摘要:

THE ROYAL SOCIETY OF CHEMISTRY Journal of the Chemical Society Fa rada y Transactions Scientific Editor Dr. Peter J. Sarre Department of Chemistry U n iversity of N otting ham University Park Nottingham NG7 2RD, UK Faraday Editorial Board Prof. I.W. M. Smith (Birmingham) (Chairman) Dr. M. N. R. Ashfold (Bristol) Prof. H. M. Frey (Reading) Dr. R. Aveyard (Hull) Dr. A. R. Hillman (Bristol) Prof. M. A. Chesters (East Anglia) Prof. K. R. Jeninings (Warwick) Dr. P. J. Sarre (Nottingham) __~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~ Editorial Manager and Secretary to Faraday Editorial Board Dr. Robert J. Parker The Royal Society of Chemistry Thomas Graham House Science Park Milton Road Cambridge CB4 4WF, UK Assistant Editors: Dr. A. Bingham, Mrs.F. J. O’Carroll, Miss S. Reynolds, Dr. R. A. Whitelock Editorial Secretary: Miss J. E. Chapman International Advisory Editorial Board G. C. Allen (Bristol) J. N. lsraelachvili (Santa Barbara) V. Aquilanti (Perugia) R. W. Joyner (Liverpool) R. S. Berry (Chicago) L. Kevan (Houston) A. Carrington (Southampton) S. Leach (Paris) M. Che (Paris) R. A. Marcus (Pasadena) M. S. Child (Oxford) Y. Marcus (Jerusalem) D. C. Clary (Cambridge) B. J. Orr (North Ryde) B. E. Conway (Ottawa) R. H. Ottewill (Bristol) G.Doggett (York) C. W. Outhwaite (Sheffield) G. Ertl (Berlin) J. S. Rowlinson (Oxford) G. R. Fleming (Chicago) J. P. Simons (Nottingham) R. Freeman (Cambridge) G. Somsen (Amsterdam) H.L. Friedman (Stony Brook) P. Stenius (Stockholm) P.Gray (Cambridge) S. Stolte (Amsterdam) E. Hirota (Yokohama) M. C.R. Symons (Leicester) R. F. Howe (North Ryde) J. Troe (Gottingerr) H. lnokuchi (Okazaki) C. Zhang (Dalian) Journal of the Chemical Society, Faraday Transactions (ISSN 0956-5000) is published twice monthly by The Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road, Cambridge CB4 4WF, UK. All orders accompanied with payment should be sent directly to The Royal Society of Chemistry, Turpin Transactions Ltd., Blackhorse Road, Letchworth, Herts. SG6 1HN, UK. NB Turpin Transactions Ltd., distributors, is wholly owned by the Royal Society of Chemistry. 1991 Annual subscription rate EC f538.00, Rest of World f619.00, USA $1269.00.Customers should make payments by cheque in sterling payable on a UK clearing bank or in US dollars payable on a US clearing bank. Air freight and mailing in the USA by Publications Expediting Inc., 200 Meacham Avenue, Elmont, NY 11 003. USA Postmaster: send address changes to Journa/ of the Chemical Society, Faraday Trans- actions, Publications Expediting Inc., 200 Meacham Avenue, Elmont, NY 11 003. Second class postage paid at Jamaica, NY 11431. All other dispatches outside the UK by Bulk Airmail within Europe, Accelerated Surface Post outside Europe. PRINTED IN THE UK. @ The Royal Society of Chemistry, 1991. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form, or by any means, electronic, mechanical, photographic, recording, or otherwise, without the prior permission of the publishers.INFORMATION FOR AUTHORS The Royal Society of Chemistry welcomes submission of manuscripts intended for pub- lication in two forms, Articles and Faraday Communications. These should describe original work of high quality in the sciences lying between chemistry, physics and biology, and particularly in the areas of physical chemistry, biophysical chemistry and chemical physics. Articles Full papers contain original scientific work which has not been published previously. However, work which has appeared in print in a short form such as a Faraday Communi- cation is normally acceptable. Three copies of Articles including a top copy with figures etc.should be sent to The Editor, Faraday Transactions, The Royal Society of Chem- istry, Thomas Graham House, Science Park, Milton Road, cambridge CB4 4WF, UK. Faraday Communications Faraday Communications contain novel scientific work in short form and of such importance that rapid publication is war-ranted. The total length is rigorously restricted to two pages of the double-column A4 format. The manuscript will be returned for reduction if this length is exceeded. For a Communication consisting entirely of text and ten references, with no figures, equations or tables, this corresponds to approximately 1600 words plus an abstract of up to 40 words. Submission of a Faraday Communication can be made either to The Editor, Faraday Transactions, The Royal Society of Chem- istry, Thomas Graham House, Science Park, Milton Road, Cambridge CB4 4WF.UK, or via a member of the International Advisory Editorial Board. In the latter case, the top copy of the manuscript including any figures etc., together with the name of the person through whom the Communication is being submitted, should be sent simultaneously to the Editor at the Cambridge address. Authors may wish to contact the Board member to ensure that he is available to arrange review of the manuscript within reasonable time. In order to avoid delay in publication, proofs of Communications are not sent to authors unless this is specifically requested. Full details of the form of manuscripts for Articles and Faraday Communications, con- ditions for acceptance etc. are given in issue number one of Faraday Transactions, published in January of each year, or may be obtained from the Editorial Manager. There is no page charge for papers published in Faraday Transactions. Fifty reprints are supplied free of charge. Any author who is publishing in Faraday Transactions for the first time is entitled to a free copy of the issue in which the paper appears. Dr. P. J.Sarre, Scientific Editor. Tel. : Nottingham (0602) 41 2550 (24 hours) E-MaiI(JAN ET) :P CZPS F@ U K. AC. N 0lT.VAX Fax: (0602) 41 2473 Telex: 37346 UNINOT G Dr. R. J. Parker, Editorial Manager. Tel. : Cambridge (0223) 420066 E-Mail (JANET): RSCl @,UK.AC.RL.GB Fax: (0223) 423623 or 420247 Telex: 81 8293 ROYAL G

