admitted. and subjected to a t.p.d. run. The t.p.d. profiles from adsorbed formaldehyde and adsorbed formic acid were also measured. The adsorptions of formaldehyde (ca. 2 Pa) and of formic acid (ca. 25 or 100 Pa) were carried out at room temperature. Desorbed gases were analysed by mass spectrometry. The spectrum was recorded on a Neva NAG-515 mass filter. To normalize the sensitivity of the mass spectrometer, which varied with pressure, a small amount of Ar was constantly leaked into the system (ca. 0.1 Pa). The relative amounts of desorbed gases were determined as the peak height relative to that of Ar.T.p.d. was performed at a rate of temperature increase of 10 K min-l. I.R. SPECTROSCOPY Magnesium hydroxide was pressed into a disc and outgassed at 1273 K in a quartz i.r. cell with CaF, windows. The adsorption procedures were the same as those for t.p.d. except that adsorption and outgassing of CO were performed at room temperature instead of at 273 K. In some experiments a mixture of CO and H, was allowed to react at elevated temperatures and then the sample was cooled to room temperature and outgassed for 30min. In other experiments D, was used instead of H,. All spectra were recorded at room temperature on a Jasco 701 G spectrometer. RESULTS T.P.D. Fig. 1 shows the t.p.d.profiles obtained when the adsorbed CO was exposed to H, at different temperatures. Without addition of H,, the profile contained two peaksG-W. WANG AND H. HATTORI 1041 300 400 500 600 700 800 900 desorption temperature/K Fig. 2. T.p.d. profiles from formaldehyde adsorbed on MgO: 0, CO and 0, H,. for CO at 438 and 568 K. On exposure to H,, these two peaks decreased and a new peak for CO appeared at 773 K. The new peak for CO was accompanied by the peak for H,. In addition, a peak for CO, appeared at ca. 943 K. Fig. 2 shows the t.p.d. profiles obtained when ca. 2 Pa of formaldehyde was adsorbed on the MgO sample. Three peaks for CO appeared at 383, 573 and 788 K. The peak for CO at 788 K was accompanied by the peak for H,. Fig. 3(A) shows the t.p.d. profiles obtained when ca.25 Pa of formic acid was adsorbed on the MgO sample. The main peak was that for CO, which appeared at ca. 600 K. A small peak for H,O appeared above 1000 K. The t.p.d. profiles varied as the amount of adsorbed formic acid increased. As shown in fig. 3(B), when ca. 100 Pa of formic acid was adsorbed, the peak for H,O became almost as large as that for CO. I.R. SPECTROSCOPY A disc of Mg(OH), was outgassed at 1273 K and cooled to room temperature. The sample was exposed to 2660 Pa of CO for 30 min and then outgassed for 30 min. Following i.r. measurements, the sample was exposed to ca. 4 kPa of H, (or D,) and heated at 553 K for 30 min. The sample was then cooled to room temperature and outgassed for 30 min. The spectra obtained are shown in fig.4. On adsorption of CO on the MgO sample, several absorption bands appeared [fig. 4(b)]. Main features were observed at 2106,2093, 1564, 1469, 1454, 1374, 1352, 1320, 1256, 1177 and 1163 cm-l. These features are essentially the same as those reported by Guglielminotti et aZ.3 As the adsorbed CO was exposed to H, (or D,) at 553 K, the features for the adsorbed CO disappeared and new features appeared at 2930,2840, 1642, 1604, 1496, 1370, 1332,1250 and 1224 cm-l [fig. 4(c)]; for D, features appeared at 2170, 2130, 2060, 1635, 1600, 1471, 1370, 1332 and 1265 cm-l, [fig. 4(d)]. Fig. 5 shows the i.r. spectra obtained when a mixture of CO (2.8 kPa) and H, (8.0 kPa) or D, (8.1 kPa) was reacted at 553 K on the MgO sample followed by outgassing at room temperature. The spectra were essentially the same as those 35 FAR 11042 .0.6 0.5 0.4 0.3 0.2 -3 2 0 . 1 a 0 Y REACTION OF CO + H, ON MgO A 300 400 500 600 700 800 900 1000 1100 desorption temperature/K Fig. 3. T.p.d. profiles from formic acid adsorbed on MgO at adsorption pressures of (A) 25 and (B) 100 Pa. 0, CO; A, H,O; 0, H, and 0, CO,. obtained by the reaction of adsorbed CO with H, (or D2) although the features were intensified. In particular, the features in the region of C-H (or C-D) stretching became distinct. After the sample had been outgassed at 773 K only two features remained, those at 1635 and 1332 cm-l. Fig. 6 shows the spectra obtained when formaldehyde was adsorbed at different temperatures. At room temperature the features at 2933 and 2838 cm-l were observed in the C-H stretching region.In the region 1200-1700 cm-l features at 1602, 1382, 1368 and 1335 cm-l were observed. On heating the sample at 573 K for 15 min andG-W. WANG AND H. HATTORI 1043 I I I I I I I I I U . I . 1 l I & ‘ “ ‘ I 4000 3600 3200 2800 2400 2000 1600 1500 1400 1300 1200 1100 wavenumber/cm -’ Fig. 4.1.r. spectra of the surface species on MgO. (a) background, (b) adsorbed CO, (c) after reaction of adsorbed CO with H, at 458 K and ( d ) after reaction of adsorbed CO with D, at 458 K. wavenumber/cm -’ Fig. 5. 1.r. spectra of the surface species on MgO. (a) after reaction of a mixture of CO and D, at 458 K, (b) after reaction of a mixture of CO and H, at 458 K and (c) as (b) after evacuation at 773 K. 35-21044 REACTION OF co + H, ON MgO i 2*0° 2860 I I I I I I I I I 4000 3600 3200 2800 2400 2 ( 1635 1 1600 1500 1400 1300 1200 1100 wavenumber/cm Fig.6. 1.r. spectra of formaldehyde adsorbed on MgO at (a) room temperature and (b) 573 K. cooling to room temperature followed by outgassing for 15 min, various spectral changes were observed. The features in the C-H stretching region were distinctly separated. The features at 1602, 1382 and 1368 cm-l decreased in intensity, and a feature at 1332 cm-l developed. DISCUSSION Appearance of new peaks for CO and H, in the t.p.d. profiles at ca. 773 K upon exposure to H, of CO adsorbed on MgO indicates that the adsorbed CO reacts with H, to form a surface species. When formaldehyde was adsorbed on the MgO sample one of the t.p.d. peaks for CO appeared at ca.773 K, accompanied by the evolution of H,. Therefore it is suggested that the surface species resulting from the reaction of adsorbed CO with H, are the same as those formed on adsorption of formaldehyde. It seems likely that the surface species are formyl groups (HCO). The surface species decomposed into CO and H, at ca. 773 K and then CO and H, were desorbed. The interaction of CO with H, on oxide surfaces has been reported for ZnO and Al,O,. Boccuzzi et aZ.12 reported that adsorbed CO strongly interacts with adsorbed H on the ZnO surface, and that an interaction occurred between H adsorbed on a given Zn ion and CO adsorbed on the next nearest Zn. However, the formation of a bond between C and H was not reported. Saussey et aZ.13 recently reported that surface formyl species are formed on ZnO by the reaction of CO with H,.Using i.r. spectroscopy, Dalla Betta and SheleP4 observed the formation of formate ions during the reaction of CO with H, on Ru/Al,O,, in which formate ions were formed on the support (A1,0,) and not on the metallic part (Ru). The i.r. spectra of the species which resulted from the reaction of CO with H, on MgO demonstrates the formation of a C-H bond, suggesting that possibly the surface species formed is the formyl group (HCO). However, the features of the i.r.G-W. WANG AND H. HATTORI 1045 Table 1. Fundamental vibration spectrum of the formate ion assignment frequencya/cm-l C-H stretch v(CH), v, symmetric OCO stretch v(CO),, v, symmetric OCO deformation S(OCO),, v, a.symmetric OCO stretch v(CO),,, v, asymmetric OCO deformation S(OCO),,, v5 280 1 1370 774 1592 1381 a The frequencies are those observed for HC0ORb.l’ Table 2.Assignments of the features observed for the surface species resulting from the reaction with H, of CO absorbed on MgO frequency/cm-l type of species assignment C-H stretch asymmetric OCO stretch symmetric OCO stretch C-0 stretch asymmetric OCO stretch asymmetric OCO stretcha 2840 1604} formate 1370 1642) bidentate carbonate 1332 1496 unidentate carbonate a Tentative assignment. spectra in the range 1700-1000 cm-l do not show the existence of formyl groups in a simple form. The i.r. spectra of the compounds containing HCO groups show C-H vibrational features at relatively low frequencies, near 2700 cm-I. However, no distinct features in this range were observed for the surface species resulting from the reaction of CO with H, on MgO.Adsorption of formaldehyde on MgO did not give the C-H vibrational features characteristic of the formyl group either. Yates and Cavanagh15 reported that formaldehyde adsorbed on A1,0, is converted into formate as the temperature is raised to 300 K. It is suggested that formaldehyde adsorbed on MgO is also converted to formate above room temperature. The generally accepted assignments of the features of formate ions are given in table 1. One set of features (at 2840, 1602 and 1370 cm-l) observed for the surface species resulting from the reaction of adsorbed CO with H, is essentially the same as that corresponding to formate ions. The assignments of these features are summarized in table 2 and shown in fig.5 . It is thus suggested that the surface species are present in the form of formate ions rather than formyl groups. There are three modes by which formate can coordinate to surface atoms as shown below (M is a surface atom).I6 Y H H C C o/---\o 04-+o I \ / 0 M M M M1046 REACTION OF CO+H, ON MgO One of the criteria used to distinguish the types of coordination is the separation between the v(CO),, and the v(CO), vibrations.18 The separation is largest for the unidentate formate (I), but there is no spectral criterion for distinguishing between type (11) and (111) coordinations. The other criterion to distinguish between the types of coordination is the direction of shift in the frequencies v(CO),, and v(CO),.~~ Only v(CO),, increases as the M-0 bond strength increases for type I coordination, while both frequencies increase for type (11) and (111) coordinations.The frequencies observed for the surface species resulting from the reaction of adsorbed CO with H, are consistent with formate of either type (11) or (111). The formation of the surface species from the reaction of adsorbed CO with H, is postulated as follows: H I 0 =.,c y- H2 + CO - I ;I ’ Ic01;- H’ H- -Mg- L M g - -Mg-0-Mg- The adsorbed CO may be a negatively charged dimeric or hexameric species which reacts with a proton resulting from the heterolytically dissociative adsorption of H,. The resulting surface species (IV) is in the form of a type (11) formate. The surface species (IV) is different from the formate which is expected to form on adsorption of formic acid as follows.H Surface species (V) is the formate adsorbed on the Mg2+ ion, while surface species (IV) is the formyl group adsorbed on the 0,- ion to take the form of the formate. Species (V) decomposed at ca. 600 K to give mostly H,O and CO and partly H, and CO,. Surface species (IV) decomposed to give H, and CO at ca. 773 K. In addition to the features characteristic of the formate, three features were observed at 1635, 1496 and 1332 cm-l. The features at 1635 and 1332 cm-l persisted on evacuation at 773 K. These could be attributed to a carbonate, which would be desorbed at ca. 943 K in t.p.d. The separation of the features arising from C-0 stretching distinguishes the modes of coordination, e.g.free carbonate ion [C0,l2-, C = 0 and bridged unidentate carbonate M-0-C carbonate ,C = O.la There would be no separation for the free ion, and the separation would be ca. 100 cm-l for the unidentate species and 200-300 cm-l for the bidentate species and the bridged carbonate. Based on the separation value, the type of carbonate observed is estimated to be in the form of the bidentate or bridged carbonate. Assignment of the features which appeared at 1490 cm-l on the reaction of CO with H, and at 1470cm-l on the reaction of CO with D, is uncertain. These features 00 ’0‘ ‘0’ bidentate carbonate M ‘0’ M-O\ M-0G-W. WANG AND H. HATTORI 1047 diminished on evacuation at 773 K. One possibility is the asymmetric 0-C-0 stretching of the unidentate carbonate, which normally shows a feature at ca.1480 cm-l. Because of the shift in the feature on replacing H by D, the species should involve an H (or D) atom. In the form of a bicarbonate, HO-COT, the asymmetric 0-C-0 stretch which appears at 1655 cm-l shifts to 1640 cm-l on replacing H by D.19 The feature at 1490 cm-1 may be attributed to a surface species of the following structure. In the C-H stretching C-D) four features may region (2800-3000 cm-l for C-H, 2000-2200 cm-l for be observed. If the surface formate were the only type of surface species containing a C-H bond present, only one C-H stretching feature would appear. Therefore, surface species other than formate should exist. However, their identification was not successful. J. H. Lunsford and P. J. Jayne, J. Chem.Phys., 1966,44, 1492. R. M. Moms, R. A. Kaba, T. G. Grosherns, K. J. Klabunde, R. J. Baltisberger, N. F. Woolsey and V. I. Stenberg, J. Am. Chem. Soc., 1980, 102, 3419. E. Guglielminotti, S. Coluccia, E. Garrone, L. Cerruti and A. Zecchina, J. Chem. Soc., Faraday Trans. 1. 1979, 75, 96. R. St. C. Smart, T. L. Slager, L. H. Little and R. G. Greenlar, J. Phys. Chem., 1973, 77, 1019. A. Zecchina, M. G. Lofthouse and F. S . Stone, J. Chem. SOC., Faraday Trans. I , 1975, 71, 1476. G. Wang, H. Itoh, H. Hattori and K. Tanabe, J. Chem. SOC., Faraday Trans. I , 1983, 79, 1373. S . Coluccia and A. J. Tench, Proc. 7th Int. Congr. Catal., Tokyo, 1980 (Kodansha, Tokyo and Elsevier, Amsterdam, 198 l), part B, p. 1 154. S. Coluccia, F. Boccuzzi, G. Ghiotti and C. Morterra, J. Chem. SOC., Faraday Trans. I , 1982,78,2111. @ Y. Tanaka, Y. Imizu, H. Hattori and K. Tanabe, Proc. 7th Int. Congr. Catal., Tokyo, 1980 (Kodansha, Tokyo and Elsevier, Amsterdam, 1981), part B, p. 1254. lo T. Ito, T. Sekino, N. Moriai and T. Tokuda, J. Chem. Soc., Faraday Trans. I , 1981, 77, 2181. l 1 G. Wang, H. Hattori, H. Itoh and K. Tanabe, J. Chem. Soc., Chem. Commun., 1982, 1256. l 2 F. Boccuzzi, E. Garrone, A. Zecchina, A. Bossi and M. Camia, J. Catal., 1978, 51, 160. l 3 J. Saussey, J. C. Lavalley, J. Lamotte and T. Rais, J. Chem. Soc., Chem. Commun., 1982, 278. l4 R. A. Dalla Betta and M. Shelef, J. Catal., 1977, 48, 11 1. l5 J. T. Yates Jr and R. R. Cavanagh, J. Catal., 1982, 74, 97. l6 K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Compounh (Wiley- l7 J. D. Donaldson, J. F. Knifton and S. D. Ross, Spectrochim. Acta, 1964, 20, 847. lo J. V. Evans and T. L. Whateley, Trans. Faraday Soc.. 1967,63, 2769. Interscience, New York, 3rd edn, 1978). L. H. Little, Infrared Spectra of Ahorbed Species (Academic Press, New York, 1966). (PAPER 3/684)