ISSN:0956-5000    

DOI:10.1039/FT99187FX001

出版商:RSC    

年代:1991

数据来源: RSC

ISSN:0956-5000    

DOI:10.1039/FT99187BX003

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

Cumulative Author Index 199 1 Aguilar, A., 37 Cunnane, V. J., 107 Hori, Y., 125 Mandal, A. B., 133 Siders, P., 51 Alberti, M., 37 Angilella, V., 203 Bear, J., 167 Cuvier, S., 167 del Mar Graciani, M., 129 Diaz Pefa, M., 93 Hutson, J. M., 209 Irnai, K., 147 Irnai, Y., 83 Marrelli, S., 167 Martin Luengo, M. A., 199 Miheeva, L. M., 137, 141 Sing, K. S. W., 185 Smith, I. W. M., 205 Sole, A., 37 Belin, C., 203 Dollirnore, D., 161 Jamie, I. McL., 73 Moya, M. L., 129 Stead, W., 73 Bernardi, M., 31 Dollimore, J., 161 Jaroniec, M., 179 Mullally, J., 87 Sutton, R. L., 101 Borovskaya, A. A., 137, 141 Donovan, R. J., 15 Jenkins, J. D., 57 Mufioz, E., 129 Suzuki, T., 179 Breen, W., 11 5 Dyke, J. M., 19 Kakhu, A. I., 57 Murata, A., 125 Tezuka, Y., 147 Broadbent, D., 161 Brotas de Carvalho, M., Ellis, A.M., 19 Eriksson, J. C., 153 Kalugin, 0.N., 63 Kaneko, K., 179 Mutomaki, L., 107 Nair, B. U., 133 Toennies, J. P., 31 Tsuchida, E., 83 185 Evans, T. A., 161 Kaneko, M., 83 Nishikawa, K., 179 Velasco, R. M., 209 Burch, R., 193 Feakins, D., 87 Kazarna, H., 147 Ogryzlo, E. A., 45 Vigil, M. R., 93 Busca, G., 175 Carlile, C., 73 Firth, S., 1 Fujiwara, Y., 179 Kendrick, J., 209 Kevan, L., 167 Porter, K. E., 57 Prager, M. J., 73 Vong, M. S. W., 199 Vyunnik, 1. N., 63 Carrott, P., 185 Gianturco, F. A,, 31 Klissurski, D., 175 Pritchard, J., 193 Waghorne, W. E., 87,209 Cassidy, J., 115 Gomez-Herrera, C., 129 Kontturi, K., 107 Puri, M., 167 Walker, Z. H., 45 Chalker, S., 193 Grigo, M., 63 Kroto, H.W., 1 Ribeiro Carrott, M., 185 Walton, T. J., 199 Cockett, M. C. R., 19 Gubsky, S. M., 63 Langridge-Smith, P. R. R., Ridley, T., 15 Warr, J. F., 205 Compostizo, A., 93 Gulaeva, N. D., 137, 141 15 Rubio, R. G., 93 White, J. W., 73 Cooper,A., 209 Gurnan, S. J., 209 Lawley, K. P., 15 Sanchez, F., 129 Wright, T. G., 19 Cooper, T. A., 1 Hadjiivanov, K., 175 Li, L., 167 Sandig, R., 63 Yoshinami, Y., 125 Couch, A. D., 9 Cox, A. P., 9 Haselden, G. G., 209 Holmes, A. J., 15 Ljunggren, S., 153 Lorenzelli, V., 175 Sato, M., 179 Schiffrin, D. J., 107 Yufei, C., 107 Zaslavsky, B. Yu., 137, 141 Crespo Colin, A., 93 Homer, J., 57 Lyons, M. E. G., 115 Sermon, P. A., 199 i The following papers were accepted for publication between 17th October and 15th November 1990: Compounds with metallic copper.Part 6.4xidative C-C bond scission of 1,2-dicarbonyl on metallic copper surface studied by surface-enhanced Raman scattering G. Xue, Q. Shen and J. Dong Force constant of transition-metal-ion-ligandbonds: a molecular-orbital calculation S. U.M. Khan and Z. Y. Zhou State and localization of nickel, cobalt and copper ions adsorbed on titania (anatase) K. Hadjivanov, D. Klissurski, A. Davydov and M. Kantcheva Surface reaction of 2-mercaptobenzimidazole on metals and its application in adhesion promotion G. Xue, X. Huang and J. Ding Influence of non-ionic polymers on solvent properties of water as detected by studies G€acid-base equilibria of slllphonephthalein and fluorescein dyes B.Y.Zaslavsky, L.M. Miheeva, N.D.Gulaeva, A. A. Borovskaya, M. I. Rubtsov, L. L. Lukatskaya and N.0.Mchedlov-Petrossyan Coulombic energy in bromide salts of mono-nitrogen organic bases J. Blazejowski and J. Lubkowski De-sintering action of phosphorus in alumina-aluminium phosphates (AAP)containing transition-metal ions (M=Cr, Mn, Fe, Co, Ni, Cu, Zn) P.J. Pomonis, D. E. Petrakis and A. T. Sdoukos Investigation of hydroxyl groups in crystalline silicoaluminophosphate S APO-34 by diffuse reflectance IR spectroscopy R. Fricke, I. Girnus, S. A. Zubkov, L. M. Kustov and V. B. Kazansky Influence of a magnetic field on line intensities in the optical spectra of free molecules D. L. Andrews and A. M. Bittner Analysis of the fast relaxation times for micelle kinetics taking into account new EMF data E.Wyn- Jones, N.Takisawa, H. Gharibi, P. Brown, M. A. Thomason, D.M. Painter, D. M. Bloor and D. G. Hall High-field NMR and molecular dynamics investigations of ALDTOL conformations in aqueous and non-aqueous solvents F. Franks, J. Dadok, R. L. Kay, J. R. Grigera and S. Ying Aggregation in dense solutions of rods M. Warner Oriented crystallization of CaCO3 under compressed monolayers. Part 1 .-Morphological studies of mature crystals S. Rajam, B. R. Heywood, J. B. A. Walker, R. J. Davey, J. D. Birchall and S. Mann Oriented crystallization of CaCO3 under compressed monolayers. Part 2.-Morphology, structure and growth of immature crystals B. R. Heywood, S. Rajam and S.Mann Fast hydrothermal ammonium exchange of zeolite Na-Y H. van Bekkum, Th. L. M. Maesen, T. G. Verburg, Z. I. Kolar and H. W. Kouwenhoven FTIR studies of carbon monoxide adsorption on platinum and palladium hydrosols: influence of pH R. P. Cooney and M. R. Mucalo Hydrogenolysis of alkanes. Part &-Modification of supported-ruthenium catalysts by vanadium pentoxide G. C. Bond and S. Flamerz Hydrogenolysis of alkanes. Part 7.-Hydrogenolysis of propane and of n-butane over Ir/TiO2 and OsDi02 catalysts G. C. Bond and R. Yahya Characterisation of Con and Corn in CoAPO molecular sieves B. Kraushaar-Czarnetzki, W. G. M. Hoogervorst, R. R. Andrea, C. A. Emeis and W. H. J. Stork Infrared spectroscopic study of silica monoliths: effect of thermal history on structure C.C.Perry and X. Li An infrared and Raman study of the adsorption of NH3, pyridine, NO and NO2 on anatase C. H. Rochester, T. J. Dines and A. M. Ward (T * radicals R3N2Hal: EPR studies of the radiolysis of alkylammonium halides M. C. R. Symons, J. B. Raynor and I. J. Rowland Benzophenone sensitization of triplet oxazine and of delayed fluorescence by oxazine in acetonitrile solution F. Wilkinson, G. P. Kelly, L. F. Vieira Ferreira, V. M. M. R.Freire and M. I. Ferreira Kinetics of pyrolysis of furan J. C. Mackie and P. P. Organ Bending energy of surfactant films A. Fogden, S. T. Hyde and G. Lundberg Bimolecular hydrogen transfer over zeolites and SAPOs having the faujasite structure J. Dwyer, K. Karim and A. F.Ojo Optical and electrical AC response of polyaniline films L. M. Peter and M. Kalaji 11 Brownian motion of rod-shaped colloidal particles surrounded by electrical double layers T. G. M. van der Ven and G. A. Schumacher A Raman spectroscopic study of titanh-supported vanadia catalysts C. Hi. Rochester, T.J. Dines and A. M. Ward Micellar growth of octaethylene glycol decyl ether N. Funasaki, H-S. Shim and S. Hada Measurement of the average emission lifetimesof the A 'C+-X' I;+ and the orange arc bands of CaO J. M. C. Plane and C-F.Nien Characterizationof chabazite and chabazite-like zeolites of unusual composition S. M, Kuznicki and K. A. Thrush The properties of concentrated colloidal dispersions R. Ottewill md J. W, Goodwin Fluorescence quenching of aromatic molecules by potassium iodide and potassium bromide in methanol-ethanol solutions J.Najbar and M. Mac Recognition of substrates by membrane potential of immobilized enzyme membranes A. Higuchi, S. Ogawa and T.Nakagawa Thermodynamicsof adsorption from solution A. M. Goncalves da Silva, V. A. M. Soares and J. C. G. Calado Early stages of Fischer-Tropsch activity on Tho2and Tho.9Ceo.102 R. C. Jarnagin and A. Grabbe The detection of chlorophosphaethyne, Cl-C=P, by photoelectron spectroscopy H. W. Kroto, T. J. Dennis, G. Y.Matti, C-Y.Mok, R. J. Suffolk and D. R. M. Walton Dielectric relaxation studies of solid solutions of monofluorobenzaldehydesin a polymer: Eyring parameters for aldehyde group rotation B. J. McClelland and A.Wigg Nickel catalysts: passivation and reactivation M. Montes, A. Diaz and A. Gil Liquid-phase adsorption of ethanol-water mixtures on NaZSM-5 zeolite with inorganic and organic binders W-D. Einicke, U.Messow, M. Heuchel, P. Brauer and R.Schollner Liquid-phase adsorption of binary ethanol-water mixtures on high-silica adsorbents with different structures W. D. Einicke, W. Reschetilowski, M. Heuchel, M. v. Szombathely, P. Brauer, R. Schollner, W. Schwieger and K-H. Bergk Evolution of the cluster size distribution in stirred suspensions R. D. Cohen Effect of Zn substitutions in YBa2Cu30~~ phases on reactivity during the ammoxidation of toluene J. C. Otamiri, A. Andersson, S. L.T. Andersson, J. E. Crow and Y.Gao Photodissociation of Ne'2 M.Larsson, L. Brostrom, S. Mannervik, R.T. Short and D. Sonnek Time-resolved evanescent wave-induced fluorescence spectroscopy. Part 1.-Deviations in the fluorescence lifetime of tetrasulphonated aluminium phthalocyanine at a fused silica/methanol interface G. Rumbles, A. J. Brown and D. Phillips Rotational spectrum of phosphonium iodide vapour A. C. Legon, N. W. Howard and G. J. Luscombe Nature of the passive film on Fe-Cr alloys as studied by '*O/SIMS: reduction of the prior film and stability to ex situ surface analysis J. A. Bardwell, G. I. Sproule, D. F. Mitchell, B. MacDougall and M. J. Graham Electronic states of some simple ethers studied by vacuum ultraviolet absorption and near-threshold electron-energy-loss spectroscopy I. C.Walker, L.Bremner and M. G. Curtis Characterisation of vanadia/silica mixed-gel catalysts by EPR spectroscopy at T = 110 K A. Wokaun, K. L. Walther and A. Baiker NMR study of ion-molecule interactions. Part 6.-25Mg NMR of magnesium@) ion solvation in aqueous acetone and aqueous acetonitrile solutions K. Miura, H. Matsuda, S. Kikuchi and H. Fukui Electronic charge distribution and Lewis acidity of surface aluminium atoms in y-AlaO3: a quantum-chemical model M.V. Shimanskaya, M. B. Fleisher and L. 0.Golender ... 111 Journal of the Chemical Society, Faraday Transactions Information for Authors Faraday Transactions is an international journal for the publi- cation of original research papers (articles) and communications concerned with the sciences lying between chemistry, physics and biology, and particularly in the areas of physical chemistry, biophysical chemistry and chemical physics.The journal is published fortnightly. There is no page charge for papers published in Faraday Transactions. Ar tides Full papers contain original scientific work that has not been published previously. However, work that has appeared in print in a short form such as a Faraday Communication or Chemical Communication is normally acceptable. Papers should be typewritten in double spacing on one side only of the paper. Four copies of text, illustrations, tables and any other matter should be sent to: The Editor, Faraday Transactions, The Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road, Cambridge CB4 4WF, UK.Faraday Communications Faraday Communications contain novel scientific work in short form and of such importance that rapid publication is desirable. The total length is rigorously restricted to two printed A4 pages. The manuscript will be returned for reduction if this length is exceeded. For a communication consisting of text and ten references, with no figures, equations or tables, this cone- sponds to approximately 1600 words plus an abstract of up to 40 words. Submission of a Faraday Communication can be made either to The Editor, Faraday Transactions, The Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road, Cambridge CB4 4WF, UK, or via a member of the Inter- national Advisory Editorial Board.In the latter case, the top copy of the manuscript including any figures etc., together with the name of the person to whom the communication is being submitted, should be sent simultaneously to The Editor at the Cambridge address. Authors may wish to contact the Board member to ensure that he is available to arrange review of the manuscript within reasonable time. In order to avoid delay in publication, proofs of communi- cations are not sent to authors unless this is specifically re- quested. Administration Receipt of a paper will be acknowledged, and the paper will be given a reference number which authors are asked to quote on all their subsequent correspondence. If no such acknow-ledgement has been received after a reasonable period of time authors should check with the Editorial Office as to whether the paper or the acknowledgement has gone astray.Editorial Policy. Every paper (except communications) will be submitted to at least two referees, by whose advice Faraday Editorial Board will be guided as to its acceptability. Papers that are accepted must not be published elsewhere except by permission of the Royal Society of Chemistry. Submission of a manuscript will be regarded as an undertaking that the same material is not being considered for publication by another journal. Conditions governing acceptance are printed in Issue 1 each year and are available from the Editorial Manager. Copyright. The whole of the literary matter (including tables, figures, diagrams and photographs) in Faraday Transactions is subject to copyright and may not be reproduced without per-mission from The Royal Society of Chemistry and such other owner of the copyright as may be indicated.Reprints. Fifty reprints of each paper are supplied free of charge on request. Additional reprints can be purchased if ordered at the time of publication. Details are sent to authors with the proofs of articles and with the letter of acceptance of communications. Free Copies. Any author who is publishes in Faraday Trans- actions is entitled to a free copy of the issue in which hisher paper appears. Notes on the Preparation of Papers 1 Manuscripts must be typed in double-line spacing, single sided on A4 paper, with margins at top, bottom and left-hand side of at least 4 cm.2 The first page should be set out as follows: (i) Name and address of the author to whom proofs and correspondence should be sent. (ii) Title of the paper, with capitals for the first letter of each noun and adjective only. (iii) Authors’ names, including one forename for each author. (iv) The address where the work was carried out; if thisis different from the current address a footnote indicating the present address of this author should be included. Present addresses of other authors are not normally given. (v) Abstract, preceded and followed by a horizontal line, and typed in double-line spacing. 3 Suitable headings and sub-headings should be used in the main text as appropriate (except for communications in which no headings are used).References should be numbered serially in the text by means of superscript arabic numerals. 4 Bibliographic references (not footnotes) should follow the main text and should have the following format: 1 R. M. Barrer and R. J. B. Craven, J. Chem. SOC., Faraday Trans., 1990,86, 545. 2 R. M. Barrer and R. J. B. Craven, in New Develop-ments in Zeolite Science and Technology, ed. Y. Mu-rakame, A. Iijima and J.W. Ward, Kodansha, Tokyo, 1986, p. 521. Journal titles should be abbreviated according to the Chemical Abstracts Service Source Index (CASSI). 5 Tables should be typed on separate sheets at the end of the manuscript. 6 Diagrams should be accompanied by a separately typed set of captions.Extensive identifying lettering should be placed in the caption rather than on lines on graphs, etc. Original art- work should be supplied wherever possible. Colour photo- graphs will be accepted only when a black-and-white photograph fails to show some vital feature. 7 Bulk information (such as primary kinetic data, computer programs and output, etc.) which accompanies papers published in Faraday Transactions may be deposited, free of charge, with the Society’s Supplementary Publications Scheme, either iv at the request of the author and with the approval of the referees or on the recommendation of the referees with the approval of the author. Details are available from the Editorial Office.Nomenclature Current IUPAC nomenclature and symbolism should be used. Attention is drawn to the following publications in which the rules themselves and guidance on their use are given: Nomenclature of Organic Chemistry, Sections A, B, C, D, E, F and H, Pergamon Press, Oxford, 1979 edn. Nomenclature of Inorganic Chemistry, Blackwell Scientific Publications, Oxford, 1990. Biochemical Nomenclature and Related Documents, The Biochemical Society, London, 1978. Compendium of Chemical Terminology: IUPAC Recom-mendatiom, Blackwell Scientific Publications, Oxford, 1987. Units and Symbols The recommendations of IUPAC should be followed. Their basis is the ‘Systkme Internationale d’Unit6s’ (SI). A detailed treatment is given in the so-called Green Book: Quantities, Units and Symbols in Physical Chemistry, Blackwell Scientific Publications, Oxford, 1988 edn.V JOURNALS OF THE ROYAL SOCIETY OF CHEMISTRY Refereeing Procedure and Policy (1991) 1.0 Contributions to Dalton, Perkin, and Faraday Transactions, J. Mater. Chem., The Analyst, J. Anal. At. Spectrom. and J. Chem. Research 1.1 Introduction This document summarises the procedure used for assessing papers submitted to the four Transactions, J. Mater. Chem., The Analyst, J. Anal. At. Spectrom., and J. Chem. Research, and provides guidelines for referees engaged in this assessment. 1.2 Subject matter Papers are submitted to the various journals according to subject matter. If a referee feels that a paper would be published more appropriately in an RSC journal other than the one suggested by the author, he should inform the Editor. The topics covered by the various journals are as follows: Dalton Transactions (Inorganic Chemistry).All aspects of the chemistry of inorganic and organometallic compounds, including bioinorganic chemistry and solid-state inorganic chemistry; the applications of physicochemical techniques to the study of their structures, properties, and reactions, including kinetics and mechanism; new or improved experimental techniques and syntheses. Faraday Transactions (Physical Chemistry and Chemical Physics). Gas-phase kinetics and dynamics; molecular beam kinetics and spectroscopy, photochemistry and photophysics; energy transfer and relaxation processes: laser-induced chemistry; spectroscopies of molecules, molecular and gas- phase complexes: quantum chemistry and molecular structure, statistical mechanics of gaseous molecules and complexes; spectroscopies, statistical mechanics and quantum theory of the condensed phase, computational chemistry and molecular dynamics; colloid and interface science, surface science, physisorption and chromatographic science, chemisorption and heterogeneous catalysis, zeolites and ion-exchange phenomena; electrode processes, liquids and solutions; solid-state chemistry (microstructures and dynamics); reactions in condensed phases; physical chemistry of macromolecules and polymers; materials science; thermodynamics; biophysical chemistry and radiation chemistry.Perkin Transactions 1 (Organic Chemistry). All aspects of organic and bio-organic chemistry. These include synthetic organic chemistry of all types, organometallic chemistry, chemistry and biosynthesis of natural products, the relationship between molecular structure and biological activity, the chemistry of polymers and biological macromolecules, and medicinal and agricultural chemistry where there is originality in the science. Perkin Transactions 2 (Physical Organic Chemistry). Physicochemical aspects of organic, organometallic, and bio- organic chemistry, including kinetic, mechanistic, structural, spectroscopic, and theoretical studies. Such topics include structure-activity relationships and physical aspects of biological processes and of the study of polymers and biological macromolecules.Journal of Materials Chemistry. The chemistry of materials, particularly those associated with advanced technology; modelling of materials; synthesis and structural characteris- ation; physicochemical aspects of fabrication; chemical, structural, electrical, magnetic and optical properties; applic- ations. The Analyst (Analytical Chemistry). Theory and practice of all aspects of analytical chemistry, fundamental and applied, inorganic and organic, including chemical, physical, and biological methods. Journal of Analytical Atomic Spectrometry. The develop- ment and analytical application of atomic spectrometric techniques.Journal of Chemical Research. All areas of chemistry. The format of this journal (one- or two-page printed synopsis in Part S, plus microform version of authors’ full text typescript in Part M) makes it particularly suitable for papers containing lengthy experimental sections or extensive data tabulations. 1.3 Procedure Each manuscript is considered independently by two referees. The referees’ reports constitute recommendations to the appropriate Editorial Board, which is empowered to take final action on manuscripts submitted. The Editor, acting for the Editorial Board, is responsible for all administrative and executive actions, and is empowered to accept or reject papers. It is the Editor’s duty to see that, as far as possible, agreement is reached between authors and referees; although the referees may need to be consulted again concerning an author’s reply to comments, further refereeing will be avoided as far as possible.1.3.1 Adjudication of disagreements. If there is a notable discrepancy between the reports of the two referees, or if .the difference between authors and referees cannot be resolved readily, a third referee may be appointed as adjudicator. In extreme cases, differences may be reported to the appropriate Editorial Board for resolution. When a paper is recommended for rejection by referees, the Editor will inform the authors and return the top copy of the manuscript. Authors have the right to appeal to the Editorial Board if they regard a decision to reject as unfair. The Editor may refer to the Editorial Boards any papers which have been recommended for acceptance by the referees, but about which the Editor is doubtful. 1.3.2 Anonymity.The anonymity of referees is strictly preserved, and reports should be couched in terms which do not disclose the identity of the writer. A referee should never communicate directly with an author, unless and until such action has been sanctioned by the Society, through the Editor. 1.3.3 Confidentiality. A referee should treat a paper received for assessment as confidential material. Information acquired vi by a referee from such a paper is not available for citation until the paper is published. 1.4 Policy The primary criterion for acceptance of a contribution for publication is that it should advance scientific knowledge significantly. Papers that do not contain new experimental results may be considered for publication only if they either reinterpret or summarise known facts or results in a manner presenting an advance in chemical knowledge.Papers in interdisciplinary areas are acceptable if the chemical content is considered satisfactory. Papers reporting results regarded as routine or trivial are not acceptable in the absence of other, desirable attributes. Although short papers are acceptable, the Society strongly discourages the fragmentation of a substantial body of work into a number of short publications; such fragmentation is likely to be grounds for rejection.The length of an article should be commensurate with its scientific content; however, authors are allowed every latitude (consistent with reasonable brevity) in the form in which their work is presented. Figures and flow-charts can often save space as well as clarify complicated arguments, and should not be excised unless they are unhelpful or really extrava- gant. If a paper as a whole is judged suitable for the Journal, minor criticisms should not be unduly emphasised. It is the responsibility of the Editor to ensure the use of reasonably brief phraseology, and to assist the author to present his work in the most appropriate format. However, referees should not hesitate to recommend rejection of papers which appear incurably badly composed.It should be clearly understood that referees’ reports are made in confidence to the Editor, at whose discretion comments will be transmitted to the author. To assist the Editor, referees are requested to indicate which comments are designed only for consideration, as distinct from those which, in the referee’s view, require specific action or an adequate answer before the paper is accepted. Referees may ask for sight of supporting data not submitted for publication, or for sight of a previous paper which has been submitted but not yet published. Such requests must be made to the Editor, not directly to the author. 1.4.1 Authentication of new compounds. Referees are asked to assess, as a whole, the evidence in support of the homogeneity and structure of all new compounds. No hard and fast rules can be laid down to cover all types of compounds, but the Society’s policy is that evidence for the unequivocal identification of new compounds should wherever possible include good elemental analytical data; for example, an accurate mass measurement of a molecular ion does not provide evidence of purity of a compound and must be accompanied by independent evidence of homogeneity.Low-resolution mass spectrometry must be treated with even more reserve in the absence of firm evidence to distinguish between alternative molecular formulae. Where elemental analytical data are not available, appropriate evidence which is convincing to an expert in the field may be acceptable.Spectroscopic information necessary to the assignment of structure should normally be given. Just how complete this information should be must depend upon the circumstances; the structure of a compound obtained from an unusual reaction or isolated from a natural source needs much stronger supporting evidence than one derived by a standard reaction from a precursor of undisputed structure. Referees are reminded of the need to be exacting in their standards but at the same time flexible in their admission of REFEREEING PROCEDURE AND POLICY (1 991) evidence. It remains the Society’s policy to accept work only of high quality and to permit no lowering of stand- ards. 1.5 Titles and summaries Referees should comment on Titles and Summaries with the following points in mind.Titles of papers are used out of context by several organisations for current awareness purposes. To enable such systems to serve chemists adequately, titles must be written around a sufficient number of scientific words carefully chosen to cover the important aspects of the paper. Summaries should preferably be self-contained, so that they can be understood without reference to the main text. 1.6 Speed of Refereeing The Editorial Boards are anxious to maintain and to reduce further if possible the publication times now being achieved. In this connection, referees should submit their reports with the minimum of delay. or return manuscripts immediately to the Editor if long delay seems inevitable.1.7 Suggestion of Alternative Referees The Editor welcomes suggestions of alternative referees competent to deal with particular subject areas. Such suggestions are particularly helpful in cases where referees consider themselves ill-equipped (in terms of specialist knowledge) to deal with a specific paper, and in highly specialized or new areas of research where only a limited number of experts may be available. If, in such a case, the alternative and the original referee work in the same institution, the manuscript may be passed on directly after informing the Editor. 1.8 Notes (Short Papers) and Letters ‘Notes’ are published in Dalton Transactions; the correspond- ing format in The Analyst and J. Anal. At. Spectrom.is referred to as a ‘Short Paper’. These articles are intended for the description of essentially complete pieces of work which are not of the length to justify a full paper. They are NOT preliminary communications, nor in any way an alternative to Chemicul Communications, for which there are additional criteria of novelty and urgency. The quality of material contained in a Note (Short Paper) should be the same as that in a full paper. Investigations arising out of some larger project but not prosecuted to the same degree are particularly appropriate for this format. A Note (Short Paper) should not normally exceed in length about 8 pages of typescript, including figures, tables, etc. It should comprise a short abstract (except in The Analyst and J.Anal. At. Spectrom.) and Discussion, but adequate experimental details are required. In J. Chem. Research, a ‘Short Paper’ is essentially of the same type. As a consequence of its length, it appears in full in Part S with no microform version in Part M. ‘Letters’, published only in Dalton Transactions, are a medium for the expression of scientific opinions and views normally concerning material published in that journal; it is intended that contributions in this format should be published rapidly. The Letters section is for scientific discussion, and is not intended to compete with media for the publication of more general matters such as Chemistry in Britain. Only rarely should a Letter exceed one printed column in length (about 1-2 pages of typescript). Where a Letter is polemical in nature, and if it is accepted, a Reply will be solicited from other parties implicated, for consideration for publication alongside the original Letter.vii REFEREEING PROCEDURE AND POLICY (1991) 1.9 Relationship with Communications Journals In cases where a preliminary report of the work described has appeared (for example in Chemical Communications), referees should alert the editor to any excessive and unnecessary repetition of material; this can arise in connection with communications journals whose restrictions on length and the reporting of experimental data are less severe than those of Chemical Communications. Furthermore, the acceptability of the full paper must be judged on the basis of the significance of the additional information provided, as well as on the criteria outlined in the foregoing sections.2.0 Contributions to Chemical Communic-ations Chemical Communications 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 in most cases by full papers in other journals, providing detailed accounts of the work. It is Society policy that only a fraction of research work warrants publication in Chemical Communications, and strict refereeing standards should be applied. The benefit to the reader from the rapid publication of a particular piece of work before it appears as a full paper must be balanced against the desirability of avoiding duplicate publication.The needs of the reader, not the author, must be considered, and priority in publication should not be allowed to determine acceptability. Acceptance should be recommended only if, in the opinion of the referee, the content of the paper is of such urgency that rapid publication will be advantageous to the progress of chemical research. The length of Communications is strictly limited; only in exceptional circumstances should it exceed one printed page (two-and-a-half to three A4 pages of typescript) and referees should be particularly critical of manuscripts longer than this. Communications do not contain extensive spectroscopic or other experimental data, but referees may ask for sight of such data before reaching a decision.The refereeing procedure for Communications is the same as that for full papers, except that rapidity of reporting is crucial in order to maintain rapid publication. The Journals Committee functions as the Editorial Board of Chemical Communications and as such acts as final arbiter in cases of dispute. 3.0 Communications submitted to The Analyst and J. Anal. At. Spectrom. Criteria for acceptance of Communications submitted to The Analyst and J. Anal. At. Spectrom. are similar to those for contributions to Chemical Communications, except that they should be concerned specifically with analytical chemistry.However Communications to The Analyst and J. Anal. At. Spectrom. are not subjected to refereeing in the usual way; a decision whether or not to publish rests with the Editor, who may or may not obtain advice from a referee. 4.0 Communications submitted to Perkin or Faraday Transactions or J. Mater. Chem. Criteria for acceptance of Communications submitted to Perkin or Faruday Transactions or J. Muter. Chem. are similar to those for contributions to Chemical Communications, except that the work will be of more specialist interest. For Perkin Communications inclusion of key experimental data is expected. Assessment is carried out by a small nucleus of referees, consisting largely of members of Perkin Editorial Board or of the Faraday or Materials International Advisory Editorial Board, as appropriate.5.0 Contributions to Mendeleev Communic- ations Mendeleev Communications, published jointly by the Royal Society of Chemistry and the USSR Academy of Sciences, is a sister publication to Chemical Communications, containing preliminary reports of the same type, in any area of chemistry. The majority of contributions are from Soviet authors. Assessment involves two stages of refereeing. Manuscripts submitted to the Soviet Editorial Office are refereed initially by a Soviet scientist. If found acceptable they are then reviewed by Western scientists chosen by the Royal Society of Chemistry. A favourable recommendation at this stage, from one referee, is sufficient authority for acceptance.If the recommendation is unfavourable, however, a second RSC referee is consulted; two unfavourable reports are required for rejection. Manuscripts submitted to the UK Editorial Office undergo this two-stage refereeing process in reverse. 6.0 X-Ray Crystallographic Work 6.1 Crystallographic papers are of two types: (A) The majority, which contain definitive data on completely refined determinations. (B) A minority which include brief accounts of structures containing feature(s) of unusual interest and where the structure solutions are clear but where (for any of a variety of reasons) the full refinement has not been completed. These are then regarded as preliminary publications, at least so far as the X-ray results are concerned.Both types of publication are appropriate for Chem. Commun.; only those of type (A) should normally appear in Dalton or Perkin Transactions. 6.2 Papers of type (A) in Dalton and Perkin Transactions should normally contain the information in their titles that an X-ray structure determination has been carried out; this is often appropriate in Chem. Commun. also, but not obligatory. Papers of type (B) need not do so if the X-ray determination forms only a minor part. Summaries should always contain this information unless the paper is of type (B) and the structure determination is not a main point of the communication. 6.3 All papers containing crystallographic determinations will be refereed by two referees, one a structural chemist.If the editor considers it advisable, the paper may also be sent to a crystallographer for comment. Referees will not normally be expected to check values of structural parameters for publication (eg. bond lengths and angles against atomic co- ordinates; this will be done after publication by CCDC or Bonn), but should still pay attention to the quality of the experimental crystallographic work. However their primary concern should be such new chemistry as is involved in the structure. 6.4 On occasions Chern. Commun. will publish preliminary accounts [type (B)] of crystal structures of unusual chemical interest. By ‘preliminary’ is meant that the data have not yet been fully refined. Sufficient supplementary data must be provided for the referee to judge whether the ‘not-fully-refined’ structure does indeed prove the desired point, and care should be taken by the referees to ensure that the authors do not overstate the case they have-for example by reporting bond ...Vlll lengths to very high degrees of apparent precision when they have poor R-factors. Such papers will always be refereed by a professional crystallographer. Authors must indicate in the paper or the supplementary data the justification for publishing without full refinement and referees should comment on whether the case for publication is convincing. 6.5 In many cases the structure referred to in Chem. Commun. will be fully refined. The Chern. Commun. can then be considered to fulfil the archival function, and the structure determination may not require further detailed refereeing when presented as part of a full paper.In the full paper, the author’s purpose will then be served by a simple reference back to the original communication. However, if the crystallography is discussed again at any length in the full paper, the data should be re-presented to the referees in full, and re-published if considered necessary. REFEREEING PROCEDURE AND POLICY ( 1991) 6.6 There may be other cases when an author wishes to publish a paper in Dalton or Perkin in which the result of a crystal structure determination is discussed, but in which details or extensive discussion are considered unnecessary. The crystallographer may even be omitted as a co-author (for example when the determination is carried out by a commercial company). If the author is able to show the referees that this procedure is appropriate, it will be allowed provided that it does not lead to unnecessary fragmentation.However, the author must provide, as supplementary information, sufficient data relating to the crystal structure determination to allow a referee to make sure that the point made is correct, and co- ordinates etc. will be deposited with CCDC (or Bonn). The brief published description of the determination should be supplemented by appropriate reference to ‘unpublished work’. ix INSTRUCTIONS FOR AUTHORS (1991) APPENDIX IUPAC Publications on Nomenclature and Syrnbolism 1.O Compilations A one-letter notation for amino-acid sequences (1968) 1.1 Nomenclature of Organic Chemistry, a 550-page Abbreviations and symbols for the description of the hardcover volume published in 1979, available from Pergamon, conformation of polypeptide chains (1969) Oxford.Nomenclature of peptide hormones (1974) Section A: Hydrocarbons Recommendations for the nomenclature of human im- Section B: Fundamental heterocyclic systems munoglobulins Section C: Characteristic groups containing carbon, hy- Protein data bank. A computer-based archival file for drogen, oxygen, nitrogen, halogen, sulphur, macromolecular structures (1977) selenium, and tellurium Nomenclature of multiple forms of enzymes (1976) Section D: Organic compounds containing elements not Nucleotides and nucleic acids exclusively those referred to in the title of Abbreviations and symbols for nucleic acids, polynuc- Section C leotides and their constituents (1970) Section E: Stereochemistry Lipids Section F: General principles for the naming of natural Nomenclature of lipids (1 976) products and related compounds Section H: Isotopically modified compounds Nomenclature of steroids (1967) Nomenclature of quinones with isoprenoid side chains 1.2 Nomenclature of Inorganic Chemistry, a 278-page (1 973) hardcover volume published in 1990, available from Blackwell Tentative rules for the nomenclature of carotenoids (1970).Scientific Publications, Oxford. Amendments (1974) Chapter 1: General aims, functions and methods Nomenclature of tocopherols and related compounds Chapter 2: Grammar (1 973) Chapter 3: Elements, atoms, and groups Carbohydrates, etc.Chapter 4: Formulae Tentative rules for carbohydrate nomenclature. Part 1 Chapter 5: Names based on stoichiometry (1969) c Chapter 6: Neutral molecular compounds Nomenclature of cyclitols (1 973) Chapter 7: Names for ions, substituent groups and Phosphorus-containing compounds radicals, and salts Nomenclature of phosphorus-containing compounds of Chapter 8: Oxoacids and derived anions biochemical importance (1976) Chapter 9: Co-ordination compounds Miscellaneous Chapter 10: Boron hydrides and related compounds Trivial names of miscellaneous compounds of importance in biochemistry (1965) 1.3 Biochemical Nomenclature and Related Documents, a Nomenclature and symbols for folk acids and related 220-page softcover manual published in 1978 by The compounds (1 965) Biochemical Society for IUB, and available from the Nomenclature for vitamins B-6 and related compounds Biochemical Society Book Depot, PO Box 32, Commerce (1973) Way, Colchester, Essex C02 8HP. The contents are as Nomenclature of corrinoids (1973) follows: General 1.4 Compendium of Analytical Nomenclature, a 280-page Nomenclature of organic chemistry.Section E: Stereo-hardcover volume published in 1987, available from Blackwell chemistry (1 974) Scientific Publications, Oxford. The contents are as follows: Nomenclature of organic chemistry. Section F: Natural Presentation of the Results of Chemical Analysis products and related compounds (1976) Solution Thermodynamics (activity coefficients, equilibria, Nomenclature of organic chemistry.Section H: Isotopically PH)modified compounds (1977) Recommendations for Terminology to be used with Isotopically labelled compounds: common biochemical Precision Balances practice Recommendations for Nomenclature of Thermal Analysis Recommendations for measurement and presentation of Recommendations for Nomenclature of Titrimetric An- biochemical equilibrium data (1976) alysis Abbreviations and symbols for chemical names of special Electrochemical Analysis interest in biological chemistry (1965) Analytical Separation Processes (precipitation, liquid- Abbreviations and symbols: a compilation (1 976) liquid distribution, zone melting and fractional crystalliz- Citation of bibliographic references in biochemical ation, chromatography, ion exchange) journals (197 1) Spectrochemical Analysis (radiation sources, general Amino acids, peptides and proteins atomic emission spectroscopy, flame spectroscopy, X-ray Nomenclature of a-amino acids (1974) emission spectroscopy, molecular methods) Symbols for amino-acid derivatives and peptides (1971) Recommendations for Nomenclature of Mass Spec-Rules for naming synthetic modifications of natural trometry peptides (1966) Recommendations for Nomenclature of Radiochemical Abbreviated nomenclature of synthetic polypeptides or Methods polymerized amino acids (1971) Surface Analysis (including photoelectron spectroscopy) X INSTRUCTIONS FOR AUTHORS (1991) 1.5 Compendium of Chemical Terminology: IUPAC Recommendations, a 454-page volume published in 1987, available in hardcover and softcover from Blackwell Scientific Publications, Oxford.1.6 Quantities, Units, and Symbols in Physical Chemistry, a 134-page hardcover volume published in 1988, available from Blackwell Scientific Publications, Oxford. 2.0 Documents not included in the compil-ations 2.1 Nomenclature of Elements and Compounds 2.1.1 Amino acids and Peptides Nomenclature and symbolism for amino acids and peptides (Pure Appl. Chem., 1984, 56, 595; Eur. J. Biochem., 1984, 138, 9).2.1.2 Analytical Reagents Guide to trivial names, trade names, and synonyms for substances used in analytical chemistry (Pure Appl.Chem., 1978, 50, 339). 2.1.3 Boron Compounds Nomenclature of inorganic boron compounds (Pure Appl. Chem., 1972,30, 681). 2.1.4 Carbohydrates Conformational nomenclature for five- and six-membered ring forms of monosaccharides and their derivatives (provisional) (Pure Appl. Chem., 1981,53, 1901; Eur. J. Biochem., 1980, 111, 295). Abbreviated terminology of oligosaccharide chains (provis- ional) (Pure Appl. Chem., 1982, 54, 1517; J. Biol. Chem., 1982, 257, 2347). Polysaccharide nomenclature (provisional) (Pure Appl. Chem., 1982, 54, 1523;J. Biol. Chem., 1982, 257, 3352). Nomenclature of unsaturated monosaccharides (provisional) (Pure Appl. Chem., 1982, 54, 207; Eur. J. Biochem., 1981, 119, 1; errata Eur.J. Biochem., 1982,125, 1). Nomenclature of branched-chain monosaccharides (provis-ional) (Pure Appl. Chem., 1982, 54, 21 1; Eur. J. Biochem., 1981, 119, 5; errata Eur. J. Biochem., 1982, 125, 1). Symbols for specifying the conformation of polysaccharide chains (provisional) (Pure Appl. Chem., 1983, 55, 1269; Eur. J. Biochem., 1983, 131,5). 2.1.5 Delta Convention Nomenclature for cyclic organic compounds with contiguous formal double bonds (Pure Appl. Chem., 1988,60,1395). 2.1.6 Elements Recommendations for the names of elements of atomic number greater than 100 (Pure Appl. Chem., 1979,51, 381). 2.1.7 Enzymes Enzyme Nomenclature (1984), published by Academic Press in hardcover and softcover editions. 2.1.8 Folic Acid Nomenclature and symbols for folic acid and related compounds (Pure Appl.Chern., 1987,59, 833; Eur. J. Biochem., 1987,168,251). 2.1.9 Glycoproteins Nomenclature of glycoproteins, glycopeptides, and peptido- glycans (Pure Appl. Chem., 1988,60, 1389). 2.1.10 Heterocyclic Compounds Revision of the extended Hantzsch-Widman system of nomenclature for heteromonocycles (Pure Appl. Chem., 1983, 55,409). 2.1.11 Hydrogen Names for hydrogen atoms, ions, and groups, and for reactions involving them (Pure Appl. Chem., 1988,60, 11 15). 2.I. 12 Isotopically Modijed Compounds Nomenclature of inorganic chemistry. Part 11. 1. Isotopically modified compounds (Pure Appl. Chem., 198 1,53,1887). 2.1.13 Lambda Convention Treatment of variable valence in organic nomenclature (Pure Appl.Chem., 1984,56,769). 2.1.14 Nitrogen Hydrides Nomenclature of hydrides of nitrogen and derived cations, anions, and ligands (Pure Appl. Chem., 1982,54,2545). 2.1.15 Nucleotides Abbreviations and symbols for the description of conformations of polynucleotide chains (provisional) (Pure Appl. Chem., 1983, 55, 1279; Eur. J. Biochem., 1983, 131,9). 2.1.16 Numerical Terms Extension of Rules A-1.1and A-2.5 concerning numerical terms used in organic chemical nomenclature (Pure Appl. Chem., 1986, 58, 1693). 2.1.17 Polymers Nomenclature of regular single-strand organic polymers (Pure Appl. Chem., 1976,48, 373). Nomenclature for regular single-strand and quasi single-strand inorganic and co-ordination polymers (Pure Appl. Chem., 1985, 57, 149).Source-based nomenclature for copolymers (Pure Appl. Chem., 1985,57, 1427). Stereochemical definitions and notations relating to polymers (Pure Appl. Chem., 1981,53,733). Use of abbreviations for names of polymeric substances (Pure Appl. Chem., 1987,59,491). Basic definitions of terms relating to polymers (Pure Appl. Chern., 1974,40,477). Definitions of terms relating to individual macromolecules, their assemblies, and dilute polymer solutions (Pure Appl. Chem., 1989,61,211). A classification of linear single-strand polymers (Pure Appl. Chem., 1989,61,243). Definition of terms relating to crystalline polymers (Pure Appl. Chem., 1989,61,769). 2.1.18 Polyanions Nomenclature of polyanions (Pure Appl.Chem., 1987,59,1529). 2.1.19 Prenols Nomenclature of prenols (Pure Appl. Chem., 1987,59,683; Eur. J. Biochem., 1987,167, 181). 2.1.20 Retinoids Nomenclature of retinoids (provisional) (Pure Appl. Chem., 1983, 55, 721; Eur. J. Biochem., 1982, 129, 1). 2.1.2 1 Steroids Nomenclature of steroids (Pure Appl. Chem., 1989,61, 1783). 2.1.22 Tetrapyrroles Nomenclature of tetrapyrroles (Pure Appl. Chem., 1987, 59, 779). 2.1.23 Tocopherols Nomenclature of tocopherols and related compounds (Pure Appl. Chem., 1982, 54, 1507; Eur. J. Biochem., 1982, 123, 473). 2.1.24 Vitamins Nomenclature of Vitamin D (provisional) (Pure Appl. Chem., 1982,54, 1511; Eur. J. Biochem., 1982,124, 223). 2.1.25 Zeolites Chemical nomenclature and formulation of compositions of synthetic and natural zeolites (Pure Appl.Chem., 1979, 51, 1091). 2.2 Terminology, Symbols, and Units, and Presentation of Results 2.2.1 General Glossary of terms used in physical organic chemistry (Pure Appl. Chem., 1983,55, 1281). xi 2.2.2 Analytical Nomenclature, symbols, units, and their usage in spectroch- emical analysis. Part VII, Molecular absorption spectroscopy, U.V. and visible (Pure Appl. Chem., 1988, 60, 1449). Part X, Preparation of materials for analytical atomic spectroscopy (Pure Appl. Chem., 1988,60,1461). Recommendations for publication of papers on a new analytical method based on ion exchange or ion-exchange chromato- graphy (Pure Appl. Chem., 1980,52,2555). Recommendations for presentation of data on compleximetric indicators.1. General (Pure Appl. Chem., 1979,51, 1357). Recommendations for publishing manuscripts on ion-selective electrodes (Pure Appl. Chem., 1981,53, 1907). Recommendations on use of the term amplification reactions (Pure Appl. Chem., 1982,54,2553). Recommendations for the usage of selective, selectivity, and related terms in analytical chemistry (Pure Appl. Chem., 1983, 55, 553). Nomenclature for automated and mechanised analysis (Pure Appl. Chem., 1989,61, 1657). Nomenclature for sampling in analytical chemistry (Pure Appl. Chem., 1990,62, 1193). 2.2.3 Clinical Physicochemical quantities and units in clinical chemistry with special emphasis on activities and activity coefficients (Pure Appl.Chem., 1984,56, 567). Quantities and units in clinical chemistry (Pure Appl. Chem., 1979,51, 2451 j. Quantities and units in clinical chemistry: nebulizer and flame properties in flame emission and absorption spectrometry (Pure Appl. Chem., 1986,58, 1737). List of quantities in clinical chemistry (Pure Appl. Chem., 1979, 51,2481). 2.2.4 Colloids and Surface Chemistry Definitions, terminology, and symbols in colloid and surface chemistry. I (Pure Appl. Chem., 1972, 31, 577). TI, Hetero-geneous catalysis (Pure Appl. Chem., 1976, 46, 71). Part 1.14: Light scattering (provisional) (Pure Appl. Chem., 1983, 55, 931). Reporting experimental pressure-area data with film balances (Pure Appl. Chem., 1985,57,621). Reporting physisorption data for gas/solid systems with special reference to the determination of surface area and porosity (Pure Appl. Chem., 1985,57,603).Reporting data on adsorption from solution at the solid/ solution interface (Pure Appl. Chem., 1986,58,967). 2.2.5 Elect rochem is try Nomenclature for transfer phenomena in electrolytic systems (Pure Appl. Chem., 1981,53, 1827). Electrode reaction orders, transfer coefficients, and rate constants-amplification of definitions and recommendations for publication of parameters (Pure Appl. Chem., 1980,52,233). Classification and nomenclature of electroanalytical techniques (Pure Appl. Chem., 1976,45,81). Recommendations for sign conventions and plotting of electrochemical data (Pure Appl.Chem., 1976,45, 131). Electrochemical nomenclature (Pure Appl. Chem., 1974, 37, 499). Recommendations on reporting electrode potentials in nonaqueous solvents (Pure Appl. Chem., 1984,56,461). Definition of pH scales, standard reference values, measurement of pH and related terminology (Pure Appl. Chem., 1985, 57, 531j. INSTRUCTIONS FOR AUTHORS (1991) Interphases in systems of conducting phases (Pure Appl. Chem., 1986,58,437). The absolute electrode potential: an explanatory note (Pure Appl. Chem., 1986,58,955). Electrochemical corrosion nomenclature (Pure Appl. Chem., 1989,61, 19). 2.2.6 Kinetics Symbolism and terminology in chemical kinetics (provisional) (Pure Appl. Chem., 1981,53,753). 2.2.7 Photochemistry Recommended standards for reporting photochemical data (Pure Appl.Chem., 1984,56,939). Glossary of terms used in photochemistry (Pure Appl. Chem., 1988,60, 1055). 2.2.8 Quantum Chemistry Expression of results in quantum chemistry (Pure Appl. Chem., 1978,50, 75). 2.2.9 Reactions Nomenclature for organic chemical transformations (Pure Appl. Chem., 1989,61, 725). System for symbolic representation of reaction mechanisms (Pure Appl. Chem., 1989,61,23). Detailed linear representation of reaction mechanisms (Pure Appl. Chem., 1989,61, 57). 2.2.10 Rheological Properties Selected definitions, terminology, and symbols for rheological properties (Pure Appl. Chem., 1979,51, 1215). 2.2.11 Spectroscopy Recommendations for publication of papers on methods of molecular absorption spectrophotometry in solution (Pure Appl.Chem., 1978,50, 237). Recommendations for the presentation of infrared absorption spectra in data collections. A, Condensed phases (Pure Appl. Chem., 1978,50,231). Definition and symbolism of molecular force constants (Pure Appl. Chem., 1978,50, 1709). Nomenclature and conventions for reporting Mossbauer spectroscopic data (Pure Appl. Chem., 1976,45,211). Recommendations for the presentation of NMR data for publication in chemical journals. A, Proton spectra (Pure Appl. Chem., 1972,29,625).B, Spectra from nuclei other than protons (Pure Appl. Chem., 1976,45,217). Presentation of Raman spectra in data collections (Pure Appl. Chem., 198 1,53,1879). Names, symbols, definitions and units of quantities in optical spectroscopy (Pure Appl.Chem., 1985,57,105). A descriptive classification of the electron spectroscopies (Pure Appl. Chem., 1987,59, 1343). Presentation of molecular parameter values for i.r. and Raman intensity (Pure Appl. Chem., 1988,60, 1385). Recommendations for EPR/ESR nomenclature and conven- tions for presenting experimental data in publications (Pure Appl. Chem., 1989,61,2195). 2.2.12 Thermodynamics A guide to procedures for the publication of thermodynamic data (Pure Appl. Chem., 1972,39,395). Assignment and presentation of uncertainties of the numerical results of thermodynamic measurements (Pure Appl. Chem., 1981,53, 1805). Notation for states and processes; significance of the word ‘standard’ in chemical thermodynamics and remarks on commonly tabulated forms of thermodynamic functions (Pure Appl.Chem., 1982,54, 1239). xii MINUTES OF THE EIGHTEENTH ANNUAL GENERAL MEETING OF THE FARADAY DIVISION The eighteenth Annual General Meeting of the Faraday Division of The Royal Society of Chemistry was held at 09.00 on Thursday 5 April at Hulme Hall, Manchester with Professor R H Ottewill OBE, CChem, FRSC, FRS in the chair. 1. MINUTES The Minutes of the 17thAnnual General Meeting, which were tabled, had been printed in Faraday Transactions and were approved. 2. THE 1989 ANNUAL REPORT OF THE FARADAY DIVISION Faraday Discussion 87 on ’Catalysis by Well Characterised Materials’ was held in Liverpool on 11 -1 3 April 1989.There were 170 participants including 55 from overseas of whom 12 were from the USSR. Altogether 17 countries were represented. The Introductory Lecture was given by Professor P B Wells (University of Hull) and Professor R W Joyner was Chairman of the Organising Committee. The Surface Reactivity and Catalysis Group was associated with this Discussion. The Faraday Division Annual General Meeting was held on 12 April during this conference. Discussion 88 was held in Oxford on 11 -1 3 September and was on ’Charge Transfer in Polymeric Systems’, and was arranged in collaboration with the Electrochemistry Group. The meeting attracted 135 participants of whom 55 were from overseas, representing 17 countries. Professor R W Murray (University of North Carolina, USA) introduced the Discussion with the Electrochemistry Group Medal Lecture.The Organising Committee Chairman was Professor W J Albery and local arrangements were made by Professor A Hamnett. Symposium 25 was held on 12-14 December at the University of Warwick on ’Large Gas Phase Clusters’. There were 100 participants including 47 from overseas representing 14 countries. The Introductory Lecture was given by Professor R S Berry (University of Chicago, USA) and the Chairman of the Organising Committee was Professor K R Jennings. The 9th Joint meeting of the Division with the Deutsche Bensen Gesellschaft fur Physikalische Chemie, the Division de Chirnie Physique de la Societe Francaise de Chimie and the Associazione ltaliana di Chimica Fisica was held in Aachen, West Germany on ’Transport Processes in Fluids and Mobile Phases’.Professor G R Luckhurst represented the Division on the Organising Committee. At the Society’s Annual Congress, 4-7 April at Hull, the Division held a symposium on ’Surfactant Interactions in Collodial Systems’ convened by Dr R Aveyard and Dr W D Cooper and at the Autumn Meeting, 26-28 September at Loughborough, the topic was ’Chemistry at Interfaces’. The coveners were Professor F Wilkinson and Dr M J Jaycock and the Statistical Mechanics and Thermodynamics Group was involved with the session dealing with polymer adsorption when Lennard-Jones Lecture (sponsored by Unilever) was delivered by Professor P G De Gennes (College de France).Two meetings were co-sponsored with sister Societies, ’Applications of Neutron Scattering in Colloid and Surface Science’, with the Society of Chemical Industry and ’Sensors and their Applications’with the Institute of Physics. The Bourke Lectures were given in May by Professor D H Levy (University of Chicago, USA) on ’Spectroscopy of Amino Acids and Peptides in the Gas Phase’ at Southampton, Birmingham and Manchester. Half and full day symposia held in London featuring endowed lectures of the RSC were: 16 February at the Scientific Societies Lecture Theatre, London: ’Kinetics and Dynamics of Gas-Surface Interactions’, (in association with the Surface Reactivity and Catalysis Group and the Thin Films and Surfaces Group of the Institute of Physics), featuring the Tilden Lecture by Professor D A King (University of Cambridge).2 November at the Scientific Societies Lecture Theatre, London: ’Frontiers in High Resolution Solid State NMR Spectroscopy’ featuring the Centenary Lecture by Professor E Lippmaa (Estonian Academy of Sciences). The Faraday Lecture on ’Platinum’ was given by Professor J M Thomas as a Royal Institution Discourse. The thirteen Subject Groups affiliated to the Division again had a most active year and held meetings either alone or with other Groups or Institutions on: 0 Batteries from the Laboratory to the Market Place: A Workshop Meeting (Electrochemistry Group with the scI) Electron Transfer Reactions (Electrochemistry Group with the Organic Reactions Mechanisms Group) Graduate Students’ Meeting (Theoretical Chemistry Group) Fundamentals of Wear (Colloid and Interface Science Group with the SCI, Institute of Metals and the Institution of Mechanical Engineers) 0 Neutron and X-Ray Scattering: Complementary Techniques (Neutron Scattering Group) Atomic Mechanisms of Mass Transport in Solids (Polar Solids Group) 0 Surfaces, Ions and Clusters (Molecular Beams Group) Spring Informal Meeting (Electrochemistry Group) 0 Electroanalysis (Electrochemistry Group with the Electroanalytical Group) 0 Developments in Gas Kinetics; New Techniques, Results and their Interpretation (Gas Kinetics Group) Materials for Non-Linear and Electro-Optics (Industrial Physical Chemistry Group with the Thin Films and Surfaces Group of the Institute of Physics) Graduate Students’ Meeting (Electrochemistry Group with the Electroanalytical Group) 0 Biologically Engineered Polymers 89 (Polymer Physics Group) Carbons and Catalysis (Carbon Group with Surface Reactivity and Catalysis Group) ... Xlll Physical Aspects of Polymer Science, 25th Anniversary Meeting (Polymer Physics Group) Inorganic Particulates (Colloid and Interface Science Group) High Temperature Superconductors (Polar Solids Group with Low Temperature Group of the IOP and the Institute of Ceramics) Industrial Applications of Advanced Materials (Polar Solids Group with the IOP) Battery Meeting (Electrochemistry Group with the Society of Chemical Industry) Synchrotron Radiation and Industrial Chemistry (Industrial Physical Chemistry Group with Daresbury Laboratory) Polymers in Motion (Polymer Physics Group) Wetting and Spreading (Colloid and Interface Science Group) Electronic Structure Calculations on Large Molecules: Novel Methods and Applications (Theorectical Chemistry Group) Spectroscopy of Neutrals, Ions and Complexes (High Resolution Spectroscopy Group with Molecular Beams Group) Materials, their Synthesis, Structure, Properties and Applications (Polar Solids Group) Applied Neutron Scattering (Neutron Scattering Group) The Marlow Medal was awarded to Dr J E Bagott (Shell International Petroleum Co.Ltd.) for his experimental contributions to chemical kinetics and spectroscopy.The collaboration with physical chemistry Societies/Divisions in Europe continued in 1989 and the President attended a meeting of Officers in Aachen in September when matters of mutual interest were discussed. Programmes of the General Discussions and December Symposium were circulated to members in February and October and the annual Newsletter was distributed in February. The Division had 4016 members in 1989. Dr D A Young retired from the Scientific Editorship of Faraday Publications in 1989 and was suceeded by Dr PJ Sarre who took up his appointment in October. The past year was one of change, planning and implementation. From January 1990 Faraday Transactions I and II were merged to form a single journal published fortnightly and called ’Faraday Transactions, A Journal of Physical Chemistry and Chemical Physics’. Other innovations were the introduction of an International Advisory Editorial Board and a Faraday Communications section in the journal.3. TREASURER’S REPORT The Treasurer said that the Faraday Division accounts formed part of The Royal Society of Chemistry accounts and a separate balance sheet was not produced. The Division had conducted its activities within budget in 1989 and the conferences during the year had shown a small surplus on direct costs. There had been a large number of applications for student bursaries and the allocation for this item had been oversubscribed.The Subject Groups affiliated to the Faraday Division were all in a sound financial position. 4. ELECTIONS TO COUNCIL Members of Council elected to take office from the Society’s Annual General Meeting in 1990 were as follows: POSITON TO RETIRE President: Professor R H Ottewill 1991 Vice Presidents ProfessorJ S Rowlinson who have served Professor D H Whiff en as President Professor P Gray ProfessorN Sheppard ProfessorA D Buckingham Vice Presidents: Professor H M Frey 1991 ProfessorJ P Simons 1991 ProfessorJ M Thomas 1991 Professor R N Dixon 1992 Professor A Carrington 1993 Professor R Parsons 1993 Professor F S Stone 1993 Ordinary Members: Professor I M T Davidson 1991 Professor K J Packer 1991 Professor IW M Smith 1991 Professor D G Hall 1992 ProfessorS Leach 1992 ProfessorA J Leadbetter 1992 Dr K C Waugh 1992 Professor M A Chesters 1993 Dr D C Clary 1993 Professor D Phillips 1993 Chairman: Faraday Editorial Board Professor I W M Smith Chairman: Standing Committee on Conferences Professor F S Stone Honorary Secretary: (Professor J P Simons) Honorary Treasurer: (Professor R Parsons) Representatives on RSC Council: Professor R H Ottewill ex officio Professor A D Buckingham Secretary: Mrs Y A Fish The President thanked retiring Council Members, Professor R Freeman, Professor C H Rochester and Professor D J Tildesley and in particular, he thanked Professor A D Buckingham for his services to the Division during his term as president. 5.FUTURE ACTIVITIES The Programme of future Faraday Division Conferences was presented and the President drew attention to Discussion 92 on ’The Chemistry and Physics of Small Metallic Particles’, to be held at the Royal Institution in September 1991, to mark the 200th anniversary of the birth of Michael Faraday. xiv FARADAY DIVISION INFORMAL AND GROUP MEETINGS Molecular Beams and Dynamics Group with CCP6 Chemical Dynamics in the Time Domain To be held at the University of Oxford on 21-22 March 1991 Further information from Dr. J. M. Hutson, Department of Chemistry, University of Durham, South Road, Durham Polymer Physics Group Polymer Physics-a conference to mark the retirement of Professor A.Keller FRS To be held at the University of ktol on35April 1991 Further information from Dr. M. J. Richardson, National Physical Laboratory,Teddington,Middlesex lW11 OLW 150th Anniversary Congress Division with Analytical Division: Electrochemical Sensors Division with Dalton Division: New Electronic Materials: Synthesis, Structure and Spectroscopy To be held at lmpenal Cdkqe, Londonon 9-1 2 April 1991 Further information from Mrs Y. A. FEh, The Royal Society of Chemistry, Burlington House, London W1V OBN Colloid and Interface Science Group Computer Simulation in Colloid Science (provisional) To be held at the Scientific Societies’ LectureTheatre, London on 15th May 1991 Further information from Dr. R. Buscail, ICI plc, Corporate and Colloid Science Group, PO.Box 11, The Heath, Runcorn WA7 4QE Statistical Mechanics and Thermodynamics Group with the British Liquid Crystal Society Understanding Self-Assembly and Liquid Crystals To be held at the University of Leeds on 3-5 July 1991 Further information from Dr. N. Boden, Department of Chemistry, University of Lee&, Led LS2 9JT Polymer Physics Group with Macro Group UK Polymer Surfaces and Interfaces II To be hdd at the University of Durham on 21-26 July 1991 Further information from Dr. M.J. Riion, NationalPlysicalLaboratory,Teddington, Middlesex lW11 OLW ~~ ~ ~~ ~ ~ Polymer Physics Group Biennial Meeting To be held at the University of Leedson 9-11 September 1991 Further information from Dr. M. J. Ridson, National Physical Laboratory,Teddington,Middlesex lW11 OLW ~~~~ ~ ~ Division with the Division de Chimie Physique de la Societe Franpise de Chimie, Deutsche Bunsen Gesellschaft fur Physikalische Chemie and Associazione italiana di Chimica Fisica Synchrotron Radiation and Dynamic Phenomena To be held in Grenoble, Franceon 9-13 September 1991 Further information from SFC/Dwisioncie Chimie Physique, 10 rue Vauquelin, F-75005 Paris, France Colloid and Interface Science Group Adsorption of Surfactants and Polymers To be held at the University of Bathon 10-1 2 September 1991 Further information from Dr.W. D. Cooper, Shell Research Ltd., Thomton Research Centre, P.O. Box 1, Chester CHI 3SH Electrochemistry Group Biennial Group Informal Meeting To be held at the University of Salford on 16-1 8 September 1991 Further information from Dr.S. P. Tyfiekl,C.E.G.B., Berkeley Nudear Laboratories, Berkeley, Gloucestershire GL1 9PB ~~~~~ ~ Division Autumn Meeting: Spectroscopy in Environmental Science To be held at the University of York on 24-26 September 1991 Further information from Professor R. E. Hester, Department of Chemistry, University of York,York YO1 5DD Polymer Physics Group with the Institute of Marine Engineers Polymers in a Marine Environment To be held at the City Conference Centre,Londonon 23-25 October 1991 Further information from Rhian Bufton, Conference Organiser, Instituteof Marine Engineers, The Memorial Building,76 Mark Lane, London EC3R 7JN Division Annual Congress: Characterisation of Solids and Surfaces To be held at UMIST, Mand.lesteron 13-16 Apnl1992 Further information from Dr.J. F. Gibson, The Royal Society of Chemistry, f3urlington House, London W1V OBN xv THE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY IN ASSOCIATION WITH THE HIGH RESOLUTION SPECTROSCOPY AND MOLECULAR BEAMS GROUPS OF THE FARADAY DIVISION GENERAL DISCUSSION No. 91 Structure and Dynamics of Reactive Transition States University of Nottingham, 25-27 March 1991 aganising Committee ProfessorJ. P. Simons (Chairman) Dr M. N. R. Ashfold Dr D. C. Clary Professor R. N. Dixon In the last few years the spectroscopy, reaction dynamics and theoretical communities have combined to mount by far the most seriousassault yet seen onthe commanding heights of the "transition state" in reactive collisions (and halfcolliiions).A parallel assault for reactive cdlisions (and half-adisions) at surfaces is gathering in the foothills. The "spectroscopy of cdlisions" is likely to grow very tap@ indeed with the advent of femtosecond techniques, resonance Raman probing and multiple photon ionisation resonantly enhanced Further information may be obtained from: Professor R. Grice Professor A KleynProfessor 1. M. Mills Professor N. V. Richardson via dissociation continuum states. The Discussion will bring togetherspectroscopists,dynamiasts (both gas phase and surface) and theoreticians, assessing the scene and setting the agenda for some years to follow. Experimental and theoretical Discussion papers are sought, particuhrly concerned with (a) probing Transition State Dynamics,and (b) Oriented Systems and Surface Reaction Dynamics.Professor J. P. Simons, Department of Chemistry, University of Nottingham, University Park, Nottingham NG7 2RD. THE FARADAY DIVISION OF THE ROYAL SOCIETY OF CHEMISTRY GENERAL DISCUSSION No. 92 The Chemistry and Physics of Small Metallic Particles The Royal Institution of Great Britain, London,l8-20 September 1991 Organising Committee Professor J. M. Thomas (Chairman) Professor A. J. Leadbetter Dr. P. P. Edwards There is a growing interest, ranging from the theoretical to the preparative, in dispersedandsupportedparbdesof minute dimensions consisting of less than ten and up to several million atoms.These so called subcdloidalspecies possess a wealth of elecaonic, magnetic, chemical and electrochemical properties. They also constitute formidable challenges in regard b evoking reliable methods for their Charaderization and the determination of their internal and surface sttuctures. The Disarssionwill bring togetherexperimentalists and theoreticians from a wide range of disciplines, the aim being to identify the key attributes of presentday knowledge and desirable future trends in a Professor F. S. Stone Professor C. R. A. Catlow fieldwhict~was begun, at the colloidal level, largely by Faraday. Only those parbdes exhibiting metallic behaviour, interpreted broadly, will be discussed. (Large, gas-phase dusters were the subject of a Faraday Symposium held in December 1989.) The meetingwill be heldin thetheatre where Faraday lectured on more than a thousand occasions in the building in Albemarle Stmet where he livedand worked for fifty years.This Discussionaincides with the bicentenary of Faraday's birth and will end with the service commemorating his life and work in Westminster Abbey on Friday,x)September. Further information may be obtained from: Mrs Angela Fish, The Royal Society of Chemistry, Burlington House, London WiV OBN. THE ROYAL SOCIETY OF CHEMISTRY, FARADAY DIVISION, SYMPOSIUM 27 The Conformations of Flexible Molecules in Fluid Phases University of Southampton 16-18 December 1991 Organising Committee Professor G. R. Luckhurst (Chairman) Dr.J. H. R. Clarke Dr. J. W. Emsley Many molecules are able to exist in a rangeof different conformations in the fluid phase with distributions which reRed and may even determine the nature of the phase. The experimental and theoretical study of the geometry of the mnformations and their distribution is a challenging task. An interdisaplinary approach involving different Dr. D. J. Osguthorpe Dr. J. Yarwood experimental techniques and strong interaction with theory is clearly requiredto stimulate advances in this field. The Symposium aims to assist in these advances by bringing together those concerned with the many aspects,both static and dynamic, of molecular flexibility in fluid phases. Contributions for consideration by the Organising Committee are welcomed. Ttles and abstracts,of about 300 words, should be submitted by 31 January 1991 to: Professor G.R. Luckhurst, Department of Chemistry, The University, Southampton, SO9 5NH England. Full papers for publication in the Symposium issue of Faraday Transactions will be required by 20th August 1991. xvi JOURNAL OF CHEMICAL RESEARCH Papers dealing with physical chemistry or chemical physics which appear currently in J. Chem. Research, The Royal Society of Chemistry's synopsis + microform purnal, include the following: AM1 Studies on the Gas-phase Pyrolysis of Phenyl Acetate lkchoon Lee, Chan Kyung Kim and Bon-Su Lee (1990, Issue 2) Quantitation of Cross-linking in Alkoxysilane Coatings NigelJ. Clayden and Peter Palasz (1 990, Issue 2) Pressure Dependence of the Hammett Acidity Function (Ho).Part 2. Aqueous Sulphuric Acid Solutions Katsuhiro Tamura, Masami Dan and Takashi Moriyoshi (1990, Issue 4) Acidity of Indolecarbxylate and Indololate Anions. A Re-examination of the &-Acidity Function Manuel Balon, Maria M.A. MUAOZ, Carmen Carmona, Jose Hidalgo and Doming0 Gonzalez (1990, Issue 5) Conformation al Energies of P 1,c3-Dimet hoxy- t-5-met h y lcyclo hexane and trans-1,dDimet hoxy cycloh exan e by 'H NMR and MMZPEOE Calculations Lars G. Hammarstrom, Ulf Berg and Tommy Liljefors (1 990, Issue 5) Analytical Study of Concentrated Chloride Media. Part 1. Solubility and Potential Catherine Sella and Denise Bauer (1 990, Issue 6) Intermolecular Calculations on kine Dimers Francisco Torrens, Jos4 Sanchez-Marinand Francisco Tomiis (1990, Issue 6) Activity Coefficients in Mixed-electrolyte Solutions at 25OC:Sodium Chlorideddium Acetate-Water and Sodium ChloriisSodium Propionate-Water Systems Miguel A.Esteso, Felipe F. Hernandez-Luis, Luis Fernandez-Merida and Oscar M. Gonzalez-Diaz (1 990, Issue 8) Thermo-solvatochromism of a Pyridinium N-Phenoxide Betaine Dye in some Binary Solvent Mixtures Romuald 1. Zalenski, lzabella Adamczewska and Christian Reichardt (1 990, Issue 9) Analytical Study of Concentrated Chloride Media. Part 2. Distribution Coefficient in Liquid-Liquid Extraction Catherine Sella and Denise Bauer (1990, Issue 9) xvii Supervision of Technical Staff An Introduction for Line Supervisors by R.Weston, Leicester Polytechnic, D. C. Norton, Ex-Chief Technician, Bromley College of Technology, M. Grimshaw, North East Surrey College of Technology This unique book forms an introduction to supervisory skills for line supervisors employed in scientific, educational, medical and industrial laboratories. Unlike other publications on supervision it is written specifically for supervisors working in laboratories and concentrates on the specific skills associated with the control of staff in scientific laboratories. The authors have considerable experience as laboratory supervisors and in teaching technical staff. Brief Contents: Organization Induction and Monitoring of Staff Training The Role of the Supervisor within the Laboratory Leadership Counselling and Discipline Organization, Planning and the Technical Supervisor Industrial Relations: the Supervisor and the Trades Unions Motivation Health and Safety Recruitment and Selection The Law and the Supervisor Salaries and Grading The Supervisor and Technology Softcover x+242 pages ISBN: 0 85186 423 6 (1989) Price: f 15.95 Customers wishina to obtain an insDection CODY of this title should contact the Sales Promotion Manager at our Cambridge address. ROYAL c$E&g{ To Order, Please write to the: Royal Society of Chemistry, Distribution Centre, Blackhorse Road, Letchworth, Herts SG6 1HN. UK. or telephone (0462) 672555 quoting your credit card details. We can now accept AccessNisalMasterCard/Eurocard. For further information, please write to the:6 Royal Society of Chemistry, Sales and Promotion Department, Thomas Graham House, Science Park, Milton Road, Cambridge CB4 4WF. UK. RSC Members should obtain members prices and order from :information Services The Membership Affairs Department at the Cambridge address above. DYNAMIC PROPERTIES OF BIOMOLECULAR ASSEMBLIES Edited By S.E. Harding, University of Nottinghamand A.J. Rowe, Universityof Leicester Special Publication No. 74 ISBN: 0 85186 896 7 Hardcover 376 pages Price f45.00 Published November 1989 To order please contact: InformationThe Royal Society of Chemistry, Distribution Centre, Blackhorse Road, Letchworth, Herts SG6 1 HN, UK. ServicesTelephone: (0462) 672555. Fax: (0462) 480947. Telex: R25372 xix

ISSN:0956-5000    

DOI:10.1039/FT99187BP005

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. CHEM. SOC. FARADAY TRANS., 1991, 87(1), 9-13 Microwave Spectrum, Structure, Dipole Moment and Excited States of (CH,),CCP Andrew D. Couch and A. Peter Cox* Department of Physical Chemistry, University of Bristol, Cantock's Close, Bristol BS8 1TS,UK The microwave spectra of four isotopic species of (CH3)3CC=P have been studied in the frequency range YO GHz. The zero-point Fverage structure has been determiaed for the heavy atoms to be: r(C, 'P) = 1.543(3)A, r(C,-CP) = 1.478(3)A and r(C,-CJmethyl]) = 1.544(3)A with LC,-C2-C3 = 109.2(1)". The ground-state dipole moment has been determined to be 1.486(3) D (1 D = 3.33564 x C m) from Stark-effect measure- ments. Two prominent vibrational satellite series have been assigned. The lowest bending vibration is deter- mined to have a wavenumber of 140(10) cm-' from the /-type doubling constant q = 1.474(4) MHz, in agreement with relative intensity measurements corrected for the effects of /-type doubling and resonance.The introduction of a C"P bond into a molecule was first achieved with the production of HCEP in a rotating arc.'q2 Since then, several RCEP derivatives have been identified and characterized by their microwave3 and/or photoelectron ~pectra.~The C=P radical has recently been observed in a carbon star5 following the laboratory measurement of its millimetre-wave spectrum by Saito et aL6 After PN, CP is only the second phosphorus-containing molecule detected in an astronomical source. (CH,),CCEP (3,3-dimethyl-l-phosphabutyne)is the most thermally stable of the phosphaalkynes and hence has been the most commonly investigated.In organometallic chem- istry it has been used as a donor ligand.7 A photoelectron spectroscopic study8 of the vapour has been used to interpret the mode of coordination of the C=P group to metals. A gas-phase structure of (CH3),CC=P has been determined by electron diffra~tion.~ Some preliminary microwave results are given within ref. 9. Here a detailed microwave investigation is presented including carbon-13 data and an accurate dipole moment. Experimental A microwave spectrometer employing 100 kHz Stark modu- lation applied to a 3 m stainless-steel X-band absorption cell was used for this study. Backward-wave oscillators were used with chart recorder display for broad band sweeps (see Fig.1) and klystrons with oscilloscope presentation were used for accurate measurements, which were made in the 9-40 GHz region. All accurate measurements were made at dry-ice tem- perature, although relative intensity measurements between room temperature and dry-ice temperature were used to dis- tinguish carbon-13 transitions from vibrational satellite spectra. A sample pressure of ca. 7 Pa was used. Radiofrequency-microwave double-resonance spectra were collected according to the method of Wodarczyk and Wilson." A Marconi signal generator (TFSOlD/l, 10-470 MHz) and rf amplifier (TF2172, 5-10 W) were used to provide the radiofrequency pumping. Stark effects were measured by applying a direct voltage shift to the baseline of the 100 kHz modulation and measur- ing the line displayed in the field-off phase." The d.c.voltage was precisely applied by a Brandenburg power supply (472 R) and checked with a Fluke 5080A digital multimeter (accurate to 0.1 %). The sample was kindly provided by Dr. Jurgen Seameitat (Inorganic Chemistry Department, University of Bristol), having been prepared by the method of Becker et al." Its purity was checked by NMR. Results Microwave Spectrum (CH3)3CC=P exhibits a typical symmetric-top spectrum, as illustrated in Fig. 1. The ground-state measurements are given in Table 1. K-splittings were resolved for the three highest J' tJ" transitions studied, working with low modu- lation fields (typically 8 V cm-I).The K = 0 components are not modulated under these conditions and were obtained from the intercepts of the D,, plots. Other K = 0 transition frequencies were either measured at high modulating fields (typically 2000 V cm-') or, for the two lowest J transitions, obtained by extrapolation from accurate Stark-effect mea- surements. The spectroscopic constants are given in Table 2. Carbon- 13 species were measured in natural abundance. The on-axis I3C species were measured as isotopic shifts rela- tive to the main species at modulation fields of 100 V cm-'. The low voltage allowed weak isotopic lines to be measured against an interfering background of vibrational satellites. The 13C spectra were positively identified on the basis of correct relative intensities (invariance with temperature dis- tinguishes them from vibrational satellites) in addition to Stark effects and frequency positions.The off-axis carbon-13 I I28 I I30 I I32 I 134 I I 36 I I 38 v/GHz Fig. 1 Broad-band microwave spectrum of (CH3)?CC=P, showing the J = 9 t8 (a), 10 t9 (b) and 11 t10 (c) transitions. Time con- stant, 300 ms; sweep rate, 50 MHz s-'; Stark field, 800 V cm-'; temperature, 205 K. Each transition has a weaker series to low fre-quency; this corresponds to transitions in ('3CH3WCH3),CCsP,the main species signal is overloading the detection system and giving a non-linear response 10 J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 Table 1 Transition frequencies for the ground vibrational state of Table 3 Isotopic shift measurements (MHz) for on-axis carbon-13 (CH,),CCP species observed frequency," obs.-calc. J' J" K v/MHz Av 8 7 26459.74 -63.72 -3.983 f0.003 3 2 0 9 922.49b -0.01 9 8 29767.13 33.63 71.65 1.868 f0.003 3.981 f 0.003 4 3 0 13 229.99' 0.00 10 9 33074.45 37.32 79.59 1.866 f0.003 3.980 f0.003 5 4 0 16 537.49 0.01 11 10 36381.80 40.99 87.54 1.863 f 0.002 3.979 0.002 6 5 0 19 844.96 0.01 12 11 39689.22 44.78 95.59 1.866 f0.002 3.983 f0.002 7 6 0 23 152.42 0.00 1.866 f0.003 3.981 f0.0038 7 0 26 459.86 -0.02 9 8 0 29 767.30 -0.02 " Measured frequency of peak absorption for main species at a Stark 10 9 0 33 074.75' 0.00 field of 10 kV m-' adopted for isotopic shift measurements.10 9 4 33 073.93 0.02 10 9 5 33 073.44 0.01 10 9 6 33 072.86 0.01 6(H-13C) = -4 x A; 6(CH,-"C) = -9 x lo-' A;10 9 7 33 072.19 0.02 10 9 8 33 071.39 0.01 6(13C=P) = -6 x lO-'A. 10 9 9 33 070.50 0.02 The C-C and C-H shrinkage corrections were taken 11 10 0 36 382.15' -0.01 from those calc~lated'~ for (CH,),CCl. The C=P shrinkage 11 10 3 36 381.63 -0.01 correction was taken from that estimated1' for C-0 in 11 10 4 36 38 1.24 0.00 COF,, a bond with a similar force constant. The harmonic 11 10 5 36 380.70 -0.02 11 10 6 36 380.07 -0.01 corrections to the rotational constant B,-B, x 47 kHzg have 11 10 7 36 379.32 -0.01 negligible effect on the present structural calculation. For the 11 10 8 36 378.46 0.00 11 10 9 36 377.47 0.00 11 10 10 36 376.36 -0.01 Table 4 Microwave transition frequencies for off-axis carbon- 13 12 11 0 39 689.56' 0.00 species (position 3) 12 11 2 39 689.32 0.02 12 11 3 39 688.97 -0.02 upper level lower level 12 11 4 39 688.50 -0.05 J K, K, J K, K, obs./MHz obs.-calc."12 11 5 39 688.00 +0.02 12 11 6 39 687.27 -0.01 12 11 7 39 686.48 0.02 8 0 8 7 0 7 26 204.8 1 0.03 12 11 8 39 685.53 0.02 8 1 8 7 1 7 26 162.56 -0.02 12 11 9 39 684.48 0.04 8 1 7 7 1 6 26 25 1.12 -0.01 12 11 10 39 683.19 -0.05 9 0 9 8 0 8 29 479.49 0.02 12 11 11 39681.90 -0.01 9 1 9 8 1 8 29 432.63 0.01 9 1 8 8 1 7 29 532.24 -0.01 " Accurate to f0.05 MHz. Transition frequency determined from 10 2 8 9 2 7 32 763.69 -0.00 -0.02Stark-effect measurements. 'Transition frequency obtained from DJK 10 0 10 9 0 9 32 753.83 plot.10 1 10 9 1 9 32 702.55 -0.02 10 1 9 9 1 8 32 813.26 -0.01 10 2 9 9 2 8 32 758.26 0.01 11 10 2 10 10 1 36 029.61 0.01 (3.3%) has an asymmetric rotor spectrum and is more remote 11 0 11 10 0 10 36 027.89 0.02 -0.05from the main species spectrum. It can be clearly observed in 11 1 10 10 1 9 36094.13 0.04Fig. 1 on the low-frequency side of each main-species J' cJ" 11 2 10 10 2 9 8 36 033.80 11 2 9 10 2 36 040.98 -0.03transition (the intensities of 13C lines relative to the main 11 1 11 10 1 10 35 972.37 -0.04 species are deceptive in this recording; the intense main 11 4 8 10 4 7 36 034.75 -0.02 species spectrum is saturating the detection system which is 11 7 5 10 7 4 36 032.57 0.01 not giving a linear response).The carbon-13 data are given in 11 8 4 10 8 3 36 03 1.72 0.04 Tables 3 and 4. 11 9 3 10 9 2 36 030.71 0.02 12 4 9 11 4 8 39 310.62 -0.02 12 4 8 11 4 7 39 310.62 -0.02 Molecular Structure 12 5 8 11 5 7 39 309.85 0.01 12 5 7 11 5 6 39 309.85 0.01 The rotational constants for four isotopomers of 12 6 7 11 6 6 39 309.05 0.01 (CH,),CC=P are shown in Table 5. We wish to calculate the 12 6 6 11 6 5 39 309.05 0.01 zero-point average structure giving a well defined geometry 12 7 6 11 7 5 39 308.17 0.01 0.01which can be compared directly with the rav obtained in the 12 7 5 11 7 4 39 308.17 electron-diffraction study.g The Kraitchman equations1 were 12 8 5 11 8 4 39 307.19 -0.00 12 8 4 11 8 3 39 307.19 -0.00solved for the three unique carbon-atom positions.The 12 9 3 11 9 2 39 306.09 -0.02 following shrinkage corrections were adopted to allow 12 9 4 11 9 3 39 306.09 -0.02 for changes in average bond lengths with isotopic 12 10 3 11 10 2 39 304.86 -0.05 substitution: 6(13CH3-C) = -7 x lo-' %i = 6(13C-C(P)); 12 10 2 11 10 1 39 304.86 -0.05 12 11 2 11 11 1 39 303.63 0.03 12 1 12 11 1 11 39 242.16 0.03 12 2 10 11 2 9 39 318.65 0.04 Table 2 Ground-state spectroscopic constants" for (CH ,)&CP 12 2 11 11 2 10 39 309.18 -0.01 12 11 1 11 11 0 39 303.63 0.03 12 1 11 11 1 10 39 374.99 0.03 12 3 10 11 3 9 39311.61 0.05 1653.7511 (8) 0.068 (3) 2.632 (5) 12 3 9 11 3 8 39 31 1.61 -0.08 Errors la. " See Table 5 for spectroscopic constants from fit. J. CHEM. SOC. FARADAY TRANS., 1991, VOL.87 Table 5 Rotational constants (MHz) and centrifugal distortion constants (kHz) for isotopic species of (CH,),CCP (CH3)3C-C~P (CH3),C- l3C=P Bo CO 1653.7511 (8) - 1651.885 (4) - A0 - - DJ 0.068 (3) - DJK 2.632 (5) - DI - - 13C substitution lying off the figure axis of the symmetric top, Kraitchman's equations for a molecule with a plane of sym- metry have been adopted; in these equations AIc -AIb = AIa has been used because of the limited knowledge of the A rotational constants for both the I3C and the main species. The accuracy of this procedure may be affected by a small change of inertial defect on substitution which will lead to a slight error in the b-coordinates of the off-axis carbon atoms.However the b-coordinates obtained in this way agree very well with those found for i~obutane,'~.'~ (CH,),CH. In addi- tion the derived I, value for (CH,),CCEP agrees well with other t-butyl derivatives;" for this comparison CH, param- eters were taken from the electron-diffraction determination,' see Table 6. The phosphorus atom was placed using the first-moment condition along the a axis. The heavy-atom skeleton was then slightly adjusted to fit 1, together with the first-moment equation. The CH, parameters were again assumed from the electron-diffraction study;' this assumption has negligible effect on the heavy-atom parameters. The carbon coordinates obtained by this procedure are only slightly changed from the Kraitchman values calculated with bond shrinkages.The digits given in brackets after the coordinates in Table 6 give the small fitting adjustments which have been made. The off- axis carbons are best fitted by allowing a tiny additional angular shrinkage of -0.01" in the LC,C,C, with 13CH3 substitution, similar to that found in the structural determi- nation of pivalaldehyde." The error estimates in the struc- ture given in Table 6 arise mainly from uncertainties in the isotopic shrinkage effects. The structure is illustrated in Fig. 2. Table 6 Average structure for (CH3),CC3P principal axis coordinates/A atom a P -2.1491 c, -0.6056 (6)" c2 0.8727 (5)c3 1.3812 (1) c4.5 1.3812 (1) C,=P c,-CP (methy1)C3-C, LC3-C2-C1 (assumed LH-C-H 1: (calc.) = 305.5958 u A' I, (obs.) = 305.5956 1; = 305.6043 b c 0.0 0.0 0.0 0.0 0.0 0.0 1.4576 (-25) 0.0 -0.7288 (12) f1.2623 ( T 22) bond length/A 1.5435 f0.0026b 1.4783 0.0027 1.5437 f0.0025 109.23 f0.13" = 108.8" and C-H = 1.08) I, (calc.) = 111.32 u A' Last dig>'s in brackets added to the best-fit coordinate give the Kraitchman value.20. (CH3)313C-CsP (I 3CH 3MCH3)zC- CpP 1649.770 (4) 1643.518 (4) - 1632.454 (4) - 4420 (20) - 0.106 (5) - 2.572 (5) - 0.014 (7) Dipole Moment The ground-state dipole moment of (CH3),CC=P has been determined via the first-order Stark effect. The displacements of the KM = & 1 and KM = f2 components at J3 t2 and 544- 3 were measured as a function of d.c. voltage.The average frequency of the positive and negative components was taken to cancel even-order contributions to the essen- tially first-order displacements. The displacements were com- pared with those of the calibrant molecule, propyne.20 A small correction of -0.4 V for the zero-basing error of the applied 100 kHz Stark modulation voltage was deduced from the calibration and corrected for in the (CH3)3CCEP data. The Stark coefficients are given in Table 7. A value of 1.4861(30)D was obtained for the dipole of (CH,),CCEP. Excited Vibrational States Two vibrational satellite series are apparent in the microwave spectrum of (CH,),CC=P (see Fig. 3). The strongest series extending to the high-frequency side clearly belongs to a degenerate vibration with a statistical weighting of (u + 1) and is assigned to the C-CsP bending mode by compari- son with the IR spectrum" of (CH,),CC=N.Indeed I-type doublets have been measured for the u = 1 state of this series using radiofrequency-microwave double resonance, as illus- trated in Fig. 4. The transitions measured and the derived value of q, the l-type doubling constant, are given in Table 8. The vibrational wavenumber, o,of this lowest mode has been estimated to be at 175 cm-' from microwave relative inten- sities against a value of 148 cm-' oia its Raman spectrum both values reported in ref. 9. We have found this situation to be quite usual for bending modes of this kind in a number of molecules,22 that the microwave relative intensities give an upper limit for the value o compared with that obtained 1.478 1.543 y> Fig.2 Average structure, bond lengths in A, bond angles in degrees J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 Table 7 First-order Stark coefficients" (MHz V-') and dipole moment (D)of (CH,),CCP 3 2 *1 3 2 f2 4 3 +1 4 3 +2 0.13562 (13) +0.016 (23) 1.8955 (18) 1.4859 (23) 0.27147 (30) -0.050 (43) 1.8971 (24) 1.4871 (31) 0.05419 (12) -0.029 (40) 1.8935 (42) 1.4843 (38) 0.10857 (16) -0.001 (41) 1.8968 (28) 1.4869 (29) pa",= 1.4861 (30) Average shifts for IKM 1 giving a zero-basing correction of -0.40 directly from vibrational spectroscopy. This results from the fact that the excited-state intensities relative to the u =0 intensities are weakened (ca.16% for v = 1) by the effects of I-type doubling and resonance. This is especially the case for I v =o 36.2 36.3 36.4 36.5 36.6 36.7 36.8 v/G Hz Fig. 3 Medium-resolution scan of Jll t10 transition in (CH,),CC=P indicating the two vibrational series in its microwave spectrum. Time constant, 300 ms; sweep rate, 4 MHz s-'; Stark field, 800 V cm-';temperature, 205 K ,-_____-19.82 19.84 19.86 19.88 19.90 v/GHz Fig. 4 Radiofrequency-microwave double-resonance spectrum of J = 6 t5 I-type doublet transitions in the u = 1 vibrational state of the lowest bending mode in (CH3)3CC=P. Radiofrequency pump, 44.2 MHz; sweep rate, 3 MHz s-;temperature, 205 K Table 8 Observed I-type doublet frequencies (MHz) and rotational constants (MHz) for the lowest bending mode in (CH,),CCP lower I-type doublet upper I-type doublet J' tJ" VlOW.3 V,pper 4 6 5 19 856.30 19 873.94 1.470 f0.008 9 8 29 784.20 29 8 10.79 1.477 f 0.006 11 10 36 402.70 36 435.16 1.475 f0.005 12 11 39 71 1.99 39 747.36 1.474 f 0.004 Average value of q = 1.474 f0.004 MHz.B* = 1655.435 (2) MHf. a See ref. 42 for definition of B*, q is four times larger than the con- stant defined in ref. 42. components. For methyl acetylene calibration: d[v]/dV =0.28619 (5) MHz V-', and intercept = -0.116 (8) MHz V. p[CH,CCH] = 0.7839 (10) D, ref. 20. heavy-top molecules such as (CH3),CC= P, where pro-nounced I-type resonance tends to occur. A value of o can also be calculated from the experimental I-type doubling constant, q, using the molecular force field.Grenier-Bes~on~,has given a formula for 4 in terms of quad- ratic and cubic potential constants which may approximately be written as: a' Be2 41 =-0, The value of a' has been estimated from experimental data24 for CF3CGCC1, a molecule with a similar force field to (CH,)3CCGP, to be 2.25 k0.15. Using this and the B, value for (CH,),CCGP, substitution into the expression for q gives a value of o = 140 (10) cm-', which is in good agreement with the Raman value. The second weaker vibrational series, extending to the low- frequency side, is the degenerate methyl torsional mode at ca. 220 cm- '. This is assignable through comparison with other t-butyl molecule^.^^-^' In addition a non-degenerate methyl torsional mode is usually found to low frequency of the ground state, but for (CH,),CC=P this is obscured between the ground state and the stronger degenerate mode.Discussion The zero-point average structure of (CH3),CC=P has been refined in the present work. The inclusion of 13C data appears to have relieved some of the correlation present in the electron diffraction study.' The C=P linkage is found to be ca. 0.008 A longer in the microwave structure which brings it more into line with other RC=P molecules (see Table 9), and in addition shows the same behaviour as in the isovalent nitriles where the CEN bond length is almost invariant with substitution. The other main difference in the determinations is the LCCC of the t-butyl group; in the present study this group is shown to open up from tetrahedral against the CEP bond, whereas the electron-diffraction structure has it closing down.While the group is reasonably close to tetrahedral the present angle accords with the behaviour of other t-butyl tops such as isobutane and (CH,),CCsN, see Table 9. Otherwise the structural determinations are in good agree-ment. The C-CP bond is lengthened relative to other C-CP bonds, but perhaps less pronouncedly than for the nitriles ; this small effect matches the insignificant dipole change in CH,C=P with t-butyl substitution. The dipole moment values of RCEP molecules (also given in Table 9) do not show any simple additive behaviour.The CsP bond is clearly less polar than the isovalent CEN. The functional group properties of C=P are also different from the CEN group with respect to complexation with transition metals; RCEP prefers to attach in a side-on manner,' whereas nitriles almost exclusively adopt an end-on configu- ration. This difference in mode of coordination has been explained in terms of the results from photoelectron spectros- J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 Table 9 Structural parameters (A and degrees) and dipole moments (D) for R-CEX R-C=X C,=X c,-cx H-CrP CH,C=P (CH 3)3CC= P (electron diffraction) 1.542 (0) 1.544 (4) 1.543 (3) 1.536 (2) 1.465 (3) 1.478 (3) 1.473 (4) H-CEN 1.156 (2) CH3C=N (CH,),CC=N FC=P 1.157 (3) 1.159 (1) 1.541 (5) 1.458 (3) 1.495 (15) CF3C=P 1 S44’ 1.460 NGC-CEP 1S44’ 1.382 H-CSC-C=P 1 .544b 1.382 CH,=CHC=P 1S44’ 1.432 PhC=P 1 S44’ 1.467 C=P(X zc +) 1.562 (0) a Present work. Assumed from CH3CEP structure.copy,* where the 71-n separation is found to be increased in the phosphaalkynes relative to the nitriles. Thus the first ion- isation potential of RC=P is associated with excitation from n(C2P) rather than non-bonding CT as for the CGN group. This allows for the possibility of the formation of complexes in which phosphaalkynes are 2-, 4- or 6-electron donors.40 The experimental results so far relate to transition-metal complexes. A logical extension of these studies would be to study van der Waals complexes involving RCEP in a free-jet expansion; many such studies have been made on nitrile systems, especially by pulsed nozzle microwave Fourier-transform spectro~copy.~~ It would be interesting to see in the dimers the extent to which the same principles of coordi- nation apply.The vibrationally excited microwave spectra warrant further study. The series to the high-frequency side of the ground state belonging to the degenerate C-C’-P bend shows the effects of I-type resonance characteristic of the heavy butyl top. As a result relative intensity measurements of the satellites under low resolution give an upper limit for the wavenumber. When loss of intensity due to the additional splitting of the excited states is allowed for empirically, good agreement is found with o = 140 (10) cm-’ obtained from the /-type doubling constant.We thank the SERC for a research studentship. We also thank Dr. Jiirgen Seameitat for the sample of 3,3-dimethyl-l- phosphabutyne. References 1 T. E. Gier, J. Am. Chem. SOC., 1961,83, 1769. 2 J. K. Tyler, J. Chem. Phys., 1964,40, 1170. 3 M. J. Hopkinson, H. W. Kroto, J. F. Nixon and N. P. C. Simmons, Chem. Phys. Lett., 1976,42, 460. 4 H. W. Kroto, J. F. Nixon, N. P. C. Simmons and N. P. C. West- wood, J. Am. Chem. Soc., 1978,100,446. 5 M. Guelin, J. Cernicharo, G. Paubert and B. E. Turner, Astron. Astrophys., 1990, 230, L9. 6 S. Saito, S. Yamamoto, K. Kawaguchi, M. Ohishi, H. Suzuki, S-I. Ishikawa and N.Kaifu, Astrophys. J., 1989,341, 11 14. 7 J. C. T. R. Burckett-St. Laurent, P. B. Hitchock, H. W. Kroto and J. F. Nixon, J. Chem. Soc., Chem. Commun., 1981, 1141. 8 J. C. T. R. Burckett-St. Laurent, M. A. King, H. W. Kroto, J. F. Nixon and R. J. Suffolk, J. Chem. Soc., Dalton Trans., 1983, 755. 9 H. Oberhammer, G. Becker and G. Gresser, J. Mol. Struct., 1981, 75, 283. 10 F. J. Wodarczyk and E. B. Wilson, J. Mol. Spectrosc., 1971, 37, 445. 11 A. P. Cox, I. C. Ewart and W. M. Stigliani, J. Chem. SOC., Faraday Trans. 2, 1975,71,504. PP ref. 0.392 2 1.499 28 1.544 (3) 109.2 (1) 1.486 a 1.543 (2) 109.9 (2) 9 2.997 29, 30 3.946 29, 31 1.536 (7) 108.4 (5) 3.97 32 0.279 33 - 34 3.44 35 0.745 36 1.183 37 - 38 - 39 12 G.Becker, G. Gresser and W. Uhl, Z. Naturforsch, Teil B, 1981, 36,16. 13 J. Kraitchman, Am. J. Phys., 1953, 21, 17. 14 R. L. Hilderbrandt and J. D. Wieser, J. Chem. Phys., 1972, 56, 1143. 15 V. W. Laurie, D. T. Pence and R. H. Jackson, J. Chem. Phys., 1962,37,2995. 16 D. R. Lide, J. Chem. Phys., 1960,33, 1519. 17 R. L. Hildebrandt and J. D. Wieser, J. Mol. Struct., 1973, 15,27. 18 P. R. R. Langridge-Smith, R. Stevens and A. P. Cox, J. Chem. SOC.,Faraday Trans. 2, 1980,76,330. 19 A. P. Cox, A. D. Couch, K. W. Hillig, M. S. La Barge and R. L. Kuczkowski, to be published. 20 J. S. Muenter and V. W. Laurie, J. Chem. Phys., 1966,45855. 21 G. A. Crowder, J. Phys. Chem., 1971,752806. 22 A. P. Cox and M. C.Ellis, personal communication. 23 M. L. Grenier-Besson, J. Phys. Radium, 1960,21, 553. 24 A. Bjerrsoth and K. M. Marstokk, J. Mol. Struct., ,1972, 13, 191; E. Augdahl, E. Kloster-Jensen, V. Devarajan and S. J. Cyvin, Spectrochim. Acta, Part A, 1973,29, 1329. 25 D. R. Lide and D. E. Mann, J. Chem. Phys., 1958,29,914. 26 D. R. Lide and D. E. Mann, J. Chem. Phys., 1961,36,965. 27 A. M. Ronn and R. C. Woods, J. Chem. Phys., 1966,45,3831. 28 H. W. Kroto, J. F. Nixon and N. P. C. Simmons, J. Mol. Spec-trosc., 1979,77, 270. 29 C. C. Costain, J. Chem. Phys., 1958,29,864. 30 B. N. Bhattacharya and W. Gordy, Phys. Reu., 1960,119, 144. 31 P. A. Steiner and W. Gordy, J. Mol. Spectrosc., 1966, 21, 291. 32 L. J. Nugent, D. E. Mann and D. R. Lide, J. Chem. Phys., 1962, 36,965. 33 H. W. Kroto, J. F. Nixon and N. P. C. Simmons, J. Mol. Spec-trosc., 1980, 82, 185. 34 J. C. T. R. Burckett-St. Laurent, T. A. Cooper, H. W. Kroto, J. F. Nixon, 0. Ohashi and K. Ohno, J. Mol. Struct., 1982, 79, 215. 35 T. A. Cooper, H. W. Kroto, J. F. Nixon and 0.Ohashi, J. Chem. SOC.,Chem. Commun., 1980,333. 36 H. W. Kroto, J. F. Nixon and K. Ohno, J. Mol. Spectrosc., 1981, 90,512. 37 K. Ohno, H. W. Kroto and J. F. Nixon, J. Mol. Spectrosc., 1981, 90,507. 38 J. C. T. R. Burckett-St. Laurent, H. W. Kroto, J. F. Nixon and K. Ohno, J. Mol. Spectrosc., 1982,92, 158. 39 R. S. Ram and P. F. Bernath, J. Mol. Spectrosc., 1987,122,282. 40 J. C. T. R. Burckett-St. Laurent, P. B. Hitchcock, H. W. Kroto, M. F. Meidine and J. F. Nixon, J. Organomet. Chem., 1982, 238, C82-84. 41 N. W. Howard and A. C. Eegon, J. Chem. SOC., Faraday Trans. 2, 1987,83,991. 42 M. L. Grenier-Besson and G. Amat, J. Mol. Spectrosc., 1962, 8, 22. Paper 0/03694C; Received 13th August, 1990