ISSN:0300-9599    

DOI:10.1039/F19848001039

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

J . Chem. SOC., Faraday Trans. I, 1984, 80, 1049-1057 Properties of a Liquid/Solid Interfacial Phase at High Pressures BY SENTARO OZAWA, MASAYUKI GOTO, KI-ICHI KIMURA AND YOSHISADA OGINO* Department of Chemical Engineering, Faculty of Engineering, Tohoku University, Aramaki Aoba, Sendai 980, Japan Received 16th May, 1983 A negative apparent adsorption onto activated charcoal has been observed for methyl alcohol, ethyl alcohol, n-propyl alcohol, n-heptane, methylcyclopentane and methylcyclohexane under pressures of 0.1-196.2 MPa and at temperatures of 298.2-348.2 K. The adsorption isotherm exhibits a straight line with a negative slope over a wide range of the liquid-phase density, indicating that the volume and the density of the adsorbed phase do not vary with pressure. The characteristic behaviour of the adsorption isotherms is interpreted as an indication that the adsorbed phase includes void spaces resulting from the differences in size among the helium molecule, an admolecule and an adsorbent pore.The linear parts of the adsorption isotherms have successfully been generalized, indicating the validity of the interpre- tation of the experimental results. Data on the pressure dependence of the interfacial tension are also presented. Although much information about the adsorption of gases on solid surfaces has been acc~mulated,~-~ little is known about adsorption of liquids on solids at high pressures.8 Experimental difficulties in measuring accurately the degree of adsorption from a liquid phase onto a solid adsorbent increases as the pressure of adsorption increases, and this has presumably led to the scarcity of high-pressure data on the adsorption of liquids.We have speculated that a liquid in contact with an adsorbent would exhibit p-V-T behaviour different from that of the pure liquid, and that this difference would enable one to evaluate the degree of adsorption. Since the pV-T relation can be measured accurately for a liquid, this method appears to be worth examining. This paper gives details of the experimental method of pV-T measurements for the liquid/solid adsorption system and then presents the data obtained. Characteristic properties of the adsorbed phase are discussed leading to a possible model of adsorption. EXPERIMENTAL PRINCIPLES The method of measuring apparent adsorption in a liquid/solid adsorption system is essentially the same as that for measuring the specific volume of a liquid.The liquid/solid system is regarded as a hypothetical fluid with an apparent specific volume (V,,,) which is measured experimentally. This apparent specific volume defines the specific volume of the adsorbent (V,) by the following formula: Cpp = [( W P J + w, V,M % + w,) (1) where 4 is the mass of the liquid in contact with the solid adsorbent of mass W, and p1 is the bulk density of the liquid. The specific volume V, thus defined is different from the specific 10491050 LIQUID/SOLID INTERFACE AT HIGH PRESSURES Fig. 1. Glass (Pyrex) piezometer for measuring the apparent specific volume of a liquidlsolid system: (A) solid adsorbent, (B) liquid adsorbate, (C) mercury and (D) glass float.volume VHe measured with helium, because the liquid used for measuring V, interacts with the surface of the solid adsorbent and so V, is an apparent volume which is affected by adsorption phenomena . The apparent adsorption is defined as the difference between the mass of the liquid introduced into the adsorption system and the mass of the liquid remaining in the dead space of the system (the volume not occupied by the adsorbent). Thus rl is expressed by r, = [wl-(h- VHe w,)pll/K v, = ( W P J + w, V,. r, = Pl(VHe - V,). (2) (3) (4) where V, is the total pyknometric volume of the adsorption system and is given by Combining eqn (2) and (3) gives Values of p1 and V,, can be measured experimentally (the present work assumes that V,, is independent of pressure).The value of V, can also be obtained from the experimental value of KPp using eqn (1). Thus one can deduce the value of r, from p V-T data for the liquid/solid system and from those for the pure liquid measured under the same conditions. MATERIALS The liquids used for the adsorption study were guaranteed reagents supplied by Wako Pure Chemical Industries Ltd. Some of them were purified further, and all purities were determined1.3 - I M m 5 kn -- 1.2 k- 1.1 0 S. OZAWA, M. GOTO, K. KIMURA AND Y. WINO ( a ) > I ( b ) X 1051 1.5 - I M m 6 A* 1.4 -- k- 1.3 100 200 100 200 P IMPa P lMPa Fig. 2. Comparison between specific volume of a pure liquid (---) and partial specific volume of the liquid contacting with adsorbent (-) for (a) MeOH and (b) n-C,: (i) 298.2, (ii) 323.2 and (iii) 348.2 K.gas-chromatographically as follows: methyl alcohol (MeOH; dried with CaO and purified to better than 99.6% by distillation), ethyl alcohol (EtOH; dried with molecular sieve 4A and purified to better than 99.9% by vacuum distillation), n-propyl alcohol (PrOH; dried with molecular sieve 4A and purified to better than 99.9% by vacuum distillation), n-heptane (n-C,; 99.973, methylcyclopentane (MCP; dried with molecular sieve 5A and purified to better than 99.9% by vacuum distillation), and methylcyclohexane (MCH; better than 99.9%). Activated charcoal (AC; Futamura Chemical Industries Ltd, type CS-120A) was used as an adsorbent. The B.E.T. surface area of the adsorbent was calculated to be 1350 m2 g-' from nitrogen-adsorption data at liquid-nitrogen temperature.The micropore volume ( -= 30 nm) calculated from nitrogen adsorption at a relative pressure of 0.9679 was 0.63 cm3 g-l. APPARATUS Details of the high-pressure system and the piezometer for measuring the specific volume of a liquid have already been reportedlo and therefore are not given here. To measure cPp the piezometer shown in fig. 1 was inserted into the high-pressure cell.lo The change in volume of the liquid contained in the piezometer and in contact with the solid adsorbent was converted into the change in position of a magnetic tip (not shown in fig. 1) attached to the top of a glass float laid on the mercury surface which separated the pressure-transmitting liquid (isopropyl alcohol) and the liquid kept in the piezometer.The position of the magnetic tip inside the high-pressure cell was measured by a differential transformer device.1° The reproducibility of the specific-volume measurements was better than - +o. 1 %. PROCEDURES Procedures for measuring cPp are as follows, and those for measuring p 1 have been reported elsewhere.lO Ca. 4 g of activated charcoal (AC) were placed in the piezometer and degassed in a vacuum (ca. 0.13 Pa) at 300 "C for 3 h. After cooling, the weight of adsorbent was accurately determined. The adsorbent was degassed again under the same conditions as above and cooled1052 LIQUID/SOLID INTERFACE AT HIGH PRESSURES 0.65 0.70 0.75 pllg cm-3 Fig. 3. Typical adsorption isotherms showing negative adsorption for (a) MeOH/AC and (b) n-C,/AC: 0, 298.2, a, 323.2 and 0, 348.2 K.in a vacuum. The piezometer containing the adsorbent was filled with the desired liquid without exposing the adsorbent to the ambient atmosphere. The piezometer was detached from the vacuum line and a predetermined amount of mercury (purified by vacuum distillation) was poured into it. The piezometer with its contents was weighed accurately and then inserted into the high-pressure cell. The amount of liquid adsorbate was obtained by difference. The specific-volume measurement was started 1 h after the high-pressure cell had reached isothermal conditions ( 0.025 K). The pressure was kept constant for 20 min, during which the displacement of the differential transformer from a stationary point to the point where it detected the magnetic tip was read twice.The measurement was carried out initially by increasing the pressure and then by decreasing the pressure. Thus four separate values of the specific volume cpp were obtained at a given pressure, serving to confirm the reproducibility of the measurement.S. OZAWA, M. GOTO, K. KIMURA AND Y. WINO 1053 RESULTS AND DISCUSSION PARTIAL SPECIFIC VOLUME OF THE LIQUID ADSORBATE With values of %pp and c(G at 0.1 MPa) one can evaluate the partial specific volume (V,) of the liquid in contact with the adsorbent using the following relation : The use of Vp instead of V,,, is convenient in order to show clearly the difference in volumetric behaviour between the pure liquid and that in contact with the adsorbent, as illustrated in fig.2 for MeOH and n-C,. The figure shows that the partial specific volume (V,, solid lines) is larger than the specific volume of the pure liquid (6, dotted lines) at any temperature or pressure used in the present work for both MeOH and n-C,. Other adsorption systems (EtOH/AC, PrOH/AC, MCP/AC and MCH/AC) showed specific-volume isotherms similar to those in fig. 2. The fact that the partial specific volume of the liquid in contact with the adsorbent is different from that of the pure liquid