ISSN:0956-5000    

DOI:10.1039/FT9918700009

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. CHEM. SOC. FARADAY TRANS., 1991, 87(1), 15-18 Optical-Optical Double Resonance (OODR) Studies of the Halogen Ion-pair States Part 2.-The f(0:) +B(0:) Transition Dipole Moment Function in I, Andrew J. Holmes,t Kenneth P. Lawley," Trevor Ridley, Robert J. Donovan and Patrick R. R. Langridge-Smith Department of Chemistry, University of Edinburgh, West Mains Road,Edinburgh EH9 3JJ, UK The electronic transition dipole moment function, pf+&?),for I, has been determined over the range 3-5 A by simulation of the dispersed fluorescence from a range of upper vibrational levels of the f(O,+) ion-pair state. jlf+B(R) displays a maximum close to Re of the upper state. The decrease in transition dipole at R <Re is discussed in terms of the changing electronic structure of the ion-pair state.The lifetime of the f(v' = 0) state has also been measured (z = 13.6f0.7ns), together with the ratio of fluorescent intensities from f(v' = 0) to the B", A and B states, enabling the f + B transition dipole function to be determined absolutely from this calibration at Re,where pUf+B= 3.0D (1 D = 3.33564x C m). Two characteristics of diatomic ion-pair states are their large Re values and small values of me compared with the ground state. Consequently, vertical excitation of ion-pair states from the ground state, whether by single-photon absorption or a coherent two-photon process accesses a restricted band of very high vibrational levels, typically u' > 100 in the case of I, .Ip3The optical-optical double resonance (OODR) techniqueL8 uia the B state in I, can, by varying the wave- length of the probe photon A,, be used to access a much wider range of ion-pair vibrational levels.At the long-wavelength limit of A,, there is good Franck-Condon overlap with u' = 0 of the manifold of ion-pair states and these transitions occur near the outer turning point of the B-state vibrational motion. At the short-wavelength limit of A2 the second transition occurs around the inner turning point of the intermediate state, making the overall two-photon process essentially a vertical one resulting in high u' levels. Probe transitions occurring near Re of the B state access u' z 5&60 in the f state. Also, because of a variety of perturbations operating in the intermediate B state, a much wider range of final electronic states is accessible with the OODR technique compared with the one-photon and coher- ent two-photon routes.As part of a continuing series of experiments in this labor- atory, the OODR technique was used in three probe wave- length regions to populate the f(0,f) state of I,. The final vibrational levels were as follows: (i) u' = 0, (ii) vr = 57 and (iii)u' = 88. In this paper we report the dispersed fluorescence from these levels, with particular emphasis on that from the highest ion-pair vibrational state. An RKR analysis of the f state up to v' = 75 has been reported by Perrot et later extended by Hickmann et al." The B-state potential is well known"," and so the new information derivable from the dispersed fluorescence is the f +B electronic transition dipole function, p12(R)over the range of R spanned by the vibration amplitude of the highest u' level.We also report a measurement of the radiative lifetime under collision-free conditions of the f state (u' = 0) and the integrated fluorescent intensities from that level to each of the f state and hence put the whole determination of p12(R)on an absolute basis. Experimental The experimental configuration has been described in detail e1~ewhere.l~Briefly, two Lambda Physik dye lasers (FL3002 and FL2002) were simultaneously pumped by a Lambda Physik EMG 201MSC excimer laser to provide the following two-colour OODR laser excitation scheme: f(O,+)'B(O3 'X(0,f) (n-17)R(69) (17-1)P(70) where n = 0, 57, 88.Iodine at a pressure of 100 mTorr was contained in a glass cell fitted with Spectrosil quartz windows. The response function of the monochromator/PM tube (Jobin-Yvon HRS 2/Hammamatsu R928 combination) was measured, fitted to a polynomial in u, F(u), which was then applied to the simulated spectra. Thus, in all compari- sons made between observed and simulated spectra, the detector response has been fully allowed for. The corrected fluorescence S(u) is in units of photons per unit wavelength interval per second and for a particular transition is related to the Einstein A coefficient by S,,(u) = F(u)u3A&) (1) The dispersed fluorescence from f(u' = 0) to the B", A and B states are shown together in Fig.1, this time digitally record- ed with the detector response correction applied. The lifetime of the f state was measured by following the time decay of the fluorescence excited by a probe-laser pulse shortened to 4 ns (as measured by a fast photodiode) using a Lambda Physik FL90 pulse compressor. The transient signals were recorded by scanning a boxcar gate 2 ns wide over the excitation pulse, P(t), and monitoring the fluores- cence at a wavelength characteristic of the f + B emission, Zn(t). P(t)was then used in a numerical convolution to model the fluorescence signal, assuming single exponential decay with a trial lifetime z: C2 = 0 and 1 ungerade valence states that carry appreciable oscillator strength.From these two extra pieces of informa-Z;:"(t) = C P(t)e~p[(t'-t)/~]dt' (2)tion we deduce the absolute magnitude of p12(R)at Re of the l and adjusting z and C to minimise the standard deviation a(T) over the elapsed time T over which fluorescence was t Present address : Lash Miller Chemical Laboratories, University monitored. This upper limit of observation was generally of Toronto, Toronto M5S 1A1, Canada. taken to be 42. However, the best-fit values of 7 showed no J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 f -> B”. illll r I I I r I I 1 275 280 285 290 295 330 335 340 345 350 A/nm Fig. 1 The complete dispersed fluorescence from I, f(u’ = 0). The intensity of the f 4B system has been reduced by a factor of 30 significant dependence on T and gave an average value of 13.6 ns.Moreover, the lifetimes determined in this way showed no dependence on the initial estimate used to start the fitting, which was obtained by the method of moments or from a simple log fit of the exponential tail of the signal. The probe pulse, the fluorescence monitored at 338 nm and the best-fit convoluted signal are displayed in Fig. 2, together with the residual, I;;b“(t)-IFy(t).A slight ringing, which had the same characteristics in all the decay curves recorded, is visible just after the probe pulse. Filtering this out numeri- cally at the data analysis stage before fitting produced no change in the best estimate of z, but reduced the standard deviation of the fit so that the uncertainty in z was kO.05 ns, which reflects the low level of shot-to-shot variation.There is no systematic deviation from white noise in the residual signal at longer times which would indicate a more compli- cated functional dependence. Fluorescent decays recorded at various pressures of iodine between 50 and I00 mTorr pro- duced no significant systematic pressure dependence of z, but there was an overall day-to-day scatter of kO.7 ns, which we take to be a realistic estimate of the uncertainty in z. Discussion Qualitative Features of the Fluorescence The complete f-state fluorescence from v’ = 0 (Fig. 1) shows three electronic systems (first identified by Heeman et al.14), terminating in the A 3111ustate at 283 nm, the B” ‘nlustate at 292 nm and the B (%&) state at 340 nm.The f B tran- sition is the only parallel one and has by far the largest Ein- stein A coefficient. The areas under the three electronic systems in Fig. 1, corrected for monochromator + detector response, give the following ratios for the three Afi coeffi-cients at the wavelengths indicated Af-,(340) :Af-Bn(292) :Af-A(282) = 1 0.014 :0.016 The f+B” fluorescence is clearly of the bound-+free type terminating on the repulsive wall of the B” state. The f +A band shows discrete structure with a spacing characteristic of the A state (0,= 95.0 cm- 15). Fig. 3 and 4 show the f 3 B dispersed fluorescence from the higher vibrational levels v’ = 57 and 88, respectively. The complete spectrum of dispersed f B fluorescence from u’ = 88 and u’ = 57 shows clearly the following features: (i) a red extremum invariant with v‘ at Are,, = 340 nm and accompanied by a peak in If,,(ii) a blue extremum that shifts to shorter wavelengths as u’ increases and is also marked by a rise in intensity of the envelope of the discrete structure, (iii) a I , 0 15 30 45 60 75 I It/ns 270 280 290 300 310 320 330 340 350 Fig.2 The fluorescence decay curve for the f+ B emission (-) A/nmafter excitation by a compressed probe pulse (---). The best-fit simu- lated decay is also shown (---), together with the residual signal Fig. 3 The observed (a) and simulated (b) dispersed fluorescence along the baseline from f(u’ = 57) J. CHEM. SOC. FARADAY TRANS., 1991, VOL.87 il 260 270 280 290 300 310 320 330 340 350 A/n m Fig. 4 The observed (a) and simulated (b) dispersed fluorescence from f(v' = 88) discrete/continuum boundary, at Ad, marked by a slight rise in I,,,that also progressively moves to the bluet with increas- ing u'. This feature is more prominent in the fluorescence from u' = 57, where it occurs at 307 nm, shifting to 287.5 nm at v' = 88 and in general becomes more prominent the lower u', until it merges with the red extremum. f+B fluorescence from u' < 25 is entirely bound --+ bound. Simulation of the Dispersed Fluorescence Four functions form the input to any simulation of dispersed fluorescence:the potential-energy functions of the upper state V2(R)and lower state, Vl(R), the transition dipole moment function pI2(R)and the detector response function F(u) (assuming this not to have been applied to the raw data).To a useful extent, the effect on the simulated spectrum of modifying p12(R)is uncoupled from changes in the two potentials, especially within the constraints imposed on the latter by fitting the known vibrational term values G, . Broadly speaking, the shape of p12(R)determines the relative inten- sities of the red and blue extrema and the relative prominence of the peak in intensity at the bound/free boundary [this feature is sensitive to the amplitude of p12(R)at large R]. Small lateral displacements of the two potentials, which would preserve G,, and Go,,,produce a first-order shift in the nodal positions of the high-frequency interference structure in Z,,(u) but only a second-order change in the envelope of this structure.The spectroscopic constants of Perrot et a1.' are valid up to u' = 78 and were used to calculate RKR turning points for the f state which were then fitted to a spline function to obtain a smooth potential for the upper state between 2.95 and 5.0 8,. In order to simulate the fluorescence from u' = 88 the upper-state potential has to be extended and the RKR turning points for that level (G, = 7717.8 cm-') were estab- lished at 2.935 and 5.150 A, assuming the outer branch of the potential is a Coulombic function between the outer turning points of u' = 78 and 88. t To the red of 1, the spectrum is an oscillatory continuum, changing at 1, to a modulated line spectrum.Because of the high density of vibrational states near the dissociation limit, the distinc- tion between a continuum and a discrete spectrum is not always clear. The shift of 1, with v' is equal to the shift in the upper-state vibrational term value, AG", . The B-state curve was similarly constructed using the RKR points of Luc" up to u' = 62, within 0.5% of D,.The inner branch of the potential had to be extended into the region of positive potential with respect to the separated atom limit. This extension down to 2.43 8, was initially modelled using the knot points of Brand et ~1.'~to 2.5 8, and the final knot point adjusted so that the phase of the f+B boundjfree structure was reproduced.The knot points finally used to describe the repulsive limb of the B state between 2.62 and 2.43 8, are listed in Table 1. Table 1 Knot points for a cubic spline reconstruction of the exten- sion of the inner wall of the B state used in simulating the f +B fluorescence 2.43 12000.0 2.50 8845.3 2.55 6884.3 2.60 5229.5 2.62 4646.9 The simulations of the f +B dispersed fluorescence spectra for u' = 57 and 88 are shown in Fig. 3 and 4, respectively. The optimum transition dipole function is given in Fig. 5. It exhibits a maximum around 3.6 A, slightly beyond Re of the f state. The fall-off in p12(R)to large R is entirely to be expected in a charge-transfer spectrum which must go to zero at R = m, and was predicted by M~lliken'~ in 1939.The transition dipole of any charge transfer must decrease as R -+0 to the united-atom value which will be small if Ryd- bergisation of the upper valence orbital has taken place. However, the decrease in p12(R)between 3.6 and 3.0 8, is presumably a consequence of a rapid configuration change from essentially a symmetrised product of I+(3P0) and I-ionic wavefunctions towards a new configuration which at small R is predominantly$ (2242), 'E-(O;). Assuming the B state to be largely (2431), 311(0,')near its equilibrium separa- tion, we see that the f-, B transition will become one-electron-forbidden when the two limiting configurations are reached. In contrast, the lower ion-pair state of 0; symmetry, the E state, acquires the dominant configuration (1432), 3n0 as R decreases and the E +B transition remains strongly allowed.However, Tellinghuisen'8 found that the E --+ B fluo- rescence reported by Rousseau and William~'~ is best fitted by a transition dipole function that peaks around 3.89 A, appreciably beyond Re of the E state, 3.65 A. Taken in iso- lation, Tellinghuisen's observation might point to an elec-tronic rearrangement in the B state at R = 3.9 8,,but there is no sign of a change in our f-+B transition dipole at this .~larger separation. Perrot et ~1 deduced p12(R)between 3.2 and 4.5 8, for the E(0:) +B transition from lifetime measure- ments for 10 < u' < 33. They found a rather abrupt change in gradient at around 3.8 8, in pE+Bwhich then remained con- stant to smaller R, at a maximum value of 3.16D.However, in neither experiment was the detector response function F(u) reported; this is an important correction because p12(R)2is proportional to the amplitude of the envelope of the dis- persed fluorescence and hence to F(u).Indeed, further work in this laboratory on the E state, using higher u' levels than those of Perrot et al., suggests that pEhBcontinues to rise smoothly between 3.8 and 3.1 A. A different description of the electronic configuration of the f state has recently been put forward by Li and Balasubrama- nian.20 From the results of ab initio calculations, they assign $ The configuration a: xi 7c: 0: is abbreviated to ($0. 1: 18 I I3.2 2.4 e..-0 c C2 1.6 a-0 .-P 0 c 0 0.8 c9 c 0.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 RIA Fig. 5 The f + B transition dipole function the dominant configuration (0442) to the f state near Re. However, this cannot be correct because the configuration (0442) is a singlet one and the 'D or 'S states of I+ are only minor constituents at the 10%-20% level (arising from spin- orbit interaction) in the product of ionic wavefunctions ~3Po)~1So).The latter is known21 to be a good approx- imation to the electronic structure down to Re.It is also clear that the transition between the configurations (0442), 'C, -+ , 311(0,') is one-electron-forbidden, whereas we have found that at Re of the f state the f -+ B transition is strong.Radiative Lifetime and Absolute Transition Dipole Simulating the dispersed fluorescence can give p12(R)only to within an arbitrary scaling factor. The absolute transition dipole can be obtained once the Einstein A,, coefficient is measured for any f-state vibrational level. This requires two types of measurements, that of the lifetime of the state and the relative intensity of fluorescence in each of the possible decay channels of that state, = '-I 'f+B/Iall channels If+, 1 (3) where, strictly speaking, the channels refer to individual vib- ronic transitions and the A coefficient is also for a single vib- ronic transition. We substitute for each I intensities integrated over urr(or the lower continuum for bound + free transitions) and thus obtain an coefficient for f(u' = 0) that is summed over all u".From the integrated dispersed fluorescence in the three decay channels of the f state (Fig. 1) we find If+B/zIall channels = 0.97, leading to A,,, = 7.35 x 10' s-l for u' = 0. From this we deduce the vibrationally aver- aged transition dipole (0 I k+B(R)10) = 2.99 D which, because the amplitude of vibration is so small may be Table 2 Knot points used to generate the f-, B transition dipole function in Fig. 5 2.80 1.950 3.10 2.472 3.40 2.949 3.50 3.000 3.60 2.922 3.80 2.406 4.40 1.296 4.90 0.780 5.30 0.450 J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 equated to P~,~(R~).Finally, then, we can calibrate the point at Re = 3.6 A on the Pf+B(R) function obtained from fitting the dispersed fluorescence to give the absolute transition dipole function shown in Fig.5. Conclusions The behaviour of the f+B transition dipole function between 3 and 5 8, has been deduced by fitting the dispersed fluorescence from several vibrational levels of the f state, ranging from u' = 0 to u' = 88 consistently with a single func- tion, P~+~(R),which peaks around 3.6 A, close to Re of the ion-pair state. The rapid decrease in Pf+B at smaller separa- tions is attributed to a configuration change in the electronic structure of the f state. It is suggested that this change is from a product of atomic ion eigenstates, I 3P0)I 'So), which is a combination of the 3110g and 'C-(O;) states in Hund's case (a), to a single dominant MO description, 'Xi.Transition from the latter to the B state (311&)is dipole-forbidden. In a separate series of measurements, the lifetime of the f state in u' = 0 was measured to be 13.6 & 0.7 ns. This allows the whole Pf+B(R) function to be defined absolutely. The dominant decay channel around Re of the f state is through the f -+ B transition at 340 nm, which is more than an order of magnitude more intense than the perpendicular (As2 = l)f + B" and f -+ A transitions, which are also recorded here. The absolute transition dipole moment Pf+B at Re = 3.6 A is deduced to be 3.0 D. We are grateful to Mr Philip Jewsbury for developing the lifetime convolution program and for his evaluation of the relative fluorescent intensities in Fig.1. A.J.H. was supported by an SERC research studentship throughout this work. References 1 A. Hiraye, K. Shobatake, R. J. Donovan and A. Hopkirk, J. Chem. Phys., 1988,88,52. 2 M. Bartels, R. J. Donovan, A. J. Holmes, P. R. R. Langridge- Smith, M. A. McDonald and T. Ridley, J. Chem. Phys., 1989,90, 6821. 3 A. R. Hoy and R. H. Lipson, Chem. Phys., 1990,140, 187. 4 J. C. D. Brand and A. R. Hoy, Appl. Spectrosc. Rev., 1987, 23, 285. 5 T. Ishiwata and I. Tanaka, Laser Chem., 1987,7,79. 6 J. P. Perrot, B. Femelat, J. L. Subtil, M. Broyer and J. Cheva-leyre, Mol. Phys., 1987, 61, 85. 7 J. P. Perrot, B. Femelat, M. Broyer and J. Chevaleyre, Mol. Phys., 1987, 61,97. 8 G. W. King and T. D. McLean, J. Mol. Spectrosc., 1989, 135, 207. 9 J. P. Perrot, A. J. Bouvier, A. Bouvier, B. Femelat and J. Cheva-leyre, J. Mol. Spectrosc, 1985, 114,60. 10 J. S. Hickmann, C. R. M. De Oliveira and R. E. Francke, J. Mol. Spectrosc., 1988, 127, 556. 11 P. Luc, J. Mol. Spectrosc., 1980,80,41. 12 S. Gerstenkorn and P. Luc, J. Phys. (Paris), 1985,46, 867. 13 R. J. Donovan, A. J. Holmes, P. R. R. Langridge-Smith and T. Ridley, J. Chem. SOC.,Faraday Trans. 2, 1988,84,541. 14 U. Heemann, H. Knockel and E. Tiemann, Chem. Phys. Lett., 1982,90, 17. 15 K. S. Viswanath, A. Sur and J. Tellinghuisen, J. Mol. Spectrosc., 1981,86,393. 16 J. C. D. Brand, A. R. Hoy, A. K. Kalkar and A. B. Yamashita, J. Mol. Spectrosc., 1982,95, 350. 17 R. S. Mulliken, J. Chem. Phys., 1939,7, 20. 18 J. Tellinghuisen, Phys. Reo. Lett., 1975,34, 1137. 19 D. L. Rousseau and P. F. Williams, Phys. Rev. Lett., 1974, 33, 1368. 20 J. Li and K. Balasubramanian, J. Mol. Spectrosc., 1989,138, 162. 21 P. Jewsbury and K. Lawley, Chem. Phys., 1990,141,225. Paper 0/03020A; Received 5th July, 1990