strongly suggests that the former consists of two different phases, a bulk phase and an adsorbed phase. Obviously the specific volume of the bulk phase should be identical to that of the pure liquid. On the other hand, the specific volume of the adsorbed phase can take a value different from that of the pure liquid.APPARENT ADSORPTION Illustrated in fig. 3(a) are the isotherms of apparent adsorption (r,) for the MeOH/C system. The bulk density (pl) of MeOH has been used as the abscissa rather than the pressure. The apparent adsorption isotherms for n-C, are similar to those of MeOH, as shown in fig. 3(b). The isotherms shown in fig. 3(a) and (b) indicate that the apparent adsorption (r,) is negative and decreases almost linearly with increase in the density of the bulk phase. However, the linear correlation breaks down as p1 approaches higher values. Isotherms similar to those shown in fig. 3 were also observed for other adsorption systems (EtOH/AC, PrOH/AC, MCP/AC and MCH/AC). The temperature dependence of the adsorption isotherm was most marked for the n-C,/AC system.On the other hand, PrOH/AC and MCH/AC showed almost no temperature dependence and the other systems showed only a small temperature dependence. VOLUME AND DENSITY OF THE ADSORBED PHASE It is well known3p11*12 that apparent adsorption (r,) is related to absolute adsorption (rt, the amount of adsorbate entrapped in the adsorbed phase) by the following formula : (6) where ps and Kds are the density and the volume of the adsorbed phase, respectively. At high densities of the bulk phase the absolute adsorption probably takes on its rl = rt(l -Pl/Ps) = &dsPs(l -Pl/ps) saturation value, i.e. = &ds& = constant. (7) With the assumption that Kds is independent of pl, differentiation of eqn (6) with respect to p1 gives This equation explains the linear adsorption isotherms with negative slopes observed experimentally.The values of Qds obtained from the slopes of the isotherms were (aq/apl)T = - Kds. (8)1054 LIQUID/SOLID INTERFACE AT HIGH PRESSURES helium molecu i - le 0 4 ( a ) ( b ) Fig. 4. Possible models for the adsorbed phase. found to depend only slightly on the liquid species used and on the adsorption temperature adopted in the present work, i.e. Gds = 0.19 cm3 g-l on average, The discussion above indicates that the density of the adsorbed phase does not vary on compressing the adsorption system in the range where the adsorption isotherm exhibits a linear dependence upon the bulk density (PI), i.e. (aps/3p)T = 0. On the other hand, the density of the bulk liquid increases on compression, i.e.(i3p1/8p)T > 0. Thus one can write ( a P I / m T ’ ( W @ 4 T . (9) Ps < P1 (10) In addition, the density of the adsorbed phase is restricted by the following relation: which is derived from eqn (6) by taking the negative adsorption (r, c 0) into account. The two relations mentioned above [eqn (9) and (lo)] state that the adsorbed phase is more rigid but less dense than the bulk phase. Usually a phase with a lower density is more compressible than a phase with a higher density. The adsorbed phase therefore appears anomalous. STRUCTURAL MODEL FOR THE ADSORBED PHASE The unusual volumetric behaviour of the adsorbed phase probably originates mainly from the difference between the molecular diameter of the adsorbate (4) and that of helium (tiHe) used to measure the volume of the dead space.Since dl > dHe, there is a void space accessible to a molecule of helium which would not be accessible to an adsorbate molecule. The magnitude of this void volume [ Vvoid = (dl - dHe)/2S where S is the specific surface area of the adsorbent] is appreciable and results in a small value of ps, even when the adsorbate molecules are closely packed in the adsorbent pore, as illustrated in fig. 4(a). In addition, another type of incompressible void space may appear when the close packing of admolecules is hindered by the limited size of the adsorbent pore, as shown in fig. 4(b). Thus one can derive a reasonable model for the adsorbed phase with a density satisfying the relations mentioned above [eqn (9) and (lo)].Do 4 W a M .A * 3 L- I 3 L- n O M 4 W a M n * . 3 -0.01 L- I 3 L- n -0.02 S . OZAWA, M. GOTO, K. KIMURA AND Y. OGINO 1 1055 0 OD 4 a M --- h -0.01 L L- I 3 L- n - 0.02 0 0.05 0.1 p10l)--P1(P*)k Fig. 5. Generalized adsorption isotherms at (a) 298.2, (b) 323.2 and (c) 348.2 K: 0, MCH; a, PrOH; 0, MeOH; A, EtOH; A , n-C, and A, MCP. POSSIBLE CHANGE IN THE ADSORBED PHASE The structure of the adsorbed phase will be preserved as long as the admolecules maintain their rigidity and the pore wall withstands the compression. Since the experimental data obtained on increasing the pressure were almost identical to those obtained on decreasing the pressure, the possibility that the pore structure disintegrates through compression can be ruled out. On the other hand, the assumption that the admolecules are rigid is not always valid, particularly at high pressures.Even a very small deformation of an individual admolecule could bring about a drastic cooperative phenomenon, viz. a phase change3 in the adsorbed phase. It is not unlikely that such void spaces as shown in fig. 4(b) are reduced by the phase change, and thus a denser adsorbed phase and a greater absolute adsorption result. This would lead to an increase in the apparent adsorption (r1), yielding an upward deviation of the1056 LIQUID/SOLID INTERFACE AT HIGH PRESSURES Table 1. Pressure dependences of the interfacial tension for several adsorption systems on activated charcoal at 0.1 MPa and 298.2 K adsorbed species (aa/dp),/ cm MeOH EtOH PrOH n-C, MCP MCH 1.57 1.39 2.16 0.96 3.1 1 3.79 7 * I pressure/MPa Fig.6. Pressure dependence of the interfacial tension difference a(p)-a(p*): 0, MCH; A, MCP; 0, PrOH; a, MeOH; A, EtOH and ., n-C,. adsorption isotherm from linearity. The non-linear parts of the adsorption isotherms shown in fig. 3 might suggest that such a phase change is actually occurring. GENERALIZED ADSORPTION ISOTHERM The relation given by eqn (8) enables us to generalize the adsorption isotherm. Thus integration of eqn (8) with the assumption that Gds is a common constant for every adsorption system gives U P ) - U P * ) = - F/,dsbl(P) ---Pl(P*)l (1 1)S. OZAWA, M. GOTO, K. KIMURA AND Y. WINO 1057 where p* is a reference pressure which in the present work is 0.1 MPa. This relation indicates that plot of [T,(p) - G(p*)] against [PI@) -pl(p*)] for all adsorption systems should fall on a single straight line with a negative slope - Kds and a zero intercept.Such plots are presented in fig. 5 , which shows that the generalization is partly accomplished. To improve the generalization would require detailed information about adsorption in the high-density region, where a phase change in the adsorbed phase presumably takes place. PRESSURE DEPENDENCE OF THE INTERFACIAL TENSION The Gibbs adsorption equation relates the surface excess (T,/SM where M is the molecular weight) to the interfacial tension (a) for any adsorption system with an inert solid adsorbent by (12) &ISM = - (1 / R T ) da/d In a where a is the activity of the bulk liquid, R is the gas constant and Tis the absolute temperature.On the other hand, the following relation holds under a constant temperature RTdlna = VLdp (13) where VL is the molar volume of the liquid. Thus one obtains from eqn (12) and (1 3) Since the values of l-,, F/; and S are known as a function of pressure, one can evaluate the pressure dependence of the interfacial tension using eqn (14) at any desired pressure below 200 MPa. Listed in table 1 are the values of (aa/ap), for the adsorption systems studied, and fig. 6 represents the variation of a(p) - a(p*) with pressure. The microscopic meaning of these interfacial-tension data is not clear, but they are nevertheless very useful in discussing the pressure dependence of a composite isotherm for adsorption onto activated charcoal from a binary solution consisting of any two species listed in the table. Details of the experimental results and a discussion of adsorption from solution at high pressures will be reported in the near future. P. G. Menon, Chem. Rev., 1968,68, 277. P. G. Menon, Advances in High Pressure Research (Academic Press, New York, 1969), vol. 3, p. 313. P. G. Menon, J. Am. Chem. Soc., 1965,87, 3057. S . Ozawa, S. Kusumi and Y. Ogino, J. Colloid Interface Sci., 1976, 56, 83. Y. Wakasugi, S. Ozawa and Y. Ogino, J. Colloid Interface Sci., 1981, 79, 399. J. Specovius and G. H. Findenegg, Ber. Bunsenges. Phys. Chem., 1980, 84, 690; 696. J. J. Kipling, Adsorption from Solution of Non-electrolytes (Academic Press, New York, 1965), p. 188. R. W. Cranston and F. A. Inkley, Advances in Catalysis (Academic Press, New York, 1957), vol. 9, p. 143. lo S. Ozawa, N. Ooyatsu, M. Yamabe, S. Honmo and Y. Ogino, J . Chem. Thermodyn., 1980,12,229. l1 A. S. Coolidge, J. Am. Chem. Soc., 1934, 56, 554. l2 A. Michels, P. G. Menon and C. A. Ten Seldam, Reel. Trav. Chim., 1961, 80, 483. 'I W. van Megen and 1. K. Snook, Mol. Phys., 1982,45, 629. (PAPER 3/794)

ISSN:0300-9599    

DOI:10.1039/F19848001049

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

J . Chem. Soc., Faraday Trans. 1, 1984,80, 1059-1068 Adsorption from Binary Solutions on Activated Charcoal at High Pressures up to 490 MPa BY SENTAROZAWA, KAZUO KAWAHARA, MASASHI YAMABE, HIDEO UNNO AND YOSHISADA OGINO* Department of Chemical Engineering, Faculty of Engineering, Tohoku University, Aramaki Aoba, Sendai 980, Japan Received 6th June, 1983 Effects of pressure on surface excesses for twelve adsorption systems each consisting of a binary solution of components A and B and activated charcoal have been studied at pressures up to 490MPa and at a temperature of 298.2 K. The following three types of pressure dependence have been observed: arA/ap < 0, a r A / @ > 0, = 0, where r A is the surface excess of component A. A comparison of these results with the difference in the pressure derivatives of pure-component interfacial tensions, (ao",ap - aa",dp = AAB), has revealed that and a r A / @ < 0 if A,, takes a large positive value, a r A / @ > 0 if A A B takes a large negative value ar,/ap = 0 if A A B is small.General criteria for determining the directions of the pressure dependences of adsorption have been derived theoretically and are utilized to discuss the experimental results. The scarcity of information about adsorption from solution onto solid surfaces at high pressures is in sharp contrast to the large number of studies reporting the results of low-pressure adsorption studies.'$ Because of this situation, we have initiated studies of high-pressure adsorption from the liquid phase and published3 details of the experimental devices and techniques used.In addition, we4 have studied the adsorption of pure liquids at high pressures up to 196.2 MPa in order to obtain data for the pressure dependence of the interfacial tension of the absorbed phase. These two studies have provided us with foundations on which to base the present work. This paper presents adsorption data for twelve adsorption systems, each con- sisting of a binary solution and activated charcoal, and shows that the pressure dependences of the adsorption isotherms can be classified into three types. A comparison between the present observations and our previous study4 reveals that the difference in the pressure derivatives of the pure-component interfacial tensions is one of the main factors determining the high-pressure characteristics of the adsorption isotherm from a given binary solution. An application of the adsorption theory proposed by Schay5 to high-pressure adsorption reveals another factor determining the direction of the pressure dependence of the adsorption isotherm.EXPERIMENTAL MATERIALS Activated charcoal (AC, Futamura Chemical Industries Ltd, type CS-120A) in a granular form (1-2 mm size) and with a specific surface area of 1350 m2 g-l (N, adsorption; B.E.T. 10591060 HIGH PRESSURE ADSORPTION ON ACTIVATED CHARCOAL I .o - I M -1.0 0 1.0 M - 1 .o r 11.0 0 -1.0 0.5 XMCP 0.1 0 XMCH Fig. 1. Pressure dependence of the surface excess r,: (a) MCP+Pr'OH/AC; (6) MCP+ PrnOH/AC, (c) MCH+PrnOH/AC; 0 , O . l ; a, 294; 0 , 4 9 0 MPa. method) and a pore volume of 0.63 cm3 g-' was used as an adsorbent after degassing.This adsorbent is the same as those used in previous ~ t u d i e s . ~ ' ~ Twelve binary solutions served as adsorbates. They were MeOH + PrnOH, MeOH + BunOH, EtOH + PrnOH, EtOH + BunOH, MCP + PrnOH, MCP + PriOH, MCP + BunOH, MCP + nC,, MCP + nC,, MCH + MCP, MCH + PrnOH, and nC, + EtOH : MeOH, EtOH, PrnOH, Pr'OH, BunOH, MCP, MCH, nC, and nC, denote methanol, ethanol, propan-1-01, propan-201, butan-1 -01, methylcyclopentane, methylcyclohexane, hexane and heptane, respectively. The purity of every reagent was gas-chromatographically determined to be better than 99.6%.S. OZAWA, K. KAWAHARA, M. YAMABE, H. UNNO AND Y. OGINO 1061 I I l l I 0 0.5 1 .o XMM~OH Fig. 2. Pressure dependence of the surface excess rA : (a) MeOH + PrnOH/AC, (b) MeOH + BunOH/AC; 0,O.l; a, 294; 0 , 4 9 0 MPa.APPARATUS AND PROCEDURE A specially designed adsorption cell, a device for the stable stirring of the high-pressure adsorption system, a device for sampling without disturbing the high-pressure adsorption equilibrium and a precision differential refractometer for sample analysis are necessary for the experimental determination of adsorption isotherms at high pressures, in addition to a conventional high-pressure generating unit. Details of these have already been rep~rted.~ Experimental procedures have also been reported3 in detail, and hence they are only briefly described below. A given amount (W/g) of the adsorbent degassed at 1.3 x 10-1 Pa and 400 OC for 3 h was mixed with a desired binary solution A+B (A and B denote the constituents of the solution) with a predetermined initial composition (xi and xg, both in mole fraction), without exposing the adsorbent to the ambient atmosphere: the mixing was carried out in the adsorption cell by breaking an ampoule containing the degassed adsorbent.The cell with its contents was accurately weighed and then pressurized. The desired adsorption pressure was maintained until adsorption equilibrium was reached (ca. 5 h), continuing the stirring of the adsorption system. A small portion of the solution in1062 HIGH PRESSURE ADSORPTION ON ACTIVATED CHARCOAL 0 0.5 1.0 xMCP -2.01 I I I I I I I I 1 0 0.5 1 .o xMCP . : - E E M O 3 q 0 0.5 1.0 XEtOH Fig. 3. Pressure dependence of the surface excess r A : (a) MCP + nC,/AC, (b) MCP + nC,/AC, (c) EtOH+Bu"OH/AC; 0,O.l; a, 294; 0 , 4 9 0 MPa.equilibrium was withdrawn keeping the pressure as close as possible to the equilibrium value. The sampling was repeated several times and the composition (xA and xB, both in mole frac- tion) of each sample was measured with the differential refractometer. The average of the composition data thus obtained was regarded as the equilibrium value. The surface excess (r,) was evaluated by where no denotes the total number of moles of the solution charged in the adsorption cell. The temperature of adsorption was maintained at 298.2 K. r A = (no/ w) (xi - XA) (1) RESULTS ADSORPTION ISOTHERMS Adsorption isotherms observed for the systems MCP + PriOH/AC MCP + Pr*OH/AC and MCH + PrnOH/AC (where AC indicates activated charcoal) are shown in fig.1. The isotherms of these systems exhibited sigmoidal shapes and the values of r M C p and r M C H decreased appreciably on increasing the adsorptionS. OZAWA, K. KAWAHARA, M. YAMABE, H. U " O AND Y. OGINO 1063 3 ,O 2.0 - I 00 - 2 1.0 E u' 1 r 2 0 -1.0 0.5 XllC, 1.0 Fig. 4. Pressure dependence of the surface excess r,: nC,+EtOH/AC; 0, 0.1; 0, 294; 0, 490 MPa. pressure. These characteristics of the isotherms were similar to that of the MCP+ EtOH/AC system studied ear lie^.^ The surface excess rMeOH for the MeOH+PrnOH/AC system and that for the MeOH + BunOH/AC system also exhibited isotherms which were sigmoidal in shape, as shown in fig. 2. These surface excesses, in contrast to those shown in fig.1, increased with pressure. The surface excess rMCq for the MCP+nC,/AC system and that for the MCP + nC,/AC system exhibited U-shaped isotherms and decreased on increasing the adsorption pressure, as shown in fig. 3. This figure also includes the adsorption isotherms for the EtOH + BunOH/AC system. The latter isotherms are also U-shaped, although the direction of their pressure dependence is opposite to those of the other isotherms shown in the same figure. IsothermsforthesystemsnC, + EtOH/AC,MCH + MCP/AC,MCP + BunOH/AC, and EtOH + PrnOH/AC showed little dependence on pressure, as is shown in fig. 4 and 5. DISCUSSION CLASSIFICATION OF ADSORPTION SYSTEMS The experimental results mentioned above show that the pressure dependences of the surface excesses can be classified into three types, i.e.(i) dr,/dp > 0, (ii) dr,/dp < 0 and (iii) dT,/dP = 0. Type (i) includes three systems, EtOH + BunOH/AC,1064 - I M - z -g 2 X -1 .o HIGH PRESSURE ADSORPTION ON ACTIVATED CHARCOAL (0 1 0 - - e m @ ,- I I I 1 I I I I I 1.01 1 -2.0 0 I I I I I I I I 0 1.0 - I M - E E = o 9 t3 --. - (a Q I I I I I I I I I MeOH + BunOH/AC and MeOH + PrnOH/AC, and type (ii) includes five systems studied in the present work and one system studied ear lie^.