ISSN:0956-5000    

DOI:10.1039/FT9918700015

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. CHEM. SOC. FARADAY TRANS., 1991, 87(1), 19-29 Gas-phase Metal Oxidation Reactions studied by Chemielectron Spectroscopy and Chemiion Mass Spectrometry: Reactions of Cerium and Lanthanum with 02(X 'Xi),02(a 'Ag) and O('P) Martin C. R. CockettJ John M. Dyke,* Andrew-M. Ellis$ and Timothy G. Wright Department of Chemistry, The University, Southampton, SO9 5NH, UK The gas-phase reactions M + 02(X 3Z;), M + 02(a 'Ag)and M + O(3P) have been studied with chemielectron spectroscopy, where M represents one of the lanthanide metals, cerium or lanthanum. Assignment of the observed bands has been assisted by mass analysis of the ions produced and by approximate kinetic modelling calculations. For the M + O,(X 3Z9) and M + O,(a 'Ag)associative ionization reactions, most of the excess energy appears as electron kinetic energy, whereas for the M + O(3P) reactions a larger fraction of the reaction energy is retained in the positive ion.In recent years, a number of gas-phase chemiionization re- actions have been studied experimentally by various methods.'-7 Some of these reactions, mainly involving metals and oxidants such as 0, , 0, O,, N,O and OH, have been investigated under crossed molecular beam conditions using mass spectrometry and, in some cases, relative and absolute reaction cross-sections have been determined.3.5-7 Optical emission experiments have also been used to probe the excited-state population of ions formed in such processes.8 However, in spite of the interest and activity in this field, few experiments have been performed to measure the kinetic- energy distribution of the electrons produced by these reac- tions.Chemielectron spectroscopy, the measurement of the energy distribution of electrons arising from chemiionization reactions, has the ability to provide detailed information rela- ting to the energetics and reaction dynamics of such pro- ce~ses.~A small number of chemiionization reactions have already been studied with electron spectros~opy,~-' and they serve to demonstrate the potential of the method. This paper describes the use of chemielectron spectroscopy to investigate a number of chemiionization reactions involv- ing cerium or lanthanum, with an oxidant [O(3P),O,(X 'Z;) or O,(a 'Ag)]. Reactions of this type have been studied in some detail previously using mass spectrometry, and in some cases reaction cross-sections have been measured.' 3-' The associative ionization cross-sections determined in ref.(13)-(15) have proved useful in this present work in that they have been used to calculate approximate rate constants at a partic- ular reaction temperature. These have been used in approx- imate kinetic simulations of the reactions of interest and provide insight into the M + 0, reaction scheme. One of the requirements necessary for a metal to undergo an oxidation chemiionization reaction is its ability to form a strong metal-oxygen bond and for the oxide to have a low first ionization energy. For example, if the simple associative ionization reaction between a metal atom and an oxygen atom, O(3P),is considered: M + O(3P)-+ MO+ + e-(1) the reaction enthalpy for this reaction can be written as AHl = -DE(M0) + Eia(MO) -f Present address: Institute of Molecular Sciences, Okazaki, Japan.1Present address : Laser Spectroscopy Facility, Department of Chemistry, Ohio State University, Columbus, Ohio, USA.where DE(M0) is the dissociation energy of the metal oxide and Ei, is its first adiabatic ionization energy. Clearly, AH, will only be negative if DE(M0) =-E,(MO). In the case of cerium and lanthanum, which are studied in this work, and some of the other lanthanides, this requirement is satisfied. Similarly, for the associative ionization reaction between a metal and ground-state molecular oxygen : M + O,(X 'Zg-) -+ MO; + e-(2) the reaction enthalpy can be written as, AH2 = -DE(MO2) + Eia(MO,) where D,(MO,) represents the dissociation energy of MO, to gve M and 0,.As O,(a 'Ag), as well as O(3P) and O,(X 'Xi), has been used as an oxidant in this work, reaction (3) must also be considered : M + O,(a 'Ag) -+ MO; + e-(3) If the same ionic states are accessed in reactions (2) and (3), the reaction enthalpy, AH3, can be written as AH3 = AH2 -To(X 'Xg--a 'A$ where To(X 3Xi-a 'Ag) is the energy of O,(a 'Ad relative to O,(X "i). In principle, for a given reaction, if the energy distribution of the emitted electrons is measured, and if AH is known, then the distribution of the available energy between the product ion and the chemielectron can be determined.Fur- thermore, the shape, position and structure of the chemielec- tron band can be used to provide information concerning the mechanism of the chemiionization process. The purpose of this study is to investigate the chemiion- ization reactions M + O(jP), M + O,(X 'Xi) and M + 0, (a 'Ag) using chemielectron spectroscopy. The gas-phase chemiionization reactions of atomic cerium with O(3P) and O,(X 'Xi) have been studied previously by mass spectro- metry using both accelerated cerium beams16 and thermal reactant^.'^*'^ For the reaction of cerium with O,(X 'Zi), two different chemiionization channels were observed, leading to the formation of CeO+ and CeO;, with the mea- sured relative cross-sections for these channels being ca.1 : 30.17 It is somewhat surprising that any reaction was observed at all since the chemiionization channels yielding CeO' + 0 + e-and CeOl + e-for the reaction of cerium with O,(X 'Zg-) can be calculated as being endothermic using currently available thermodynamic data (although the error on some of the calculated reaction enthalpies is large). The chemiionization cross-section at thermal energies for the reaction of cerium with atomic oxygen to give CeO' + e-, has been measured to be ca. 300 times that of the associative ionization reaction between cerium and O,(X 'Xi) to give CeO; + e-.15 The gas-phase chemiionization reactions La + 02(X 'Xi) and La + O('P) have also been studied by mass spectrometry using both accelerated lanthanum beams' and crossed beams of thermal reactants.I5 A number of cross-sections for the La + 02(X 'Xg-) reaction scheme have been rnea~ured'~ under thermal conditions and it has been found that the cross-section for the associative ionization reaction produc- ing Lao; + e-[reaction (2)] is ca. two orders of magnitude larger than that of the corresponding cerium reaction. Addi- tionally, the measured cross-section for the La + O('P) associative ionization reaction [reaction (l)] at thermal ener- gies is ca.300 times larger than that for the La + 02(X 'Xi) associative ionization reaction. Experimental All of the chemielectron spectroscopic studies described in this work were performed on a single-detector, high- temperature photoelectron spectrometer, described else-where.18-'' The metals (Cerac Inc., 99.9%) were evaporated using a radiofrequency induction heating system2' and the metal vapour was reacted with the oxidant gas in a reaction cell which allowed partial pressures of 10-5-10-3 Torr to be used.Electrons produced from the reaction were energy-analysed and then detected using a channeltron electron multiplier. As described previously,''-l2 an He I photon source was used to produce photoelectron spectra in order to calibrate the kinetic energy scale of the chemielectron spectra obtained. The photon beam was blocked using a shutter arrangement, which could be operated from outside the vacuum system, when a chemielectron band was scanned.This enabled calibrated chemielectron spectra to be recorded without contributions from photoelectrons. Cylindrical furnaces were used for the evaporation of the metals and these were machined from either molybdenum or tungsten. They were fitted with a molybdenum extension piece and supported by an alumina rod.20 The extension I VC -J J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 pieces were used to keep the alumina/molybdenum or aluminaltungsten junction away from the high temperatures experienced inside the r.f. coil region, thus preventing re-action between the molybdenum or tungsten with the alumina.20 The only other component in the furnace region was a molybdenum or tantalum radiator, which was sup- ported from the extension piece.2o921 No ceramic components or insulating carbon felt were used anywhere in the furnace region, thus ensuring that reactive gases such as CO or CO, were not generated within the reaction environment.Oxygen atoms were generated by a microwave discharge of pure molecular oxygen flowing in a boric-acid-coated glass inlet tube. This resulted in a mixture of O('P), O,(X 'Xi) and 02(a 'Ag) being present. In order to distinguish between con- tributions to the product electron and ion signal intensities arising from reaction of 0 atoms and 02(a 'Ag) with the reactant metal, the 0 atoms were deactivated, with little change in the 02(a 'A,) partial pressure, using glass wool placed in the discharge tube. Any resulting change in either the chemielectron spectrum or the chemiion mass spectrum could then be attributed to the loss of the 0 atoms in the reactant beam.The 02(a 'Ag) and 0 atom yield under a given set of discharge conditions could be determined by recording the He I photoelectron spectra of molecular oxygen, discharged oxygen, and discharged oxygen passed through glass wool present in the inlet tube (see Fig. 1). From the known He I photoionization cross-sections at right angles to the photon beam,,, the experimental band intensities could be used to yield relative partial pressures of O,(X 'Xi), 02(a 'A,) and O('P) under a given set of con-ditions. In order to assist in the assignment of an experimental chemielectron spectrum, it was valuable to identify the ionic products of the chemiionization reaction.This was achieved by fitting an independently pumped quadrupole mass spec- trometer to the front of the ionization chamber of a multi- detector photoelectron spectrometer.21 In these experiments the electron-impact source of the mass spectrometer was switched off, enabling the exclusive detection of ions produc- ed by chemical reactions. In order to prevent the reactants from contaminating the quadrupole rods, they were mixed in a reaction cell outside the mass spectrometer housing and the I I A-I I 1 I 14 13 12 11 14 13 12 11 , 14 13 12 11 EIeV Fig. 1";), (b)O,(X'ZC,),02(X(a)foreV ionization energy region 11.&14.0photoelectron spectra recorded in the IHe and (c)02(X 'C,) and 02(a 'AJ 02(a 'A,) and O(3P) J.CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 product ions were focused into the ion source of the mass spectrometer using an ion lens. A typical 'chemiion' experiment is carried out in much the same way as a chemielectron experiment: the metal is evapo- rated from an r.f. inductively heated molybdenum or tung- sten furnace and is then crossed with a beam of oxidant at right angles, in the reaction cell. The three oxidants used, O,(X 'Xi), O,(a 'A,) and O('P), were prepared as described above and again introduced into the reaction cell via a boric- acid-coated glass discharge tube. In contrast to the chemiel- ectron experiments, where the geometry of the discharge tube required that separate experiments be performed with and without the glass wool present in the discharge tube, in a chemiion experiment all three oxidant mixtures could be pro- duced during the course of a single experiment by running the discharge either on the side of the glass wool furthest from the ionization chamber or on the side closest to the ionization chamber to obtain mixtures of O,(X 'Xi) and O,(a 'A,), or O,(X 'Z,-), O,(a 'Ag, and O('P), respectively [see Fig.l(c) and (b), respectively]. Pure 02(X 'Xi) could obviously be obtained by simply switching off the discharge [see Fig. l(a)]. Unfortunately, the relative yields of O,(X 'Xi), O,(a 'A,) and O('P) could be monitored for a chemiion experiment only by performing a separate photo- electron experiment on the multidetector instrument without the ion lens assembly in place, whereas for the chemielectron experiments, which were performed on the single-detector instrument, the relative yields of the three oxidants could be continually monitored during the course of a particular experiment. With the above procedure, the relative partial pressures of the three oxidants for a given discharge configu- ration could be calculated in a chemiion experiment in the same way as in the chemielectron experiments from the known photionization cross-sections of O,(X 'Z,-), O,(a 'A,) and O('P) at the He I photon energy.,, From both the chemielectron and cherniion spectra obtained, the relative reaction cross-sections can be calcu- lated, once the relative partial pressures of O,(X 'Xi), O2(a1A,) and O('P) are known, by measuring the relative chemielectron band and chemiion signal intensities under the discharge on and discharge off conditions (see Fig.1). The method used to obtain relative cross-sections will be described later. Results and Discussion Ce +02(X 'Zr) Previously we have described" the study of the Ce +0, (X 'Xi) reaction with chemielectron spectroscopy in some detail. The band shown in Fig. 2, reproduced from ref. (12) for comparison with Fig. 3, was attributed to electrons arising from the associative ionization reaction of cerium with O,(X 'Xi) [i.e. reaction (2) with M =Ce]. This assign- ment was achieved on the basis of evidence from a number of sources: (a)mass analysis of the ions, (b)approximate kinetic modelling calculations and (c) measurement of the electron and positive ion concentrations as a function of oxygen pres- sure in the reaction region.Ce +O(3P);Ce +02(a 'Ag) The chemielectron spectrum recorded at a furnace tem-perature of 1300 K, for the reaction between Ce and a mixture of O,(X 'Xi), O,(a 'A,) and O('P), generated in a microwave discharge in flowing oxygen gas, is shown in Fig. 3(a). With the microwave discharge on, a sharp band was recorded, with the photon source off, having a maximum at photon source (He11 off on I 0; (b4Zg) m U EIeV Fig. 2 Chemielectron spectrum recorded for the Ce +O,(X "C,) reaction 0.13 k0.06 eV and a tail extending to higher kinetic energy. At higher temperatures in the range 1350-1450 K, a second band with a maximum at 1.83 & 0.10 eV electron kinetic energy was observed which was much broader than that of the low-kinetic-energy band (Fig.3). When the microwave discharge was switched off at furnace temperatures >1450 K, both bands disappeared to reveal the band due to the reac- tion between cerium and ground state molecular oxygen (Fig. 21, having an intensity at least two orders of magnitude smaller than the bands recorded with the discharge on. However, when the microwave discharge was switched off at lower furnace temperatures, the intensity of the chemielectron band due to the Ce +O,(X 'Xi) reaction was below the detection limits of the spectrometer.This band was observed only when the furnace temperature was increased to 1450-1600 K. Unfortunately, no reliable band intensity ratios could be calculated from the observed band intensity changes because of the small number of data points available However, the qualitative changes observed in the chemielec- tron band intensities can be compared with the behaviour observed in the corresponding ion spectra. When the dis- charge was switched on, the chemiion mass spectrum showed a large increase in the intensity of the CeO+ signal together with a smaller increase in the intensity of the CeOl signal. The actual intensity changes will be presented later together with a description of the method used to calculate relative reaction cross-sections from the chemiion intensity changes.It is clear that the two chemielectron bands observed with the discharge on must be due to chemiionization reactions between Ce and O('P), and Ce and O,(a 'A,). However, with oxygen atoms deactivated only the feature centred at 1.83 & 0.10 eV was present, thus enabling assignment of this band to a chemiionization reaction between Ce and O,(a 'A,). Consequently, the band at 0.13 0.06 eV can be assigned to the associative ionization reaction between Ce and O('P), producing CeOf (see Fig. 3). It seems, from the J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 Id lamp off on 300 ;e+0'3p1 - Iln ln C 4- 3 0 I ' 0 I 1 lump-lampoff on off on I CC) +-I--+ II 0" D) -I I II 3~10~ I I I I Irrm Ic I I I I I 0 I I I I I I I 7-0 1 2 3 0 1 2 3 4 0 1 2 3 1, EIeV Fig.3 Chemielectron spectra recorded for the reactions Ce +O(3P)and Ce +O,(a 'Ag) at furnace temperatures of (a) 1300 K, (b)1370 K and (c) 1450 K dependence of the intensity ratio of the two chemielectron bands seen in Fig. 3 on temperature, that the Ce +O,(a 'Ag) and Ce +O,(X 'Xu-) reactions have significant activation energies whereas the Ce +O(3P) reaction does not. This point will be discussed later. Neither of the two bands assigned as arising from the Ce +O,(a 'Au) and Ce +O(3P) reactions exhibited vibra- tional structure. The associative ionization reaction [reaction (l)] between cerium and O(3P) is exothermic by 2.86 k 1.78 eV and consequently this amount of energy should be avail- able to the products of the reaction in the form of internal excitation energy and kinetic energy. Clearly, if the product electrons are emerging from reaction (1) (with M = Ce) with a kinetic energy of only 0.13 & 0.06 eV (at the band maximum), then the CeO' ion must be taking up the excess 2.73 5 1.84 eV of energy in the form of either electronic or vibrational excitation energy at this point on the chemielec- tron band.Since no vibrational structure is resolved in the chemielectron spectrum for the Ce +O('P) reaction, it seems likely that the product ion CeO' is formed in a highly vibra- tionally excited state, resulting in transitions from an autoionizing state, CeO*, to the closely spaced vibrational levels on the ionic state potential curve (Fig.4). The separation of the zeroth vibrational levels in O,(a 'Au) and O,(X 'Xu-) is 0.98 eV.23 Assuming that the band maximum of the electron kinetic energy distribution mea-sured for the Ce +O,(X 'EL) associative ionization reaction [reaction (2)] of 0.90 0.04 eV is a lower limit of the exo- thermicity of this reaction, then the energy available to the products of the Ce +02(a 'Ag) associative ionization re-action [reaction (3)] must be at least 1.88 Ifi 0.04 eV. Clearly then, the measured value for the band maximum of 1.83 & 0.10 eV electron kinetic energy for this reaction indi- cates that the product CeOl ion is in the same electronic state as that of the CeO; formed from the Ce +02(X 3E9) reaction.A schematic potential-energy diagram describing the Ce +O,(a 'Au) reaction is presented in Fig. 5. In practice, although qualitative information could be obtained from the chemielectron spectral intensity changes, relative reaction cross-sections could be calculated only for the chemiion experiments. Nevertheless, it was evident from the measured differences in the chemielectron band intensities at the highest experimental furnace temperatures, that the cross-section for the Ce +O(3P)reaction [reaction (l)] was several hundred times larger than the cross-section for the Ce +O,(X 'Xi) reaction [reaction (211 (consistent with the cross-section ratio measured in a molecular-beam study of ca.330 :1 15) and was of the same order of magnitude as the cross-section for the Ce i-O,(a 'Au) reaction [reaction (3)]. Ce++0 CeO reaction coordinate Fig. 4 Schematic potential-energy diagram for the Ce +0,(3P) reaction J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 reaction coordinate Fig. 5 A schematic potential-energy diagram for the Ce + O,(a 'Ag) reaction. The dotted curve represents schematically the position of the excited state obtained from the Ce + O,(X 3C,' reaction In order to relate the chemiion signals with the chemielec- tron signals, the relative reaction cross-sections for pro-duction of a particular ionic product (in this case either CeO' or CeO;) were calcuiated from the chemiion signal intensity changes as a function of discharge conditions (ie.discharge off or discharge on, with and without glass wool). Representative chemiion mass spectra for the three experi- mental conditions are shown in Fig. 6. In Fig. qa) the dis- charge is of€ and the ratio CeO: : CeO' recorded from spectra of this type averaged from all the recorded experi- mental mass spectra, is (6 f1) : 1. A chemiion mass spectrum obtained with the discharge on is shown in Fig. 6(b).The intensity of the CeO' signal is now greater than that of the CeO; signal. In fact, the measured ratio CeO; : CeO', aver-aged from a large number of experimental mass spectra of which Fig. 6(b)is representative, is 1 : (3 l),corresponding to a ca.two-fold increase in the CeOz intensity and a ca. a) I' 172 40-fold increase in the CeO' intensity over the respective CeO; and CeO' intensities measured with the discharge off. In Fig. qc), glass wool is present in the discharge tube to deactivate the O(3P). The average CeO; :CeO' intensity ratio measured under these conditions is (1.4 & 0.5) :1. This arises as a consequence of a ca. two-fold increase in the CeOi intensity and a ca. 10-fold increase in the CeO' inten-sity over the respective CeO; and CeO' intensities measured with the discharge off. In order to calculate the relative reaction cross-sections for these reactions, the intensity changes described above must be combined with the relative partial pressures of 02(X 'Z;), O,(a 'Ag) and O(3P)present in the reaction region for each of the discharge configurations.This may be accomplished by application of the following expression to a particular set of experimental data: 14Pij = z; I where a: denotes the relative cross-section for the chemiion- ization channel involving reactant i and resulting in the pro-duction of ion a; Pjjis the relative partial pressure of oxidant i, under the discharge configuration j; and Z; is the normal- ized intensity of the signal due to ion a, under the discharge configuration j. Use of this expression yields a chemiionization cross-section for production of CeOl from the Ce + 02(a 'Ag) reaction of (15 f7) : 1, relative to Ce + Oz(X 'Zg-). Similarly, the chemiionization cross-section for production of CeO ' from the Ce + O('P> reaction is (170 f60) : 1, relative to CeO' produced from Ce + 02(X 'Xi).Additionally, the Ce + 02(a 'Ag) reaction was found to produce a significant amount of CeO' and the corresponding relative cross-section for this reaction route was calculated as (40 & 20) : 1. As the data were obtained at furnace temperatures of 1450 & 50 K, these results compare we11 with the general behaviour exhibited in the chemielectron experiments at a furnace temperature of 1450 50 K, although it was not pos- sible to quantify the chemielectron data in terms of relative reaction cross-sections for the reasons mentioned earlier. However, the chemiion intensity ratios for production of CeO' and CeO; do not compare well with the observed behaviour in the chemielectron experiment performed at the 156 I 1ceojce$ 156 1 171 171,I I 15i A Fig.6 Chemiion mass spectra recorded for the reactions (a)Ce + O,(X %;), (b)Ce + O,(X 'Xi), Ce + O,(a 'Ag) and Ce + O('PP) and (c) Ce + O,(X 'ZC,) and Ce + O,(a 'A,) lower furnace temperature of 1300 50 K. At this tem- perature it was not possible to measure the intensity ratio of the chemielectron bands assigned to the as_sociative ioniza- tion reactions Ce + O('P) and Ce + O,(X 'Xi) because no electrons were observed when the discharge was switched off, although the ratio is expected to be high. Clearly, at the lowest temperatures at which chemielectron spectra were recorded (1300 K), a significant difference exists between the results obtained from the chemielectron spectra and those obtained from the chemiion mass spectra.This can be explained by considering the effect of the activation energies for the Ce + O,(X 'Xi) and Ce + G,(a 'Ag) reactions on the reaction cross-section evaluated from experimental signal intensities obtained at diflerent temperatures. At the lowest temperatures that reasonable chemielectron spectra could be recorded for the Ce + O('P) reaction [see Fig. 3(u)], the available thermal energy would be insufficient to overcome the activation energy barrier of the Ce + Oz(X 'Xi) and the Ce + 02(a 'Ag) reactions. This would result in only one band being observed in the chemielectron spectrum, assignable to electrons emerging from the Ce + O('P) associative ioniza- tion reaction.At higher temperatures [Fig. 3(b) and (c)], the activation energy barriers for the Ce + O,(X 'Xi) and Ce + O,(a 'Ag) reactions are overcome and three chemielectron bands would become evident in the chemielectron spectrum [although the band due to Ce + O,(X 'Xi) is still at least one order of magnitude weaker than the bands arising from the Ce + O,(a 'Ag) and Ce + O(3P)reactions]. Obviously, much of this discussion is qualitative. Neverthe- less, the chemiion spectra are consistent with the assignment of the chemielectron spectra and assist in the understanding of secondary reactions that occur in the reaction cell under the reaction conditions.'2 La + O,(X "Z;) The chemielectron spectrum obtained from the reaction between lanthanum and O,(X 'Xg-)at a furnace temperature of 1600 & 50 K is shown in Fig.7. A band of some structural complexity was recorded with the photon source off with a maximum measured as 1.03 & 0.20 eV. The large error 10 I I I I I I I I I I I I Iln I I ln Ic C I 3 I I 0 I I I I I I 1 2 3 EfeV Fig. 7 Chemielectron spectrum recorded for the reaction La + O,(X 3q) J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 quoted for the position of the band arose because the band maximum moved in position from 1.25 eV (at an estimated furnace temperature of 1500 K) to 0.88 eV (at a furnace tem- perature of ca.1700 K). In all experiments, as the furnace temperature was increased, so the position of the chemielec- tron band maximum moved to lower kinetic energy. This point will be discussed later. The band is more symmetrical in shape than the band obtained for the Ce + 02(X 'Zi) reaction and vibrational structure attributable to two different series was resolved in this band (see Fig. 7). The average separations of the com- ponents of these two series (denoted X and Y in Fig. 7) were measured at 670 30 and 655 f30 cm-'. The chemiion spectrum recorded for this reaction again showed signals due to the metal monoxide and metal dioxide ions (Fig. 8), but in contrast to the Ce + O,(X 'Xi) chemiion spectrum,12 the Lao' signal was found to be more intense than the Lao; signal at oxygen partial pressures in the reac- tion region of between and Torr.The experimen- tal variations of the Lao' and Lao; signal intensities with the oxygen partial pressure in the reaction region between 0 and 0.8 x lo-' Torr are shown in Fig. 9(u) and (b),respec-tively. At oxygen pressures of <6 x Torr, Lao: increases with oxygen pressure, but as the pressure is increased above this value a gradual levelling off is observed. Similar behaviour is also noted for the Lao' pressure plot, although the scatter in the data makes it difficult to ascertain the precise point at which this levelling occurs (see Fig. 9). The oxygen pressure dependence plot for the chemielectron signal, presented in Fig.9(c), exhibits approximately linear behaviour over the full range of oxygen pressures considered. At oxygen partial pressures in the range (2.0-2.5) x lo-' Torr, the Lao; :Lao+ ratio changes in favour of the dioxide ion. The experimental oxygen pressure dependence plot over this range is presented in Fig. 10. Unfortunately, it was not possible to measure chemielectron intensities at oxygen partial pressures >1 x lo-' Torr. However, the combined effect of these results is that it is difficult to associ- ate either Lao: or Lao' uniquely with the observed chemi- electron signal and hence to identify the main chemiioniza- tion reaction responsible for the observed chemielectrons. Clearly, evidence from other sources is required to achieve an assignment of the La + O,(X 'Xi) chemielectron spectrum and this is presented later. A kinetic model of the La + O,(X 'Zg-) system can be con- sidered as it may provide some evidence towards identifying a dominant reaction channel. 139La0+ 155 I 39~ao; 171I I I I 10' 170 165 16 0 155 tTl/Z Fig.8 Chemiion spectrum recorded for the reaction La + 0, (X 'ZC,); the oxygen partial pressure used was ca. 7.5 x Torr J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 0.2 0.6 0.8 / I I I I 0.2 0.4 0.6 (c 1 / d.2 0.4 0.5 0.8 P(02)/mTorr Fig. 9 Experimental oxygen dependence plots of (a) [Lao'], (b) [Lao:] and (c)[e-], for the La + O,(X 'Zg-) reaction in the oxygen pressure range 0.8 mTorr.All points have been normalised to the largest value As in the Ce + O,(X 'Zg-) case1, eight possible reactions were considered in constructing a reaction scheme for the La + O,(X 'ZgJ system (Table 1). Unfortunately, fewer thermochemical data are available for the La + O,(X 'Z;) system than the Ce + O,(X 3Z9) system. Nevertheless, use can be made of the data in ref. (12) and (24H31) together with a value of 9.5 f 1.5 eV for the first ionization energy of Lao,, estimated by comparison with the known value for CeO,,,* to obtain reaction enth- alpies for each of the reactions (4H11). Comparison of the reaction enthalpies obtained with those calculated for the Ce + O,(X 'Z;) system12 shows very similar trends. Reaction (4) would seem to be the favoured chemiionization channel for reaction of La with O,(X 'Zg-) since it is the only che- 1 I I I 0.5 1.0 1.5 2.0 2.5 P(O,)/mTorr Fig.10 Experimental oxygen pressure dependence plots for [Lao:] (+) and [Lao'] (0)for the La + O,(X 'EL) reaction, in the oxygen pressure range 0-2.5 mTorr Table P Reactions considered in constructing a reaction scheme for the La + O,(X 'ZC,)systema *Hr lev klcm molecule-' s-' La + 0, -+ Lao: + e- -1.08 f 1.95 1.5 x lo-'' (4) La+O,+LaO+ +O-La + 0, -+Lao+ + 0 + e-La+o,-+La+ +O; +0.41 f 0.32 +5.11 f 0.10 + 1.87 f 0.40 9.0 x 1.0 x 1Q-18 1.0 x lo-', (5) (6)(7) La + 0, --+ LaQ + 0 Lao+ + 0, -+Lao: + 0 La+O-+LaO+ +e-Lao: + La -+ Lao' + La0 -3.08 1.18 +2.00 f 0.68 -3.29 f 0.32 -5.29 f 2.36 1.0 x lo-" 1.0 x 10-14 6.2 x lo-'' 1.0 x lo-" (8) (10) (9) (11) See text for further details.miionization route which is sufficiently exothermic to produce electrons under thermal conditions. However, as in the Ce + O,(X 'Z;) case, the neutral reaction (8) is highly exothermic and the 0 atoms produced would be expected to undergo an associative ionization reaction with La [reaction (9)], to produce Lao+ and e-. As in the Ce + O,(X 3Zg) case,I2 the rate constants for reactions (4), (9,(6) and (9) were calculated from the available cross-sections measured by Fite et d.," assuming a reaction temperature of 1080 K. Also as in ref. 12, the remaining rate constants were estimated by using the calculated reaction enthalpies as an approximate guide.The reaction time after which ion and electron intensities were measured was esti- mated as 4 ms, from flow-rate measurements. Simulations were therefore carried out for reaction times in the range 1-8 ms. A plot of the computed concentrations of Lao:, Lao' and e- as a function of reaction time in the range 1-8 ms and at initial partial pressures of the reactants, La and O,, of 1 x Torr is shown in Fig. ll(a). The vapour pressure of lanthanum at 1600 K was obtained from ref. 29. For reaction times of up to 8 ms, the Lao+ concentration was found to be substantially greater than the Lao: concentration, which at least qualitatively agrees with the relative intensities of the two ions observed in the experimental chemiion spectrum recorded for pressures <8 x lov4Torr in the reaction region (Fig.10). Plots of the individual contributions of reactions (9) and (11) to the total Lao' concentration and of reactions (4) and (9) to the total electron concentrations are presented in Fig. ll(b). No plot for the contributions of individual reac- tions to the Lao: concentration is presented because reac- tion (4) is the only reaction to contribute significantly to the production of Lao:. These plots indicate that one of the major sources of both Lao' and electrons is reaction (9). Note that reaction (5) also contributes to the total Lao' concentration but to a far smaller extent than reactions (9) and (11) and for this reason has been excluded from Fig. 1 l(b). Additionally, although reaction (6) will in principle con- tribute to the Lao+ and electron concentrations, it has been excluded from Fig.ll(b) because its contributions were found to be small [although it has been included in the full kinetic scheme used to obtain the results presented in Fig. 1l(a)]. The simulated dependence of the product ion and electron concentrations on the initial partial pressure of oxygen proved to be more revealing. The computed pressure depen- dence plot of the Lao;, Lao' and e- concentrations over the oxygen pressure range 04.1 Torr is shown in Fig. ll(c). Clearly, the dominant ion is the monoxide ion up to 0.01 Torr. Thereafter the La0 'concentration decreases, while the Lao; concentration continues to increase until the Lao; :Lao+ concentration ratio changes in favour of the dioxide ion at 0.06 Torr.This behaviour is in good qualitat- ive agreement with the behaviour seen in the experimental pressure plots (see Fig. 10). J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 0 0 a 0 0 0 a 02468 .' '**1 , [Lao+] . t electrons 0 t (9) 04°. 0 0 0. 0 0 0 0 0 0 0 . '1.+ . 0' 1 I 2 4 6 8 0 5 10 reaction time/ms Fig. 11 (a)Computed plots of [Lao:] (x), [Lao+] (+) and [e-] (0)";) O,(X+reaction time, for the Laofas a function reaction. (b) Computed plots of the contributionsof the reactions involved in the La + O,(X 'XC,) reaction to production of [Lao'] and [e-1, as a function of reaction time.(c) Computed oxygen pressure dependence plots for [Lao:], [Laof] and [e-] for the La + O,(X 'X;) reaction in the oxygen pressure range &lo0 mTorr. See text for details of the method used It would be sensible at this stage not to draw any firm conclusions from the results of the kinetic model discussed above. Nevertheless, it does prove to be a useful exercise in building up to a qualitative picture of the overall reaction scheme. For example, it seems reasonable to conclude that the chemielectron band in Fig. 7 is due to electrons from the two competing associative ionization processes (4) and (9), since neither reaction has been shown to be sufficiently dominant under the reaction conditions used, to enable an assignment of the chemielectron spectrum exclusively to one reaction.Since reaction (9) is almost certainly contributing to the chemielectron band shown in Fig. 7, the obvious way of investigating this is to record the chemielectron spectrum for the reaction between lanthanum and oxygen atoms by producing the latter using a microwave discharge in molecu-lar oxygen. La + O(3P);La -k O,(a 'Ag) The chemielectron spectrum recorded for the reaction between La and a mixture of O,(X 'Xi), O,(a 'Ag) and O(3P),is shown in Fig. 12. With the microwave discharge on, two chemielectron bands were recorded, with the photon source off, having band maxima at 0.67 & 0.06 and 2.41 & 0.06 eV electron kinetic energy. No vibrational struc-ture was observed on either band and the average intensity ratio of these bands was measured as (3.2 0.5) :1, respec-tively. No correction for the transmission function of the spectrometer has been made in the relative 'nand intensity ratio quoted.When the discharge was switched off, both bands disappeared to reveal the chemielectron band at 1.03 f 0.20 eV. The band at 0.67 & 0.06 eV has a measured signal intensity ca. 100 times greater than that of the band observed with the discharge off, whilst the band at 2.41 f0.06 eV is ca. 30 times more intense than the Sand observed with the discharge off. All intensities quoted were measured at oxygen partial pressures in the reaction region of up to 10-Torr. The ion spectrum recorded with the discharge on shows an increase in the Lao+ signal intensity compared with that observed with the discharge off, but no apparent increase in the Lao; signal intensity was observed.This is somewhat surprising, as it might be expected that a reaction between La and O,(a 'Ag) would produce a significant increase in the dioxide ion partial pressure, as was observed in the cerium case, owing to an associative ionization reaction between the two reactants. Secondary reactions (9H11) will, however, affect the observed Lao: and Lao+ signal intensities. The conclusion that was drawn from these observations was that the two bands centred at 0.67 f 0.06 eV and 2.41 f 0.06 eV in Fig. 12 must arise from reactions between I I 0 1 2 3 EjeV EjeV Fig. 12 Chemielectron spectra recorded for La + O(3P)+ O,(a 'ALP)and (b)La + O,(X "Cg) the reactions (a) J.CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 La and 0(3P),and La and 02(a 'Ag). In order to assign these bands positively, the experiment was repeated with glass wool in the discharge tube. The result of this experiment was that only the band centred at 2.41 & 0.06 eV was observed when the discharge was switched on (Fig. 13), thus allowing assignment of this band to a chemiionization reaction between lanthanum and O,(a 'Ag). The band centred at 0.67 f0.06 eV electron kinetic energy in Fig. 12 can now be confidently assigned to reaction (9), and the band centred at 1.03 f0.20 eV in Fig. 7 can be assigned to reaction (4).In view of the fact the measured maximum cjf the band assigned to reaction (9) (0.67 & 0.06 eV) lies within the lower kinetic energy side of the band in Fig.7, it seems likely that reaction (9) is contributing to the chemielectron band recorded from lampoff on I I I I I 1 2 3 4 5 6 E/eV Fig. 13 The chemielectron La + O,(a 'Ag) reaction spectrum recorded for the Lao+ 155 171 I 0;La the La + O,(X %;) reaction in Fig. 7. This would be consis- tent with the consecutive reactions (8) and (9) occurring in the reaction_cell under the conditions used. The ratios of the discharge on : discharge off chemielectron band intensities can provide approximate estimates for the relative chemiionization cross-sections for each of the re-actions contributing to the observed chemielectron band envelopes recorded in Fig.7, 12 and 13. The method used is identical to that described earlier for the calculation of the relative reaction cross-sections from cherniion spectral inten- sity changes for the analogous reactions involving cerium. In this way the relative chemielectron reaction cross-sections, 0, for each of the reactions La + 02(X 'Z;), La + O,(a 'Ag) and La + O(3P) have been calculated as 1 : (80 L-20) : (750 180) at an oxygen pressure in the reaction cell of Torr. Some discrepancy occurs when the cross-sections are evaluated from the chemiion mass spectra. Representative chemiion mass spectra for the three discharge conditions are presented in Fig. 14. In Fig.1qa) the discharge is off and the average intensity ratio of Lao,+ : Lao+ is 1 :(3 1) at a pressure in the reaction cell of CQ. Torr. A typical chemiion mass spectrum obtained with the discharge on [see Fig. l(b)] is presented in Fig. 14(b). The intensity of the Lao+ signal has increased to the point where it is ca. 26 times more intense than the LaOf signal. This arises from a ca. nine-fold increase in the Lao+ signal intensity. In Fig. 1qc) glass wool is present in the discharge tube to deactivate the O(3P).The measured intensity ratio Lao; : Lao+ averaged from a number of spectra of the type presented in Fig. 1qc) is 1 :(6 4) which is accounted for by a cu. two-fold increase in the Lao+ intensity over that recorded with the discharge off.These intensities convert, using the method described pre- viously, to relative reaction cross-sections for production of Lao+ from the three reaction routes, La + O,(X %,), La + 02(a 'Ag) and La + O(3P),of 1 : (6 f4) : (35 f15), respec-tively. For production of Lao:, relative cross-sections of 1 : (1.0 f0.1) for La + O,(X ?Zg-) and La + O,(a 'Ag) were obtained. Although these are qualitatively in agreement with Lao+ 155 I Lad 155 I mlz Fig. 14 Chemiion mass spectra recorded for the reactions (a)La + O,(X 'Xi), (b)La + O,(X 'XE-),La + O,(a 'A,) and La + O(3P)and (c) L,a + O,(X 'Xi) and La + O,(a 'A,) J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 the chemielectron-derived values, they are quantitatively very different.One reason for this large difference may arise because the ions observed in the chemiion experiments are extracted from the reaction cell under relatively high pressure conditions and their concentrations will be affected by sec- ondary reactions. For example, Lao; produced by reaction (4) will react rapidly tlia reaction (11). This would reduce the observed Lao; signal relative to Lao+. Also, if the mass- spectrometric results were obtained at a temperature some- what higher than that used to obtain the chemielectron spectra, as appears to be the case experimentally, then the effects of activation energy differences between the three asso- ciative ionization reactions, La + O,(X 'Z;), La + O,(a 'Ag) and La + O(3P),would become evident on comparison of the derived cross-section values from the two experimental methods.At the higher temperatures, the activation energy of the La + O,(X 'Z;) reaction is overcome and this reaction contributes more to the overall reaction scheme with the dis- charge on relative to the La + O(3P) and La + O,(a 'Ag) reactions. At lower temperatures, the La + 02(a 'Ag) and La + O(3P)reactions dominate to a much larger extent over the La + O,(X 'Xi) reaction and this is reflected in the cal- culated chemielectron reaction cross-sections. It should also be remembered that the two sets of data were recorded on two separate instruments and this potentially introduces a number of instrumental factors which may affect the validity of direct comparisons.Comparison of the available reaction enthalpy and the product electron kinetic energy, measured from the cherni- electron band maximum, for the reactions of lanthanum with O,(a 'As) and O(3P) shows similar trends to that of the cerium case discussed earlier. The associative ionization reac- tion (9) between La and O(3P)is exothermic by 3.29 f0.32 eV and so this energy should be available to the products of the reaction. In this case, the observed chernielectrons are emerging from reaction (9) with a kinetic energy of 0.67 f0.06 eV at the band maximum, which leaves an excess of 2.62 f0.38 eV of energy to be taken up by the product Lao+ ion in the form of internal excitation energy, for the autoionizing transition corresponding to the band maximum.The relationship between the reaction enthalpy (as calcu- lated from available literature data) and the experimental electron kinetic energy can be investigated in more detail by measuring the onset on the high electron kinetic energy (e.k.e.) side of the chemielectron band as well as the band maximum. This has been done for the six reactions investi- gated and the values obtained are listed in Table 2. The high e.k.e. onset, when reduced by the excitation energy in the reactant channel at the temperatures used (estimated for the reactions studied as 0.15 eV), will provide a lower limit for the exothermicity of the reaction studied, as the Franck-Condon factors in the region of the onset will be very low and the true high eke.onset may not be observed. If the M + 0 values listed in Table 2 are considered first, it can be seen that for the both the La + 0 and Ce + 0 reac-tions, the high e.k.e. onset is less than the expected reaction exothermicity by >1 eV. This indicates that even at the chemielectron band onset an appreciable fraction of the re- action exothermicity is retained in the ion. For the La + O,(X 'Z;) and Ida + O,(a 'Ag) reactions, although the errors on the calculated heats of reaction are larger, it appears that at the high e.k.e. band onset almost all of the reaction enthalpy is converted to electron kinetic energy. A similar situation is expected to occur in the analo- gous cerium reactions, although in this case the high e.k.e. onsets measured for the Ce + O,(X 'Xi) and Ce + O,(a 'Ag) reactions are inconsistent with available reaction enthalpies (the AH values evaluated from available literature data are too positive by at least 1.5 eV).The only conclusion that can be drawn from these measurements is that the thermodyna- mic data used to calculate AH for the Ce + 02(X 'Xi) and Ce + 02(a 'AJ reactions listed in Table 2 are unreliable. In ref. 12, the atomization energy of CeO, and the first adiabatic ionization energy of CeO, was used to calculate AH for the Ce + O,(X 'Zi)reaction. Having assigned the main chemielectron band recorded for the reaction between lanthanum and O,(X 'Xi) to the associative ionization process which produces Lao;, atten-tion can now be given to the two vibrational series observed in this band which must be associated with vibrational modes in Lao;.If the Lao; and Lao; states involved in the autoionization process possess a C,, or D,, equilibrium geometry, then the asymmetric stretching mode, v3, can be ruled out as a contributor to the structure seen in the band, as it will be forbidden in single-quantum transitions. To our knowledge, no experimental values for the two single-quantum-al!owed modes, v1 and v2, have been reported in the literature for either Lao; or Lao,. However, inspection of the results of matrix-isolation infrared spectroscopic studies on the dioxides of other lanthanide sug-gests that the bending mode, v2, of Lao; will be of the order of 200 cm-' or less.Since the average separations of the components in the two vibrational series (655 f30 and 670 f30 cm-') are much larger than this value, the two series cannot be assigned to progressions in both the v1 and v2 modes. A more plausible explanation is that both series correspond to excitation in v1 but that they differ by one quantum in the v2 mode. This suggestion is supported by the fact that the two series are displaced from one another by 250 fSO cm-', which is in reasonable agreement with the estimate made above for the value of the v2 mode in the ion of 200 cm-'. Alternatively, if the Lao: state has a C, or C,, equilibrium geometry (i.e.an La-0-0 structure), then all three vibrational modes would be expected to contribute to the observed chemielectron band.As with the Ce + O,(X 'Xi) associative ionization reac- tion,' the classical turning-point mechanism may be invoked in order to explain the vibrational structure in terms of a potential-energy diagram and a schematic diagram for the lanthanum plus oxygen chemiionization reaction is shown in Fig. 15. A rather shallow potential-energy curve describing the autoionizing state of neutral Lao, can be used to explain Table 2 Comparison of the high e.k.e. band onsets, band maxima and calculated heats of reaction for the reactions studied in this work reaction band maximum/eV high e.k.e. band onseta/eV heat of reactionb/kJmol-' Ce + 02(X 'Xi) -+ CeOi + e-0.90 f0.04 Ce + O,(a 'Ag)-+CeOi + e-1.83 f0.10 Ce + O('P) -+ CeO+ + e-0.13 f0.06 La + 02(X 'XC,) -,Lao; + e-1.03 f0.20 La + 02(a 'Ag) -,Lao: + e-2.41 f0.06 La + O('P) -,Lao+ + e-0.67 f0.06 a This work.Calculated from available thermochemical data (see text). 2.4 k 0.1 +0.49 f0.42 3.1 f0.3 -0.41 f0.42 0.9 -t 0.1 -2.86 f1.78 1.5 f0.1 -1.08 f1.95 3.3 f0.1 -2.06 f1.95 1.8 f0.1 -3.29 & 0.32 J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 >. P C a¶-m.-c a¶I a reaction coordinate Fig. 15 Schematic potential-energy diagram for the La + O,(X "c,)reaction the shifts in band maximum of the electron kinetic energy distribution for small changes in k,T. As can be seen in the schematic diagram in Fig. 15, as the temperature increases, the turning point moves to a smaller reaction coordinate and the autoionization transition takes place to a point on the ionic potential-energy surface higher in energy.The result of this is that the band maximum of the electron distribution moves to lower kinetic energy as the temperature increases, as is observed experimentally. As with cerium, it is clear that in order to achieve a more complete understanding of the overall reaction scheme involved in these chemiionization reactions, more precise kinetic and thermodynamic data are needed together with a temperature dependence study of the reaction cross-sections. In conclusion, cerium and lanthanum show a number of similarities in the chemiionization reactions studied. From the measured onsets on the high-kinetic-energy side of the chemielectron bands for the M + O,(X 'Xi) and M + O,(a 'Ag) reactions, most of the reaction exothermicity is con-verted into electron kinetic energy.Alternatively, for the M + O('P) reactions studied, the electron carries away only a fraction of the available energy and most of the excess energy is retained in the product ion. Experiments are currently in progress on the reactions of other lanthanides (Pr, Nd, Sm, Eu and Gd) with the three oxidants used in this work. Avail- able thermodynamic data show that on moving across the lanthanide series the heats of reaction of the associative ion- ization reactions become more positive. It will, therefore, be of interest to investigate how the excess energy is partitioned between the chemielectron and the product ion in each case.This work was supported in part by the Air Force Office of Scientific Research (grant no. AFOSR-89-035 1) through the European Office of Aerospace Research (EOARD), United States Air Force. M.C.R.C. and T.G.W. are grateful to the SERC for research studentships and A.M.E. acknowledges support via a CASE studentship with CEGB. The assistance of Dr. M. Feher in the early stages of this work is also acknowledged. References 1 R. H. Burton, J. H. Brophy, C. A. Mims and J. Ross, J. Chem. Phys., 1980, 73, 1612. 2 T. Mochizuki and K. Lacman, J. Chem. Phys., 1976,65,3257. 3 G. J. Diebold, F. Engelke, H. U. Lee, J. C. Whitehead and R. N. Zare, Chem. Phys., 1977,20,265. 4 R.Haug, G. Rappenecker and C. Schmidt, Chem. Phys., 1974, 5, 255. 5 R. B. Cohen, C. E. Young and S. Wexler, Chem. Phys. Lett., 1973, 19,99. 6 H. H. Lo and W. L. Fite, Chem. Phys. Lett., 1974, 29, 39. 7 R. B. Cohen, P. Majeres and J. K. Raloff, Chem. Phys. Lett., 1975,31, 176. 8 J. S. Winn, in Gas Phase Chemiluminescence and Chemiionization, ed. A. Fontijn, North-Holland, Amsterdam, 1985. 9 N. Jonathan, A. Morris, M. Okuda and D. J. Smith, J. Chem. Phys., 1971,55, 3046; J. M. Dyke, J. Chem. Soc., Faraday Trans. 2, 1987,83, 67. 10 J. M. Dyke, M. Feher and A. Morris, J. Phys. Chem., 1947, 91, 4476. 11 J. M. Dyke, A. M. Ellis, M. Feher and A. Morris, Chem. Phys. Lett., 1988, 145, 159. 12 M. C. R. Cockett, J. M. Dyke, A. M. Ellis, M.Feher and A. Morris, J. Am. Chem. Soc., 1989, 111, 5994. 13 W. L. Fite, H. H. Lo and P. Irving, J. Chem. Phys., 1974, 60, 1236. 14 J. C. Halle, H. H. Lo and W. L. Fite, J. Chem. Phys., 1980, 73, 5681. 15 W. L. Fite, T. A. Patterson and M. W. Siegel, U.S. Air Force Report AFGL-TR-77-0030, 1976;Hanscom AFB, U.S.A. 16 C. E. Young, P. M. Dehmer, R. B. Cohen, L. G. Pobo and S. Wexler, J. Chem. Phys., 1976,64, 306; 1976,65,2562. 17 H. H. Lo, unpublished results, quoted in ref. 14. 18 J. M. Dyke, N. Jonathan and A. Morris, in Electron Spectros- copy, ed. by C. R. Brundle and A. D. Baker, Academic Press, London, 1979, vol. 3. 19 J. M. Dyke, N. Jonathan and A. Morris, Int. Rev. Phys. Chem., 1982, 2, 3. 20 D. Bulgin, J. M. Dyke, F. Goodfellow, N. Jonathan, E. Lee and A. Morris, J. Electron Spectrosc. Relat. Phenom., 1977, 12, 67. 21 J. M. Dyke, A. Morris, G. Josland, M. P. Hastings and P. D. Francis, High Temp. Sci., 1986, 22,95. 22 W. J. van Der Meer, P. van Der Meulen, M. Volmer and C. A. De Lange, Chem. Phys., 1988,126,385. 23 K. P. Huber and G. Herzberg, Molecular Spectra and Molecular Structure: Constants of Diatomic Molecules, van Nostrand, New York, 1979. 24 P. Coppens, S. Smoes and J. Drowart, Trans. Faraday Soc., 1967, 63,240. 25 J. Kordis and K. A. Gingerkh, J. Chem. Phys., 1977,66,483. 26 W. C. Martin, R. Zalubas and L. Hagan, Atomic Energy Levels -The Rare Earth Elements, NBS, Washington, D.C., 1978. 27 R. J. Ackerman, E. G. Rauh and E. J. Thorn, J. Chem. Phys., 1976,65,1027. 28 E. Murad, US Air Force Report AFGL-TR-77-0235, 1977, Hanscom AFB, USA. 29 A. N. Nesmeyanov, Vapor Pressures of the Chemical Elements, ed. R. Gary, Elsevier, Amsterdam, 1963. 30 M. Kaufmann, J. Muenter and W. Klemperer, J. Chem. Phys., 1967,47,3365. 31 R. L. DeKock and W. Weltner, J. Phys. Chem., 1971,75,514. 32 S. B. Gabelnick, G. T. Reedy and M. G. Chasanov, J. Chem. Phys., 1974,60, 1167. Paper 0/02017F; Received 8th May, 1990