^ They are MCP + nC,/AC, MCP + nC,/AC, MCH + PrnOH/AC, MCP + PriOH/AC, MCP + PrnOH/AC and the earlier MCP + EtOH/AC systems. Type (iii) includes four systems, nC, + EtOH/ AC, MCP + BunOH/AC, MCH + MCP/AC and EtOH + PrnOH/AC. Note that the direction of the pressure dependence of the surface excess is reversed if constituent A is replaced by constituent B.FACTORS GOVERNING THE PRESSURE DEPENDENCE OF ADSORPTION PURE-COMPONENT INTERFACIAL TENSIONS According to current theories of adsorption,6 the behaviour of an adsorption isotherm of an adsorption system consisting of a binary solution and a solid adsorbentS. OZAWA, K. KAWAHARA, M. YAMABE, H. UNNO AND Y. OGINO 1065 Table 1. Values of A A B for various adsorption systems at 0.1 MPa and 298.2 K EtOH MeOH EtOH MeOH MCP MCH MCP MCP MCH MCPb MCP MCP nC7 BunOHa BunOH PPOH PrnOH EtOH BunOH MCP Pr'OH" PrnOH PrnOH EtOH nC7 nC," - 1.61 - 1.43 - 0.77 -0.59 - 0.43 0.1 1 0.68 0.95 0.95 1.63 1.72 2 . 1 5 < 2 . 1 5 a Values of (da0/8p), for BunOH and Pr'OH were interpolated to be 3.0 and 2.16, respectively, using an empirical correlation between the molecular collision diameter and (i3ao/i3p),. This system is included in this table in order to discuss the pressure depen- dence of rA although its isotherms have already been rep~rted.~ The relation (aatcg/+), < (aa;C,/i3p)T was suggested by the correlation mentioned above, although it was not possible to obtain numerical value of (aaiC&),. MeOH+ BunOH M C P + n C, -2 .o - 1 .o 0 1 .o 2 .o V L A 10- 9cm v - ---7 '-- - v --/ car,/ ap)<O car,/ap '0 ( a r,/ a,) = o Fig.6, Relation between ( X A / a p ) T , a-nd A A B : ( J ) systems with ( a r A / a p ) , , > 0, ( 1 ) systems with (arA/i?@)T,X < 0, (C ) systems with (arA/i$)T.s = 0; (*) adsorption data were reported in ref. (3). is related to the difference in the pure-component interfacial tensions.We have postulated that the difference in the pressure derivatives of pure-component interfacial tensions (01, and 00,) might be one of the principal factors governing the pressure dependence of the surface excess. In other words AAB, defined by (aa",C)p-aat/ap),, has been suspected of governing the pressure dependence of the surface excess for the adsorption system A + B/AC. The pure-component interfacial tension data needed to substantiate this view are already a~ailable,~ giving the values of A,, listed in table 1. The close relation between ( X A / 8 p ) , , zA and A,, is best illustrated by fig. 6, which shows that the surface excess rA varies little with pressure when the value of AAB is small and falls to a region around the origin of the coordinate representing AAB.On1066 HIGH PRESSURE ADSORPTION ON ACTIVATED CHARCOAL . . . . . . . . . . . . . I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . a . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I . . . . . . . . . . i a r e a 1 1-1 are: I I . . . Fig. 7. Graphical expression of the criteria for high-pressure behaviour of adsorption isotherm : ( x ) indicates a representative point P with its projection P upon the AAB coordinate. the other hand, r A varies with pressure when IAABl is large: r A decreases if AAB is positive but increases if AAB is negative.Note, however, that whether there is a distinct boundary between the region of AAB where (WA/3p)T,xA = 0 and the region where (arA/ap)T,xA # 0 is uncertain. PARTIAL MOLAR VOLUMES IN THE BULK PHASE AND IN THE ADSORBED PHASE According to Schay5 the equilibrium adsorption from a binary solution (A + B) onto (2) where aj denotes the activity of the component j in the bulk phase, a; denotes the activity o f j in the adsorbed phase, Sj denotes the molar area of surface occupation of the adsorbed component j, r denotes the ratio S,/SA and KAB is the equilibrium constant of adsorption. solid adsorbent can be described by (ai/aA)r(a13/a&) = exp [SB(a",aa",)/RT1 = KAB Eqn (2) is equivalent to (yk/yA)r (YB/Y&) (Xa/XA)r (xB/xb) = exp [SB(ak-a",/RT] (3) where y j and y; are the activity coefficients forj in the bulk phase and in the adsorbed phase, respectively, and xi is the'mole fraction o f j in the adsorbed phase.By differentiating eqn (3) with respect to pressure, one obtains (a In Kx/ap)T, XA = (S,lRT) (aai/ap-aoi/aP)T + (a In KY'/aP)T, X A (4) where Kx = (xa/~,)~(x~/x;) and Ky = (~k/yA)~(yB/y&). The left-hand side of eqn (4) represents the pressure dependence of the surface concentration of A, sinceS. OZAWA, K. KAWAHARA, M. YAMABE, H. U " O AND Y. OGINO 1067 In other words, the sign of the right-hand side of eqn (4) is equal to that of The relation expressed by eqn (4) therefore means that the pressure dependence of the surface excess, ( a r A / 8 p ) T , x A , is determined by two factors, i.e.(SB/RT)(a&/@ -a&/+)T, xA and (a In K~l/tIp)~, zA. Obviously, the former factor is essentially equivalent to A A B mentioned in the preceding section. The latter factor introduced here by the theoretical treatment can be transformed into a more understandable form using the following relations : @ ~ A / ~ P ) T . xA* and P yi = r;* exp ( q / R T ) d p j P* where rj* and yi* denote y j and y i at a reference pressure p* (0.1 MPa), respectively, and 5 and are partial molar volumes o f j in the bulk phase and in the adsorbed phase, respectively. The result of transformation is CRITERIA FOR THE PRESSURE DEPENDENCE OF SURFACE EXCESSES Taking above-mentioned discussion into account, one can readily obtain general criteria for the direction of the pressure dependence of surface excess.Namely, (X'A/ap)T,xA > 0 if AVBlB/SB-AV'AtA/SA > A,, (8) (arA/i3p)T,zA < 0 if A V B f B / S B - A V A f A / S A < A,, (9) ( a r A / a p ) T , x A = 0 if AVBtB/SB-AVA*A/SA = A A B (10) where A VAfA = PA - FA and A VB'B = if;3 - FB. Obviously the partial molar volumes are replaced by the corresponding molar volumes when the adsorption system is ideal. A graphical expression for the criteria derived above greatly facilitates the discussion. This is best demonstrated by fig. 7. The pressure dependence of any adsorption system can be characterized by a point (the representative point) in this figure. Every point in area I in the figure represents an adsorption system whose surface excess increases with pressure, and that in area I1 represents an adsorption system in which the surface excess decreases with pressure. Points on the line LL' represent an adsorption system with zero pressure dependence.The empirical relation between A,, and ( a r A / @ ) T , s A shown in fig. 6 can be regarded as illustrating the projections of the representative points of the adsorption systems studied upon the abscissa (the A A B axis). According to this view there is no distinct boundary between values of A A B where (arA/t&)T,xA = 0 and where (arA/ap)T,ZA. # 0. An adsorption system can be pressure dependent even when A A B is small if the absolute value of ArBtB/SB-AV'AtA/SA for this system is large. This rationalizes the experimental result that (IAAB()EtOH+prnOH > ((AAB()MeOH+pr"OH even though the former system is pressure independent and the latter pressure dependent. Adsorption theory thus gives a reasonable interpretation of the experimental results. Unfortunately, however, the exact positions of the representative points for the adsorption systems studied remain to be determined. Adsorption isotherms and their pressure dependences will have to be obtained with extremely high accuracy in order to evaluate partial molar volumes of the adsorbed phase. These are beyond the scope of the present work.1068 HIGH PRESSURE ADSORPTION ON ACTIVATED CHARCOAL J. J. Kipling, Adsorption from Solution of Non-electrolytes (Academic Press, London, 1965). Adsorption at the Gas-Solid InterJace, ed. J. Rouquerol and K. S. W. Sing (Elsevier, Amsterdam, 1982). S. Ozawa, K. Kawahara and Y. Ogino, High-pressure Science and Technology, ed. K. D. Timmerhaus and M. S. Barber (Plenum, New York, 1979), vol. 1, p. 593. S. Ozawa, M. Goto, K. Kimura and Y. Ogino, J . Chem. SOC., Faraday Trans. I , 1984, 80, 1049. G. Schay, in Surface and Colloid Science, ed. E. Matijevic (Wiley-Interscience, London, 1969). D. H. Everett, in Adsorption at the Gas-Solid Interface, ed. J. Rouquerol and K. S. W. Sing (Elsevier, Amsterdam, 1982), p. 1 . (PAPER 3/932)

ISSN:0300-9599    

DOI:10.1039/F19848001059

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

J. Chem. Soc., Faraduy Trans. I, 1984, 80, 1069-1081 Gas Relative Permeability in the Capillary-network Model BY DAVID NICHOLSON Department of Chemistry, Imperial College, London SW7 2AY AND JOHN H. BTROFWULOS Physical Chemistry Laboratory, Democritos Nuclear Research Centre, Aghia Paraskevi, Athens, Greece Received 16th June, 1983 A theory for the permeability to a gaseous phase of a network of capillaries containing a second adsorbate phase is derived in terms of equations previously developed for the flow of gas in regular capillary networks with radii randomly selected from a distribution. The role of blind, as well as through, pores is considered. Numerical calculatisns, using a matrix- inversion method, are used to evaluate the theory. Excellent agreement is found at low volume fractions V , of the adsorbate, but divergence becomes marked as V , + 1.This can be attributed to the inability of the theory to account for the non-regular nature of a random network in which some capillaries are filled by condensed adsorbate. The flow of vapour through a porous material which is partly blocked by a condensed phase is of interest both for fundamental studies of pore structure and for a number of practical applications, e.g. the passage of gas through soils or porous electrodes, the properties of building materials etc. The problem can also arise in studies of vapour flow through porous materials under conditions where substantial condensation of the vapour can occur within the pores. In such circumstances it is difficult to determine the effect of pore blockage on the overall transport.One method of gaining insight into this effect is to introduce a stationary condensed phase into the porous material and then measure the flow of a second sorbate which remains gaseous within the pores under the conditions of the experiment. Experiments of this kind are difficult and few studies have been carried Moreover, even when the effort to construct this type of experiment has been made, there still remain a number of uncertainties which cannot be accounted for precisely, including those associated with the various features of the pore structure and with the possibility of interdiffusion hindrance by the vapour of the condensed phase. In similar circumstances computer studies of model systems have proved to be valuable, since they are not subject to the above complications and are thus capable of elucidating specific, well defined aspects of the flow behaviour.With the aid of insights gained in this way it may then be possible to identify those factors which are likely to be the most significant in the more complicated real systems. In earlier work we have carried out computer calculations on a capillary network as a model porous The network was entirely characterized by two variables, the degree of connectivity (measured as the number of capillaries meeting at a junction, nT) and the distribution of capillary radiusflr), which was randomly assigned to the network. Although useful data were obtained in this way, the calculations are lengthy. Accordingly, a more compact representation of the results was sought.Building on 10691070 GAS PERMEABILITY IN THE CAPILLARY-NETWORK MODEL the earlier work of Schopper et aL7 we were able to construct a convenient, semi-empirical equation for this purpose.8 This treatment was applied to a wide range of network connectivities and model distributions (both symmetrical and skewed) and was shown to be capable of giving a satisfactory representation of accurate numerical solutions for the network flux equations within the limitations imposed by this model. Exceptions to this were noted in the case of highly skewed bimodal distributions in networks of low connectivity. Notwithstanding the limitations imposed by this extreme case, the semi-empirical equation, which has as its only input the moments of a distribution function and the network connectivity, should generally prove to be very helpful within the context of the network model.In the present work we report the results of computer calculations of the relative permeability of networks partially blocked by a wetting adsorbate phase, and show how this semi-empirical treatment may be developed for the description relative permeability behaviour. Note that in what follows we use the term relative permeability in referring to the experimental conditions described above. This term is in fact generally applicable to cases in which both phases flow simultaneously, and the limiting case considered in this work would be more rigorously described as the end-point relative permeability. THEORY We consider the application of our previous treatment of flow in capillary networks to the case where the component capillaries are partly filled by an adsorbate phase.The theory encompasses blocking both by adsorbate films and by condensed adsorbate. It also makes allowance for the presence of blind pores, i.e. capillaries which, although they do not carry flux, are accessible to adsorbate in the same way as the remaining capillaries, which are referred to as ' through-pores'. In the model the capillaries have circular cross-sections of radius Y. At given relative pressure p / p o (= y ) those capillaries having radii less than some critical value rcrit (which is determined by y ) are completely blocked to the flow of gaseous phase by adsorbate. The remainder are partly filled by an annulus of condensed adsorbate of thickness t which also restricts the flow of gaseous phase. It will be assumed that t is a continuous function of the relative pressure y and of the capillary radius.The form of this continous function and the significance of continuity (which cannot be strictly valid of course, since we are dealing with molecular phenomena) will be given further consideration below. It is sufficient to mention here that this assumption is not of primary importance in the present context. Only the open cores of the through capillaries can carry flux; however, as the relative pressure increases and a greater proportion of the capillaries become totally blocked, some of these will be isolated from the external source or sink and can therefore no longer contribute to transport.When the fraction of totally blocked through-pores reaches the percolation threshold for the network, flow of the gaseous phase is no longer possible. The network is assumed to represent a porous medium of cross-sectional area U and thickness I. The pore radius distribution for the whole medium isfM(r) with Y c (ra,rb). If the total number of capillaries which make up the network is N , the distribution of through-capillaries (subscript T), blind capillaries (subscript B) and the total distribution are related by NfM(Y) = N13 fB(') -k NT fT(r)* (1) The moments of these distributions are defined by 'b rna = ( rn ) a = jra rnfa(r)dr (a = M, B, T). ( 2 4D. NICHOLSON AND J. H. PETROPOULOS 1071 Similarly the moments applying to the supercritical capillaries are defined by ‘b aria = ( a n ) , = J rnfa(r)dr (a = M, B, T).rcrit We shall also require moments defined with respect to capillaries partly filled by adsorbate of thickness t and having a core radius rC = r - t: In all these definitionsf, is normalized such that Jr: fa(r) dr = 1. The total amount of immobile phase present in the network can be expressed as a fraction of the total available volume and is designated V,. It can be found from the fractional volume of adsorbate in the partly filled capillaries, 6, where J rcrit / J r , and the fractional volume, K, in the subcritical capillaries, given by ‘wit K = Jra r2fM(r) d r / r r2fM(r) dr. ra On adding eqn (4a) to (4b), and using eqn (2) and (3), we have where eqn (1) has been used to obtain the second expression. The porosity and internal surface area (per unit volume) of the through-pore parts of the porous medium represented by the network at a relative pressure, y, of the adsorbate phase are given, respectively, by &TO) = N X h (YE)T/ UZ (6 4 and ATCy) = 2Nn h (r,,)T/ UL (6 b) where h is the length of the individual capillaries forming the network.The flux through the network, at a relative pressure y , can be written either in terms of the network permeability, P&), or in terms of the diffusion coefficient as (7) Jb) = upNb) AcL/z = urcgNT(Y) DgT(y) &T(Y) Ac;/z where ACi/Z is the concentration gradient across the porous membrane represented by the capillary network. The diffusion coefficient is that for pure Knusden flow through a single capillary of infinite length, oriented in the direction of flow (the reference medium), whose radius is equal to the mean effective radius 2&,/AT of the part of the network carrying flux.Thus O,*b) = 4ET(y) c/3 A T b ) (8) where a is the mean molecular speed of the flowing gas. The structure factor K ~ N T ~ ) is the ratio of the diffusion coefficient calculated for the network at relative pressure, y , to that of the reference medium, and therefore reflects the differences in flow1072 GAS PERMEABILITY IN THE CAPILLARY-NETWORK MODEL behaviour between them. It is the central object of attention in discussing flow in real or model porous materials. It is useful to write this as a product of normalized structure factor zgNT and a factor I C ~ which accounts for any anisotropy and for the dimensionality of the network.The normalized structure factor for the network ~ z ~ N T may be written in terms of the core radius Y, and the 'relative resistivity' RNT of the network, discussed in earlier work.8 Here we add the subscript T to RN calculated in ref. (8), to emphasize that it applies to the through-pore network. Thus RgNTb) = ICo ICgPT(Y)IRNT(Y) (9) where K~~~ is the corresponding structure factor for a parallel bundle of capillaries oriented in the direction of flow with the samef,(r) as the network; it is given by8 ICgPTb> = (Yg)T (rc>T/(rf>k (10) In practice the properties of the through-pores cannot usually be measured directly, and the effective radius in eqn (8) would not be known.Instead the effective radius for the whole medium, including blind pores, is the measureable quantity. The flux expressed by eqn (7) must, however, be invariant, and I C ~ N T is replaced by IC~NM, the structure factor for the whole medium, in order to maintain this invariance. A comparison of eqn (7) and (8) with analogous equations in which these substitutions have been made then leads to the relationship The subscript M, introduced in eqn (l), is normally omitted but is retained here in order to stress that both blind- and through-pores are included. Combining eqn (6) to (1 1) we obtain for the permeability at relative pressure y , The quantity of interest here is the relative permeability, P,, which from eqn (12) is For a network of capillaries having uniform radius, we have from eqn ( 5 ) and since RNa = 1 in this case, eqn (13) reduces to the simple result P, = (1 - l93'2.(14) It is our object to examine the relationship between P, and V , for the more general case considered here. In our previous work we have shown that RN(0) can be written as an expansion in the moments of the distribution function as RNT(0) = 1 + 9AT ZZT - 9L3T z3T + 1 5L4T ZIT + 18L22, ZgT * * - (15) where the subscript T has been added here to emphasize that only capillaries in the through-network contribute. The moments znT are defined by znT = [(rn>T- (Y)?l/(r)'? (16)D. NICHOLSON AND J. H. PETROPOULOS 1073 and are a measure of the nth-order dispersion of the radius distribution of the through-network. It will be useful to recall the following relationships : (17) fT(r) dr = -&'(z) dz = f,@> dp where z = p/(p) - 1 and p = r / r m with rm = g(ra + rb).Since r = (r) (1 + z) it follows from the last two equations that The quantity AT in eqn (15) is a measure of the connectivity of the through-network in which nTT capillaries meet at a junction; it is given by AT = 2/nTT. (20) The expressions for A3T, AdT and A22T are those determined as A,, A, and 1222, respectively, in ref. (8), using networks having different types of radius distribution and a variety of connectivities; they may be written as d3T = 2.9A;- 1.9AT d4T = 6.402; - 6.711; + 1. 3,AT A22T = -9.9,&+ 14.7dk-3.73AT. R N T b ) = + 9AyT c2 - 9A3yT c3 + 1 5A4yT c 4 + 8l22yT (21 a) (21 b) (21 4 (22) When y # 0, RNT(0) must be replaced by an analogous expression Here the subscript y on AnT implies that network connectivity will be altered when some of the capillaries become blocked to flow as y increases.Under such conditions we may tentatively assume that 2, is = 2T/aOT (23) with aOT given by eqn (2b), and the remaining AnyT then follow from eqn (21). Calculation of A,, according to the ansatz of eqn (23) is equivalent to assuming that the reduction of the effective 'local' connectivity is the same at each individual junction and equal to the mean value in the network as a whole. Since nTT is a small integer it is likely that the local connectivity will vary considerably when an appreciable proportion of the capillaries become blocked to flow. Eqn (23) would therefore be expected to be satisfactory when AT is small or when the number of junctions is very large.The dispersion c in the modified distribution can be put in the form where z = t/rmT, and c is defined with respect to the mean radius of the open through-pores in order to satisfy the requirement that cl = 0. Thus the moments of (p-T) (= pc) are given by an equation analogous to eqn (3): (P:)T = spbP?fT@) Pa dP with (pa, Pb) corresponding to (ra, rb). 36 FAR 11074 GAS PERMEABILITY IN THE CAPILLARY-NETWORK MODEL The expansion coefficients in in eqn (22) can be expressed in terms of pc by (26 4 Often it is assumed, in considering adsorption and capillary condensation in pores, that t is a function only of y and does not depend on pore width. We shall examine this simplifying assumption below, but first of all we develop the more general case where t = t(r,y).It is sufficient to assume that t can be written in the form of a, quadratic polynomial in l/p whose coefficients B,, B, and B, will be functions of y ; we therefore write = tb)/rmT = B, + B,/p + BJp2. (27) Then (p - z), which appears in eqn (25), can be written as where B,, = 1, Bk for k > 3 is found from B,, . . . , B, and the summation indices are determined by k = k,,k,+l ..., k,+3n (30 4 with k , = t ( n - 1) (3n+2) (30 4 and j = n,n-1, ..., -2n. (30 4 It is convenient to write eqn (29) in terms of a set of moments anT which are closely related to znT defined by eqn (18) and which in turn can be defined by 'b Zcrit anT = J znfT(z)dz. (31) If a general expression for znT can be derived for a given distribution function, the same expression can also be used to calculate anT simply by replacement of the lower limit z by zCrit.Comparison of eqn (31) with (2b) shows that aoT = aoT. In terms of anT eqn (29) then becomes where k and j are given by eqn (30a)-(30c), amT by eqn (31) and jCm has the extended definition (33 4 j ! (j-m)!m! j 3 O:jCm = ( - j - m - l ) ! ( - ( - j - l)!m! j -c O:jCm = (33 b) Note that the coefficients B, are functions of the relative pressure through eqn (27). The components of the theory can now be summarized. The relative permeabilityD. NICHOLSON AND J. H. PETROPOULOS 1075 is given by eqn (13), in which RNT(y) is given by eqn (22). The expansion coefficients, Cn, in eqn (22) are written in terms of pc in eqn (26) and A,, is given by eqn (23).