ISSN:0956-5000    

DOI:10.1039/FT9918700019

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. CHEM. SOC. FARADAY TRANS., 1991, 87(1), 31-36 Approximating Relative Rotational Energy Transfer in Molecular Collisions F. A. Gianturcot and J. Peter Toennies Max-Planck-lnstitut fur Stromungs forschung , Bunsenstr . 10,03400Gottingen, Germany Maurizio Bernardi Department of Chemistry, The University of Rome, Citta Universitaria ,00185,Rome, Italy The rotational energy transfer (RET) processes taking place in an atom-diatomic gaseous mixture are analysed in detail by carrying out collision dynamical calculations of the relevant state-to-state cross-sections. Approx- imate schemes for the dynamics are tested along the series of He, Ne and Ar rare gases in collisions with N, molecules, using the IOS and CS approximate models. The corresponding average relative energy transfer (ARET) is found to be a slow function of collision energy and can be used as an efficiency index for the systems at hand.The IOS approximation, however, grossly overestimates this index especially for the more anisotropic interactions. Other dynamical models, like the hard-ellipsoid impulsive model, are also found to have the same shortcomings as the IOSA scheme. It is therefore concluded that approximate treatments of relative rotational energy-transfer processes are capable of only near quantitative predictions for systems that show very low efficiency for such processes, and fail markedly when larger amounts of energy can be transferred during collisions. The present conclusions will be of use in analysing scattering experiments. 1.Introduction Collision-induced energy-transfer processes in molecular systems are of fundamental importance to several research areas in chemical physics and molecular physics.1-4 For example, the interpretation of sound absorption effects due to alternate compressions and expansions in the gaseous mixture’ requires an evaluation of a characteristic rotational relaxation time, z~~~,simply related to one special type of gen- eralised cross-section :6 where N is the number density, uAB the relative velocity of the partners, and the rotational relaxation cross-section 0(0001/A)AB, further discussed in ref. 7, is related to the amount of relative rotational energy transfer in collisions between the partners A and B. In this study we have examined a specific series of inert gases interacting with N, for which the potential-energy surface (PES) involved has been obtained earlier from an ab initio model approach.* This PES has been shown to predict successfully several transport properties and dynamical observable~.~~~~It should therefore provide a fairly realistic description of the regions relevant to RET collisional pro- cesses.Different dynamical approximations have been con- sidered and used for the production of degeneracy-averaged, partial-integral cross-sections and for the construction of pos- sible indicators of the relative efficiency of RET collisional processes. Such global indices can be used to indicate how well a specific PES is likely to direct energy into rotational motion and how such specific energy deposition effects vary with kinematical conditions dictated by the bath properties and by the partners masses.It is the aim of the present analysis to test the quality of the dynamical approximations for assessing the reliability of such indices. 2. Dynamical Models A broad range of prescriptions is currently available for com- puting cross-sections, elastic and inelastic, total and partial, 7 Permanent address: Dept. of Chemistry, University of Rome, Rome 00185, Italy. integral and differential, by using quantum, semiclassical or classical Various approximations can thus be used to reduce the full dimensionality of the coupled equa- tions (CC) which describe the nuclear relative motion, and their validity for various systems has been extensively tested in the last few years.”.’2 What has become by now the most common procedure for reducing the dimensionality of the CC equations relies on two distinct physical approximations. (i) The scattering is considered to be dominated by the repulsive part of the inter- molecular forces, with the attractive region playing a lesser role, so that the trajectories are distorted mainly by the repul- sive potential with little effects coming from coupling of dif- ferent impact parameters.This approximation is usually called the centrifugal sudden approximation (CSA) of Kouri and McGuire.13 (ii) The amount of transferred energy is small in comparison with the total energy available to the colliding partners.Thus, the channel wavevectors are assumed not to change in going from one final state to another. This simplifi- cation is therefore termed the energy sudden (ES) approx- imation.14 The combined use of approximations (i) and (ii) leads to the so-called ISOA decoupling scheme.’ ’ The great attraction of such a scheme is its capability of predicting the full ensemble of state-to-state cross-sections from a simple ‘golden rule’ formula” that expresses them as a weighted sum of cross-sections from the initial ground state : where ji, j, are the initial and final rotational quantum numbers and the j, index takes on all values which are allowed by the triangular relation required for each of the Clebsh-Gordan coefficients that constitute the weighting factors, these being zero for odd (ji+j, +j,).For weakly interacting systems, such as He colliding with diatomic targets, the CSA is very good down to collision energies of the order of the well depth, which are typically 20-30 K. Thus, most of the error from computing IOSA inelastic cross-sections is expected to come from the ES assumption, i.e. from disregarding the change in relative velocity resulting from the inelastic collision. J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 Table 1 Vll P He Ne 7.43 7.37Rm Q 6.73 6.68 BTT potential parameters for rare-gas-N, interactions (all values in a.u.) VL AP = PI1 -P, Ar He Ne Ar He Ne Ar 8.15 6.43 6.38 6.89 1.o 0.99 1.26 7.36 5.73 5.68 6.14 1.o 1.o 1.22 x 10-~ 0.74 1.78 3.00 0.98 2.28 5.10 0.24 0.50 2.1 6.50 6.19 6.95 5.54 5.25 5.80 0.90 0.94 1.09&la Rlp2; 6.22 5.80 6.56 5.29 4.87 5.53 0.93 0.93 1.03 6.06 5.61 6.41 5.15 4.68 5.41 0.91 0.93 1 .00 Rto3 Rtpl = turning point at Ecoll= 21.48 meV.R,,, = turning point at Ecoll= 74.24 meV. Rtpj = turning point at Ecol,= 124.08meV. Given the capabilities of the approximate IOSA approach, one interesting quantity that could help us to analyse the relation between anisotropic interactions and RET processes is the average relative energy transfer (ARET) defined as follows: @& = (1AE, a,,,(E) E if which is referred to a particular initial state lj,) of the target molecule and where AEif is the energy difference between the initial and final states. From an eqn.(3) for each lj,) one can therefore obtain a relative energy transfer parameter, (AEIE), that provides information on the efficiency of the RET process for a particular system and at a specific collision energy E. Note that the state-to-state cross-sections of eqn. (3) provide relative probabilities weighting differently the various AEif contributing to the average value. In other words, the cross-sections provide the dynamical factors while the AEif represent the structural factors of the RET process. 3. Anisotropic Surface The reason for choosing He, Ne and Ar colliding with N, molecules for this study is related to the established quality of the anisotropic potentials for these systems, which have been tested by us and found to be quite realistic" in describing several different transport properties of their gaseous mix- tures.'o,ll They have been predicted using the Tang-Toennies potential model16 which dampens the long-range ab initio dispersion terms individually using a universal damping function and adds to it a simple Born-Mayer repul- sive form.The resulting potentials will be called BTT potentials' and their more relevant features are listed in Table 1, which also contains the values of the classical turning points at three different collision energies and for the two geometries which give rise to the definition of the ellip- soid axes: the corresponding difference values are also listed in the last three columns of the table.It is interesting to note that the general smallness of the deformation is confirmed by these values, as they correspond to ca. 12-13% of the average minimum positions R, for each system. The deformation also varies relatively little with the relative collision energy E, thus suggesting fairly steep walls for all the three repulsive regions, and that the deformation does not increase significantly in going from He to Ne. This somewhat counterintuitive result is confirmed by the R, values, which also do not increase in going from He to Ne, and by the very similar behaviour found for the three rare gases along the series of interactions with H, .17 4. Average Energy Transfers Decoupled Quantum Results One interesting question concerning the quantities defined in section 2, especially that given by eqn.(2) and (3), is to verify the general accuracy of the approximations contained in the computing schemes. We already mentioned that for weakly interacting van der Waals (vdW) systems, where the average well depth is often much smaller than the collision energies, the CS approximation is very realistic especially for compar- ing partial-integral cross-sections. Thus, one first check is provided by a direct comparison between IOSA and CS cal-culations for the same quantity. This is shown in Fig. l for Ar-N, ,the most anisotropic system we have examined here. Two different collision energies are shown and the relative -aCSA)/aaA,error, (aIOSA is plotted as a function of the amount of transferred rotational energy.The partial inelastic cross-sections were computed starting from the ji = 2 level. All open channels and two closed channels were needed for convergence of the cross-sections to within ca. 2%. In the case of the IOSA results, all the cross-sections are by defini- tion associated to 'open' channels and therefore the summa- tion in eqn. (3) included all contributions that were necessary to make the partial cross-sections converge to <1%. One clearly sees that the IOSA calculations overestimate the rota- tional inelasticity and always produce cross-sections that are too large for the present systems. The consequences of the weighted sum, employed to gener- ate the ARET index of eqn.(3), are shown in Fig. 2 for the same system for ji = 0. The IOSA results are significantly larger than the CSA data at all collision energies, this effect being the greatest at collision energies below 10 meV where the IOSA values are a factor ca. 15 larger. This general trend is in keeping with previous comparisons with calculations on similar systems." On the other hand, as the collision energy increases, both sets of calculations are fairly parallel and tend to limiting values nearly independent of the collision energy. This fact underlines the dominance of the purely repulsive features of the interaction at higher energies. It also suggests h // b OF ' ' 9 '' ' 50 ' ' I 100 I I ' '150 I " AE,,,lmeV Fig. 1 Percentage deviations between partial inelastic IOSA and CSA cross-sections, as a function of rotational energy transfer.All cross-sections originate from the j = 2 rotor state and were calcu- lated for the BTT potential function of ref. 8 for two different colli- sion energies. 0,E = 0.146 eV; a,E = 0.204 eV J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 3.00 Ar-N2 (ji=O) 2.50 BTT potential 2 .oc \%2 1.50 klL \IOSAQv 1.oo 0.50 0 Fig. 2 Computed average relative energy transfers (ARET) as a function of collision energy (in meV) for the Ar-N, system. 0,IOSA calculations; 0,CSA calculations. The PES is the BTT function of ref. 18 that in the limit of high energies (AEIE) might serve as an indicator of RET efficiency in that particular system.If one now turns to the weakest of the interaction poten- tials discussed here, i.e. the He-N, system, one is able to see that the IOSA model fares much better in yielding inelastic cross-sections as shown in Fig. 3 and 4. Fig. 3 shows the partial-integral cross-sections, computed within the IOSA and CSA scheme, compared in terms of relative percentage error as a function of the amount of rotational energy trans- ferred. The corresponding j, values are also shown in the upper abscissae. In these calculations the HFDl potential function of ref. 19 was used. Since this potential shows similar trends as the BTT model but somewhat different magnitudes we also expect similar trends in cross-section error.The main features of both potential functions are listed in Table 2. The IOSA results consistently overestimate RET efficiency and do so more markedly at lower collision energy and for more anisotropic potentials. Thus, the average energy trans- fers evaluated by IOSA calculations are all likely to be grossly in error unless the system under consideration happens to exhibit a low efficiency for RET processes Fig. 3 Same as for Fig. 1, but for the He-N, system. The employed PES was that of ref. 19. Ei (in meV) shown on figure 33 He-N? (ji=O) \ HFDl potential \ \ ob \\ 0 0 50 100 EtotlmeV Fig. 4 Same as for Fig. 2 for the He-N, system and computed oia the potential function of ref. 19. H,IOSA; 0,CSA through either an unfavourable mass ratio or a weak angular potential aniso tropy .On the other hand, since the IOSA and CSA values of the ARET index from eqn. (3) appear to be scaled by factors largely independent of collision energy, at least in the range examined in Fig. 2 and 4, (AEIE) from the simpler dynami- cal approximation of the IOSA scheme can still be used as a qualitative index of ARET efficiency. The larger its value, in fact, the more likely the system at hand is to require more sophisticated calculations for estimating reliably its relax- ation properties. An extensive comparison of different methods for obtaining the latter quantities has been present- ed elsewhere." An example of the qualitative value of such IOSA calcu- lations is shown by the results presented in Fig.5. The rela- tive energy-transfer indices are plotted as a function of total collision energy in meV for a series of projectiles from He to Ar. The interplay of mass factors and anisotropy features of the interactions in producing the three different curves pro- vides some interesting insight :(i) the average relative energy transfer (ARET) changes slowly with collision energy, as dis- cussed before, and therefore the limiting values of the curves could be used to characterize the ARET efficiency in each system. There is a low efficiency for He projectiles (ARET value ca. 0.10) and a markedly larger efficiency for Ar-N, systems (ARET value ca.0.60);(ii) note that the above values change in going from Ne to Ar by a factor of ca.1.8, which is Table 2 Comparison of some characteristic potential parameters for He-N, as obtained from two different functional forms (all values in a.u.) vo V, v4 HFD1" BTTb HFD1" BTTb HFD1" BTT* Rm 6.90 7.00 8.05 7.80 -8.80 a 6.22 6.10 7.80 7.19 -8.24 E x lop4 0.79 0.80 0.04 0.13 -0.003 turning points Rt,, 6.09 6.50 4.84 5.54 1.25 0.96 Rt,,: 5.57 6.22 4.40 5.29 1.17 0.93 RtP3 5.35 6.06 4.20 5.15 1.15 0.91 a From ref. 19. From ref. 8. RIP1= turning point at Ecoll= 21.48 meV; Rlp2 = turning point at Ecoll= 74.24 meV; RIp3= turning point at Ecoll= 124.08meV. 34 120 J " ~ 1 " " I " RG -N2(ji =O) 100- BTT potential - '\ OU-Lt 8 1 0 10 50 100 EtotlmeV Fig. 5 Average relative energy transfer [eqn.(3)] computed via the BTT potential functions of Table 1 for He, Ne and Ar colliding with N, .The ARET indices are shown as a function of collision energy (in meV). e, HeAr; A,Ne; ., larger than the simple increase due to reduced mass changes (1.4). This suggets that the increased anisotropy of the latter system with respect to the former (see Table 1) plays a signifi- cant role and makes 10s results correspondingly less reliable; (iii) the visible increase of the ARET index in going from He to Ne, with similar anisotropic interactions, as seen from the features of Table 1. Thus, one expects the changes to be due mainly to mass factor changes with little effect coming from dynamical factors. The reduced mass ratio Ne-N,/He-N, is ca.3.4 and the corresponding ratio of the IOSA limiting values from Fig. 5 is CQ. 3.3. Hard-ellipsoid Model Simpler, approximate procedures which could yield accept- able ARET values for a given system have been used by experimentalists and it is interesting to examine here the clas- sical expressions obtained within the hard-ellipsoid model of anisotropic interaction and the impulsive picture of the RET dynamic^.^ 1-23 Such quantities have often been obtained to make comparisons with experiments.22 One important observable determined by scattering experi- ments is the dimensionless relative recoil velocity gi defined as: gg(jf)= (1 -AEif/E)"2 (4) for a given relative energy transfer, AE,,/E between rotor states ji and j,.Within the above classical model one finds that the angular distribution of g;(jf) is given for ji = 0 by g*(jf, 0)= [(l -Q2 sin2 0)l/'+ Q cos 0](l + Q)-' (5) where 0 is the c.m. scattering angle. Although eqn. (5) is strictly correct only for ji = 0 it is only weakly dependent on ji so long as the impulsive approximation holds. The dimen- J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 sionless parameter Q is defined as: 2PQ = Be -(AR)'h where AR in this model is simply the difference between the location of the two turning points, in the C,, and C,, orien-tations, at the given collision energy. If one now expresses the constant Be via the moment of inertia of the isolated mol- ecule, I, then eqn. (6) can be rewritten as: (7) where all quantities are in atomic units.The above parameter is the ratio between some 'dynamical' moment of inertia for the interacting system, p(AR)', and the 'static' moment of inertia of the isolated molecule, I, which simply provides a structure factor inversely related to the rotational spacings in the target. It is interesting to note that the Q parameters for the systems at hand tend to a limiting value, as the collision energy increases, which is only slowly dependent on the colli- sion energy itself. In particular, the asymptotic values of Q are very close to the corresponding relative energy-transfer limit values given by the IOSA calculations discussed before and presented in Fig. 5. This surprising result can be gleaned from the quantities shown in Table 3 for the present systems and suggest that, in spite of its ultrasimple evaluation, the parameter Q is roughly related to the ARET index discussed before.By making use of eqn. (4) and (5) one quickly recovers an expression for the relative energy-transfer value as a function of the scattering angle 0and the parameter Q : (%)(@) = 2Q x [l + Q sin2 0-cos o(1 -Q2 sin2 0)1'2](8) The above quantity, within the hard-ellipsoid model for the interaction, classically corresponds to the relative RET when going from ji to j, in the target states. It has a maximum at a particular angle OR, for which the relative recoil velocity exhibits a rainbow maximum. The corresponding relative average energy transfer can therefore be obtained by simple integration : The quantity given by eqn.(9) does not depend any more on a particular ji and j, state since the above integrated value is now taken to represent the overall energy transfer coming Table 3 Computed Q parameters [eqn. (6)] and computed IOSA average relative energy transfer values [eqn. (3)] for He, Ne and Ar inter- acting with N, 124.08 0.099 0.104 0.34 0.33 0.55 0.58 108.57 0.099 0.106 0.34 0.33 0.56 0.60 74.24 0.101 0.117 0.34 0.35 0.58 0.67 42.01 0.104 0.135 0.35 0.39 0.62 0.79 21.48 0.108 0.165 0.35 0.45 0.67 1.05 10.98 0.111 0.232 0.36 0.55 0.7 1 1.71 6.2 1 0.1 14 0.346 0.37 0.74 0.75 2.76 J. CHEM. SOC. FARADAY TRANS., 1991, VOL.87 Table 4 Computed integrated relative energy transfer (AEIE), and maximum relative energy transfer, (AE/E)max,as given by eqn. (9) and (11) for He, Ne and Ar projectiles interacting with N, 124.08 0.35 0.33 108.57 0.35 0.33 74.24 0.36 0.33 42.02 0.36 0.34 2 1.48 0.38 0.35 10.98 0.39 0.36 6.2 1 0.39 0.37 from all scattering angles during the impulsive process and therefore from all possible ji and j, values. It is obviously related to the Q parameter for each system under consider- ation and can be simplified according to the values of Q. Thus when Q < 1 then however, for Q -1 one obtains instead and, finally, for very large Q values (Q %-l)(AE/E) approaches a constant independent of Q, A further index of the efficiency of the energy-transfer process can be obtained by taking the value of the relative RET index at the angle at which it is largest.Thus, one can write: = 4Q(1 + Q)’ (T)max with the limiting values: z l.O(Q x 1) From the above derivations it is seen that the simplified relation for the (AE/E)maxgiven by eqn. (12a) in the limit of small Q values” coincides with the integrated expression of eqn. (lOa),the latter being valid also in the limit of small Qs. Thus, the Q parameter turns out to be directly related to the average relative energy transfer index discussed before. The fact that its actual values as a function of collision energy are close to those obtained via the IOSA calculations for eqn. (3) confirms more quantitatively this relation. It also tells us, however, that Q must also overestimate the ARET values as is done by the IOSA scheme and that, consequently, its relia- bility as an efficiency index is greater the smaller its value becomes.The situation of Q 41 corresponds in fact to either small reduced masses or to small potential deformations from the spherical shape features for which the IOSA calculations were shown before to be more reliable and closer to the more sophisticated CS results. The behaviour of the integrated quantity of eqn. (9) and of the maximum value of eqn. (11) as a function of collision 0.94 0.76 1.26 0.92 0.94 0.76 1.26 0.92 0.94 0.76 1.29 0.93 0.94 0.76 1.33 0.94 0.95 0.77 1.39 0.96 0.97 0.78 1.44 0.97 0.98 0.79 1.47 0.98 energy, is shown in Table 4 for all the systems discussed in the present study.5. Conclusions The anisotropic interactions examined above, and the specific systems for which we have computed average relative energy transfer parameters (ARET), have allowed us to look more closely at possible indices of efficiency that could be obtained by approximate dynamical calculations of the ARET values. As shown by the present results, the energy sudden approximation invariably produces upper bounds to the more correct coupled states calculated inelastic cross-sections. As a result the IOSA average values overestimate the energy transfers. On the other hand, we have shown that the sudden approximation implicit within the IOSA scheme is most sensitive to the short-range anisotropy of the inter- action, while including the full anisotropy only indirectly, through the angle-averaging procedure. Thus, when the RET efficiency indices are examined along a series of atom-molecule systems with similar characteristics, their IOSA esti-mates are increasingly more in error the more the system anisotropy and reduced mass increase along the series.The alternative index Q was shown to be related to a simple classical model for calculating average energy transfers and was given as a ratio between a ‘dynamical’ factor and a ‘structure’ factor in eqn. (6) of section 4. The Q parameter obtained there was; however found to be nearly coincident with the previously examined IOSA quantities of the present systems.Thus, attempts at using it in comparison with experiments22 must have the same care as that needed to evaluate the general reliability of IOSA cross-sections. This result was never explicitly shown in previous dis-cussions.20-2 In conclusion, our study of approximate methods for evaluating average energy transfers in neutral, weakly inter- acting systems has shown the following features: (i) any of the approximate treatments of the dynamics considered here tends to overestimate inelastic processes and therefore pro- duces ARET indices which are usually too large; (ii) the energy dependence of such approximate quantities is usually correct and tends to change very little with the collision energy E.Thus some fixed limiting value could be chosen to represent a given system and to qualitatively tell us how effi- cient RET processes are in that system; (iii) for systems where the efficiency is low, either because of the mass factor or of anisotropy features of the interaction, the approximate model discussed here are usually good and provides quantitative estimates of the RET efficiency in thermal and near-thermal collisions. We are grateful to Dr. M. Venanzi for several useful dis- cussions on the IOSA cross-sections and to Dr. A. Palma 36 J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 for helping us with the CS calculations. We thank Dr. St. Schlemmer for critically reading the manuscript. F.A.G.thanks the van Humboldt-Stiftung for financial support during the summer of 1989, when this work was completed. 10 11 12 13 F. A. Gianturco and M. Venanzi, J. Chem. Phys., 1989,91,2525. F. A. Gianturco and A. Plama, J. Phys. B, 1985,18, L519. F. A. Gianturco, M. Venanzi and M. Faubel, J. Chem. Phys., 1989,90,2639. P. McGuire and D. J. Kouri, J. Chem. Phys., 1974,60,2488. 14 R. T. Pack, J. Chem. Phys., 1978,68,1585. References 15 R. Goldflam, S. Green and D. J. Kouri, J. Chem. Phys., 1977,67, 4149. 1 2 3 4 5 6 7 See, e.g. F. A. Gianturco, in The transfer of Molecular Energy by Collisions, Springer-Verlag, Berlin, 1979. G. D. Billing, Introduction to the Theory of Inelastic Collisions in Chemical Kinetics, University Press, Copenhagen, 1978. M. Faubel, Adu. At. Mol. Phys., 1983, 19, 345. F. A. Gianturco and M. Bernardi, J. Phys. Chem., 1989, in the press. P. G. Kistemaker and A. E. de Vries, Chem. Phys., 1975,7,371. H. van Houten and B. I. M. ten Boch, Physica A, 1984,128,371. L. Monchick, A. N. G. Pereira and E. A. Mason, J. Chem. Phys., 1965,42, 3241. 16 17 18 19 20 21 22 23 K. T. Tang and J. P. Toennies, J. Chem. Phys., 1984,80,3726. K. T. Tang and J. P. Toennies, J. Chem. Phys., 1981,74, 1148. S. Chapman and S. Green, Chem. Phys. Lett., 1984,113,436. R. R. Fuchs, F. R. W. McCourt and A. J. Thakkar, J. Phys. Chem., 1984,88,2036. F. A. Gianturco, M. Venanzi and A. S. Dickinson, J. Chem. Phys., 1990,93, in the press. P. L. Jones, U. Hefter, A. Matheus, J. Witt, K. Bergmann, W. Miiller, W. Meyer and R. Schinke, Phys. Rev. A, 1982,26,1283. U. Buck, Comm. At. Mol. Phys., 1986,17, 143. D. Beck, U. Ross and W. Schepper, 2.Phys. A, 1979,293,107. 8 M. S. Bowers, K. T. Tang and J. P. Toennies, J. Chem. Phys., 9 1988,88,5465. F. A. Gianturco, M. Venanzi and A. S. Dickinson, Mol. Phys., Paper 0/02452J; Received 1st June, 1990 1988,65, 563.