(p,?), is calculated from eqn (32). We note that although eqn (1 5) onwards have been written in terms of quantities related to the through-pore distributionf,, it would be equally valid, in view of eqn (13), to replace these by equations relating to the distribution fM for the whole medium. RESULTS AND DISCUSSION The theory derived in the preceding section takes account of blind- as well as through-pores. It is apparent that iffB has a very much smaller mean radius thanf, but a large total volume, < may change without affecting P,. This and other aspects of networks containing blind pores will be investigated in a future publication. Here we confine our attention to networks in which only through-pores are present.ADSORPTION ISOTHERMS In order to implement the theory it is necessary to introduce a relationship between t and r ; this is most conveniently done by using the relative pressure y as a parameter. The filling of the capillary spaces as y increases can be envisaged, as already mentioned, as proceeding via the build-up of annular layers of adsorbate on their walls, followed by condensation when t reaches some critical value. This simplified picture is not altogether satisfactory, since any degree of filling which is intermediate between an integral number of monolayers would correspond to a core of varying cross-section. However, the approximation of replacing this by a core having the mean radius, corresponding to a non-integral number of layers, should not be severe and does not affect the model calculations presented here.We have used a thermodynamic model for adsorption in a predominantly meso- porous system. Adsorption into a single cylindrical capillary of radius r can be described approximately by the equationg (34) Cl c2 lny=-+- r - t tS in which C, = -(yV,/RT), where y is the surface tension and V, the molar volume of the adsorptive. For N, at 77 K, for example, C, = -0.477 nm. If the adsorption is onto a plane surface, the first term in eqn (34) is zero and the remaining equation is the FHH or slab-theory isotherm, in which C, is a negative constant and s would typically have a value of ca. 2.5, although a value of s = 3 is frequently used. For N, at 77 K, C, = -0.218 nm3 with s = 3. The derivation and validity of eqn (34) has been discussed el~ewhere.~ Here it is sufficient to point out that the equation is suitable for the description of the combined adsorption-condensation process occurring in the filling of mesopores. The constant Cz is related to the difference between the strength of the adsorbate-adsorbent interaction and that of the adsorbate-adsorbate inter- action. Therefore when (- C,) is large eqn (34) gives rise to typical type-IV isotherms, and as (- C,) decreases this behaviour goes over to that of type-V isotherms. In order to evaluate the theory presented here, we have adopted the N,/77 K parameters mentioned above, which give rise to type-IV isotherms with an initial steep rise in the amount adsorbed at y 4 0.1.It is likely that eqn (34) is not adequate for the description of adsorption into micropores (diameters Z four adsorption molecular diameters).In principle a suitable extension into this size range could be made, but here we have avoided this difficulty by choosing pore distribution functions from which such small radii are absent. 36-21076 GAS PERMEABILITY IN THE! CAPILLARY-NETWORK MODEL I I I I 1 I 0.90 0.70 2 3 Lk 0.50 L- 4- 0 Q. 0.3 0 0.10 0 Fig. 1. Relative permeability (P,) and related quantities plotted as a function of the volume fraction adsorbed (Q. ‘Fne relative permeability from numerical solutions of the flux equation by matrix inversion, calculated at various values of V,, is shown as the open circles. The full line associated with these points was calculated from the theory given in the Theory section and shows positive deviations at higher values of V,.The other properties represented are distin- guished by the following symbols: x , reciprocal connectivity, AY [ =(2/nT),]; +, fractional volume of adsorbate in the filled (i.e. subcritical) capillaries, 5 ; A, fractional volume of adsorbate in the partly filled capillaries, 4; 0, relative pressure, p / p o (adsorption isotherm). The distribution function was triangular symmetric [type D of ref. (5)] with a width parameter CT and mean radius 4.0 nm. The initial connectivity of the network was nT = 18(A = 0.1 1). Eqn (34) shows that the thickness t at any relative pressure will be a function of r. Even for integral values of s a convenient analytical expression is not available, and would obviously be out of the question for non-integral s.We have therefore used an iterative procedure to evaluate t(r), based on the algorithm It is found that accuracy to five significant figures can usually be achieved in fewer than five iterations; negative values of the left-hand side, which correspond to the onset of condensation, need to be trapped. t exhibits a monotonic decrease with increasing r and, as already mentioned, can be fitted to a quadratic function in (1 / r ) . In practice, since there is an asymptote at t , = (C2/lny)1/S, it is convenient to fit ( t - t,) initially. Standard deviations are typically of the order of 0.01 with t and r in units of C,, and the rather weak dependence is indicated by the typical convergence ratios for the coefficients in eqn (27), IB,/B,I = IB,/B,I x The critical radius rcrit, below which condensation occurs at a given relative1077 0 0.20 0 .LO 0.60 0.80 1 .O Fig.2. As fig. 1 . for connectivity nT = 8 (A = 0.25). vs pressure, may be found from eqn (34) using the condition (a lny/~t),,,,, = 0. A suitable iterative algorithm is This was used to find tcrit for a given value or r, and the corresponding y alue bcrit(r, tcrit)] was then calculated from eqn (34). This iteration was carried out over a range of values of 1 / r in intervals of 0.01 C, to give ycrit [O. 15, 1 .O], and 1 / r fitted to a 4th-order polynomial in ycrit with standard deviation better than O.OOICl-l. It is of course only necessary to perform this calculation once for a given set of parameters C,, C,, s.For the N,/77 K parameters used here it is found that Cl/r = 0.5928 - 1.139~+ 1.2735~2-0.8822y3+0.160y4. (37) NUMERICAL SOLUTION OF THE NETWORK FLUX EQUATION A number of calculations were carried out using the Gaussian elimination method applied to partially blocked networks. The main features of this type of calculation, which solves a set of linear equations for the junction concentrations, have been discussed el~ewhere.~-~ A few special features relating to the present work should be mentioned. this can have an effect on K ~ ~ . It was also noted, however, that the relative permeability was far less sensitive to network size than the absolute value, and that, for example, 6 x 6 x 10 networks gave values of P, very close to those obtained in 10 x 10 x 30 networks when no condensation occurs.The smaller size could therefore be used for An investigation of size dependence showed that, as anticipated from earlier1078 GAS PERMEABILITY IN THE CAPILLARY-NETWORK MODEL 0.90 0.70 0.50 0.30 0.10 0 0.20 0.1 0 0.60 0.80 1.0 Fig. 3. As fig. 1. for connectivity nT = 4 (A = 0.5). vs calculations on three-dimensional networks (n, = 6, 12, 18), thereby considerably reducing the demand on computer time for these calculations. The two-dimensional calculations (n, = 4,8) are very much faster, and 15 x 30 networks were used in these cases. Radius distribution functions were chosen from the set of model functions (A-F) previously inve~tigated;~ in this work only type D (triangular symmetric) and type C (triangular with negative skew) were investigated in detail.These were chosen because they resemble the type of function found in practice,l* but the theory is not specifically limited to any particular distribution function. Two values of the mean radius ( ( r ) = 4.0 and 10.0 nm) were used. RESULTS FROM NUMERICAL COMPUTATION A number of 4 against curves were calculated by matrix inversion. These results are shown as open circles in fig. 1-3 and by circles in fig. 4 and 5. The general trends observed were very similar and are amply represented by the examples shown in fig. 1-5. Two regions can be distinguished: (i) that in which no capillary condensation occurs and the gaseous flux is hindered only by reduction in the effective diameter of the flow channels and (ii) that in which the structure of the open network is modified in a random manner owing to the blockage of flow paths by condensate.The properties of the networks in these regions are illustrated in fig. 1-3 by plotting 5, 6 and A,, against V, [see eqn (4), (23), (26) and (5), respectively]. It is clear that P, is very close to being linearly dependent upon V, until V, becomes appreciably greater than zero. Departure from linearity occurs when 6 increases. Since a large amount of the immobile phase then condenses into the cores of the partly filled capillary, the increase in V, tends to offset the concomitant decrease in P, causing P, to deviate in a positive direction from the initial linear section at the higher nT values. At the lowest connectivity (nT = 4), on the other hand, the deviation is in the negative direction.D. NICHOLSON AND J.H. PETROPOULOS 1079 1 0 0 0.1 0.2 0.3 vs Fig. 4. Effect of distribution function shape. Connectivity nT = 6; distribution width parameter, 0 = 0.3; mean radius 10.0 nm. Points from numerical calculations; full line from theory; broken line, straight line of slope -3/2 (see text). (a) Skewed triangular distribution with weighting towards the lower radii [type C of ref. ( 5 ) ] ; (b) (displaced 0.1 to the right) symmetrical triangular distribution. 1 .o 0 0.2 0.4 0.6 0.8 vs Fig. 5. Effect of mean radius in a network with connectivity nT = 8. The distribution function is triangular symmetric (type D) with width parameters = 0.6: 0, mean radius 4.0 nm; 0, mean radius 10 nm.1080 GAS PERMEABILITY IN THE CAPILLARY-NETWORK MODEL The skew and symmetric distributions show only minor quantitative differences, as shown in fig.4. A change of 0 from 0.2 to 0.6 for either distribution produced a relatively small effect which is best understood in terms of the linearised theory given below. Finally, a change in mean radius from ( r ) = 4.0 to 10.0 nm, although it obviously has a large effect on the relative pressure at which condensation begins, does not have a strong effect on the plot of P, against (fig. 5). It should be stressed that these conclusions apply to initially regular networks of randomly selected radii. In differently constructed networks changes in these parameters may result in more marked variation in P,. These will be discussed in a later publication.COMPARISON WITH THEORY In the sense that good agreement can be found between the results and the approximate theoretical treatment given in the Theory section, the latter can be said to be very successful up to the point where substantial capillary condensation occurs. However, it is clear from fig. 1-3 that as soon as some of the network capillaries are blocked, divergence between the two calculations becomes marked. The most likely explanation for this divergence is that the theory of ref. ( 5 ) cannot be successfully extended to describe the properties of non-regular networks by means of the simple assumption Ay = A/a, introduced in eqn (23). This assumption implies that when a number of capillaries are removed at random from these networks, the lower connectivity in some regions would be offset by higher connectivity in others.Clearly this conjecture has only limited validity. Note also that since the model of ref. (5) is based on the parallel-capillary model, which corresponds to A = 0, the expression developed for R , converges when A 6 1 but tends to break down as A + 1. Indeed the percolation probability of a network falls to zero at the serial-model (A = 1) limit, and therefore the permeability of an infinite network would also be expected to do so. This is borne out by the lower P, values obtained from the numerical calculations. Support for this explanation also comes from a comparison between the results for networks of different connectivity. It is clear from a comparison of fig. 1-3 that the agreement between numerical results and theory diminishes as the initial connectivity is decreased. However, in the network with the highest connectivity (n, = 18, I = 0.1 1 1) deviations between theory and numerical calculations can be observed well before I, = 0.5, suggesting that departure from regular network configuration might be important. LINEARIZED THEORY The observation that the theory given earlier is accurate at low V, and that P, against V, plots are apparently linear in this region suggests that a simplified formulation should be possible in which only those terms which are first order in V, are retained. This can be done if it is assumed that t is independent of p at low V,. After rather lengthy algebra it can be shown that P, reduces, in this case, to the expression where (39) Eqn (38) shows that departure of the P, against V, plot from the limiting slope of - 3/2 [see eqn (14)] will depend both on the width of the radius distribution and the connectivity of the network.D. NICHOLSON AND J. H. PETROPOULOS 1081 R. Ash, R. M. Barrer and R. Sharma, J. Membr. Sci., 1976, 1, 17. R. Ash and A. D. Hamilton, to be published. N. K. Kanellopoulos and J. H. Petropoulos, J. Chem. Soc., Faraday Trans. I, 1983,79, 517. D. Nicholson and J. H. Petropoulos, J. Phys. D, 1971, 4, 181. D. Nicholson and J. H. Petropoulos, J. Phys. D, 1975, 8, 1430. D. Nicholson and J. H. Petropoulos, J . Phys. D, 1973, 6, 1737. 1968, 16, 277. D. Nicholson and J. H. Petropoulos, J . Phys. D, 1977, 10, 191 1 . D. Nicholson and N. G. Parsonage, Computer Simulation and the Statistical Mechanics of Adsorption (Academic Press, London, 1982), chap. 7. lo S. J. Gregg and K. S. W. Sing, Adsorption, Surface Area and Porosity (Academic Press, London, 2nd edn, 1982). ’ J. R. Schopper, Geophys. Prospect., 1966, 14, 301 ; M. Rink and J. R. Schopper, Geophys. Prospect., (PAPER 3/1021)