ISSN:0956-5000    

DOI:10.1039/FT9918700031

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. CHEM. SOC. FARADAY TRANS., 1991, 87(1), 3744 HBO Osculating Complex in the B(=P) + OH(*lI) Reaction M. Alberti, A. Sole and A. Aguilar Departament de Quimica Fisica , Universitat de Barcelona, Marti' i Franquds I, 08028 Barcelona, Spain The dynamics of the B(2P) + OH(211) reaction have been investigated by means of the quasiclassical trajectory approach, using a reasonable potential-energy surface model derived from theoretical information at the MP3/6-31G**//HF/6-31G** level. The properties analysed can be interpreted on the basis of the main topological fea- tures of the potential-energy surface used. Reaction was found to occur via two possible mechanisms: (a) the boron atom inserts into the 0-H bond to form a collision complex which subsequently breaks apart, and (6) the boron atom abstracts the oxygen atom from the OH molecule, the former mechanism being much more reactive than the latter.At low collision energies, a collision complex is formed whose lifetime is less than the rotational period, but as the collision energy increases, a transition to the formation of a more long-lived complex occurs. 1. Introduction The compounds formed in the oxidation of atomic boron are very interesting from the viewpoint of their role in com-bustion processes.' Boron chemistry has been an important and fruitful research area for many years.2 Among the unstable compounds with boron-oxygen multi- ple bonds, HBO has been known experimentally since 1971.3 More recently, Kawashima et d4,'have suggested, from a study of the reaction of diborane with nitrogen monoxide, that HBO exists as an intermediate species in the process leading to boroxin.Theoretical studies at MND06*7 and ab initio7-" levels indicate that HBO and HOB structures belong at two minima on the doublet ground-state surface of the HBO system, the latter being metastable with respect to the former. Although there is no experimental evidence about HBO formation in the B(2P) + OH ('n)reaction, it is interesting to carry out a study of this reaction in order to ascertain both the possible influence of the minima in the collision mode and their effect on the attacking path of boron atom on the OH molecule, which can be inserted into the OH bond to form a complex, which then breaks apart to give BO + H after sampling the deep HBO potential well.Alternatively, the boron atom can attack the oxygen atom from the end of the OH molecule and, by sampling the deep HOB potential well, this leads to BO + H formation. We have recently performed a preliminar quasiclassical 3D trajectory study of the B(2P) + OH('n) system" by using a Sorbie-Murrell' ' analytical fitting of the potential-energy surface. Results indicate that this system is fairly reactive and that the reaction features observed are sensitive to the pre- sence of the potential wells. This work presents the results of a detailed study on the reaction dynamics of B(2P) + OH('H). Section 2 shows the main characteristics of the potential-energy surface and the computational procedure.Section 3 describes the results obtained, giving particular attention to the dependence of reaction cross-sections, angular distributions, complex life- times and energy disposal on initial collision energy, as well as their comparison with other trajectory calculations on potential-well surfaces. The most important conclusions are given in section 4. 2. Method 2.1 Potential-energy Surface To carry out the B(2P) + OH(211) trajectory calculations we have used a Sorbie-Murrell-like potential energy surface described elsewhere.7, '' These functions are particularly useful in dynamic calculations since they do satisfy asymp- totically the system dissociation limits, as well as the properties of the surface stationary points.This method has already been applied successfully to other triatomic systems.'2v'3 The analytical function was obtained by adjusting the geometry, dissociation energy and harmonic force field of the HBO minimum. In this way, standard ab initio calculations were carried out at HF/6-31G** level, and the valence-electron correlation was incorporated using the Mraller-Plesset treatment terminated at third order (MP3). l4 The analytical potential-energy function reproduces accu- rately the HBO linear minimum and almost as well the disso- ciation energy of the HOB minimum (error < 0.11%). The barrier found in the isomerization path is somewhat overesti- mated (ca. 1.9%),7*" but owing to the high exoergicity of the B0(2Z+)+ H(2S) channel it seems that this discrepancy will have little relevance.Fig. 1 shows potential-energy contours with the BO bond length fixed at the corresponding equilibrium distance of the HOB minimum. This figure gives an approximate representa- tion of the two minima and of the isomerization path lying between them. Fig. 2 shows potential-energy contours for the 6.0 - 4.0 - - 2.0 - - 0.0- - -2.0 - - -4.0 - - '--' I' -6.0 I I I I I I I I I I I 38 0.4 1.4 2.4 3.4 4.4 RBn/A Fig. 2 Contour map for the perpendicular approach of the B atom to the middle of the 0-H bond. The arrow indicates the lowest- energy curve (-12.0 eV with respect to dissociated atoms), the contour interval is 1.0eV perpendicular insertion of the B atom into the middle of the 0-H bond, while Fig.3 displays the potential-energy surface for B attack to the oxygen at the B-0-H angle given. From Fig. 2 and 3 we can see that the potential surface is repulsive for the near collinear approach of B atom to the OH molecule, and attractive for atomic boron insertion into the 0-H bond. Fig. 4 gives a schematic representation of the minimum-energy path for the B(’P) + OH(%) -+ BO(’C+) + H(2S)reactive process. 4.4-- ‘$3.4- a? 2.4-- 1.4 1.4-- - 0.44 I I I I 0 . 4 1 0.4 1.4 2.4 3.4 4.4 0.4 1.4 2.4 3.4 4.4 Ron/A RonIA 4.4 3.4 2 cy 2.4 1.4 J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 2.2 Computational Procedure for Quasiclassical Trajectories Dynamical calculations have been performed according to the quasiclassical 3D trajectory method’ ’vl as implemented in an improved versionI7 of Hopper’s program.18 The numeri- cal integration of Hamilton’s equations of motion has been carried out with the two Runge-Kutta algorithms: order- predictor (fourth order) and order-corrector (eleventh order).The integration step has been fixed at 2.5 x s. The accuracy of the numerical integration has been verified by checking the conservation of the total energy and angular momentum along every trajectory, and making back-integrations over random samplings of trajectories. The B(’P) + OH(211)reaction has been studied, with the OH molecule being in the most populated rovibrational level at 300 K (u = 0,J = 2), for a wide range of collision energies.The influence of the vibrational and rotational energies on the reactivity has been investigated by considering respec- tively the (v = 1, J = 2) and (u = 0, J = 0, u = 0, J = 6 and u = 0, J = 8) rovibrational levels at a collision energy value of 0.04 eV, which is very close to the averaged thermal energy at 300 K ((ER) = 0.0388 eVt). At fixed collision energy, the maximum impact parameter (bmax)for every initial state of the OH(211) molecule has been calculated by running sets of 100 trajectories with a fixed impact parameter (b). All trajectories have been calculated starting from an initial separation of 7 A between the B atom and the OH centre of mass, because at this distance the inter- action energy can be neglected with respect to both the rela- tive energy and the internal energy of the diatomic molecule.Owing to the strong reactivity of the B(’P) +OH(%) process, it has not been necessary to calculate a great number of trajectories in order to obtain results with good statistics. The error in the total reactive cross section for each one of the initial conditions is less than 3%. 3. Results and Discussion 3.1 Total Reactive Cross-section and Thermal Dependence of the Rate Constant The total reaction cross-section at each collision energy, ER , is given by: where N, is the number of reactive collisions at ER collision energy, and N is the number of trajectories calculated at the same ER value.Table 1 gives the calculated reaction cross section values 0.4‘0.41 (ar)at several collision energies for the most populated rovib- 0.4 1.4 2.4 3.4 4.4 0.4 1.4 2.4 3.4 4.4 rational level (u = 0, J = 2) of the OH molecule at 300 K. ROH/A From the table we can observe how or(ER)decreases as the Fig. 3 Contour map for the approach of B to the oxygen of OH. relative energy increases. This behaviour is in good agree- B-0-H angler: (a) 40, (b) 70, (c) 100 and (6)180. The contour ment with that expected from the topology of the potential- interval is 1.0 eV energy surface, which does not present any barrier above the asymptotic reactant valleys along the minimum-energy reac- -tion path. ”-” The reaction cross-section values obtained 0 E have been fitted by the analytical function: a,(E,) = CEid (2)BO + H 7-c / linear least-squares fit of In a, vs.In ER, [C =/ d = 0.131. The thermal rate constant, k,(T),for the B(2P)+ OH(’ll) reaction can be obtained in the usual way once the excitation Fig. 4 Schematic representation of the MP3/6-31G**//HF/6-31G** energy profile for the B(’P) + OH(211)reaction t 1eV z 1.602 x J. where the constants C and d have been determined by a 35.52 A2 (eV)d, J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 Table 1 Reaction cross-sections’ and maximum impact parameters at several collision energies for the reaction B +OH(u =0, J =2) +BO +H 0.005 728 505 5.70 70.8 f1.7 0.025 496 368 5.00 58.3 f1.5 0.03 663 49 1 4.90 55.9 f1.3 0.04 749 566 4.70 52.4 1.1 0.06 667 484 4.70 50.4 +_1.2 0.08 900 667 4.60 49.3 f1.0 0.1 604 459 4.50 48.3 k 1.1 0.2 636 454 4.45 44.4 f1.1 0.3 844 600 4.30 41.3 f0.9 ~ ‘The error limits on the reaction cross-section are given by the sta- tistical error o,[(N -N,)/NN,]”2.N is the total number of calcu- lated trajectories and N, the total number of reactive trajectories. function has been calculated : where the average value is taken over the Maxwell-Boltzmann distribution of relative velocities at temperatue, T.Thus, for the cross-section function given by eqn. (2), k,(T) becomes : k,(T)=C(8/Xp)1’2(kBT)1’2-dI‘(2-d) (4) where r is the well known gamma function. Therefore, the dependence of the rate constant on temperature is given by: k, =C‘T0.37 (5) where the absence of any activation energy is evident.3.2 The Effect of Vibrational and Rotational Excitation on the Reaction Cross-section From the results shown in Table 2 we can see that the vibra- tional excitation of the reactants (u =0 -,u =1) does leave nearly unchanged the reaction cross-section value at 0.04 eV of collision energy but, at the same time, it can be observed that reaction probability (given by NJN) decreases when b,,, increases. Thus, analysis of the set of reactive trajectories seems to indicate that the behaviour observed, in respect to the effect of the reactants’ vibrational energy on the reaction cross section, may be due to two compensating effects: when changing from u =0 to u =1, b,,, increases form 4.7 to 5.0 A, in accord with the enlargement of the classical vibration amplitude but, simultaneously, reaction probability decreases, probably owing to the fact that the majority of reactive tra- jectories explore the zone corresponding to the HBO minimum, which can be classified as a late well on the poten- tial surface.It is well known from collinear ~alculations’~ that in this kind of potential-energy surface vibrational exci- tation is not very effective in promoting the reaction. In the Table 2 Reaction cross-sections and maximum impact parameters at ER =0.04 eV for the reaction B +OH(u, J) +BO +H (0, J) (070) (0, 2) (09 6) (0, 8) (1, 2) N 647 749 634 627 599 Nr 444 566 489 484 404 bmaxlA 5.2 4.7 4.8 4.9 5.0 o,/A2 58.3 1.5 52.4 1.1 55.8 f1.2 58.2 f1.3 53.0 f1.5 ‘The error limits on the reaction cross-section are given by the sta- tistical error ar[(N -N,)/NN,]’/2.N is the total number of calcu-lated trajectories and N, the total number of reactive trajectories. 39 present case, both effects lead to a near invariance of 0,from the reactants’ vibrational excitation. Table 2 also shows that, for the smaller J values calculated, 0, decreases as rotational energy increases, but from a certain value of J, 0, then increases quickly with increasing rotation- al energy and the dependence of or on J shown in Fig. 5, seems to suggest that a,(J) presents a minimum G,,,~,,(J)for J x 3. Similar rotational effects have been observed in many other systems.The initial decline in o,(J) can be explained on the basis of the diminution in the time that the B +OH system spends in the preferred orientation along the entrance channel collision. Thus, the most favoured orientation is disrupted by the rota- tional motion of the diatomic molecule. At low rotational excitation, orientational effects play an important role because of the entrance valley anisotropy.22 A large rotation- al energy makes all orientations accessible with equal prob- ability, so we can observe that a,(.!) increases after the initial decline. 3.3 Insertion vs. Abstraction An approximate analysis of the reaction mode has been carried out by monitoring the evolution of the three inter- nuclear distances and the HOB angle in a sample of reactive trajectories at several collision energies.This analysis makes evident that the majority of reactive trajectories are of a non- direct type (ca. 85%). Table 3 shows the average elapsed time for the reactive trajectories, (z), the average impact param- eter (b) and the overall rotational period of the complex HBO, zR.19 It can be seen that the (z>/zR ratio increases with collision energy. This seems to indicate that the time during 59.0 I 58.0 57.0 56.O 55.0 b‘ 54.0 53.0 52.0 51.0 50.0 0.0 2 .o 4 .O 6.0 8.0 J Fig. 5 Reaction cross-section as a function of J. E, =0.04 eV, u =0. The trajectory results (m) are connected by an interpolating solid line Table 3 Average values (b), (T), T~ and (T)/TRa at several collision energies for the reaction B +OH(u =0, J =2) +BO +H 0.005 3.6277 1.3543 0.8608 1.5733 0.025 3.1617 0.7802 0.4417 1.7663 0.03 3.1506 0.7196 0.4046 1,7785 0.04 3.0906 0.6629 0.3572 1.8558 0.06 2.9773 0.6044 0.3028 1.9960 0.08 3.1203 0.5663 0.2502 2.2634 0.1 3.068 1 0.5333 0.2276 2.343 1 0.2 2.7768 0.4672 0.1778 2.6277 0.3 2.6968 0.4389 0.1495 2.9358 See text.J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 which the three atoms are close enough for a collision complex to form increases with the collision energy. From the internuclear distance analysis it was possible to distinguish between insertion and non-insertion (abstraction in the case of reactive trajectories) of the boron atom into the OH bond, which was achieved by monitoring the minimum energy attained along the trajectory (Vmin)and by taking into account the geometry of the supermolecule at this point.A trajectory was considered to have followed an insertion path if the value of the minimum potential energy attained was less than that corresponding to the HOB minimum, and if the geometry of the supermolecule at this point showed that R,, > R,, , R,, . Results for the insertion/abstraction classi- fication at several collision energies are given in Table 4. It can be seen that the majority of reactive events correspond to the insertion mode, and that the N,(ins)/N,(abs) ratio increases with the collision energy.Individual trajectory analysis revealed that almost all trajectories that explored the potential surface near the HBO minimum were reactive, while of those exploring the surface near the HOB minimum approximately 35% become reactive at the lower collision energies, and 22% at the higher ones. The temporal evolution of the three interatomic distances and the BOH angle were examined using a sample of about one hundred reactive trajectories at ER= 0.04, 0.06, 0.08, 0.1, 0.2 and 0.3 eV for B(2P)+ OH(211, u = 0, J = 2). All the reactive trajectories analysed show a non-direct col- lision mode, insertion processes being much more complex than the abstraction ones. Moreover, a study of the collision time revealed that the insertion process lasts longer than the abstraction one, this difference becoming more pronounced at the higher values of collision energy. Some typical reactive trajectories are depicted in Fig.6. 3.4 Angular Distribution in Products The B0(2C+) angular distribution from the B(2P) + OH(211) reaction, in the centre of mass framework, has been analysed on the basis of the polar reactive cross-section (PRCS),4,(&, u, J, @), defined as:16 where ANr is the number of reactive trajectories scattered in the angle between 0and 0+ A@ for the initial condition (ER,u, J) and N is the total number of trajectories calculated for the same condition. Fig. 7 displays the total (insertion plus abstraction) PRCSs for all relative collision energies studied, the OH molecule being in the rovibrational level (u = 0, J = 2).At the lower Table 4 Insertion us. abstraction" for the reaction B + OH (u = 0, J = 2) -+ BO + H at several collision energies EJeV N 0.005 728 505 111 38 469 467 0.025 496 368 59 20 350 348 0.03 663 491 85 29 474 472 0.04 749 566 93 32 536 534 0.06 667 484 86 26 459 458 0.08 900 667 115 36 633 631 0.1 604 459 75 22 439 437 0.2 636 454 83 20 435 434 0.3 844 600 113 25 577 575 " Ninsis the total number of insertions, Nabsis the total number of trajectories that explore the potential-energy surface near the HOB minimum and Nr, and Nr,ins are, respectively, the number of reac-tive abstractions and insertions.collision energies, the backward peak is more pronounced than the forward one but, as collision energy increases, the distribution becomes more nearly symmetrical. Analysis of the angular distribution shows a predominance of backward scattering for the abstraction process, while the forward/ backward ratio for the insertion mechanism takes similar values to the global one, the PRCSs being very similar to those shown in Fig. 7. In Table 5 are given values of the forward/backward ratio for the overall process as well as for the insertion and abstraction mechanisms. It is well known that, if a long-lived complex is formed as the reactants approach, the rotation of the complex can destroy the memory of the original direction of the velocity vector, and the products will then be scattered symmetrically around 0= 90°.19-23The PRCSs found are not symmetrical except at the higher collision energy values.This could indi- cate that a transition to formation of a more long-lived complex occurs at the higher energy collision values investi- gated. In order to evaluate collision complex lifetime for each set of initial conditions, we have subtracted the time during which the three atoms are far enough apart that no super- molecule would be formed, from the total time spent in the reactive trajectories. In this way, we have found that distribu- tion time for the insertion process can be written as: P(t)= exp(-t/z) (7) where z is the lifetime of collision complex.Such a depen- dence is the same as would be expected when applying the osculating complex model (OCM),'9,23-2S in which the angular distribution I(@) is factorised into a symmetric forward-backward scattering term and an osculating one S(@, O,,,), showing the deviation from bimodal l/sin 0 dependence of the differential reactive polar cross-section, which is defined as : In the OCM, the osculating term is given by: exp(-O/@,,,)+ exp[ -(2n -@)/@,,,I (9)S(@, @,,,) = 1 + exp(-27r/OoSc) where : zR being the overall rotational period of the collision complex. Table 6 gives both collision complex lifetime and its rotational period as a function of the collision energy for the most populated rovibrational level of the OH molecule at 300 K.According with the values of O,,, obtained, shown in Table 6, the PRCS becomes more and more isotropic as E, increases due to the lessened influence of the osculating term [eqn. (9)] in the angular distribution. 3.5 Energy Distribution in Products The analysis of the energy distribution in the B0(2C+)mol-ecule generated by the B(2P) + OH(211) reaction, has been Table 5 Different forwardbackward ratios (f/b) for the reaction B + OH(u = 0, J = 2) -+ BO + H at several collision energies E,/eV 0.005 0.025 0.03 0.04 0.06 0.08 0.1 0.2 0.3 f/b 0.68 0.69 0.55 0.62 0.67 0.63 0.77 0.96 0.95 f/babs 0.44 0.18 0.46 0.32 0.30 0.29 0.22 0.36 0.35 f/bin, 0.70 0.73 0.56 0.65 0.70 0.66 0.81 1.09 1.03 f/b denotes the overall process, f/babs the abstraction mechanism and f/bins the insertion reaction mode.J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 V.7 I 0.0 4.4 8.7 0.0 2.9 5.8 tirne/10-l3s ti~ne/lO-’~s 8.1 0..=. 90.0 g a 4.4 90.0 :O< Cm 0.0 0.6 0.0 0.0 3.5 7.1 0.0 6.5 13.1 time/10-13 s tirne/10-13 s Fig. 6 Typical reactive trajectories of the W2P)+ OH(’lI) reaction (u = 0, J = 2). (a)ER= 0.04 eV; u’ = 6, J’= 19 (insertion).(b)ER = 0.06 eV; u’ = 11, J’ = 17 (insertion).(c) E, = 0.1 eV; u’ = 6, J’ = 53 (abstraction).(d)E, = 0.2 eV; u’ = 1, J’ = 27 (insertion). The continous line indicates the temporal evolution of the three internuclear distances and the dotted line shows the B-0-H angle evolution made considering the distribution of the accessible energy (Eacc)among the different energetic forms in products.For each initial condition (ER,u, J), E,,, is defined as where Eintrepresents the reactants’ internal energy and A&is the difference between the electronic dissociation energies of products and reactants. A summary of the results obtained for the different initial conditions and for the insertion pro- Table 6 Lifetime (7) and rotational period (z,“) of the collision complex ER/eV @,,/rad 0.04 3.17 0.06 3.75 0.08 4.55 0.1 5.37 0.2 7.05 0.3 10.40 7dPS 0.1803 0.3572 0.1806 0.3028 0.181 1 0.2502 0.1945 0.2276 0.1994 0.1778 0.2475 0.1495 cesses, which are the most important in the energy range studied, is given in Table 7.This table shows that the fraction of the total energy appearing in product translation is larger than that which can be expected from simple statistical argu- ments. Except for the three highest collision energy values studied, the average kinetic energy of the products remains at ca. 64.5% of the total available energy, irrespective of initial conditions, so that a large fraction of the available energy appears as kinetic energy of products and it remains essen- tially unaffected by changing the initial conditions. This observation disagrees with statistical predictions from which one would expect the average recoil energy to increase quite slowly with total energy, while the fraction of the total energy that appears in product translation decrease^.^^ Only for the higher values of the collision energy do our results seem to agree, in some extension, with statistical predictions.Thus, in addition to the observed tendency for a more symmetric rep- resentation of the polar reactive cross-section, a variation in energy partition occurs at the higher collision energy values investigated. Nevertheless, at each collision energy studied, the vibrational distributions for the insertion process are @,,” has been calculated from eqn. (10). nearly statistical. From the vibrating-rotor model the J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 50.0 50.0 -I a L8 40.0-L m . i. U d% 30.0-Q aN .a-20.0-sz -a --2 10.0-. d (a) 0.0 .,. I. I. I.,.W.I. .\ 0 20 40 60 80 100 120 140 160 180 40.0 1I a 40.0 2BU f 30.0 Q N oa r 20.0 0 (d) nn I 4"."I., -,-,-,-, ., -, .. 0.0 .,.#.~.~.,~~.#.~' 0 20 40 60 80 100 120 140 160 180 0 20 40 60 80 100 120 140 160 180 01" 40.0 , a g 30.0 30.0-J a L ,Q20.0 20.0 --c I Ib0 1 r 0 r5 10.0---IP P'o.oclL (e 1 0.00.0 --,,..11--,,..ll..bb--,,--,,..11--I 0 20 40 60 80 100 120 140 160 180 @I0 Fig. 7 Angular distribution of the BO molecule at several collision energies, v = 0, J = 2. E,/eV: (a)0.04, (b)0.06;(c)0.08,(d)0.1, (e)0.2, (f)0.3 expected distribution, Po(Jv),is given by :' squares show the statistical population predicted by the vibrating-rotor model while the black lozenges are the trajec-POU;) = (1 -jV,yq1 -j,y (12) tory results.In respect to the products' rotational energy (Eiot)comingwhere from the insertion process, from Fig. 9 it can be seen that ELo, f, = E"/E,CC (13) varies linearly with E,, which is consistent with the fact that the total angular momentum appears primarily as product The BO vibrational population as a function offv (the frac-rotational excitation. Moreover, insertion processes give a tion of vibrational energy) is displayed in Fig. 8. The open hot rotational distribution, which becomes more sharply J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 43 0.3 0.2 0.1 0.0 0.0 0 1 2 3 4 5 6 7 8 9 1011 121314 0 1 2 3 45 6 7 8 9 1011 121314 V' V' 0.2 a* h 0.1 -e rn 2icR 0.1eE$c I 4 4 'a 0.00.0 191.1'1'1'1'1'1'1'1 'I' 0 12 34 5 6 7 8 9101112 0 1 2 3 4 5 6 7 8 9 1011 121314 V' V' 0.2 h 2 v CL 0.1 0.0 0 1 2 3 45 6 7 8 9 10 11 121314 0 12 3 45 6 7 8 91011121314 V' V' Fig.8 Analysis of the BO vibrational distribution. (+) Trajectory results; (a)statistical population as predicted by the vibrating-rotor model ; u =0, J = 2. E,/eV: (a)0.04.(b)0 06. (c) 0.08,(40.1, (e)0.2, cf)0.3 peaked at the higher collision energies, with a JkaXvalue that differs strongly from the corresponding value for a thermal distributioii at 300 K. From t$e calculated energy distribution in the recoil of the productc it is remarkable that energy randomization is not achieve: it all during the lifetime of the collision complex.Both tr, :islational and rotational distributions can be explained in terms of the statistical adiabatic model for bending vibrational and rotational motions of the collision complex, such as have been found in other reactions where a light atom flies away from the complex.26 4.Conclusions In this paper we report the dynamics of the elementary reac- tion B(2P)+OH(2H)--+ BO(2Cf)+H(2S)by using the quasi- 44 Table 7 Energy distribution in products" for the reaction B + OH --+ BO + H at several collision energies' 0.005 3.873 64.5 26.5 9.0 0.025 3.893 64.6 25.6 9.8 0.03 3.898 64.6 25.2 10.2 0.04 3.908 64.6 25.2 10.2 0.06 3.928 64.5 25.0 10.5 0.08 3.948 64.4 24.8 10.8 0.1 3.968 65.6 23.6 10.8 0.2 4.068 64.5 21.8 13.7 0.3 4.168 63.5 21.1 15.4 0.04 3.907 60.3 29.7 10.0 0.04 3.995 65.0 24.9 10.1 0.04 4.037 66.1 23.4 10.5 0.04 4.411 58.7 31.2 10.1 O/i'Ek, Ebib and E;,, are given as of E,,,.'Only for products formed throughout the insertion reaction mode; see text. classical trajectory method. An empirical analytical Sorbie-Murrell potential-energy function has been con-structed for the ground doublet electronic state of the HBO system from ab initio calculations data, which include valence-electron correlation. Results arising from the reaction mode analysis tend to reflect the main topological features of the potential-energy surface used in trajectory calculations. We have observed that a,(&) decreases as the relative energy increases in the energy range studied and this behaviour is correlated with the absence of a potential barrier in the entrance channel.Vibrational excitation from v = 0 to u = 1 does not modify significantly the reaction cross-section value at E, = 0.04 eV. This effect can be tentatively explained in terms of the pre- sence of a late well on the potential-energy surface. The strong dependence of the reaction cross-section (at fixed colli- sion energy) on the rotational contents of the diatomic mol- ecule can be interpreted in terms of both orientational and energetic effects. The reaction was found to proceed via two possible mecha- nisms. The boron atom can either abstract the oxygen from one end of the OH molecule, or insert itself into the OH bond to form a collision complex that subsequently decom- poses into products.The classification of abstraction and insertion processes was made by monitoring the minimum potential energy attained in the course of each trajectory, the former being much more reactive than the latter. Evidence for collision complex existence was obtained by analysing the polar reactive cross-sections, lifetime and product energy distributions, and the osculating complex model was used to interpret the PRCSs. These complexes seem to be short-lived, their lifetimes being, in fact, shorter than their rotational periods at low collision energies, but the lifetimes do increase with collision energy.A transition occurs from an osculating short-lived complex at low collision ener- gies to a more long-lived one at higher energies. The energy distribution in products gives further evidence of collision complex formation. Even though the vibrational- state distribution in the product BO molecule was statistical for the insertion reaction, this was not necessarily the case for the rotational and translational distributions. The authors are grateful to the Computer Center of Barcel- ona University for supporting the computer time on the IBM 3090 computer machine. We also thank. Dr. R. Sayos for the use of the program to fit the analytical potential-energy func- tion and for many interesting discussions. J. CHEM. SOC. FARADAY TRANS., 1991, VOL.87 0.3 0.0 0.1 0.2 0.3 0.4 ERW Fig. 9 Dependence of E,,, on E,. y = 0.3535 + 0.9662~.R = 0.99 References 1 G. J. Green and J. L. Gole, Chem. Phys. Lett., 1980,69,45. 2 J. DeHaven, M. T. O'Connor and P. Davidovits, J. Chem. Phys., 1981,75,1746. 3 E. R. Lory and R. F. Porter, J. Am. Chem. SOC., 1971,93,6301. 4 Y. Kawashima, K. Kawaguchi and E. Hirota, Chem. Phys. Lett., 1986,131,205. 5 Y. Kawashima, Y. Endo, K. Kawaguchi and E. Hirota, Chem. Phys. Lett., 1987, 135,441. 6 M. Alberti, R. Sayos, M. Gonzalez, J. M. Bofill and A. Aguilar, J. Mol. Struct. (Theochem.), 1988,166, 301. 7 M. Alberti, Ph. D. Thesis, Dept. Quimica Fisica, Universitat de Barcelona, 1990. 8 N. L. Summers and J. Tyrrell, J. Am. Chem. SOC., 1977,99, 3960.9 J. Tyrrell, J. Phys. Chem., 1979,83,2906. 10 S. Sakai and K. D. Jordan, J. Phys. Chem., 1983,87,2293. 11 A. Sole, R. Sayos, J. M. Lucas, M. Gonzalez, X. Gimenez, M. Alberti and A. Aguilar, in Studies in Physical and Theoretical Chemistry, ed. R. Carbo, Elsevier, Amsterdam, 1989, vol. 62, p. 535. 12 J. N. Murrell, S. Carter, S. C. Farantos, P. Huxley, and A. J. C. Varandas, Molecular Potential Energy Functions, John Wiley & Sons, New York, 1984. 13 R. Sayos, M. Gonzalez and A. Aguilar, Chem. Phys., 1990, 141, 401. 14 C. Merller and M. Plesset, Phys. Rev., 1934,46, 618. 15 R. N. Porter and L. M. Raff, in Dynamics of Molecular Colli- sions, Part B, ed. W. H. Miller, Plenum Press, New York, 1976, p. 1. 16 D. G. Truhlar and J. T. Muckerman, in Atom-Molecule Colli-sion Theory, A Guide for the Experimentalist, Plenum Press, New York, 1979, p. 505. 17 M. Gonzalez, J. M. Lucas, R. Sayos and A. Sole, unpublished work. 18 D. G. Hopper, Program No. 248, QCPE, Indiana State Uni- versity, Bloomington, 1974. 19 P. A. Whitlock, J. T. Muckerman and E. R. Fisher, J. Chem. Phys., 1982, 76,4468. 20 M. Gonzalez, A. Aguilar and M. Gilibert, Chem. Phys., 1989, 131, 335. 21 M. Gonzalez, A. Aguilar and M. Gilibert, Chem. Phys., 1989, 131, 347. 22 N. Sathyamurthy, Chem. Rev., 1983,83,601; H. J. Loesch, Chem. Phys., 1986, 104, 213; 1987, 112, 249; C. A. Boonenberg and H. R. Mayne, Chem. Phys. Lett., 1984, 108,67; H. R. Mayne and S. K. Minick, J. Phys. Chem., 1987,91, 14.3. 23 Molecular Reaction Dynamics and Chemical Reactivity, ed. R. D. Levine and R. B. Bernstein, Oxford University Press, Oxford, 1987. 24 S. Stolte, A. E. Proctor and R. B. Bernstein, J. Chem. Phys., 1976,65,4990. 25 M. K. Bullit, C. H. Fisher and J. L. Kinsey, J. Chem. Phys., 1974, 60,478. 26 J. M. Farrar and Y. T. Lee, J. Chem. Phys., 1976,65, 1414. Paper 0/03017A; Received 5th July, 1990

ISSN:0956-5000    

DOI:10.1039/FT9918700037

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. CHEM. SOC. FARADAY TRANS., 1991, 87(1), 45-50 45 Kinetics of the Reaction of Silicon with Gaseous Bromine Zane H. Walker and Elmer A. Ogryzlo” Department of Chemistry, University of British Columbia, Vancouver, B.C., Canada V5T I Y6 The chemical etching of intrinsic polycrystalline silicon has been studied at Br, pressures between 0.05 and 30 Torr, and at temperatures ranging from 540 to 600°C. In this temperature range, and at a pressure of 1 Torr and above, the etch rate is given by: Retch= 10’0.2i0.4 nm min-’ exp-[(132 f7 kJ rnol-’)/Rq x p1/2 -109.8i2.4 nm min-’ exp-[(140 40 kJ mol-’)/Rfl. Deviations from this rate law at low pressures were found to be consistent with a reaction mechanism which involves the dissociative adsorption of Br, on the silicon surface.A kinetic analysis of the pressure and temperature dependences of the etch rate yields a value of 108.9i1.3 nm min-’ Torr-’ exp-[(lo8 & 22 kJ mol-’)/R7-l for the first-order rate-controlling rate constant at low pressures and 10’0.2i0.4 nm min-’ Torr-’/’ exp-[(132 f7 kJ mol-’)/Rfl for the half-order ‘composite’ rate constant which is observed at high pressures. The possibility that bromine atoms are produced in the gas phase, and that these species are responsible for the observed etch rates is also considered. The atom reaction and the molecule reaction proceed through the same intermediate, and the relative contribution from the two processes depends on how far the concentrations of these species are from their equilibrium values.The reaction of a gas with a solid to produce gaseous pro- ducts (called gas-phase etching) is a conceptually simple process whose mechanism remains essentially undetermined. This includes the simplest example of such a reaction in which a diatomic molecule (e.g. Cl, or Br,) reacts with an elemental solid (e.g. C or Si) under conditions where the pro- ducts are volatile. Because of the dearth of kinetic data for these systems, no one has yet attempted to formulate any general principles which might govern such processes. The unanswered questions include : what are the mechanisms of the etching reactions? What elementary reactions are involved? Do these change with changing pressure? Are the activation energies directly related to the bond dissociation energies of the halogen molecules? If rate measurements can be made over a suficiently large range of pressure and tem- perature some of these questions can be answered by a simple determination of the orders of the reactions and the tem- perature dependences of the rate constants.For the reaction of Br, with silicon there appears to be only one earlier kinetic study. In 1982 Sveshnikova et a!.’ reported an investigation of the Br, etching of Si(100) at tem- peratures between 490 and 550°C in which they found that the etch rate increased with increasing pressure at low Br, pressures but reached a limiting value at presssures between 5 and 15 Torr. They interpreted these observations by assuming that the first step in the reaction is a reversible adsorption of Br, on the reacting surface, and that pressure independence results when the surface is saturated with adsorbed Br, .In this study intrinsic polycrystalline silicon has been exposed to Br, pressures ranging from 0.05 to 30 Torr and at temperatures ranging from 540 to 600°C. Our results differ markedly from those of Sveshnikova et al. and suggest a different etching mechanism. Experimental The etching reactor consisted of a quartz tube 2.5 cm in dia- meter and 30 cm in length. A 15 cm length of the reactor was heated using a length of heating tape connected to a Variac. The samples were mounted on a holder consisting of an 18 mm x 14 mm x 9 mm block of single-crystal silicon, the surface of which was polished in order to ensure good thermal contact between the holder and the sample being etched. The holder, which was thermally oxidized in order to produce a thin protective oxide layer, was supported via a 4 mm quartz tube.Temperatures were measured during etching experiments with a thermocouple probe which was passed through this quartz tube into the centre of the silicon sample holder. A quartz spring was employed to hold the samples in place and to ensure good thermal contact with the sample holder. Polycrystalline silicon films 350 nm thick, deposited on a thin (50 nm) silicon oxide layer on top of single-crystal silicon wafers were used in the etching studies. Prior to etching, samples ranging in size from 0.04 to 0.09 cm2 were washed in dilute HF then dried under nitrogen before being loaded into the reactor.Once loaded the system was pumped down via cryostatic and roughing pumps, then heated via the heating tape to the desired temperature. The system was brought to atmospheric pressure with He after etching. A.C.S. reagent-grade bromine (Aldrich), with a quoted purity of 99.5%, was used in our preliminary etching experi- ments. The results obtained using this grade of bromine showed significant scatter and hence this reagent was deemed unsuitable for our quantitative etch-rate measurements. High-purity (99.99+%) bromine (Aldrich) was obtained and yielded improved reproducibility and much reduced scatter. The bromine was loaded into the cold finger of a 5 dm3 glass bulb.The cold finger was immersed in a C0,-methanol bath, which solidified the bromine, while being pumped on to remove any trapped air. A bromine gas flow of 12 sccm was maintained throughout all experiments. Br, pressures and reactor temperatures were chosen such that reasonable etch rates were obtained. Etch rates between a few nm min- and 800 nm min-’ were found to be most easily measurable. At the upper limit of these etch rates, with a Br, pressure of 30 Torr and a temperature of 600”C,ca. 3 x lo4 Br, molecules undergo collisions with each surface silicon atom before it is desorbed as product. Therefore the reaction of the Br, with the Si produced no concentration gradients in the system even under the most extreme reaction conditions.Etch rates were measured in sztu using laser interferometry by following the sinusoidal intensity of the He-Ne laser 633 nm radiation reflected from the stationary SiO, surface and the receding Si layer which is being etched. These etch rates were occasionally verified by laying down 0.5 mm wide SO, stripes on the surface and measuring the absolute etch depth achieved for a determined length of time using a Tencor pro-filometer which measures the height of the etched surface relative to the masked surface. 46 Results Br, pressures between 0.05 and 30 Torr and temperatures between 540 and 600°C were found to etch intrinsic poly- crystalline silicon at measurable rates. When low-purity bromine was used a loss of reflectivity from the surface was observed as the etching progressed.Surface profilometry of the samples after etching indicated that the variation in surface roughness was equivalent to ca. 10-15940 of the total etch depth. Since this surface roughness disappeared when high-purity bromine was used, it was clearly associated with impurities present in some of the commercially available bromine. The impurity level also affected both the magnitude and the reproducibility of the etch rate. Slightly slower but much more reproducible etch rates were observed with pure bromine. Since the most probable impurity in the bromine was another halogen such as chlorine or iodine, small amounts of these gases were added to the reactor to deter- mine whether they were responsible for the more rapid reac- tion. Although the addition of IBr, had no measurable effect on the reaction rate, the addition of C1, appeared to produce a slight increase in the etch rate. It is possible that all the effects observed with impure bromine can be attributed to a combination of impurities such as oxygen, water and chlo- rine, with an inhibition caused by the first two and an accel- eration resulting from the last.The pressure dependences of the etch rates at temperatures of 540, 570 and 600°C were determined using the high-purity Br, and are presented in Fig. 1. The etch rates increase in a non-linear fashion with increasing pressure, but do not appear to level off, even at 30 Torr. The order of the reaction with respect to Br, was determined by plotting In Retch us.In(Br, pressure). A least-squares linear regression of these plots indicated slopes, and hence orders, equal to ca. 0.6, sug-gesting a square-root dependence of the etch rate on Br, pressure. This possibility is verified in Fig. 2, where plots of the etch rate us. (Br, pressure)'/2 reveal a linear relationship at pressures above one Torr. These lines can be represented by Retch= C,(Br, + C, (1) The slopes (C,) and intercepts (C,), along with their associ- ated errors as determined from a weighted linear least-squares fit assuming an uncertainty of k5Y0 in the etch rates, are listed in Table 1. Discussion Fig. 1 shows that the isothermal etch rate continues to increase with increasing Br, pressure and does not level off P/Torr Fig.1 Etch rate of polycrystalline silicon as a function of the bromine pressure at: A,540"C; 0,570"C; .,600"C J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 aW--600 r I .-C E g 400-1 0 1 2 3 4 5 P/Torr' Fig. 2 Etch rate of polycrystalline silicon us. (bromine pressure11'2 at three temperatures. The straight line represents a least-squares fit through the high-pressure points (i.e. 1 Torr and above). .,540°C; 0,570 "C; .,600 "C even at pressures of 30 Torr. This is contrary to the observa- tions of Sveshnikova et a!.,' who reported a levelling off of the etch rate for Br, etching of Si(100) and Si(ll1) at pres- sures of 4-15 Torr for temperatures ranging from 490 to 550°C.A representative set of data corresponding to Br, etching of Si(100) at 550°C has been reproduced from the paper of Sveshnikova et al. and is plotted in Fig. 3 as etch rate vs. Br, pressure. The Br, concentration dependence of the reaction was represented by where Kimis the reaction rate in the zero-order region, a is a constant at a particular temerature, and P is the Br, partial pressure. Thus at low Br, pressures, the reaction should become first order with respect to Br,, while at high Br, pres- sures a transition to zero order should be observed. However, in the In Retchus. ln(Br, pressure) plot of their data as shown in Fig. 4, the resulting function can be represented by two straight lines. One corresponds to a zero-order regime encompassing the three highest pressure points ; the secand includes the remaining data points, yielding a slope, and hence an order of ca. 0.6, in agreement with our results.There do not appear to be data points corresponding to a first-order regime required by their interpretation. The apparent sudden change in reaction order at a Br, partial pressure of 15 Torr in their system is difficult to explain with the infor- mation provided. Note that in a previous paper, Sveshnikova et aL2 reported a 63% increase in etch rate for Si(100) at 450°C upon going from a Br, pressure of 34 to 180 Torr. This indicates that the saturation limit reported in their second paper may be a consequence of their experimental technique rather than a fundamental characteristic of the reaction.The linear relationship observed between etch rate and (Br, pressure)'", as illustrated in Fig. 2, points to a mecha- Table 1 Experimentally determined values for C, and C, as a func- tion of temperature T/"C slope, CJnm min-' Torr-1/2 intercept, C2/nm min-.' 540 570 44+1 94 * 2 -5.0f0.4 -16.3 1.0 600 166 f3 -25.0 f1.0 J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 500 00 0 0 -400 r I C 0 .--E 300 E .. 0 Y It“ -0 2 4 6 8 10 12 14 16 18 20 P/Torr Fig. 3 Etch rate of Si(100) us. bromine pressure at 550°C taken from the paper by Sveshnikova et al.’ nism consisting of at least two steps. In the first, bromine molecules are reversibly dissociated to produce atoms.In the second, the atoms react with the silicon in a first-order rate- controlling process. The first step could happen on the silicon surface or in the gas phase. We will first consider the mechan- ism in which the dissociation occurs on the silicon surface. Reversible Dissociative Adsorption (RDA) Mechanism This mechanism can then be written: k2 Br, (4)-SiBr, where Br, is the gas-phase bromine molecule and Brads is a Br atom adsorbed on what is probably a partially bromin- ated surface. The first step represented by eqn. 3,-3 is the reversible dissociative adsorption of Br, on such a silicon surface. The second step, represented by eqn. 4 is the reaction of an adsorbed bromine atom (Brads) to produce a species which is either gaseous product (where x = 1, 2, 3, or 4 depending upon the temperature) or some precursor which forms that product in a subsequent non-rate-determining step.It follows that the etch rate in units of nm min-’ (i.e. units which are independent of surface area) is determined by the rate of reaction 4, i.e. h r .r 6 E E \r0 Y 4‘ 1 -2 -1 0 1 2 3 In(Pporr) Fig. 4 In Retchus. ln(bromine pressure) for data presented in Fig. 3 Once a constant etch rate is established, the rate of formation of Brads is equal to its rate of removal, i.e. Solving this quadratic equation for Brads yields Substituting this expression into eqn. 5 and rearranging leads to At low bromine pressure, where the term within the square root approaches 1, and eqn.8 reduces to i.e. the reaction becomes first order with respect to bromine pressure as the dissociative adsorption equilibrium breaks down. At high bromine pressures, l6(klk- l/kg)PBr29 I, and eqn. 8 reduces to This rate law is consistent with the empirical rate law obtained at high pressures given by eqn. 1. By using the slopes and intercepts listed in Table 1, values for the first- order constants which control the etch rate at low pressures, k,, and the half-order constants which control the etch rate at high pressures, (k,/k-,)ll2kz, have been calculated and are listed as a function of temperature in Table 2. The temperature dependences of the low-pressure first-order rate constant k, and the ‘composite’ half-order rate constant (k,/k-,)112k, which are listed in Table 2 can be used to obtain the activation energies for these processes.We assume an Arrhenius form for that temperature dependence, i.e. k = A exp[-EJRT] (12) where R is the gas constant, T is the temperature in K, A is a temperature-independent pre-exponential and E, is the acti- vation energy. Note that since the experiments were carried out at constant pressure, E, is really an activation enthalpy. However, the two quantities differ only by RT for a first- order reaction and (1/2)RT for a half-order reaction. This is within the experimental uncertainty in E,. The Arrhenius plots for these two rate constants are shown in Fig. 5. The Table 2 Experimentally determined values for the low-pressure first-order rate constant k, and for the ‘composite’ half-order rate constant (k,/k-,)‘/’k, as a function of temperature T/”C 540 k,/nm min-’ Torr-’ 97 * 12 (kl/k-,)1’2k2/nm min-’ Tort-”’ 44+1 570 136 k 14 94 f2 600 276 k22 166 f3 48 lines drawn through.the points yield the following equations : 1--10'8.9* 1.3) nm min- 1 Torr-1 x exp-[(108 f22 kJ mol-')/RT] (13) (kl/k-l)1/2k, = i0(10.2*0.4)nm min-' Torr -'I2 x exp -[(132 f7 kJ m~l-~)/RTl (14) The 108 kJ mol-' activation energy for k, is 86 kJ mol-' below the bond-dissociation energy of Br,.If k, represents the dissociative adsorption of Br, on the surface, the value is not unreasonable. The potential-energy curve for this first reversible step is shown on the left-hand side of Fig.6. The energy level of the adsorbed atoms @Brads) must lie some dis- tance below the transition state at 108 kJ, but is undeter- mined relative to the reactants, and is therefore represented by a dotted line. From the activation energy of the half-order composite rate constant we can draw the right-hand potential-energy curve in Fig. 6, since the peak lies 132 kJ above the reactants. Once again the position of the Brads is undetermined, although it must naturally lie at one half the energy of 2Br,,,. For reasons that will be discussed later, Brads is located at 14 kJ mol-'. There is, however, one pos- sible criticism of this interpretation of the data, and that is the magnitude of the pre-exponential in k,.It is more than two orders of magnitude greater than the collision frequency of Br, with the surface. However such 'greater than collision frequency' pre-exponentials have also been observed in silicon etching by Br3 and C14y5 atoms and in the chemical- vapour deposition of silicon.6 In view of the common I I I I 4 1.16 1.18 1.20 1.22 lo3 KIT Fig. 5 Arrhenius plots for the high-pressure rate constant, (k,/k-1)1/2kz, ;and for the low-pressure rate constant, k,, I 140 I I 7 120 1 A i \ products '0' reaction coordinate Fig. 6 Potential-energy curve for the rate-controlling steps in the etching of silicon by gaseous bromine J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 occurrence of such large pre-exponentials, even in the reac- tions of atoms, it would appear that they are not necessarily an indication of an incorrect mechanism, but simply that the reaction is not an 'elementary step', i.e.it is a composite of at least two elementary steps. In the present reaction the steps would be the reversible physisorption of a Br, molecule on the halogenated silicon surface followed by its dissociation on the solid surface. There is a second difficulty which is common to the other two mechanisms which we will discuss shortly. In an earlier study the activation energy for the reaction of atomic bromine with silicon' was determined as 62 2 kJ rnol.-'. The position of Br(g) is fixed at 97 kJ mol- 'relative to Br, on the potential-energy diagram shown in Fig.6. It seems rea- sonable to assume that the reaction of atoms will proceed through the same transition state which occurs in the second step of the Br, reaction (and lies at i32 kJ mol-' relative to Br,). However, in that case, provided Br(g) is in equilibrium with Brads the activation energy for the atom reaction should be 35 kJ mol-' (132 minus 97). Although the above RDA mechanism provides a rational- ization of the observed half-order kinetics, it is not unique in that regard. A mechanism in which gas-phase atoms are formed and then react with the surface is equally capable of accounting for a square-root dependence on the Br, pressure. For this reason we will consider two alternative mechanisms in which the Br, is dissociated either in the gas phase or on the reactor walls before the gas stream encounters the silicon surface.Gas-phase Dissociation (GPD) Mechanism At the pressures used in our experiments the gas-phase disso- ciation of Br, is a bimolecular process and can be represented by reaction 15 and the reverse reaction is a termolecular process represented by -15. Br, + Br, k; ,Br + Br + Br, (15,-15) k' If this is followed by a reaction with the silicon surface which is first order in Br atoms, i.e. k;Br + Si -products (16) the reaction becomes half order with respect to Br, molecules when reactions 15 and -15 maintain an equilibrium concen- tration of Br atoms (Br ) over the substrate surface. By cal- eq.culating Breq concentrations (which range from 1.5 to 3% over this temperature range) from thermodynamic data,' and making use of an independently determined rate constant k; for the reaction of Br atoms with intrinsic polycrystalline ~ilicon,~it is possible to determine the etch rates expected from Breq.These calculated etch rates are plotted us. Br, pressure for temperatures of 540, 570 and 600°C in Fig. 7. Also included are the experimentally measured Br, etch rates. From this plot it would appear that an equilibrium concen- tration of gas-phase Br atoms could more than account for the observed Br, etch rates. The fact that the predicted rates at the higher temperatures lie above the experimental points is puzzling. It is not difficult to show that to fit the data the activation energy for the atom reaction would have to be ca.30 kJ in the range 540400°C rather than the value of 62 kJ found in the range 300460°C. This is conceivable because, as mentioned earlier, the pre-exponential for the rate con- stant governing the atom reaction exceeds collision fre- quency. This indicates that the reaction does not consist of an elementary step, and therefore could have a rate-controlling step with a smaller activation energy in a different tem- perature range. J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 14ooc 1200 1 4 I I I I. 0 5 10 15 20 25 30 35 PEirJTorr Fig. 7 Etch rates of polycrystalline silicon as a function of bromine pressure at three temperatures.Solid lines represent etch rates based on calculated equilibrium Br-atom concentrations and experimen- tally determined Br atom reaction rate ~onstant.~ A, 540°C; e, 570 "C; .,600 "C Regardless of how accurately the experimental points fit the predicted etch rate, because of the requirement of a three- body collision in the reverse reaction -15, the time required to reach equilibrium through reactions 15, and -15 is rela- tively long. The time required to reach l/e of the Br,, is given by expression 0.193 t= Br,, kf-,Br , (17) Using the literature value for k'-,* at 600 "C of 1.01 Torr-' s-' and a Br, pressure of 1 Torr, a time of 7 s is calculated. In the present system at this temperature and pressure, the Br, flow velocity is 90 cm s-', resulting in a residence time for Br, molecules of 0.08 s in the heated region of the flow reactor before arriving at the silicon surface.Since the con- centration of atoms must be near equilibrium if half-order kinetics is observed, it is clear that the formation of Br atoms in gas-phase collisions is too slow a process to account for our results. We therefore considered the possibility that the atoms are dissociated on the reactor walls in the heated region upstream from the sample. Wall-catalysed Dissociation (WCD) Mechanism This mechanism can be written: k; Br, +wall 2Br +wall (18,-18) k" * k;Br +Si -products (19) Reaction -18 is relatively fast compared to reaction -15 at low pressures and can be a significant pathway for Br-atom recombination under these conditions.Although the rate constant kfLl has not been reported, values for the recombi- nation coefficient, y, of 6 x and 4 x loT4have been determined.'.'' Provided we write k'; wall =k,, kfL wall = kPl, k; =k, and Br =Brads the set of eqns. 5-14 developed for the previous mechanism applies to this one. The rate con- stants have a somewhat different interpretation. (1) The rate constant k; must have an E, characteristic of the disso- ciation of Br, into gas-phase atoms (i.e. 194 kJ mol-'). (2) The pre-exponential for k'; contains a measure of the surface area available for dissociating Br, .This area is orders of magnitude larger than the surface area of the sample being etched and could then account for the large pre-exponential factor experimentally determined for k';.There are, however, a number of inconsistencies in this mechanism that appear when the activation energies are examined. First, k;' is the rate constant for the dissociation of Br, into gas-phase atoms (with the wall only acting as a third body) and should have an E, equal to the bond energy of Br,, namely to 194 kJ mol-'. Experimentally, the E, for k'; was found to have a much lower value of 108 & 22 kJ mol-'. The second incon- sistency is common to the other two mechanisms discussed above. As in the SCD mechanism k; is the rate constant for the reaction between a gas-phase Br atom and the silicon surface. E, can be calculated for k; from the composite rate constant k; (k';/k'L 1)1'2, assuming E, = 194 kJ mol- for k;' and an E, of zero for k': (halogen atom-wall recombination is known to have a weak temperature dependence").Using these values, E, =35 & 7 kJ mol- is calculated for k; .This is ca. half the 62 f2 kJ mol-' independently determined for the reaction between a gas-phase Br atom and the same poly- crystalline silicon surface.j There are two additional pieces of evidence that argue against the WCD mechanism and thus support the RDA mechanism. In a recent study of the etching of n-type poly- crystalline silicon (dopant number density of 5 x lo'* cm-3) by Br,,', the etch rate was measured in the present flow system, and then in a P,O, poisoned wall reactor with only a 3.5 cm portion of the reaction tube heated upstream from the sample.The heavily doped silicon reacts 2-3 orders of magni- tude faster than intrinsic silicon permitting the use of lower temperatures between 300 and 450°C and hence a shorter heated region could be used. The effect of wall poisoning was to decrease the rate of reactions 18 and -18 by several orders of magnitude,' and a shorter heated region signifi- cantly reduced the extent to which the reactions occurred. These changes should have combined to dramatically decrease the Br atom concentration formed in the stream before it reached the silicon surface, and the etch rate should have been dramatically reduced. However, no diffference in the etch rate between these two systems was measurable.Only the RDA mechanism is consistent with this observation. In a recent mass-spectrometric study of the interaction of Br, with a (100)silicon surface, Jackson et ~1.'~showed that if Br, is adsorbed on a silicon surface at room temperature and then the surface is heated up to 1000 K under high- vacuum conditions, the products desorbed from the surface are found to be Br,, Br, SiBr, SiBr,, SiBr, and SiBr, in yields that are a complex function of the desorption tem-perature. The activation energy for that desorption step was found to lie between 119 and 186 kJ mol-'. The conclusion that we wish to draw from this work is that there is a reac- tion between molecular bromine and silicon in the absence of gas-phase bromine atoms, and that reaction occurs through an intermediate that yields gaseous silicon compounds, bromine atoms or bromine molecules in three 'desorption reactions' that have very similar activation energies.This is consistent with the RDA mechanism and with the relative activation energies for the three desorption steps that were calculated on the basis of that mechanism. Conclusion We conclude that although an equilibrium gas-phase atom concentration from the thermal dissociation of Br, could account for the observed etch rates, such an equilibrium con- centration is not formed in the gas stream prior to its encounter with the silicon chip. Why then is the etch rate predictable from an atom concentration which does not occur in the system? There is, we believe, a reasonable answer to this question.As discussed, both k, and the rate constant for the reaction of Br atoms have Arrhenius pre-exponential factors which suggest that they are not elementary steps but probably have as their first steps the adsorption equilibria described by eqn. 20,-20 and 21,-21: BrZ(g) k4 L -BrZ(ads) (20,-20) k-4 k3 Br(g) Br(ads) (21721) k-3 In a system in which all species are in equilibrium, both Br,(g) and Br(g) would be in equilibrium with Brads. When Br(g) is present in concentrations far in excess of the equilibrium values (i.e. with the microwave discharge on) that species is in equilibrium with Brads through reactions 21, -21. Whether reaction -3 followed by -20 contributes to the loss of Brads depends on temperature and pressure.Insofar as it does it should have the effect of decreasing the observed etch rate, and lowering the observed order of the atom reaction. On the other hand, when the Br(g) concentration is much smaller than the equilibrium value, the equilibrium concen- tration of Brads can be maintained by the two equilibria 20, -20 and 3, -3. The extent to which reaction -21 contrib-utes to the loss of Brads depends on temperature and pressure. Insofar as it does it should have the effect of decreasing the observed etch rate and raising the observed order of the Br, reaction. Therefore if either Br(!) or Br2(g) are present in concentra- tions below the equilibrium values, the rate constants mea- sured for the etching reaction could be lower than those which would be measured if all species were present in their equilibrium concentrations.Hence there could be a fortuitous agreement between the etch rates calculated on the basis of assumed equilibrium concentrations of either species. However, these etch rates could be below the etch rates which occur when all species are present in their equilibrium concentrations. J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 We conclude that the etching by ‘gaseous bromine’ can occur through the dissociative adsorption of Br, on the react- ing surface, and/or through the dissociation of Br, in the gas phase. Both paths produce the same reactive intermediate which we have labelled Brads and it is therefore possible for this species to be in equilibrium with either bromine atoms or molecules, or with both.It is this adsorbed atom that reacts with the silicon surface in a rate-controlling step to yield a gaseous product. We thank Bell Northern Research for samples of poly-crystalline silicon. Z.H.W. is grateful to the Natural Science and Engineering Research Council of Canada for a graduate fellowship. References 1 L. L. Sveshnikova, S. M.Repinskii and A. B. Posadov, Poverkh. Fiz., Khim., Mekh., 1982, 8, 134. 2 L. L. Sveshnikova, V. I. Donin and S. M. Repinskii, Sov. Tech. Phys. Lett., 1977, 3, 223. 3 Z. H. Walker and E. A. Ogryzlo, submitted for publication. 4 E. A. Ogryzlo, D. E. Ibbotson, D. L. Flamm and J. A. Mucha, J. Appl. Phys., 1990,67, 3 11 5. 5 Z. H. Walker and E. A. Ogryzlo, submitted for publication. 6 T. J. Donahue and R. Reif, J. Appl. Phys., 1985,57,2757. 7 A. J. Downs and C. J. Adams, Comprehensive Inorganic Chem- istry, ed. A. F. Trotman-Dickenson, Pergamon Press, Oxford, 1973, ch. 26. 8 J. K. K. Ip and G.Burns, J. Chem. Phys., 1969,51,3414. 9 E. A. Ogryzlo, Can. J. Chem., 1961,39,2556. 10 M. A. Clyne and A. R. Woon-Fat, Trans. Faraday Soc., 1973,69, 412. 11 P. G. Ashmore, A. J. Parker and D. E. Stearne, Trans. Faraday SOC., 1971,67, 3081. 12 Z. H. Walker and E. A. Ogryzlo, unpublished. 13 R. B. Jackson, R. J. Price and J. S. Foord, Appl. Surf Sci., 1989, 36,296. Paper 0/02312D; Received 23rd May, 1990

ISSN:0956-5000    

DOI:10.1039/FT9918700045

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

J. CHEM. SOC. FARADAY TRANS., 1991, 87(1), 51-56 Monte Carlo Calculation of Energy and Electron Transfer in a Monolayer of Spheroids Paul Siders Chemistry Department, University of Minnesota Dututh Minnesota 55812,USA Orientation-sensitive reactions have been modelled within equilibrium Monte Carlo distributions of reactants in a monolayer. The model consists of hard oblate spheroids with their centres in a plane. Configuration averages of a dipole-dipole energy-transfer rate constant and of an electron-transfer atomic-orbital overlap were calcu- lated. The apparent reaction order is greater than two over most of the density range studied and is sensitive to orientational ordering. This article reports calculations of ensemble-averaged rate constants for energy- and electron-transfer in a model mono- layer of oblate spheroids. The constant-NPT Monte Carlo method ‘,2 was used to generate equilibrium configurations of 128 hard oblate spheroids.The spheroids’ centres were con- fined to a plane, but the spheroids were free to rotate. This model was chosen to represent, albeit crudely, a monolayer of large aromatic molecules such as chlorophylls, porphyrins or polycyclic aromatics. The spheroids represent the repulsive interactions between such large, flat molecules. Orientational ordering in the monolayer may affect the kinetics of orientation-sensitive reactions such as electron and energy transfer. Energy transfer between substituted porphyrins in a monolayer has been studied experimentally by Tweet et Gonen et ~l.,~and Frackowiak and Szur- kowski6 The work of Gonen suggests that energy-transfer fluorescence quenching is sensitive to an orientational phase transition in the monolayer.Orientational ordering substan- tially affects energy transfer in the present model system. This article reports bimolecular rate constants calculated as equilibrium averages of the orientation- and distance- dependent unimolecular rate constants. The rate constant is legitimately calculated from an equilibrium configuration for reactions which are fast relative to translational and rotation- al diffusion, and for the initial behaviour of slower reactions. Forster’s dipole-dipole energy-transfer rate constant’ was taken to describe energy transfer. For electron transfer, the approximate theory of Brocklehurst was used.8 The density dependence of the calculated bimolecular reaction rates implies a reaction order greater than two over most of the density range.Reaction orders for surface reactions have been found to differ from those for bulk reactions in other model systems. The origin and significance of such differences are discussed, for example, by Silverburg and Ben-Shaul’ and Kopelman. Thermodynamics Periodic boundary conditions were used in the Monte Carlo calculation. The 128 spheroids reside in a rectangular basic cell. Its height and width fluctuated independently. The inter- particle potential is simply a hard-body pair potential. Dipole-dipole interaction is not included, as that would make the already long calculations much longer, and would require non-trivial consideration of temperature.The analytic spheroid overlap criterion of Vieillard-Baron’ was used. Initial configurations were equilibrated until no serial correlation in density was found at the (estimated) 99% confidence level, after which configurations were saved for analysis. System-size effects on density and orientational order parameters were found to be small. Pos-sible system-size dependence of rate constants was not studied. The calculation and the phase diagram of the system have been discussed in detail elsewhere.” In Fig. 1 are shown pressuredensity data for three spher- oid shapes. Spheroid shape is described by a/b, where a is the semi-major and b the semi-minor axis.For reference, Fig. 1 also shows data for spheres, obtained in the limit a/b = 1. The close-packing particle density po = 1/2d3(ab); the same as for hard ellipses in a plane. The spheroids order orientationally in two steps. Their principal symmetry axes tend to tip into the plane at low density. This ordering occurs to some extent at all non-zero densities, and is nearly complete by the density at which the second orientational ordering occurs. At higher density @/po 0.6 for a/b = 6 and 0.45 for a/b = 10) the symmetry axes align into a nematic. This second ordering appears to occur continuously over a range of densities, with no hyster- esis or density discontinuity. The rates of energy and electron transfer are sensitive func- tions of the distance between donor and acceptor, so the radial distribution function is relevant to reaction rate.The radial distribution function is shown in Fig. 2 for low, moder- ate and high densities. As density increases, pair density moves in closer to the minimum separation, rzl = 2b. Also, a peak near this minimum separation appears at high density. The orientational ordering that accompanies increasing density makes close approach possible. The density depen- dence of the radial distribution function is about the same as PIP0 Fig. 1 Pressure us. density calculated by the constant-NPT Monte Carlo method; alb = 512 (O), 10 (A)and 1 (spheres)(0)6 (a), 1'1"'"" Fig.2 Radial distribution function at low, moderate and high density; a/b = 6; p/po = 0.699 (A),0.396 (a),0.187 (0) usual: as density increases g(,) moves in towards contact and develops sharp structure. Theory Let us suppose a microscopic rate constant, k,, for the reac- tion between two molecules. This unimolecular rate constant is assumed to be a pair function, dependent on orientations of the two potential reactants, and on their separation. The rate constant, k,, is a unimolecular rate constant in the sense that two potential reactants constitute a reactive super-molecule. The macroscopic, observable, rate constant for electron transfer or energy transfer is assumed to be bimolecular. That is, rate/area = kbip2 (1) where p is the spatial number density of reactant molecules.The rate per unit area must reflect the average of k, over the pair configurations. Let p(,)(X,, X,) be the probability density of reactant pair, where Xi = (ri, Q,); i = 1, 2. Vector ri is in the (x, y) plane and locates the centre of spheroid i, and Qi is the set of Euler angles (a, x = cos p) specifying the orientation of spheroid i. The normalization of p(,) is sj P(~)(X~,X,) dX, dX, = N(N -1).The reaction rate is rate = /lk.,X,, x21P(2)(x1,X,) dX, dX, (2) The relative translational coordinates r21= r, -rl are sum- cient for k,. Also, p(,)(X1,X,) = ~(,)(r,~, where the Ql, o,)~, normalization of ~(,)(r,~,o,,a,) is N -1. Integrating over rl in eqn. (2) yields eqn. (3) rate/area = ku(r,l, Q,, a,)p j j 1 x ~(~)(r21,a,,QJ dQ, dQ, (3) Comparison of eqn.(1) and (3) shows that 1 j1kbi = ('/PI ku(r21~'1, '2) x p(,)(r,,, Q,, Q,) b,, dQ2 dQ, (4) The pair distribution function in eqn, (4)is the thermodyna- mic limit of a sum of delta functions located at spatial and angular coordinates chosen from an equilibrium ensemble. J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 1"~(~'(r,~,Q,, a,) = lim -1 1 6(r2, -[rj -ri]) N-.m N i=l j+i x 6(Q, -Ri)6(Q2 -Qj) (5) The rate constant, k,, may be long-ranged. Monte Carlo simulation yields pair-distribution data only for r2, up to half the smaller dimension of the basic Monte Carlo cell, ca. J(N/p), although the pair data could be extended to larger r,, using periodic images.Long-range behaviour of k, is handled as follows in this paper: for small separations, r2, ((r, where rc is a cut-off distance, k, is averaged over explicit pair data. For large separations, the radial density is supposed uniform and angular distributions are supposed independent of separation. (The cut-off distance chosen for calculations reported in this paper is r, equal to half the minimum basic cell height.) Explicitly, it is assumed that p(,) (r21,Q2, 0,)= P~(~)(Q,,a,) for r21 > r,. The large-r,, angle distribution p(,)(R,, Q,) is normalized to unity. The radial integral in kbi may, of course, be divided into two parts, the first over r2, < rc and the second over r21 > I-,. The small-r,, integral is evaluated directly from Monte Carlo pair data.In eqn. (6), the Heaviside unit step function H(x) = 1 if x > 0; 0 if x < 0. I" ku(r21,Q2 3 Ql) dr21+ j5[j 1I1rzi II >rc x P(~)(Q,,Q,) dQ2 dQ, (6) The large-r,, radial integral of k, was evaluated approx- imately, as follows: the large-r,, radial integral of k, was evaluated analytically. The large-r, angular distribution p(,)(Q,, Q,) was estimated by counting particle pairs separ- ated by somewhat less than the cut-off distance. [The interval J(;) < r2,/rC< 1 was chosen arbitrarily.] Then the angle integral of the product of the large-r,, angle distribution and the radially integrated k, was evaluated numerically. Great accuracy was not required, because the second term is small. That is, kbi is expected (and found) to be dominated by the sum over explicit pair configurations. For particles near the edges of the basic Monte Carlo cell, the cut-off circle of radius r, is likely to extend beyond the boundaries of the cell.In order to avoid large edge effects on kbi: (i) rc is kept smaller than half the height of the Monte Carlo cell; and (ii) distances I( rji 11 from particle i are com- puted both within the cell and to the neighbouring periodic images of the cell. Another way kbi might be calculated is using the minimum- image convention. For this method, the delta-function repre- sentation of p',) is used, but in the sum overj, 11 rj -ri 11 is the minimum distance between particle i and particle j, where j may be in the basic Monte Carlo cell or in any of the neigh- bouring periodic images of the basic cell.Eight periodic images of the basic cell are considered: i.e. four adjoining its edges and four its corners. The minimum-image method pro- vides no pair data for r,, greater than the smaller edge length of the Monte Carlo cell. Any long-range contributions to kbi are simply neglected. The bimolecular rate constant calcu- lated using the minimum-image convention is denoted k:'. (rji for minimum 11 rji)11, Qj,Ri) (7) In Table 1, minimum-image, cut-off and 'complete' [i.e. eqn. (6)]bimolecular rate constants are compared at several J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 Table 1 Minimum-image, cut-off and complete kbi lo4 tail alb PIP 0 k:‘ kcu,-off correction kbi 512 0.376 0.688 0.688 7.3 0.689 6 0.342 13.8 13.8 9.1 13.8 10 0.358 157.0 157.0 0.24 157.0 All rate constants are given in units of (Rg/ta4). pressures and spheroid eccentricities.All three constants are found to be numerically identical. The ‘tail correction’ is the second term in eqn. (6), the ‘cut-off’ constant is the first term, and kbi is the sum of the two terms. Long-range contributions appear negligible, but were calculated anyway for all kbi reported in this article. Tail corrections for the electron- transfer squared overlap are even smaller than for these energy-transfer constants. The formulae above describe calculation of kbi for a single N-particle Monte Carlo configuration. In practice, a large number (M >, 160) of configurations are used.The reported kbi is the average over configurations. The configurations used are separated by lo00 Monte Carlo moves, where a ‘move’ is an attempted translation and rotation of all N par-ticles, as well as an attempted change in height and width of the basic Monte Carlo cell. Reported uncertainties are 95% confidence intervals, calculated under the assumption that the M configurations are uncorrelated. This assumption is reliable at low pressure but becomes increasingly less reliable as pressure increases. Energy-transfer Rate Constant Macroscopic, bimolecular energy-transfer rate constants are calculated using Forster’s expression for the unimolecular rate constant. where k^ = cos $21 -3 cos 412 cos (8) In eqn.(8), z is the mean lifetime of the energy donor, R, is a critical transfer distance and rZ1is the distance between the centres of the reactants. The orientation dependence of k, is contained in the factor IC,where 4121is the angle between the transition dipole moment vectors of molecules 1 and 2, while $2 and iP1are the angles between these respective vectors and the line joining the centres of molecules 1 and 2. The hard-spheroid interaction potential does not include any dipole moments. A dipole is added to every spheroid after calculating configurations, in order to model dipole- dipole energy transfer. Any effects these dipoles would have had on spheroid configurations are neglected. In each spher- oid, a dipole is placed perpendicular to the spheroid’s prin- cipal symmetry axis, and in the horizontal symmetry plane. Orientational distributions are uniform with respect to rota- tion of spheroids about their principal symmetry axes, so Forster’s rate constant is used averaged with respect to such rotation, as in eqn.(9). -k, = C3/(2~)1(~0/~21)6(k-2) (9) where (K2> = C1/(2742111m,a27 a17xz, Xl,Y2 7 71) dY2 dY1 Calculated energy-transfer rate constants are shown in Fig. 3, for a/b = i. The rate constant increases smoothly with density, presumably because the average donor-acceptor separation diminishes as density increases. For spheroids of this eccentricity, no nematic ordering occurs, at least not in the density range graphed.12 For spheroids with a/b = 6, nematic ordering does occur.The range over which this ord- ering roughly occurs is p/p, = 0.5-0.7. The bimolecular rate constant is nearly constant during alignment of the spheroids. Fig. 5 shows kbi for still more eccentric spheroids. For a/b = 10, kbi depends even more strongly on density. As for a/b = 6, kbi tends to level off during nematic alignment. Angle-averaged Energy Transfer Both translational packing and orientational ordering occur simultaneously and apparently continuously as density increases. In order to investigate the relative importance to kbi of these two changes, an angle-averaged rate constant has been calculated and is graphed in Fig. 6. Also shown in Fig. 6 (labelled ‘correlated ’,filled circles) are the complete kbi results of Fig. 4.The orientation dependence was removed from k, PIPQ Fig. 3 Energy-transfer rate constant for alb = 3 I I I ~ ,=, , , , I , , , I , , ,] 0 0.0 0.2 0.4 0.6 0.8 The transformation from Euler angles (a,p = cos-x, y) to P/pQ Forster’s angles ($21, 42,41)and the explicit result of the Fig. 4 Energy-transfer rate constant for alb = 6. Nematic ordering integration over y2 and y1 are given in ref. 13. occurs from p/p = 0.5-0.7 300-- F 200- W W> 9'- an 100- f f I -L o!a1 0.0 ' I € ' [ ' 0.2 ' ' I ' 0.4 ' ' I ' 0.6 ' ' I 0.8 PIP0 Fig. 5 Energy-transfer rate constant for a/b = 10. Nematic ordering occurs from p/p = 0.3-0.5 by replacing the correct IC,with its isotropic angle average, 3.Results of this calculation, graphed as the upper points in Fig. 6, show that the overall density dependence of kbi is not directly the result of orientational ordering. Of course, orien- tational ordering is inextricably tied to density increase and hence to donor-acceptor separation and kbi. Uncorrelated Approximation The pair probability density p(,) reflects correlations between reactant orientations, and between orientation and trans- lation. At low density, either low total density or low reactant density, such correlations must be small. For simplicity, pair correlations may be neglected, except that reactant overlap (that is, spheroid overlap, not electronic overlap) will not be permitted.Explicitly, the pair distribution is approximated as a product of single-particle distributions. P'2'(r21, Ql, Q,) = PC1 +f21(r21, a,9 Ql)lf(Q,)f(Q,) (10) wheref,, is the Mayerffunction which equals -1 if particles 1 and 2 overlap, zero if they do not. The single-particle orien- I I I @I I olo 012 014 0;6 OI8 PIP0 Fig. 6 Comparison of the best (correlated, a)and two approximate calculations [angle-averaged (ic2 = and uncorrelated, A] of the3) .,energy-transfer rate constant; a/b = 6 J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 tational distribution function is f(Q). The solid angle Q = (a,x). The angle o! is measured relative to anem,along which the nematic director lies. In this uncorrelated approximation (uncorrelated except for the short-range correlation rep-resented by f,1), the bimolecular rate constant becomes kubY = s 5J Gzl, 01,Q2)L-1 +f21(r21, Q2 9 Q1)l Xf@,)f(Ql) dr21 dQ2 dQ1 (11) The integral in eqn.(1 1) is evaluated numerically. Results are shown in Fig. 6, for a/b = 6. In the high-density limit f(n)+ d(Q -Q,,,), the Mayer function,f,,, becomes a simple function of r21 and 8, where r2, = (rzl, 8). The closest-approach distance is (2ab)/ (b2cos2 8 + a' sin' @,I2. For the case of the Forster energy- transfer k,, the integrals in eqn. (11) can be evaluated analytically, yielding kinc = 225nnR:/(2"zb4). For a/b = 6, this high-density uncorrelated limit is 32Rg/(za4), which appears to be less than the correct (correlated) high-density limit.Electron-transfer Rate Constant In this work, the unimolecular electron-transfer rate constant, k,, is taken to be proportional to the squared overlap 11 S,, )I2 of two atomic orbitals, one centred in the donor spheroid, and one in the acceptor spheroid. This is the atomic-overlap approximation introduced by Brocklehurst' in his early study of orientation and distance effects in elec- tron transfer. More recently, the approximation was used in a study of electron transfer between randomly orientated donors and acceptors in three dimension^.'^ The approx- imation is a great simplification, and certainly errs quantitat- ively. However, it is consistent with the intent to explore general qualitative behaviour in the context of a simple model.The present work uses atomic 2p orbitals aligned with the spheroids' principal symmetry axes. Calculated orienta- tion effects depend strongly on the one-electron orbitals chosen. The present calculations are relevant to molecules for which cofacial electronic interaction is much stronger than edge-to-edge interaction. Time-dependent quantum perturbation theory suggests that the unimolecular rate constant would more accurately be taken to be proportional to the square of an electronic coupling matrix element, say H21.The assumption in Brock- lehurst's work, and here, is that H,, is proportional to the overlap of donor and acceptor electronic wavefunctions, S, ,, and that S,, may be computed with atomic (hydrogenic) orbitals. While it might be practical to calculate H,, itself at some simple level of electronic-structure theory, such an effort would ill match the highly approximate hard-body intermolecular potential used in the Monte Carlo calcu- lations. In this work, the bimolecular electron-transfer rate con- stant is calculated as the ensemble average of k,, and the orientation dependence of k, is represented by S;,.The effect of density on the ensemble-averaged S;, is, of course, only part of the effect to be expected on kbi. Other non-electronic factors such as solvent reorganization energy also influence kbi and are density-dependent. In addition, dynamic factors such as diffusion and solvent caging are density-dependent. Such dynamic factors are uniformly neglected in this work, which assumes that reaction occurs from near-equilibrium reactant distributions.The many factors involved in the theory of electron-transfer kinetics have been reviewed recently, by Newton and Sutin," for example. J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 Table 2 Semi-major axes and orbital exponents for spheroids 1 .o 0.5000 9.15 2.5 0.6786 12.52 6.0 0.9086 21.39 10.0 1.1856 27.14 The orbital exponent 5[t,hZprz r exp (-@)I is a parameter in the present calculation. The only physical constraint on 5 is that it must be large enough to ensure weak donor- acceptor orbital overlay. For the present work, < has been chosen so that 95% of the 2p, probability density is within the spheroid's surface.These values of 5, listed in Table 2, are large enough so that a-type p overlap (as well as n-type p overlap, of course) is a decreasing function of separation r2,, for all rzl greater than the minimum contact distance 2b. The 2pa overlap is a decreasing function of r2, for <rzl > 4.466, so cb > 2.23 is sufficient to ensure that a-type overlap decreases beyond contact. The 5s in Table 2 give 5b z 3, except in the spherical a/b = 1 case. If b were scaled to rep- resent the half-height of an aromatic n-electron cloud, say b = 2 A, then 5 z 0.7 kl,which is reasonably consistent with observed distance dependences of electron-transfer rates. Calculated squared overlaps are plotted in Fig. 7 for three spheroid shapes.The general density dependence of (Si,) is remarkably like that of kbi,even though these functions differ radially and orientationally. Magnitudes of khi and (Si,) are, of course, very different, and depend differently on donor- acceptor shape (i.e. on a/b). Magnitudes are sensitive to the 5-4-A-c " mc3-0 7-2-1-0:o 012 0:4 0:6 0:8 PIP, Fig. 7 Electron-transfer squared pp overlap; a/h = 10 (@), 3 (H), 6 (4 Table 3 Low-density limits electron 104<s:,) 10 2.83 0.019 6 1.24 0.081 2.5 0.34 1.1 1.01" 0.099 2.4 1.oo 0.0982b 2.39' a For a/b = 1.01, a = 0.5010 and 5 = 9.16. Precisely, n/32. Precisely, (4t9 + 34t8 + 136r' + 404t6 + 1416r5 + 5340t4+ 15180t3 + 28170t2 + 308705 + 15435)[n e~p(-25)]/(81005~). distance-scaling parameters R, (for kbi) and 5 (for (S; which were treated differently for energy and electron trans- fer.Low-density Limit As p +0, the pair probability densities approach uniform dis- Q2,tributions: ~(~)(r~,,Q,) +pH(r21-rcont)/16n2. For energy transfer, ,-ran,-JJJJ 4% -0) d(a, -8) dx2 dx,. The contact distance, denoted rcont, is a function of the orien- tation angles and can be evaluated numerically, using the overlap function. Performing the angle integrations by rec- tangular quadrature yields the results in Table 3. Discussion Both energy- and electron-transfer rate constants are density- dependent within the present model. At low density, the rate constants are small and the reaction appears to be second order.At intermediate densities the rate constants indicate approximately third-order density dependence. The order diminishes to nearly two during nematic ordering of the spheroids. This density dependence evidently follows from the density dependence of the radial pair distribution function. The dis- tribution function generally favours smaller separations as density increases, so the bimolecular rate constant increases with density. However, during nematic ordering the shape of the radial distribution changes, so that the probability of small separations grows little until nematic order is estab- lished. This change in radial distribution is shown in Fig. 8 for a/h = 6. The two intermediate densities there lie in the nematic-ordering region.The small-separation peak near r2, = 3b actually diminishes as density increases, throughout the small density range over which khi and (St,) are nearly constant. The similarity of the density dependence of the two bimolecular rate constants is surprising, because the unimolecular rate-constant expressions differ considerably, both radially and angularly. Evidently, both rate constants c?'0.61 4 4 lo 4 6 8 r2 1/b Fig. 8 The radial distribution function at short range, for densities near the transition to nematic order; a/b = 6; p/po = 0.699 (@), 0.650 (+), 0.613 (W), 0.568 (0) 56 J. CHEM. SOC. FARADAY TRANS., 1991, VOL. 87 are predominately controlled by the radial distribution, which is itself tied to orientational ordering.4 5 G. L. Gaines Jr., A. G. Tweet and W. D. Bellamy, J. Chem. Phys., 1965,42, 2193. 0. Gonen, H. Levanon and L. K. Patterson, Israel J. Chem., I thank Professor Thomas Bydalek for helpful discussions of the high-pressure behaviour of kbi. This work was supported 6 1981,21,271. D. Frackowiak, J. Szurkowski, S. Hotchandani and R. M. LeBlanc, Mol. Cryst. Liq. Cryst., 1984, 111, 359. by the Donors of the Petroleum Research Fund, adminis- tered by the American Chemical Society. The use of comput- ing equipment provided through the AT&T University Donation Program is gratefully acknowledged. 7 8 9 10 11 T. Forster, Discuss. Faraday SOC.,1959,27, 7. B. Brocklehurst, J. Phys. Chem., 1979, 83, 536. M. Silverberg and A. Ben-Shaul, J. Stat. Phys., 1988,52, 1179. R. Kopelman, Science, 1988,241, 1620. J. Vieillard-Baron, J. Chem. Phys., 1972,56,4729. 12 P. Siders, Mol. Phys., 1989,68, 1001. References 13 14 M. He and P. Siders, J. Phys. Chem., 1990,94,7280. R. P. Domingue and M. D. Fayer, J. Chem. Phys., 1985, 83, M. P. Allen and D. J. Tildesley, Computer Simulation of Liquids, 2242. Oxford University Press, Oxford, 1989. 15 M. D. Newton and N. Sutin, Annu. Rev. Phys. Chem., 1984, 35, J. P. Valleau and G. M. Torrie, in Modern Theoretical Chemistry 437. 5A: Statistical Mechanics, ed. B. J. Berne, Plenum Press, New York, 1977, ch. 5. A. G. Tweet, W. D. Bellamy and G. L. Gaines Jr., J. Chem. Paper 0/03226C ;Received 18th July, 1990 Phys., 1964, 41,2068.

ISSN:0956-5000    

DOI:10.1039/FT9918700051

出版商:RSC    

年代:1991

数据来源: RSC