ISSN:0300-9599    

DOI:10.1039/F19848001069

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

J. Chem. SOC., Faraday Trans. 1, 1984, 80, 1083-1088 Hydrogenation of Ethylene over Chromia-Alumina Catalysts Electron Spin Resonance and Kinetic Studies BY CHARLES A. MCAULIFFE* AND FATHY M. ASHMAWY~ Chemistry Department, University of Manchester Institute of Science and Technology, Manchester M60 1QD Received 17th June, 1983 The hydrogenation of ethylene over chromia-alumina catalysts containing known amounts of Cr2+ and Cr3+ has been investigated at temperatures between - 78 and 200 "C. The activity of the catalyst depends strongly on the presence of Cr3+ (determined by exr.), in as much as increasing Cr3+ concentration is accompanied by a considerable increase in the rate of hydrogenation. Preliminary kinetic studies show that the reaction rate is zeroth order in hydrogen at - 78 "C and first order at higher temperatures ( > 0 "C).The e.s.r. signal ascribed to Cr3+ (g = 1.986) undergoes some changes under the reaction conditions when exposed to a 1 : 1 mixture of hydrogen and ethylene at 100 "C. The chromia-alumina catalyst system has been the subject of many investigations. Its high activity for the hydrogenation of alkenes and the dehydrogenation of paraffins is well kn~wn,l-~ but the nature of the active sites is still open to speculation.6-10 We have reported previouslyll on the effect of the adsorption of ethylene, hydrogen and water vapour on the e.s.r. spectra of reduced chromia-alumina catalysts. More recently12 we have investigated the activity of a series of catalysts containing known amounts of Cr2+ and Cr3+ on alumina, where it has been found that Cr2+ plays an active role in the dehydrogenation of ethane.The present study was undertaken in an attempt to characterize further the active sites in the hydrogenation of ethylene over chromia-alumina catalysts containing known amounts of Cr2+ and Cr3+. Electron spin resonance has been used to follow changes in the signal ascribed to Cr3+ before and after ethylene hydrogenation. Preliminary kinetic results are also reported here. EXPERIMENTAL The preparation of the catalysts (1.0, 3.75 and 5.02 wt % Cr203 on alumina) and the procedure adopted for the purification of materials were essentially those described previous1y.l' In the present work two types of reduction were carried out at 500 "C and 1 atm$ for 8 h followed by outgassing at 500 "C for 1 h (hydrogen flow rate was 50 cm3 min-l): (a) catalysts reduced in carefully purified and dried hydrogen (Cr2+ produced); this will be referred to as reduced catalysts; and (b) reduction with hydrogen saturated with water vapour (only Cr3+ produced) (this was achieved by allowing hydrogen to flow through a water trap maintained at room temperature), the saturation of the hydrogen with water vapour being sufficient to reoxidise Cr2+ to Cr3+;12-14 this will be referred to as oxidized catalysts.t On leave of absence from the Faculty of Science, University of Tanta, Tanta, Egypt. $ 1 atm = 101 325 Pa. 10831084 HYDROGENATION OF ETHYLENE Table 1. Results of e.s.r. and rate measurements at -78 "C catalyst composition (wt % CrzO,) type of surface Cra+/sites Cr2+ /si tes (g catalyst)-' a (g catalyst)-' 1 .o 1 .o 3.75 3.75 5.02 5.02 reduced oxidized reduced oxidized reduced oxidized 3-52 x 1019 1.44 x 1020 3.2 x 1019 7.5 x 1019 6.75 x 1019 - 2.18 x 1020 2.19 x 1020 3.05 x 1020 - - 8.6 x 1019 Cr2+ rate (%) /mol min-l m-2 47.8 2.0x 10-8 - 3.6 x lop8 34.4 8.1 x 13.0 x 28.3 22.0 x 10-8 - 34.6 x - a Calculated from the number of spins from the doubly integrated areas of the em-.signals. Calculated from the difference in the number of spins upon reduction of Cr3+, according to O'Reilly and Ma~1ver.l~ The oxidation-reduction treatments were carried out in a silica reaction vessel attached to an all-glass system which could be evacuated to lop6 Torr.? The rate of ethylene hydrogenation was followed by measuring the pressure change in a constant-volume reaction vessel; the materials (hydrogen + ethylene) were circulated over the catalyst using a glass electromagnetic circulating system. The gases were thoroughly mixed before contacting the catalyst.The product, ethane, was identified by mass analysis using an A.E.I. MSlO spectrometer. The e.s.r. measurements were carried out on a Decca XI spectrometer at room temperature. A standard sample of CuSO, - 5H20 was used in order to measure the spin concentrations, which were reproducible to better than 10%. RESULTS AND DISCUSSION E.S.R. AND CATALYTIC ACTIVITY Table 1 summarizes the results of ethylene hydrogenation at -78 "C and the surface composition of the catalysts. It is clear from the results in table 1 that a good correlation exists between the catalytic activity and the number of Cr3+ ions present.Two observations are worth recording. First, the most active catalyst (5.02 wt % Cr,03) gave a figure of 3.05 x lo2* Cr3+ sites (g catalyst)-' and shows the highest rate of hydrogenation; secondly, comparison between columns 3 and 6 shows that the oxidation of Cr2+ to Cr3+, for a given catalyst, is accompanied by a substantial increase in activity, which strongly suggests that the Cr3+ sites thus formed are active in the hydrogenation reaction. In further support of this two series of experiments were performed. In the first series the effect of added water vapour on the reduced and oxidized catalysts was tested (2.0 x mol water were adsorbed per g catalyst at room temperature followed by outgassing at room temperature). The catalyst was almost completely poisoned by exposure to water.However, the initial activity could be almost completely restored by subsequent heating of the catalyst in dry hydrogen at 550 "C followed by outgassing at the same temperature. The adsorption of water may proceed according to the following mechanism : H2O (gas) OH- HO- Cr3+02- -+ HO- Cr3+ OH-. t I Torr = (101 327/760) Pa.C . A. McAULIFFE AND F. M. ASHMAWY 1085 I 1 0 1000 2000 3000 LOOO 5000 G Fig. 1. Electron spin resonance spectrum of chromia-alumina (5.02% Cr203). (a) Catalyst reduced with dry H, at 500 "C (Cr2+ and Cr3+ produced). (6) Catalyst after ethylene hydrogenation at 100 "C. 0 1000 2000 3000 LOOO 5000 G Fig. 2. Electron spin resonance spectrum of chromia-alumina (5.02% Cr,O,).(a) Catalyst reduced with H,+water vapour at 500 "C (only Cr3+ produced). (6) Catalyst after ethylene hydrogenation at 100 "C. It is known that Cr3+ is a hard acid and 02- is a hard base;16 therefore, one would expect water, which is also a hard base, to adsorb at Cr3+ sites, thus completely blocking access to the active sites. In a second series of experiments an attempt was made to look at the e.s.r. signal ascribed to Cr3+ (g = 1.986) under reaction conditions. Both the reduced and the oxidized catalysts were exposed to a 1 : 1 mixture of hydrogen and ethylene at 100 "C. The results are shown in fig. 1 and 2. Examination of the e.s.r. spectra of the reduced sample, fig. 1, shows an increase in the number of Cr3+ sites after hydrogenation.This increase (by ca. 30%) is probably1086 HYDROGENATION OF ETHYLENE 2L 22 20 18 16 1L s? E 6 12 -.- Q 10 8 6 L 2 0 (1:4) (1 :3) (1 : 2 ) I I I ~ I I ~ ~ ~ I ~ J 1 2 3 L 5 6 7 8 9 1 0 1 1 1 2 time/min Fig. 3. Rates of hydrogenation of ethylene at - 78 "C on 5.02% chromia-alumina at different initial hydrogen pressures. Numbers in parentheses show the ratio of ethylene to hydrogen initial partial pressures. caused by oxidative adsorption of hydrogen on Cr2+ to give Cr3+ (H;), which has been proposed as occurring on c h r ~ m i a l ~ - ~ ~ and on chromia-aluminall at I00 "C. On the other hand, the e.s.r. signal of the oxidized catalyst was reduced by ca. 15% when exposed to a 1 : 1 mixture of hydrogen and ethylene, fig. 2. This could be due to some ethylene molecules remaining irreversibly adsorbed on Cr3+ sites.ll KINETIC MEASUREMENTS Preliminary kineticmeasurements show that the initial rate ofethylene hydrogenation is temperature dependent.At -78 "C the reaction rate is given by u = kPR2 P k 2 H I u = kPk, PEpH4* while at higher temperatures (> 0 "C) the reaction rate is given byC. A. MCAULIFFE AND F. M. ASHMAWY 1087 20 19 F 18 17 16 15 14 -" 13 3 12 X g- 11 2 10 - 9 8 2 7 6 5 4 3 2 1 - / - 78°C I 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 time/min Fig. 4. Plot of first-order kinetics of ethylene hydrogenation at different temperatures on 5.02% chromia-alumina. The zeroth-order rate of the reaction in hydrogen at -78 "C is demonstrated in fig. 3, while fig. 4 shows a plot of the overall first-order kinetics at different temperatures.Note that at higher temperatures (> 0 "C) the kinetics of the reaction suggest considerable saturation of the active Cr3+ sites with ethylene, as demonstrated by the zeroth-order rate in this reactant and the e.s.r. spectra, fig. 2, with the result that the rates of hydrogenation are lower at higher temperatures than at -78 "C. To summarize, catalysts containing known amounts of Cr2+ and Cr3+ on alumina were prepared and their activities were tested in the hydrogenation of ethylene, which showed that the preferred active site is Cr3+. Kinetic studies show that the rate of hydrogenation is temperature dependent and the e.s.r. measurements reveal consider- able changes in the intensities of the signal ascribed to Cr3+ under reaction conditions.R. L. Burwell Jr, A. B. Littlewood, M. Cardew, G. Pass and C. T. H. Stodart, J. Am. Chem. SOC., 1960,82, 6272. L. L. Van Reijen, W. H. Sachtler, P. Cossee and D. Brouwer, Proc. 3rd Znt. Conf. Catalysis (North Holland, Amsterdam, 1964), vol. 2, p. 829. R. L. Burwell Jr, G. L. Haller, K. C. Tayler and J. F. Read, Ado. Catal., 1969, 20, 8. J. Masson and B. Delmon, Proc. 5th Znt. Conf. Catalysis (North Holland, Florida, 1972), vol. 1, p. 183. F. M. Ashmawy, J. Appl. Chem. Biotechnol., 1977, 27, 137. V. Indovina, A. Cimino and M. Valid, 6th Znt. Conf. Catalysis (The Chemical Society, London, 1976), vol. 1, p. 216. ' J. Masson, J. Bonnier, P. Duvigne and B. Delmon, J. Chem. Soc., Faraday Trans. I , 1977,73, 1471. Y. S. Khodakov, P. A. Makarov, G. Delzer, and Kh. M. Minachev, J. Catal., 1980,61, 184. Y. Iwasawa and S. Ogasawara, Chem. Lett., 1980, 127.1088 HYDROGENATION OF ETHYLENE lo P. P. M. M. Wittgen, C. Groenenveld, P. J. C. J. M. Zwaans, H. J. B. Morgenstern, A. H. van l1 F. M. Ashmawy and H. M. Steiner, J. Chem. SOC., Faraday Trans. I , 1977, 73, 1646. l2 F. M. Ashmawy, J . Chem. Soc., Faraday Trans. I , 1980, 76, 2096. l3 F. D. Richardson and J. H. Jeffers, J . Iron Steel Inst., London, 1948, 160, 261. l4 C. G. Maier, U.S. Bur. Mines, Bull., 1942, 436. l5 D. O'Reilly and D. MacIver, J . Phys. Chem., 1962, 66, 276. l6 F. Basolo and R. G. Pearson, Mechanism of Inorganic Reactions (Wiley, New York, 1967), chap. 1 . l7 D. Dowden and D. Wells, Proc. 2nd Int. Con$ Catal. (Technip, Paris, 1960), vol. 2, p. 1499. Is R. L. Burwell Jr and K. S. Stec, J. Colloid Interface Sci., 1977, 58, 54. Heughten, C. J. M. van Heumen and G. C. A. Schiut, J. Catal., 1982, 77, 360. P. W. Selwood., J. Am. Chem. Soc., 1966, 88, 2676. (PAPER 3/ 1028)

ISSN:0300-9599    

DOI:10.1039/F19848001083

出版商:RSC    

年代:1984

数据来源: RSC